Telecommunications

Telecommunications in Europe
COMMUNICATION AND SOCIETY

edited by George Gerbner and Marsha Siefert
IMAGE ETHICS
The Moral Rights of Subjects
in Photographs, Film, and Television
Edited by Larry Gross, John Stuart Katz,
and Jay Ruby
CENSORSHIP
The Knot That Binds Power and Knowledge
By Sue Curry Jansen
THE GLOBAL VILLAGE
Transformations in World Life and Media in
the 21st Century
By Marshall McLuhan and Bruce R. Powers
SPLIT SIGNALS
Television and Politics in the Soviet Union
By Ellen Mickiewicz
TARGET: PRIME TIME
Advocacy Groups and the
Struggle over Entertainment Television
By Kathryn C. Montgomery
TELEVISION AND AMERICA'S CHILDREN
A Crisis of Neglect
By Edward L. Palmer
PLAYING DOCTOR
Television; Storytelling, and Medical Power
By Joseph Turow
THE VOX DEI
Communication in the Middle Ages
By Sophia Menache
THE EXPORT OF MEANING
Cross-Cultural Readings of DALLAS
By Tamar Liebes and Elihu Katz
TELEVISION IN EUROPE
By Eli Noam
TELECOMMUNICATIONS IN EUROPE
By Eli Noam
Telecommunications
in Europe
Eli Noam
New York Oxford

OXFORD UNIVERSITY PRESS
Oxford University Press
Oxford New York Toronto
Delhi Bombay Calcutta Madras Karachi
Kuala Lumpur Singapore Hong Kong Tokyo
Nairobi Dar es Salaam Cape Town
Melbourne Auckland
and associated companies in
Berlin Ibadan
Copyright 1992 by Eli Noam

Published by Oxford University Press, Inc.,
198 Madison Avenue, New York, New York 10016-4314
Oxford is a registered trademark of Oxford University Press
All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,
electronic, mechanical, photocopying, recording, or otherwise,
without the prior permission of Oxford University Press.
Library of Congress Cataloging-in-Publication Data

Noam, Eli
Telecommunications in Europe / Eli Noam.
p. cm.(Communication and society)
Includes bibliographical references and index.
ISBN 0-19-507052-6
1. TelecommunicationEurope. 2. Telecommunication policyEurope.
I. Title. II. Series: Communication and society (New York, N.Y.)
HE8084.N63 1992 384'.094dc20 90-22969
9876543
Printed in the United States of America
on acid-free paper
Preface
This book, together with its companion, volume, Television in Europe, cov-
ers many subjects and many countries. It is therefore necessary, at the outset,
to state what it is and what it is not.
The book is an analysis of the rise and beginning fall of the telecommuni-
cations monopolies in Europe. These Post, Telegraph, and Telephone admin-
istrations (PTTs) are vast and powerful; in many countries they are the largest
employer, investor, and buyer. They often run vast financial savings and pay-
ment systems, and even provide services such as bus transportation in some
countries. Their physical presence is ubiquitous, and their services reach daily
into almost every home and office. But despite the importance and scale of
these institutions, social scientists have paid little attention to them. 1
One reason for such lack of interest is that the subject of telecommunications
seems forbiddingly technological. Actually, most of the issues in telecommu-
nications are quite accessible once they are stripped of needless jargon. Another
reason is that for a long time most experts had grown up within the traditional
system and mutually reinforced each others' views. Only recently did expertise
begin to diversify, and more detached judgments became possible.
To understand the future one must know the past. Hence, this book deals in
some detail with the historical and political context. This is done especially in
Part I which describes the dynamics of changes and the forces that transform
the traditional PTT system.
Part II is given to discussions of individual countries. All Western European
countries are analyzed, hence this section serves as somewhat of a handbook
on Western European countries' telecommunications systems, and provides a
bird's-eye view of the continent. At present, telecommunications knowledge is
highly segmented along territorial lines. It has proved to be an enormous un-
dertaking to treat so many countries and write a book that tries to transcend
national systems and go beyond theorizing. Some priorities had to be set, and
several of the larger countries received a more detailed treatment. Eastern Eu-
rope, whose societies have only recently rejoined the European mainstream and
the dynamics of its telecommunications evolution, is treated only briefly.
In the discussion of some countries, I have selected particular issues for
emphasis. For example, procurement of equipment by a PTT is given greater
attention in the chapter on Germany; industrial policy is emphasized in the
chapter on France; and the United Kingdom chapter looks at a PTT reorgani-
vi Preface
zation and its aftermath. Thus, the country chapters, taken together, also pro-
vide a mosaic of policy issues across Europe.
Part III discusses several special topics in European telecommunications, such
as industrial policy, the struggle with American interests, internationalization
of conflict, and vertical integration into new fields. Value-added services, in-
tegrated networks, transborder data flows, and international services are further
examples of new policy issues. The last part of the book analyzes the future of
networks, the instability of the international system, and the modularization of
networks. It concludes with an outlook on the future policy agenda and the
emerging new open network environment, which is very different from the
traditional one.
The book is not a comparative study in the sense of contrasting different
countries' approaches to a problem and evaluating the best of them. This is
done only occasionally, both among European countries and with respect to the
United States. More interesting than the comparative approach"which PTT
procurement system is best?"is the institutional one"how and why did
Germany's procurement system evolve? How and why is it changing?"
Although the treatment of an entire subcontinent with some specificity is
worthwhile, it is also difficult. In each country there are numerous experts who
know more about the subject matter for their own country than any single
author could. Hence, it is easy to be critical of the broader analysis by refer-
ence to superior knowledge of the details in a subplot. I have tried to deal with
the specifics by investing much time in visits; in reading academic, trade press,
governmental, and consulting literature; and in engaging in voluminous corre-
spondence. Drafts of the chapters were sent in each country to a dozen or more
individuals and institutions, sometimes in several rounds, and most of the com-
ments and corrections were incorporated. Often the replies and corrections from
the same country were contradictory, mirroring the policy debates in those
countries themselves. Even after the corrections, I am sure that national experts
will be able to identify some inaccuracies. Some may be based on sources, of
which many are unavoidably secondary rather than primary; others will be due
to updates to information presented in the book. Some perspective is necessary.
Although a change in the particulars may be of critical significance to those
directly involved, the broader picture is rarely affected by it.
In numerous contacts with PTT officials, I have become used to defensive
reactions to any suggestion that the traditional arrangement was not solely aimed
at optimally serving the public's need for a telecommunications infrastructure.
One form of response was dismissal of alternative views as having an "Amer-
ican perspective" (i.e., not being grounded in the traditions of another coun-
try). It is, of course, true that all institutions are historically grown. Indeed,
that is the very premise of this book. But PTT officials, who are typically
electrical engineers turned civil servants, are rarely familiar with the genesis
and evolution of their own systems. Their perspective tends to be ahistorical.
Indeed, a second type of response, at tension with the first, has been that the
discussions in this book are relevant to the past but are "dated" in that things
have recently changed, perhaps due to a reorganization, a new technology, or
Preface vii
a new directive yet. This is as if one negated the relevance of childhood and
youth to an evaluation of an adult. Episodes along the way do not become
irrelevant to the understanding of an ongoing institution simply because the
rules have recently been changed. A third and related type of response has been
the flight into detail. It is this expert-oriented view of the world, favoring the
trees over the forest, that has kept discussions of telecommunications policy
out of the public realm.
Another form of defensiveness is to consider an evaluation that contains a
critique of the monopoly system as adversary and possibly inspired by business
designs. I have no such ties or commercial engagements. Nor do I ignore the
significant historical contribution made by PTTs to the evolution of their coun-
tries' communication infrastructure. PTTs were very important in the first phases
of network evolution; but as this book argues, telecommunications are moving
into a third stage, that of the pluralistic network. Its origin lies in the success,
not the failure, of the traditional system. The unravelling of that system by
various centrifugal forces cannot be contained, and we are merely at the begin-
ning of that process. Such a conclusion may be unwelcome, but it is not ad-
versary.
This is not a book about U.S. telecommunications. To be critical of a prac-
tice in a European country does not necessarily mean to approve of its U.S.
counterpart. But it is of course true that the perspective of this book has bene-
fitted from participation in American telecommunications policy and research.
I should clarify that much of it has been written or conceived before I took a
leave as a professor at Columbia University and as director of its Center for
Telecommunications and Information Studies (now the Columbia Institute for
Tele-Information: C.I.T.I.), and before I was appointed to the New York State
Public Service Commission. Nothing in this book reflects the views of that
regulatory body (whose international involvement is, at any rate, negligible).
The PSC and its staff of 650 regulate intrastate telecommunications for 10
million lines in the state, as well as electricity, gas, and water. It is in frequent
disagreement and litigation with the Federal Communications Commission in
Washington.
Thus, the views of this book do not defend the views of official Washington.
The New York State PSC, for example, instituted low-income telephone ser-
vice, based on a blueprint by the author, that provides basic telephone service
for about 1.5 million poor households for $1 per month, with similarly low
installation charges. This makes telephone service virtually free for poor peo-
ple. I mention this because one of the self-images of traditionalists in telecom-
munications policy is that only they are socially concerned. A large body of
literature sees change in telecommunications as a challenge by large business
interests, many of them multinational firms, to existing socially motivated and
home-grown arrangements. The reality, however, is much more complex. Do-
mestic monopolies are less rooted in social benevolence than its defenders claim.
The imperatives of an information-based economy and a rapidly developing
technology do not leave unscathed the institutional structure under which the
informational nervous system of society functions, and it makes little sense to
viii Preface
slay the messenger of that news. Changes came to the surt'ace first in America,
and subsequently reached Europe and Japan. It would be surprising if this pro-
cess were not to continue for decades and if at its end the various national
systems were not very different from those of today.
At the conclusion of this project, thanks are owed to a large number of people;
only a few can be recognized in these pages, and none should be held respon-
sible.
The project was made possible by financial assistance from the German Mar-
shall Fund of the United States and later from the Gannett Foundation. Peter
Weitz and Gerald Sass deserve the credit for supporting the work and keeping
it alive as its scope grew.
At Columbia, a lively collection of student assistants, editors, cite and quote
checkers, and typists participated in the project, later helping to update the
information. They were supervised first by Christopher Dorman and then by
Richard Kramer, who contributed in many valuable ways. Assisting them were
Hiiseyin Bayazit, Andrew Blau, Theresa Bolmarcich, Laura Bulatao, Dawn
Chang, Paul Chew, Andrew Day, Peggy Danneman, Sherry Emery, Christine
Enemark, Valere Gagnon, Rhonda Harrison, Christopher King, John Kollar,
Jessica Lee, Junno Lee, Catherine Lim, Alfred Lucas, Michael McManus, Kurt
Miller, Altagracia Miranda, Erica Simmons, Wendy Stryker, My-Phuong Tran,
and Mark Young. Barbara Martz was enormously helpful with press and arti-
cles from many sources. Roberta Tasley and Douglas Conn, as managers of
the Columbia Center, provided the necessary organizational structure. Aine
Ni'Shuilleabhain supervised the last and strenuous phase of the project. Rachel
Toor, Herbert J. Addison, and Steve Bedney of the Oxford University Press
guided the project to conclusion.
Among academics, private experts and public officials, both at home and
elsewhere, I am grateful to Per Klitgaard Andersen, Elena Androunas, Fran-
cisco Pinto Balsemao, Johannes Bauer, Marino Benedetti, Hans Bergendorff,
Catherine Bertho, Martin Bullinger, G. Buyck, Tom Byrnes, Farrell Corcoran,
Andrea Costa, James Cowie, William Drake, Martin Elton, Beth Eres, Ugo de
Fusco, Paul W.J. de Graaf, Bruce Greenwald, Karl Erik Gustafson, Jaakko
Hannuksela, Mario Hirsch, Johann Hjalmarsson, Wolfgang Hoffmann-Riem,
Janos Horvat, Elfriede Hufnagl, C. Jacobaeus, Herman Cohen Jehoram, Tuen
de Johngh, Tanya Kiang, Leon Kirsch, Christian Kobelt, Ismo Kosonen, Geo-
vanni Lanza, Michael Latzer, S. C. Littlechild, Giovanni Maggio, Pennetti
Mannisto, Horst Edgar Martin, Fergus McGovern, Heikki Myllo, Gosta Neo-
vius, Godefroy Dang Ngyuen, Sam Nilsson, Mogens Kuhns Pedersen, Francois
Pichault, G. Russell Pipe, Gerard Pogorel, Remy LeChampion, Sheizaf Ra-
faeli, Anthony Rutkowski, Shaul Hai, Joachim Scherer, Donald Smullin, Charles
Stabell, Matthias Steinmann, Jan Thurmer, Sylviane Toporkoff, Rudolf Trach-
sel, Sergio Treves, Jeremy Tunstall, Thierry Vedel, Timo Viljakainen, Lennart
Weibull, Christian von Weizsacker, Eberhard Witte, and Glenn Woroch. To
those whose names have been inadvertently omitted, my apologies.
Preface ix
Most of all, I owe this book to my wife Nadine Strossen, champion of free
speech and human rights in America and abroad; to my mother, who provides
the bridge across the Atlantic; and to the memory of my father.
New York E.N.

November 1991
This page intentionally left blank
Contents
I TRADITION AND CHANGE
1. Public Telecommunications: A Concept in Transition 3

2. The Establishment of the PTT System 7
3. Network Tipping: The Rise and Fall of the Public Network
Monopoly 26
4. Forces of Centrifugalism 43
5. Defense of the Telecommunications Monopoly 55
6. Policy Directions 61
II THE TELECOMMUNICATIONS SYSTEMS OF EUROPEAN COUNTRIES
7. Germany 69
8. The United Kingdom 103
9. France 133
Telecommunications in the Benelux and Alpine Countries
10. The Netherlands 169

11. Belgium 178
12. Luxembourg 184
13. Switzerland 186
14. Austria 195
Telecommunications in Scandinavia and the North Atlantic
15. Sweden 203

16. Finland 212
17. Norway 220
18. Denmark 224
xii Contents
19. Iceland 230

20. Ireland 232
Telecommunications in the Mediterranean Countries and Eastern Europe
21. Italy 239

22. Spain 251
23. Portugal 258
24. Israel 262
25. Turkey 267
26. Greece 270
27. Telecommunications in Eastern Europe 274
III BATTLEFRONTS IN TELECOMMUNICATIONS POLICY
28. The International Organizations of Telecommunications 293

29. Brussels Takes On the Traditional System 305
30. Telecommunications Policy as Industrial Policy 312
31. Transatlantic Trade Friction 322
32. International Telecommunications Services 331
33. The Economics of ISDN Integration 350
34. The Political Economy of ISDN 360
35. Value-Added Networks and Services 369
36. Videotex 382
37. Transborder Data Flows 392
IV THE FUTURE OF TELECOMMUNICATIONS
38. Networks in the Future 407

39. Toward a Modular Network 412
40. Telecommunications Liberalization: An Expansionary Process? 417
41. Regulatory Tasks for the Future: Interconnectivity in the Pluralist
Network 425
42. The Pluralist Network 432
Notes 441
References 457
Index 497
I
TRADITION AND CHANGE
1
Public Telecommunications: A Concept
in Transition
The Monopoly System
A number of far-reaching changes in telecommunications policy originated in

the United States. Because many were passed under a conservative political
regime, they were usually viewed in Europe as the product of American busi-
ness interests, wrapped in a Chicago free market economic ideology. But more
recently, Japan and several European countries have begun to adopt similar
policies, or at least to discuss changes that previously seemed unthinkable. This
indicates that the changes in telecommunications go deeper than the nature of
the respective governments in power and that they reflect a more fundamental
transition.
For a century, telephony throughout Europe had been a ubiquitous, central-
ized, hierarchical network operated by a monopolist. The operating entity was
usually a state administration for post, telephone, and telegraph (PTT), though
in some instances, the post was separated from telecommunications. The entire
arrangement was therefore known as the PTT System. In the United States,
AT&T fulfilled much the same function in telephony. Western Union, the tele-
graph monopolist, and the Post Office Department (later the U.S. Postal Ser-
vice) added the T and the P. The physical and organizational structure of PTTs
was hierarchical. Major policies were set by technical experts, largely outside
public scrutiny. To be sure, the arrangement served the important goal of in-
terconnecting society and operated as a mechanism of redistribution. It was not
merely a technical system, but a social, political, and economic institution based
on the sharing of resources and the transfer of benefits toward favored groups,
often the economically weak and almost always the middle class and farmers.
The origin of the centralized network system for communications preceded
electronics by centuries and was embedded in the emergence of postal monop-
olies. In 1505, the Hapsburg emperor Maximilian granted exclusive mail-
carrying rights to what one would call today a multinational company, the
Taxis family of Italy. Although this concession proved to be an unexpectedly
rich source of revenue to the Hapsburgs, who shared in the profits, it also
required vigilant protection from the incursion of mail systems, of which there
was a multitude (Dallmeyer, 1977). Neighboring Prussia went one step further,
3
4 Public Telecommunications: A Concept in Transition
and in 1614 established a state-run postal monopoly (Stephan, 1859). Thus, the
PTT system was born as a creation by the absolutist state for the absolutist
state. Later, this system was rationalized as based, depending on one's point
of view, on economies of scale, national sovereignty, cross-subsidies, or public
infrastructure needs; but the early creators of the postal monopoly system were
quite forthright in their mission to generate profits for the state and its sover-
eign. The postal system became a major source of revenue, at a time when
European rulers had insatiable needs for cash. As the goose that would lay the
golden eggs, the system was ardently protected through the centuries against
encroachment by private competitors and by other states.
When the telegraph emerged in the nineteenth century, it was rapidly inte-
grated into the monopoly. Later, much was made of the military importance of
state control over telegraphy. Although this consideration may have been sig-
nificant for the major powers, it was less relevant for the many other countries
that also banned private telegraphs. Even for the larger states, the strategic
importance of the new medium did not really require operations by the state
itself any more than it did for overseas mailan important service in the era
of imperialismwhere private delivery under contract was regularly used.
When the telephone made its appearance in 1876, it too was soon integrated
into the state monopoly once its financial viability became clear. Here official
histories claim that the purpose was to bring telephony to rural areas neglected
by commercial interests. This is true in some cases, but in other instances (for
example, Norway), the historical record is quite different, and does not lead to
find an aim of spreading telecommunications service across the country. At the
same time, telecommunications were integrated into an international system of
collaboration, with the official goal of technical coordination, but also, from
the beginning, with a cartel agenda on prices and service conditions.
For almost a century, a tightly controlled system of telecommunications was
in place in most developed countries. Its structure was supported by a broad
coalition that provided political support in return for a share in the monopoly
rents. This rent-seeking coalition can be termed the "postal-industrial com-
plex." It encompasses the government PTT as the network operator and the
private equipment industry as its supplier, together with residential and rural
users, trade unions, the political left, and the newspaper publishing industry,
whose postal and telegraph rates were heavily subsidized. The traditional sys-
tem worked in particular to the benefit of the equipment industry, which was
provided large markets by huge PTT procurements, especially after World War
II. These markets were also almost always protected from foreign competition
by buy-domestic policies. Within most industrialized countries, equipment
manufacturers often collaborated in formal or informal cartels that set prices
and allocated shares of the large PTT contracts. The structure of telecommu-
nications in the United States, although private, was not all that different, be-
cause it was a near-monopoly, with a full integration of network operation and
equipment manufacturing. Its corporate ideology was shaped by AT&T's pa-
tron saint, Theodore Vail, himself a former postal man as the head of the U.S.
Railway Mail Service.
Public Telecommunications: A Concept in Transition 5
The system was profitable and reassuring for insiders, and its inefficiencies
were hidden by the general downward trend in the cost of electronic technology
(which was due, on balance, more to computer and component firms than to
traditional telecommunications firms.) The PTTs also set standards for equip-
ment in a way that would often discourage or delay outsiders, and they often
collaborated with favored domestic firms in equipment development and in ex-
port promotions.
In Switzerland, for example, the PTT in 1984 set standards for cordless
telephones by issuing a fifty-five page specifications manual and requiring no
less than forty duplex channels and automatic scanning. Because of these rules,
which were allegedly passed to protect the users, only one company, a Swiss
one, could meet the standards quickly (and not surprisingly, since it had played
a major role in developing the rules). Moreover, the manufacturer's estimated
price to the PTT for a set was almost $800, and monthly rentals came to about
$15 (Wolf, 1983). At the same time, one could buy a simpler but perfectly
adequate cordless telephone in the United States for less than $75. Swiss users
resorted to buying cheaper but illegal foreign equipment in the many stores
where they were marked "for export only."
Although the traditional monopoly systemthough not always its execu-
tionoperated to the economic advantage of equipment firms, its principle long
enjoyed broad public approval from the constituent groups of the postal-indus-
trial coalition. The monopoly followed concepts of public service: universal,
accessible, affordable, and redistributive. As a public service, telephony was
outside the mechanism of the market, even in otherwise free-economy coun-
tries. Any change in that status was bound to be controversial; an expansion of
the realm of the market into the realm of rights and politics has always been a
painful transition.
Thus, a partial relocation of telecommunications from the public domain into
the economic one was objectionable to many. Indeed, the single most powerful
argument in defense of the traditional centralized system is a value preference
for the principle of state ownership in infrastructure as distinguished from the
"scientific" arguments that a monopoly is necessary for engineering and eco-
nomic reasons.
PTTs were not entirely frozen in their institutional development. After many
decades of stability, some of them were transformed in the late 1980s into
"PTOs", public telecommunications organizations. The new designation con-
notes a separation from postal services, and change from a governmental ad-
ministration into a state-owned semi-independent organization. Furthermore,
some new types of activities such as value-added services and mobile telephony
were permitted to be offered by private firms. But these changes do not prove
that a major reorientation had taken place, and that the problem of state mo-
nopoly had been resolved in most countries.
The exaggeration of the extent of actual change can lead to bureaucratic
doublespeak. For example, the Danish government, in creating the PTO
teleDenmark, declared, "As a consequence of decisions made at the EC level,
there will be competition within all spheres of telecommunications in the next
6 Public Telecommunications: A Concept in Transition
few years, apart from telex, ordinary telephones, radio-based mobile services,
satellite services, the infrastructure and the use of telecommunications network
for broadcasting. . . ." (Ministry of Communications, 1990). In other words,
"everything" will be competitive, except for the remaining 95%. In almost
every country there still exists a tight monopoly over transmission infrastruc-
ture, switched services, and voice servicethe vast majority of telecommuni-
cations activities. The liberalized exceptions are minor in comparison, and, in
the case of value-added services, consist of sophisticated computer-based activ-
ities in which the PTTs had only a limited participation. Similarly, the separa-
tion from postal service and the creation of a more independent status repre-
sents no reduction in economic power. To the contrary, postal services had
become a financial drag on telecommunications. Independence permitted a
branching into new activities, including the vertical expansion into manufactur-
ing, and the horizontal expansion internationally. Regulation was left to tiny
government departments that were hard-put to control some of their countries'
largest and most complex organizations.
This is not to say that no change has taken place. The traditional PTT system
was stable for a century. But in the 1980s pressures emerged, just as they did
in the television sector, that could no longer be contained. This was accompa-
nied by harsh political disputes. By the end of the decade, change was in the
air everywhere. The postal and telephone monopolies were being reorganized
in most countries; some competitive suppliers were emerging; intraorganiza-
tional private networks were increasingly taking traffic from the public net-
work; and the European Commission in Brussels was hectoring the national
governments to loosen restrictions. The actual extent of the change should not
be exaggerated, given the inertia of the past; but once the process is set in
motion, further transformation is inevitable. This process will continue, and
will lead to a telecommunications infrastructure that is very different from to-
day's.
These are the themes this book discusses, both in general and in country-
specific terms.
2
The Establishment of the PIT System
Early Postal Systems
The key institutions of European telecommunications are the PTTsthe gov-

ernment administrations of posts, telephone, and telegraph. These formidable
organizations did not emerge overnight, but are the result of centuries of strug-
gle. Most Europeans are not aware of the historical background that led to the
formation of a monopoly system and to its institutions that are part of their
environment.
Pressures for liberalization of the monopoly system are not a recent phenom-
enon. The history of postal monopoly from the Middle Ages onward has been
one of the efforts to establish and defend monopolies against the competition
of private services and rival governments. These drawn-out rivalries put into
question the assertion that postal serviceand later telecommunications, which
were usually merged into the postal systemis a "natural monopoly" whose
economic and technical characteristics secure it from competitors. It was only
in the late nineteenth and early twentieth centuries that governments had finally
successfully suppressed attempts at rival service and consolidated postal powers
in an enforceable fashion. To understand how the monopoly institutions emerged,
it is necessary to take a look at their history.
Postal services go far back. In antiquity, the Medes and the Persians had
systems for conveying messages that used human shouts transmitted from hill
to hill, as well as routes of runners. Xenophon observed that the Persian kings
had established rider routes with stations at which horses were changed. The
Romans set similar routes. In the thirteenth century, Marco Polo described
regularly spaced "poeste" on China's main roads, where imperial messengers
would change horses. He reports that there were 10,000 posts in the Mongol
Empire with 300,000 horses. "The thing is on a scale so wonderful and costly
that it is hard to bring oneself to describe it" (Collins, 1960). Messages moved
at 25 miles a day, but urgent dispatches could be speeded up to an incredible
400 miles a day. The Spaniards observed similar Inca runner systems in Peru
in the sixteenth century.
In medieval Europe, the origins of commercial postal service were messen-
ger services run by such disparate institutions as city-leagues, guilds, religious
orders, and commercial operators. In many instances, traveling merchants and
journeymen (particularly butchers, for some reason) carried messages. Over
7
8 The Establishment of the PTT System
time, regular routes developed, centered around inns at which messengers would
rest. It became customary for patrons to leave messages for delivery and trans-
portation at these "posts," and for messengers to exchange horses, thereby
considerably speeding up transportation. Two basic forms of organized postal
service emerged. The first was private service, subject to governmental author-
ity. Its prime example was the Thurn-and-Taxis postal system. The second was
a state-operated postal service, such as the ones in Prussia and in England that
became the direct ancestors of the modern European PTTs. These two types of
systems and their struggle with each other will now be described.
Private Postal Service: The Taxis System

The Tassis family had already provided courier services in Italy for several
generations, when in the fifteenth century the Hapsburgs required better com-
munications for their far-flung European possessions. King Maximilian I turned
in 1489 to Janetto Tassis, appointing him as exclusive postmaster. In 1501,
Phillip the Fair of the Spanish branch of the Hapsburgs similarly appointed
Janetto's brother Francesco Tassis, now Germanized as Franz von Taxis, to
rule as the captain and master of the post in the Netherlands. Five years later,
Taxis also received a concession to run postal courses between Spain and the
emperor in Austria, at his own initiative and risk (Dallmeyer, 1977).
From the beginning, the Taxis were obligated to transport state letters at no
charge as a benefit to the Hapsburgs, and to contribute to the imperial coffers.
This was the origin of the system of cross-subsidies that still survives today;
from the outset the Taxis family consciously used it as an argument for its legal
monopoly. Their system also received operational privileges, including the use
of the postal horn (which established priority on the highway), freedom from
local road and bridge tolls, royal protection for its messengers, and the right to
requisition local help.
The Taxis monopoly franchise proved unexpectedly profitable. One com-
mentator observed in 1748:
Everybody considered such [postal] institutions as misguided, and nobody could
imagine that the letters and items of merchants and other people would generate so
much postal revenue that horses, wagons, coachmen, and postal servants would be
supported. But as soon as the German merchants became aware that they could
obtain the [information] on the rate of exchange, the tariffs and the prices of all
goods, for little money, without therefore having to travel to Antwerp, Brussels,
etc., the new post of Taxis drew such an incredible number of letters together, that
the von Taxis had such a rich surplus of money from the postal service, as a
medium-size German principality could barely generate [von Beust, 1748].
Now the government authorities' appetite was whetted. The previously ig-
nored postal service was recognized as a rich source of revenue, and this led
to bitter disputes about the rights of regional principalities and of the free cities
to operate their own systems. It was readily understood that the coexistence of
The Establishment of the PIT System 9
rival services would lower profitability. The Taxis were supported by the Ger-
man emperor, who advanced the legal theory, profitable to both, that the right
to grant a postal franchise was an imperial prerogative. But this was frequently
challenged or ignored, and the Taxis had to work ceaselessly at suppressing
other private messenger services and governmental postal services of smaller
principalities and cities.
In particular, during the period of the Dutch rebellion against Spain, the
Taxis system of a loose family confederation operating under various govern-
ment concessions unraveled, and competing local messenger guilds emerged.
Since competition posed a threat to the imperial revenues, Emperor Rudolf II
initiated in 1579 what has since become second nature to European govern-
ments: He prohibited private competition and appointed a blue-ribbon commis-
sion to study postal reforms. The commission included two members of the
Fugger familybankers to the emperors and large-scale merchants (in today's
terms, user and creditor representatives). The commission, unsurprisingly, re-
stored control to the Taxis. In 1595, Leonhard von Taxis was appointed impe-
rial postmaster general. He and Emperor Rudolf II sent notices around to sup-
press the "lesser, butcher and messenger posts."
But competitors and rival authorities kept disputing the monopoly privileges
granted to the by now Count von Taxis. Duke Frederick of Wurttemberg an-
notated the emperor's missive, in his own hand: "Because there is no obliga-
tion, one must not obey, as we will not do, but ask Your Majesty to place your
posts elsewhere, as it has been in past days, so it shall remain" (von Beust,
1748). The states also accused the Taxis monopoly of unreliable service and of
excessive charges. Yet subsequent Hapsburg monarchs renewed the Taxis rights.
Charles V even extended them, after his coronation in 1630, for his entire
realm, over which the sun did not set. The Thirty Years War complicated the
Taxis fortunes by introducing commercial, religious, and diplomatic intrigues
of major proportions. Emperor Ferdinand III, whose war finances depended
upon Taxis contributors, was induced again to proclaim the prohibition of rival
messenger service where the Taxis operated, even though this antagonized many
cities and principalities that still operated their own systems.
The eighteenth century was the zenith of the private Taxis system. The Em-
peror's financial needs continued, of course, and the head of what had been a
humble family of messengers was now promoted to hereditary imperial prince,
head of the House of Thurn and Taxis. But this concealed the decline of the
system: The powers of the emperors were waning, and with them the founda-
tion of Taxis power. The Thirty Years War had created a more decentralized
governmental structure in Central Europe, and the territorial rulers no longer
recognized the imperial postal monopoly claim. Whatever the constitutional
legalities, the territorial rulers controlled roads over which postal service had
to pass. Furthermore, the emperor's arguments for a monopoly system were
much weakened when he established state-owned posts in his own Austrian
domains.
These disputes over postal monopolies lasted for centuries. The major states
unilaterally established their own posts on whatever routes they found profit-
10 The Establishment of the PIT System
able. In today's language, they engaged in "cream skimming," though it was

done by governments against a private operation. Eventually, Taxis operated
mostly in the smaller or less desirable territories.
The Development of the Classic PIT: The Prussian Post
The Prussian Post established the institutional model of a state postal adminis-
tration for much of Europe and the rest of the world. Its early history is one of
a centuries-long struggle against the private Taxis system. And although most
of this rivalry is long forgotten, traces of antagonism are no doubt left within
its institutional subconscious.
The direct ancestor of today's giant German Bundespost Telekom system
started out as tiny semigovernmental service in Brandenburg (Prussia's predecessor)
with twenty-four messengers. Initially, the system operated only for govern-
mental use. Soon, however, some private service was provided on the initiative
of the postmasters who directly shared in the profits. In 1649, the Prussian
government formally took over the system and expanded routes, relay stations,
postmasters, and agreements with foreign governments.
Government monopoly did not necessarily mean state provision of postal
services. Local postmasters were not paid a salary, but could levy charges on
letters. Private concessionaries that were compensated by the local postmasters
provided much of the actual transportation of the mail. They were required to
keep to the established schedules; fines were levied for tardiness.
Soon rivalries flared up with the messenger services of free cities who were
prohibited to pass through Prussian territory. Danzig messengers armed them-
selves; pitched battles broke out and led to temporary warfare that also involved
other states. The Taxis were able to buy rights of transit through Prussia, but
were prohibited from picking up or distributing any letters on Prussian territory.
The state's monopoly was extended to transporting coach passengers, an ar-
rangement directly leading to today's railroad monopolies.
Where state monopolies were not enforced, a wide diversity existed. In 1695,
postal customers in the Free City of Hamburg could choose among local postal
offices affiliated with Sweden, Denmark, Prussia, Brunswick, Thurn and Taxis,
Holstein, Mecklenburg, Saxony-Gotha, and Nuremberg, plus various private
messenger and delivery services. In 1712, the General Prussian Postal Order
was issued, reiterating the governmental monopoly. On occasion, private op-
erators were licensed, but only on routes or for time periods that were noncom-
peting. They were usually not permitted to change horses and were otherwise
restricted. However, the suppression of private service providers was not easy.
As Heinrich Stephan, later the Prussian Postmaster General, writes:
The civil and military administrations were instructed to carefully seek the main-
tenance of the state postal rights. If the efforts in that respect were not always as
successful as desired, one should not overlook that the total elimination of the
previously very extensive private-messenger institutions is a difficult task, whose
Figure 2.1. Revenues, Expenditures, and Profits of the Prussian Postal Service.
(Source: Stephan, 1859, pp. 130, 297-299.)
successful solution could have been reached only gradually. (Stephan, 1859, p.
51).
The postal system became highly profitable. In 1662, Prussian postal profits
were 7000 Thalers on 10,000 Thalers expenses. Twenty years later, profits had
quadrupled to 29,058 on 22,902 expenses, a substantial margin. And in 1688,
they were 39,213 on 40,758 of expenses (Stephan, 1859, p. 59). See Figure
2.1. The profits of local postmasters were similarly considerable. Neumann,
the postmaster of Konigsberg, enjoyed an annual income of about 2000 Tha-
lers. To put these figures in perspective, in 1623 a meal of about four courses
in a Berlin inn "suitable for persons of nobility" cost about 1/4 of a Thaler;
day labor cost 1/24 of a Thaler, plus meals. Messengers were compensated at
1/12 of a Thaler per day, plus about 1/5 of a Thaler for an average day's
mileage.
In subsequent years, the profits of the Prussian postal service continued to
increase. These profit figures are still understated, because they do not include
the sale of postmasterships.
What were the profits used for? In 1699, the first year for which a breakdown
is available, profits were a remarkable 85,000 Thalers on 68,000 expenses. Of
the net gain, 45,000 Thalers went to the monarch individually, 6000 to his
wife, 6000 to the crown prince, and 4000 into construction of the elector's
Berlin residence. Three thousand Thalers were used to support the sciences
(Stephan, 1859, p. 130). Thus, almost two thirds of the considerable profits of
the post directly benefitted the royal family, and 4 percent went into what we
would call today research and development.
During the forty-six years of Frederick the Great's rule, the postal service
generated 20 million Thalers, much of it going directly to the king, who de-
pended considerably on this revenue. In addition, appointment to a postmaster-
ship was accompanied by payment to the royal coffers. The king paid an ex-
traordinary amount of attention to the details of the postal service. Given the
rich flow of revenues, protection against postal competition became an impor-
tant part of foreign policy.
Seeking still higher returns to defray the cost of the Seven Years War, Fred-
erick the Great decided to raise postal rates considerably and to expand the
state monopoly to packages. A group of French fiscal specialists organized the
changes. Users and foreign postal administrations initially refused to pay the
higher rates, but to no avail. For the postal civil servant Heinrich Stephan (later
a celebrated postmaster general), writing in a later age in which postal admin-
istrators were still guilelessly proud of making profits for the state rather than
skillful in understating them, this was proof of the strength of the monopoly:
In the seventeenth century when private companies still existed, "a general
increase in the letter mail rate would have been a signal for the ruin of the
postal institution, and for the reemergence of a whole number of private postal
institutions" (Stephan, 1859, p. 294). But this was no longer true in the eigh-
teenth century. The power of the absolutist state made it possible to increase
rates that a century before would have been unsustainable because of competi-
tion.
Internally, however, the increase of postal rates was controversial. Freder-
ick's ministers, in a report to their sovereign about the decline of Prussian
commerce, ventured respectfully that the postal rate increase was a contributing
factor. This 1766 report is a remarkable document in its aversion to mercantil-
istic restrictions and its support for a free-market philosophy, in the same vein
as Adam Smith's Wealth of Nations, published a decade later. The report is a
direct ancestor of today's pleas by various ministries of economics or industry
in favor of less restrictive postal policies. And it was similarly controversial.
Within twenty-four hours, the king strictly prohibited his ministers from engag-
ing in any discussion along these lines. But it turned out that the king's min-
isters, accused by him of ignorance, malice, and corruption, had a better un-
derstanding of economics than their sovereign, since the total revenue of the
postal service declined for some time. Even the absolutist monarch was unable
to abolish demand elasticity. Moreover, several foreign postal services avoided
Prussia as a transit route. Thus, whatever competition that did exist for routes
exerted itself. In time, revenues recovered. But to Heinrich Stephan, the archi-
tect of the telecommunications monopoly system a century later, one of the
lessons of the initial setback must have been that rate coordination between
neighboring countries was necessary to protect monopoly power, just as do-
mestic exclusivity was its precondition.
In the 1720s, several European states contemplated the establishment of a
cooperative postal arrangement that would permit a route from Amsterdam all
the way to Danzig and Petersburg, entirely bypassing Prussia and establishing
a major alternative to its postal service. Prussia stood to lose 75,000 Thalers a
year in revenue. It took all its diplomatic, secret service, and commercial ef-
forts to stifle the establishment of the rival system, but the competitive pressure
also led to considerably improved postal service on the threatened route. Most
important, it entered into an alliance with its long-time rival Taxis, which was
similarly threatened by the proposed new system. In 1723, the former compet-
itors reached a treaty against the common enemy; they agreed upon a code of
conduct, and divided routes among themselves. Though they hardly needed the
reminder, they also promised to be vigilant in their suppression of independent
postal carriers. The relations between the Prussian and the Taxis systems be-
came even closer, when Alexander von Taxis, elevated to prince in 1754 be-
cause of his enormous wealth and financial support of the emperor, also sought
to buy a seat in the imperial electoral college, which included the handful of
the highest nobles of the realm. Frederick the Great supported him in return
for postal concessions. But the agreement proved short-lived, because of the
Seven Years War that engulfed Europe. Taxis eagerly displaced Prussian posts
when Frederick's war fortunes were low, and when Prussia turned the tide of
war, it too expelled the Taxis postmasters and substituted its own.
During the Napoleonic Wars, the Prussian Post and the Taxis system were
again in conflict over their rights. When the Holy Roman Empire came to an
end in 1806, the position of the Taxis post, based upon the grant by the em-
peror, had lost its foundation. In a conflict with strong secular powers, it was
pushed aside, regardless of its performance as a carrier.
After the Vienna Congress in 1815, the increased flow of traffic led to a
large number of bilateral agreements modeled after the Prussian-Taxis arrange-
ments, which regulated the relationships of postal services. In the following
decades, coordinating bodies developed both for post and for the newly devel-
oped telegraphy. A regional postal union was created in 1850 and was soon
expanded to telegraphs. It provided the model for later European and later
global collaboration of state telegraphs and telephones.
In 1866, the Thurn und Taxis Reichspost, which still existed in fifteen small
states, had shrunk to 15 percent of the size of the Prussian Post. In that year,
Prussia went to war with Austria over hegemony in Central Europe, and occu-
pied Frankfurt, the Taxis's headquarters. Thus ended the 350-year-old Taxis
system. The family, left with a huge fortune but no postal routes, was hence-
forth relegated to an active role in the society pages.
Von Stephan, Germany's postmaster general and later state secretary and
minister, played a remarkable role in the development of European postal and
telecommunications institutions. Born in 1831, as the son of a tailor, he began
working as a postal scribe and rose rapidly (Grosse, 1931). A man of unusual
breadth, he wrote monographs on the history of transportation and postal ser-
vice, contributed such innovations as the postcard, and restructured German
and international postal service. Stephan became postmaster general and inte-
grated the German postal systems; later he was instrumental in the international
agreements that led in 1874 to the establishment of the International Postal
Union.
In 1876, the newly established German state decided to expand its backward
telegraph system. Von Stephan, a postal man, was appointed to head the tele-
graph office. Within a short time, he had merged the two services, despite
opposition from the more technical telegraph personnel. Von Stephan also rap-
idly introduced the telephone. Its purpose, however, was distinctly different
from that of its American counterpart, where the telephone became established
in private homes and businesses. The Reichspost viewed the telephone as an
extension of state telegraphy, to be used in rural post offices, where trained
telegraph operators were economically infeasible.
In the German state of Wurttemberg, which maintained its independent postal
system even after the German unification, the American Bell Telephone Com-
pany sought a government concession to provide telephone service in the cap-
ital city of Stuttgart. Impatient after bureaucratic delays, the Bell representative
started wiring without a license until he was stopped by the police. The postal
authorities soon took over. This was Germany's only brush with private tele-
phone service for more than a century, until the 1990s. In 1892, the compre-
hensive telephone authority of the state was cemented into law.
An account of the further history of German telecommunications is provided
in Chapter 7 on Germany.
The Establishment of the PTT Monopoly in Britain
In the United Kingdom the monopoly for postal and telecommunications ser-
vices was also far from "natural," but rather the outcome of fierce economic
and political battles for control. This struggle for a British postal monopoly and
its extension to telecommunications will be discussed in greater detail than the
similar developments that took place in other European countries.
British postal service began with informal messengers for the royal house-
hold. In 1481, Edward IV created a more organized route system for govern-
mental use with stations 20 miles apart. It was not a monopoly, however.
Private letters had to be carried by a variety of messengers until 1512, when
Sir Brian Tube, the first recorded English postmaster general, opened the Royal
Post to nonofficial letters. The system stagnated during the reign of Queen
Elizabeth (1533-1605). In 1590, John Lord Stanhope was made hereditary master
of the posts. Stanhope obtained a national monopoly and subcontracted specific
postal routes to local postmasters, often for substantial sums. These postmas-
ters, frequently innkeepers, also held the local monopoly over the hiring out of
horses to travelers. For some services, particularly international ones, the offi-
cial postal service had to compete with the rival Foreigner's Post and the Mer-
chant Adventurers' Post (Hemmeon, 1912).
In 1591, a royal proclamation affirmed the government's claim to monopoly
over postal services to foreign countries. But in 1619, a rival operator by the
name of de Quester, attracted by the potential profits, obtained from the King
the monopoly rights for "foreign parts," leading to protracted litigation and
intrigue. Eventually, de Quester transferred his interests to Thomas Withering,
who revolutionized English postal history in 1635 by transforming a tottering
system whose profits accrued to licensed private operators into a rich source of
revenue for the state (Clear, 1940, pp. 21-32). Withering, supported by Charles
Fs royal proclamation, brought remarkable operational planning to his postal
reorganization, which endured for more than two centuries. Withering spear-
headed a postal reform that regularized and speeded up service considerably,
to up to 120 miles in twenty-four hours. In the process of reform, however,
Withering made enemies: Within two years he was dismissed and replaced by
the two powerful secretaries of state, Lords Coke and Windebank (unlike With-
ering members of the High Nobility) and by a wealthy creditor to the king.
Petitions by the London merchants to retain Withering's control were to no
avail.
Under Cromwell's Commonwealth, the farming out of postal routes was re-
sumed. The postmaster-generalship itself cost an annual 10,000. After the
Restoration, Lord Stanhope resurfaced with claims, but the monopoly patent
was eventually given to the Duke of York, in a grant estimated by Parliament
as worth 21,500 per year. Another 5000 of postal income was assigned by
the king to his mistresses and favorites. Slowly, postal service improved with
political stability. But the Royal Post still operated primarily on a handful of
great roads emanating from London. "Cross-and-bye" posts linking provincial
towns were rare, and neither urban nor rural service was provided. These could
be offered by private "common carriers." Disputes between the official and
private posts led to physical casualties and forced the government service to
cut its rates substantially.
In 1680, a London entrepreneur, William Dockwa, set up a remarkable pri-
vate urban mail system, the London Penny Post, which created at once a level
of service unsurpassed ever since: hourly collection at 179 postal points, four
daily deliveries for residences, and six to eight for business centers. The charge
per letter and parcel was a uniform 1 penny for up to 1 pound. Postage was
payable in advance and was credited by stamping the letters, a novelty. Dock-
wa's venture required a risky initial investment of 10,000; but when it proved
highly profitable, the Duke of York muscled in with the help of the courts.
Dockwa was pushed aside, and compensated with a minor pension, while a
multitude of courtiers with no connection to the postal system drew large rev-
enues. The Duchess of Cleveland, for example, was given 4700 a year. In
contrast, the universities of Edinburgh and Glasgow received a meager subsidy
of 210 each from the postal revenue. The lease of postal offices generated
crown income of 65,000 in 1685. In 1650, when a London municipal postal
service was proposed whose profits would benefit the poor, Parliament sup-
pressed it.
More fortunate than Dockwa was another private businessman, Ralph Allen,
who obtained permission in 1721 to serve the cross-and-bye posts linking pro-
vincial towns in return for 6000. After two years of heavy losses, the service
Figure 2.2. Revenues, Expenditures, and Profits of the Post Office in Great Brit-
ain. (Source; Hemmeon, 1912, pp. 245, 247, 252.)
turned very profitable. (Allen died with an estate reportedly worth 500,000.)
After 1765, major road improvements were undertaken, which led to the estab-
lishment of stage coach service by private operators who soon undercut the
official postmasters by also carrying mail. The Royal Post set up its own rival
"mail coaches" in 1784.
Postal service was highly profitable, as Figure 2.2 illustrates. Until the mid-
nineteeth century, for each pound of expense there were more than one or even
two pounds in pure profits. Nevertheless, the appetite of the exchequer for
more revenues was unlimited; the urban Penny Post became the Two-Penny
Post in 1801, and soon thereafter the Three-Penny Post. Interurban and inter-
national rates were similarly increased. Part of the reason for the rate increases
was the need to finance the Napoleonic Wars. Another significant reason was
the need to cover the increasingly costly exemptions from postagethe frank-
ing privilegeof members of Parliament, which at times degenerated into ex-
tensive rackets operated by members for their constituents and supporters. The
revenue drain was estimated at 140,000 for the year of 1763 alone. Another
major revenue loss was the virtually free mailing of newspapers.
Throughout the seventeenth and eighteenth centuries, royal proclamations
reaffirmed the government's monopoly over the lucrative postal service. Their
very repetition suggests a steady challenge of the exclusivity of mail service,
particularly when it was expensive and unreliable. Many of the private common
carriers conveyed some letters over the official routes, and government regula-
tions were enacted literally to slow them down. Later, the 1855 Annual Report
of the British Postmaster General argued forthrightly: ' 'The object of the Crown
in establishing this letter post was probably quite as much the formation of a
profitable monopoly as the accommodation of the public" (Stephan, 1859).
The system became intolerable for a country in the midst of the industrial rev-
olution. In its late stages, the monopoly was widely evaded. Private parcels
crammed with letters were sent by coach to the major cities for redistribution
and delivery. Stagecoach guards traveled with pockets full of letters which they
posted in their city of destination (Edwards, 1879, p. 87). In one location,
calculations showed that only one letter in fifty went through the official post
office. One witness freely admitted violating the postal monopoly about 20,000
times and being caught only once. The postmaster of Manchester testified that
probably half of all letters between Liverpool and Manchester were transported
illegally. Another witness estimated that more than five-sixths of all letters
between London and Manchester bypassed the postal service.
Thus, postal reform was not only proposed to aid commerce, but also to
reduce the role of private services.
In 1840, the postal service underwent major change. Sir Rowland Hill, a
noted reformer of taxation and education, though no postal expert, issued a
private report in 1837 advocating that postal service be a tool for promoting
economic activity rather than an instrument of revenue generation (Hill and
Hill, 1880; Smyth, 1907). The key to reform was a significant reduction of the
postal rate for a regular letter to a distance-insensitive penny, creating the fa-
mous "penny post." Another feature was the use of postage stamps for pre-
payment of charges. Th'ese measures radically reduced the cost of handling a
letter. Before the postal reform, the tariffs for a regular letter were, for ex-
ample, 4 pennies for distances up to 15 miles or 9 pennies for a destination
between 18 and 120 miles away. Hill expected traffic volume to increase and
eventually offset the revenue lost due to rate reduction. In the first year after
reform, traffic volume more than doubled, and the number of routes increased.
Postal revenues dropped, however, from 12.4 to 11.4 million. It took ten
years for the old revenue level to be regained, and almost twenty years to reach
the former profit level. When the post office showed a 7 million deficit, the
Whig government, which had instituted the "supply-side" reform as a measure
of its initiative to liberalize trade, was compelled to resign from office. The
reform, nevertheless, greatly improved postal service in Britain. It also secured
the previously discredited state monopoly system by making it uneconomical
and illegal to compete against the low penny rate.
The Struggle for a Telegraph Monopoly in Britain
The telegraph was introduced into the United Kingdom almost at the same time
as the postal monopoly was made secure. Edward Cooke and Charles Wheat-
stone filed for a patent in 1837, four months before Samuel Morse registered
his own system in the United States. In contrast to Morse's system of dots and
dashesan early variant of digital transmissionthe British telegraph relied on
an analog system in which a needle pointed to letters of the alphabet. The

Cooke-Wheatstone system was easier to operate, but much more expensive in
its capital equipment. It originally required five wires, eventually reduced to
two, as opposed to Morse's one-wire system. Despite the technical differences,
Cooke and Wheatstone successfully prevented Morse from obtaining a British
patent.
Construction of the telegraph commenced in Britain earlier than in the United
States. The first operations, undertaken by the privately owned Electric Tele-
graph Company, proved profitable. Subsequently, the Magnetic Telegraph
Company entered the market in a competitive fashion and maximum rates fell
from 10 shillings to 4 shillings; but because of the high elasticity of demand,
total industry profits increased (Brock, 1981). Eventually, the two companies
entered coordinating agreements.
The telegraph network expanded rapidly: Between 1851 and 1855, the Elec-
tric and International Telegraph Co. and the British and Irish Magnetic Co.
alone tripled their mileage and multiplied their business volume by 700 percent.
In 1861, a rival firm, the United Kingdom Telegraph Company (UKTC), en-
tered the market with a low-rate service. The new company was vociferously
opposed by the earlier entrants, but under the competitive regime rates dropped
to less than 1 shilling for those routes that were served by UKTC. In 1865, the
new company was integrated into a new cartel and rates were fixed at 1 shilling
for under 100 miles and 2 shillings for distances above 200 miles.
Soon, however, the private cartel character of the telegraph was challenged,
particularly by the newspaper publishers, for whom telegraph rates had become
a large expense. Because they expected lower rates and better service under
public ownership, newspaper owners began calling for nationalization. They
were supported by the British Post Office, which argued that the telegraph was
related to the mail-carrying function, on which it had a monopoly. Proponents
of nationalization claimed that duplication would be eliminated, that smaller
towns would receive service, and that telegraph employees would be better
paid. Parliament, persuading itself that it could reduce telegraph rates and at
the same time compensate the cartel for future monopoly profits, in 1868 au-
thorized a takeover of telegraphs by the Post Office. Consistent with the goal
of self-financing, it did not appropriate funds for compensation. The telegraph
companies received a settlement of twenty times the 1868 profits, or about 10
million, which was four times as high as had been originally estimated. This
created a public furor, as well as difficulties in repayment, and deeply influ-
enced attitudes toward telephone policy in the coming decades.
The press came out well ahead, with new rates at one tenth of the regular
charge during off-peak, and one seventh of the regular charge during regular
times. Its traffic volume was prodigious. By 1895, the press transmitted two
fifths of the total number of words sent over the telegraph, generating an esti-
mated revenue loss of 1 million. The number of telegraph offices was greatly
expanded and salaries increased. Overall, the financial position of the telegraph
was shaky, and by undercutting its revenue the introduction of the telephone
threatened to make the government policy toward the telegraph look even worse
than it already did.
In 1850, a submarine telegraph cable was laid from England to France, but
it failed after a short time. A transatlantic cable project was started by the
Atlantic Telegraph company but was soon abandoned. In 1858, another trans-
atlantic cable between the United States and Britain resulted in exchanges of
messages between Queen Victoria and U.S. President Buchanan, but the cable
broke down almost immediately and could not be revived. Eventually, another
cable provided more durable service.
The Struggle for Telephone Monopoly in Britain
The establishment of a British telephone monopoly was also anything but nat-
ural. When the telephone arrived in Britain, its spread was slow compared to
the United States and several continental European countries. This was not due
to a lack of public attention to the new technology; indeed, from the beginning
the telephone created a sensation. Alexander Graham Bell gave demonstrations
to audiences that overflowed into the streets, and he was received by Queen
Victoria for a private showing. Government users, however, were conservative
in adopting the new medium. The chief constable of Exeter, for example, did
not seek to have an office telephone installed until 1901 (Perry, 1977, pp. 69-
96). The advent of the telephone was not enthusiastically received by the Post
Office, which had only recently taken over the telegraph. Its chief civil servant
questioned the practicality of the new instrument. In 1878, the postmaster gen-
eral, Lord John Manners, declared that there were no plans for the Post Office
to use telephones as a supporting part of telegraphy.
Since the telephone was regarded a luxury, its development received no priority.
And indeed, the telephone was expensive. Limited local service cost three times
as much as employing a maid or messenger boy. It could pay for a household's
annual expenses for coal, firewood, and electricity. When in 1902 the London
County Council protested against high telephone rates, The Times editorialized,
"When all is said and done the telephone is not an affair of the millions. It is
a convenience for the well-to-do and a trade appliance for persons who can
very well afford to pay for it" (Perry, 1977, p. 75).
The growth of the telephone industry was further affected by patent disputes.
Both Bell and Edison had received British telephone patents and were soon
involved in litigation (Meyer, 1907). Both companies offered the patent rights
to the Post Office, but negotiations failed. The parties then jointly formed the
United Telephone Company in 1880. In the meantime, the government tried to
include authority over the telephone in a pending telegraph bill. This proved
unsuccessful in Parliament, but a court of law soon held that the original 1869
Telegraph Act encompassed telephones.
The postmaster general then offered to license the United Telephone Com-
pany if it waived its right to appeal the court decision; a royalty payment of 10
percent was required. The Post Office concession did not include the power to
erect poles or use public rights-of-way. For this it was necessary to obtain the
local authority's or private property holder's permission. The concession was
not exclusive, and additional licenses could be granted to other companies as
well as to the Post Office itself. Furthermore, the provision of long distance
telephone communications was severely restricted in order to protect the gov-
ernment's investment in the telegraph system. For the same reasons, public pay
telephones were precluded in most instances. When a local company in Man-
chester wanted to provide public pay telephones at 2 cents per conversation, it
was required to charge 24 cents, the equivalent of the charge for a twenty-word
local telegram.
Similar protectionist restrictions were put on messenger services' use of the
telephone. In 1891, messenger companies were forced to be licensed under the
following conditions: that they not use the telephone, that they pay a minimum
license fee per year, that they affix a 2-cent stamp on every letter delivered by
messenger, and that no messenger take more than six letters at a time from one
center. Moreover, messenger companies were prohibited from using a messen-
ger call box with telephone attachment, an American invention that could sum-
mon a messenger.
United Telephone applied for and received a license for London. For other
areas, it set up regional firms that it provided with patent rights in return for
equity. These subsidiaries then applied for Post Office operating licenses.
In 1882, when a competitor to the United Telephone Company applied for a
license, the application was granted by postmaster general Henry Fawcett, one
of the leading economists of the time. Fawcett favored competition and free
trade between private firms and between the private firms and the government.
He opposed nationalization and abolished restrictions such as size limitations
of exchange areas; he also permitted the company to engage in some long
distance transmission.
In 1884, the Post Office began to install telephone exchanges itself, despite
the opposition of the Treasury, which did not want the government to engage
in competition against private firms. The Post Office also eliminated the pro-
hibition of private interurban trunk lines, and interconnected several local ex-
changes. With a growing national network, the United Telephone Company
began a consolidation, eventually absorbing twelve subsidiaries into the Na-
tional Telephone Company (NTC).
By 1892, telephones were installed in about 400 cities by the NTC. The
company suffered, however, from problems in acquiring way-leaves (rights-of-
way). Finally, in 1892, partly in response to pressure from the newspaper pub-
lishers, the government improved the company's rights-of-way situation by a
quid pro quo that sought to protect the ailing long-distance telegraph service,
which could barely cover costs after Parliament had reduced its charges by half.
In return for the NTC's right to acquire rights-of-way, the Post Office sought
full control of long-distance telephone service. Under pressure and dependent
on the government (the telephone company's long-distance lines were strung
along government railroad tracks), the company sold its trunk lines to the Post
Office in 1896.
However, the agreement required the company to obtain the consent of the
local authority for each right-of-way. This turned out to be an insoluble prob-
lem and made the agreement virtually useless. The slow expansion of the
national trunk system created another bottleneck. The Post Office and the Trea-
sury were apprehensive about incurring financial obligations. Trunks were laid
only if local authorities guaranteed adequate revenues for the construction and
maintenance of the new line. The company was forced in many instances to
issue that guarantee.
Local authorities demanded approval rights for telephone service partly to
obtain financial payments. Many cities refused to give rights-of-way, because
some wished to operate a municipal telephone service themselves. Circuitous
routings became necessary, and there was no access to certain areas. For ex-
ample, in 1885 the NTC was unable to reach the Middle and Inner Temple
areas in London in which most barristers had their offices. The arrangement
also delayed the introduction of so-called metallic circuits which were of sub-
stantially higher quality than single-wire circuits, and of underground cables.
The situation was further complicated by the emergence of the Duke of Marl-
borough as a promoter of a rival venture, the New Telephone Company. Founded
in 1884, it had obtained a Post Office license but was not actually operating.
In 1891, the Duke began to campaign for his company with several letters to
the Times, in which he offered to provide London subscribers with metallic-
circuit telephone service for a flat rate of 12 guineas for an unlimited number
of calls, a rate much lower than the existing one. In 1892, stock for the New
Telephone Company was offered to the public at a fairly high price. The Na-
tional Telephone Company bought one-third to keep leverage over the potential
rival. But when the time came to make the promises real, the Duke published
yet another letter in the Times, this time repudiating the possibility of low-cost,
high-quality telephone service. Referring to himself as having "bleated a good
deal" about the lower-priced telephone (Meyer, 1907, p. 102), he vaguely
referred to future efficiencies that would establish the conditions he had de-
scribed in promoting his venture. The remainder of the New Telephone Com-
pany's shares were subsequently acquired by the National Telephone Company
for $2.2 million although actual property value was estimated at $1 million or
less.
Beyond the blue-sky aspects of these securities transactions, some of their
long-run effects were to unreasonably raise expectations. Public opinion, once
made to believe in cheap flat rate service, saw higher rates as an expression of
private monopoly power. The tide began to shift toward the advocates of na-
tionalization.
The House of Commons appointed a select committee in 1895 to report on
the feasibility of municipal telephone service. The government's own attitude
ran at first from lukewarm to negative. The postmaster general, Arnold Morley,
declared in a parliamentary debate in 1895 that telephony was not a specific
municipal responsibility, as were gas and water. He went on to predict: "gas

and water were necessaries for every inhabitant of the country; telephones were
not and never would be." 1 He further explained patiently to Parliament that
"in a town like London, Glasgow, or Belfast, an effective telephone service
would be practically impossible if the large majority of houses were furnished
with telephones, so great would be the confusion caused by the increased num-
ber of exchanges." But in 1898, another select committee concluded that the
municipalities should provide effective competition to the NTC. The govern-
ment accepted this recommendation, and authorized municipal telephone ser-
vices as well as Post Office exchanges in the 1899 Telegraph Act. The Act
also precluded the NTC from entering areas in which it had not already been
operating.
This policy proved an utter failure. Of the 1300 local authorities that were
eligible to engage in telephony, only six actually installed telephone exchanges,
one of which was sold almost immediately to the NTC. In London the Post
Office made little headway with customers against the NTC.
Its competitive strategy in ruins, the government pursued outright takeover.
In 1901, the company capitulated and agreed that upon the expiration of its
license in 1911, its London facilities would be purchased by the postmaster
general at the cost of reconstruction (i.e., much below market cost, minus
depreciation). The purchase price was fixed by the Railway and Canal Com-
mission and arbitrators on the basis of replacement cost without allowance for
goodwill or future profit. This principle of government acquisition was subse-
quently extended in 1905 to the entire country, also to become effective in
1911. For the few areas for which licenses had been granted to extend beyond
1911, the postmaster general agreed to buy the unexpired license from the
company. The entire system was nationalized for a mere $12.5 million. Even
so, this purchase price proved highly controversial in Parliament. The Post
Office also acquired the meager band of municipal systems. In the end, only
the Hull telephone system was left independent, and has remained so until
today.
Thus, after a long struggle, the unified state monopoly system of post, tele-
graph, and telephone, all under one roof, was complete and would stay so for
almost three quarters of a century.
The PTT System at Its Maturity
Developments similar to those described for German and British telecommuni-

cations took place in most other European countries; details can be found in
the respective country chapters.
In time, the Post, Telephone, and Telegraph administrations that were estab-
lished throughout Europe achieved enormous power, partly because telecom-
munications became ever more important and partly because the PTTs ce-
mented their monopolies through an extensive system of economic and political
Tje Establishment of the PTT System 23
alliances with key constituencies in society, in a system that can be termed the
"postal-industrial complex."
Today PTTs are among the largest employers in their countries. In Germany,
for example, the combined Bundespost had the single largest work force in the
nation, some half a million employees, even before unification. A similar num-
ber are on its retirement rolls. Add to that 1 million of their families and rela-
tives, and one has a substantial voting bloc. Employees in these systems are
usually reasonably well-paid and secure in their positions, typically as civil
servants. To some extent, the lower-skilled but more numerous postal employ-
ees benefit at the expense of their telecommunications colleagues, but they also
provide political backing. In several countries, such as Sweden or Spain, postal
and telecommunications operations have always been separate. More recently,
several countries have divided the functions of telecommunications operations
from the regulatory functions and have often given a greater independence to
the former than under a state administration; they often use the term PTO (Pub-
lic Telecommunications Organization, or Operator) instead of PTT. But they
are still state controlled virtual monopolies. These changes will be discussed
further in Chapter 5.
Closely linked with PTT employees are the workers in the large equipment
supply and electrical industries, who are well aware of the threats to their jobs
due to the migration of electronics industry production to the Far East. The
trade union movement in general is rooted by history and membership in the
manufacturing sector of the economy. The interests of employees of telecom-
munications users are less represented, partly because white collar service in-
dustry unions, although they exist, have played a lesser role in the trade union
movement.
Unions have traditionally strong interests in supporting the principle of PTT
monopoly. First, as employees of huge enterprises, they are fearful for the
potentially lower employment levels and wage rates of a competitive regime.
The example of the United States, where deregulation of the airlines and the
divestiture of AT&T led to a reduction in employment and to attempts at roll-
backs in wage settlements, is certainly a deterrent. Second, as members of the
socialist wing of the body politic, unions often held a political preference for a
nationalization of key industries. There was a strong general feeling across the
political spectrum that a critical part of the infrastructure, particularly one with
such future importance to the information economy, cannot be entrusted to
business owners and managers dedicated to the profit motive and to the protec-
tion of their positions of power in society.
Beyond those factors is the notion that the PTT is an important planning tool
in the development of information technologies, and as such must be under
government control. Additional arguments are that a PTT monopoly is an aid
in social redistribution because it provides subsidized telecommunications ser-
vices to the poor and to inhabitants of outlying areas.
Apart from their own positions of direct influence, much of the PTTs' power
arises from allowing other groups of society to share in the benefits of their
monopoly. One such group is equipment manufacturers, typically very large
24 The Establishment of the PTT System
private companies (Siemens in Germany, CGE and Thomson in France, GEC

and Plessey in the United Kingdom, BTMC in Belgium, Italtel in Italy, Philips
in the Netherlands, NEC and Fujitsu in Japan, and Ericsson in Sweden). In
most European countries the market share of the largest four manufacturers in
total telecommunications equipment is above 90 precent. Such companies are
among the most potent European firms; for a long time they set the tone for
the private sector's telecommunications policy preferences within general in-
dustry associations. In the equipment markets, PTTs fill the role of a monop-
sonist, or primary buyer. The maximum of joint profit for both monopsonist
(who is a monopolist supplier of the final product) and a group of oligopolists
usually lies in some form of cooperative behavior. The PTTs are instrumental
in coordinating the industry, an arrangement that can be advantageous to sup-
pliers, who as a result need not compete vigorously against one another.
A variety of barriers were set to protect this cooperation, including a reluc-
tance to procure foreign equipment, a coordinated development of new tech-
nology, and PTT-organized equipment standards. One consequence of this pro-
tective system was that European prices were said to be 60-100 percent higher
for switching equipment and 40 percent higher for transmission equipment than
prices in North America (OECD, 1983), or even 200 percent higher (see table
on page 330).
The office equipment manufacturers, new computer companies, and data
processorsthe grouping that may be termed the "second electronic indus-
try"were partly outside the postal-industrial complex. In combination with
the service industry, these firms were a potential counterweight. But in time
the PTTs drew them into their orbit, especially by assuming a key role in
domestic industrial policy. This role made the PTT an important financial backer,
valued customer, domestic protector, and international promoter in high-tech-
nology markets. It could channel development contracts to domestic industries
and undertake tests of such technology. It could also coordinate R&D among
manufacturers and provide nontariff protection and export advantages. The PTTs
thus assumed some of the costs of the early part of the learning curve, and in
effect subsidized the development of products that were then offered in the
world market, in a way similar to the role of defense spending in the U.S.
The scope of telecommunications grew to enormous proportions. In the Eu-
ropean Community alone, in 1986 there were 173 million telephone customer
lines served by telecommunications organizations employing 922,000 people
(excluding the more labor intensive postal services), annually generating 64.8
billion ECU (about $77 billion), and investing 22.5 billion ECU (about $27
billion). Average penetration of main lines per 100 population in 1990 was
37.75, ranging from 17.8 in Portugal to 56.0 in Denmark.
The broad coalition of government bureaucracy, private equipment suppliers,
labor unions, intellectuals, good-government advocates, the political left, the
press, the poor, and the rural areas is formidable. The other side includes ser-
vice industry users who are denied certain communications services, and the
"unborn technological generations." Organizations of large users within Eu-

ropean countries, although they exist and are joined in umbrella organizations,
are relatively weak.
For all its strength, the traditional system became subject to a host of forces
of change. The dynamics of this transformation constitute the subject of the
next chapter.
3
Network Tipping: The Rise and Fall of
the Public Network Monopoly
According to John Kenneth Galbraith, the "great advantage of being in the

same world as the United States is that it reveals to other countries the plea-
sures and horrors that will afflict them only a few years hence." 1 Although
such a generalization must be taken with caution, it suggests that some recent
American change in telecommunications may be part of a broad and general
trend and may be relevant to circumstances as they evolve elsewhere. More
recently, several other industrialized countries have begun to adopt similar pol-
icies. Japan, for example, has seriously contemplated splitting its near monop-
olist NTT into local-exchange and long-distance companies and giving them a
greater distance from equipment suppliers. Britain, Germany, and the Nether-
lands have lowered entry barriers. Other countries are considering similar changes.
Much of Latin America is engaged in the privatization of telecommunications.
These developments raise the question of whether change may have explana-
tions more fundamental than the nature of the respective governments in power.
Of course, there are unique aspects to any country, and they will keep national
telecommunications systems to some extent distinct. But the variations should
not obscure central themes that repeat themselves elsewhere.
There has been little attempt at a broader interpretation of the formation and
transformation of networks that can explain the dynamics of change, and this
is the subject of this chapter.2
The discussion is more theoretically oriented, and includes some equations
that readers can skim without loss of the general argument. We begin with a
look at the general concept of networks.
The Concept of Network
Networks are an important concept in society and economy. They abound as

physical facilities, such as those of electric utilities, communications, and
transportation. They also exist as relational systems, such as those of "old
boys," political supporters, and intelligence agents.
The term network is old; in the Bible's King James translation, it is used by
26
Network Tipping: The Rise and Fall of the Public Network Monopoly 27
the Supreme Regulator: "And the Lord spake unto Moses, saying . . . You
shall also make it a grating, a network of brass . . ." (Exodus XXVII, 4). In
the original Hebrew, the word is reshet (net), similarly used today for telecom-
munications and other networks.
The term is used by most academic disciplines, and with a variety of mean-
ings. Chemists apply it to arrangements of molecules (Zacharisen, 1932, pp.
38-42); biologists, to cell structures (Knox, 1830, p. 214); mathematicians, to
topology (Klingman and Mulvey, 1981); electrical engineers, to distribution
systems (for high voltage), or for circuit configurations of components (for
weak voltage) (Kami, 1986, pp. 1-4).
Operations researchers use a network terminology to solve shortest-path
problems, maximum-flow models, and optimal routing (Elmaghraby, 1970, pp.
1-3). Computer scientists apply the term for computer interconnections in
hardware and to implementation algorithms in software.
In the social sciences, political scientists use the concept of networks in
discussing hierarchies, interactions, gatekeepers, and policy communities
(Richardson et al., 1985, pp. 6-8). For sociologists and social anthropologists,
networks are a major way to see the world; a basic point is that the nature of
linkage affects behavior (Barnes, 1954; Bott, 1971; Boissevain, 1979). Soci-
ologists speak of network dyadsinterpersonal linkage between two persons in
which each is indebted to the other, similar in some ways to the exchange
relation of economics.
Among the social science disciplines, economists have probably paid less
attention to networks. There is no body of analysis for the network concept.
Somewhat related is work on market structure by some industrial organization
theorists (Baumol et al., 1982). Closer are public choice theories of group
formation, discussed in the following section. Other writings on networks are
by Noam (1988), Heal (1989), and Economides (1989).
Corresponding to different disciplines' use of the network concept, econo-
mists also approach concrete applications differently. Thus, when it comes to
telecommunications networks and network policy, several ways of thinking can
be distinguished. They are the golden calves worshipped by different profes-
sional denominations.
For technologists a primary organizing concept is economies of scale and
their first cousin, standardization. Economists, on the other hand, worship at
the altar of competition, mostly to the trinity of structure, conduct, and perfor-
mance. The increasing disenchantment with this view is represented more in
academia than in the regulatory environment.
Lawyers in this field often judge policy issues in terms of conflict of interest,
which translates here to potential for cross-subsidies. This perspective is partic-
ularly developed in the United States (hence, the AT&T divestiture).
Finally, many other social scientists as well as most politicians and journal-
ists organize reality in telecommunications policy around the concept of income
distribution (i.e., around the question of who pays and who receives and of
what factors of political power lead to such distribution).
All these concepts are legitimate but have been carried by their proponents
28 Network Tipping: The Rise and Fall of the Public Network Monopoly
beyond their explanatory power. Used single-mindedly, they have degenerated

from tools of analysis to rallying slogans in policy disputes.
Perhaps the greatest common failing of these traditional organizing ways of
looking at telecommunications principles is that they concentrate on "supply-
side" analysis. That is, they look at the subject from the angle of production
and producers: AT&T versus MCI; Intelsat versus Cable & Wireless; value-
added networks versus basic carriers, and so on.
It is not surprising that this approach would be taken. After all, policymakers
deal primarily with carriers, technologists deal with networks, and economists
deal with competitors. Moreover, journalists love a horse-race angle to their
coverage. But this supply-oriented perspective obscures the other side of the
coinwhat could be called a demand-side telecommunications analysis. At its
most basic, one should not think of telecommunications primarily as a service
produced by carriers, but as an interaction of groups and subgroups in society,
facilitated by service suppliers called carriers. The supply structure, if left to
its own devices, is a reflection of the underlying interaction of communication
users with each other, whether in one all-encompassing "user coalition" or in
several user groupings. A ubiquitous public network that interconnects every-
body with anybody, under a single organizational roof, is technically and finan-
cially merely one arrangement out of several.
Thus, one should not view the liberalization of telecommunications as a pol-
icy of primarily permitting the entry of suppliers. Just as important, though less
obvious, it is the easing of exit, by some partners, from a previously existing
"sharing coalition" of users that has become confining.
Integration and centrifugalism are two basic types of forcescall them com-
munalism versus particularism, order versus diversitycommon to many social
processes. Telecommunications are only one instance for the widespread ascen-
dancy, in recent years, of centrifugalism in previously shared arrangements.
Wherever one looks, people break up all kinds of social networks of interaction
and form new ones.
Examples abound. They include public education, mass transit, dispute res-
olution, pension systems, health provision, electrical power and gas distribu-
tion, stock exchanges, and so on. It is beyond the scope of this chapter to
discuss the factors underlying these transformations. But if we focus on tele-
communications, we can look at its key institutional relation, the shared net-
work, and analyze why its unity breaks down.
Theories for the Emergence of Multiple Networks

A number of explanations have been offeredexplicitly or implicitlyfor the
demise of monopoly in telecommunications. There are four major types of the-
ories.
Technological Explanations
The technological perspective comes in two variants:
1. "More powerful technology leads to new transmission options, thereby

to competition, and thus to the breakdown of monopoly."
2. "The merging of telecommunications and computing technologies breaks
down traditional barriers separating different industries and undermines
monopoly power."
These views are typically held by technologists and are influential in an en-
gineering-oriented industry such as telecommunications. But they are not suf-
ficient as explanations; otherwise one would observe a diversity of physical
networks also emerging in, say, France, Australia, or Mexico. After all, the
same transmission and switching technologies are available anywhere on the
globe.3 Yet their impact on network structure has varied and provides no evi-
dence for technological determinism. Technological change provides a precon-
dition for change, but it is not a sufficient one.
Political Explanations
There are three related political explanations, all using the perspective of coun-
tervailing powers:
3. "In the information age, a telecommunications monopoly becomes too
powerful and its scope needs to be limited."
4. "Government regulation proves incapable of controlling a monopoly and
is therefore replaced by policies encouraging a competitive industry struc-
ture."
5. "Large business users successfully fight the monopolistic restrictions."
The problem with these views is that the introduction of a multiplicity of
carriers is only one policy option out of several. An alternative response to
political power or regulatory inefficiency might well be a stricter or more ef-
fective regulation, as would be nationalization or a size reduction along geo-
graphical and/or functional lines while maintaining monopoly. Thus, it is not
clear why competition is the necessary remedy to monopoly power.
Nonsustainability Explanations
Another view is that a monopoly, even if efficient across its multiple products,
cannot protect itself from entry into some lines of business:
6. "The diversification of telecommunications makes it difficult for any one
provider to serve all submarkets without competitive entry."
This view is essentially that of an economic nonsustainability theory ad-
vanced by Baumol et al. (1982), applied to telecommunications (Woroch, 1990).
It can explain the emergence of entrants for new products of a multiproduct
firm, but it does not adequately cover competition in traditional core markets
of a telecommunications monopolist, unless one accepts very restrictive as-
sumptions (Shepherd, 1983).
Market Structure Explanations

There are two variants to the market structure approach, one passive and the
other active:
7. "Monopoly's inefficiency eventually leads to the emergence of competi-
tion."
8. "Competition is a policy chosen to enhance efficiency and technological
development."
These views are held by many economists. Theory 7 is similar to the views
of Milton Friedman on the impermanence of privileged economic arrange-
ments. Theory 8 expects governments not simply to wait for competitors' entry
but to institute proactive and pro-competitive policies (e.g., the United King-
dom and Japan, where competition was introduced from above). These two
views share the premise of inefficiency of monopoly. In other words, a multi-
carrier market structure is believed to be emerging because of some failure of
the traditional system. Yet this assumption is at odds with the reality of net-
work performance in those countries where structural changes in networks are
most rapid. If inefficiency were the causal force for rival entry, Egypt or Mex-
ico (to use two almost random examples) should have introduced competition
long before the United States and Japan, whose networks were among the most
advanced and ubiquitous networks in the world even before embarking on their
liberalizing policies.
It has always exasperated the proponents of the traditional network system
to be told that their problem was inefficiency. This clashed with their observa-
tions of economies of scale, benefits of long-term technological planning, and
effectiveness of end-to-end responsibility. Thus, explanations premised on the
inadequacies of the monopoly system are not persuasive.
A New Approach: The Tipping of Network Coalitions

None of these eight theories for the emergence of multiple networks provides
an adequate explanation, although they all contain truth and in the aggregate
hold explanatory power.
In contrast, a ninth and alternative view is proposed, that of the dynamics of
group formation. The thesis is not based on the failure of the existing system,
but rather on its success. Changes in technology, politics and cost are merely
enabling a more fundamental shift of coalitions.
9. "The breakdown of monopoly is due to the very success of the traditional
system in advancing telephone service and in making it universal and
essential. As the system expands, political dynamics take place that lead
to redistribution and overexpansion. This provides increasing incentives
to exit from a sharing coalition, and to an eventual 'tipping' of the net-
work from a stable single network coalition (the public network) to a
system of separate subcoalitions."
This view of success undermining its own foundations is, from the monop-
oly's perspective, deeply pessimistic, because it implies that the harder
their efforts and the greater their success, the closer is the end to their special
status. As in a Greek tragedy, their preventive actions only assure their
doom.
A Model of Networks4
It may be useful to ask why there is usually only one public telephone network
in a country. It is not for the interconnectedness of all participants, since this
justification would lead one to have only one large bank for all financial trans-
actions. Interaction does not usually require institutional integration. This was
Adam Smith's major insight. To distinguish telecommunications from this ob-
servation by labeling it infrastructure requires us to define that term in a way
that is not vacuous or circuitous, an almost impossible task.
Another explanation is "natural monopoly." Although such a monopoly may
exist for a local exchange area, the examples of the United States, Canada,
Denmark, and Finland prove that a widespread horizontal integration of local
exchange areas is not required. Even if it were, one must ask, why these econ-
omies end miraculously at the national frontier. Has there ever been a national
monopolist that asked a larger neighbor to take over its system in order to reap
the benefit of economies of scale? Moreover, it has not been established that
an integration of local and long-distance service is based on economies of scale.
These services are institutionally separated in several countries.
Perhaps the best way to look at a network is as a cost-sharing arrangement
between several users. Fixed costs are high, marginal costs low, and a new
participant, C, helps incumbents A and B to lower their cost. In this respect a
network is similar to a "public good," such as a swimming pool or national
defense. But although there is basically only one national defense system, there
are many types of arrangements for swimming pools. A user may want to share
the pool with a few dozen families, but not with thousands. A pure public good
admits everyone, a pure private good admits only one. But there is a wide
spectrum between the pure private good and the pure public good (Buchanan,
1965). A telecommunications network is one intermediate example. It is not a
private good yet it does not meet the two main conditions for a public good:
nonrival consumptions and nonexcludability. In fact, nonexcludability has to
be established as a legal requirementthe universal service obligation. What
has been happening in recent years to telecommunications, and what goes by
the more dramatic labels of divestiture and deregulation, is largely a shift in
the degree of its intermediate position, a shift toward the direction of private
good.
We shall now develop, in a stepwise fashion, a model for network evolution
and diversification to explain the breakdown of the network monopoly.
Figure 3.1. Stages in Network Expansion.
The Basic Model5
Let the total cost of a TC network be given by a cost function that depends on
the number of subscribers.6
(1)
Let an individual's utility be given by a function u(P, n), where P is the price
for network usage and n is the number of network members.7 We assume that
the more members there are on the network, the better it will be for an individ-
ual subscriber, other things remaining equal (including network performance
and price),8 that is, we assume that network externalities exist. This means that
(8u/dn > 0), though at a declining rate (82u/8n2 < 0).
We assume that the network membership is priced at average cost (i.e., that
users share costs equally). Then price P is also a function of n: P = P(n).
(This assumption will be dropped later.9)
These relations can be shown schematically in Figure 3.1, where u(ri), the
benefit of participation in a network, is steadily increasing with network size,
though at a declining rate. The price of participating in the network
is declining, at least at first.
Critical Mass
Subscribers will find it attractive to join a good-sized network, because total
costs are shared by many, making average costs low, while the number of
subscribers n adds to utility. This can be seen in Figure 3.1, where the utility
of joining a network rises at first. Conversely, where the network is small, the
average cost is high, and externalities are small. In that range, below a "critical
mass" point HI, a network will not be feasible, unless it is supported by exter-
nal sources. We define critical mass as the smallest number of users such that
a user is as well off as a nonuser u(n) = P(n).I0
To reach n\ requires a subsidy of sorts, either by government or by the
network operator's willingness to accept losses in the early growth phases of
operations. The strategic problem is to identify in advance a situation in which
such a break-even point n\ will be reached within the range n < N, where N
= total population. Such a point may not exist, and subsidies would have to
be permanent to keep the network from imploding. We shall return to the crit-
ical mass issue later in the section entitled "Subsidies for Reaching Critical
Mass."
Private Optimum
Through the cost-sharing phases of network growth, the earlier network users
can lower their cost by adding members. However, at some point average cost
AC increases in the range beyond the point n0 where AC = f(n).
Beyond n0, expansion becomes unattractive for the members of the network
for cost reasons because (for example) they are in more remote locations with
lesser population density, and thus are more costly to serve. But some further
expansion would be accepted by the network members, since newcomers be-
yond the low-cost point would still add to utility. This will be up to the point
2 where the total derivative with respect to the number of users is equal to
zero.
(2)
which holds where the marginal utility of an added subscriber is
(3)
The optimal point lies in the range of increasing AC. Graphically, n2 would lie
where the two derivatives are of equal size, u'(P) = '() Left to themselves,
the existing subscribers of the network would not accept members beyond n2,
the private optimum. From size 0 to size n2. the network is in its cost-sharing
stage.
Social Optimum
From a societal point of view, however, the optimal network size in an equal-
price system may diverge from the private optimum.
Assume social welfare given by the sums of utilities.
(4)
so that its derivative
(5)
(6)
Since u(n) > C(n)ln below the point of intersection n 4 , social optimum 3 is
greater than private optimum n 2 . (It should be noted that the same size will be
chosen by an unconstrained monopolist that sets the price at P = u(n^) to
exhaust consumer surplus.)
What is the implication? Left to itself, the network will cease growth beyond
2. at least as long as costs are equally shared. Existing network subscribers
would not want to admit newcomers beyond 2 - Latecomers beyond that point
add cost, because they raise AC and add fewer externality benefits. [In reality,
of course, networks are "public" and not able to reject members (B. Green-
wald, 1990, communication).] The socially optimal size 3 will not be reached
by itself, but requires some external direction (e.g., governmentally mandated
expansion and/or a differentiated pricing scheme) or some internal politics of
expansion, which will be described later.
Entitlement Point and Universal Service Obligation

To understand the politics of government-directed network expansion, let us
assume that political decisions are made by a mechanism in which majorities
rule. When private optimum size 2 < N/2, there are more people outside than
inside the network, and there are positive net benefits (i.e., H(n 2 )-AC(n 2 ) >
0). A majority consisting of N 2 network outsiders would therefore outvote
the 2 network insiders and require the opening of the network to additional
members. This would be the case up to the point where network size reaches
N/2, at which point the network insiders would have grown to a majority and
would resist further growth. Beyond N/2, then (or where n2 s N/2 and a ma-
jority against expansion exists from the beginning), a politically directed growth
will occur if the coalition of network insiders can be split by aligning the re-
maining outsiders N/2 with some of the insiders who are offered a more favor-
able share of cost (i.e., by price discrimination). It can be shown that this
coalition formation will lead to an overexpansion of the network. The dynamics
of such price discrimination and its impact are discussed in greater detail later
in this chapter, in sections entitled "Political Price Setting and Redistribution,"
"Monopoly and Expansion," and "Network Tipping."
Politically directed growth beyond private optimum n2 can be termed an
"entitlement growth," because it is based on political arguments of rights to
participate in the network where average net benefits are positive (encouraging
attempts of entry) while marginal net average benefits are negative, leading to
attempts at exclusion. In economic terms, the argument is made to expand the
network at least to where C'(ri) = P = u(n), leaving fixed costs to be distrib-

uted unequally, for example, by a Ramsey pricing rule. When the marginal net
benefits are positive, there is no need to resort to the language of entitlements,
since growth is self-sustaining and sought by network insiders. It is only be-
yond that point that entitlements, rights, and universal service rights (i.e., ob-
ligations by the network) become an issue. We can thus define n-i as the "en-
titlement point."
This way of analyzing entitlements serves to clarify the often asked question;
"For which services will universal service be extended?" The analytical an-
swer is to those services that
1. have grown beyond minimum critical mass, and
2. have reached, through self-sustained growth, a private optimum, beyond
which further growth is not internally generated because marginal aver-
age net benefits are zero, but where
3. average net benefits are positive (and therefore encourage demand for
entry), and
4. the number of those excluded is sufficiently large to lead to an opening
by means of internal and external redistributive politics.
Exit from the Network

If 2 < N, with N being the total population, some people would be left out of
the network. But as discussed in the previous section, a government would
require for the network to be open to other users. Yet there may well be a point
where the network is expanded to an extent that, given its increasing marginal
cost, a user is better off by not participating. We define n4 as the "exit point"
(i.e., the largest n such that the indifference exists between dropping off the
network and sharing in the cost of supporting the expanded network).
(7)
This exit point may lie beyond the total population, 4 > N. But this seems
unlikely under an average-pricing scheme, because the last subscribers may
impose a heavy burden on the rest of subscribers, and the subsequent departure
of some subscribers would lead to further reduction in the utility of the remain-
ing members and may induce a secondary exodus. Thus, assuming n^ < N, a
government's aim to establish a truly universal service, without resorting to a
subsidy mechanism or price discrimination, is likely to be infeasible. In other
words, a universal service policy is likely to depend on a redistributive policy.
In the range of size 2 to 4, the network is in its second phase, the redistri-
butive stage.
Political Price Setting and Redistribution

We have so far assumed that universal service is something imposed externally
by government. In this section, however, it will be shown that the internal
dynamics of network members will take the network toward universal service
and toward its own disintegration.
As has been shown, a network will cease to grow on its own after private
optimum n2. But this conclusion was based on a pricing scheme of equal-cost
shares. Yet there is no reason for such equal-cost shares to persist if they are
allocated through a decision mechanism that permits the majority of network
users to impose higher-cost shares on the minority. (This assumes that no ar-
bitrage is possible.) Unequal prices and a departure from cost could be ration-
alized benignly as merely "value-of-service" pricing (i.e., higher prices for
the users who value the telephone greatly).
Suppose for purposes of the model that decisions are made through voting
by all network members." Let us assume at this stage that all users are of
equal size (or that voting takes place according to the number of lines a sub-
scriber uses, which in terms of the model is the same thing) and that early
network users have lower demand elasticity for network use. The determinative
vote is provided by the median voter located at n/2. A majority would not wish
to have its benefits diluted by a number of beneficiaries larger than necessary.
This is the principle of the "minimal winning coalition." Its size would be
n/2 + 1.
A majority will establish itself such that it will benefit maximally from the
minority. The minority that can be maximally burdened are the users with less
elastic demand for telephone service, which are the early subscribers. But there
is a limit to the burden, given by utility curve u(ri). If price gets pushed above
u(n), subscribers will drop off. Hence, the majority (2/2 + 1) will burden the
minority (n2/2 1) with a price up to positive utility, and they will bear the rest
of the cost. The minority's price P\ will be such that
(8)
12
The majority's price will then be
(9)
This, then, is the redistributory outcome, assuming no discrimination within

majority and minority.
Monopoly and Expansion

Such redistribution, however, is not a stable equilibrium. Before network size
2 was reached (once the critical mass threshold was crossed) by voluntary
association. Additional members were not admitted, because they lowered util-
ity for the incumbents. But with internal redistribution, several things happen.
There are now incentives for the minority network members to exit the network
and form a new one in which they would not bear the redistributory burden.
This is economically feasible if the minority is larger than critical mass, n^ll
> n\. Even when this is not the case, the minority can band together with
those beyond network size n-i who desire telephone service but were previously
excluded.
This exit would deprive the majority of the source of its subsidy and is
therefore held undesirable. The only way for the majority to prevent this "cream
skimming" or "cherry picking" is to prohibit the establishment of another
network, both by those wanting to leave the original network and by those not
admitted to it in the first place because they are beyond 2 - Thus, a monopoly
system and the prevention of arbitrage become essential to the stability of such
a network system.
At the same time, the model predicts that the network will expand beyond
2. For the majority, there is added utility from added network members, while
most of its cost is borne by the minority, who are also willing to bear a greater
burden as the network expands due to its greater utility. The majority will
therefore seek expansion beyond 2- The cost to the majority is only that the
subsidy by the minority must be shared with more network participants. There-
fore, the majority would admit new members up to the point ns where marginal
utility to its members is equal to the marginal price increase due to the diluted
subsidy.
But this is not the end of the story. With expansion to 5. the majority is
now 5/2 rather than n2/2 (i.e., larger than before), and it can tax a larger
minority (s/2) than before. Hence, the expansion process would take place
again, leading to a point 's > ns. This process would continue until an equi-
librium is reached.
n*5 is the point up to which the network will grow under the internal dynam-
ics described earlier. The greater the marginal utility from added network mem-
bers, and the smaller the marginal costs, and the greater the fixed costs are, the
larger n*5 will be.13
Network Tipping
As this process of expansion takes place, the minority is growing too. The
likelihood that its size will increase beyond the point of critical mass n\ is
increased, and the utility of its members, given the burden of subsidy, may
well be below that of membership in a smaller but nonsubsidizing alternative
network. We have so far assumed that there is only one network and that a
user's choice is whether to join or not. Suppose there are no legal barriers to
the formation of a new network. In that case, a user's choice menu is to stay,
to drop off altogether, or to join a new network association. Assume that the
new network has the same cost characteristics as the traditional network. (In
fact, it may well have a lower cost function for each given size if there has
been accumulated monopolistic inefficiency in the existing network and rent-
seeking behavior by various associated groups.) Then minority coalition mem-
bers would find themselves better off in a new network B, and they would
consider such a network, abandoning the old one. The only problem is that of
transition discontinuity. A new network, in its early phases, would be a money-
losing proposition up to its critical mass point n ' i . 1 5
The point where exit becomes possible, given the redistributory burden that
keeps utility just balanced with price, is the point n beyond where
(10)
which is at the critical mass point. Hence, the minority will strive to exit the re-
distributory network once its size is more than twice the size of critical mass.16
The majority may attempt to alleviate these pressures to exit by reducing the
redistributory burden and thus keeping the minority from dropping out. But
that means the network size 7 would not be optimal to the majority anymore,
and members would have to be forced out. And this, in turn, would reduce its
majority, so that it would have to drop the subsidizing burden from at least
some minority members as the n/2 point separating the majority from the mi-
nority shifts to the left.
This means either higher burdens on the shrinking minority frustrating the
purpose of bribing it into stayingor still less benefits for the majority if it
wants to keep the network from fragmenting. Such a disequilibrium process
will continue up to the point where network size n = 2 n\ (i.e., where the
minority may be too small to create a self-supporting new network). One might
call this the effect of potential exit by the minority, and it results in a lessened
redistribution of newcomers to the network in order to keep the first entrants
inside.
It might be argued that in a new network, internal redistribution based on
coalitions might exist. But this is not likely. Once the possibility of exit is
established, each burdened subgroup can simply join another network. Thus,
internal redistribution will happen only if a network is unique, and thus if a
burdened user will not readily switch into another network.
Unequal User Size

We have assumed so far that network voters are of equal size. In reality, how-
ever, some users are much larger in terms of lines n than others. The minority's
position would be further weakened if voting were governed by a principle of
"one subscriber, one vote" rather than the "one line, one vote" previously
assumed.
Suppose users are ordered according to size on Figure 3.1; in other words,
the largest users are those that joined the network first. This is not unrealistic,
since users with great needs for telecommunications are likely to have been the
first to acquire a telephone, and early subscribers had the longest time to ex-
pand usage. Let us further represent the distribution of lines n for a user v by
(H)
where A > 0, a > 1.
The median voter (or median account) is v/2 and its preferences govern. But
the network size provided by the users arrayed to the left of such median user
is larger than those to the right.
In Figure 3.1, , the median account, is to the right of n/2. In other words,
the median voter whose preferences govern is at a network size greater than
the median point of the network size. The more the distribution of lines is
skewed (the larger the coefficients A and a) the further to the right is ,. And
the more skewed the distribution, the more likely it is that the voting minority
will reach by itself a size beyond the critical mass point.
Interconnection
The process of unraveling of the existing network would commence even ear-
lier if a new network has the right to interconnect into the previous one, be-
cause in that case it would enjoy the externality benefits of a larger reach nA
+ nB, while not being subject to redistributory burden. This is why intercon-
nectivity is a critical issue for the establishment of alternative networks, as the
historical U.S. examples demonstrate, from the Kingsbury Commitment in 1913
to MCI Execunet in 1976 and today's ONA and collocation proceedings.
Since the benefits of network reach remain, a subscriber's exit decision is
cost-driven, and takes place if C(nB)lng < c(n)ln. With redistribution under the
primary network, but not in the new network B, the test would be, for a sub-
scriber /', if c(nB)/ns < PI (n).
Would there exist, for any subnetwork, internal redistribution based on coa-
litions? Once the possibility of exit is established, each burdened subgroup
could join another network. Thus, internal redistribution will happen only if a
network is unique to its users.
Network interconnection means that the network still centers around a soci-
ety-wide concept of interconnected users. But it consists now of multiple sub-
networks that are linked to each other. Each of these subnetworks has its own
cost-sharing arrangements, with some mutual interconnection charges. Inter-
connection facilitates the emergence of new networks and lowers entry barriers.
But given entry, it may reduce competition by establishing cooperative linkages
instead of end-to-end rivalry (Mueller, 1988). Interconnection is a useful con-
cept, because it responds to the frequent claim that a single network is neces-
sary for universal reach. This assertion is clearly incorrect. Interaction does not
usually require institutional integrations. Otherwise, we would have only one
large bank for all financial transactions, as mentioned. But as the next section
will show, it may also lead to market failure in the establishment of the original
network.
Subsidies for Reaching Critical Mass
We have mentioned that waiting for demand to materialize prior to the intro-
duction of a network or network service may not be the optimal private or
public network policy. Demand is a function of price and benefits, both of
which are in turn functions of the size of the network. Hence, early develop-
ment of a network may require internal or external support to reach critical
mass.
This suggests the need, in some circumstances, to subsidize the early stages
of the networkup to the critical mass point n\when the user externalities
are still low but cost shares are high. These subsidies could come either from
the network provider or its membership as a start-up investment, or from an
external source (e.g., a government) as an investment in "infrastructure," a
concept centered around externalities. The question is how the internal support
is affected by the emergence of a system of multiple networks.
The private start-up investment in a new form of network is predicated on
an expectation of eventual break-even and subsequent positive net benefits to
members. But if one can expect the establishment of additional networks, which
would keep network size close to n\, there would be only small (or no) net
benefits realized by the initial entrants to offset their earlier investment. This
would be further aggravated by interconnection rights, because a new network
could make immediate use of the positive network externalities of the member-
ship of the existing network that were achieved by the latter's investment. Hence,
it is less likely that the initial risk would be undertaken if a loss were borne
entirely by the initial network participants while the benefits would be shared
with other entrants who would be able to interconnect and thus immediately
gain the externality benefits of the existing network users, but without con-
tributing to their cost sharing. The implication is that in an environment of
multiple networks that can interconnect, less start-up investment would be un-
dertaken. It pays to be second. A situation of market failure exists.
How could one offset this tendency if it is deemed undesirable? Patents are
one solution. If a contemplated new network arrangement is technologically
innovative, it might obtain a patent protection for some period. Where a service
is innovative but not patentable, one might create a "regulatory patent" for a
limited period of protection, or the initial approval (where necessary) might be
accelerated. Similarly, interconnection rights might be deferred for a period, or
joint introductions might be planned that eliminate the first entrant penalty. But
these measures would also reduce the usefulness of alternative networks and
hence could lead to the dynamics of political expansion, redistribution, and
break-up described earlier. Another approach is to require access charges that
cover the externality benefits for interconnection. But this requires market power,
since otherwise no price above cost could be sustained. And if that market
power exists, it could be exercised in a restrictive fashion.
It is quite possible that none of these measures would be as effective in
generating the investment support as a monopoly network that can reap all
future benefits. This would mean that the private and social benefits of net-
works in the range between n\ and 4 would not be realized. In such a situa-
tion, there may be a role for direct outside support, such as by a government
subsidy. At first, this may seem paradoxical. One would expect a competitive
system of multiple networks to be less in need of government involvement than
a monopoly. But there is some economic logic to this. Just as the subsidies to
individual network users that were previously internally generated by other net-
work users would have to be raised externally (through the normal mechanism
of taxation and allocation) if at least some users are still to be supported, so
might subsidies to the start-up of a network as a whole have to be provided

externally, also through taxation and allocation, where network externalities as
well as start-up costs are high enough to make the establishment of a network
desirable.
Social Welfare and Multiple Networks
If network associations can control their memberships, stratification is inevita-

ble. They will seek those members who will provide them with the greatest
externality benefitsthose with many actual or potential contacts. Furthermore,
they will want to admit low-cost, high volume, low risk customers as club
members. Thus, different affinity-group networks and different average costs
will emerge.
But what about social welfare in such a differentiated system? The traditional
fear is that the loss of some cost sharing and externalities brought by a second
network would reduce social welfare. But the news is not necessarily bad.
Where mutual interconnection is assured, one can keep the externalities benefits
(and even increase them) while moving down the cost curve toward a lower
AC. Furthermore, the cost curves themselves are likely to come down with the
ensuing competition.
The welfare implications of the formation of collective consumption and pro-
duction arrangements are analyzed by theorists of clubs (Schelling, 1969; Tul-
lock, 1971; Rothenberg, 1976; Tiebout, 1956; McGuire, 1972).
Optimal group size will vary according to the dimension to be optimized: It
depends on the ratio of marginal utilities for different dimensions, set equal to
the ratio of transformation in production, and is in turn related to size (Buch-
anan, 1965, pp. 4-5). But this does not imply that one should keep networks
nonubiquitous and unequal. Financial transfers can be used. However, it is
generally not Pareto-efficient to attempt income transfer by integrating diverse
groups and imposing varying cost. It is more efficient to allow subgroups to
form their own associations and then to redistribute by imposing charges on
some groups and distribute to others. The set of possible utility distributions
among separate groups dominates (weakly) the set of such distributions among
integrated groups (McGuire, 1972, p. 124). User group separation with direct
transfer is more efficient than the indirect method of enforced togetherness with
different cost shares. In other words, differentiated networks plus taxation or
another system of revenue shifting such as access and interconnection charges
can be more efficient than monopoly and internal redistribution.
Conclusion
The theoretically based analysis of the model means that a network coalition,
left to itself under majority-rule principles, expands beyond a size that would
hold together. Such an arrangement is therefore inherently unstable. It can stay
together only as long as arbitrage is prevented, as long as the minority cannot

exercise political power in other ways, and as long as it has no choice but to
stay within the burdensome network arrangement.
But beyond that point, the over expansion policy creates incentives to form
alternative networks. And the more successful network policy is in terms of
achieving universal service and "affordable rates," the greater the pressures
for fracture of the network.
This chapter has shown that networks become unstable with expansion be-
yond their optimal size to participants, and because a democratic decision-mak-
ing mechanism establishes expansion through a redistributive mechanism. Sta-
bility can be maintained only through coercion, which becomes increasingly
difficult to maintain.
The analysis also shows that a subsidy is necessary in the early stages of a
network to reach critical mass and that such a subsidy is less likely once the
monopoly network is supplemented by other networks with the right to inter-
connect. In such a situation, market failure may well exist that would reduce
start-up investment in innovation. This suggests that in a competitive network
environment where interconnection exists, alternative subsidy mechanisms must
be established if new types of networks are to reach critical mass.
Thus, the very success of network expansion bears the seed of its own de-
mise. This is what might be called the "tragedy of the common network" [in
the Greek drama sense of unavoidable doom, and borrowing from the title of
Hardin's classic article "The Tragedy of the Commons" (1968) on the deple-
tion of environmental resources]. In the words of Alfred North Whitehead:
"The essence of dramatic tragedy is not unhappiness. It resides in the solem-
nity of the remorseless working of things." In the case of telecommunications
the tragedy is that the breakdown of the common network is caused not by the
failure of the system but by its very successthe spread of service across
society and the transformation of a convenience into a necessity.
4
Forces of Centrifugalism
The model of Chapter 3 identified an evolution of network development with

three stages:
1. The cost-sharing network. Expansion of the network is based on the logic

of spreading fixed costs across a large number of participants, beyond a
take-off point ("critical mass"), and of increasing the value of network
connectivity.
2. The redistributory network. At this stage, the network grows beyond the
size that is optimal to its original members, through a politically directed
expansion caused by the formation of internal coalitions that lead to trans-
fers from some users to others.
3. The pluralistic network. Beyond a certain point, the cohesion of the uni-
tary network breaks apart because the dynamics of expansion and redis-
tribution lead to a divergence in the interests of its participants that can
no longer be reconciled within one network. The results are exit, forma-
tion of new networks, and the emergence of a federation of subnetworks.
The network has progressed to its "tipping point," where its cohesion
breaks up and a multinetwork system emerges.
These trends have a logical progression. At first the network expands be-
cause it makes economic and technical sense. Later it expands because it makes
political sense. But as the network provider succeeds in providing full service
to every household, it also undermines the foundation of its exclusivity.
Most countries in the world are still in the phases of the cost-sharing or
redistributory networks. In Europe, first stage countries are Portugal, Turkey,
Greece, and the nations of Eastern Europe. All others have reached substantial
universal penetration of telecommunications for a number of years, and it is
here that the transformation of the network system has begun progressing toward
its third stage, which the United States entered first, followed by Japan. Since
there is a strong correlation of economic growth and telephone penetration
(roughly an additional telephone per $50,000 GNP), countries with high eco-
nomic growth are likely to progress rapidly through the first two stages. Even-
tually, through economic growth and the instabilities transmitted from the more
advanced nations, many other countries will move into a pluralistic network
environment and will be faced with the policy issues inherent in such a system.
43
44 Forces of Centrifugalism
We have seen how the centralized and hierarchical system of the traditional
network, despite its public popularity, has been subject to forces of centrifu-
galism that have undermined its stability. The present chapter looks at the con-
tributing factors that make this breakup more likely. Technology is one such
factor, though one should not exaggerate its contributions. It is not microwave
and satellite transmission that has made long-distance telephone competition
possible. Several other factors have contributed to the disintegration of the cen-
tralized monopoly.
The Service Economy and Increases in User Size
The driving force for the restructuring of telecommunications has been the phe-
nomenal growth in user demand for telecommunications, a force that in turn is
based on the shift toward a service-based economy. The largest users of tele-
communications are major providers of services; they include corporate head-
quarters, banks, insurance firms, airlines, health delivery organizations, engi-
neering and consulting firms, law offices, and media organizations. The growth
of these service industries in highly developed countries has been due in part
to diminished competitiveness in traditional mass production, where newly in-
dustrialized countries have excelled. It was also partly due to their large pool
of educated people skilled in handling information. These advantages were
reinforced by productivity increases in information transactions through com-
puters and advanced office equipment. Information-based services, including
headquarters activities, therefore emerged as a major comparative advantage of
developed countries. Manufacturing and retailing, at the same time, became
far-flung and decentralized.
Within the service sector, the number of people performing functions related
to information and its creation, processing, and manipulation has grown partic-
ularly large. According to one study (Beniger, 1986), 37 percent of the work
force in 1980 was engaged in information jobs, 22 percent in industrial jobs, 2
percent in agriculture, and 29 percent in other services. Another study asserts
that 54 percent of the American work force and 63 percent of all working time
are now devoted to information work (Strassman, 1985, p. 56). An earlier
study found a similar trend (Porat, 1978).
As electronic information transmission (i.e., telecommunications) became in-
creasingly important to the new services sector, it turned into a major expense
item. For instance, for Citicorp, America's largest bank holding firm, telecom-
munications are the third largest cost item, after salaries and real estate. Con-
sequently, the purchase of communications capability at advantageous prices is
much more important than in the past. Price, control, security, and reliability
have become variables requiring organized attention. This, in turn, led to the
emergence of the new breed of private telecommunications managers whose
function is to reduce costs for their firms, establishing for the first time sophis-
ticated telecommunications expertise outside the postal-industrial coalition. These
managers aggressively seek low-cost transmission and customized equipment
Forces of Centrifugalism 45
systems in the form of private networks of power and scope far beyond those
of the past (Schiller, 1982). These private networks, which start out as systems
of dedicated lines leased from the monopoly network carriers, can become large
and far-flung. Some require hundreds of skilled technicians and managers to
operate and administer, and they are carving out slices from the public net-
work. It does not take a large number of private networks to have an impact.
In the United States, for example, the largest 3 percent of users typically ac-
count for 50 percent of all telephone revenues. These activities are spearheaded
by private firms but are not exclusive to them: Nonprofit institutions (e.g.,
hospitals and universities) and public organizations (e.g., state and local gov-
ernments) are actively pursuing similar cost-reduction strategies. The U.S. gov-
ernment, for example, awarded in 1989 a $25 billion contract for a "private"
network to serve its civilian activities (FTS-2000).1
The Saturation of Basic Service
The model of Chapter 3 argued that the expansion of a single network can
eventually lead to its fragmentation. What has been the path of such expansion?
For a long time the primary policy goal of most industrialized countries was to
establish a network that would reach every household; this also benefited the
supplying industry. Even in highly developed countries, the achievement of
substantial network penetration is a very recent phenomenon. In West Ger-
many, penetration in 1960 was 12 percent. In 1980, it was 75 percent (Schulte,
1982); and in 1990 it was getting close to 100 percent (Pfeiffer and Wieland,
1990). In France, it was 6 percent in 1967 (Guerard and Lafarge, 1979), 54
percent in 1983 (AT&T, 1983), and 97 percent in 1990 (Steckel and Fossier,
1990). Similar trends existed in almost all European countries. Hence, the im-
position of the costs of the last subscribers is a fairly new development, and
responses to it are only now beginning to work themselves out. In the United
States, universal service was substantially achieved some twenty-five years ear-
lier.
An Activist Role by the Equipment Industry
Once universal penetration was reached, the industry had to reorient itself,
because its activity level would otherwise have dropped dramatically.
Having been successful in spreading telephony, the supplying industry too
became a victim of its own success. It was left with several strategies:
1. Upgrade. After achieving universal penetration, the equipment industry
advocated an upgrade of the network. This means an accelerated supply
push rather than demand pull. It also means moving into videotex, inte-
grated services digital networks (ISDN), integrated broadband networks,
and cable television as ways to provide the industry with procurement
contracts.
2. Export. Increased attention to international activities can substitute for the

shrinking basic domestic market. However, many of the more interesting
markets in industrial and industrializing countries are protected by their
governments, which use the network as a way to promote a domestic
electronic industry. The results are trade frictions around the world and
eventually partial opening of national markets in order to achieve rec-
iprocity. This reduces the traditional territorial compartmentalization of
the industry and loosens the close relation between equipment industry
and network monopoly. One of the positive consequences of this is that
prices fall even beyond the gains of technology.
3. Target users as equipment buyers. Perhaps most important in the long
run, manufacturers turned to large users as a market for equipment. Whereas
in 1975 virtually 100 percent of United States capital investments in tele-
communications equipment were made by the carriers, that figure had
dropped to 67 percent by 1986. About $15 billion was invested by non-
carriers, mostly large users (Crandall, 1988). Such equipment includes
PBXs, multiplexers, concentrators, network management equipment, sat-
ellite and microwave facilities, etc.
Users increasingly assumed control over the network segments closest to
themfirst, over equipment on their premises; second, over the wiring seg-
ments in office and residential buildings. It was the natural next step that they
began to share in a full array of telecommunications services within their build-
ing through "shared-tenant services," which shifted some of the switching from
the public exchange to a common PBX and moved transmission from the public
network to a private one. Next, users moved to switching and to local area
networks (LANs) for high-volume links serving the data flows within an orga-
nization.
The implication is that the equipment industry, in the past a protector of the
traditional monopoly, has increasingly become part of the process of creating
alternative networks.
Reductions in Equipment Costs and Increases in Productivity
Another factor leading to new networks is that the underlying economics of

transport and switching have shifted the cost curve for telecommunications con-
siderably downward. A unit of communications has become much cheaper to
transmit and switch. The cost of a network drops as electronic and photonic
equipment becomes cheaper, more powerful, and less expensive to operate. In
the United States, the price for switches paid by one telephone company, for
example, came down in price per line from $230 in 1983 to $144 in 1988.2
Manpower requirements declined considerably,3 and productivity increased.
Similarly, the price per meter of fiber fell from $7 in 1977 to 23 cents in
1988, and its transmission capacity increased enormously. LEDs dropped from
$2000 a few years before to less than $30 in 1990. According to some predic-
tions, in three to five years fiber will be cheaper to install than copper (Elton,
1991; Egan, 1990).
In terms of the model, as the cost curve drops, the critical mass point n\
shifts to the left to a smaller minimal size, and it becomes easier to start an
alternative network. The growth of large users also means that it takes fewer
of them to reach any given network size n. This reduces transaction costs of
organizing and coordinating a new network club and makes it possible for a
smaller number of users to enjoy the economies of scale.
Diseconomies of Scale and Scope
Economic and technological development lead to increased specialization and

to a divergence rather than convergence of options. In telecommunications, the
rapid technological development has spawned a large number of applications.
For examples see the chapter on value-added services. It has become difficult
for the traditional network to keep up with all these options and customized
needs. Thus, the growth of technological and operational alternatives undercuts
the economies of scale and scope once offered by the centralized network. This
will be discussed in detail in the chapter on ISDN.
The Emergence of the "Second" Electronic Industry
Throughout the 1960s and 1970s, many traditional telecommunications firms

gradually lost their technological preeminence. Insulated from competition and
secure in their profits, they had not been highly successful technologically rel-
ative to their resources. Most missed out on the development of computers or
tried unsuccessfully to enter that industry. Their lack of success came despite
major national efforts and subsidies for the development of computer and semi-
conductor components. (This will be discussed in the chapter on industrial pol-
icy.)
Over time, an independent computer and component industry evolved in most
developed countries, forming a "second" electronics sector. These firms, which
were based on direct relations with users without the mediation of the PTTs,
provided an element of a new coalition that began challenging the traditional
postal-industrial complex. They were allied with the large service industry users
in a "services-information industry coalition." In the United States, this grouping
includes, for example, American Express, IBM, Time-Warner, TWA, Silicon
Valley firms, and Citicorp. Defending the traditional system was primarily AT&T,
but this was not enough to stem the tide. Hence, the victory of the services-
information coalition over the traditional forces was inevitable in the United
States. (For a discussion of the AT&T divestiture, see Cole, 1990.)
In Britain, the new coalition was slower to gather than in the United States,
and the defense of the traditional industrial sector was more tenacious and ideo-
logical. But the balance of power swung in the 1970s.
The British electronics industry was not successful internationally, particu-

larly when one subtracts Britain's former colonies as a market. On the other
hand, British service industries such as banking, insurance, trading, publishing,
and media were doing well. London, along with New York and Tokyo, was
the major center for international services, and it was the preferred European
headquarters of non-European firms. Although the Thatcher government argued
for the deregulation of telecommunications largely on the grounds that it would
make British high technology more competitive, the most important effect of
its policy was to help make London the center for European business transac-
tions. This is a role with which Britain, given its traditions, was comfortable.
A similar story can be told for the Netherlands.
In Japan, where the "first" telecommunications industry transformed itself
better than anywhere else into the "second" electronics industry, the changes
were smoothest, since the private industry did not stand to lose while the ser-
vice sector was gaining strength. Reform was accomplished as a continuation
of industrial policy. The national monopoly carrier, NTT (Nippon Telephone
and Telegraph), was privatized, and competition was introduced under the
prodding of the Ministry of International Trade and Industry (MITI) without
the bitter public conflicts of the United States or Western Europe (Ito, forth-
coming; Nambu, forthcoming; Nagai, forthcoming; Komatsuzaki, 1986).
In several continental European countries, such as France, the service sector
is relatively weak in comparison to manufacturing. French banks, most of which
were nationalized, do not play the same role in international business as Lon-
don banks. On the other hand, the industrial sector has been close to the heart
of the French left. The weakness of the service sector can be explained in part
by the traditional socialist emphasis on the production of goods, with its pro-
letarian connotations, in contrast with the more middle-class rooted white-collar
service activities such as finance and international trade. It also reflects the
traditional electoral base of socialist parties. The emphasis on high-technology
manufacturing also fits neatly into France's traditional concern with national
autonomy, which appeals to the political right.
Increasing Cost of Incremental Subscribers
Over the years, low-cost subscribers have been added to the network earlier
than high-cost subscribers. As the network reaches universality, connecting the
last members increases cost. In the Bell system, the average capital investment
cost per new telephone grew steadily (in 1982-1983 dollars) (Telecom Fact-
book, 1986):
1945: $1928
1955: $2050
1965: $2580
1975: $3960
1985: $46244
Forces of Centrifugaiism 49
Upward Drift of Cost Curve of the Old Network

Costs and efficiencies of networks are a question not simply of engineering but
also of market structure. The existing network, operating as an exclusive ar-
rangement, tends to drift upward in terms of cost for a given technology vin-
tage. This can be exacerbated by regulatory arrangements that lead to wrong
incentives, such as to overcapitalization (the Averch-Johnson effect). Some in-
dicators for this are the cost reductions achieved by U.S. companies when
competitive pressures started to make themselves felt.
For example, AT&T's equipment installation and maintenance cost, at the
time of divestiture, was estimated at $61 per hour as opposed to $33 for IBM
and $28 for MCI.
The implications are that a new network, unencumbered by the accumulated
high-cost attributes of the old one, could operate on a lower cost curve even in
the absence of technical progress.
Pluralism in Network Usage
Another related factor contributing to the increase in specialized telecommuni-

cations networks is the growing number of groups in society that interlink via
telecommunications. As the communications needs of groups have become spe-
cialized, private user clusters have emerged. Early examples are travel agents
and airlines, automobile parts suppliers, and financial institutions, all of which
have established group networks that combine some economies of scale with
customization. Thus, pluralism of association has led to group communica-
tions, which we can locate conceptually somewhere between private and public
network activities.
Again, these trends are discernible in all Western industrial countries. They
go back to the changing nature of the production process itself. Michael Piore
and Charles Sabel (1984) coined the term industrial divide for points where the
path of technological development itself is at issue. In such periods the very
nature of technology is determined by the social and institutional realities which
shape economic institutions far into the future. Accordingly, we may be at a
second industrial divide, the first being the emergence of mass industrial pro-
duction. Today we are witnessing difficulties stemming from the transformation
of the economy from an industrial base to a service base, and to smaller-scale
production units. Adam Smith illustrated the importance of economies of scale
in production in his famous example of the needle factory. Size made it pos-
sible to deploy specialized tasks and skills and to gain productivity increases.
An advanced economy based on mass production was the synonym for capital-
ist-era production for both Smith and Marx, and an even more advanced econ-
omy, or so it seemed, would be based on still more mass production, This
policy became the "corporatist" interpretation of political economy, justifying
policies that care for and protect those economic entities that are already large
and powerful, and allying the public interest with that of its largest economic
organizations. The PTT system is a classic example of this approach.
But the mass production view of economic advancement needs revision. Sig-
nificant activity takes place in small and medium-sized firms, and it is difficult
to reconcile their role with the mass production theory.
One way to do so is to take note of the clustering of small firms. The ability
to provide specialized services through clusters of firms creates economies of
scale too. One need not think of the economic process as having to take place
under a single roof or ownership structure. The entrepreneurial function is to
bring together factors of production, not necessarily to own them. Thus, in
terms of neoclassical economic theory, whether the services are performed by
employees or outside contractors, by owned or leased capital is immaterial,
though Marxist economists will not agree.
Hence, the logic of continuing specialization may lead to clustering of firms.
Instead of large-scale vertical integration, a functional specialization may lead
to a disintegration of the traditional functions of a firm. Within those special-
ties, supplier firms may have horizontal economies of scale even though their
size may be moderate because of their high specialization. In the aggregate,
these firms form a production and service complex that can be large, but it is
not under one corporate control. The easy availability of telecommunications
and information technology plays a key role in this development, because it
makes decentralization possible. Telecommunications networks provide the glue
that keeps such a production system together. At the same time, this trend leads
to specialized network functionalities and equipment options, which in turn
accelerate the evolution of telecommunications into a pluralist system.
Provider Specialization
As the information flow requirements of large users become larger, they also
become increasingly specialized. Equipment offered by numerous vendors per-
mits many configurations to accommodate the requirements and procedures of
organizations. It is no longer as necessary to forgo benefits of specialization in
order to benefit from cost sharing.
By their very nature, the traditional networks provide standardized and na-
tionwide solutions, carefully planned and methodically executed. In the old
days, sharing a standardized solution was more acceptable to network mem-
bers, because the consequential loss of choice was limited and outweighed by
the benefits of the economies of scale gained. As the significance of telecom-
munications grew, the costs of nonoptimal standardized solutions began to out-
weigh the benefits of economies of scale, providing the incentive for nonpublic
solutions. Furthermore, some users aggressively employed a differentiation of
telecommunications services as a business strategy to provide an advantage in
their customers' eyes, and affirmatively sought a customized rather than a gen-
eral communications solution. Although these considerations are most impor-
tant for large firms, the differentiation of communications needs for small busi-
ness and residential users has been moving rapidly as well.
The desirability of opting out of the traditional sharing arrangement depends,
among other things, on traffic densities. Where these are high, private networks
of different types emerge, depending on the degree of traffic specialization. For
high-density general-purpose use, private networks emerge, but for low-density
general purpose use the shared public network may last longest. For interme-
diate density, "virtual private networks" provide an intermediate option. For
specialized services, the network types to emerge, depending on traffic densi-
ties, are strategic private networks, closed-use group networks, and value-added
networks. These classifications5 can be charted as follows:
Network Specialization
Traffic
Density General Specialized
Dense Private networks Strategic private
networks
Intermediate Virtual private Closed-use group
networks networks
Low-density Public networks Value-added net-
works
As traffic densities and usage specialization grow, the core of the traditional
public network (the lower left corner) shrinks.
Telephony has gone far beyond simple switched voice connections. Many
"value-added services" have been developed and introduced, including voice
mail, videotex and audiotex, and electronic message interchanges. Concep-
tually, most advanced telecommunications services can be analyzed as four
layers superimposed on each other: basic transmission, data packet transmis-
sion, generic services, and applications packages. Although in many instances
several of these layers can be integrated within the same carrier, they need not
be. Thus, when a bank customer uses an automatic teller machine, several
functionally different service providers may be involved on the same layer, and
several other firms may provide communications on different geographical seg-
ments. The underlying banking transaction, in turn, may trigger interbank elec-
tronic transfer networks of similar complexity. Such structure calls for special-
ized service providers and packagers. This will be described in greater detail
in the chapter on value-added services.
Technology of Alternative Transmission Paths
With multiple transmission technologies, total control has become increasing-

ly difficult for one organization. In large countries, PTT exclusivity in long-
distance communications is already sustained more by politics than by eco-
nomic or technological advantages. For local distributionthe segment

exhibiting the greatest characteristics of "natural" monopolyseveral different
transmission systems have emerged, including coaxial cable television net-
works, stationary cellular radio, microwave multipoint distribution, metropoli-
tan areas, fiber-nets, and infrared transmission (Noam, 1986). For the PTT to
control all the transmission paths means to assert control in new areas, some
of which may be already occupied by other actors, such as cable network op-
erators or satellite organizations. Alternative transmission routes have become
economically feasible, opening new channels of communications and upsetting
traditional patterns, much as advances in navigation and ship building resulted
in Columbus and Vasco Da Gama upsetting traditional shipping routes.
Distance Insensitivity of Transmission Cost
The attractiveness of global transactions has increased as cost has become less
sensitive to distance (Pool, 1990). For satellite transmission, in particular, the
marginal cost with respect to distance is virtually nil. Fiber-optic links have
also lowered distance-sensitive costs. As a result, communication flows can be
routed indirectly to circumvent regulatory barriers and restrictive prices. This
will be discussed in Chapter 32.
If a country's PTT exercises restrictive policies, its firms will be disadvan-
taged internationally, and foreign firms may choose not to domicile themselves
in the country. Similarly, companies in countries with those restricted policies
may pressure for change when they learn about options available elsewhere.
The challenge to the monopoly is that once it is breached by one country,
the change may become irreparable. If the United Kingdom, with its liberali-
zation policy, or one of the smaller European countries finds it profitable to
change its policies and become a communications "haven," it becomes diffi-
cult to contain the dynamics. There are strong incentives for arbitrage if a
carrier tries to charge above cost. This will be discussed in Chapter 40.
Merging of Technologies
Other challenges to the centralized network arise when traditional telecommu-

nications networks and equipment increasingly become contiguous and over-
lap with previously separate sectors, such as computers, office equipment,
and broadcasting. One response for PTTs has been to enter new fields, such
as cable television, master television antennas, electronic publishing, com-
puter utilities, electronic mail, modems, answering machines, and so on. Such
moves lead to conflict with interests that were not previously part of the postal-
industrial sector, leading to unsecured boundary regions and to confrontations
with other powerful government entities.
Governmental Industrial Policy
Governmental industrial policies and regional economic collaborations also have

implications for telecommunications monopolies. In most developed countries,
governments have established support programs in electronics. While the PTTs
usually have been an important practical and regulatory part of this effort (Nora
and Mine, 1980), such industrial policy has created a support structure for their
future rivals. For example, several European countries finance satellite devel-
opment programs in order to enhance their electronic and aerospace industries
through civilian programs. The development of such satellites establishes the
imperative for their actual operation and their financial viability, which in turn
opens the door to potential future use by new types of carriers. In other coun-
tries, such as Britain, cable television is supported as part of technological
development, with the goal of establishing optical switched-star cable systems.
These networks could be used in the future for more general telecommunica-
tions purposes. Elsewhere, various local governments are advancing so-called
teleports as part of regional developments, thereby providing the facilities for
bypassing the public network.
Regional Collaboration
Regional and supranational economic collaborations like the European Com-

munity have challenged protective domestic arrangements between industry and
government. The European Commission has overturned a whole range of do-
mestic and collaborative policies defended by PTTs. This is discussed in the
chapter on international organizations.
A Breakdown in Intra-PTT Cohesion and Labor Support
The traditional organizational integration between the postal and telecommuni-

cations components of the PTTs has become subject to strain. Increasingly,
telecommunications employees have backgrounds and salary expectations sim-
ilar to those of workers in the high-technology sector, yet they are constrained
individually by civil service standards and collectively by the burden of the
postal service's deficits. At the same time, PTT managers have begun to rec-
ognize that a change in organizational status can give them substantial flexibil-
ity and independence. Therefore, they have become willing to transform tele-
communications operations from government administrations into government-
held corporations. Thus, the growing significance and attractiveness of the elec-
tronics sector led PTT employees and managers to begin defining their interests
as different from the collective PTT and to seek a more advantageous non-civil
service status in a semi-independent private-law organization separate from the
governmental regulatory organization, especially if they could maintain most
of their civil service benefits at the same time. As its constituent parts began
to redefine their interests, the PTT's traditional coalition weakened.
This chapter has described the centrifugal forces which have led, after a long
period of stability, to changes in Europe. The policy options will be the subject
of the next chapter.
5
Defense of the Telecommunications
Monopoly
As telecommunications moved in new directions, the PTTs were increasingly

faced with the task of justifying the existence of a rigid monopoly in essentially
market-oriented economies, delivering a service that could be provided by the
private sector. This is not, after all, a situation where the public subsidizes
unprofitable but socially desirable enterprises such as railroads. The arguments
in favor of the traditional system will be discussed in what follows. It is im-
portant to stress at the outset that the issue is not the existence of publicly
supplied telecommunications services, but rather their exclusivity.
The Need for Social Subsidies
It is often believed that a monopoly is necessary to ensure socially meritorious

goals such as service for the poor. The fulfillment of this obligation necessitates
subsidization, which must come out of revenues from more profitable telecom-
munications services. Liberalization would attract competitors to precisely those
areas and customers that are most profitable, and such "cream-skimming" would
reduce PTT profits and cut into the cross-subsidies of their socially desirable
services. But it is elementary in the analysis of public finance issues to separate
conceptually the mechanism of taxation from that of distribution. There are
various ways of providing subsidies to support universal services. Alternative
mechanisms for the allocation of government subsidies are readily used for the
distribution of most other goods and services. Such subsidies can be distributed
through welfare or social service programs, cash payments, entitlement stamps,
tax incentives, or producer subsidies. Therefore, it is not necessary that tele-
communication services be supplied by a monopoly to ensure their adequacy.
Nor is supply by a governmental entity required. Even if there is a role for
government in assuring the maintenance of universal service, for example, as
a "supplier of last resort," this does not prove the need for total monopoliza-
tion. Railroads, for example are frequently government owned, yet they com-
pete with private truckers, barges, airlines, and, often, supplementary private
55
56 Defense of the Telecommunications Monopoly
railroads. Nor is there any reason for the source of a subsidy to be the telecom-
munications system itself. It can be provided from general revenues and raised
through general and progressive taxation.
The usual response to these points is that they are politically unrealistic.
Advocates of the monopoly argue that the political system would not permit
massive cross-subsidies if they were transparent to the government and the
legislature of a country. This is probably true in Europe, as it is in the United
States. But the argument is at heart antidemocratic. If European governments
and legislatures, faced with a transparent choice, choose to cut back some sub-
sidies in favor of other social goals, they merely exercise their role of setting
priorities for public expenditures. (In the author's view, some such subsidies
can be justified; he drafted New York State's lifeline program, entitling 1.5
million poor households to telephone service at $1 per month. See, addition-
ally, the arguments for a critical mass subsidy to initiate a service, provided in
Chapter 3.)
The cross-subsidy rationale is normally used in support of basic services for
disadvantaged categories of users or areas. But similar arguments have also
been advanced to support new and advanced services such as data transmission
and value-added services, with the claim that new technological services gen-
erally lose money initially, and their introduction therefore necessitates some
form of critical mass subsidy. But this argument can easily become circular. A
government monopoly is defended on the grounds that it is necessary to en-
courage the introduction of new services that would not otherwise be supplied
by private firms. But the private firms cannot enter because they are not al-
lowed to undercut the monopoly that is deemed necessary to support the new
services.
Many people sincerely believe that universal service depends on a public
monopoly. This, however, is not correct. First, for most of the monopoly pe-
riod there was nothing close to universal service in existence, and investments
were targeted highly unequally. In France, for example, telephone service was
offered far more extensively in the Paris region than in the provinces (see Fig-
ure 9.1, p. 141). At the turn of the century, Berlin and New York City had
similar telephone densities (one telephone per twenty-nine persons in New York,
and one for forty-three in Berlin). But in the rural state of Iowa, telephone
density was twice as high as in New York State (one telephone per nineteen
people), with 170 different telephone companies and cooperatives providing
service, whereas rural districts in Germany had one phone per 500 people, less
than one-tenth of the Berlin density (Brock, 1981). In 1960, the percentage of
German farmers with telephones was an extraordinarily low 9 percent (Schulte,
1982). In 1980, it had gone up to 80 percent. In France, it was 9 percent in
1968, 73 percent in 1980, and 93 percent in 1985 (T. Vedel, 1988, commu-
nication). In the United States, according to the Rural Electrification Adminis-
tration, 96 percent of all farms had telephones in 1988. In typical farm states
such as Iowa and Kansas, telephone densities were higher, with 95.1 and 95.2
per 100 households, than the national average of 92.4 (Federal-State Joint
Board, 1988), even after a more liberalized telecommunications regime. 1
Defense of the Telecommunications Monopoly 57
Natural Monopoly
A serious argument for PTT exclusivity is that communications services are a
"natural monopoly," that is, that single-firm production is cheaper, at every
level of output, than multifirm production. This is true when average (and mar-
ginal) costs are continuously declining with output. In that case, there are effi-
ciency advantages to the largest network the market will bear. But if the un-
derlying characteristics are truly naturally monopolistic, there should be little
or no danger of competitive entry, since new entrants, unless they could sup-
port losses for a sufficiently long growth period, would inevitably fail. What
then is the purpose of prohibition of entry if economic barriers would be enough?
The counterargument is that, although theoretically no entry would be prof-
itable, in reality the rate structure is based on social goals, which leads to tariffs
above cost for some services in order to provide for the subsidization of others.
These high-profit services would then become the object of "cream-skimming"
in a liberalized environment, even though it would be economically inefficient
to provide them outside the monopoly. This reasoning links the cross-subsidy
argument with the natural monopoly argument.
The most sophisticated analysis along these lines was advanced in the United
States as part of new industrial organization theory linked to William Baumol
and others such as Elizabeth Bailey, Gerald Faulhaber, John Panzar, and Rob-
ert Willig. These writers demonstrated that in a multiproduct setting (i.e., where
a firm supplies an array of related items), such a firm may have overall econ-
omies of scale, but under certain conditions its position may become "unsus-
tainable" because particular product lines are vulnerable to competition, even
in the absence of cross-subsidies (Baumol et al., 1982). Such loss of the mo-
nopoly would result in a loss of economies of scale and scope, that is, of
size and of joint production. This implies that even in a system free of cross-
subsidies, a natural monopoly could be subject to competitive entry. The ar-
gument is effective in the abstract, but it is based on certain theoretical and
empirical assumptions that have been criticized (Shepherd, 1983).
There are other problems with the economies-of-scale argument. The first is
the question of whether they actually exist beyond some point, a question that
is a theoretical and empirical morass. Most older U.S. and Canadian econo-
metric studies of economies of scale, in the days of the AT&T monopoly, seem
to show that they are of a modest size (Meyer et al., 1980). A review of several
studies finds that, excluding one AT&T study, the range of scale elasticities for
the Bell system was between 0.98 and 1.24 (Mantell, 1974). Another study
(Nadiri and Schankerman, 1981) finds a much larger value of around 1.8, though
it was declining over time. An AT&T submission in the antitrust law suit (AT&T,
1976) shows a great range (0.74-2.08). Much lower results were found for
Bell Canada (0.85-1.11) (Dobell et al., 1972; Fuss and Waverman, 1977). The
FCC concluded, based on most of these studies, that the existence of econo-
mies of scale was not definite.
Furthermore, even if the existence of economies of scale is accepted, they
Network size
Figure 5.1. Economies of Scale and Shift of Cost Curve.
must also be shown to offset the inefficiencies of monopoly and regulation. For
example, regulation of rate of return has been shown to create an incentive
for overcapitalization (the Averch-Johnson effect). There is also the matter of
simple, accumulated high-cost operations. The operating costs of MCI per
revenue-minute were estimated as 24.8 cents, only three-fourths of the costs of
the much larger AT&T, which were estimated as 31.8 cents per revenue-minute
at the end of its near-monopoly phase (Chrust, 1985, p. 10).
These numbers indicate that economies of scale cannot be observed in the
abstract but must be considered with reference to the competitive pressures on
efficiency under which they are realized. The shape of average cost curves may
be downward-sloping, but the actual location of the cost curves is not indepen-
dent of different institutional arrangements and their dynamics. In other words,
one must distinguish between movement along a cost curve and a shift of the
cost curve itself. The latter, induced by competition, may be greater than the
lost economies of scale.
In Figure 5.1, for example, the average cost in the monopoly network is
declining, suggesting benefits of economies of scale.
At size H I , the cost is C\. However, suppose that competition is permitted.
A second network enters, and its rivalry leads to a lowered cost function at
each level of production, for either network, depicted as A ' , B'. Let us assume
that both networks split the market equally, that is, that the network size of
each is n\IT.. At that size, the cost of each is 2, which is lower than the cost
in the monopoly network.
This does not even take into account the increased demand due to lowered
prices and increased marketing efforts of the competitors, which should move
average cost further downward on the competitive-cost curve. In other words,
the existence of economies of scale does not dispose of the question of lower
cost.
Defense of the Telecommunications Monopoly 59
Furthermore, if the realization of economies of scale were the governing

issue for telecommunications policy, why would they coincide with the size of
a country? Should not logic suggest supranational carriers who can realize still
greater economies? Why not a Nordic or a Benelux PTT? Why not serve Den-
mark through the German Bundespost, just as Germany supplies many of Den-
mark's automobiles? These arrangements do not exist for a variety of reasons,
in particular because of the loss of sovereignty. But that simply means that
economies of scale, even where they exist, are only one factor among several
and are regularly disregarded for a variety of political or control reasons.
Technical Integrity
Yet another argument in favor of monopoly is that it is needed to assure the

technical integrity of the network. Liberalized entry into network services and
equipment could permit users to utilize equipment that might damage the net-
work and thus impose negative externalities on others.
In the equipment field, the need to protect the integrity of the network is not
a persuasive argument. It was used by AT&T for many years to prevent even
the attachment of nonelectrical equipment, such as plastic noise guards on tele-
phone sets. These restrictions were struck down in 1956 in the Hush-A-Phone
decision. Similarly, the German Bundespost at one time prohibited emergency
telephone number stickers supplied by private parties from being put on tele-
phones. Harm to the network can be prevented by establishing safety standards,
just as for other products.
Compatibility
A close cousin to the technical integrity argument is the claim that a monopoly
is necessary to ensure that all participants can actually communicate with each
other. If several carriers supply services and equipment, technical incompati-
bility could fragment the network, leading in time to the need to have multiple
telephone instruments on one's desk. This argument is flimsy at best. Standards
setting takes place in numerous industries without the need for monopoly pro-
duction. In telecommunications, much of it already takes place through various
international bodies. There are also economic incentives to compatibility. But
where these are not adequate, legal requirements and technical standards of
interconnection and access can assure that there will not be fragmentation into
incompatible networks.
Comprehensive Planning
Telecommunications, it is argued, requires a complex and long-term planning

process, for which a market mechanism is not well suited. This claim tends to
view competing carriers as tiny operations rather than as large corporations

whose very essence is their ability to engage in long-range planning. Further-
more, it assumes that the traditional dominant carriers would cease to engage
in planning and that an overarching plan could not be fashioned by a govern-
ment in the absence of a full monopoly. It is true that planning becomes more
complex as the number of participants increases. On the other hand, there is
likely to be a more dynamic planning and execution process in a nonmonopol-
istic setting. This issue is discussed in Chapter 39 on the open network system,
where the necessity of a move toward a modularized network system is argued.
Communications as a Responsibility of the Government Infrastructure
Another argument in favor of PIT monopoly is that control over communica-

tions, because it is an essential infrastructure service, is the responsibility of
the government. This view holds it to be undesirable to leave the sector under
private control, since control over a country's electronic highways would afford
unacceptable private power. Telecommunications is considered a quintessential
public service, as is transportation over roads, canals, and railroads. This ar-
gument does not usually distinguish between a government assuring such ser-
vices and a government providing them. For example, the state should assure
that no part of the population will starve. But this does not mean that the
government itself must grow the food, and still less that it must control all food
stores or supermarkets.
Still, the view that private power over communications is to be avoided is a
legitimate one. And although it might be said that competition would prevent
one firm from becoming predominant and that competitive rates would make
manipulation of communications difficult, these are essentially market-oriented
arguments, and they are based on value judgments. For those who support
public ownership of the means of production and a reduced role for the private
sector, they will not be persuasive. Clearly, if government ownership per se is
a preferred value, arguments made for the pursuit of efficiency or for the use
of alternative means of achieving redistributory goals are inadequate. Advo-
cates of state control over communications see liberalization as an increase in
the political power of business, as a subordinating of communications to the
profit motive, and as a potential threat to national sovereignty. One can dis-
agree with these views or with the reality of the threats, but not with the legit-
imacy of holding them.
6
Policy Directions
Policy Options
Given the forces of change, what are the policy directions? It is necessary to
distinguish between several very different reform concepts that are frequently
intermingled under the vague heading of "deregulation."
Corporatization
Under corporatization, the status of telecommunications operation is trans-
formed from a state administration (i.e., part of the civil service) into a publicly
owned corporation. Such a course is usually advocated to encourage efficiency
and flexibility. The monopoly status is not touched by corporatization as such,
though once the close link to the government is severed, a process is set in
motion that makes further changes more likely. Sometimes the corporatized
entity is described as a "private" firm, in the sense that it may be organized
under private law provisions (which determine its status in, for example, con-
tract and labor law), but that description confuses legal detail with the reality
of control, which is still governmental. Corporatization is usually accompanied
by the creation of a regulatory mechanism of government, typically as part of
the PTT ministry.
Privatization
A privatization policy may follow corporatization as the government sells part
or all of the shares in the telecommunications entity to private investors. Pri-
vatization, too, need not affect the monopoly status. In the United States, AT&T
was private and a near monopoly for a very long period.
Ownership per se does not determine economic performance. For example,
the nationalized French banks have performed quite well. More relevant is the
extent of state political interference. Such interference can take place in nomi-
nally private organizations such as Italy's SIP. In contrast, a government or-
ganization such as France Telecom has established significant independence
regardless of its legal status.
61
62 Policy Directions
Liberalization
Liberalization involves a more pronounced change in market structure. By opening
equipment supply and services to entry, liberalization changes the nature of
markets and the power of the established suppliers. Privatization is more dra-
matic, but liberalization is by far the more important policy. Advocates of
telecommunications reform tend to combine and sometimes confuse these two
policies.
Devolution
A prime example of devolution is the divestiture of AT&T in America. It is
not necessary for either liberalization or privatization, although it addresses the
problems of competitive barriers and market power. A devolution can also be
part of liberalization, where some segments of the market are opened up, and
others are not. If different regulatory treatment of carriers active in both open
and restricted markets is sought, a policy of devolution may make sense. De-
volution is also possible without liberalization. In Denmark and Italy, for ex-
ample, the public network had been segmented but monopolistic.
Deregulation
Despite the frequent usage of the term deregulation, it is an imprecise concept

and is usually used as a synonym for liberalization, that is, for a lowering of
entry barriers or other restrictions. But it is also used to mean a reduction in
red tape and government involvement. This second meaning can easily be at
odds with the first: Partial liberalization as a process tends to complicate things
much more than an outright prohibition and can lead to a more extensive set
of rules to address the new problems. To analogize: It is much simpler to ban
all trucks from using a highway than to permit some of them to use the road.
In the latter case, one must establish rules on weight, speed, time of day, and
so on, and enforce them.
Deregulation is often a euphemism for laissez-faire, a term discredited by
nineteenth-century robber barons. In America, deregulation has a more benign
sound, since aversion to the heavy hand of government regulation has been a
theme on which a wide range of the political spectrum can agree as a general
proposition, though rarely in a concrete case. In America, the accelerated pen-
etration of electronic technology into the telecommunications sector coincided
and interacted with an intellectual and political move toward more laissez-faire.
In Europe, the new technology was similarly available, but the ideological re-
ceptivity to new institutional arrangements was different. For most Europeans
the trend of economic history had been toward increased forms of public con-
trol. The political left took the scientific inevitability of this trend as dogma;
the conservative right, though parts of it were fighting against public control,
was in doubt of its own long-term prospects in stemming the trend.
Policy Directions 63
The American historical experience, for a long time, followed the expected
path from relatively unbridled laissez-faire capitalism to a regulatory system
that kept steadily but unspectacularly expanding in the decades following the
Great Depression and World War II. The unusual aspect of recent develop-
ments in U.S., U.K., and Japanese regulatory policy is their reversal of this
historical trend.
Within the spectrum of European policy response, one extreme is the United
Kingdom, whose government, under former Prime Minister Margaret Thatcher,
supported a free-market economy. The British government brought about sep-
aration of the telecommunications monopoly British Telecom from the postal
services, and reorganized it as a privately owned corporation subject to some
competition.
While Britain was consciously attempting to raise its high-technology stan-
dards through market forces, with the government attempting to create an en-
trepreneurial environment, French policies increased the governmental role. The
Socialist government set development of a high-technology electronics industry
as a national priority and nationalized much of the French electronics and tele-
communications equipment industry to gain a lever for the achievement of this
goal. This policy, only partly reversed by conservative government, had the
support of a large part of the public and intelligentsia. The effect is that the
French actually created, for a time, a state-owned analog of the old AT&T
system: a vertically integrated complex of equipment manufacturing coupled
with a telecommunications transmission monopoly and an R&D laboratory. At
the same time that the Bell telecommunications monopoly in the United States
was being segmented into several component parts, the French did the opposite
and integrated, for the first time, the major elements of telecommunications
under one ownership (though the French state is admittedly a very large um-
brella whose spokes may be at odds with one another). Equipment manufacture
was partly privatized again and the telephone administration was made more
independent in 1990. But the state and its affiliated institutions are strongly in
charge.
The telecommunications policy of the Federal Republic of Germany lies
somewhere between the liberalization of the United Kingdom and the nation-
alization of France. For a long time, the Deutsche Bundespost was loath to
relinquish its monopoly power over domestic and international telecommuni-
cations, and it has labored to protect the status quo. Only reluctantly, under
the prodding of other parts of the government and following the Dutch ex-
ample, were institutional reforms moved forward in 1988. The telephone ad-
ministration was separated from postal matters and corporatized as a state or-
ganization, with greater operational separation from the PTT Ministry, which
now functioned as the regulatory agency. Similar courses were taken by the
Netherlands, Denmark, and several other countries.
These four positionscorporatization, privatization, liberalization, and de-
volutionconstituted the primary policy menu of European countries in the
1980s. Yet for all their differences, European strategies also had an important
common element: assigning the national telecommunications carrier a role as a
64 Policy Directions
locomotive for national high-technology development. This is discussed in the

following section.
Political Telematique: The New Role of the PTTs
For a long time, the mission of the postal-industrial complex in developed

countries centered around the achievement of a high degree of penetration.
"Universal service" was the organizational goal, and it happily brought to-
gether the interests of the manufacturers, users, and politicians. With the goal
of universal service largely achieved through effective collaboration of PTTs
and industry, a new organizational ideology needed to be articulated, both to
instill a sense of purpose internally and to legitimize the continuation of the
institutional regime externally. The new views were influentially expressed in
the 1978 French Nora-Mine report. This government report concerned itself at
length with the threat to national sovereignty posed by the government's lack
of control over electronics, that industry of the future whose paragon, IBM,
was viewed as threatening French sovereignty:
As a controller of networks, the company would take on a dimension extending
beyond the strictly industrial sphere; it would participate, whether it wanted to or
not, in the government of the planet. In effect, it has everything it needs to become
one of the great world regulatory systems [Nora and Mine, 1980, p. 72].
How was the government to deal with IBM, "one of the great actors on the
world stage"? The growing interaction between computer technology and tele-
communications made control possible. Although governmental influence over
the computer industry was limited, the latter's overlap with telecommunica-
tionsover which the state traditionally had controlprovided the state with a
lever of power.
[National governments need to] strengthen their bargaining position with a solid
mastery of their communications media. . . . However, the difficulty lies even
more in the fact that no country can play that role alone. . . . But the internation-
alization of the stakes means that today no economic Gallicanism is sufficient to
keep Rome out of Armonk. Independence would be vain and as easy to outflank
as a useless Maginot Line if it were not supported by an international alliance
having the same objectives [Nora and Mine, 1980, pp. 72-73].
Nora and Mine coined the term telematique for the sector that is also var-
iously referred to as IT, C & C, informatics, or compunications. Their analysis,
which may be termed political telematique, became extraordinarily influential
as PTTs embraced its notions, which assigned to them a central role in high-
technology policy and in the preservation of the national interest against Amer-
ican (and later Japanese) economic and technical interests. The equipment in-
dustry was also supportive since the Nora-Mine report created a presumption
in favor of government subsidies and protectionism as a matter of preserving
national sovereignty.
Policy Directions 65
Political telematique's view of liberalization is negative. In 1984, the influ-

ential French daily newspaper Le Monde carried the Nora-Mine theme forward
in a series of lengthy articles assessing U.S. deregulation (le Boucher and Qua-
trepoint, 1984). It viewed the United States as engaged in two wars: militarily
against the Soviet Union and industrially against Japan, with the advanced tech-
nologies of computers and communications vital factors in both battles.
To win this international war, the United States deregulated and broke up
AT&T. At first glance this may be surprising. "Why smash this power [AT&T]
in the middle of a war against Japan?" (Le Monde, 1984). The answer is,
"Deregulation of communication in the U.S. has as its main goal to give
American industry a good 'kick in the pants,' in order to get it to start a con-
quest of the rest of the world."
Given the energizing effects of divestiture and deregulation, one would ex-
pect the United States to prefer to be the sole custodian and beneficiary of such
a system. Nevertheless, Le Monde saw the United States as proselytizing the
rest of the world. Opening the American equipment market to foreign imports
is part of a U.S. export offensive. International liberalization gives the United
States several advantages. It reduces the communications costs of multinational
(i.e., American) users and it pries open the European equipment market. Once
"liberated," European telecommunications would be captured by American firms.
"Would not abandonment of state control over communications cause them to
fall under the control of IBM?" Having posed the issue in such a way, the
analysis leads to advocacy of political solutions, such as international agree-
ments and domestic restrictions. Liberalizing change is described as profoundly
threatening to French and European economic and sovereignty interests, and
therefore requires energetic containment by government.
The past several chapters of this book have laid out a more theoretical or
general framework for the evaluation of telecommunications. The next sections
will analyze, in much greater detail, the past, present, and future of the various
national systems.
II
THE TELECOMMUNICATIONS
SYSTEMS OF EUROPEAN
COUNTRIES
7
Germany
History of the Reichspost
For a discussion of the early history of German postal services, telecommuni-

cations, and institutions, the reader is referred to Chapter 2, in which the emer-
gence of the country's PTT system is discussed.
The history of German telecommunications can be said to have begun in
1832, when Prussia introduced an optical telegraph system that connected Ber-
lin with Koblenz on the Rhine. The system was designed for official dispatches
only, and a petition by the Berlin Chamber of Commerce to transmit stock
market information was emphatically rejected (Allentier, 1973). Later, civilians
were permitted to use the telegraph during slack periods, as an effort to supple-
ment state revenues.
After the invention of the electrical telegraph, Werner Siemens' newly estab-
lished firm was instrumental in promoting a Prussian governmental line from
Berlin to Frankfurt in 1848, where the first German parliament was then meet-
ing.
In governmental debates on the establishment of Prussian telegraph service,
strict state control and monitoring over civilian messages was held necessary
to protect the postal service, to supervise politically oriented telegraph mes-
sages, and to limit "unfair stock market maneuvers."
Most German principalities introduced their own state telegraph service. In
1850, the German Telegraph Union was created for technical and operational
coordination. In 1876, the newly established German state decided to expand
the backward telegraph system. The government appointed Heinrich Stephan,
director general of the postal system, to also head the telegraph office. Within
a short time, Stephan merged the two services, despite some opposition from
the telegraph personnel.
The invention of the telephone included German participation. An early ver-
sion was conceived by Philipp Reis (1834-1874), a teacher who demonstrated
the principles of voice transmission in 1861. His device, however, never worked
in practice.
In 1877, news of Bell's invention of the telephone reached Berlin. After
reading about it in Scientific American, Stephan immediately wrote to Boston
to obtain sample equipment and then personally experimented with it. In a
handwritten memorandum to Bismarck, he predicted a great future for the tele-
69
70 Germany
phone. But the purpose of German telephony was distinctly different. Stephan
viewed the telephone as an extension of state telegraphy, to be used in rural
post offices where trained telegraph operators were not economically feasible.
It took Emil Rathenau, a noted Berlin industrialist with a particular interest in
applications of electricity, to establish the concept of public telephony on users'
premises.
Because Bell's telephone was introduced in Germany three months after the
German patent law was passed on July 1, 1877, Bell's invention fell between
the cracks and received no patent. Thus, the manufacture of Bell equipment
was free in Germany, and Werner Siemens began producing and improving it
(Siemens, 1949, p. 199). Through their companies, Siemens and Rathenau
henceforth became the main protagonists in German telecommunications equip-
ment manufacturing. Together with Stephan, they created German telecommu-
nications, from the beginning a close alliance of governmental and business
interests.
Stephan called upon Rathenau to organize the Berlin business community for
the first telephone exchange in Germany. Rathenau, one of the more interesting
figures in German economic and political history, agreed to work on the intro-
duction of telephone service in Berlin (Goetzler, 1976, p. 6). After consider-
able difficulty, the Berlin exchange opened in January 1881, connecting eight
subscribers. Official German accounts of the introduction of telephone service
traditionally describe the Berlin exchange as the country's first. Actually, the
first system was established in Mulhouse in what was then German Alsatia,
and by private rather than postal initiative. However, not wishing to be up-
staged by its reluctant province, the Berlin authorities denied the Mulhouse
system an official permit to operate until after the Berlin state system was
opened.
Local exchange areas were consolidated into territorial districts, and after
1884 these were interconnected by long-distance trunks. Telephone service ex-
panded, and Germany had 10,000 local post offices with telephone service by
1890. Besides these state-run telephones, however, there was little penetration.
But the spread of electronic outposts into the countryside was immensely wel-
come, and the Reichspost, and Stephan at its helm, became enormously popu-
lar.
In 1882, the Reichspost established conditions of private interconnection with
the public network. Ten years later there were more than 2000 private lines
connecting with the public system, and almost 3000 purely private telephone
systems within private properties such as factories. In 1892, the comprehensive
telephone authority of the state was established by law.
The next decades witnessed steady improvements of technology, services,
and penetration of the telephone. In 1886, the first international public tele-
phone communications were established between a town in Alsace and Basel,
Switzerland. A trial for an American-made automatic exchange was inaugu-
rated in 1900, and German equipment came on line in 1908. In 1909, wireless
telegraphy was introduced.
After the initial push into rural post offices, the development of telephony in
Germany 71
Germany was much more centered in the cities than in the United States. In
1902, overall telephone density in the United States was one telephone for each
thirty-four persons, as compared to a much lower density of one telephone per
128 people in Germany. But for the cities, the ratio was more similar. The
number of people per telephone in New York was thirty-nine, in Chicago,
thirty; and in Boston, nineteen. In Berlin the density at the same time was
forty-three persons per telephone. It was fifty-two in Hamburg, seventy-seven
in Frankfurt, and thirty in Stuttgart. But in rural areas the matter was quite
different. In New York State, telephone density was one telephone per thirty-
one inhabitants; in Illinois, twenty-four, and in Iowa, a largely agricultural
state, one per nineteen. In Germany, on the other hand, one eastern Prussian
district had one telephone for 634 inhabitants, and in many others the density
was less than one per 500 (Holcombe, 1911, p. 432).
During the years of the Weimar Republic, technology, subscribership, and
services all steadily expanded. In 1923, subscriber trunk dialing was introduced
in Bavaria. Intercontinental telephone service to the United States was inaugu-
rated in 1928.
Rampant postwar inflation made the Reichspost dependent on government
subsidies, until the 1924 Imperial Postal Finances Act (Reichspostfinanzgesetz)
established financial autonomy and control by an administrative council rather
than parliament.
When the National Socialists assumed power in 1933, Adolf Hitler appointed
W. Ohnesorge as state secretary (later upgraded to minister) in charge of the
Reichspost.'
History of the Equipment Industry
The histories of German public telecommunications and of the private Siemens

firm are closely intertwined. From its humble beginnings, Siemens rose to be-
come a giant in international telecommunications, thoroughly dominating most
aspects of telecommunications equipment in Germany and one of the world's
largest electrical and electronics groups (Dodsworth, 1987). Siemens is the largest
electronics and second largest industrial firm in Germany. The company's
equipment is manufactured in more than two dozen countries, and its employ-
ees worldwide number about 376,000, with 10 percent in research and devel-
opment (Appel, 1990). Siemens has a strong presence in components manufac-
ture and information systems and is an integrated supplier of information
technology. Outside of telecommunications, its products include power-gener-
ating equipment, nuclear power technology, and medical apparatus. Siemens
makes over 200,000 products through some 300 divisions (Fuhrman, 1990, p.
98).
Siemens has been involved in telecommunications since its inception in 1847.
Entering the telegraph business, it initially bet on the Wheatstone pointer tele-
graph (Siemens, 1949, p. 1-29). But when an address by the king of Prussia,
transmitted to the 1848 Frankfurt parliament by the Morse telegraph in a quarter
72 Germany
of an hour, took the Wheatstone system seven hours, postal authorities shifted
their preferences. Siemens adopted the new equipment but found himself al-
most immediately embroiled in a public debate about the effectiveness of his
buried telegraph lines. After he published a pamphlet that criticized the state
authorities' construction method, the firm lost all governmental contracts and
almost went out of business. Fortunately, other governments, particularly Rus-
sia's, had their own needs for telegraph equipment and the firm was thus kept
alive.
Siemens and his brothers set up operations in several major European coun-
tries (Scott, 1958). Werner was the "Berlin Siemens," and the driving force
of the main firm, Siemens & Halske. (The related firm, Siemens & Schuckert,
manufactured locomotives and heavy electrical equipment). A gifted engineer
and a courageous entrepreneur, he was also a political liberal of sorts who did
not hesitate to have himself elected to the Prussian Parliament as an opposition
deputy, despite the fact that the government was the major customer of the
fledgling company. His brother, Carl Wilhelm, who lived in London and even-
tually became Sir William Siemens, was the main figure in the separate British
firm Siemens Brothers. Heinrich, known as the St. Petersburg Siemens, looked
after the family's substantial Russian business. A close relative, Georg Sie-
mens, was the founder and head of the Deutsche Bank, which rapidly became
the most important financial institution in Germany and which still remains the
Hausbank of the firm.
With his relatively liberal views, Werner Siemens was similar to two other
major figures in the German electric industry, Emil Rathenau, the founder of
AEG, and later Robert Bosch, founder of the Bosch Company (Heuss, 1946).
AEG-Telefunken became the second most important German telecommunica-
tions company. After its decline in the 1970s, its position was assumed by the
increasingly diversified Bosch, a very successful firm that began in automotive
component supply and consumer durables and increasingly moved into tele-
communications. The electronics industry was progressive and internationalist
in comparison with the heavy industry that set the image of German business.
The relationship between Rathenau and Siemens was characterized by both
competition and cartel behavior. Rathenau founded the German Edison Com-
pany in 1883. According to an agreement, Siemens stayed out of electric dis-
tribution, which was reserved to Rathenau's firm, which in turn bought all of
its equipment, except for light bulbs, from Siemens. Rathenau introduced elec-
tric lighting into Germany by creating a complex supporting structure. His firm
was transformed into AEG (Allgemeine Electricitaets Gesellschaft, or General
Electric Company), with the Siemens firm as one of its major investors. When
alternating current was developed as a rival to the untransformable direct cur-
rent, both Siemens and Rathenau fought its introduction vehemently, just as
Edison did in the United States against Westinghouse, and just as unsuccess-
fully.
At Stephen's request, Siemens started to produce unpatented telephone
equipment. When Michael Pupin, a Columbia University professor, developed
the so-called Pupin coil, which greatly improved the amplification of signals,
Germany 73
Siemens acquired the German patent rights for a large sum and defended the
patent against challengers, which included the Reichspost. He told the inventor,
"As far as I know, this is the highest price that has ever been paid for a
mathematical formula; it goes far beyond the hundred oxen of the Pythagoras.
The Post Office knows very well why it led us in such a dance last year, for it
is now its turn to pay" (Siemens, vol. I, 1957, p. 181). Previously only one
among several competing firms for local systems, Siemens now acquired tech-
nical leadership in long-distance telecommunications equipment through the Pupin
license.
Siemens still had no dominance in switching, however. When the Reichspost
experimented with the introduction of the Strowger automatic telephone ex-
change, a clear advance over the hand-operated exchanges, Siemens waved the
flag to discourage the purchase of American equipment. Serious about auto-
matic switching, the Reichspost set up a consortium of companies to develop
such equipment. The consortium included Siemens despite its lack of experi-
ence in the area, but it eventually became the driving force and acquired the
German rights for the Strowger patent, and improved it. This was the beginning
of Siemens dominance in German switching equipment. As a family member,
Georg Siemens, wrote, "Siemens and Halske had created for themselves, with
their [exchange] systems, a technical monopoly" (Siemens, vol. II, 1949, p.
53). To deal with this situation, the Reichspost united other exchange equip-
ment companies in joint ventures.
Competition, cartelization, and the breaking of cartel agreements continued
to characterize relations among Germany's main electric equipment producers.
The "Big Three" of power generation agreed that they would explore and
inform each other of each request for a bid over a specified value. They also
agreed that they would bid predetermined prices that varied only slightly from
each other, and that they would permit the contracting firms to give price dis-
counts only if the order threatened to be lost to an outsider. This agreement
operated under the code name V.C., for Vertrauliche Corresponded, or con-
fidential correspondence. Similar relations existed in the production of incan-
descent lamps. According to Georg Siemens, "Due to senseless price slashing,
it had come to the point where lamps which twenty years earlier cost twenty
Marks, cost the same amount in Pfennig, and this could not go on. Thus in
1904, the 'Sales Organization United Incandescent Lamps' was established in
Berlin, a syndicate in which Siemens & Halske and AEG held a quota of 21
percent of total production" (Siemens, vol. II, 1949, p. 240). To tighten cartel
cooperation, a joint firm for incandescent lamps (Osram) was established in
1919, into which Siemens, AEG, and Auer brought their production facilities
and patents. Another selling cartel syndicate was founded in 1911 for insulated
wires. As Georg Siemens describes it, the wire cartel "claimed in its favor that
it kept its members to a strict adherence to the standards of the Association of
German Electro-technical Engineers, and provided them in return with a fairly
undisturbed independently wealthy existence" (Siemens, vol. II, 1949, p. 24).
Perhaps the longest-lived and most successful cartel in the German electrical
sector was the cable cartel, which dates back to 1876 and a large cable order
74 Germany
by the Reichspost to Siemens & Halske and Felten & Guilleaume. That cartel
set supply quotas for members in 1914 and established an enforcement office
that instructed each firm what price it could quote for an order. Any violation
led to a large contractual penalty. If a firm exceeded its quota, it was instructed
to quote high prices to avoid winning further contracts. The system operated
very successfully under strict cartel management which had access to relevant
financial records of the firm.
At AEG, meanwhile, Emil Rathenau was succeeded as president by his son
Walter. In addition to being a company executive, Walter was a novelist, es-
sayist, designer of the value-added tax (which fifty years later was adopted in
Europe), and head of Germany's wartime economic planning. After World War
I, the younger Rathenau gave a speech that discussed critically the reasons for
Germany's defeat. Even though he had been an essential figure in the economic
mobilization during the war, his speech led to vituperations against him that
were frequently anti-Semitic. According to Georg Siemens, "Particularly in the
circles of heavy industry, a virtual boycott against AEG was commenced, and
a number of large and important orders, which it would have received other-
wise, went to various competitors, primarily the Siemens-Schuckert firm" (Sie-
mens, vol. II, 1949, p. 283). AEG's central administration, aghast at its pres-
ident, forced him to resign. He then entered the political arena and served as
minister of foreign affairs until he was assassinated in 1922 by right-wing stu-
dents.
In the early development of the radio, Siemens and AEG were competitors.
Later, Emperor Wilhelm II, who was obsessively interested in naval matters
and therefore in the development of radio for navy and merchant marine, ex-
pressed his displeasure that two German groups were competing against each
other and weakening themselves in relation to the Marconi company. The mil-
itary authorities preferred a uniform solution. This led to a great deal of polit-
ical pressure on Siemens, and in 1903 to the formation of a joint company with
AEG. The joint company became the renowned Telefunken company, eventu-
ally part of AEG. The pioneer of German broadcasting, Hans Bredow, was a
director of Telefunken before he moved to the Reichspost, which he then es-
tablished as the undisputed ruler of the airwaves.
Telefunken held rights to an extraordinarily important amplification tube un-
der the Lieben patent. Without it no other company could produce an effective
radio set. For a while, Siemens, AEG, and Telefunken all sold the identical
product with different exteriors, under the protective umbrella of Reichspost
licensing.
In telephone equipment, Siemens produced a 2500-subscriber central battery
telephone exchange in 1909, the first PBX in 1912, automatic trunk dialing
equipment in 1923, and long-distance public telephones in 1929. In 1930, the
development of a full-wire transmission technique made possible a nationwide
telephone service, and in 1933 the company established the first telex network
in the world.
During the Nazi period, the firm's leader, Carl Friedrich von Siemens, kept
some distance from the regime, but the company benefitted from the rearma-
Germany 75
ment that took place. The war caused major destruction to Siemens plants, and
a large portion of its facilities in Berlin was dismantled by the victorious Rus-
sians. Both Siemens & Halske and Siemens & Schuckert were restructured,
and most of the headquarters was moved to Munich. The firm soon resumed
its place as the foremost name in the German electronics sector.
The Present German Telecommunications Industry
In addition to Siemens and AEG, the important German equipment suppliers

are Standard Elektrik Lorenz (SEL), an ITT and later Alcatel subsidiary; AEG-
Telefunken's subsidiary Telefonbau und Normalzeit, renamed Telenorma and
acquired by Bosch; Deutsche Telephone und Kabel Werke (DeTeWe), a Sie-
mens subsidiary that is also a part owner of Telenorma; ANT, with a heavy
Bosch participation; and Tekade, a Philips subsidiary.
The German telecommunications industry is particularly dependent on ex-
ports, which represent nearly 30 percent of sales. This is larger in percentage
terms than any other large industrial nation, except for Holland and Sweden.
According to the electronic industry's trade association, Germany in 1983 was
the largest exporter of communications products in the world (ZVEI, 1983, p.
58). Nonetheless, the government determined in that year that German infor-
mation technology was lagging in the world market (Neumann, 1986). Simi-
larly, in its 1983 annual report, the Bundesbank, Germany's central bank, found
Germany's export of high-technology goods to be inadequate. 2
In the 1960s Siemens tried to participate in the new world of computers, but
with mixed success. It produced computers under RCA's licenses but was left
in the lurch when RCA left the computer market. It then joined the Dutch
Philips and the French CII in Unidata, a joint venture, which, however, also
soon faltered. Next, it joined forces with the Japanese Fujitsu for IBM-com-
patible machines, and in 1974 it acquired the firm Computer GmbH, an unsuc-
cessful joint venture of AEG-Telefunken with Nixdorf (Malerba, 1985, p. 177).3
During the 1970s, Siemens seriously miscalculated in its development of a
semielectronic space division switch, which became obsolete before comple-
tion. It then intensively developed the successful digital EWSD switch. EWSD
has operated in South Africa since 1980 and in Germany since 1982. The EWSD
system is used for very large exchanges as well as for the German mobile radio
network and is the cornerstone of German ISDN efforts. In digital develop-
ment, Germany lagged somewhat behind the French, who began experimenting
with digital switches in 1966.
Siemens has also established a presence in integrated broadband fiber net-
work development and was in charge of two "BIGFON" broadband network
pilot projects. It is also a leader in electronic mail (teletext) systems.
In 1985, German government procurement was about 9 percent of total sales
for Siemens. The percentage was much higher in telecommunications, where
the Bundespost accounted for almost one-third of Siemens's 1986 sales of about
$4 billion.
76 Germany
For its American marketing, Siemens established a corporation in Boca Ra-

ton, Florida, to adapt its digital switch to American specifications, including
software that permitted centrex service. The undertaking was difficult and ex-
pensive: Siemens spent $200 million to adapt its products and had to pay Bell-
core, the local Bell companies technical arm, $10 million for equipment testing
(Dodsworth, 1987). To help its entry into the American market, Siemens con-
sidered a collaboration with the large and vertically integrated American tele-
communications company GTE. The arrangement would have permitted the
two companies to share development costs. The negotiations were on and off
for almost two years. But during this period, Siemens successfully courted
several American regional Bell companies and no longer needed GTE. In the
end, a 1986 agreement created a company into which GTE brought its Ameri-
can transmission equipment manufacturing as well as its manufacturing subsi-
diaries in Belgium, Italy, and Taiwan, thus providing Siemens with an entry
into these markets. Siemens controlled 80 percent of the new company and
paid GTE $400-$500 million. The American switching business of the two
companies was left outside the venture, however. This transaction was not much
noted at the time, because of the commotion over ITT's almost simultaneous
sale of control over its international equipment operations to the French firm
CGE. Together, these two transactions resulted in the large-scale withdrawal
of the two major American firms most active in the European telecommunica-
tions equipment market.
In 1989, Siemens purchased the PBX manufacturer Rolm from IBM, thus
ending the U.S. computer giant's frustrating attempt to establish a telecommu-
nications presence that had led to losses of $100 million in 1987 and 1988.
In 1990, Siemens restructured itself drastically, creating fifteen operating groups
and two divisions from the original seven. It continued acquisitions with the
American Bendix Electronics, the French computer firm IN2, and especially
the German computer company Nixdorf. It also participated in a hostile take-
overan unusual event in Europewith GEC of the U.K.'s Plessey, discussed
in the following chapter on U.K. telecommunications. But the collaboration
with GEC was not an easy one. The merger led to its assuming control in North
America of OPT Stromberg-Carlson, a supplier of switches to independent tele-
phone companies. It gained a major contract from Bell South, and was on its
way as the third supplier of the U.S. market.
AEG-Telefunken has been less prosperous than Siemens. The firm was ac-
tive in technical development, but since the mid-1970s was engaged in a con-
stant battle for survival. Losses were severe in home electronics and office
equipment, and it managed to lose money even in construction projects in Saudi
Arabia and Iraq. The firm would have gone bankrupt in 1983 without major
government involvement. Daimler-Benz, Germany's major industrial firm, sub-
sequently acquired a majority share.
Standard Elektrik Lorenz (SEL), now an Alcatel subsidiary, was acquired by
ITT in 1925 from AT&T's Western Electric. It is one of Alcatel's largest West
European units, with an annual volume of operations of about $2.2 billion.
Switching and transmission products accounted for 48 percent of sales in 1989.
Germany 77
The company had 21,000 employees in 1989 and was successful in worldwide
home electronics business, in particular through its television sets and digital
television system DigiVision.
SEL played a major role in the development of ITT's System 12 digital
switch. With the Bundespost's order for System 12 in 1982, SEL obtained
potentially 30-40 percent of the German digital exchange market.
Bosch, long a leader in automotive parts and household durables, is con-
trolled by a private charitable foundation. Like SEL, it is located in the Stutt-
gart area. It acquired Telenorma and its 18,000 employees from AEG, and
added more than 40 percent of ANT (a company with 6500 employees and a
leader in digital broadband switching, communications satellites, fiber optics,
and cellular telephones). Bosch also owns the mobile radio company Blau-
punkt, which is also active in automobile video display systems, television and
stereo, and Btx terminals. Bosch increased its investments in Telenorma, JS
Telecommunications, in France; Hasler, in Switzerland; and Telettra, in Italy
and created a mobile communications division. Taken together, the Bosch group
is a contender for the number two position in German telecommunications
equipment.
Nixdorf Computer was perhaps the greatest challenge to the established order
of Germany's electronic industry. The company was founded by Heinz Nixdorf
in a factory basement in 1952 with an investment of $6000, and it has remained
innovative and untraditional. By identifying market niches in the computer field,
it left behind large companies such as Siemens, Triumph-Adler (formerly of
AEG), and Kienzle. For example, Nixdorf had a large share of the German
market in computerized banking equipment and automated teller machines. In
1983, Nixdorf introduced Germany's first digital PBX, and by 1989, telecom-
munications accounted for over 10 percent of its revenues.
Nixdorf was also one of Europe's largest software developers. Four thousand
of its 20,000 employees are directly involved in the development of software
and provided about half of total revenues. Unlike other European computer
companies, Nixdorf also moved aggressively into the American market.
In 1978, Volkswagen made a bid to acquire a controlling share of Nixdorf
but was rebuffed. Eventually, however, Nixdorf's undercapitalization and tech-
nical ambitions caught up with the company. After its 1989 losses of close to
DM 1 billion, it was forced to sell a 51 percent stake to Siemens. Siemens then
merged its own computer operations into Siemens-Nixdorf Informations Sys-
tems, which became Europe's largest computer company.
Telecommunications Since 1970
Expansion of Service
The "democratization" phase of German telecommunications achieved univer-

sal service in a remarkably short time. Between 1970 and 1986, telephone
density increased from 14.3 to 43.8 lines per 100 population (DBF, 1987).
78 Germany
Figure 7.1. Telephone Main Lines per 100 Households by Occupational Group
in Germany. (Source: Schulte, 1982.)
The 1970s, in particular, was a time of telephone expansion, engineered by

the Social Democratic government, which was in power for the first time in
forty years. In 1970, there had been about 9 million main stations, but by 1980
there were more than 22 million. The use of telecommunications services per
subscriber household rose along with the greater penetration. The average
household increased its expenditure on telecommunications from 0.3 percent of
its total consumption in 1960 to 1.9 percent in 1980.
As can be seen in Figure 7.1, telephone penetration varied by economic and
social class (Schulte, 1982). Workers (skilled and unskilled) had the lowest
penetration rate: in 1980, 58 percent of unskilled workers and about 70 percent
of skilled workers had telephones. These figures are not particularly high. In
contrast, the rural population is well supplied, with-a penetration rate for farm-
ers of about 80 percent in 1980. As recently as 1970, less than 20 percent of
unskilled workers had a telephone. Skilled workers were only slightly better
off, and farmers had a penetration rate of only about 30 percent at that time.
In other words, their penetration had almost tripled in a decade.4 In 1960,
Germany 79
penetration had been less than 10 percent for these social groups! But for white-
collar employees too penetration was 84 percent in 1980, in contrast to 42
percent in 1970 and about 15 percent in 1960. Thus, for the vast majority of
German households, the telephone is a fairly recent acquisition. In Germany, a
sense of appreciation still exists toward the supplier of the service, the Bundes-
post, for making it possible to join the telecommunications network.
With the task of universal service largely completed, the next phase in tele-
communications, starting in the early 1980s, was to refocus priorities to in-
creasingly sophisticated telecommunications services, and this meant toward
the needs of large users, whatever the official philosophy proclaimed.
New Services
The usage and density of telex service in Germany are high. Germany's telex
network is the world's largest, three times denser than that of the United States,
and had thirty-nine telex accesses per 10,000 inhabitants in 1988. In 1988,
there were about 155,000 telex subscribers, down from the 1986 figure (ITU,
1990). Most likely, this number will continue to decline because of the emer-
gence of fax and electronic mail, known as teletext. (Teletext should not be
confused with the broadcast text service of the same name).
Teletext was introduced by the Bundespost after seven years of development
in 1981, following the recommendations of the blue-ribbon Commission on
Communication Techonology (KtK) panel. Growth was disappointing.
ISDN and videotex in Germany are discussed in other chapters of this book.
A few details will suffice here. In 1986, Germany conducted two ISDN trials
in Stuttgart and Mannheim. In addition, the Bundespost tested, after 1983,
fiber-optic broadband networks under the name of BIGFON, providing to users
two digital telephone lines, data and text services, two to four centrally switched
TV channels, twenty-four stereo audio channels, and a videophone video ca-
pability. 5
Regular ISDN service was introduced in 1989 at a Hanover trade fair. The
DBF signed up its first ISDN applications customer, a retail drugstore chain in
August 1989, but subsequent user demand was low (Schwarz-Schilling, 1990,
p. 58). The DBF therefore created a "U2000" network strategy, shifting em-
phasis from ISDN transmission to an intelligent network. ISDN plans were
expanded to include the distribution of 50,000 microcomputer boards to stim-
ulate demand (Gilhooly and Schenker, 1989).
The Deutsche Bundespost as an Institution
By any calculation, the Bundespost is a gigantic enterprise. The German Mo-

nopoly Commission observed that the DBF's 1980 budget of DM 55.4 billion
was significantly above that of Nordrhein-Westfalen, the largest German fed-
eral state, which encompasses nearly one-third of the West German population.
Its revenues were also much higher than those of any German manufacturing
80 Germany
firm. This comparison is even more impressive, since most of the Bundespost's
activities are domestically achieved.
Similarly significant are profits. The Monopoly Commission report found a
real profit of approximately DM 5 billion in 1980. It noted that, according to
the calculation of the Association of Postal Users, another DM 0.75 billion
ought to be considered profit, which would result in a profit margin of 15.8
percent. Total telecommunications profits are even larger if one disaggregates
them from postal services, which have been substantially in the red for many
years.
Telecommunications accounted for 70 percent of total DBF's revenue, with
postal services 26 percent and banking and financial services 4 percent.
By its own estimate, the Bundespost's share in Germany's GNP more than
doubled between 1960 and 1982, from 1.4 percent to 3.4 percent. In 1982, the
Bundespost provided secure jobs for more than 500,000 employees (about
216,000 in telecommunications). Not counting the effect of technology devel-
opment on exports, its procurement orders and investments provided jobs for
an additional 200,000 people (DBF, 1982a). Together with the hundreds of
thousands of retirees, relatives, and neighbors, this means a very large voting
constituency. In 1988, the combined DBF had DM 52.5 billion in revenues
and DM 16.9 billion in investments. The Postal service reported losses of DM
2.26 billion (Blau, 1989; Moschel, 1988; ITU, 1990).
Article 87 of the Basic Law (constitution) of the Federal Republic of Ger-
many postulates governmental control of telecommunications. The Postal Ad-
ministration Law of 1954 established the Bundespost as the organization serv-
ing the communication needs of Germany (Elias et al., 1980). The
Telecommunications Facilities Law of 1928 gave the Bundespost what it inter-
preted as the sole right to erect and operate telecommunications facilities. The
exceptions were private facilities not crossing property lines (and, under certain
conditions, connections between separate properties of the same owner) and
certain government authorities (such as the state railway Bundesbahn and the
armed force Bundeswehr) which are allowed to run their own internal com-
munications operations. The Minister of Postal Affairs could, in principle, grant
waivers to other providers, though this had not been done for many years.
Overall responsibility for the Bundespost lies with the Federal Minister of
Post and Telecommunications, who is a cabinet member and typically an influ-
ential party figure.
The DBF operated a substantial telecommunications technical center (FTZ)
in Darmstadt, a central approval office for telecommunications in Saarbruecken
(FZZ), two engineering colleges, a scientific institute, and more than 18,000
post offices throughout the country. Until 1982, the Bundespost also provided
a far-flung postal bus service that carried 281 million persons during its last
year of operation. For a long time, it was the largest bus operation in Europe
(Stuebing, 1982). The Bundespost had no monopoly over its bus routes but
could establish them virtually everywhere, while private competitors could run
lines where the Post buses already operated only by demonstrating that addi-
tional service was necessary.
Germany 81
The Bundespost also operates a huge postal financial system. In Germany,

the typical method of paying bills is not through a bank check but through a
postal pay order. Several billions of such financial transactions take place each
year. The Postal Savings Bank is a major thrift institution. In 1986, it had 20.9
million accounts in a country of 61 million and held an average aggregate credit
balance of DM 35 billion (DBF, 1987). The postal reform of 1989 made finan-
cial operations a semi-independent public corporation.
The bulk of Bundespost profits were raised by telephone services. Telegraph
services were subsidized at between 2 percent and 5 percent of cost, and postal
services covered only 86.2 percent of their very large costs in 1986 (DBF,
1987). By far the largest relative subsidy is extended to newspaper delivery,
which covered only 47.8 percent of its cost. Newspaper publishers thus benefit
substantially from the telecommunications sector subsidy of the postal service.
In 1982, 8111 newspapers were approved for the low-cost newspaper rate.
In addition to the internal subsidies mentioned, there were explicit levies on
the Bundespost. In 1982, over DM 4 billion were transferred to the govern-
ment. In percentage of total expenditures, the levy amounted to 7.1 percent in
1950, 6.0 percent in 1970, and 9.6 percent in 1982. In comparison, deprecia-
tion charges were only 50 percent more, and expenditure on outside capital
was substantially lower (DBF, 1982b). In 1986, the levy was DM 4.8 billion,
or 11 percent of profits (DBF, 1987). While the levy was a drain for the Bun-
despost, it also created a special interest of the government in its profitability.
In consequence, the Finance Ministry was a champion of institutional status
quo in discussions of DBF reform.
In addition to its role in high-tech development, one of the roles of the DBF
is its ability to function as a counter-cyclical investment vehicle. In 1982, with
German industry in general recession and its investment level decreased, BDP
investment was at a record level of 12.7 billion DM, and it was by far the
largest investor in Germany. 6 Furthermore, the DBF's procurement targeted
"preferred suppliers" which are often located in economically weak regions or
employing disabled individuals (Henneman, 1984, pp. 70-77).
Procurement Practices
Genera/
Procurement practices of the DBF have been a particular bone of contention,
both internationally and domestically. The German Monopoly Commission found
equipment prices to be generally higher domestically than in North America,
and the equipment was of a costly design.
In 1960, only 25 percent of total procurement was subjected to at least a
limited bidding process. This increased over the next twenty years to 40 per-
cent. A set-aside process was used in competitive procurement: Bids for part
of the quantity were solicited first, with the resultant price the upper limit for
the remaining quantity. The problem is that the supplying companies could
82 Germany
tacitly agree on a price leader and then set a high "market price" for the
second stage in which they would participate. In 1976, the Cartel Office im-
posed fines totaling DM 1 million against twelve telephone equipment manu-
facturers and twenty of their managers, after finding that in 1974 and 1976 they
had rigged bids for DBF terminal equipment.
The DBF's renowned research and development arm, the Fernmeldetechn-
isches Zentralamt (FTZ), played a key role in the process through its power
over the selection, licensing, and development of equipment. Almost 60 per-
cent of the DBF's procurement contracts were awarded by the FTZ. Tradition-
ally, it collaborated with a chosen supplier to develop products, with the other
manufacturers acting in effect as subcontractors to the primary firm.
In 1963, the German Federal Accounting Office issued a report criticizing
the Bundespost's procurement practices. A two-volume DBF response asserted
that it saw "no reason and no possibility for a fundamental change" (Scherer,
1985, p. 79). The theme of liberalization was then sounded by academics
(Mestmacker, 1980). Similarly, the German Monopoly Commission, in a major
report, found in 1981 that the DBF's practices contributed to the rigidity of the
supply market structure by its reliance on uniform equipment technology, and
a de facto concentration of procurement (Monopolkommission, 1981). The
principle of uniform technology (Einheitstechnik) was a key element in the
Bundespost's procurement practice. All suppliers had to conform to an agreed-
upon technology for a product; the four main development firms (Entwicklungs-
firmen) shared patents that were obtained in the course of the product devel-
opment. The Bundespost defended this policy, citing its advantages for tech-
nical planning, in particular in the postwar reconstruction, where it reduced
duplicative R&D. For digital switching, the Bundespost modified its principles,
but it had in any event no realistic alternative. It also opened procurement of
public packet switching systems to international bidding, and Northern Tele-
com, despite its views about domestic market restrictions quoted earlier, ob-
tained the order.
Procurement of Central Exchanges: A History

The closed nature of the procurement system, at least in the past, is observable
in the central exchange market. Here the Bundespost accounts for practically
all domestic purchases. Traditional suppliers included Siemens, with 40 percent
of the market; SEL (ITT, later Alcatel), with 30 percent; DeTeWe, with 14
percent; and Telenorma, with 10 percent. These figures understate the impor-
tance of Siemens, which also owns a substantial share in DeTeWe. Further-
more, Siemens had developed the switching system, was the system leader,
and had assured itself a dominant position through agreements with other sup-
pliers (Monopolkommission, 1981).
The basic procurement system for switching equipment was created in the
1920s, when the Reichspost introduced its "postal dial system," developed by
Siemens with a smaller partner. The Reichspost, not wanting to depend on a
single supplier, pressured Siemens to license its technology to the predecessor
Germany 83
of today's SEL (Mix & Genest), as well as DeTeWe, Lorenz, Berliner, and
other manufacturers, in return for an assured procurement quota. Starting in
1924, together with its small partner, Siemens received a set quota of 60 per-
cent of the procurement for the next five years. The agreement was renewed in
1927. In 1933, with the National Socialists in power, the nine major telecom-
munications equipment manufacturers, some of them now under new Aryan
ownership, entered into the so-called Telecommunications Domestic Contract
with each other. This contract governed all telecommunications transactions
with state agencies and private users. They carved up the market among them-
selves, with the Siemens group receiving 54 percent; the Standard group (which
consisted of Mix & Genest, C. Lorenz, and Berliner), 36 percent; and De-
TeWe, 10 percent of the Reichspost procurement volume. The contract was to
run for ten years. The Reichspost was notified of the agreement and stayed
within its provisions.
For the six firms that survived World War II, the quotas were maintained,
now justified by the need to stabilize the market where overcapacity existed
and to permit national planning. After 1951, the Bundespost notified the central
exchange manufacturers, by then three in number, to plan their production ca-
pacity each for a specific volume. In the following years, the watchdog Federal
Accounting Office began to show concern about the arrangement. The Bundes-
post insisted that even though the participants acknowledged the nature of its
coordination, it was only an internal administrative matter and did not include
an agreement between the Bundespost and firms or between the firms them-
selves. After 1954, the exclusive circle was expanded to include the firm T&N
(later renamed Telenorma), following that firm's legal pressures.
In 1955, the Bundespost decided to substitute its System 55 with the new
HMD switch developed by Siemens. Because it did not wish to depart from its
practice of using only one type of equipment, rival equipment by SEL was not
adopted. However, the DBF required Siemens to grant licenses to the other
three switch manufacturers. The decision to move to the new switch was un-
dertaken without involvement of the Bundespost's Supervisory Council. As part
of these transactions, in 1955 a highly confidential agreement was signed by
the Bundespost and the four industry manufacturers. The preamble of this "Heads
of Agreement" document stated that the Bundespost wished in the future to
include T&N in switching equipment contracts. The firms then agreed to a
quota of 46 percent for Siemens, 31.3 percent for SEL, 15.2 percent for De-
TeWe, and 7.5 percent for T&N. For trunk switches, a quota of 55 percent for
Siemens, 32 percent for SEL, 10 percent for DeTeWe, and 3 percent for T&N
was agreed upon (Scherer, 1985, p. 441). The firms also agreed not to contest
each others' market shares.
In subsequent years the Bundespost, with considerable legal agility, denied
the existence of quotas, but it largely adhered to the figures of the Heads of
Agreement. These market shares, combined with the Bundespost's detailed three-
to five-year projections of its total equipment needs, made reasonably accurate
production planning by the firms possible. In practice, orders for equipment
came up from the regional telecommunications directorates, together with pref-
84 Germany
erences for equipment suppliers. These orders were then aggregated, and, if
necessary, centrally modified to reach the target quotas. In other words, the
preferences of the Bundespost's own regional directorates could be disregarded
in order to meet the industry's agreement. Although the practices were criti-
cized by the Federal Accounting Office, the Bundespostdefying reality, eco-
nomics, and common senseinsisted that the quotas were only representing
real market shares and that its procurement practices did not establish any eco-
nomic power relationship or market concentration.
In the mid-1970s, as residential telephone service neared saturation and the
economy suffered from recession, telephone expansion dropped considerably.
In 1971 the waiting list for telephones had been 600,000, but by 1974 it had
shrunk to 17,000. Equipment volume contracted, and workers were laid off.
Alternative export markets were not easily available. The reliance on the safe
and traditional electromechanical switching technology, by then twenty years
old, left the manufacturers without a product that was competitive in the world
market.
In 1975, the Bundespost started to introduce price competition by permitting
limited changes in the procurement quota, according to prices offered. As a
result its bargaining position strengthened. The system provided for a quota
increase of 2 percent above that of the previous year for the lowest bidder. If
the Bundespost considered the lowest bid to be still too high, it could withdraw
the entire offer. If other firms lowered their prices to that of the lowest bidder,
they would share the remainder of the market among themselves, absorbing the
2 percent reduction of market share.
One problem in the adoption of the procurement method was that the small-
est and the most recent of the suppliers, T&N, behaved like a maverick, in-
sisting on a larger share than before. This forced the Bundespost to ask for the
blessing of the Federal Cartel Office for the scheme, which it received. T&N,
however, did not give in, and argued that violations of German and European
anticompetitive laws were taking place and demanded a quota of 15 percent.
This led again to intense negotiations with the four firms, the Bundespost, the
Cartel Office, and the Ministry of Economics. The firms and the Cartel Office
reached an agreement, but the Postal Ministry, in an uncharacteristic procom-
petitive stance, decided to challenge it. Eventually, a compromise was reached
that protected T&N's market share from an undue retaliatory drop. In 1977,
the Bundespost cemented T&N's cooperation by giving it the largest increase
in its share, from 7.5 percent to 8 percent.
The development of first-generation electronic switching proved to be a dis-
mal chapter in the joint R&D efforts of Bundespost and private industry. In a
major miscalculation, a fully digital development was considered and then dropped
in favor of a semielectronic space division system. The project foundered, be-
cause technical progress occurring during its development continuously led to
changes in the project goals, and because the firms had problems coordinating
the development among themselves. The system was intended to be introduced
in 1977 or 1978, but these target dates were pushed back to 1981 or 1982. It
became clear that the TDM digital technology was considerably superior to the
Germany 85
one on which the Bundespost and the manufacturers had bet. In 1979, Bundes-
post Minister Kurt Gescheidle pulled the plug on the development program. He
declared that technical development hit the Bundespost and its suppliers like a
"natural catastrophe" but that the Bundespost had learned from its mistakes,
and as a result, it would now closely follow technical developments in the
international telecommunications markets. The large losses in the project were
borne chiefly by the equipment industry.
After the dropping of the ill-fated analog system, the Bundespost initiated a
crash program in digital switching. It approached the major manufacturers to
submit new development models, though only up to three models would be
tested by the Bundespost, and of those, at most two would be chosen, with
cutover dates in 1984-1985. This led initially to three proposals, by SEL (ITT),
by Siemens and its affiliate, and by DeTeWe, T&N, and Tekede. The latter
group, possessing no digital exchanges of their own, did not meet the May
1982 deadline to continue the project, and thus only two models were presented
to the Bundespost. A competitive selection was not instituted, and both the
EWSD system of Siemens, and ITT's System 12 were accepted in 1983.
Customer Premises Equipment
Similar cartel arrangements existed for PBXs. Since the first private branch
exchange (PBX) was introduced in Germany in 1900, the penetration of PBXs
was substantial. There are more than a million PBXs in Germany, connecting
about 10 million extension stations. About 60 percent of total traffic volume
originated from or reached a PBX (ZVEI, 1983). The DBP's market share was
about 40 percent of units, but less in DM volume.
Until 1934, the market for PBXs was relatively free. All private suppliers
could enter and install equipment on private premises, though the Reichspost
was also a supplier. Competition led to technological development, but it also
led to severe price wars. In particular, the Reichspost was being undercut by
private competitors in a variety of ways. Until 1934, only compatibility stan-
dards were set for PBXs. In 1934, however, the PBX manufacturers considered
the price competition to be "ruinous" and set out to stabilize it. This was
accomplished through the establishment of uniform technical requirements and
performance standards.
The private suppliers agreed to conform to uniform conditions of supply as
well as to the prices of Reichspost-supplied PBXs. Any new supplier had to be
admitted to the production of PBXs. In other words, the Reichspost achieved
the elimination of competition that would undercut its own position while pro-
viding an umbrella for a price-fixing agreement among private suppliers.
The 1934 PBX rules also explicitly aimed to slow down technological de-
velopment. They sought to "cut back the overdeveloped(uberspitzte)technol-
ogy to the actual needs and to the associated reasonable extent, and to assure
that in further development only pertinent considerations would be determina-
tive" (Wittiber, 1934, p. 184, footnote 48).
In the 1980s a government report described the historical growth of this
86 Germany
regulatory system since the 1930s and found that "The result was a cartel
situation which divided the market among the suppliers. This did not come
about through prices, but through . . . standards regulation" (Mestma'cker,
1980, p. 185). The DBF bought, leased, and resold customer terminal equip-
ment. Traditionally, it was the sole supplier of simple terminal equipment. It
also provided, though not exclusively, PBXs, modems, and other more sophis-
ticated equipment, including telefax equipment. It had exclusive rights for ser-
vicing telex equipment but did not supply it.
Problems with this arrangement arose due to the Bundespost's triple function
in the telecommunications field: It regulated, operated, and competed. It was
the primary telecommunications regulatory authority of the country, with the
powers of government. At the same time, it was also the predominant network
operator: In equipment supply, it competed with private firms. The tension
between these several functions led to the appointment of a Kommission Deutsche
Bundespost in 1969-1970, which drafted a new charter, including the organi-
zational separation of political leadership and supervision, on the one hand,
and the actual operation, on the other hand. This bill was not passed by the
Bundestag (Elias et al., 1980).
The industry's attitude toward the Bundespost as a supplier was complex.
The DBF is, after all, its major customer and often a friendly supporter of the
industry's technical development and exports. But as a supply competitor, it
enjoyed advantages they resented: The Bundespost's equipment licensing au-
thority provides it with information about its own competitors' products; it had
no risk of bankruptcy; it did not have to set its pricing according to costs; and
it was able to subsidize its products internally for market penetration and deter-
rence of competitors (Monopolkommission, 1981).
The legal framework supported any DBF pricing policy. When equipment
was offered at a loss, it could be justified as a fulfillment of a public obligation;
when it was profitable, it was said to be part of the Bundespost's function to
balance its budget or to raise revenue for the federal budget.
In the customer terminal equipment market, there are eighteen firms, of which
the usual four (Siemens, SEL, T&N, DeTeWe) are the "development firms,"
while the others are "follower firms." According to the Monopoly Commis-
sion, these four have protected their technology through hundreds of patents.
The commission maintained that the "overlap [of patents] is obviously toler-
ated with mutual good will. Apparently the comprehensiveness of patent li-
censes is not utilized offensively to restrict competitors, but rather serves de-
fensively to secure individual production programs. The development firms show
a common interest in a peaceful coexistence in technological competition"
(Monopolkommission, 1981, para. 131).
The remaining eight suppliers of terminal equipment had a substantial 60
percent of the terminal set market. Those companies are fairly obscure and
depend on the Bundespost's good graces. They have no domestic distribution
system and have no need for one, since the Bundespost is their predominant
German customer.
The industry's ardent defense of the Bundespost suggests the symbiotic re-
Germany 87
lationship between the network monopoly and the equipment suppliers. The
trade association's (ZVEI) position on the role of the Bundespost has been:
"The German telecommunications industry is convinced that the functional in-
tegrity of the public telecommunications network, the nationwide availability
of services under the same conditions, the compatibility of communications
networks with all terminal devices, and the resulting freedom for all users to
communicate can be guaranteed only by a network monopoly" (ZVEI, 1983,
p. 43). The industry also defended the Bundespost's role as a distribution sys-
tem for small producers. "Participation of the Deutsche Bundespost in the ter-
minal equipment market is extremely advantageous for small and medium-sized
communication companies, which cannot meet the costs of a nationwide sales
and services network of their own, and therefore supply their product directly
to the Deutsche Bundespost. This allows them to achieve a notable share of
the market" (ZVEI, 1983, p. 45). This was in conflict with other industries,
where medium-sized producers were able to market their products, whether
through their own sales organization or through agents and intermediate whole-
salers, rather than having to depend on a state organization.
A complex bidding system was employed for terminal equipment that pro-
vided for greater competitiveness than for switching equipment. In the mid-
1970s the Bundespost charged suppliers with a breach of the price and quantity
agreement. They were fined, and the DBF tightened its procurement rules.
Immediately, the new system proved unworkable. The Siemens bid was by far
the lowest, and only SEL agreed to match it. The other ten firms refused and
made their high prices a David vs. Goliath issue for the survival of small and
medium sized companies. With business organization, unions, and politicians
being involved, the Bundespost retreated and announced its willingness to ac-
cept not the highest, but rather the third highest price as its guide. This was
accepted by eight of the ten companies, but the small firms were still unhappy
about the loss of the previously cozy arrangement. With the Federal Cartel
Office joining in to support the existence of a diversity of firms, the Bundespost
retreated still further.
The Bundespost strongly opposed the liberalization of terminal equipment on
legal grounds, based on a Telecommunications Order that declared terminal
equipment to be part of the public network. The order was confirmed by the
Federal Constitutional Court in 1978. The DBF also argued on technical grounds
that the separation of network and terminal equipment was impractical, since
terminal equipment fulfills a network protection function. To a certain extent
this is true, though other means of protection can be easily established. The
logical consequence of the Bundespost's view was that much of the office
equipment, as it becomes "smart" and connected to the network, could be
claimed to be subject to the Bundespost's regulatory jurisdiction and potential
control.
Conflicts about the Bundespost's role in equipment supply go far back. In
later years, controversy arose over telefax service initiated by the Bundespost
in the 1970s. The Bundespost wanted to ensure that it would play a major, if
not exclusive, role in the expected success of telefax equipment. The matter
88 Germany
came before the Administrative Council of 'he Bundespost, which was con-
scious of the potential controversy with office equipment manufacturers. The
union representatives on the board were unanimously opposed to a provision of
telefax equipment by private suppliers only and considered exclusive provision
by the Bundespost a viable alternative.
Within a short time, a rare public debate concerning the liberalization of
equipment began. In particular, the Federal Ministry of Economics championed
a liberalization, and in 1977, it opposed Bundespost participation in advanced
terminal equipment. It argued that such participation was not necessary, that it
was potentially discriminatory against its competitors which were also its sup-
pliers, and that it encouraged internal subsidies that distorted competition. The
Bundespost countered that operational experience with equipment was essential
for a network provider to be able to improve services. The Economics Ministry
pointed out, however, that the Bundespost could easily acquire equipment for
its own use and in that way gather operational experience. The Conference of
the State Economics Ministers took up this matter in 1978 and came out on the
side of the Federal Ministry of Economics. When that ministry conducted a
public "telefax-hearing" in August 1978, Siemens, the largest manufacturer,
favored a role of the Bundespost as equipment supplier.
The Ministry of Economics softened its position and demanded that (1) there
be a ceiling on market share of the DBF for telefax equipment at 15 percent
and (2) the DBF not be able to refuse its customers use of competing suppliers'
equipment. Eventually, the minister for posts and telecommunications agreed
but did not make a commitment to a specific percentage of market share, except
that it would not be a "market-dominating" position. In German cartel law
this terminology implies a share of less than 20 percent.
Equipment cases were a constant source of friction. Because much of the
terminal equipment was Bundespost owned, a slower rate of technological change
in the network and in terminal equipment was suspected, since the Bundespost
had to consider whether innovations would make its own equipment base ob-
solete. Under such an argument, network performance and design become the
prisoner of the existing terminal equipment base. But it is also true that change
in network performance could make much of the terminal equipment base ob-
solete, and that is an argument for an integrated provision of network and
terminal services.
In a study commissioned by the Monopoly Commission, three economists
Carl Christian von Weizsacker, Jurgen Miiller, and Giinter Kniepsargued that
the Bundespost's participation in terminal equipment was not a serious prob-
lem, because it already held a monopoly in telecommunications services, which
are closely complementary to equipment. Since the Bundespost did not produce
terminal equipment itself, it had an interest in seeing terminal equipment as
inexpensive and as widely distributed as possible, creating a positive contribu-
tion to its network monopoly. The terminal equipment was characterized by a
supplier oligopoly, which led to prices that were above cost. Therefore, the
participation of the Bundespost could lower prices; its large economies of scope
and scale allowed it to underprice the other market participants. Thus, the DBF
Germany 89
should actually be required to participate in the terminal market (Knieps et al.,

1981).
The Monopoly Commission decided that the question of whether economies
of scope would be more beneficial than the Bundespost's exclusion from the
market was largely an empirical one, and best resolved by making competition
a "discovery procedure" (Snow, 1983). In the telefax market, competitors noted
that the Bundespost's rental fee was only half as high as that of private firms,
allegedly because of cross-subsidization. To avoid this problem in the future,
the Monopoly Commission discussed, though it did not recommend, the isola-
tion of parts of the Bundespost's activities in separate independent subsidiaries
(DBF, 1985).
In arguing against equipment liberalization, the Bundespost claimed that the
"result of the liberalization of the terminal equipment market in the United
Stateswith the exception of the telephone handset for the simple telephone
main stationis a situation that has existed in the Federal Republic for a long
time" (DBF, 1981, p. 97a). This statement shows a lack of familiarity with
the terminal market in the United States that existed by 1980. The 1956 Hush-
a-Phone and the 1968 Carterfone decisions had made equipment connections
possible. Although AT&T attempted to impose restrictions by requiring expen-
sive buffer protection devices, the FCC overrode AT&T's rear-guard resistance
and adopted instead a system that simply defined minimum safety standards
that manufacturers had to meet.7 Under this system there is no testing or ap-
proval. Competitors have twenty days to contest a filing, and responses are due
within ten days. A ruling on such disputes, which is rare, takes generally less
than two months.
Competition in equipment supply for customer terminals had markedly pos-
itive effects in Germany. Until the mid-1970s there was no particular reason
for the Bundespost, given its monopoly position, to seek customers and to
engage in marketing for terminal equipment. As an official in the Ministry for
Post and Telecommunications noted, the demand for telephones far exceeded
the supply. In addition, "Since the telephone, by itself, was already a status
symbol, subscribers did not have particular demands concerning form, color,
or performance" (Wolf, 1982).
In the 1970s, however, economic recession, market saturation, and the emer-
gence of a gray market in terminal equipment brought a reduction in demand
for terminal equipment, leading, as the Bundespost noted, to "serious conse-
quences for the German Bundespost and its suppliers. . . . New ways had to
be pursued to move quickly from this low point. The German Bundespost tried
for the first time to reach the customer and more than in the past to respect his
desires and to fulfill them more exactly" (Wolf, 1982, p. 26). This led to
market research, an advertising campaign, and changes in the design, color,
performance, and price of telephone equipment. The surcharge for telephones
of a color other than gray was eliminated. As another Bundespost official ac-
knowledged, "Through the increasing supply of unlicensed equipment, often
with considerable defects, a certain pressure had already developed in the mar-
ket. The Deutsche Bundespost recognized this trend" (Schulte, 1982, p. 328).8
90 Germany
The Bundespost fiercely protected its equipment monopoly. At one time, its
advertising subsidiary informed potential customers of other advertising agen-
cies that the Bundespost's permission was necessary to affix a sticker contain-
ing an emergency number and an advertising message to the telephone set. The
competitors sued. The lower courts found that this behavior was an attempt to
extend the postal monopoly into the competitive realm.
This case corresponds to the American Hush-A-Phone decision, in which the
Bell System had similarly tried to prevent a trivial nonelectrical attachment to
its equipment but was rebuffed in court. In Germany, however, a 1980 court
decision accepted the prohibition of stickers on telephone equipment owned by
the Bundespost. "The prohibition of attaching stickers serves therefore the se-
curity and maintenance of an undisturbed telecommunications operation . .
(DBF, 1981, para. 76).
Despite (or because of) such defensive measures, pressures on the Bundes-
post grew. In 1981, the state economics ministers created a working commit-
tee, Deutsche Bundespost and Telecommunications Monopoly, and charged it
with proposing a modification of telecommunications equipment law. The pro-
posals asked for a DBF exclusion from equipment provision, except when tech-
nical, economic, and reasons of national economy, or other interests of society
required it. In 1982, partly because of diffuse criticism against its dual role of
supplier in competition with private firms and as the approval authority, the
DBF established a new central office for telecommunications licensing (ZZF)
with a semi-independent status and a location in Saarbriicken, at some distance
from the Bundespost establishment in Bonn and Darmstadt.
The DBF was slow to approve modems and similar devices. Until 1986 only
the DBF could sell modems. Its modem monopoly was broken after a 1985 EC
Commission intervention. Even after 1986, the DBF continued to drag its feet
on accepting the technical specifications approval of different modem suppliers.
It wanted data communication to use its public data networks. Similarly, up to
1989, the DBF would not allow standard fax boards on the telecommunications
network since it did not want the fax machines permanently connected (Pfeiffer
and Wieland, 1990, p. 84).
In 1982, the European Commission joined the chorus, issuing a complaint
against the DBF regarding modem and baseband equipment, citing it in viola-
tion of Articles 85 and 86 of the Treaty of Rome. European legal provisions
specifically, Articles 37, 85, 86, and 90have some applications to the behav-
ior of postal authorities. Article 37 requires member states of the European
Community to treat their trading and distribution monopolies so as to eliminate
potential discrimination among countries. This article has exceptions for rea-
sons of governmental function, security, and consumer protection. Article 90
applies competitive requirements to companies imbued with the public interest.
This area and the application of antimonopoly rules require much more judicial
clarification. (For an analysis, see Scherer, 1985, and Mestmacker, 1980). The
DBF subsequently modified its practices. In 1986, it also agreed that modems
could be sold by manufacturers directly to users rather than relying on the
bulky SEL or Siemens modems it had provided. These restrictions, coupled
Germany 91
with the Bundespost's own marketing, were important, since only 59,000 mo-
dems were in use in the entire country at the time, in comparison with 300,000
in the United Kingdom and 115,000 in France (Bruce, 1986). The decision had
to be made by the German Cabinet, so strong was the resistance of the Bun-
despost. The DBF, however, was left with the powers of setting standards and
approving all modems, an arrangement that could cause considerable delay.
(The firm Deutsche Fernsprecher Gesellschaft, for example, waited more than
a year before receiving approval for a modem). Furthermore, modems con-
nected to the Btx videotex service remained under Bundespost monopoly.
In 1985, the European Commission also forced the Bundespost to give up
its monopoly over cordless telephones. A Siemens cordless phone sold for $665,
whereas similar Japanese models cost $79 in the United States (Schares, 1989,
p. 137).
The American government, prodded by its industry, was critical of German
procurement practices. Negotiations over liberalization of the German telecom-
munications market reached some agreement about a reduced domestic orien-
tation by the Bundespost, as well as tenders that permitted a streamlining of
the approval process, greater ease for leased lines, and liberalized specifications
for PBXs.
As part of the 1989 reforms, Germany's equipment market, including first
sets, was officially liberalized on July 1, 1990. The Telecommunications In-
stallation Act (Fernmeldeanlagengesetz) specifies a "no harm to the network"
standard for type approvals.
Reform Debates
For a long time there was no public discussion of the Bundespost's role, in
marked contrast to the intense discussion in Germany over cable television and
media policy in general.
More than interest group politics is at work in Germany. The ideology of a
mixed economic system goes back a long time and has intellectual roots in an
important school of thought, Gemeinwirtschaftslehre (Public Economics), which
argues for government involvement in economic activities (Thiemeyer, 1983;
Snow, 1983). The Gemeinwirtschaftslehre goes back to Adolph Wagner, who,
in "Wagner's law," postulated the necessary growth of the public sector (Wag-
ner, 1887).9
Reform discussions of German postal affairs are not new; organized attempts
to study the system and recommend reforms go back at least to 1579, when
Emperor Rudolf II appointed a blue-ribbon panel (including representatives of
large postal users and financial institutions) to report on the Thurn and Taxis
postal monopoly. Four hundred years later similar questions were still being
studied. In 1969 a commission was set up to investigate the organization and
role of the Bundespost. The panel proposed the separation of political supervi-
sion from management of the Bundespost and recommended that a presiding
body of five, without the traditional civil service status, manage the PTT in the
92 Germany
manner of a private company, though with social obligations as well as finan-

cial goals. But when the government attempted even such a modest reorgani-
zation, it was blocked, partly because there was no agreement on the extent of
participation of the postal unions in the proposed governing boards. Later, pres-
sure by the postal unions prevented a reemergence of the proposal.
The general reformist spirit of the early 1970s, when the Social Democrats
held the federal chancellorship, led to the appointment of a Commission on
Communications Technology (KtK), at the initiative of the Social Democratic
government official Prof. Horst Ehmke, who happened to be simultaneously
minister of postal affairs and minister for science and technology. [It is inter-
esting to observe that in Germany, academic researchers have been more ac-
tively involved in issues of telecommunications policy than they were else-
where in Europe (Hoffmann-Riem, 1984).] The KtK was set up in 1974 to
formulate proposals for the development of a "socially desirable and econom-
ically feasible telecommunications system for the Federal Republic of Ger-
many." The commission was chaired by Eberhard Witte, an economist at the
University of Munich (Commission for the Development of the Telecommuni-
cations System, 1976). In the area of telephony, the report touched institutional
concerns only lightly; liberalization of equipment supply or the opening of
communications services to carriers other than the Bundespost were rejected.
But the KtK also examined media policy issues in cable television. As a result,
the federal states, traditionally in charge of broadcasting, initially refused to
cooperate with the committee, and political maneuvering was required to effect
a compromise.
The KtK report established general guidelines that were adopted by the gov-
ernment cabinet in 1976 and became the foundation for the subsequent estab-
lishment of pilot cable projects to test the technical, economic, and social fea-
sibility of the new cable medium. Perhaps most importantly, they led to the
creation of politically acceptable institutional models. The report also encour-
aged the development of a data network and recommended the establishment
of a planning body for telecommunications outside the postal authority. The
latter, however, was never implemented.
In 1977, the German Constitutional Court dealt a negative ruling to propo-
nents of reform by holding that the Bundespost's exclusive rights for the con-
struction and operation of telecommunications facilities covered the entire tele-
communications sector, including transmission, switching facilities, and terminal
equipment, and that it encompassed all new forms of information transmission
in network and user segments.
Nevertheless, in the early 1980s, the German Monopoly Commission, an
advisory commission to the government, launched an investigation of the Bun-
despost and issued a report on the Bundespost's role in telecommunications
(Monopoly Commission, 1981). This report was unusual because it was the
first time that the commission investigated the market power of a public enter-
prise. It traced in detail the problems of telecommunications monopoly, and
forcefully advocated liberalization in some areas. Even where it recommended
Germany 93
maintenance of the existing policy, it did so typically as the lesser of the evils,
and rarely without a fairly negative analysis of the existing system.
The leading spirit behind the report was Professor Ernst-Joachim Mest-
macker, who served at the time as the commission's chairman. He thought it a
"mercantilistic view of the state" to believe that the public interest is ade-
quately served by a monopoly enterprise simply because it is the state that
operates it (Mestmacker, 1980, p. 196).
The Monopoly Commission concluded that economies of scale were not par-
ticularly important for the connection of network and terminal equipment but
that the DBF's involvement with the terminal equipment market had negative
effects for market structure and competition. The commission also determined
that the argument of natural monopoly had to be doubted, based on the Amer-
ican experience, particularly in long-distance communications. However, it was
better, at least for a while, not to go ahead with a development of parallel
network infrastructure, given the necessary large investment. But the commis-
sion held that specialized, privately supplied value added services should be
permitted on Bundespost facilities. Direct network competition could be per-
mitted with the development of technology, in case of a potential deterioration
of DBF services, and a reduction of internal subsidies to some of the DBF's
services. Given its own scathing criticism of the existing arrangement, the com-
mission's policy conclusions were somewhat modest and probably represented
an adjustment to the political realities.
Also in 1981, the German Bundestag, in a major effort to formulate public
policy in this field, created an Inquiry Commission on New Information and
Communications Technologies. The commission had as its mandate the explo-
ration of all international aspects of the new technologies. The major political
parties appointed their communications experts to the commission. For ex-
ample, the Christian Democrat representative was Christian Schwarz-Schilling,
who soon thereafter became minister for post and telecommunications. The
parties also appointed seven nonparliamentary experts from the outside, includ-
ing Wolfgang Hoffmann-Riem, a law professor and head of the media research
center Hans-Bredow Institut in Hamburg.
The commission was active in its two years of existence. But the heavy party
politicization of media issues which is prevalent in Germany as in much of
Europe, prevented constructive proceedings. Another set of conflicts existed
between the commission and the federal states. During the commission's ten-
ure, the government changed hands, and elections were held sooner than ex-
pected. No final report could be agreed upon by the deeply divided commis-
sion. A lengthy preliminary report was supported by parts of the commission
and eventually printed as a publication of the German parliament. The report
reflects what must have been frustrating proceedings. Dissenting opinions, even
on purported factual matters, appear frequently, as do references to the lack of
agreement on many of the questions. Even on the simple statement that tech-
nologies not only provide "opportunities" but also "could lead to dangers,"
the commission could find no agreement. The commission discussed the liber-
94 Germany
alization of the postal monopoly's innovative communications servicesspe-

cifically, the Monopoly Commission recommendation that the network should
be opened up by admitting more specialized networks and giving general per-
mission to lease resale lines in order to increase competition. It listed several
arguments for such service competition, such as encouraging an improved uti-
lization of resources in telecommunications. This more efficient resource utili-
zation would not negatively affect the economies of scale of the telecommuni-
cations network, since the DBF would remain the sole network carrier. Allowing
private suppliers to explore and develop new telecommunications services not
yet offered by the DBF, including information and communications services,
was seen as improving the potential for innovation. And the Bundespost would
still be needed in the initial phases of development to subsidize new services.
The commission approvingly cited the German Constitutional Court's deci-
sion that private parties can be prevented from offering only lucrative services,
whereas the Bundespost, because of its legal obligation, must also supply less
lucrative services. The commission therefore concluded that the "Post is enti-
tled to exclude such 'cherry-picking' by private competitors" (Inquiry Com-
mission, 1984, p. 180).
The commission was also negative about competition in basic networks.
"Permitting competition would easily lead to an economically inefficient du-
plication of investment" (Inquiry Commission, 1984, p. 174). The report con-
ceded that it was empirically unclear whether telecommunication is indeed a
natural monopoly and admitted that there had been no investigations of that
question for the Federal Republic. However, empirical studies of economies of
scale for the United States, Canada, and Great Britain were cited that suggested
that the telecommunications sector as a whole was a natural monopoly, al-
though the economies of scale had decreased over time. (This literature, some
of which was commissioned in support of embattled service monopolists, is
more ambiguous than the commission's reading.)
The commission's stance on the Bundespost's role in equipment markets was
evasive. After a description of the complexity of the equipment licensing pro-
cess, the report concluded that: "Whether there are real [distorting] effects
cannot be determined" (Inquiry Commission, 1984, p. 185). The report stresses
that the network needs to be protected technologically, and users of equipment
need to be protected from malfunctioning or incompatible equipment they have
acquired.
Faced with criticism, the Bundespost began to act more flexibly and to bend
its principles quietly, as long as its monopoly was left intact. It began to permit
some resale of capacity in data lines. Resale of voice transmission was still out
of bounds, though the DBP seemed to have tolerated it in some instances.
Resale was formally limited to 50 percent of the lines' capacity, although this
limit was impossible to enforce. But criticism persisted. The free-market posi-
tion, as explained by a member of the subsequent blue-ribbon reform commis-
sion, was essentially that "Reputable expert opinions commissioned by the
Bundespost itself lead to the conclusion that the state-run enterprise also earns
a rather mediocre grade in an international comparison. . . . It pointed out
Germany 95
that the Federal Republic lagged far behind in the number of public telephones,
with 162,000 compared with the 1.7 million in the US and 828,000 in Japan,
and that it had for a long time no itemized billing, cordless phones, call-wait-
ing, call-forwarding, etc." (Moschel, 1988, p. 3).
In 1984, the government published another report to serve as a conceptual
document for the advancement of information technology. The report, Regi-
erungsbericht Informationstechnik 1984, stressed the importance of govern-
mental initiative in order to keep German information technology competitive
(Lange, 1984).
In the 1980s, even the Social Democratic Party, which had faithfully sup-
ported the Bundespost monopoly, began to criticize its rate of upgrading of the
network. Its General Secretary, Peter Glotz, himself a media expert, criticized
the Bundespost's low level of investment in digital switching (Glotz, 1983),
and urged a rapid development of the Bundespost networks. As Glotz asserted,
if no modern public networks would be offered, private networks would emerge
on leased lines, which would provide the same efficiency benefits, but would
be less sensitive to social politics and other requirements such as data protec-
tion.
The Reform of German Telecommunications
Partly in response to pressure for liberalization from the United States and the
European Community, a commission was established in 1985 to make recom-
mendations for the reform of telecommunications policy. The panel, chaired
by Eberhard Witte of Munich University, who had also headed the KtK, was
composed of twelve members who represented the four major parties and a
spectrum of economic interests. From the beginning, the commission was sub-
ject to criticism. To some it was a threat to the established order; to others it
was a legitimization of minimalist reform steps which largely served the Bun-
despost's interests. It took almost one year to set up the panel, and another two
and one-half years until it produced its report in September 1987.
Internally, the commission was split in various ways. The free market or
liberal wing consisted of the Free Democratic Party representatives, a law pro-
fessor, the banks' representative, and the president of the Federal Associate of
German Industry. In the past, the telecommunications industry interests had
generally defined the position of the German industry association. The orien-
tation of the association's president was thus significant, because it indicated
that users and other electronic firms' interests had gained in importance and
influence.
The traditionalist wing consisted of the trade unionist, the Social Democratic
delegate, and in key questions, the representative of the Bavarian Christian
Democrats. That conservative party favored the status quo, because it wished
to satisfy its large rural population, and perhaps because Munich was the head-
quarters of Siemens, which also did not wish to see an upheaval in well-estab-
lished relations.
96 Germany
The commission was split six to six over the key question of whether to
permit competition in physical networks. Chairman Witte, who had two votes,
elected not to break the tie, which was likely to have been in favor of network
competition, for pragmatic reasons: Even if a majority had recommended such
competition, the minority would have included the representatives of three of
the four major parties, representing more than 80 percent of the parliamentary
deputies. Political realities had to be taken into account if the recommendations
were to be adopted. The economically liberal opponents believed that the rec-
ommendations had to reflect what was best, not just what was most realistic
politically. The approach ultimately taken by the commission majority on this
issue was to initiate a process and then to expect its dynamics to lead in future
development. Thus, it recommended a number of modest pro-competitive steps
that would, in time, its majority hoped, obtain a momentum of their own and
lead to a more significant change in German telecommunications.
It is important to note that the commission found that the existing telecom-
munications monopoly was not enshrined in, or protected by, the German Basic
Law. Structurally, it recommended a separation of telecommunications from
postal services, and a separation of "sovereign tasks," or regulation, from the
entrepreneurial tasks of organization. It recommended that regulatory tasks be
vested in the ministry and others reside with operational organization Deutsche
Telekom, which would have greater operational and legal independence. Deutsche
Telekom would also have greater flexibility in compensation, payment meth-
ods, and deployment of labor. Deutsche Telekom would retain its monopoly
on the basic network and on voice telephony. But data, text, and other non-
voice communications would be open to competition from the private sector,
using the Telekom network. It would also lose its monopoly on the basic tele-
phone set, and the terminal equipment market would be fully liberalized. The
commission also proposed that tariffs be set by the Ministry of Communica-
tions and be more in line with costs, thus eliminating tariff distortions. Greater
freedom would be granted in the interconnection of leased lines with switched
service. In a departure from existing controversial practice for leased lines,
Telekom would develop a tariff structure not based on volume-sensitive charges.
Satellite-delivered slow-speed data would be opened to transmission by others,
another first.
The commission also recommended that the government reevaluate market
conditions every three years, with the option of permitting the establishment of
competing networks if the market was not developing satisfactorily (Witte, 1987).
The commission thus provided, at least rudimentally, for a previously missing
mechanism of policy change.
Most of the recommendations were accepted by PTT Minister Schwartz-
Schilling and by the government in a somewhat more restrictive bill. It rejected
a conditional DBF monopoly but allowed the following: value-added services,
a second cellular network, usage-sensitive leased-line tariffs, liberalized termi-
nal equipment, liberalized type approval, slow-speed satellite service for data,
and limited resale. It imposed special obligations on private carriers (Moschel,
1988, p. 17). But the new telecommunications organization was not to be fully
Germany 97
freed under its own board of directors, and its name was slightly but pointedly
changed to DBF Telekomindicating its ties to the Bundepost. It must con-
tribute 10 percent of income to the Federal budget. (The total that the DBF
paid to the government through its annual contribution in 1988 amounted to
5.25 billion DM, approximately 10 percent of its annual income). After the
reforms, the three different sectors of the DBF are subject to taxes, as if they
were private enterprises. Furthermore, an "infrastructure council" with mixed
state-federal membership was established to respond to the desire of several
states to have input into telecommunications policy and to address their fear of
neglect by a profit-oriented Telekom.
After bitter debates and opposition by the trade unions, the Law on Restruc-
turing the Bundespost passed on July 1, 1989.
The objective of the 1989 reform was to eliminate the structural conflict
between the regulatory and the production activities of the DBF and to encour-
age participation and competition from the private sector for new services and
hardware. To accomplish this, a separation of responsibilities was mandated.
Managerial responsibilities over the PTT were taken away from the Ministry of
Posts and Telecommunications, so that the ministry could in the future concen-
trate on regulatory issues without any conflict of interest. The ministry, which
would retain cabinet representation, would now set and implement standards,
establish access rules, allocate spectrum, license entrants, and so on.
The 1989 Reform (the Poststrukturgesetz) created three distinct public cor-
porations, for telecommunications, for postal service, and for the PTT's tradi-
tional financial services. Each of these corporations would have its own board
of directors, its own published balance statements, and its own organizational
structure. Further, these new entities would have the legal responsibilities of
private corporations to their customers and suppliers, rather than their previous
protection of being part of the German state.
The chairpersons of each of these distinct entities meet in a single Deutsche
Bundespost directorate, which is responsible for consolidating the individual
balance sheets of the different PTT entities, coordinating service between the
different groups, and handling joint labor compensation issues. Each of these
entities, however, was still required to make a profit on its own.
The staff members of the new operating entities are still civil servants, and
as such retain their traditional protections from dismissal. It remains difficult
to reward employees for merit.
The telecommunications entity, DBF Telekom, retained exclusive right to all
voice transmission. Since this service amounts to 90 percent of all the telecom-
munications business in Germany, it is still in control of the most important
and profitable sector. Further, in part because of monopoly conditions, it can
charge above cost for these services. The rationale for allowing DBF Telekom
such profit was that it would be able to invest this money into infrastructural
changes, including fiber optics and to cross-subsidize service to more remote
regions in the country.
No alternative physical network can be set up except under very special
circumstances (internal use of public administration or transportation concerns,
98 Germany
or between singly owned parcels of land no more than 25 kilometers away from
any of the other parcels of land). Many users complained that the new law still
allowed DBF Telekom to move slowly in servicing their needs. In response,
Schwartz-Schilling drafted stronger legislation in 1990 aimed at reducing DBF
Telekom's ability to block competitors with limited access and high tariffs.
Competition for nonvoice transmission began shortly thereafter. Germany
already had 185,000 cellular subscribers in 1989, connected to .DBF's then
monopoly service. But in May 1989 the government solicited bids for the con-
struction of a second (D2) cellular network, based on the European GSM stan-
dard. Several American firmsNYNEX (with BT and Daimler-Benz), Bell
South (with Siemens, Olivetti, and Shearson-Lehman-Hutton), and Pacific Te-
lesis (with Mannesmann and Deutsche Genossenschaftbank)sought licenses.
Mannesmann, a large steel and engineering conglomerate, led the winning con-
sortium, but its service was delayed by a year of negotiations with DBF Tele-
kom over interconnection and carriage tariffs. A third and digital, in principle,
mobile telephone network was decided upon in 1991.
In 1991, the ministry granted 22 companies licenses for private satellite net-
works that did not interconnect with the public network. Services above 15
kbps required special approval. Most systems are, at least in part, over the 15
kbps threshold. Licenses were also awarded for paging and radio data.
Although it is important to note the DBF's new services, a perspective must
be kept. Most of the DBF's telecommunication revenues derived from tradi-
tional telephone service. Of the DM 38.4 billion of telecommunications reve-
nue, traditional telephone service made up DM 34.2 billion, or 90 percent.
Data services (including videotex) and telegraph were DM 2.67 billion (7 per-
cent), and radio and other telecommunications sources were DM 1.5 billion.
Of the data service figure, 40 percent was earned by leased lines (Pfeiffer and
Wieland, 1990, p. 11), until rates were drastically rebalanced.
East GermanyGerman Democratic Republic
Despite the vast differences in political systems, East Germany's telecommu-

nications operations were organized in some ways similarly to those of West
Germany prior to the latter's 1989 reform. Telecommunications were provided
by a monopoly PTT "Deutsche Post" (DP), which had a standard two-part
structure of Directorates and regional offices. The PTT was subject to the Min-
istry of Posts and Telecommunications, or "MPF," which controlled central
strategic and regulatory offices and exercised management, coordination and
planning functions over the full range of communications technologies.
In the period immediately prior to unification, the Ministry dominated a ver-
tically integrated structure of state owned companies that provided telecom-
munications services, manufactured telecommunications equipment, and in-
stalled cable, switching and terminal equipment. This structure included two
classes of entities. The PTT directly owned the VEB Fernmeldebau, which
provided installation services. (A VEB or "volkseigner Betrieb" was a state-
Germany 99
owned company of the type that predominated in the German Democratic Re-
public's centralized economy.) The MPF also controlled the "Kombinat Na-
chrichtenelektronik," a state cartel that organized and directed all fourteen of
the VEBs that manufactured telecommunications equipment. In addition, the
Minister headed a government commission that coordinated the various state-
owned networks and directed their interconnection with the public network, and
the MPF controlled a public frequency commission that oversaw the allocation
of radio spectrum space.
In addition to its telecommunications services, the DP provided extensive
postal services, such as distribution, billing and marketing for all newspapers
and magazines. It also was responsible for radio and television receivers and
transmission stations, as well as all studio production equipment. The PTT's
control over public communication was so extensive that if individuals needed
a microphone for any public purpose, they had to rent it from the PTT (Neu-
mann, 1990).
Before 1989, the telecommunications equipment industry had been under the
control of a manufacturing industry ministry. Under that arrangement, the PTT
was forced to report its investment requirements to a central planning bureau-
cracy that allocated resources based on its own political priorities instead of on
the needs of the telecommunications sector. Telecommunications services were
considered a nonproductive part of the economy and were thus assigned a low
priority. Conversely, equipment export was a high priority, and approximately
75% of all telecommunications goods were sent abroad, especially to other East
European countries. In 1989, the telecommunications industry was put under
the control of the MPF in a controversial effort to improve economic perfor-
mance and turn its productive capacity to domestic uses (Neumann, 1990).
The 1985 Law on Posts and Telecommunications gave the state the exclusive
right to provide postal and telecommunications services. While some leased
line networks were operated by businesses that had obtained waivers to do so
from the MPF, the only nonpublic telecommunications systems were those op-
erated by the army and the police. All terminal equipment was provided by the
PTT, except for a small number of PBXs that received a special waiver. There
was no private telecommunications equipment market.
Although the GDR had one of the most advanced telephone systems in the
Eastern Bloc, the infrastructure was inadequate and outdated. There were only
very basic services. In 1965, there were 1.6 million telephones, with a waiting
list of 84,000 people. The waiting list for phones grew to 660,000 in 1983 (A.
Rutkowski, 1990, communication; Kelly, 1990, Figure 2). By 1990, there were
1.8 million main exchange lines, twice as many as 1969, serving 4 million
telephones. Nevertheless, only one out of seven households had a phone, and
party lines were common. Moreover, the few residential phones available were
unevenly distributed. In East Berlin, 50 per cent of homes had service, while
in other cities, such as Dresden, only one out of nine homes had service. By
1990, the official waiting list was approximately 1.2 million people (almost as
many people as had telephones), and was growing twice as fast as installed
main lines. At the previously planned rate of construction, the backlog would
100 Germany
have taken at least ten to fourteen years to fill (U.S. Department of State, 1990,
p. 51); some estimates suggest as much as a twenty year delay. Moreover, as
in many Eastern European countries, the list probably did not include discour-
aged potential users.
Telegrams were a significant form of public communication. In 1965, 8 mil-
lion domestic and 3 million international telegrams were sent from 6000 telex
connections. In 1989, those figures were 11 million and 3 million telegrams,
respectively, and 18,000 telex connections.
The East German network structure prior to unification was composed of
1,500 local networks, 2,700 local switching centers, and 182 switching centers
at the trunk level. While almost all of the main exchange lines were connected
to automatic exchanges, just 20 per cent could be switched automatically for
international calls. Almost a quarter of the local and trunk switches were an-
cient, installed between 1922 and 1934, while 42.6 per cent were installed in
1950. Another 28.1 per cent were installed between 1963 and 1965. Part of
the switching system was so old that it had already been fully depreciated more
than twice. At the local and trunk levels, the switches were analog and electro-
mechanical; 25 per cent of all East German switches were crossbar. Most of
the transmission system was analog, although since the mid-seventies PCM-
systems were used. Fiber optic cable amounted to less than 1 percent of the
transmission network.
The network could barely support facsimile and data transmission, and there
was neither a circuit-switched nor packet-switched data network. The first mo-
dem use was reported in 1969 when eight modems were put on line. By 1983
there were 500. There was an operator-assisted data network that connected
1,500 subscribers, and another 3000 subscribers leased lines for data transmis-
sion. (The PTT had planned to build a packet-switched network since the
mid-eighties and had signed an agreement with Siemens to do so, but CoCom
restrictions on the packet-switching technology prevented the project.) As a
result, only a few hundred fax machines and modems were in use before uni-
fication, while over 10,000 prospective subscribers were on the waiting list for
data connections. Mobile and value added services were unavailable. There
were 18,000 telex connections, transmitting text at 50 bps over thirty to forty
year-old technology, and 3,000 data circuits with a capacity of 2.5 kbps (Gart-
ner and Habenicht, 1990, p. 13). The only well-developed part of the telecom-
munications network was that operated by the GDR's secret police for surveil-
lance and domestic control (Hafner, 1990).
The German political unification in 1990 created an urgent need to upgrade
East Germany's antiquated infrastructure. In December, 1989, a Joint Govern-
ment Commission was created by the East and West German governments to
manage the process of unification for all aspects of telecommunications. The
government of Prime Minister Hans Modrow, still a Communist party official,
decided to shift the PTT toward financial independence from the government
and make it a public enterprise. By early 1990, the MPF concluded that it
would move toward the West German organizational structure in an effort to
secure necessary support. In March 1990, a non-Communist East German gov-
Germany 101
ernment was elected. Soon thereafter, it replaced the old PTT structure with
three public enterprises along the same lines as West Germany.
Simultaneous with the negotiations on economic and currency union, the
East and West German Ministers of Posts and Telecommunications negotiated
a Joint Declaration that outlined the plan for integration. Targets were for the
East German network to be on par with West Germany's by 1997. DBF Tele-
kom planned to install 7.3 million new subscriber lines, 10 million miles of
fiber-optic and copper cable, 68,000 public phones, 360,000 facsimile connec-
tions, and 50,000 packet-switched connections by that time, and add 2000 dig-
ital central office switches. The newly united PTT planned to add 100,000 East
German subscribers in 1990, 300,000 in 1991 and over a million annually
thereafter. In contrast, the growth rate before unification had been approxi-
mately 60,000 subscribers per year. In addition, DBP Telekom, which needed
to rely on satellite and cellular services to accommodate the extra load as the
landlines were being installed, planned to add 300,000 mobile phone connec-
tions. The quick availability of cellular service relieved some of the pressure
on the standard network, which experienced a 500 percent increase in traffic
within eight weeks following unification (Hafner, 1990; Gartner and Habenicht,
1990).
Before unification, consideration was very briefly given to the possibility of
privatizing the PTT. While such a move could be done fairly straightforwardly
through legislation (as opposed to the situation in West Germany, where pri-
vatizing the DBP would probably require a constitutional amendment), this
option was not pursued. The interest in unifying the East and West German
systems took precedence. Instead, the Eastern DP was merged into the Western
entity. Almost all employees were retained (the exceptions were primarily the
numerous secret police workers within the PTT whose major function was to
wiretap calls and open letters).
Another option was to operate the East German network as a separate re-
gional company interconnected with the Western DBP Telekom. Such an ar-
rangement would have made the transfer of investment funds and expertise
from the West to the East more difficult, though creative alternatives could
have been devised. But the tradition of national monopoly was too strong for
such an option to be seriously considered.
The merger of East Germany's telephone system into DBP Telekom was not
a smooth one. Soon, East Germany's 130,000 postal and telecommunications
employees went on strike for a more rapid equalization of pay levels. DBP
Telekom, meanwhile, was asked by the government to contribute an additional
1.3 billion annually to the federal budget for four years, to help pay for the
general cost of unification. This was beyond the substantial amounts that were
needed to upgrade East Germany's telecommunications. It was also difficult to
move West German technical employees to the East, where their skills in dig-
ital communications were needed. Another problem was how to integrate the
two dozen East German private networks, a legacy of the GDR's separate min-
isterial and state industries' systems, into the national system. One of them was
acquired by the West German chemical industry for its own use. Service up-
102 Germany
grade in the East was slow, leading to some calls in the East for the entry of
alternative service providers. In an effort to speed up development, the govern-
ment initiated turnkey projects in which SEL, Bosch, DeTeWe, Deutsche
Aerospace AG (Daimler-Benz), and Siemens would construct full-scale local
networks in various cities. These substantial burdens gave ammunition to those
traditionalists in the Bundepost and the Ministry who had never been happy
about liberalization, and who could argue for slowing down its pace. Indeed,
had German unification burst upon the scene only one year earlier than it did,
it is probable that the entire Bundespost reform would have been put on hold.
The Monopoly Commission in Bonn, on the other hand, looking at the same
set of problems, advocated the relaxation of monopoly to accelerate the devel-
opment in the East. Thus, the old controversies continued.
In 1991, DBF Telekom invested about $4 billion in eastern Germany, and
anticipated the need for an additional $100 billion before the year 2000. These
huge financial demands took place at the same time that revenue-eroding lib-
eralization was being introduced by the telecommunications ministry for ter-
minal equipment, value-added and mobile services, interconnection of private
leased-line networks, satellite data, and even resale and switching. The com-
pany was also criticized for having some of Europe's highest rates for leased
lines.
DBF Telekom, run by Helmut Ricke, a manager hired from outside the or-
ganization, responded with a massive rebalancing of tariffs, systematically re-
ducing leased-line and long-distance rates, and raising rates for local service.
It moved into value-added services, forming a joint venture with IBM. Still,
the company was limited by law from certain activities, for example, directly
offering service in the newly opened countries of Eastern and Central Europe.
In consequence, the government in 1992 gave serious consideration to changing
DBF Telekom's status, and privatizing up to 49 percent of its shares through
an international offering. Such plans, unthinkable only a few years earlier,
demonstrate how far the dynamics of change have taken hold in transforming
Heinrich von Stefan's monolithic state instutition.
8
The United Kingdom
The Reorganization of British Telecommunications
The early history of British telecommunications, up to their nationalization in

1911, is provided in Chapter 2.
In the years following nationalization, the telephone system developed only
very slowly. In 1912, the first public automatic exchange was opened; in 1916
amplification for long distance service was introduced; the Strowger switching
system was adopted in 1922 and remained the mainstay well into the 1980s
(Foreman-Peck, 1985, pp. 215-28). Demands for network investment were
made in 1920 by a Parliamentary Select Committee, the Telephone Develop-
ment Association (formed in 1924), and the Liberal party. But the General Post
Office, the Treasury, and the Union of Postal Workers (though not the technical
unions) resisted change until a 1932 Act of Parliament demanded reorganiza-
tion. An opposition group, the "Memorialists," was formed to press for tele-
communication reform. It had the support of 320 MPs. A commission was set
up (the Bridgeman Inquiry) that recommended moderate change. But the Post
Office maintained its inefficient structure; even in the face of a substandard
system, the postmaster general declared that he could not see how additional
capital could be more usefully spent.
A structural reorganization of telecommunications was accomplished only in
1969. It transformed the General Post Office from a governmental department
subject to Treasury and parliamentary interventions to a more autonomous pub-
lic corporation. But it also reaffirmed the Post Office monopoly over telecom-
munications. While Post Office autonomy was the primary objective of the
1969 Act, it remained an elusive goal because of the frequent interventions of
the Conservative government that came to power shortly thereafter. Govern-
ment cutbacks in capital investment as well as limitations on telecommunica-
tions tariffs forced the Post Office to operate under increasing financial strains,
and to implement unpopular rate increases in 1975 (Morgan, 1987).
During that time, the Post Office's level of service also attracted strong crit-
icism. In response, the government in 1977 commissioned another panel to
propose reforms (Littlechild, 1983b). These recommendationsthe Carter Re-
portwere relatively modest: to separate the Post Office into its postal and
telecommunications parts, with both still operating as monopolies. The Post
Office brushed off the proposals, feeling unthreatened. But times had changed.
103
104 The United Kingdom
The traditional equipment industry-labor-rural coalition could not prevail any-

more against Prime Minister Margaret Thatcher and her inner circle's determi-
nation to change the system after their electoral victory in 1979 (Locksley,
1983).
Under Thatcher, who dominated British politics through the 1980s until her
fall in 1990, the right wing of the Conservative party, rather than the more
moderate followers of former Prime Minister Heath, set economic policy. Their
economic perspective was one of classic economic liberalism. They were alarmed
by the decline of the British electronics industry; for example, Britain's share
in telecommunications equipment trade had fallen from 25 percent in 1960 to
5 percent in 1980. In pursuit of economic growth, the government sought a
major redirection of British industry from traditional and stagnating industries
to those with a strong knowledge baseelectronics, information, biotechnol-
ogyall closer to the white-collar interests of Tory followers than to the tra-
ditional smokestack firms with their unionized workforce.
There were three distinct elements to the Conservative government's tele-
communications policy: (1) a separation of telecommunications from the Post
Office; (2) a liberalization of terminal equipment, service competition, and en-
hanced services; and (3) the privatization of the public network provider.
In 1980, Sir Keith Joseph, a member of Thatcher's inner circle, announced
to Parliament the government's intention to end the monopoly on telecommu-
nications. The government soon presented a far-reaching bill that became the
British Telecommunications Act 1981. The Act separated telecommunications
from the Post Office and established a new state-owned corporation, British
Telecom (BT), to operate telecommunications. The Act also ended the near
monopoly over terminal equipment. Before 1981, the Post Office supplied,
leased, and maintained nearly all telephone equipment, with the exception of
PBXs of more than 100 lines. The 1981 Act maintained the supply monopoly
for the first telephone set only, although that set could now be privately ser-
viced. Terminal equipment still required technical approval from the Depart-
ment of Industry, which delegated standard setting authority to the British Stan-
dards Institution (BSI). The Secretary of State for Industry appointed the Brit-
ish Electrotechnical Approvals Board (BEAB), which set up a laboratory, the
British Approvals Board for Telecommunications (BABT), to undertake testing
to determine whether consumer apparatus met BSI standards.
The 1981 Act also permitted either the Secretary of State for Industry in
consultation with BT, or BT itself, to grant licenses for competitive networks.
Almost immediately, and with active government prodding, a consortium in-
cluding the telecommunications company Cable and Wireless (C&W), Bar-
clay's Merchant Bank, and British Petroleum (BP) founded Mercury Commu-
nications as the first alternative carrier for long-distance network services. In
time, Mercury became fully owned by C&W.
Other changes rapidly followed the Telecommunications Act 1981. The gov-
ernment decided to sell a majority of BT's shares, and BT itself was internally
restructured as a business rather than as a civil service administration. In a
White Paper announcing its intention to privatize BT, the government argued
The United Kingdom 105
that private management would increase efficiency and permit easier access to
capital markets. Run as a public corporation since 1969, BT was subject to
restrictive rules on borrowing, and had to self-finance approximately 90 percent
of its investments, an extremely high rate (Department of Industry, 1982). By
1982, after two years of energetic prodding, BT's per capita investment was
still only two-thirds that of France's and just over half that of West Germany,
and revenue per employee was $4200, compared with $4900 in France, $6200
in West Germany and $7900 in the United States (Journal of Commerce, 1983).
A second telecommunications bill was introduced in 1982 to implement the
privatization goals of the White Paper. Although the bill was not immediately
passed because of the general election, it was reintroduced after the Conserva-
tives formed the new government.
As part of the privatization terms, the government forgave or assumed 2.9
billion of BT's long-term debts and shouldered pension obligations of 1.25
billion. Together, 4.15 billion of liabilities were taken off BT's books, a huge
amount in relation to the 4 billion for which 51 percent of the entire enterprise
was sold.
One of the major obstacles to privatizing BT was its sheer size. The sale of
BT for about 4 billion accounted for 30 percent more than the total amount
raised for all U.K. companies in 1983, itself a ten-year high. It was therefore
feared that the floating of BT shares would disrupt the capital markets. This
led briefly to the suggestion of an AT&T-style divestiture whereby BT would
be sold in parts, either by functionlocal, long-distance, international, and
business servicesor along geographic lines, or both. Prime Minister Thatcher
allegedly favored this idea for a brief time. In the end, 51 percent of an undi-
vided BT was offered to the public. To assure the success of the issue, the
government chose a fairly low share price. Opponents argued that the Tory
strategy was to make it expensive and unpopular for the next government to
renationalize BT. The Labour opposition called for a renationalization, to be
financed by an exchange of shares for nonvoting bonds; their initial price would
be the stock issuing price and would thus not reflect appreciation of shares.
And an appreciation there was. On the first day of its public offering, the
price of BT's shares almost doubled. But the success of the public offering also
had its cost in terms of subsequent policy flexibility, because it created a large
constituency opposed to deregulatory actions which might reduce BT's profit-
ability.
Prior to privatization, British Telecom had almost a quarter of a million
employees, most of whom belonged to one of BT's several labor unions. The
largest of these was the National Communications Union (NCU)formerly the
Post Office Engineering Union (POEU); others include the Union of Com-
munications Workers, the Society of Telecom Executives, and several smaller
unions.
BT's unions strongly opposed privatization and competition. Their opposi-
tion was based on fears that labor would bear the brunt of the 25 percent cost
savings that BT's chairman, Sir George Jefferson, publicly projected for the
first three-year period. The unions also felt that opening the market to compel-
ing networks was a step toward eroding BT's monopoly position and profits,
in which they indirectly shared. They also stressed the negative impact on res-
idential service and R&D and feared a change in their civil servant status.
Consequently, the unions fought the licensing of a competing carrier and staged
work protests against the new company. Union fears proved not unfounded. In
its first year as a private company, BT reduced its staff by 17,000 employees,
mostly by attrition. However, although the number of BT's employees de-
clined, its total labor costs increased (Oftel, 1985a).
BT's unions argued that the government should reestablish the monopoly
over the trunk network and the first telephone set, integrate Mercury into BT
at market value, and, most importantly, maintain control of 75 percent of the
company's shares. They also advocated filling 50 percent of BT's board with
employee representatives, with other directors coming from consumer organi-
zations, other industries, major users, and the government.
Oftel: The New Regulatory Framework
It would be simplistic to believe that the privatization of BT alone would make

it more efficient and responsive to its customers. A public monopoly is subject
to political supervision, and therefore potentially more accountable than an un-
restricted private monopoly. Hence, in the absence of a competitive environ-
ment, a regulatory mechanism is essential to complement privatization. Ac-
cordingly, the government established a regulatory body to perform that function,
and named it the Office of Telecommunications (Oftel).
The government also remained BT's largest shareholder, though it an-
nounced in 1991 further privatization of its shares. Though the Thatcher gov-
ernment promised not to assert its right as a shareholder in commercial deci-
sions or even to vote at shareholder meetings, this promise does not bind future
governments. Furthermore, the government can affect BT's actions through the
Department of Trade and Industry and the Monopoly and Mergers Commis-
sion.
BT is also obligated under its license to provide universal telephone service,
a public emergency system, and public coin telephones where they previously
existed. It also must permit terminal equipment to be connected to its network,
provided that such equipment conforms to technical standards.
Oftel is headed by a director general of telecommunications (DGT), who is
appointed by the secretary of state for industry for a five-year term of office. It
resembles the Federal Communications Commission in the United States but
lacks the FCC's formal independent status and its power over broadcast com-
munications. In 1988, Oftel's staff numbered 116 in comparison with the FCC's
1900, to which one should also add the American state public utility commis-
sions' considerable personnel (The New York and California commissions, which
are the largest, have more than 600 persons, though they are responsible for
regulating other utilities as well. The FCC also regulates broadcasting.)
Unlike the commissions in the United States or Canada, Oftel has no direct
authority to set rates and can only make recommendations. However, almost
from the beginning Oftel has expanded the scope of its investigation into rate
issues under its mandate to assure the provision of "good services to consum-
ers" and to insure that BT does not abuse its monopoly power.
Oftel supervises existing licensees and makes recommendations on applica-
tions for new licenses. Licensees include British Telecom; Hull Telephone
Company, the only independent local telephone company; Mercury, the new
long-distance carrier; Cable & Wireless, its parent, with many international
involvements; value-added networks; Vodaphone and Cellnet, the cellular ser-
vice companies; and other telecommunication service providers.
Despite Oftel's activity, the Department of Trade and Industry has retained
significant authority over telecommunications, including spectrum allocations,
cable television, technical licensed satellites, and resale of BT capacity (S.
Littlechild, communication, 1989). Customer and manufacturer concerns are
also channeled to Oftel through Advisory Committees on Telecommunications
in England, Scotland, and Wales (Manning, 1988).
Consumer and advocacy organizations have criticized Oftel's powers to pro-
tect consumers as inadequate, claiming that Oftel can only require BT or other
communications carriers to respond to grievances. Actually, Oftel can also sue
in court or alter and amend the operating license; and although Oftel's power
to protect consumers directly may be limited, they are far larger than those of
the earlier consumer protection body, POUNC, which had a primarily consul-
tative role. But Oftel's procedure is not open to the public when it comes to
information. British Telecom, for example, was not required to make public
the information on revenues, profits, costs, and quality performance that was
provided to Oftel for a determination of its price formula.
Bryan Carsberg, a professor of accounting at the London School of Econom-
ics, was appointed as the first director general of Oftel in 1984.' Carsberg
established a more active supervision over the telecommunications industry than
many expected, helping, for example, to prevent a venture between BT and
IBM on value-added networks, forcing an interconnection policy on BT that
was favorable to Mercury, and requiring BT to change its accounting system
to make it more possible to detect cross-subsidization. However, Oftel drew
criticism when it involved itself in BT's procurement practices. Carsberg rec-
ommended that BT's purchase of digital exchanges from a second source
("System X") be limited to 20 percent, for at least three years. This looked
like a protectionist measure that favored the "System X" of the British firms
GEC and Plessey. In public, BT rejected Carsberg's "recommendation" for a
voluntary purchase quota, though as a practical political matter it went along
with the allocation.
One important initial decision involved the basic choice of regulatory tech-
nique (Littlechild, 1983a). An interministerial working group recommended
linking the BT license to a maximum rate of return on capital and a specific
rate of return for the local, long-distance, and international services, with a
share of excess profits returned to consumers. As long as the rates of return
were not exceeded, specific prices would not be subject to control. The U.S.
experience with rate-of-return regulation had demonstrated the problems of this

method, including the difficulty of calculating the rate base and allowable ex-
penses; incentives for overcapitalization (gold-plating), known as the Averch-
Johnson effect; disincentives against risky investment, since they are not re-
warded by a higher rate-of-return; inadequate incentives for efficient and lean
operations in general; and absence of price control within broad rate of return
categories. The chief economic advisor to the prime minister, Professor A. A.
Walters, advocated an alternative incentive mechanism to keep prices low and
raise output, under which regulators would project a five year revenue growth
rate. Overfulfillment of this target would lead to a lower company tax rate,
while underfulfillment would lead to a higher rate, and therefore to a penalty.
Although Walter's method included no price control and made the government
a de facto partner in monopoly prices, it reduced the incentive for monopolistic
restrictions of output. It remains questionable, however, whether the adminis-
trative problems of this method would be less complex than those of rate-of-
return regulations. Profits need to be calculated, which is always a slippery
matter. Still more difficult to define is output, particularly in an industry where
the number of services and their quality dimensions are large, complex, and
changing.
Professor Stephen Littlechild of Birmingham University suggested as a third
option substituting partial control over prices for control over profits. For those
services in which BT was dominant, tariffs would be increased only by the rate
of inflation, minus a specified percentage. Hence, a real reduction in telephone
rates would be built into the system, based on the expectation of advancing
technology and productivity. The Littlechild approach can be a useful transi-
tional technique, since it is simpler to handle and implement than the other
proposals. However, it is unlikely that the pricing of monopoly telecommuni-
cations services could continue for an extended period of time to be based on
long-past prices. Any evaluation of the price formula was likely to involve
reference to profitability (and eventually did). Another problem is the need to
regulate service quality, since BT might seek to cut costs and raise profits
through service quality deterioration. Furthermore, what should be the correct
productivity offset? If it is too low, BT reaps a windfall; if it is too much, it
will suffer financially. Littlechild prudently did not wish to resolve all of these
questions, but proposed a five year transitional period for his method, after
which the entire system would be reconsidered.
In 1983, the government, having weighed the various possibilities, accepted
the recommendation. It decided eventually on a formula for overall changes, a
"tariff basket" based on the retail price index (RPI) inflation rate, minus 3
percent annually for the productivity offset. Furthermore, BT assured that rates
for consumers would not rise by more than 2 percent above inflation. Subse-
quently, the pricing formula was incorporated into BT's license.2
Oftel modified the price formula in 1989, raising the productivity offset from
3 percent to 4.5 percent. In discussing the proper RPI discount, Oftel tacitly
admitted that price cap regulation was predicated upon rate of return calcula-
tions. BT's rate of return was consistently above the RPI from 1983 to 1989:
Year Rate (%)

1983 19.3
1984 17.5
1985 19.2
1986 20.2
1987 21.4
1988 21.3
1989 21.6
Oftel also redefined the basket to include residential, business, domestic long-
distance, and directory assistance services, and forced BT to introduce a re-
duced tariff for infrequent users. BT was able to retain its RPI + 2 limit on
installation connection and rental charges. Critics of BT argued that rates had
simply been lowered where competition from Mercury existed and raised for
other customers. Local calls had risen 35 percent from 1984 to 1986 and peak
rate trunk calls fell 32 percent in the same period (Gist, 1990, p. 47; Dixon,
1990b). In 1990, BT raised line rental charges 11 percent and introduced cus-
tomized tariffs.
The Changing Network Environment
CaWe and Wireless
With British liberalization, Cable & Wireless (C&W) became one of the most
interesting telecommunications carriers in the world. Until the 1980s, it was a
sleepy postcolonial governmental enterprise. But thereafter it became the first
truly global telephone company, linking the world's major trading centers.
C&W, which was nationalized in 1947, used to operate Britain's interna-
tional telegraph service as well as telecommunications services in many of Brit-
ain's overseas possessions. After decolonialization, the company continued to
operate domestic public telecommunication services in more than two dozen
countries, and the international communications of 37 countries. Many of C&W's
overseas operations were joint ventures with local governments or local private
interests.
In 1981, the Conservative government privatized more than half of the com-
pany; in 1985, the remainder of C&W was sold in the second largest stock sale
in British history after the BT sale earlier that year.
Privatization made it possible for C&W to expand rapidly and aggressively.
As a government company it needed Treasury approval to spend more than 10
million, but as a private firm it could freely invest in new projects. C&W's
profits increased rapidly from 90 million on sales of 500 million in 1981, to
360 million on revenues of 1.3 billion in 1985 and 500 million in revenues
of 2.2 billion in 1990.
The centerpiece of C&W's operations and profits is in Hong Kong, where it
operates the franchise for international telephony and holds a controlling inter-
est in the Hong Kong Telephone Company, of which it sold a 20 percent share
in 1989. The firm also has an agreement with the People's Republic of China
to upgrade the Chinese telecommunications network in the nearby provinces,
in Beijing, and in the Yangtze Delta, and has a 49 percent interest in the
Shenda telephone company in a Chinese special economic zone next to Hong
Kong. C&W also owns 75 percent of telephone operations in Macao and op-
erates international and domestic service in Bahrain, the commercial center of
the Persian Gulf region, through a firm in which it has a 40 percent interest.
In 1990, C&W had operations in forty-eight nations, including most Caribbean
countries.
In the United States, C&W operates through TDX Systems and has become
the fourth largest long-distance company. Another C&W activity is its joint
venture PTAT for a private submarine fiber-optic cable between the United
Kingdom and the United States. C&W also participates in the trans-Pacific
cable venture IDC, and increased its share to 17 percent. It is also represented
in domestic Japanese service through Fair-way. These pieces fit together as part
of a strategy to link the four major financial centers in the worldLondon,
New York, Tokyo, and Hong Kongwith Bahrain and Singapore as additional
link-up possibilities. Aside from the PTAT and North Pacific cables, C&W also
backed the Asiasat project and co-won a mobile license in Germany (Cable &
Wireless, 1990, p. 8). C&W also entered in 1991 telecommunications in Swe-
den through the first competing public network service. Tele2 AB was owned
40 percent by Cable & Wireless and 60 percent by Kinnevik, and used fiber
lines belonging to the Swedish railroad. C&W also established facilities man-
agement centers in continental Europe.
Mercury
In the U.K., C&W is the sole owner of Mercury, the alternative to BT in long-
distance service. Originally, C&W's partners in Mercury were British Petro-
leum and Barclay's Merchant Bank. Although Mercury was modeled on the
U.S. company MCI, there are great differences between the two. Whereas MCI
was a small, maverick firm that entered the market by prevailing in court over
the opposition of both AT&T and the federal authorities, Mercury was born
with three silver spoons in its mouth and the government as its godparent.
Mercury emerged through a classic insider deal rather than on the basis of
competitive bidding. Furthermore, because the Conservative government staked
the credibility of its telecommunications on Mercury's effectiveness, it pro-
tected the company by giving it an exclusive license to compete with BT until
1990. There are, however, similarities to MCI. Both initially stated that their
intention was only to provide limited service offerings, in particular leased
lines, aimed at users underserved by the monopoly. However, their ambitions
grew rapidly, and they sought to gain many additional customers by intercon-
necting into the public network and by providing switched public domestic and
international long-distance service.
Mercury's intentions raised the stakes for BT considerably. To both BT and
AT&T it seemed unfair that they should be required to let competitors use their
carefully nurtured local distribution network. The notion of common carriage,
however, is based on the premise that service must be provided to anyone who
pays the posted price. Although British Telecom was understandably not enthu-
siastic about its new competition, Mercury's presence provided the basis for
further liberalization of BT's operations, which was a positive prospect for the
large company.
After Mercury's initial license was granted in 1982, it quickly established a
microwave network service within London, and connected this network by dig-
ital microwave with Birmingham and Manchester. It also began construction
of a national fiber-optic trunk system centered in Birmingham. In London it
made substantial use of subway tunnels and acquired rights to lay cable through
the underground cable network of the London Hydraulic Power Company.
Mercury, as a subsidiary of C&W with its substantial international presence,
also established itself abroad. In 1983, it received permission to provide inter-
national service. In 1984, transatlantic service to the United States was intro-
duced. In 1987, Mercury also penetrated the continental European market when
it arranged to exchange public telecommunications traffic with Italy. This was
troublesome for BT, which derived about 20 percent of its profits from inter-
national service.
To meet its switching equipment needs, Mercury went outside the traditional
British manufacturers GEC and Plessey to Northern Telecom in Canada. In
1986, Mercury entered an agreement with the largest domestic computer com-
pany, ICL (through their parents C&W and STC), for a joint venture in spe-
cialized data communications and value-added services.
Although Mercury's regular license, granted in 1984, is similar to British
Telecom's, it has several important differences: (1) Mercury does not have to
fulfill BT's universal service obligations; (2) it does not operate a full national
system and is under no obligation to do so; (3) it cannot provide maritime
services other than for offshore installations; and (4) it is not price regulated.
Mercury started its public switched operations in Britain on May 15, 1986,
with long-distance rates that were about 15 to 20 percent lower than BT's,
despite the latter's anticipatory tariff reductions. Mercury's goal was to reach a
5 percent market share by 1990 (about 8 billion) and a much larger share of
large user business.
By 1987, 37 percent of major telecommunications users in Britain were us-
ing Mercury for some of their service. Of eighteen financial institutions in the
City, seventeen became Mercury customers. The share of large users in BT
revenues is about 30 percent. It was not merely a matter of price: large users
were seeking to be less dependent on one supplier, an important factor in a
country as prone to strikes as Britain. But when it came to small and medium-
sized users Mercury was lagging.
BT's primary leverage over Mercury is through interconnection into its local
distribution network, and the regulatory determination of that interconnection
relation is therefore critical. The question of how much BT can charge Mercury
involves murky conceptual and accounting issues. Additionally, technical as-
pects such as the numbering system, the points of interconnection, the quality
of service, the number of digits to be dialed, and so on, have to be considered.
BT has some reasons to be less than fully cooperative, and Mercury has an
incentive to cry wolf and seek an advantageous interconnection arrangement.
Without a period of protection, Mercury argues, it could not compete with BT,
and its failure would undermine the entire base of government pro-competition
policy. It would be embarrassing (at least for the Conservatives) to see Mercury
fail. One government minister commented:
If we opened up to free competition, there is a danger that British Telecom would
be able to wipe the floor with all the tiny competitors. We think the method of
introducing a little competition begins with the Mercury rival network. It is our
duty to look after Mercury, to nurse it. (Jason, 1985, p. 4)
Hence, Mercury has a certain leverage over British policy makers that is dis-
proportionate to its economic power. With a Labour government in power, the
reverse is possible and Mercury could be choked to death by "technical" reg-
ulatory decisions rather than by decisions debated and passed by Parliament.
The problem of interconnection, and the respective licenses of BT and Mer-
cury, came before Oftel. In 1985, Bryan Carsberg decided on a framework for
the interconnection of Mercury and BT, a ruling that was considered a major
success for Mercury. The ruling stipulated that BT had to provide Mercury
with local interconnections at both ends of a telephone connection, set the com-
pensation that Mercury must pay BT, and established a time schedule of pay-
ment for one-half of the cost of providing the additional capacity. BT must also
provide full international interconnections for Mercury.
The issue of fair interconnection is complicated. In the United States it has
led to two decades of dispute and was a major issue leading to the AT&T
divestiture. The Justice Department, and with it Judge Harold Greene, con-
cluded that one could not expect a local monopoly to provide genuinely non-
discriminating access to its long-distance competitors. The divestiture estab-
lished the principle of complete and equal access and allowed users to choose
a "primary long-distance carrier." But the question of the cost for such access
to the local network, whether by AT&T or by its competitors, precipitated
fierce battles between long-distance carriers and local exchange companies, and
among long-distance carriers themselves.
The issues are hardly trivial. The total of access charge payments that long-
distance companies must pay to the local exchange companies can comprise
more than half of their entire revenues; they are also a major source of revenue
for local exchange companies (Noam, 1986). In the United States, according
to some calculations, the actual cost of access to the local company was ap-
proximately $0.03 per minute, but charges to the long-distance carrier were
between $0.07 and $0.08 per minute. This substantial gap exists because, in
most views, long-distance calling has subsidized local telephony. In a compet-
itive environment, a substantial markup for access above cost creates an incen-
tive for long-distance carriers to "bypass" the local public switched network
entirely and reach users directly. Especially for large users it begins to make
economic sense to lease local circuits from the local telephone carrier to the
long-distance carrier, which are not governed by the access line charges, or to
create communication links entirely outside of the local telephone network.
At the heart of this complex matter is the question of the allocation of the
joint costs, largely of a fixed nature, that are incurred to make the provision of
both local and long-distance services possible. Conceptually and practically,
the allocation between the two types of services is difficult, and in some views,
arbitrary. After much dispute, the FCC decided to resolve this problem by
imposing a flat-rate "user charge" to be borne by end users. In theory, this
charge would be partially offset by lower long-distance charges, as carriers
would pass on their savings from reduced access charges to end users. The
economic logic was that since flat-rate user charges were not usage sensitive,
users could not avoid them by decreasing the number of telephone calls they
made through the local public switched network. Hence, it would reduce the
incentive for "uneconomic" (i.e., purely regulation-induced) bypass.
A related question is whether a differentiated access charge should be set for
different long-distance carriers for access to the local exchange network. (AT&T
had to pay more than its competitors.) Once customers could use all carriers
under equal terms, unequal access charges for different carriers appeared to be
unequitable. However, because the alternative long-distance carriers did not
have the economies of scale of AT&T, equal access rates could have, in effect,
made them uncompetitive. Therefore, the alternative carriers asked for at least
a temporary handicapping of AT&T in their favor.
Mercury reached profitability in 1989. However, its total income was 1.6
percent of BT's and its shares of international traffic (4.3 percent), inland phone
calls (0.7 percent), and leased lines (7.6 percent) were miniscule. Mercury
invested a total of 825 million from 1981 to 1989, less than one-third of BT's
investment for the year 1990 (Arlandis and Gille, 1989).
Mercury's status was reviewed in the 1990 'duopoly review,' leading to fur-
ther opening of telecommunications and to potential competition for Mercury.
This review is discussed at the end of this chapter.
Value-Added Networks and Resale

After the government announced in 1980 that it intended to reduce BT's mo-
nopoly, Professor Michael Beesley of the London Business School, a transpor-
tation expert, was commissioned by the government to study the economic
implications of permitting third parties to offer services over BT's network.
The main focus of Beesley's inquiry was the feasibility of value-added network
services (VANs) and their benefit to the customers, as well as any resultant
effects on BT's pricing and profitability. Beesley expanded the analysis to in-
clude the pure resale of network capacity as well.
Beesley received many complaints that BT's monopoly had delayed new
services and new technologies, such as digital technology, packet switching
services, and memory telephones. The various restrictions, including the pro-
hibition of resale and sharing, were criticized. They inhibited usages such as
direct debit arrangements between customers and banks, computer access for
storing and forwarding messages, and the extension of radio paging services to
include storage and forwarding of messages. All these services were impossible
because they required the sharing of leased circuits. BT permitted a number of
value-added services as long as they did not require private switching or com-
pete with BT's actual or planned services. Other proposed services, however,
were turned down. In Beesley's views, these restrictions were only the tip of
the iceberg because innovative services do not flourish in the abstract; they
await opening of the rules under which business can take place (Beesley, 1981).
BT estimated that unconstrained reselling of capacity would result in a net
loss of revenue from domestic calls of 30 million for 1984-1985. With more
generous assumptions for potential "cream skimming," it calculated a net rev-
enue loss of 110 million. If these figures are taken as the lower and upper
ranges, they represent between 0.4 percent and 1.5 percent of BT's gross rev-
enues and between 3.2 percent and 12 percent of its profit. A 12 percent re-
duction of profit is hardly a small loss. Beesley, however, did not share this
apprehension, for he believed those estimates to be very "vulnerable." Even
using BT's estimate, Beesley found that revenue loss would amount at the
maximum to only 6.4 per residential customer, a loss that would be offset by
about an 11 percent increase in the residential rate. This would, however, be
accompanied by gains to consumers in reduced long-distance rates and by po-
tential innovations in services. Beesley concluded that resale should be permit-
ted without use restriction because it would encourage ingenuity in meeting
customers' demands with favorable effects on productivity and exports.
Beesley also expanded the analysis to international service. BT was highly
sensitive on this issue, considering the great profit contribution of international
traffic. One study estimated that a hypothetical London to New York private
satellite link would cost 5300 per year, in contrast with the BT tariff for such
a connection of about 50,000. For approximately the same distance, a leased
line coast-to-coast in the United States costs 4500 per year, again less than
one-tenth of BT's price. The pricing of international calls seemed to be close
to its revenue maximizing high. According to Beesley, for a three-minute U.S.-
to-U.K. call the mean expected elasticity was -0.936. For a U.K.-to-U.S. calls,
it was -1.094. These figures are close to the maximizing 1.0, which indicates
revenue vulnerability to a price reduction induced by competition or resale.
Beesley recommended easing all restrictions on the offering of all services
on the BT network, including resale. He further recommended that constraints
on BT's pricing be reduced or eliminated so that rates could move toward
costs. BT, in turn, should be free to enter nonvoice markets as a competitor,
provided safeguards were in place to prevent unfair competition. Domestically,
leased rates ought not to discriminate on the basis of total usage. Internation-
ally, however, differentiated rates could be instituted. And this liberalization,
Beesley argued, should be seen in the context of a possible rival entry into
transmission and switching, which he advocated.
BT opposed the recommendations of the Beesley Report, pointing again to
revenue losses. BT's chairman, Sir George Jefferson, conceded that "change
in the way telecommunications services are provided in Britain is inevitable;

BT is not seeking to preserve the status quo . . . " (British Telecom, 1981).
But he also warned that "precipitate implementation of the Beesley recommen-
dations [would] cause irreversible damage . . . Further detailed studies are
necessary."
In addition, BT waved a warning flag in the direction of equipment manu-
facturers: "British Telecom is concerned about the possible impact on its sup-
pliers . . . If it is faced with the competition from separate networks, it may
be unable, as a result of the activities of its competitors to make long term
decisions that support the British national telecommunications industry" (Brit-
ish Telecom, 1981).
In July 1981, the government partly accepted the recommendations and an-
nounced that it would permit private suppliers of VANs to receive licenses but
that it would not presently permit the resale of leased line capacity, since this
could undermine universal service. The government did provide, however, for
the possibile resale in the future.
In 1984, the government changed course by deciding that resale of services
would be permitted after July 1989. In the meantime, it liberalized the leasing
of shared capacity on privately owned networks and the third-party use of pri-
vate circuits, thus coming close to resale.
In 1989, the DTI approved an Oftel report that allowed simple resale of
capacity leased from BT, Mercury, or Hull. The United Kingdom was the first
European nation to allow leased line resale. DTI began issuing licenses to pri-
vate network operators and lifted restrictions on the number of connections to
other private networks. British Rail, Istel, and Racal quickly established virtual
network services, private pay phones, and other services on their leased lines.
In response, BT raised leased-line tariffs by 11 percent.
Even the Post Office returned to offering telecommunications services. It
started National Network (Natnet) to resell 30 percent of the capacity on its
own private network. The Post Office network carries 12 million calls yearly
from 25,000 users and 4500 terminals to 200 locations throughout the United
Kingdom (The Economist, 1990a; Purton, 1990).
Beesley had indicated the difficulties in distinguishing simple resale from
value-added services. In the United States, where the FCC's First, Second, and
Third Computer Inquiries grappled with a related problem, experience demon-
strated the fluidity of the separation. The British approach to VAN licensing
has been to try to proceed pragmatically, without getting into philosophical or
semantic disputes. VANs are required to offer more than just transmission. The
question of whether packet switching could be considered a value-added service
was particularly controversial. Against the opposition of BT, the government
recognized packet switching as a value-added service, thus allowing licensed
competitors to provide this service.
The primary British regulatory structure is based on the separation of the two
categories of service: "basic conveyance" and all other services, which are
defined as "value-added." The definition of basic conveyance service is "the
conveyance of a message by means of a telecommunications system to a single
destination . . . in the same form that it was received without any additional
services having been provided . . . " (Department of Trade and Industry, 1985,
p. 16). Drawing on the experience of other countries, especially the United
States, the government wisely regarded its definitions as only temporary.
More specific rules for VANs were set in 1982 (Department of Trade and
Industry, 1985). By February 1987, 841 VANs were licensed, operated by 221
different companies (Department of Trade and Industry, 1987). Of these, the
most popular were store and retrieve systems (112), mailboxes (90), protocol
conversion between incompatible computers and terminals (90), customers' data-
bases (66), deferred transmission (63), user management packages (58), view-
data videotex services (62), wordprocessor/facsimile interfacing (46), multiad-
dressing routing (56), and speed and code conversion between incompatible
terminals (49). Other VANs include automatic ticket reservations, conference
calls, long-term archiving, secure delivery services, telesoftware, and text ed-
iting.
VANs with a volume of more than 250,000 per year were subject to rules
that prevent the establishment of a dominant market position. These limitations
were aimed at BT and IBM. Under its license, BT must provide services na-
tionwide, unlike some of its VAN competitors. BT is also subject to rules that
prevent a cross-subsidy of its VANs out of its other services.
In 1986, the VAN rules were further modified to encompass managed data
networks, which had thrown the separation between basic service and VANs
into disarray, thus illustrating the complications of partial liberalization and
partial approval requirements. The legislation was again changed in 1987 when
the VANs license was replaced by the Class License for Value Added and Data
Services (VADS). Under the new classification, those wishing to run under
license from Oftel, other than major service providers, need no longer register
with Oftel.
In 1984, BT and IBM teamed up and announced their intention to establish
a joint VAN venture for data network management services, and applied for a
value-added license under the name JOVE. This set off a strong domestic pro-
test in Britain, and about 100 computer and communications companies regis-
tered their opposition. They were concerned with the reliance on IBM's SNA
architecture, which they feared would threaten British industry and government
development toward an open systems interconnection (OSI). The two partners
argued that their venture allowed for an OSI protocol standard, but by then the
opposition had grown to a clamor. There was also much concern expressed
about the feasibility of competition if two dominant firms in closely related
markets were permitted to link together. With Britain actively introducing com-
petition into its telecommunications and computer fields, the government felt
that such a move would have been counterproductive. Not only would it make
competitive entry more complicated, but it would remove the potential rivalry
between IBM and BT.
Other critics of the JOVE venture feared that it would permit IBM to achieve
some measure of control over BT. They pointed out that IBM-UK had six
employees in 1951 and 16,000 by 1984 (Bird and Huxley, 1984). Given the
determined opposition, the British government, through the Department of Trade

and Industry, rejected the application in October 1984 but left the door open
for either company to offer such services on its own.
In response to JOVE, several other VAN operators consolidated. The United
Kingdom's two leading competitors, ICL and General Electric Information Ser-
vices, merged their VAN operations into International Network Services Lim-
ited (INS) in 1987. In 1989, one of the United Kingdom's largest VAN oper-
ators, Istel, was acquired by AT&T.
Hull Telephone Department

The Hull Telephone Department (HTD), the only independent local telephone
company in Britain, received its original license in 1902 to operate in an area
already served by the private National Telephone Company. Four years later,
Hull Telephone was the only municipal company that resisted acquisition by
the Post Office. The offer was rejected by the Hull chairman's tie-breaking
vote. In 1914, Hull Telephone purchased the National Telephone Company
plant from the Post Office and has renewed its original license every twenty to
thirty years since then. With the privatization of British Telecom and the crea-
tion of Oftel in 1984, a new regulatory system was established, and Hull re-
ceived another twenty-five-year license. Hull as a city, and with it HTD oper-
ation, is controlled by the Labour party. In the first eighty years of its existence,
HTD had only four general managers, all of them exhibiting unusual political
longevity.
Hull Telephone is officially known as the City of Kingston-upon-Hull Tele-
phone Department. It serves a population of 375,000 with a total of 128,000
lines and fourteen telephone exchanges. The company has semielectronic ex-
changes and digital System X switches. Its main exchange is located adjacent
to the British Telecom exchange to which it is connected through a hole in the
wall.
In contrast to the United States, local service in Hull is not subsidized from
long-distance revenues that originate in its territory. Originally, Hull Telephone
paid 10 percent of its revenues to the Post Office and retained 5 percent of the
trunk and international revenues that it initiated. After 1979, both types of
charges were eliminated and Hull instead paid the government a license fee of
150,000 per year. After liberalization and BT privatization, Hull still pays a
license fee and has negotiated with British Telecom for a different division of
revenue involving both companies. Hull customers have split usage of long
distance service fairly evenly between BT and Mercury, offering a potential
model for the rest of Britain.
HTD telephone service has been cheaper than that of BT; Hull provided an
unlimited local call at 5 pence in 1985, whereas BT provided for that price
local service ranging between thirty seconds and two minutes, depending on
the time of day. Hull's fee for residential line rental per year was 50, and 84
for businesses, compared to 65 and 102 for BT (The Economist, 1985). Hull's
local service is profitable, in contrast to BT's. But it is cheaper to provide local

service in a city of Hull's size than in rural areas.
Cellular Telephony
With Oftel especially concerned about competition in new service, mobile radio
received special attention. Professor Bryan Carsberg was interested in protect-
ing new services offerings against BT dominance and concluded that "BT should
not be a network operator, either directly or indirectly, or have more
than a minority share in providers of service for new p.m.r. [public mobile
radio] networks." A public telecommunications operator running a public mo-
bile radio system would have "unmatchable advantages over competing sys-
tems" (Oftel, 1985b, p. 2).
After applications were solicited, licenses were granted to two providers,
Cellnet and Vodaphone, with the obligation to cover almost the entire country
by 1989. Cellnet is a consortium of BT and Securicor, a security services com-
pany that had also been active in telemetry services; Vodaphone is a joint ven-
ture of Racal, Millicom, and Hambros. The competitive structure that had been
set up gave advantages to early starters, and the two companies rushed head-
long into the service. Cellular telephone service became operational in 1985,
drawing a very strong demand.
Cellnet and Vodaphone, however, cannot sell directly to users, they must go
through service providers who resell to the public in return for a commission
of approximately 15 percent. The terms for dealing with these service providers
must be published and must be equal for all, with the exception that the terms
can be volume-related. For service providers, the commission is approximately
15 percent. In practice, the principals of the two cellular telephone operations
set up their own independent retail service providers to deal with the public:
BT established BT Mobile Phone, and Racal set up Racal Vodac. Importantly,
however, these companies cannot receive advantages over other service provid-
ers. Service providers are allowed to market both networks; and resellers, in
turn, can employ independent agents for transactions (Fuller and Mitchell, 1986).
In 1987, there were about fifty reselling organizations, including established
firms such as Marconi and Motorola. Also involved in the sale or resale of
equipment and installation was a legion of local agents, ranging from garages
and telephone stores to office equipment suppliers. This structure permitted the
emergence of vigorous competition, although mostly for equipment packages
rather than for "air time" tariffs. Because the retailers share in the profits from
cellular phone calls, they often sell equipment at low cost to increase the sub-
scriber base. In 1990, mobile phones were available for as little as $300. There
were 650,000 subscribers.
In the competition between the two network operators, Cellnet started with
the significant advantage of BT's resources, an established telephone network,
and existing mobile telecommunications operations with an established cus-
tomer base. With the privatization of BT, it also had the good will of millions
of British shareholders. Racal's Vodaphone, on the other hand, was an un-
known entity. Its advertising slogan was "Racalthe largest company you've
never heard of" (Raggett, 1986).
The relation between two companies moved from hostility to cooperation,
partly because there was enough business for both companies' services in the
initial phase. BT provides interconnection service for Vodaphone, thus creating
a delicate relationship. In the technical field, the companies must cooperate to
achieve the required full cellular interconnection that would permit "roaming"
between the two systems. By 1990 there were nearly 800,000 cellular users,
the second highest number in Europe behind Sweden, split evenly between
Cellnet and Vodaphone. Each had invested some 200-300 million in their
networks and each was gaining 15,000 new subscribers monthly. Even with
success in gaining subscribers, there was substantial user criticism with the
performance of Racal (22 percent dissatisfaction in a survey) and Cellnet (32
percent dissatisfaction.
In 1989, the United Kingdom moved to the forefront of mobile communica-
tions by licensing two entirely new forms of operation, Telepoint and PCN.
Telepoint or CT-2 (cordless telephone second generation) consists of hand-held
units that can be used within 200 meters of a base station.
Telepoint service operators install low-cost base stations ($180-$500, de-
pending on estimate). In 1988, the DTI awarded four Telepoint licenses to
Ferranti, Phonepoint (BT, France Telecom, STC, NYNEX, and the German
DBF), Callpoint (Mercury, Motorola, and Shaye), and BYPS Comms (Bar-
clays, Philips, and Shell).
CT-2 employs a frequency division multiple access system, as opposed to
the EC's Digital European Cordless Telecommunications (DECT) standard, which
uses time division multiplexing and which was especially promoted by Erics-
son. The DTI asked for the European Telecommunications Standards Institute
(ETSI) to determine the appropriate standard, GSM or DECT, for the U.K.
system after operators were unable to reach an agreement (Green, 1990b).
Personal Communications Networks (PCN), is another form of mobile com-
munications. It employs very small mobile transceivers (though at low power,
with less reach and less mobility than cellular).
DTI received eight applications for PCN service and granted licenses to three
consortia in 1989: Mercury PCN (C&W, Motorola, and Telefonica), BAe (British
Aerospace, Millicom, PacTel, and Sony), and Unitel (STC, Thorn EMI, US
West, and the DBP). BT's Cellnet and Racal's Vodaphone did not receive li-
censes but will be allowed to configure their systems for PCNs. Each consor-
tium aimed to invest $1~$2.5 billion for service available by 1992 (Sims, 1989;
Oftel, 1990a, p. 1).
Optimistic projections of 12-30 million Telepoint and PCN users by the year
2000 were quickly scaled back. For Telepoint, only 3500 transmitters were
installed in 1990, compared with 86,000 public call boxes and 358,000 private
pay phones (Lynch and Hayes, 1990). By 1991, Telepoint had proved to add
little value to existing cellular service, and all but one of the operators had
gone out of business. PCN did not fare much better. PCN figures were adjusted
downwards and yearly losses were foreseen until the turn of the century.
Cable Television as a Telecommunications Carrier

Cable television policy in the United Kingdom is dealt with in the companion
volume on European television. However, the effects of cable television and
broadband transmission on the telephone system and its providers will be briefly
discussed here.
Traditional telephone and cable television transmission cross paths in two
ways. First, broadband cable systems have the potential to be used for the
distribution of traditional and new telephone services. Mercury had unspecified
plans to use local cable networks to link customers to its long-distance system.
To the extent that arrangements such as these are possible, competitive local
distribution may emerge (Noam, 1982).
To provide more than fixed line service, cable distribution requires some
switching ability, such as a star network architecture, as opposed to the tree-
and-branch systems in cable television. In its 1983 specification for cable licen-
ses, the British government required that a certain number of cable franchises
were to be in the star configuration. This proved to be a decision that under-
mined the prospective viability of many proposed cable systems. The second
way in which cable television and the traditional telephone services overlap is
through BT's own role in the installation and operation of cable systems. Such
involvement in cable television is in marked contrast to the United States (Noam,
1985).
In defining the role of BT (and to a lesser extent of Mercury) in cable tele-
vision transmission, the government was faced with a dilemma. On the one
hand, if BT and Mercury were to be granted an active role in laying cable and
in local operation even on a common-carrier basis, BT's power would be in-
creased even further, and thus the entry of new telecommunications carriers
and services would be less likely. On the other hand, the availability of the
two companies as actual or potential operators would put pressure on all other
companies for efficient performance and could also lead to cabling of marginal
areas. The government adopted a compromise position. It permitted the entry
of BT and Mercury, either independently or in consortia with other companies
as potential licensees. Because of the centralized licensing mechanism, it was
quite unlikely that British Telecom would receive a predominant number of the
licenses, at least under a Conservative government, even if it could point to
some superior economies.
At the same time, the government sought to protect BT and Mercury's tra-
ditional lines of business from competition by cable operators, because this
would tend to have a negative effect on the national public system and weaken
Mercury's position as the "official" competitor to BT. Cable operators are
permitted to offer a variety of interactive services, but voice and much of data
communications remained exclusively the domain of BT, Mercury, and Hull.
Similarly, data communications in major business centers are the exclusive

province of the three carriers, although a cable operator could offer them in
conjunction with BT and Mercury. In 1987, Oftel granted temporary licenses
to two cable television companies to offer telephone service in cooperation with
Mercury Communications. The role of cable carriers became an even more
interesting question after many cable franchises were granted in 1989 and 1990
to American phone companies such as NYNEX, US West, and PacTel, which
proposed to offer basic telephone service and interconnect cable networks with-
out using BT or Mercury networks. At the same time, Mercury and BT were
pressing for the right to offer cable TV service. The government's 1991 White
Paper increased cable companies' flexibility in telecommunications. It also kept
BT out of their turf until at least 1997.
Britain Breaches the European Cartel
Britain's liberalization policy put the United Kingdom into conflict with its
partners in the European Economic Community and the European Conference
on Post and Telecommunications (CEPT), the organization of European PTTs.
This conflict was exemplified in the "British Telecom Case" before the Euro-
pean Court, a case that illustrates the British dilemma of reconciling its Euro-
pean role while pursuing a telecommunications policy somewhat different from
its partners.
Britain had about 100 private message forwarding agencies, which receive
or transmit telex messages from customers who themselves do not have a telex
subscription (Dumey, 1983). At first operating only within the United King-
dom, several agencies then expanded their message relaying service into con-
tinental Europe, North America, and Asia. This was profitable because British
Telecom's international telex rates were quite low, particularly to North Amer-
ica, creating an incentive for continental European users to route their telex
traffic via London when sending their telex messages across the Atlantic. Ini-
tially, firms went directly through their British subsidiary's office if they had
one. For users without related U.K. branches, the telex forwarding agencies
started to fulfill the same function. In some instances, the agencies offered
superior service, including a money-back guarantee if the message was not
transmitted within a certain time.
The PTTs, facing the loss of revenue and an emerging competitive price
regime, fought back by mustering the rules of CEPT and the CCITT that "har-
monize" the PTTs behavior. CCITT telegram recommendations required PTTs
to block telegram and telex messages that were sent to forwarding agencies for
transmission in order to "evade" full charges of the complete route. In 1975,
the British Post Office, still operating under the traditional policy guidelines,
clamped down on the forwarding agencies and required messages to be charged
a rate by the forwarding agencies equal to the tariff that would have been paid
for a direct telex route bypassing the United Kingdom. But it was impossible
to enforce this provision. The telex bureaus had no incentive to check on rates
between third countries. In early 1978, the provision was dropped. It was,
however, replaced by other and stricter rules prohibiting telex agencies from
providing international services for their customers altogether when messages
were in data form (from computer to computer) and were received through the
telephone lines and then converted to telex, facsimile, or other visual form.
One of the telex agencies lodged a formal complaint with the Commission
of the European Communities in June 1979. Proceedings took place in 1980,
while the British government reorganized its telecommunications system and
set up British Telecom. BT at first maintained the previous prohibitions. The
case before the European Commission therefore proceeded, and the decision
was announced in December 1982 (Official Journal, L360, 1982, p. 36). In its
decision, the commission found BT's rules to be a violation of Article 86 of
the Treaty of Rome. Ironically, however, BT had withdrawn these restrictions
two months earlier, in light of Britain's changed attitude toward service com-
petition; the issue was thus moot as far as the British situation was concerned.
In its decision, the commission found that the Post Office, and later BT, had
abused their position as a statutory monopoly. The 1976 restrictions would
have required telex agencies to discriminate in their rates for equivalent trans-
actions according to the country of the customer or the country of destination.
The 1978 restriction was similarly discriminatory, because it prohibited telex
traffic both to and from countries, including EEC members. It also found that
the prohibition of use of a combined telephone line and telex link for computer
data transmission imposed restrictions on the development of a new market and
new technologies, as well as on the efficient use of existing facilities, thus
restricting interstate European trade.
The British government, when it was still opposing the telex bureaus, used
Article 90(2) of the Treaty of Rome in its defense, contending that this section
exempted public enterprises from the EEC competition rules. That section,
however, contains a broad principle that applies competitive rules of the treaty
to public monopolies where important services of "general economic interest"
are at stake. Hence, where the PTT's obligation is to cooperate with other
PTTs, the commission found that the "development of trade must not be af-
fected to such an extent that it is contrary to the interest of the community,"
even if this resulted in the inability of the PTT to fulfill its duties. Thus, the
limitation of Article 90(2) did not apply.
Though the court ruled against BT's provisions of 1976 and 1978, it imposed
no fine, taking a number of factors into account: BT had acted originally under
the pressure of other European PTTs; it had not profited by these rules in the
sense of additional revenues; and it had suspended and not enforced the prohi-
bition during the commission proceedings (Dumey, 1983).
Thus, the commission announced that the restraints upon telecommunications
services, even when undertaken within a concerted European policy, were in
violation of the European antitrust provisions. This could have been the end of
the story, but the commission's principle was too important to be left unchal-
lenged. The British government, by now firmly embarked on a course of lib-
eralization and possibly coveting the role of London as a communications hub,
was in no mood to appeal the decision, but other European countries were. The
Italian government therefore took up the defense of BT's lost virtue and chal-
lenged the commission's decision before the European Court of Justice (case
41-83) in an appeal that had implications far beyond telecommunications. It
was the first time that a Common Market member state had appealed a decision
of the commission in an individual competition case. Moreover, it was the first
time that a member state appealed a decision that concerned a company over
which it did not have direct jurisdiction.
The Italian government made several arguments. It claimed that BT's actions
were part of the exercise of legislative power, rather than entrepreneurial activ-
ity, it claimed that these activities were regulatory and public law activities
essential to BT to accomplish the task of providing telecommunications ser-
vices domestically and internationally. Furthermore, Italy argued that BT, as a
member of the ITU, was required to implement international regulations adopted
by the ITU (such as the 1947 Atlantic City Convention) such that international
agreements must be honored even if the effects were contrary to the EEC treaty
rules. Furthermore, according to Article 222, the treaty "does not prejudice the
system of property ownership within the Member States." Italy argued that the
legality of BT's action in its capacity as monopoly could not be brought into
question without questioning the legality of the monopoly itself. Instead, the
claim continued, the British government needed to protect itself from unfair
competition by cream-skimming enterprises that jeopardized the economic in-
tegrity of the telecommunications systems.
The commission had declared the regulations against telex message forward-
ing to be anticompetitive and in violation of Article 86 of the Treaty of Rome.
In response, Italy argued that the Community's competition rules did not apply
to monopoly telecommunications services authorities and that regulatory activ-
ities of public companies should not be considered as an activity of an "under-
taking" within the meaning of Article 86.
The European High Court of Justice announced its decision in March 1985,
two years after the case was appealed to it and ten years after BT's regulations
of 1975 and 1976 (Court of Justice of the European Communities, 1985). The
court firmly rejected all of Italy's claims, thus putting an end to the legal at-
tempt to block third-party traffic in telex. In all likelihood, similar arbitrage
services by a European country functioning as a communications hub will thus
be upheld against PTT cartel prohibitions.
The Equipment Sector
The British telecommunications service monopoly was traditionally closely linked

with the country's equipment industry. In consequence, the industry took a
decidedly unenthusiastic view of the liberalization and privatization of its main
customer, British Telecom.
A tradition of oligopoly pervaded the early history of British equipment man-
ufacturing. Before 1920, cable procurements by the Post Office were allocated
among themselves by the members of a cartel, the Telephone Cable Manufac-

turers Association (TCMA) (Foreman-Peck, 1985, p. 224). In the 1920s, cartel
agreements were reached with the Post Office to limit the number of orders
submitted to companies outside of the association's members. Later, during the
Depression, the Post Office pressured TCMA members to reduce prices and
cost by concentrated production. The TCMA, in return, asked for a ten year
supply agreement and a promise not to go outside of the cartel, which was
agreed upon.
Similar market-stabilizing arrangements were made for switching equipment.
After World War I, the GPO decided to concentrate on the Strowger system
produced by the Automatic Electric Company. Four companiesAEC, STC,
Siemens (Britain), and GECentered into an agreement to share orders during
the years 1923-1928. In 1928, the four companies committed themselves to a
"bulk supply" agreement for another five years, at prices already agreed upon.
A committee of the manufacturers themselves determined the various manufac-
turing quotas. Similar bulk supply agreements (BSAs) existed for other types
of equipment. When they were abandoned, the effects on procurement prices
were often dramatic. For example, when the BSA for batteries was abandoned
for competitive contracts in 1956, prices fell by almost 50 percent. Similarly,
when the BSA for telephone cords was abolished in 1953, prices decreased by
more than 50 percent within five years. The BSAs also had the effect of reduc-
ing incentive for research and development, since the market shares were al-
ready fixed.
In the 1960s, the Post Office Engineering Union (POEU) issued a report
entitled The Telephone Ring, which described the cartel of British equipment
manufacturers (the "ring") and the decline of telephone exports from Britain.
The report pointed out that telephone dials made by ring companies cost almost
twice as much as the same dial made by a continental European company.
More recently, the protective impact of the BT supply monopoly, which has
kept prices high, has also been obvious in the PBX market. Prior to liberali-
zation of PBXs with less than 100 lines, the cost of a PBX line in London was
650, compared to 200 in Dublin for similar equipment (Stapley, 1981). Mo-
nopoly also affected diversity. When BT chose not to compete in the market
for PBXs above 100 lines, a wide array of products became available.
BT's newly independent equipment practices were viewed with apprehension
by the manufacturing firms. The Telephone Equipment Manufacturers Associ-
ation (TEMA) accused BT of a variety of unfair trade practices(Connections,
1985). When BT announced its intention to purchase the Canadian PBX maker
Mitel, TEMA vociferously opposed this vertical integration into manufacturing
and argued to the Monopolies and Mergers Commission, with some justifica-
tion, that it would strengthen BT's power in terminal equipment and threaten
British PBX manufacturers. Against the advice of Oftel, the commission never-
theless recommended an approval of the merger, albeit with some tough con-
ditions attached: structural separation, absence of cross subsidies, and no BT
purchases from Mitel for the public network or for supply to end users at least
until 1990. But even these conditions were softened by the government when
it approved the merger, though with a ceiling on procurement from Mitel. BT

promised not to acquire more Mitel equipment for the public network than it
had bought in 1985 in terms of value. Since PBX prices have fallen, however,
this amount has actually permitted a quantitative increase. In 1989, the DTI
allowed BT to increase its procurement of PBXs from Mitel. Mitel's financial
performance kept deteriorating, and British Telecom eventually tried to sell it,
although it found no buyers after two years of efforts.
In response to complaints from manufacturers that equipment liberalization
was being slowed by lengthy approval procedures, Oftel convened the Birtwis-
tle Committee, which in 1987 suggested eliminating any mandatory test. BT
opposed these changes. Even after CPE equipment and handsets were liberal-
ized, BT controlled 70 percent of the U.K. market for equipment.
The General Electric Company (GEC) is the largest electrical and electronics
company in the United Kingdom (Locksley, 1982). Traditionally, it has had
strong ties with the government through its telecommunications, electric gen-
eration, and defense procurement, and it is British Telecom's major supplier.
GEC is involved in the development of the System X digital switch, and it
markets private PBXs, partly using the technology of Northern Telecom and
NEC. The company also designs and manufactures semiconductor components
and electronic office equipment. GEC also sought entry into cable television
franchise operations.
Plessey is the second major telecommunications equipment manufacturer. It
was the lead developer of the System X digital switching system. The company
is also active in the defense aerospace industries and in microelectronic com-
ponents. Other Plessey products include satellite ground terminals, radio relay
systems, telex, packet switching systems, and coin-operated telephones. It ac-
quired the American switching manufacturer Stromberg-Carlson, subsequently
controlled by Siemens, and is a part owner of Scientific-Atlanta, a major Amer-
ican producer of satellite and cable television equipment.
In 1985, Plessey's profits declined dramatically. At the same time, the com-
pany had to carry the losses of Stromberg-Carlson. It sought a major American
military communications procurement order, jointly with Rockwell Interna-
tional, and even persuaded Prime Minister Thatcher to raise the matter person-
ally with President Reagan. What Mrs. Thatcher reportedly had not been told
was that the Plessey/Rockwell bid was almost twice as high as the bid submit-
ted by the U.S. firm GTE with the French Thomson-CSF. Plessey consequently
lost the bid.
Plessey's major headache was the digital exchange System X, jointly devel-
oped with GEC. System X development had cost 800 million and was far
behind schedule.
In 1985, GEC offered about $1.7 billion in a takeover bid for its rival Ples-
sey. Plessey resisted and counterproposed its taking over GEC's System X
operations. Both companies argued that all they wanted was to end duplication
of telecommunications equipment development. Overall, GEC was in a much
better position, with a large cash surplus and a strong management team headed
by Lord Weinstock. GEC was four times larger then Plessey, with an annual
revenue of 8.9 billion and 165,000 employees, compared to Plessey's 2.1

billion annual revenue and 37,000 employees.
GEC's bid for Plessey divided the government. Both GEC and the Depart-
ment of Trade and Industry argued that it would require a large company to
compete in world markets. But the Monopolies and Mergers Commission and
the Ministry of Defensea major customerobjected. The commission con-
cluded that although the merger would result in cost savings for System X
production, it would reduce competition for PBX transmission equipment and
traffic control systems. Moreover, the commission found that the System X
rationalization could be achieved without a full-scale merger between the firms,
but with a merger of only the System X part of their business. Based on a more
sober assessment of the export potential, it found excess capacity in System X
production and did not believe that the loss of competition between these two
System X manufacturers would outweigh the cost savings of rationalization.
Thus, it virtually recommended a joining of System X interests, but opposed
approval of the overall merger (Monopolies and Mergers Commission et al.,
1986). Plessey and GEC's telecommunications interests subsequently merged
in 1987 to become GPT. Both companies maintained equal ownership in the
joint venture.
In early 1989, GEC itself became the target of merger talks, with various
interested parties involved in purchase discussions, including Plessey, STC,
Siemens, and others. Later that year, GEC combined with Siemens in a suc-
cessful hostile takeover of Plessey, despite protests from its largest customer,
BT. Hostile takeovers in Europe are extremely rare. Political concerns were
also raised over Plessey's defense contracts and radar systems falling under
Siemens' control. Siemens was expected to gradually phase out GPT's System
X technology, and the partnership was not an easy one.
The other major telecommunications equipment supplier is the Standard
Telephone Company (STC), for many years a subsidiary of ITT (Young, 1983).
STC was originally the British manufacturing outpost of AT&T. In 1925, it
was sold to ITT, at that time a fledgling outfit of two Virgin Islands brothers
with great ambitions. AT&T was under pressure in the United States for having
international operations that, it was claimed, were subsidized by U.S. tele-
phone subscribers. A deal was struck for $30 million, the same price for which
the United States had purchased the Virgin Islands only eight years earlier. The
name of the British subsidiary company was changed to Standard Telephone
and Cable, Ltd. STC became a mainstay of British telecommunications manu-
facturing. One of its specialties is submarine cables, where the company holds
a large share of the world market. It designs, manufactures, and installs these
cables and has moved into use of fiber-optic technology. The company also has
a major presence in maritime radio.
In the 1970s and 1980s, successive parts of STC were offered by ITT to the
public, and by 1982 ITT's stake was only 35 percent and STC had become, at
least briefly, a British-controlled company. ITT withdrew partly because it needed
the money for operations and acquisitions and partly because STC had always
been seen as an American Trojan horse, a liability in a politicized business
such as telecommunications equipment. The company minimized the public

visbility of its ITT connection.
With Plessey and GEC, STC was a partner in the original development con-
sortium for the digital-exchange System X. But it was later dropped, partly
because of its American ties and the conflicts with the ITT System 12. In
recompense, it received a large contract for the less advanced TXE4 switches.
In 1984, STC acquired the major British computer manufacturer ICL for
about $550 million after quietly buying up about 10 percent of the shares and
then bidding for the company. The British government feared that significant
British computer technology would leak into the United States through ITT.
Thus, a separation of technology sharing was arranged. STC's acquisition of
ICL was also sensitive since ICL represented the British government's major
efforts of promoting a British computer industry. Only three years earlier, the
British government provided ICL with a 200 million loan guarantee to rescue
it. It also favored this company in British governmental procurement contracts.
Part of ICL's problem was to move from being a classical technology-driven
firm to a demand-oriented firm concentrating on specialized systems.
To finance the ICL acquisition, STC issued securities which dropped disas-
trously from 190 pence in the spring of 1985 to 86 pence a few weeks later.
Amid the unfavorable publicity from the London financial community, STC's
chairman and chief executive resigned. Northern Telecom acquired first 27 per-
cent of STC, and the rest of the firm in 1990. It then sold 80 percent of ICL
for 743 million to Japan's Fujitsu.
Thorn-EMI is another main electronics manufacturer. In the equipment mar-
ket, in partnership with the Swedish manufacturer L. M. Ericsson, it offered
Ericsson's AXE switch. That switch, modified as "System Y," became the
second source for BT, next to the System X. BT chose Thorn-Ericsson to
ensure itself a consistent supply of switches and to exert competitive pressure
on its System X suppliers, GEC and Plessey. Thorn-Ericsson emerged from a
pool of eight international candidates to fill this role.
Telecommunications Services
The introduction of Marconi's wireless telegraphy around the turn of the cen-
tury greatly worried those who held shares in British submarine cables. The
Post Office, which after 1889 operated submarine cables between Britain and
its European neighbors, also felt negative effects. In an effort to forestall the
development of radio, the Wireless Telegraphy Act of 1904 was passed, pro-
hibiting installation and operation of wireless telegraphy stations without li-
cense by the postmaster general. Although proponents of the Act claimed they
were acknowledging the importance of wireless to national security, it is clear
that the government feared the emergence of a new monopoly, which it would
then have to purchase. An application in 1905 by the private General Interna-
tional Telegraph and Telephone company for a commercial license for wireless
service within Britain was refused because it was in direct competition with the
government telegraph monopoly. The government eventually signed agree-

ments with the Marconi interests for wireless transmission to North America
and Italy. In 1920, long-distance ship-to-shore radio telegraph service was in-
troduced; in 1927, long-wave radio telephone service to the United States was
begun; and in 1932, ultra-shortwave radio telephone links were put into do-
mestic service.
Following World War II, Britain greatly increased its international telecom-
munications links. In 1956, the first transatlantic telephone cable connected
Britain with Canada. Six years later, the first satellite ground station was opened
to connect with Telstar. Domestic direct-dial long-distance service was intro-
duced in 1958, and in 1963, international direct-dial service was introduced. In
1976, the first TXE4, a large electronic telephone exchange, began operating.
Two years later, Europe's first fiber optic cable link was installed as an integral
part of the public network. International traffic accounts for 1.3 percent of
U.K. revenues and nearly doubled from 1983 to 1988. The most frequently
called nation is the United States which receives 23.3 percent of calls, far
ahead of Germany (10.9 percent) and France (9.5 percent) (Staple, 1990, p.
30).
Given its island location, it is not surprising that Britain is a major user of
submarine cable. In 1986, there were twenty-five cables operating between Britain
and the Continent. The first submarine fiber-optic cable began service in 1985
between the United Kingdom and Belgium, and operating at 280 Mbps. Seven
submarine cables connect Britain with the Irish Republic, the Channel Isles and
other offshore islands. Britain also has two cables to Canada, two cables to the
United States, and shares five other cables to the United States.
From the beginning, Britain was a primary site for satellite communications.
A satellite earth station was constructed in Goonhilly, selected for its far west-
ern locationimportant for the earlier low-altitude satellite. The Goonhilly sta-
tion aerials are vast, up to 30 meters in diameter. A second station, Madley, is
even larger. For both locations, Marconi was the prime contractor.3
The main data service on the public switched network is Datex, which has
medium speed. Packet switching was introduced in 1981. Private circuits for
high-bit rate digital services exist under the names of Kilostream, Megastream,
and Satstream. With Finland, BT was the first in Europe to implement video-
tex, which is described in greater detail in Chapter 36. But in 1990, only 95,000
terminals were hooked up to BT's Prestel system, and 55 percent of those
connections were by business users (Woollacott, 1990b, p. 41). BT was first
with videoconferencing and was an early entrant into third-party-provided mass
message service. Public facsimile service is offered via Bureaufax in coopera-
tion with the Post Office.
Among European countries, Britain has the lowest cost for the lease of a
private circuit from Europe to the United States. A 1985 study found the line
cost to be twice as high in West Germany as in the United Kingdom, 48,000
versus 24,000 (Garnham, 1985). By 1990 the gap had narrowed somewhat,
with the cost of a voice-grade circuit over 50 percent higher to Germany than
to the United Kingdom.
BT began ISDN service on a trial basis in 1985. The service, Integrated

Digital Access, had limited applications but was aimed at determining user
requirements for eventual ISDN service. BT upgraded its IDA service in 1990
with ISDN 2, a CCITT-compatible service offering 2B + D channels. BT's ISDN
2 tariffs, like those in France and Germany, are priced the same as regular calls
but include a one-time connection charge and yearly rental fees.
As in all developed countries, telegraph service is in decline. In 1983 alone,
international telegraph service decreased by 23 percent, and telegram service
lost 20 million. A large computer-based international telex exchange opened
in 1982. London became Europe's telex hub because its rates to North America
were substantially lower than those on the Continent. Telex bureaus emerged
to handle the retransmission of telex traffic from Europe to North America.
The continental PTTs opposed this process, leading to the important British
Telecom decision of the European Court of Justice, In 1988, the United King-
dom had the most telex terminals (111,500) and lowest rates of any European
country (ITU, 1990).
"Chatlines" became popular after 1986, but they also caused consumer pro-
tection problems. In response to consumer criticism, Oftel imposed restrictions
in 1988 that effectively terminated such services. Complaints stemmed from
both the high cost and the nature of these services. Without itemized billing,
consumers could not keep track of the number or type of such calls; employees
could amass large bills for their employers, as could children for their parents.
To curb such abuses, the corporate consumer was forced to incur the expense
of devices that block calls to specific exchanges. These devices, however, were
not available to residential consumers. In addition, message content also proved
to be a source of controversy. Many consumers felt that Oftel should take
action against message services whose content could be characterized as soft
pornography. At the same time, however, Oftel also received complaints that
some advertisements for message services suggested pornographic content without
delivering it.
Addressing these complaints, Professor Carsberg of Oftel encouraged the
establishment of the Independent Committee for the Supervision of Telephone
Information Services (ICSTIS) in 1986 to monitor message content. In addi-
tion, in July 1988, Carsberg advocated controls on Chatline services that would
provide access to these lines only to consumers who actively requested them
(Oftel, 1988). In 1989, Oftel instituted Codes of Practice for Chatlines, which
restricted access to minors and taxed service providers to establish a fund for
subscribers who faced large bills because of unauthorized use.
Telecommunications Service Improvements

After privatization and reorganization, BT embarked on a long-term moderni-
zation program estimated at over 20 billion. But it found the sailing rough.
System X switches were delayed and flawed in several London installations.
Oftel generally received so many complaints that it considered instituting dam-
Table 8.1 Performance Comparison of New York Telephone Co. and

British Telecom
New York Telephone British Telecom

Operator response Average, 4 seconds 87% within 15 seconds
Long distance blocking <l% 3.6%
Service orders filled 92% within 5 business days 62.2% within 8 business days
Complaints to Company per Line 0.04 0.22
Complaints cleared 75-80% within 24 hours 74% within 5 hours; 90.2%
within 2 days
Source: BT, and communication to the author by NYT. I988.
age liability against BT. A BT line averaged a technical problem every two
years, ten times the rate of the Bell companies in the United States. Even BT
conceded the fault rate to be two to three times higher than that in the United
States (Hudson, 1987).
Overall penetration in 1988 was forty-two main lines per 100 households.
As a private firm subject to some competition, BT's sensitivity to its customers
increased. Business customers, in particular, benefitted from service improve-
ments and rate reductions. The waiting time required to install private circuits
in the City of London, used to be several months, was rapidly shortened. Per-
formance profit centers were created within the company. Management em-
ployment contracts began to include performance and profitability clauses and
were limited in duration in order to break a civil-service, lock-step salary en-
vironment. A sales force with major-account managers was created in 1980 to
help the company protect and increase its business.
The U.K.'s liberal telecommunications policy and relatively lower interna-
tional telephone rates helped attract large users. The Ford Motor Company, for
example, set up the communications center for its European operations in the
United Kingdom. Nevertheless, there was much dissatisfaction with BT's ser-
vice. In an effort to rebut the criticism directed at it for service quality, in 1987
BT announced a quality strategy that included expenditures of over $400 mil-
lion to improve both international and domestic service. In response to the
complaints, Oftel resumed its publication of survey figures on service quality.
Oftel's 1990 Annual Report noted and increase in complaints, from 23,800 in
1988 to 31,650 in 1989, despite increases in percentage of faults repaired in
one day from 65 percent to 86 percent (Oftel, 1990a, p. 8).
Table 8.1 shows the comparison in quality of service between British Tele-
com and New York Telephone. The quality performance of New York Tele-
phone does not rank high among the more than thirty telephone companies in
New York State. However, Table 8.1 shows that its service quality is above
BT's.
According to Sir Brian Carsberg, head of Oftel, "The largest area of com-
plaints is billing. People say they could not have run up the amount that ap-
pears on their bill . . . In the United States where they have itemized billing,
this is a relatively small area of complaint" (Connections, 1985). BT began to

offer itemized billing when it introduced the System X digital switches, but set
an extra charge for this service.
The service quality of public coin telephones was poor. In 1983, BT pro-
vided more than 350,000 pay phones, many of them rented on private prem-
ises. BT was not enthusiastic about this service. According to its accounting,
it lost more than 77 million on it in 1982-1983, about 20 percent of its total
profits. The requirements of BT's license led it to upgrade service. The com-
pany embarked on making the instruments more vandal-proof, a goal that is
particularly important, since in some areas vandalism accounted for 75 percent
of the cost of pay phone maintenance.
Nevertheless, the state of public telephones was for a long time the single
worst aspect of British telephone service. In response to a 1985 survey by the
Daily Mail that showed almost 60 percent of public telephones out of order at
any given time, Oftel commissioned its own study, which found a still extraor-
dinary rate of 50 percent. Over two years of effort aimed at improving this
statistic produced progress. At the end of 1987, Oftel found 23 percent of
public phones out of order, a number that declined to less than 10 percent by
mid-1988 (Oftel, 1988). Another area of complaint about BT concerned its
rates. Local telephone charges in 1988 were among the highest in Europe,
although business and residential service baskets were slightly below OECD
averages. In 1990, BT raised residential rates 9 percent after Oftel refused
steeper increases. Business rates remained unchanged, but line rental charges
for home users increased 12 percent. In 1989, Oftel began an investigation into
excessive international rates.
BT followed an increasingly international strategy and expanded its global
presence by buying in 1989 a 22 percent stake in the second-largest U.S. cel-
lular service provider, McCaw, for $1.5 billion, thereby gaining access to 30
percent of the U.S. mobile market. It also bought an 80 percent stake in the
paging firm Metrocast in 1988, and it acquired McDonnell Douglas's data com-
munications operations, Tymnet, for $355 million in 1989. In 1986, it acquired
Dialcom from ITT. Through Tymnet, BT also purchased Edinet, Ontyme, and
a 25 percent stake in Japan's Network Information Service Co. BT also bid for
a German D2 cellular license. Soon, BT's international investments led to crit-
icism of its modernization programs in the United Kingdom as inadequate.
Furthermore, the fate of its global investments was mixed. Metrocast had to be
closed; McCaw's debts escalated to over $5 billion, and Mitel was heavily in
the red. BT shifted away from owning equipment and further foreign carriers,
and into the servicing of major users and their global networks. It pursued
various international alliances and established Syncordia, an Atlanta-based
managed data service.
Despite some liberalization of entry, BT still vastly predominates almost all
British telecommunications. Although in the past it played no role in manufac-
turing, its acquisitions after privatization indicate a trend toward vertical inte-
gration.
Mercury's status was reviewed in the 1990 'duopoly review.' After Oftel
submitted a report in 1990, the review was opened to public comment. Over
200 groups responded. In 1991 the Department of Trade and Industry, which
oversees Oftel, presented its findings in Competition and Choice: Telecommu-
nications Policy for the 1990s, a title reminiscent of the government's parallel
efforts in television liberalization through the broadcasting White Paper (Home
Office, 1988). The duopoly review decision signalled a move from managed
competition to a more open system, but with a more level playing field.
Equal access provisions were envisioned only in two stages, and not before
the 1992/3 review of BT tariffs, since BT (and some user groups) had argued
that the costs (300 million) and technical problems of developing switching
capability for multiple operators were extensive. The decision also changed
British Telecom's tariff structure; it included international services in the tariff
basket after a one-time 10 percent reduction, and it tightened the price cap to
RPI-6.25 percent from RPI-4.5 percent. While cable, satellite, and cellular firms
received greater flexibility (and interconnectivity) to offer telecommunications
services, there was no symmetry. BT was prohibited from offering cable tele-
vision service until at least 1997.
The reaction to the White Paper was mixed. Some welcomed its further
liberalization and viewed it as a challenge to BT's market power. Others thought
it too lenient on BT, perhaps in order to protect the value of government shares
prior to full privatization. They also saw Mercury as the loser since it faced
new competition in its core business of serving large users. BT also received
the right to grant special tariff packages to large users. On the other hand, the
consortia which were expected to offer telecommunications services (such as
British Rail, British Waterways, Racal, British Aerospace, and the Post Office)
were hampered by the unresolved interconnection issues. And few cable oper-
ators actually had networks in place that might have offered local services.
Thus, British telecommunications policies have created an entity similar to
the predivestiture AT&T in the United Statesa private, dominant, and regu-
lated carrier with a limited competition in the long distance field. Such a system
had been considered problematic in the United States (even without BT's do-
mestic quality problems), and led to further liberalization and divestiture. Anal-
ogously it was unlikely that the transformation of British telecommunications
had reached a stable equilibrium.
9
France
Early History of French Telecommunications
Throughout the 1980s, no West European government controlled telecommu-

nications more comprehensively than France; French state organizations were
the principal operators, developers, customers, export promoters, and, for a
time, primary equipment producers. This statist role has a long history. Orga-
nized French communications date back to the thirteenth century, when the
University of Paris operated a messenger service. In 1464, Louis XI established
a government courier service system for official use, and by 1622, there were
also regular postal routes for the public. In 1725, the government declared the
entire transportation system, including postal service, to be the exclusive do-
main of the state. Since most private operators leased postal routes from the
state, postal routes became a significant source of state revenues. In 1790, the
transportation of persons and parcels, but not of letters, was partially liberal-
ized. In 1794 (the third year of the Republic), interested parties were allowed
to provide postal services, except for the delivery of letters and small packages.
However, after losing postal service revenues to the emerging competition, the
government soon imposed a 10 percent tax on the receipt of private posts
and raised letter rates considerably, in some instances by more than 500
percent.
Always in need of revenue, Napoleon reestablished the governmental mo-
nopoly in 1804 by a decree that lowered postal rates somewhat but kept them
considerably above the original rates. Napoleon also centralized the postal ad-
ministration and established a system of postal inspectors in every department.
This governmental monopoly was zealously enforced, with 750 violations pros-
ecuted in 1805 alone.
The high rates, which the Bourbons maintained after the Restoration, pro-
vided considerable profits for the state. In 1847, profits were FFr 17.8 million
on expenses of FFr 35.5 million. According to Heinrich Stephan, later Ger-
many's postmaster general, the French profits were the highest of all European
postal administrations (Stephan, 1859, p. 630).
In the eighteenth century, French, German, and British inventors tried to
create reliable high-speed communications. Claude Chappe, a one-time sem-
inarian and later a physicist, developed a novel optical telegraph system; in
1792, his brother, Ignace, a member of various revolutionary bodies, success-
133
134 France
fully introduced a proposal to initiate a route from Paris to Lille. Chappe called
his invention the tachygraphe ("rapid writer"), but this fortunately was changed
into the term telegraphe for "distance writer," and this Graeco-Latin desig-
nation has stuck.
The Paris-Lille telegraph route was completed in 1794. Signals were con-
veyed by various positions of mechanical arms on windmill-like stations lo-
cated on hilltops at intervals of 6-12 miles. In 1803, the route reached Brussels
and a few years later, Amsterdam. In 1798, another Chappe brother established
a line to the German and Swiss borders. Lines were extended to Milan in 1805,
and Venice in 1810. The Napoleonic Wars made rapid communications impor-
tant. After the Bourbon Restoration, the systems continued to expand within
France. By 1845, the system encompassed 535 optical telegraph stations link-
ing Paris with twenty-nine cities. The Chappe optical telegraph system
and variant systems were also used in England, Scandinavia, Prussia, and
Austria.
A message from Paris to Lille, a distance of 225 kilometers, was relayed by
twenty-two stations and required, in theory, only several minutes for transmis-
sion. But the Chappe system was complicated and required a skilled operator,
usually trained and supervised by military officers. The British optical system
was simpler but slower. Signals on the Chappe telegraph could be transmitted
at a rate of one signal every sixteen seconds. Because of this time-consuming
procedure, elaborate codes were developed for standard phrases and expres-
sions. Codes also served to protect the secrecy of the official messages. Trans-
mission at night or in bad weather was not possible and the system was plagued
with problems. It was exceptional for a telegraph line to transmit even six
telegrams of about twenty words a day. The entirety of this network could
handle a maximum of 7000 dispatches annually. It was therefore not surprising
that military, diplomatic, and administrative messages of the state claimed ab-
solute priority and that the system was operated by a board of the Ministry of
the Interior. Initially, no private messages at all were transmitted on the sys-
tem. In 1833, Alexander Ferrier created a private company to respond to pri-
vate demand. Although his effort failed, it demonstrated the interest in devel-
oping a private telegraph.
In 1837, such private efforts were outlawed by a law that declared the tele-
graph a government monopoly within the expanding postal monopoly. Private
communications networks were discouraged primarily to inhibit the exploitation
of financial information by speculators. A French legislator argued: "Govern-
ments have always kept to themselves the exclusive use of things which, if
fallen into bad hands, could threaten public and private safety: poisons and
explosive are given out only under the state authority, and certainly the tele-
graph, in bad hands, could become the most dangerous weapon. Just imagine
what could have happened if the passing success of the Lyon silk workers
insurrection had been known in all corners of the nation at once" (Brock,
1981). The law provided substantial fines for unauthorized transmission. Suc-
ceeding governments of varying political persuasions have never significantly
France 135
deviated from this principle. The French government monopoly over telecom-
munications was thus established by law even before the introduction of electric
telegraphy. When the latter arrived, the French government was far from en-
thusiastic. An explicit ban on private electric telegraph lines was quickly de-
clared when the advent of the railroad led to an attempt to create a private
telegraph line between Versailles and Saint Germain. Although private entry
was thus made impossible, the government was reluctant to enter electrical
telegraphy itself because it feared undermining its own elaborate optical tele-
graphic system. The advent of the electric telegraph threatened state power, as
one minister complained: "No, the electric telegram is not a sound invention.
It will be always at the mercy of the slightest disruption, wild youth, drunk-
ards, bums, etc. . . . All that is unnecessary with the electric telegraph are
those destructive elements within only a few meters to a wire over which su-
pervision is impossible. . . . The visual telegram, on the contrary, has its
tower, its high walls, its gate well guarded from inside by strong armed men"
(Allentier, 1973, p. 100).
Eventually, however, the French government constructed an electrical tele-
graphic system which was operated and controlled by a unit of the Interior
Ministry until 1878, primarily for governmental and only secondarily for public
use. In 1853 the optical telegraph was discontinued, its low capacity and rela-
tively high cost of operation being factors in its rapid demise.
In 1852, in an important departure, Louis Napoleon Bonaparte (Napoleon
III) allowed private users to access the telegraph system. Control was gradually
relaxed, though state control was still used against opposition newspapers and
coded messages (Bertho et al., 1984), to name two examples.
When it came to international submarine telegraph cables, England had, until
1880, virtual control, and French colonies often had to use English lines to
communicate with Paris. Given her rivalry with Britain, France found such
dependence intolerable. Meanwhile, several attempts by French companies to
enter the submarine cable market ended in failure or in buyouts by American
or English companies. The Ministry of Posts and Telegraph began to invest in
several cable companies. After 1879, the government operated its own cables
to Algeria and Africa. The Compagnie Franchise au Cable Telegraphique (CFCT)
was formed in 1889 and operated privately until its nationalization in 1945.
The Establishment of Telephony
The telephone debuted in France at the Paris World Fair of 1878 (Holcombe,
1911). In contrast to Germany, where it was the state that seized upon the new
invention with enthusiasm, in France the private sector played a significant role
in the establishment of the new medium. The Third Republic was pushing for
economic liberalism in areas of public services, usually through concessions to
private companies which would shoulder the burden of investment and risk.
The telephone was included in this framework, and in 1879 the government
136 France
announced its decision to award private concessions. But all construction had
to be supervised by state engineers.
Concessions were not exclusive and lasted only five years; a 10 percent roy-
alty payment was required, and the government could purchase all the tele-
phone equipment it needed from the concessionaires at an agreed upon or ar-
bitrated price. At the same time, there was no rate regulation. The concessionaires
thus had both the incentives and the ability to try to recoup their investment as
quickly as possible. In 1879, licenses were issued to the firms of Edison, Gower,
and Blake-Bell. Before construction began, however, these franchises were
merged into the Societe Generale de Telephone (SGT). This process of consol-
idation was based on cartel advantages as well as the Paris municipality's wish
to avoid multiple and unsightly wire networks.
The first Paris exchange opened in 1881. Development however, was slow.
Service was limited in quality and expensive. The two viable policy options
were either to increase the government's role or to allow telephone companies
to expand and develop. The government instead chose to increase restrictions
on private operations, without strengthening its own role as an operator by
committing resources.
In 1882, the government determined that the SGT was inclined to provide
services only to the dozen or so large French cities. The National Assembly
was persuaded to support governmental telephone construction in various
medium-sized cities. But the annual budget allocation for this endeavor was a
paltry FFr 250,000.
In 1884, the French government renewed the existing private licenses over
strong opposition, but without a clear-cut policy of its own. Soon it became
desirable to link the local exchanges by long-distance connections. The tele-
graph authority, sensing a threat, began construction, starting in 1885 between
Paris and Rouen, and in 1887 between Paris and Brussels.
In subsequent years, French policy continually changed. In 1887, Grannet,
the minister in charge of telegraphy, introduced legislation to strengthen private
telephony to alleviate the backward condition of French telephony. His succes-
sor, however, opposed the plan. With some justification, the public believed
the French telephone system had become a complete morass in less than ten
years. By 1889, the Grannet plan was rejected, largely on the argument that
private telephony would jeopardize the financial soundness of the state tele-
graph system (Holcombe, 1911).
Backed by a coalition of dissatisfied business users, small towns, and leftist
republicans, the government soon decided to take over the entire network. Op-
ponents argued against too much state power and against support for what was
considered a luxury. But the French National Assembly approved nationaliza-
tion in 1889, at the end of the second five-year concession period, fired by the
national enthusiasm of the revolution's centennial. The companies' installers
were given the option of setting up installation businesses of their own, origi-
nating a system of private installation firms still in use today, and whose origin
is thus not a liberalization but a nationalization. 1
France 137
Nationalization and Its Aftermath
Since the SGT refused to surrender its property willingly, it was literally taken
over by force, with compensation paid after the fact. At the time, there were
only 8500 subscribers, of which only 2000 were outside Paris (Nouvion, 1984).
After extensive court action, the company received about FFr 11 million, twice
the amount the telegraph authority had been willing to pay, but less than the
company had demanded.
Having taken control of the telephone and telegraph, and having placed it
under the Office of the Under secretary of State for the Postal and Telegraph
Service, the government now had to figure out what to do with it. Because the
telephone held no priority in general economic development, the government
was not prepared to make the significant investment necessary to finance con-
struction. It had, over two generations, built up an extraordinarily high general
debt burden and was reluctant to increase it.
Instead, it devised a system in which potential subscribers and municipalities
were forced to extend interest-free loans that would eventually be repaid from
the profits derived from their own receipts. Futhermore, subscribers were re-
quired to purchase telephone sets themselves to save the state's money. This
was the origin of France's subscriber equipment policy, which was presented
in the 1980s as an example of liberalism.
This system of financing was also expanded to long-distance transmission.
Local systems resembled cooperatives in that they united the first group of
subscribers, who paid for the construction of the network. As in other cooper-
ative ventures, the admission of newcomers and the potential integration of
systems into the national network presented problems.
Financing proved problematic when it came to the replacement of obsolete
equipment or making improvements. By 1900, this situation was referred to as
a "telephone crisis." The system was congested, unreliable, and expensive.
There were still only 30,000 telephones in the entire country! By comparison,
in 1909, there were 27,000 telephone lines alone in the 100 largest hotels of
New York City (Attali and Stourdze, 1977, p. 106).
A number of improvements were achieved under A. Millerand, a Socialist
who became minister of the PTT. Millerand established a new system of fi-
nancing telephone expansion out of public revenue rather than subscriber charges,
but he was unsuccessful in obtaining appropriations from the legislature. Even
the French business community shortsightedly opposed additional budget allo-
cations. The Chamber of Commerce argued that construction had to be financed
not by government budget but by greater internal economies of PTT operations.
Millerand was dropped from the cabinet.
In the early years of the century, service remained abysmal. There was only
one line between Paris and Marseilles, and during one twelve-month period
(1905-1906), it had 204 interruptions with an average duration of 14.5 hours.
Between Paris and Lyon there were five lines, with 550 interruptions of an
138 France
average duration of ten and a half hours (Holcombe, 1911, p. 302). Local
exchanges were enormously congested. In 1905, operators in Paris took almost
two minutes on average to make a local connection.
By 1906, the government acknowledged the problem and a special law passed
to transfer the Post and Telegraph Department from the Ministry of Commerce
and Industry to the Ministry of Public Works. It also authorized a budget of 19
million francs but required that the funds be spent within the same fiscal year,
leading to insufficient planning and excessive costs. Millerand, back in the
government, instituted a labor relations reform for PTT workers, featuring eight-
hour days, security of tenure, overtime payment, and full payment during ill-
ness.
The destructiveness of World War I further deteriorated the telephone net-
work. Industry was not up to the task of rebuilding it. French equipment mak-
ers were merely manufacturers of general electrical machinery, small compa-
nies, or subsidiaries of foreign firms. In 1920, when the Paris network was to
be upgraded, several foreign firmsWestern Electric, Siemens, Ericsson, and
ITTsought the business. ITT bought two French manufacturing firms, created
a large research laboratory in Paris, and studied the needs of the French net-
work in depth. As a result of its commitment, it was awarded the major French
orders for automatic central offices. But this foreign dominance created resent-
ment in the 1930s and after World War II, leading to a French industrial buildup
against ITT's position.
The first automatic exchange opened experimentally in 1913 in Nice using
the Strowger system. The PTT also ordered semiautomatic systems from the
French company Le Materiel Telephonique (LMT) then owned by AT&T and
using Western Electric's rotary technology. The first such exchange opened in
Angers in 1915 (Nouvion, 1984, p. 80). However, the spread of the automatic
exchange was slowed by World War I and the economic problems in its after-
math. Crossbar exchanges were not introduced in Paris until 1964. Six years
later, the electronic time division switch, the E10, was introduced, leapfrog-
ging a generation of technology.
During the period of economic conservatism of the 1920s, the magnitude of
the investments required for the telephone network sparked a debate over its
denationalization. Two proposals emerged from the debate: ITT, eager to put
down roots in the French market, offered to take over and operate the entire
national network as it did in Spain; the second proposal was from the Societe
Industrielle du Telephone (SIT), a predecessor of the telecommunications op-
erations of CGE.
In 1921, a law was advanced to study denationalization. Louis Deschamps,
who advocated privatization of various government monopolies, including the
telephone, served briefly as PTT minister. Because of the project's difficulty
and opposition from PTT employee associations, his successors did not pursue
his course. From 1923 onward, the PTT was instead directed to operate like an
administration with industrial and commercial purposes, under a separate bud-
get (subject to parliamentary approval) and required to cover its costs.
In 1920, the PTT awarded the concession for international radio-electric links
France 139
from France to the private company Compagnie Sans Fils (CSF), which created
Radio France for this purpose. The concession to CSF caused a vigorous de-
bate. Because the international radio-electric link posed promising opportuni-
ties, the labor unions and the political opposition were against privatizing this
sector.
FIT relations with Radio France were generally poor because radio links of
the PTT and Radio France (SRF) were at times in competition. A FFr 60 mil-
lion station opened in 1923, allowing France to build a wireless network that
included PTT links to Africa and Indochina, communications links to central
and northern Europe, and Radio France transmission to most of the world.
Meanwhile, the PTT slowly upgraded the domestic telephone network, with
the first long distance cables in France using Pupin coils put into service in
1924. Still, the average wait for an interurban connection was five hours. To
address this problem, the government adopted a separate budget for the PTT as
an annex to the general budget, giving it access to special loans. In addition,
it approved another modernization plan for the telephone network. Between
1924 and 1934, the number of subscribers increased annually by an average of
45,000, almost double the previous rate, though growth was still excruciatingly
slow.
The automation of the Paris network started in 1925 and led to major indus-
trial struggles over procurement contracts. The competing firms included the
French Compagnie des Telephones Thomson-Houston, and the Societe Indus-
trielle du Telephone (SIT), both of which proposed a Strowger system. LMT,
now acquired by ITT, proposed a rotary system, and the Ericsson subsidiary
proposed its own method. The French Compagnie Generate de Telegraphic et
de Telephonic was also in contention, but its system depended on a Siemens
patent, and the French government did not want Parisian telephones to rely on
German technology.
At the time, ITT had experienced several setbacks in Europe: Germany and
England chose the Strowger system, and although Spain opted for exclusive
use of ITT's rotary system in 1924, the Spanish could have changed their minds
if the French had not also chosen that system. With so much at stake, ITT
hedged its bets and bought the Thomson-Houston telephone division in 1925
(later to become CGCT), giving it a stronger position for the French procure-
ment. This was accomplished just as Thomson-Houston patented its new R6
switching system.
The PTT awarded the project to LMT (the ITT subsidiary) in 1926. The
firm's promise to build a large manufacturing facility and to release its pro-
cesses to other companies designated by the PTT weighed heavily in the deci-
sion. These companies were the Societe Grammont, which withdrew in 1931
because of financial problems, and Ericsson.
The ITT contract, however, did not eliminate demand for the R6 system.
Rural exchange contracts went to Thomson-Houston, ITT's other subsidiary,
which was considered "more French" than LMT. The PTT also extended a
license for the R6 to SIT. This period marked the low point for French switch-
ing firms and the high point for ITT. Through its two subsidiaries it dominated
140 France
this key sector. A few years later, in 1932, the large French electrical firm
Compagnie Generale d'Electricite (CGE) took control of SIT and sought un-
successfully to forge links with the American firm Automatic Electric of Chi-
cago. This marked the beginning of the CGE's long march against ITT. With
substantial help from the French government, it eventually took control of ITT's
worldwide telecommunications equipment operations in 1986, more than half a
century later.
Between 1924 and 1934, the number of lines increased an average of 7.6
percent annually. Because of budgetary restrictions, this rate fell to 2.6 percent
between 1935 and 1939. The number of lines served by automatic systems rose
from 3.6 percent in 1926 to 45.6 percent in 1938. The corresponding 1938
figure for Germany was 84.9 percent, and for the United Kingdom, 54 percent.
While telephone penetration in France rose, it was still very low: 3.7 per 100
inhabitants in 1938, compared to 4.6 in the UK, 15 in Germany (Nouvion,
1983, p. 84), and 15.1 in the United States (U.S. Department of Commerce,
1939). Even that figure for France overstates the actual situation, because of
the imbalance in favor of Paris. Only a handful of the French departments had
a telephone density above five telephones per 100 inhabitants. In many, the
density was less than two.
World War II destroyed parts of the French telephone network, though some
of the long distance telephone network had been expanded by the German oc-
cupiers for military use (Bertho et al., 1984). From 1941 to 1944, the Vichy
government consolidated various research centers into one organization, the
CNET, as part of a modernization effort. After Liberation, telecommunications
fell behind the rest of Europe in rebuilding because of inadequate investment
funds. The network became a drag on France's otherwise rapidly developing
economy. This is not to say that there was no innovation; there was progress
in switching, transmission, and telex. France was the first European country to
use coaxial cable for long distance trunks (Paris to Toulouse, in 1947). France
was also first to use such a cable for a totally automatic service, starting in
1952. Yet demand for telephone lines vastly outstripped supply. Telex ad-
vanced partly through the exertions of the firm Sagem and the liberal equipment
policy for telex and PBXs. Complaints became endemic.
In 1969, the PTT admitted that the average long-distance call had to be
placed three times before it found a clear circuit, an efficiency rate of less than
40 percent. According to a Western Electric report, only one call out of four
during business hours was completed on the first attempt. The report also claimed
that in 1969 "it was not unusual for a telephone subscriber to wait anywhere
from 30 minutes to more than an hour and a half for a dial tone and more than
two days to get a call through" (Western Electric, 1978, p. 1). Under pressure
to improve this performance, the government budgeted FFr 45 billion for tele-
communications development in its Sixth Plan (1971-1975). This amount was
ten times greater than that provided in the Fourth Plan, but it was still far from
adequate.
Telephone distribution was also quite unevenly spread across the country.
Figure 9.1, from a Direction General de Telecommunications (DGT) staff study,
France 141
Telephone density per 100 inhabitants

Figure 9.1. Telephone Diversity and Usage in France: Paris versus the Regions.
(Source: Guerard et al., 1979.)
shows the vast difference in telephone density between Paris and other cities
and regions of the country (Guerard et al., 1979).
On the horizontal axis, the telephone density is mapped. Vertically, tele-
phone usage per main line is traced for the year 1977. As can be seen, the
telephone density for Paris was almost twice that of major areas of activity in
the provinces, which were, in turn, much higher than in the countryside.
The Great Leap Forward
By the mid 1970s, French telephone performance had deteriorated to the point
that it became one of the campaign issues in the 1974 presidential election that
brought to power Valery Giscard d'Estaing, a right-of-center finance minister
with a technocratic image and an interest in telephone issues. Giscard d'Estaing,
as an Assembly deputy, had introduced legislation in 1967 to establish the
telephone administration (DGT) as an independent entity. In 1974, the bitter
joke was that half of the country was waiting to get a telephone installed, and
the other half was waiting for a dial tone. Only 5 million lines served a popu-
lation of 52.6 million, whereas the United Kingdom, with roughly the same
population, had twice the number of telephone connections. In 1974, telephone
densities were extraordinarily low. For blue-collar workers they were only 10
percent; for junior executives and clerks, 22 percent; and for the agricultural
professions, 19 percent (Logica, 1979).
In 1975, the French government set a national priority of modernizing and
expanding the domestic network, while at the same time using telecommuni-
cations as a base to develop a high-technology industry strong enough to com-
pete in the world market. President Giscard d'Estaing launched a five year plan
aimed at giving the entire population access to telephone service, with empha-
142 France
sis on bringing the particularly low quality of telephone service in the provinces
in line with the standards enjoyed in Paris. The number of subscribers quadru-
pled in eight years to over 20 million in 1983, aided by the Vlth and Vllth
Plans, which allocated FFr 405 billion and FFr 120 billion, respectively, for
telecommunications (Scientific American, 1983).
Telecommunications were set as the first priority in the Vllth five-year plan
for the French economy. Its goals were to reach parity in density with West
Germany and the United Kingdom and to reduce the wait for a telephone from
16.4 months in 1974 to 0.5 months in 1982. In equipment, accelerated conver-
sion to electronic switching was planned, both for increasing the effectiveness
of the French system and as a basis for export (Connaughton, 1982).
With money and attention, French telecommunications were remarkably
transformed. For the first time in their history, they were a national priority
supported by the state and by now a largely indigenous industry. In 1976 alone,
2 million lines were added. This number increased to 2.7 million in 1979, and
almost 3 million in 1980. In 1981, the number of new installations began to
plateau and decline. This was an impressive performance, but less unusual in
the European context than is often believed. Other European countries were
also rapidly increasing their investment in telecommunications infrastructure at
a similar pace. French telecommunications investments, as a percentage of the
total national investment, were 3.36 percent in 1981. In Germany, without
telecommunications being proclaimed a national economic priority, it was 3.20
percent at the same time. Italy spent 3.26 percent, and the United Kingdom
spent 3.5 percent (Benedetti, 1983).
France Telecom
Until 1989, the Ministry of Post, Telegraph and Telephone (PTT, later PTE,
when it also included the space portfolio) controlled French telecommunica-
tions, operating through the Direction Generale de Telecommunications (DGT),
renamed France Telecom (FT) in 1988. Postal service was provided by another
PTE General Directorate, the DGP (later La Poste). The DGT, established as
an autonomous directorate in 1946, employed about 166,000 people (about 7.2
per 1000 lines) and was composed of twenty-two regional directorates, which
were in turn divided into operational subregions. Some of its operations were
carried out through a subsidiary or a separate entity, and others through a mixed
investment firm controlled by the state but constituted according to private
company law. In 1985, these subsidiaries were placed under the control of the
state holding company Compagnie Generale de Communication (COGECOM).
Two dozen of these subsidiaries operated in foreign countries or ran satellite
ground stations and submarine cables. In 1989, COGECOM had revenues of
more than $1.8 billion. Other subsidiaries operated in France, including Entre-
prise Generale des Telecommunications (EOT), for advanced terminal equip-
ment, and France Cables et Radio for new services to large users. Telesystemes
France 143
provides computer services and software. Transpac is the packet-switched data

network.
Before the reforms of 1990, the annual telecommunications budget was part
of the overall state budget and therefore had to be passed by the National
Assembly; long-term program authorization and advance credit approval, how-
ever, reduced the uncertainty of the process. In the 1960s, telecommunications
investment had to be self-financing, which proved inadequate, particularly when
rates were restricted and the number of lines was low. These limitations were
eased, and the self-financing ratio was reduced to about 40 percent. Telecom-
munications obtained financing through the public Caisse Nationale des Tele-
communications (CNT). It also obtained some financing from private finance
companies which purchased telecommunications equipment specified by the DGT
(later France Telecom) and then leased it to the latter for ten to fifteen year
periods; the equipment then reverts to the DGT upon payment of a 3 percent
residual value fee (Voge, 1986).
After 1982, the DGT was subject to large financial transfers to the general
government budget and was required to contribute to the French electronics
industry. French telephone users thus directly subsidized the electronics indus-
try, in addition to contributing indirectly through procurement and R&D sup-
port. In 1985, the DGT contributed $2.3 billion out of $12.2 billion in revenue
to nontelephone operations of the government, including the Post Office Direc-
torate, which received $0.5 billion, and the general budget, which received
$0.3 billion (Morley, 1986).
The Organic Law of 1959 provides for domestic tariffs to be set by cabinet
decree. Long distance traffic was heavily profitable whereas basic subscription
and local service was being subsidized. According to one study by DGT econ-
omists, 1983 revenues were FFr 8.7 billion for basic subscription, FFr 6.9
billion for local message units revenues, and FFr 20.5 billion for long-distance
service (Curien and Brunetiere, 1984). The ratios of revenues to costs were,
respectively, 0.4, 0.79 and 2.3. For a long time, local telephone traffic was
not time-measured, long-distance calls were highly differentiated by distance,
and off-peak pricing was rudimentary (Pautrat, 1984). After 1984, the network
has distinguished between four levels of price, depending on the time of day
(red, white, black, and light blue tariffs).
The DGT and later France Telecom did not manufacture equipment, but
supported and guided most of the telecommunications R&D in France through
the Centre National d'Etudes des Telecommunications (CNET). The CNET was
derived from PTT labs that had existed before World War II. CNET was cre-
ated to increase independence from foreign firms and provided an authority to
regulate the competition between PTT suppliers.
In 1957, a Department of Research on Electronic Machines was created within
CNET. Its purpose was to apply developments in electronics to telephone
switching. The first prototype of electronic switching appeared in the early
1960s and led directly to Alcatel's dominance, since that firm was closely allied
with CNET. In 1989, CNET had 1500 researchers and a budget of $350 million
(Steckel and Fossier, 1990, p. 38; Arlandis and Gille, 1989, p. 37).
144 France
Services
Partly because of its fairly recent major expansion, the French telephone system
achieves high technical standards. Development was particularly successful in
both digital switching and transmission. In 1989, 71 percent of all lines were
electronically switched, with the rest using crossbar switches; rotary type switches
had been totally phased out. 2 In 1989, nearly 4 million electronic lines were
put into service. By 1989, 58 percent of switches were digital, 12 percent
analog, and 30 percent crossbar. Total network digitalization was set for 1996
(Steckel and Fossier, 1990, p. 22).
France was a leader in public packet-switched networks. Available since 1978,
Transpac is the world's largest packet-switched network. Subscribers were es-
pecially attracted by the distance-independent tariffing. Private companies are
involved in the network, which the GCE subsidiary SESA designed. Transpac
was originally operated through a mixed public-private economic organization.
To spread risk, maintain commitment, and increase input, Transpac users were
also shareholders. Later only 3 percent remained in private hands. In 1985,
Transpac was placed under the control of the state holding company COGE-
COM.
Transpac had 5.5 million users (including the 5 million subscribers of the
French Minitel service) and 86,000 direct access points in 1991, carrying 2
billion characters per month. It also introduced a range of new services, includ-
ing Atlas 400, a X.400 electronic data interchange, and a mass market elec-
tronic mail system, Minicom. Transpac also moved internationally by acquiring
a 15 percent stake in the U.S. VAN Infonet, together with several European
PTTs. It entered ventures with the German DBP Telekom and the Swiss and
Danish PTTs. In 1991, Transpac began to offer service in the United Kingdom,
as part of an international expansion program. Other data services are Transfix,
Transcom (a medium-speed public switched service), and Transdyn (a satellite
service). These services were bundled and upgraded as Transmic.
France Telecom moved toward ISDN, known in French as RNIS (Reseau
Numerique a 1'Integration de Services). In 1987, the DGT introduced its ISDN
service Numeris in Brittany. By 1989, 18 million telephone subscribers had
potential access to Numeris and several dozen companies were developing ap-
plications in partnership with France Telecom.
Limited resale of leased line capacity was allowed for data services in 1987,
but tariffs were set to prevent arbitrage. Sources were divided into Category I
(under 3.5 Mbps) and Category II networks. All networks required a license.
By 1990, some 120 applications were filed but only one authorization was
granted for a Category II network. Restrictions on VANs were further lifted in
a 1989 regulatory ruling, which allowed automatic authorizations of Category
I (Roussel, 1989d, p. 13). VAN providers were still limited to a single connec-
tion to the public network, ruling out network bypass, and they must support
OSI protocols for data transmission (Steckel and Fossier, 1990, p. 10).
French policy toward VANs has been somewhat reluctant. When the Euro-
France 145
pean Community proposed in 1989 to liberalize data service provision, France

protested, suggesting that network operators be given authority to issue licenses
and levy access charges on private VANs operators. France also led the oppo-
sition to the B.C. Commission's Open Network Provision.
On the other hand, France is a leader in public videotex. Its Teletel system
is accessed by 5 million Minitel terminals installed and distributed free of charge
by France Telecom. This is discussed in detail in chapter 36 on European vid-
eotex.
In contrast to data transmission, where France is strong, mobile communica-
tions had a slow start. In 1982, cellular radio was introduced in Paris. In 1985,
the system still had only ten repeaters. Matra, in collaboration with France
Telecom, manufactures the systems.
In 1987, a second mobile communications license was awarded to a consor-
tium that included the water utility Compagnie Generale des Eaux (also active
in cable TV); Racal, of Britain; and Bell South, one of the seven regional
telephone holding companies in the United States. This consortium, the Societe
Franchise du Radiotelephone (SFR), installed a network in direct competition
with the France Telecom system, using equipment by Nokia and Alcatel
(Chamoux, 1988, p. 6).
In 1990, cellular penetration was still low, at 0.27 percent of the population,
compared with 1.25 percent in the United Kingdom and 3.68 percent in Swe-
den. France Telecom's Radiocom 2000 had over 90 percent of the 170,000
subscribers. The DRG threatened to intervene in 1990 if France Telecom could
not agree with SFR on interconnection arrangements to the GSM mobile stan-
dard.
France Telecom also introduced a telepoint cordless service (Pointel) based
on the U.K. CT2 system (Dawkins, 1990a). In paging services, a closely re-
lated duopoly exists. Telecom Systemes Mobiles Alphapage service grew to
200,000 in 1989. TDF-1 Radio Service, in which control was acquired by
France Telecom, had only 18,000 subscribers in 1989, partly because of the
high cost ($700) of its paging devices (Telecom France, 1990, p. 5). Both
services are offered through COGECOM, France Telecom's holding company.
Network improvements were aided by steady increases in productivity. Turn-
over per employee rose steadily by over 10 percent from 1984 to 1987 before
leveling off in 1988. France Telecom's productivity was consistently higher
than that of British Telecom (Arlandis and Gille, 1989, p.8).
The Restructuring of the Telecommunications Industry
As part of a high-technology renaissance, the Vllth Plan of 1975 set the devel-
opment of France as an international leader and exporter of telecommunications
equipment as a main goal of French telecommunications policy. Since the gov-
ernment wanted to retain domestic control over these key industries and the
major subsidies they received, government policy aimed to minimize foreign
competition in the electronics industry in France. Thus, the government finan-
146 France
cially backed the involuntary acquisition by Thomson-CSF of ITT's share in

LMT and 40 percent of Ericsson's French subsidiary, SFT. This restructuring
in 1976 raised the domestic equipment market share of French firms from 43
to 71 percent; and for public exchange switches alone, from 39 to 80 percent
(Connaughton, 1982, p. 14.)
When Frangois Mitterrand was elected president and a Socialist government
came to power in 1981, telecommunications seemed to be a perfect industry
for its goals. It was a sector that already had a strong tradition of state control,
long-range planning, and close cooperation between government and industry.
Thus, the new government, like its predecessor, seized upon telecommunica-
tions as a vital sector. But there was an important difference of emphasis: the
more technocratic Giscard d'Estaing era was the golden period of telecommu-
nications engineers. They were accorded status, investment funds, and relative
operational freedom to expand the network. Under Mitterrand, telecommuni-
cations were not a goal in and of themselves, but one tool for a more general
modernization of society and the economy. As a result, telecommunications
policy became more political and less technocratic; engineering bureaucrats
yielded the center stage to empire builders straddling the public and private
sectors. The focus of governmental policy moved from the network to its sup-
plying industries.
The General Commissariat of Planning assembled a study group to investi-
gate the economic effects of DGT's investment on employment, internal trade,
and economic growth. Using a multisectoral econometrics model, the commis-
sion attempted to gauge the effect of an increase in telecommunications invest-
ments each year between 1983 and 1988 (Bienaim and Picaud, 1984). The
DGT passed the test of investment effectiveness. Politically, it reasserted its
autonomy and achieved predominance in its planning, development, and fi-
nancing functions in the flliere electronique, the technological-economic-
political complex of electronics.
Initially, the Directorate for Electronic Industries and Data Processing (DIELI),
a board of the Ministry of Industry Planning, supervised government planning
in the electronics industry. It was in charge of setting guidelines for the devel-
opment of the French electronics industry, but the relation of those plans to
telecommunications was not clear. In time, the DGT asserted increasing influ-
ence.
The Socialists had come to power on the campaign slogan of "nationaliza-
tion, decentralization, planification, and democratization." Of these goals, they
accomplished the first and pursued the third in telecommunications. The second
and fourth goals inspired the creation of local radio, television, and cable tele-
vision. On the other hand, the reorganization of the telecommunications sector
served to further concentrate the industry rather than to accomplish the declared
fourth goal of decentralization.
Despite the substantial governmental efforts to expand the role of the French
electronics industry, it claimed only 5 percent of the world market in 1981-
1982. Without the domestic market, French firms had only 1.4 percent of the
France 147
world electronics market, including France's former colonies. France had only
0.3 percent of the U.S. electronics market (Scientific American, 1983)
Furthermore, government plans had generated little success for French con-
sumer electronics, office equipment, and computers. In 1982, Bull amassed
losses of FFr 1.35 billion, while IBM France posted identical profits.
To improve this situation, the incoming Socialist government's five year plan
of 1982 provided as one of its four main objectives the development of French
expertise in electronics. At that time, telecommunications had a turnover of
about FFr 26 billion (i.e. sales revenue), and comprised more than one quarter
of France's filiere electronique, the chain of technical, social, and economic
activities associated with the electronics industry. The investment target for the
five year-period (governmental, private, and academic) was FFr 140 billion
(approximately $21 billion) for the five-year period. This was approximately
$4 billion annually, a sum comparable in magnitude to IBM's $2 million R&D
expenditure at the time (Darmon, 1985, p. 79).
For a long time, France maintained a policy of promoting "national cham-
pion" firms and projects of high prestige and visibility such as the Concorde,
the first light water nuclear power plant, the largest solar energy furnace, and
a high-speed train. Le Monde dubbed these projects the "new cathedrals" (Crane,
1979). A major theoretician behind the Mitterrand government plans and poli-
cies was Jacques Attali, counselor to the president, who described his French
economic model as "based on a mathematical model influenced by linguistics
and psychoanalysis."
The Socialist government implemented its policy in three major steps. The
first was the nationalization of the electronics sector. The second was a restruc-
turing that, in effect, assigned specialized functions to the various nationalized
companies: CII-Honeywell-Bull (CII-HB) was designated for computers; Al-
catel for telecommunications and office automation; Thomson and Matra for
semiconductors; and Thomson for consumer electronics. The third step was a
five-year, $20 billion government development plan for electronics, which was
only partly implemented.
Industrial planning designated the electronics area as the hoped-for "loco-
motive" for French industry. Economic models of the demand for goods and
services corresponding to the growth of telecommunications services supported
this industrial plan. Studies concluded that telecommunications represented one
of the best investment choices for the stimulation of production (Bonan et al.,
1985). Nationalization and government control were justified by the charge that
French private business had not been stoking the locomotive hot enough. Ac-
cording to Jean-Pierre Chevenement, then the minister of research and industry,
"If capitalism had worked the way it was supposed to, things might be differ-
ent todaybut capitalism did not work" (Locksley, 1983, p. 134). Similarly,
for Pierre Dreyfus, the French minister of industry and one-time head of the
Renault automobile company, "private enterprise in France does not take risks;
it is chilly, timid, shy, against taking risks even when they are necessary for
new products and long-term needs" (Delamaide, 1982, p. 167).
148 France
Still, France had fared relatively well in the global recession of 1981, and it
is not obvious why the governmentalready the dominant force in the elec-
tronics industry through the DOT, military procurement, and R&D financing
had to target this sector for nationalization.
Americans speak of nationalized industries, but the French call them na-
tional industries. France has a well-established tradition of national industries,
including firms such as the tobacco monopoly, Seita; the advertising agency
and communications firm, Havas; the automobile firm, Renault; and all the
major banks, in addition to the more conventional state activities in heavy in-
dustry, utilities, and transportation. The state's role in industrial policy dates
back at least to Colbert, under whom state enterprises such as the Gobelins
tapestry and Sevres porcelain works were established in 1681. Three centuries
later, in 1982, Prime Minister Pierre Mauroy celebrated this tradition by pro-
claiming "nationalization is a form of the French genius" (Delamaide, 1982,
p. 167).
The industry is enormously dependent on government procurement. In 1988,
France Telecom provided a market of $2.5 billion, or 45 percent of production,
for the French telecommunications industry, whose exports accounted for 15
percent ($800 million), with the French private sector and the military together
adding another 40 percent (Steckel and Fossier, 1990, p. 6). The industry is
also highly concentrated. Of the forty companies comprising the sector, the
largest five accounted for 90 percent of total production. Three of these five
were nationalized: CGE (including CIT-Alcatel), Thomson (Thomson-CSF and
LMT), and CGCT. Additionally, the computer manufacturer CIT-Honeywell-
Bull, and the defense and electronics firm Matraboth increasingly moving
into telecommunicationswere nationalized. Altogether, the French state be-
came an owner of three quarters of the telecommunications equipment busi-
ness. Several smaller groups remained in private hands, such as G3S, which
included CSEE and SAT (Societe Anonyme de Telecommunications Radio-
electrique et Telephonique) affiliated with the Dutch company Philips and ac-
tive in microwave, radio, and data communications and switching. The com-
pany Jeumont-Schneider produced electronic PBXs.
The nationalization of the electronics industry was part of a larger pattern.
By 1982, 30 percent of the country's entire industry was state owned, and the
government had holdings in more than a thousand firms. These firms controlled
95 percent of bank assets and about half of all French business activities (Mon-
sen, 1984). But by the end of 1983 the franc had been devalued three times,
and nationalized companies had lost an estimated $4.5 billion. Whereas ini-
tially they retained considerable flexibility under Industry Minister Dreyfus,
who as head of Renault had learned the benefits of decentralization, they were
subsequently burdened by much more interference under his successor, the more
left-leaning Jean-Pierre Chevenement. This led to a rebellion of company pres-
idents, who appealed directly to Mitterrand. Laurent Fabius, the pragmatic bud-
get minister, was reassigned to the ministry of industry. After becoming prime
minister of France in 1984 at the age of thirty-seven, Fabius renewed a flexible,
France 149
decentralized policy and reestablished the priority of high-technology fields over

aging industrial sectors such as steel.
The initial nationalization costs about $6 billion. But the losses that had to
be carried mounted from $226 million in 1980 to $4.6 billion in 1982, and
$4.2 billion in 1983. Thus, the annual subsidy of losses was more than two-
thirds higher than the initial costs to the government of taking over the com-
panies. Under the pressure of these losses, the government began to emphasize
financial accountability. Meanwhile, it reduced capital infusions, issued non-
voting shares on the Paris stock exchange, and even encouraged venture capital
markets. At the same time, the opposition, both Gaullists and Giscardists, for-
mulated sweeping economic platforms in which they advocated reduced gov-
ernmental ownership and regulation, greater tax incentives, lower taxes, and
some forms of deregulation. Former President Giscard d'Estaing defined the
goal as seeking a "liberal economysomething that has never existed in France
before" (including under his presidency).
The direct government aid package for the electronics industry expanded
from $1 billion in 1983 to $1.2 billion in 1984, a level maintained in 1985.
This did not include substantial support through DGT procurement and R&D.
In both 1984 and 1985, Bull received $100 million. Thomson received $100
million in 1984 and $130 million in 1985. It also shared R&D costs with the
state. At CGCT, the government's equity participation grew from $300 million
in 1984 to $450 million in 1985.
Eventually, ending state companies' deficits became a priority and French
companies rapidly cut jobs. Alcatel-Thomson eliminated 1700 jobs, LMT 1300,
and Thomson 4000. Personnel in the telecommunications industry steadily di-
minished. There were 94,000 people employed in 1977 and 77,000 in 1983, a
25 percent decline (Darmon, 1985, p. 88). This was not altogether surprising.
After the years of vast expansion of the telephone network, the industry expe-
rienced surplus capacity and inevitable decline once the backlog was elimi-
nated. In 1985, export orders also fell below the levels of the previous two
years. Some reprivatization was considered. Edith Cresson, the new minister
for trade and industry, and prime minister after 1991, began to argue that na-
tionalization targeted parent companies and not necessarily their subsidiaries,
which could have substantial private elements.
Losses continued in the 1990s. In 1991, the government had to promise more
than $1 billion of support to the state companies Groupe Bull and Thomson.
Bull's 1990 deficit was $1.2 billion, and the government promised $700 million
to alleviate the computer company's problem. Thomson, in similar dire straits,
received $350 million. France Telecom had to contribute toward those amounts.
All in all, the Socialist experience was sobering. It dampened the govern-
ment's high-technology ambitions and challenged traditional socialist ideology.
This was important, given France's considerable influence in the world of ideas,
particularly in the developing world. The question was whether the retrench-
ment of the French Socialist government was only a temporary measure, dic-
tated by the reality of then-existing conditions, or whether it could in fact be
150 France
viewed as a historic reorientation of the idea of state control. Conservative

thinkers began to receive a respectability that they had not enjoyedwith the
possible exception of Raymond Aronsince World War II. A number of best-
sellers by writers Jean-Franc,ois Revel, Guy Sorman, and Francois de Clossets
popularized the idea that French society had become too state-controlled, too
full -of privilege and constraints, and that as a result it could not grow and
innovate. They advocated certain forms of deregulation and denationalization.
On a more philosophical level, Andre Glucksmann, a leader of the "new phi-
losophers" and a former Maoist, analyzed the problems of traditional socialist
ideology in modern industrial society and what he referred to as the "empti-
ness" of the Mitterrand goal of modernization. These views began to set the
stage for the temporary return to power in 1986 of conservative government,
which rescinded some of the nationalizations, a path not contested by the So-
cialists when they regained a majority in 1988. But first, we will discuss the
industry itself.
The French Telecommunications Industry
Following nationalization, the Socialist government restructured the telecom-

munications industry with the goal of streamlining and rationalizing it, in the
hope of reducing its heavy losses and increasing exports. An important question
is whether the state's interventions deterred the industry or encouraged it. There
can be no simple answer here to this issue central to business-government
relations in general. But one can observe that the government's interventionist
attitude, complemented by the strong polarization of French politics, leads to
strong policy swings and creates uncertainty and risk for managers. Conse-
quently, they invest time and effort in building supportive coalitions. Where
governmental support is the ultimate variable for success or failure, securing it
will be a priority. Conversely, governments and parties in a polarized political
environment need the support of powerful economic entities. This symbiotic
process leads in time to politicized managers and managerial politicians. It is
an environment conducive to empire builders, but not necessarily to product
innovators. Its effect may be offset by the protection and funding that arise
from government actions, though such help can also be extended through a
hands-off policy rather than by microinvolvement in management. In the case
of French digital switching, much of the technology was state-developed; the
market was a state market; the risk was the government's, and export deals
were subsidized in various ways as part of foreign commercial policy. The role
of the managers, thus, became one of dealing with and manipulating such ac-
tions. The two main players in this game were CGE and Thomson.
Compagnie Generate d'Electricite (CGE) is the largest French industrial
company outside of the automobile and oil sector. Founded in 1898, it had
1989 sales of about FFr 143.9 billion, and FFr 4.9 billion in net income. The
CGE group includes over 100 subsidiaries in electronics, electric generation,
France 151
transmission, publishing, transportation, engineering, and construction. The

group's organizational structure is decentralized to the extent that some of its
subsidiaries have stocks that are independently traded on the Bourse. It has a
major presence in two key governmental infrastructure projects in France: nu-
clear power generation and telecommunications equipment. In 1982, CGE was
nationalized, and in 1987 it was reprivatized. To deal with the problem of
raising capital as a nationalized firm without a priority in the government bud-
get, CGE issued "participatory certificates" to private investors. The nation-
alized CGE was one of only three such French companies that were profitable
(the other two were Saint Gobain and Matra). In the first year of its nationali-
zation, the group spent $2.4 billion on R&D, half of which went into telecom-
munications and informatics (Vasseur, 1984).
Foreign firms had long dominated the French market. In 1939, as discussed
earlier, Thomson-Houston, controlled by ITT, had 31 percent; LMT, also an
ITT firm, had 41 percent and SFT Ericsson had 23 percent. French firms sup-
plied the remainder: Societe Industrielle des Telephones (SIT) (3 percent) and
L'Association des Ouvriers en Instruments de Precision (AOIP) (2 percent)
(Bertho and Nouvion, 1986, p. 60). In stark contrast, by 1985 a single state
firm, the merged company Alcatel-Thomson, had 84 percent of the market and
exerted a major influence on the remaining 14 percent of then state-controlled
CGCT, which survived largely for political reasons.
Key dates for the foreign and French victories, respectively, are 1925 and
1976. In 1925, in what became known as the battle of the Carnot exchange,
multiple firms competed for the Paris area automatic exchange equipment or-
der. Paris was one the last European capitals to automate its telephones. Sev-
eral systems had been introduced earlier elsewhere in France. In 1913,
Thomson-Houston installed an American Strowger unit in Nice under license;
in 1915, AT&T's LMT cut over a rotary system in Angers; in 1924, SFT-
Ericsson similarly installed its own Strowger system in Dieppe: Britain's GEC
and Germany's Siemens offered other Strowgers. French suppliers were
Thomson-Houston, Grammont, and SIT, and SIT had no technology licensing
agreement with a foreign developer. Within a short time, the fledgling ITT
acquired LMT from AT&T's Western Electric, and it acquired a minority in-
terest in Thomson-Houston. The PTT chose the ITT rotary technology and its
price offer, provided that equipment was produced in France by LMT, licensed
to SFT Ericsson and to Grammont, and backed up by an ITT lab located in
Paris. Defeated, Thomson-Houston sold its telecommunications divisions to
ITT in 1925, where they became the foundation for what was named in 1939
CGCT, ITT's other French subsidiary. ITT subsequently also won the contract
for medium-sized and rural exchanges. Its successful R6 was actually an earlier
Thomson-Houston development. With the advent of the Great Depression,
Grammont also exited from the switching business; CGE briefly participated
with SFT and SIT, but by 1932 French firms played a minimal role in the field
(Aurelle, 1986). ITT dominated directly, and whatever Ericsson produced was
under ITT license. More important to the government than switching was long-
152 France
distance transmission, which had strategic importance in light of the potential

for war with Germany. Hence, the government encouraged the strengthening
of the firm SAT, which in time became part of today's CGE.
After World War II, cross-bar switches were introduced on an experimental
basis. In 1955, the ITT Pentaconta was chosen, to be manufactured by LMT
and CGCT. Ericsson's CP 400 was selected for medium-sized exchanges and
licensed to AOIP and CIT (which had evolved from SIT in 1946). This license
was the seed for the global CGE-Alcatel empire, since CGE soon acquired
CIT, after being unsuccessful with its own electromechanical switch. CGE also
took a substantial minority interest in SFT-Ericsson. The French firms were
aided by the PTT's encouragement of re-creation of the Socotel, a cartel for
switching firms dating back to the end of the Vichy era, when several French
firms allied themselves for mutual protection against postwar competition from
foreign firms. The market was stable until the mid-1970s, when the electronic
switching generation was decided upon. CGE, through its telecommunications
arm, CIT-Alcatel, offered a D-10 system. Joining with Ericsson in the firm
Citerel, it also promoted the Ericsson AXE. ITT's system was the E l l . Other
competing firms were Thomson CSF (offering Northern Telecom's Sp-1); SAT
(with Siemens EWS); and TRT (a Philips subsidiary). Thomson's reentry into
switching was remarkable insofar as it had just concluded a market division
agreement with CGE in 1969, under which it concentrated on components,
low-tension technology, consumer electronics and professional materials. CGE,
according to the agreement, would specialize in high-tension, telecommunica-
tions, and electric power equipment. But the government persuaded Thomson
to participate in telecommunications. The government, intent on a greater and
more diverse French presence, also forced Ericsson and ITT in 1976 to transfer
to Thomson their French firms, SFT and LMT. Meanwhile, CGE was still
looking for a foothold, but it had the critical advantage of a close alliance with
the PTT's technology laboratory, CNET. This relationship was forged in the
late 1950s by CGE's Ambroise Roux. Later, CNET developed the concept of
time division multiple switching (TDM), and CGE created the applications in
close collaboration with CNET engineers. Thomson, on the other hand, devel-
oped its MT20 without such governmental ties (Bertho, 1987) and chose the
space division multiplex technology. Of course, the DOT chose TDM as its
preferred technology, as advised by CNET. Just then the government estab-
lished expansion of the network as a top investment priority, thereby creating
a huge order boom.
By 1976, Thomson's long-running commitment to manufacturing switches
had become equivocal; the DOT strongly encouraged it to set aside its agree-
ments with CGE and reenter the telephone industry. Research on digital switch-
ing was undertaken, culminating in 1982, when Thomson's MT system became
operational.
ITT, in the early 1980s, was the only meaningful foreign presence left; it
still supplied PBXs and central office equipment through CGCT. But not for
long. In 1982, the Socialist government nationalized CGCT. In that period
CGE and Thomson-CSF also took over AOIP, a workers' cooperative for cen-
France 153
tral office equipment and PBXs, which had been unable to make the transition
to electronic switches and maintain its 10 percent share of the market. The
Socialists did not want AOIP to die. It also preferred the DOT to have several
equipment manufacturers to choose from. The government had already inter-
vened in 1979, requiring Thomson and CIT-Alcatel to hire part of AOIP's
personnel, to use its factories, and to manufacture a switch developed by the
AOIP, the URA 2G. CIT-Alcatel already had an equivalent product, and it
could dodge the order, but Thomson had to comply. Some of the difficulties
Thomson experienced in the early 1980s can be traced to its forced adaptation
to the URA 2G (Darmon, 1985, p. 171).
CGE's telecommunications, on the other hand, prospered. The emergence of
digital exchanges coincided with the great expansion of the French network.
Together, they made CGE the European leader in TDM switching. In 1977,
the DOT opted for the CIT-Alcatel E-10 as the centerpiece of its large-scale
expansion. To reduce CIT-Alcatel's bargaining strength, the DOT also desig-
nated the Thomson MT system as its second source. CGE's Alcatel produces
TDM switching systems, submarine cable systems, PBXs, and PCM equip-
ment. Its subsidiary, Alcatel Electronique, is involved in peripheral equipment.
Another subsidiary, SESA, is a leader in packet switching networks (Quatre-
point, 1984). CGE was also active in transmission lines and cable television
through its subsidiary Cables de Lyon. It also became dominant domestically
in nuclear reactor construction.
The second major player in the telecommunications equipment field, already
repeatedly mentioned, was Thomson-CSF a part of the Thomson Group ag-
glomeration of more than 100 companies and subsidiaries. Under a variety of
labels, it held over 30 percent of the European Community's color television
market. In 1987, it also acquired RCA's consumer electronics division in ex-
change for some of its own subsidiaries (particularly those involved with med-
ical equipment), and became the world's second largest TV manufacturer.
Thomson owned the major German television set manufacturersAEG, Tele-
funken, Nordmende, and Sabaaccounting for about one-quarter of the Ger-
man market.
The Socialist government nationalized Thomson in 1981. To facilitate its
nationalization plan, it pledged $900 million in financial support to Thomson
for telecommunications development. Alain Gomez, a graduate of ENA and
the Harvard Business School and a former leftist, was appointed to replace the
previous group of managers, whose career backgrounds had been predomi-
nantly in the military and who were known for a slow moving management
style. Of the top eleven managers under Thomson's former regime, only two
had come from civilian industry. Five top executives, including the president,
hailed from the navy. The company focused on military business and was less
profitable in consumer markets. When it became a state-owned enterprise, it
was directed to focus on consumer markets. The government hoped to avoid
repetition of Thomson's failure to exploit the potential of video disc players.
In 1972, the company had been at the forefront in developing the video disc
player, but rather than applying the technology commercially, it sold it as a
154 France
tool for teaching the maintenance of military equipment, without promoting its
applications in consumer electronics. Eventually, Thomson left video disc pro-
duction and licensed its technology to a Japanese firm, which produced the
machines at half the French cost (Pierrand, 1984). (Of course, it was in good
company, since Philips and RCA were similarly unsuccessful.) Thomson is
also the largest French producer of electronic components. In 1987, Thomson
merged its semiconductor activities with those of the Italian firm SGS, then a
member of the STET group.
As part of its nationalization activities, the French government bought out in
1982 ITT's remaining French telecommunications companies, mainly CGCT.
ITT valued its French companies at $375 million but got much less in compen-
sation. The government argued that CGCT was unprofitable, but this was a
self-fulfilling prophecy. ITT lost money in France because its equipment, which
sold well around the world, was rejected by the DGT in favor of French switches.
As a result, the government could point to the company's relative lack of com-
mercial success and argue for a low compensation for its expropriation.
The government assured CGCT's allocated share of about 16 percent of the
DGT procurement. This was augmented by government contributions of FFr
188 million to CGCT's capital. Still the company asked for FFr 1 billion in
various forms.
But things kept changing. In 1983, Gomez reached a transfer agreement with
George Pebereau, director-general of the CGE, handing over all of Thomson's
civil telecommunications activities in exchange for CGE's departure from mil-
itary procurement until at least 1990.3 Thomson was heavily in the red, having
not only devoted itself to developing the MT system, but spread itself using
over different switching systems (Darmon, 1985, p. 95). The reorganization
was by no means forced on the companies by the government. Gomez and
Pebereau persuaded Laurent Fabius that France's telecommunications needed
consolidation and "rationalization" in order to be internationally competitive.
The restructuring was unprecedented in its size or scope in Europe. Its main
aspect was that CGE and Thomson swapped major assets. After the exchange,
a new telecommunications group was created that combined the telecommuni-
cations businesses of Thomson and CGE, controlled by the CGE subsidiary
CIT-Alcatel. Thomson, for its part, received CGE's electronics component
business, military division, and consumer electronics operations. There were
several reasons for the reorganization. It attempted to simplify the existing
structure (i.e., to reduce competition 4 ) and generate economies of scale. But
more important, it attempted to stop the substantial financial losses at Thom-
son, which had been FFr 2.2 billion in 1982, with no end in sight.
The DGT, however, opposed consolidation because it did not wish to be
dependent on a single equipment supplier. In its view, rivalry between com-
petitive suppliers would benefit technological innovation (a perspective it did
not usually extend to its own operations) and make the DGT less dependent on
one company. For the previous right-of-center government, it had been axio-
matic to provide for multiple French suppliers; it had encouraged Thomson-
CSF to enter by having it absorb ITT's LMT and Ericsson's interests. To pre-
France 155
vent the CGE-Thomson reorganization, the DOT argued that a single domestic
supplier could only provide about 65 percent of these needs, whereas two na-
tional suppliers could provide about 90 percent. Thus, it concluded that the
consolidation had negative implications for the French balance of trade and for
French jobs by opening up the domestic market to foreign competitors.
In the end, Fabius accepted the Gomez and Pebereau proposal and consented
to financial support: In 1984, FFr 1.7 billion in capital endorsements and FFr
250 million in public credits were extended toward the reorganization, with FFr
500 million earmarked for 1985. Ironically, Thomson got the same amount of
government money for leaving the field as it had unsuccessfully requested in
order to stay in the telephone business (Darmon, 1985, p. 96).
The agreement led to the creation of Alcatel-Thomson, a manufacturer with
sales of FFr 12 million and 40,000 employees. The combination represented
70 percent of French domestic telecommunications production and ranked at
that time fifth in world telecommunications production. Alcatel-Thomson con-
trolled 85 percent of public switching, 75 percent of electromagnetic transmis-
sion, 6 percent of cable, and 60 percent of private telephone sets.
The Ascendancy of CGE and the Expulsion of American Firms
In the past, the American and French markets for telecommunications equip-
ment had barely overlapped. In the mid-1970s the U.S. Department of Com-
merce found that imports were only 3-4 percent of total telecommunications
equipment sales in France, and of those, the U.S. share was only 1 percent of
the total, made up primarily of teleprinters, modems, microwave components,
and test equipment. (This, however, did not include ITT's substantial activities
in France; its ownership was American but its production and R&D European.)
The DGT created or continued cartels with leading manufacturers (U.S. De-
partment of Commerce, 1975, p. 3). Sotolec was formed in 1947 for compa-
nies in transmission equipment. Two years later, Socotel was similarly estab-
lished for the switching equipment industry, continuing an arrangement of the
Vichy regime. Members of the Sotolec cartel included the firms LMT, SAT,
TRT, CGE, CIT-Alcatel, and Thomson-CSF, and their main purpose was to
"prohibit ruinous and savage competition." Though sometimes characterized
as paper organizations, these cartels controlled equipment specifications, pat-
ents, and development. CNET maintained technological control over suppliers
and at the same time provided suppliers with access to design specifications
and to each others' patents on a royalty-free basis.
French firms similarly found much of the U.S. market difficult to reach. But
with the AT&T divestiture, the close link between the Bell Operating Compa-
nies and AT&T, which had assured AT&T dominant access to the world's
largest market for telecommunications equipment, was severed, and French
producers began eyeing this opening with great hopes. In consequence, CGE's
Alcatel subsidiary worked hard at breaking into U.S. markets.
French telecommunications equipment exports to the United States grew from
156 France
4.4 percent of the French export market in 1983 to 19.3 percent in 1988. Total
French exports in 1988 were Ffr 10.156 billion, more than double the 1983
figure. Exports to Western Europe, on the other hand, declined relatively from
17.3 percent to 13.8 percent. However, Alcatel's success was limited mostly
to the small independent telephone companies. Eventually, the firm reduced its
U.S. switching activities for lack of a major market success.5
During the same period, AT&T found itself unable to crack the French mar-
ket. This led to an important agreement in 1985 between AT&T-Philips (APT)
and Alcatel involving a partial mutual opening of each others' markets. Behind
the agreement was the question of what to do with the remaining French public
switch manufacturer, CGCT, which the French government had nationalized
from ITT in 1982. CGCT was losing large sums of money, but it had one
major asset, the traditional allocation of 16 percent of the French public switch-
ing market, or approximately 300,000 lines per year, as a second source for
the DOT. CGCT was initially given the task of manufacturing the Thomson
MT digital switch under a license, but it was in no position to develop and
upgrade it for the future. The question then of who would be the second source
supplier to the DGT therefore remained unresolved. One would think that the
DGT itself would be the one to decide with what other firm it would like to
deal. However, CGE, which already held 84 percent of the market, was in fact
negotiating with a foreign consortium about whether it should be admitted to
the market as CGE's competitor, in return for export benefits to CGE. CGE
was, in effect, selling the share of the French market that, for political reasons,
it did not supply. Specifically, CGE agreed that APT would receive CGCT's
16 percent market share for its 5-ESS PRX switch, to be manufactured in
France by CGCT and adapted to French standards. CGE would receive several
types of benefits: AT&T would supply marketing and technical assistance in
adapting its El0-5 switch for North American use; it would include it in its
product line and pay indemnities if sales did not reach a specified amount.
Second, Philips would transfer the microwave equipment manufacturing of its
French TRT subsidiary into a joint venture controlled by CGE, while AT&T
would undertake to buy at least $200 million of such microwave transmission
equipment over four years. In other words, Philips would relinquish control
over one of the few activities competitive with CGE in France, and AT&T
would help sell the products in the United States. Lastly, CGE would receive
$100 million.
CGCT, unenthusiastic about AT&T, preferred to deal with Siemens or
Ericsson, but for CGE, AT&T was a more compliant partner. Given the close
European collaboration between France and Germany, including the heavy in-
volvement of French companies in German consumer electronics, a company
such as Siemens, once it had a toehold in the French market, could not be
easily subjected to political protectionism. In contrast, any AT&T involvement
in France could receive much greater government scrutiny and future pressure,
since public opposition against it could always be easily organized.
The agreement required government approval and was criticized on a variety
of grounds. First, French trade unions sought guarantees, fearing that AT&T
France 157
involvement in CGCT might lead to a reduced labor force. More generally,

they and others were concerned about the entrance of an American high tech-
nology company into France, with possibly irreversible results for French in-
dustry.
Siemens began to pursue the possibility of substituting for AT&T in the 16
percent CGCT quota. The DOT supported the APT deal, while Siemens argued
along political lines for a "European solution," even though Philips, AT&T's
partner, was Dutch. Other French ministries did not want to antagonize Ger-
many. The matter soon reached the highest political levels, with discussions
between West German Chancellor Helmut Kohl and French Prime Minister
Jacques Chirac, warnings from the American ambassador, and hints of retalia-
tion from FCC Chairman Mark Fowler. Because of DGT's preferences, APT
seemed to have the inside track. But then the ITT-CGE merger burst on the
scene and changed everything.
In July 1986, CGE entered into an extraordinary agreement culminating in
its taking control of the telecommunications operations of the American-based
conglomerate ITT. ITT's international telecommunications equipment activities
had been originally acquired in 1925, when AT&T left the international mar-
ket. Despite its American headquarters, the company had only a limited equip-
ment presence in the United States, and in the telecommunications business it
was a multinational company without a real home base.
But the company, which had become a far-flung conglomerate in the 1960s
and 1970s under Harold Geneen's leadership, had subsequently corne upon
hard times. Led by Rand Araskog, it needed money, and it was unable to
sustain the large funds necessary to extend its System 12 digital switch to the
U.S. market or to finance development of the next generation of technology.
Meanwhile, CGE's Alcatel had difficulties of its own, especially in export
sales; it had almost no public switch sales in the rest of Europe, and most of
its international sales were political arrangements with former French colonies
or part of foreign aid packages. ITT's international presence combined with
CGE's financial resources made them a formidable combination. A deal was
arranged.
The complex agreement provided that ITT merge its telecommunications
equipment, office automation, and consumer electronics divisions into a hold-
ing company. CGE brought in it its own Alcatel equipment subsidiary. ITT
held a 37 percent share of the new firm, whereas CGE, together with other
European entities, controlled the rest. CGE had 53.6 percent, Societe Generate
de Belgique, 5.7 percent, and Credit Lyonnais, 1.7 percent. ITT received $1.3
billion, and the holding company assumed $800 million of ITT's debt. In 1990,
ITT sold an additional 7 percent to CGE for $640 million.
The new company was unambiguously called Alcatel, the name of CGE's
telecommunications subsidiary. Management and organization was set up in
decentralized fashion, with some support staff functions in Brussels, and nom-
inal headquarters in Amsterdam.
One of the new firm's major problems was that it had no less than three
different digital switches, each incompatible with the others and each with its
158 France
own problems. ITT's System 12 has advanced fully distributed processing, and
when widely deployed, the intercommunication between its different modules
can cause overloads. The MT20 switch was transferred to Alcatel from Thom-
son in 1983 and found few fans, even in France. Alcatel's own E10 switch
itself was losing international market share and needed modernization. Alca-
tel's main challenge, however, was not technical, but political and managerial:
It had to hold together a wide array of subsidiaries functioning under divergent
policy environments, management styles, and product lines.
The newly merged firm became the second largest international telecommu-
nications firm after AT&T, amassing $7 billion in assets, almost $10 billion in
sales, and 150,000 employees. It supplied a huge 42.5 percent of all European
public telephone switches. CGE heralded the agreement as establishing, for the
first time, a large-scale European telecommunications firm, but few Europeans
got enthusiastic over such a French state-dominated arrangement. CGE claimed
that it required the merger to give it sufficient size to operate the economies of
scale necessary for success. In other words, it asserted that it took almost half
the European market to be economically viable.
This deal finally established the French victory over ITT after more than half
a century of struggle. Georges Pebereau had been its strategist. But soon after
his triumph, the new conservative government replaced him with Pierre Suard.
The deal put into question CGE's separate arrangement, also negotiated by
Pebereau, with AT&T. Now, with CGE inheriting much of ITT's footholds in
many other European countries, it was likely that these countries would expect
greater reciprocity. In return, the 16 percent market share that had been allo-
cated to AT&T might be more usefully employed to assuage one or several
European countries, notably Germany. In other words, it was time for CGE to
reassess its contract with AT&T. The tug-of-war grew acrimonious. Within the
French government, the DGT preferred APT, while other ministries did not
wish to antagonize Germany, which was backing Siemens. The rival companies
successively sweetened their bids. In the end,the French government chose the
Swedish Ericsson as a neutral compromise, together with the French defense
firm Matra, which thus gained a foothold in its country's telecommunications.
Subsequently, Alcatel Chairman Pierre Suard strongly opposed AT&T's in-
volvement in Italy. Suard stated that the goal of European telecommunications
policy must be to maintain and improve the leadership of European industry.
In 1989, CGE had 210,300 employees, an increase over the 192,000 in 1982,
but this concealed significant staff cuts prior to and following its integration of
ITT's international telecommunications operations in 1987. For example, Al-
catel cut its new acquisition's work force by 25,000 to 125,000 and closed
seven of ITT's eighteen national research labs (Tully, 1989). As CGE became
the second largest manufacturer of central switching equipment worldwide, ac-
tive in over 100 countries, it lent the domestic French industry a major inter-
national presence.
Alcatel's opposition to the AT&T/Italtel partnership contributed to its acqui-
sition of Italy's Telettra in 1990. Alcatel thereby bought Telettra's 33 percent
market share in tranmission equipment in Italy and a 45 percent share of the
France 159
same market in Spain. Combined revenues of Alcatel and Telettra were $10.9
billion in 1989 compared with total of $11.1 billion in 1990. But in an effort
to reduce anti-competitive vertical integration, the European Commission re-
quired Spain's Telefonica telephone monopoly to divest itself first of its 21
percent share in the Spanish Alcatel subsidiary and of its 10 percent share in
Telettra, as well as open its purchasing policy. Alcatel merged its satellite op-
erations with Aerospatiale and bought the transmission equipment division of
U.S. giant Rockwell.
The French Computer Industry
The early mainstay of the domestic French computer industry was Groupe Bull,
which began in 1931, when a French group, partly owned by the Callies fam-
ily, cousins of the Michelin tire owners, bought the patent for punched card
machines from the Norwegian inventor Fredrik Bull. Its first significant product
was a tabulator-printer that successfully competed against the French subsidi-
ary of IBM, which had been established in 1920 (Mclnnes, 1964).
In 1952, Bull introduced its first computer, reputedly more innovative than
IBM's most advanced 604 computer, though IBM was just then starting in the
computer business. Bull's revenues rose tenfold in the decade from 1952 to
1962; exports grew 30 percent annually, and its labor force increased sixfold
to 14,000. Bull became the darling of the Paris Bourse, with its shares reaching
a price of 1470 francs by 1960, although its earnings never exceeded 3.80
francs per share.
Trouble soon followed, as IBM's more advanced 650 computer started to
recapture market share from the Bull models. Bull fell further behind when
IBM introduced its 1401 transistorized computer. While trying to recoup its
losses with the high-priced and innovative Gamma 60 computer, Bull's produc-
tion and development lagged behind schedule. The machines had chronic prob-
lems, and Bull sold or leased only fourteen of them. IBM came out with a still
more advanced series, the 7070 model (Bransten and Brown, 1964, p. 154).
Stymied in its products development, Bull sought help in America, entering a
ten-year agreement for the sale of RCA's computers in France under the Bull
label. It also obtained access to RCA's patents and production knowhow and
soon marketed RCA's Model 301 in France, promoting it as a French product
competing with those of the foreign IBM (some of whose products were man-
ufactured in France for export to other European countries). Unfortunately for
Bull, RCA soon left the computer business altogether.
Despite the success of its shares, Bull was undercapitalized and met with
economic difficulties. Instead of reducing their control to raise equity, the own-
ers took on large debts. When they faced the necessity of raising equity, the
stock had begun to drop and it was too late. Between 1959 and 1962, total
indebtedness increased fivefold, exceeding capital and reserves. In desperation,
Bull restructured, to gain access to the Swiss financial markets, and sold off its
share in Olivetti-Bull, a computer venture with Olivetti in Italy.
160 France
By 1963, the French government urged the state banks to assume control of
Bull. Bull's chairman, Callies, sought links with the American company Gen-
eral Electric, at that time a computer maker with European ambitions, as pref-
erable to control by the French government. Although GE agreed to buy 20
percent of Bull, the two companies had not reckoned with President Charles de
Gaulle, who personally ordered the deal stopped, believing it contrary to the
French interest in its computer industry to have an American firm share own-
ership. But after the government failed to line up French support, it was, em-
barrassingly, forced to permit an even larger participation of GE in Bull (49
percent) than the one it had previously vetoed, and at a lower sale price,
amounting to a virtual control of Bull by the American company. The French
government insisted, however, on segregating GE's control from Bull's defense
contracts, which at that time involved developing France's nuclear capability.
The deal was rationalized as part of a strategy of teaming up with another
American giant against IBM.
Not surprisingly, General Electric soon experienced major headaches with
Bull. Though it was clear that Bull was overstaffed, GE had committed itself
to not reducing employment. When it was confronted with reality, a period of
labor strife ensued. GE also canceled two ongoing computer developments at
Bull. Eventually, GE left the computer business entirely, with Honeywell as-
suming its share in Bull, to form the Honeywell-Bull company.
French concerns over computer "sovereignty" were not ill founded. The
French were shocked by the U.S. State Department's refusal of an export li-
cense for large scientific computers to the French atomic energy commission
for use in H-bomb research. Realizing its vulnerability, the government for-
mulated its Plan Calcul in 1966 and 1967, among other projects, merging two
French-owned computer manufacturers, CAE (owned by CSF and CGE, both
important telecommunications manufacturers) and SEA (a subsidiary of the
Schneider industrial group) to create the firm CII (Compagnie Internationale
pour 1'Information). Another element of the Plan Calcul was financial support
to establish consortia for the development of electronic technology. Bull, the
main computer firm, was excluded from the Plan Calcul because of its Ameri-
can links. A computer czar, with the title of "Delegate Generate," was to
oversee the projects and the government procurement, although the position
largely became a public relations affair.
Plan Calcul delineated goals for. the development of scientific computing,
leaving much of commercial office computer market to IBM. This misjudged
the explosive growth of business applications of computers, while failing to
recognize the innovativeness of the American Control Data Corporation (CDC)
in the scientific computing field.
The product strategy behind the Plan Calcul also missed other developments.
The emergence of time-shared use of computers, linking terminals to a power-
ful central computer, led the Plan to predict a future with only a few giant
mainframes and consequently to set product development priorities toward de-
veloping such machines. The subsequent trend was almost the opposite; mini-
and microcomputers proliferated, and IBM was not the dominant firm in that
France 161
market. On the other hand, one benefit from the Plan Calcul was to provide
digital technology skills to Thomson and to CGE, which were later applied to
digital telecommunications switches. This link was not foreseen in the Plan
Calcul, but it was its major success.
A second Plan Calcul was instituted in 1971, again amidst much publicity:
in absolute terms, however, the support levels were small. For the first Plan
between 1966 and 1971, government support had been $140 million. For the
second five years (1971-1976) support was $263 million.
CII had been nursed along with infusions of government capital. But after a
while, a strategic rift developed between CH's two major owners, Thomson-
Brandt and CGE, then rivals in the telecommunications field. The French gov-
ernment, Thomson, and CII were in favor of CH's participation together with
Philips and Siemens in the European computer venture Unidata, designed as a
competitor to IBM. However, CGE wanted a link with the second semi-French
computer company, Honeywell-Bull, and refused to put up money for Uni-
data.
By 1976, neither CII nor Honeywell-Bull were doing well; with some gov-
ernment pressure and financial help, they were merged into CII-Honeywell-
Bull, which was 53 percent French and 47 percent owned by Honeywell. The
new firm had an extraordinarily complex structure to assure participation by
major French electronic firms, by the French government itself (9.5 percent),
and by Honeywell. Recognizing the importance of orders, the government
promised $800 million of public sector procurement within the next four years,
to provide necessary economies of scale. If that target was not met, major
contractual damages would have to be paid to the company by the government.
It turned out to be expensive for the French public sector to buy the French
computers to replace IBM systems. In particular, the cost of converting soft-
ware turned out to be much higher than the cost for computers themselves. For
the government, this more than doubled the amount it had planned to spend on
CII-Honeywell-Bull computers.
With the Mitterrand nationalizations, Honeywell was forced to reduce its
involvement in Bull substantially. But because its participation was still deemed
to be desirable because of its access to advanced technology and the American
market, it retained 19.9 percent of Bull and entered into a ten-year marketing
and technology agreement with it. The French government also invested 1.5
billion francs in Bull. In 1986, Honeywell exited from computer manufacturing
altogether.
In the 1980s, Groupe Bull's guiding philosophy shifted. It had not been able
to manufacture its DPS7 computer, which was on the drawing board since
1979. Bull began to rely on cooperative agreements with other firms, including
Honeywell, Convergent, Trilogy, and NEC. Such a strategy brought with it the
risk of transforming Bull into a marketer of others' products, while contradict-
ing notions of French autonomy in the computer field that underlie much of the
government's financial efforts.
Bull was the twelfth largest computer manufacturer in the world. In the Eu-
ropean market for minicomputers, it was second, with 11 percent of the total.
162 France
It was weaker in low-end minis and especially mainframes; the latter market
was dominated by IBM (57 percent) and Siemens (15 percent), leaving Bull
with a 6 percent share. In 1988, Bull received 18 percent of its revenues from
U.S. operations (Schenker, 1989a, p. 13). It then sought a greater share of the
U.S. market and acquired in 1989 the computer division of Zenith, an Ameri-
can manufacturer in financial difficulties because of its television manufactur-
ing.
Groupe Bull continued to receive some $150 million annually from the French
government (Petersen, 1989, p. 80). But its finances were in trouble. At the
turn of the decade, Bull suffered heavy losses. In 1990, Bull's incurred a $1.2
billion deficit, necessitating a $700 million government contribution together
with France Telecom. Bull also sought minority participation by NEC. The
Japanese company already provided it with mainframe technology.
Other Efforts in Industrial Policy
Sem/cont/ucfors
In 1982, French companies accounted for only 5 percent of their domestic
microprocessor and MOS/memory market (though this number rises to 45 per-
cent when foreign companies producing in France are included). To increase
the French share has been a major development priority for the government's
filiere electronique industrial policy in the 1980s, an "action program for the
electronics sector" aimed at creating at least 1000 jobs. To attain its aims, the
government encouraged the creation of two new French-American ventures,
Matra-Harris Semiconductors (MHS) and Eurotechnique, by Saint-Gobain and
National Semiconductors (Malerba, 1985). Other major French microcompo-
nent manufacturers were Thomson and Schlumberger; The latter acquired the
American company Fairchild, a pioneer in integrated circuits that had lost its
leadership position partly because of the flight of some top personnel into en-
trepreneurial start-ups. Schlumberger played an important role for a while, headed
by the legendary Jean Riboud, an intimate of President Mitterrand. It was at
one time called the best managed enterprise in the world, but Riboud died in
1985 and the company was beset by problems in its declining oil drilling busi-
ness. Heavy losses on its Fairchild semiconductor operation led Schlumberger
to sell it in 1986.
A second "Components Plan" (Plan Composants) was adopted in 1983,
seeking to promote specialized components in telecommunications, military ap-
plications, and MOS (metal oxide) technology. The plan's economic goals in-
cluded raising French volume of production to 4.65 billion francs in 1986. The
plan's budget was FFr 3.2 billion (FFr 700-800 million annually), mostly
benefitting four companies. The government's telecommunications research arm
CNET also played an important role in this area.
In 1987, Thomson merged its semiconductor operations with those of SGS,
France 163
the microchip firm controlled by the Italian government's STET group, to form
Europe's second largest microchip firm after Philips.
Communications Satellites
In cooperation with Germany, France developed an experimental satellite sys-

tem called Symphonic, launched in 1974 and 1975. The next generation of
satellites, Telecom-1, was decided upon in 1978 and contracted from Matra
with the limited involvement of Alcatel and Thomson CSF. Telecom-1A and
- I B are used for domestic civilian and military applications as well as for trans-
mission of telecommunications and television programs to France's overseas
possessions.
This domestic French satellite must be distinguished from the European ECS
satellite, produced by a European consortium with British Aerospace as prime
contractor, administered by the European Space Agency (ESA), and operated
by the European satellite organization Eutelsat.
Telecom- 1's coverage of most of Europe allows it to provide links beyond
the French borders. Two agreements were signed in 1981 with Germany, per-
mitting it to offer domestic national communications in German territories us-
ing part of the system's capacity. A second agreement was signed in 1982 with
Eutelsat for offering international European links. A 1980 agreement between
France and Germany led to the launch of the TDF-1 direct broadcast satellite
in 1988. TDF-2 was launched in 1990.
The Restructuring of France Telecom
France Telecom itself continued to widen the scope of its activities. It had
155,000 employees and revenues of $166.6 billion in 1989. It had 27 million
access lines, forty-five main lines per 100 persons. It established international
partnerships in Germany, Spain, Italy, and Singapore (the latter cable project
involved a $600 million fiber optic cable from France through the Middle East
to Singapore), and joint projects with Germany's DBF Telekom, a Phonepoint
bid with BT, as well as participation in Infonet and a Pacific Cable with other
partners (Steckel and Fossier, 1990). After the Chirac conservative government
came to power in 1986, Gerard Longuet, a leader of the right-of-center Repub-
licans, became PTT minister. Although the conservative government took an
early and active role in privatizing the broadcasting field, it moved more care-
fully in telephony. It promised to propose legislation that would establish pri-
vate competition in the telecommunications field by the end of 1987.
Marcel Roulet, a PTT veteran, was appointed as the Chirac government's
director of the DOT. He recommended ending the DGT's dual role of network
operations and regulation and establishing the DOT as a state enterprise rather
than as a government administration. Under this plan, the industrial policy role
of the DGT would move to the Ministry of Industry and some of the DGT's
164 France
competitive subsidiaries would be partly privatized. At the same time, the DOT
would be free, under such an arrangement, to participate in joint ventures with
private companies and to purchase its equipment more freely.
Longuet was succeeded by his party colleague, Frangois Leotard. He ex-
panded the High Authority of the Audiovisual Sector, an agency created by the
Socialists a few years earlier to supervise broadcasting, into a National Com-
mission for Communications and Liberties (CNCL), with some regulatory pow-
ers also over telecommunications. In October 1986, the ministry set up the
Mission a la Reglementation to handle postal and telecommunications regula-
tion. Jean-Pierre Chamoux headed this task force, developing new policy ini-
tiatives along these lines, but many of its proposals were slowed down when
the labor unions threatened disruption and presidential elections interfered.
The government did not recommend an opening to competing service provid-
ers in transmission. This left the liberalization of value-added networks services
(VANS) as a major focus. Since 1987, VANS had been legalized and usually
needed only to notify the PTT administration to commence operations. The
decree distinguished between closed, user group networks and universal net-
works open to third parties.
The Chirac government moved more actively in the reprivatization of the
electronics industry. CGE was sold, undivided, to the public in May of 1987
for $3.7 billion. Thus, a private near-monopoly in telecommunications equip-
ment succeeded the public one. Next was CGCT, which was sold to Ericsson
and Matra in January 1988. The privatizations of Thomson and Bull were de-
layed after the stock market crash of 1987 and were later abandoned after the
presidential elections confirmed Mitterrand in 1988.
Furthermore, it was more difficult to privatize painlessly the two money los-
ing companies. In 1990, Thomson had a deficit necessitating a $350 million
contribution by the government. Bull's deficit was $1.2 billion. Alcatel's profit
for the same year was $750 million.
Chirac's efforts at privatization had been opposed by Paul Quiles, former
oil company engineer and minister of defense, though not particularly well
versed in telecommunications issues. A savvy politician, Quiles became PTT
minister in the Socialist government that soon returned to power. At the same
time, the PTT became the Ministry of PTEPost, Telecommunications, and
Spacereflecting the growing importance of satellite to network communica-
tions.
In 1988, the regulatory and operational functions of the DGT were split into
the Direction de la Reglementation Generale (DRG), and France Telecom, the
new name for the DGT. The DRG (the former Mission a la Reglementation)
has authority over type approval, value-added services, and the electromagnetic
spectrum (Roussel, 1989b). Its first head was Bruno Lassere, formerly legal
counsel to the DGT, while Marcel Roulet was appointed the first director gen-
eral of France Telecom. The DRG also represents France in international ne-
gotiations. Authority over France Telecom's most significant decisions such as
network investments, public service obligations, and management remained un-
France 165
der the control of the PTE through the Direction des Services Publiques (DSP).
The DRG had only authority over competitive services, which was still limited
to a handful of VANs and a small second cellular service provider, and its
powers to bring competition to the French market were limited.
In 1989, Quiles commissioned an analysis of French policy from civil ser-
vant Hubert Prevot. Originally a defender of the DOT monopoly, Quiles stated
publicly that growth should not be constricted by "archaic structures." Quiles
also had the support of France Telecom's management, which had complained
about political control (Dawkins, 1990b). The Prevot report recommended the
formal separation of post and telecommunications services, and the removal of
France Telecom's equity from the Treasury to create a more autonomous and
competitive entity. The autonomy would also change the civil servant status of
employees. This resulted in union opposition. Led by the Communist CGT
labor federation, one fifth of PTT workers went on to strike in 1990. Despite
the national debate they triggered, the proposed reforms passed the National
Assembly with surprising ease. In the Senate, however, the proposal's detrac-
tors were many. The strongest criticism came from a group of conservative
senators led by Gerard Larcher, a strong supporter of former PTT minister
Longuet. They issued a scathing report that charged Quiles with being weak-
kneed in his proposed reforms. The report claimed that, despite gaining auton-
omy, France Telecom would still answer to the government in setting tariffs
and before making large investments or purchases. The senators found partic-
ular fault with the program's hiring provisions, labeling them too rigid to stop
the chronic "brain drain" from management ranks.
The Larcher committee also proposed that a larger contingent of lawmakers
be chosen for the proposed twenty-one-member France Telecom board and that
more autonomy be given to the board than was recommended by Quiles. The
Senate conservatives also wanted to create a European Telecommunications
Foundation to promote the continent's interests in telecommunications.
After promises of job security for employees, the government's reform pro-
posals passed the National Assembly and Senate in May 1990. The law pro-
vided that France Telecom be separate from La Poste and managed by a twenty-
one-member board, including representatives of French ministries and users
groups. France Telecom's president is nominated to a three-year term by the
government. It must negotiate its budget with the government and set rates
according to a price-cap scheme, instead of the past 6 percent rate-of-return
system. In addition, the COGECOM subsidiaries, including TDF, Transpac,
and France Cables et Radio, were all removed from the control of the Finance
Ministry.
Paul Quiles was succeeded in 1991 by Jean-Marie Rausch, and in 1992 by
Emile Zuccarelli, head of a small leftist party and a mayor in Corsica, but
without a telecommunications background. The position of PTE Minister was
downgraded to a junior minister reporting to the Minister of Finance, Economy
and Budget.
Limited proposals to liberalize terminal equipment and value-added service
166 France
markets were introduced in 1990, and France Telecom's monopoly over basic
telephone, telex and national network services was reaffirmed. V-sat licenses
were issued in 1991 to France Telecom, but also to BAeCom, Alpha Lyracom,
Reuters, and PolyCom. A second mobile license was awarded to SFR.
Thus, the first steps in what will be a continuing journey toward an industry
structure of greater openness had been initiated, though they still represent pri-
marily a strengthening of the traditional operator, liberalizing its ability to func-
tion successfully rather than altering the structure of the communications sec-
tor.
Telecommunications in Benelux
and the Alpine Countries
10
The Netherlands
Historically, the Netherlands has been a center for transportation, insurance,

banking, and trade. The port of Rotterdam is the largest in the world and a
major hub for the oil business. Much of the extensive European Rhine and
Meuse barge traffic reaches Holland through inland shipping. In addition, the
country accounts for a quarter of international trucking within the EEC (PTT
Telecommunicatie, 1983). Holland is also the headquarters for three of the
largest multinational companies in the worldShell, Philips, and Unilever. In
1980, the sum of the country's imports and exports was 60 percent higher than
its entire GNP, a ratio five to six times higher than that of the United States or
Japan, and twice as high as those of other European countries (Wieland, 1986).
It is therefore not surprising that the Netherlands targeted its telecommuni-
cations system to help establish itself as a major European economic center. In
1985, the government decided to reorganize the PTT by forming a semi-inde-
pendent telecommunications company separate from the basic network pro-
vider. This moderate reform, which went into effect in 1989, marked the sec-
ond European PTT transformation, following the more far-reaching changes in
the United Kingdom.
The Dutch reorganization was particularly significant because it served as a
model of corporatization for other European countries. The extensive competi-
tiveness in the United States made its system institutionally too distant to be
directly applicable. The British policy, favoring privatization and the establish-
ment of a competitive carrier, was also too radical for the continental countries.
The Japanese model suffers from the lack of transparency of Japanese govern-
mental and economic processes. Holland, on the other hand, is a close and
respected neighbor, and its policies cannot be easily dismissed as attempts at
hegemony. The Dutch PTT reform was therefore watched with particular inter-
est by other European countries.
The Dutch PTT

A national postal service existed in the Netherlands since 1799, when it was
created to supplant the private postal services (Kingdom of the Netherlands,
1981). Telephone service was begun in 1881, when the private Nederlandse
Bell Telephoon Maatschappij opened in Amsterdam. Other companies followed
169
170 The Netherlands
elsewhere, operating under governmental concessions (Dabbs et al., 1982). Pri-

vate companies also established long-distance interexchange telephones starting
in 1888. However, the government took over the long-distance licenses when
they expired in 1897 in order to protect the revenues of state telegraphy; sub-
sequently, the Government Telephone service was formed.
In the meantime, as a consequence of monopolistic charges levied by private
local companies on subscribers, most municipalities co-opted local telephone
service. In the largest citiesAmsterdam (1896), Rotterdam (1896), and The
Hague (1903)the municipal governments built completely new systems rather
than purchase the deteriorated Bell plant (Holcombe, 1911). Starting in 1907,
the PTT collected the local networks under its aegis, and by 1927, it operated
all but the networks in Amsterdam, Rotterdam, and The Hague. Eventually,
the PTT gained full national control over telephone service when the last three
municipal systems were forcibly incorporated in 1940 during the German oc-
cupation in World War II.
Strictly speaking, the PTT, as defined by the 1904 Telegraphs and Tele-
phones Act, was not a statutory monopoly; in practice, however, it played that
role ever since it absorbed the remaining municipal systems. Formerly called
the Staatsbedrijf der Posterijen, Telegrafie en Telefonie, the Dutch PTT pro-
vided telecommunications as well as postal and financial services. It was the
country's largest employer, with 29,000 of its 100,000 employees working in
telecommunications.
The PTT was controlled politically by the minister of transport and public
works and financially by the minister of finance. Its governance was subject to
the PTT Raad (Council), a board that included all major political parties, and
to provincial PTT Kamers (PTT chambers), representing the provincial govern-
ments, the PTT unions, and users. Since the PTT depended on the government
for approval of its budget, it had to conform to governmental economic policy,
such as restrictions on price increases. As a result, a government ordinance
was required to change charges for rental equipment, for example.
The PTT was affiliated with Nepostel, a technically independent foundation
that provides Third World consulting and assistance, and with CASEMA, a
central cable television antenna system company, of which it held 51 percent.
CASEMA's other major shareholders were the national broadcast transmitting
organization NOZEMA (22 percent), which provided radio and TV broadcast
transmission services in the Netherlands, and several of the country's semi-
independent broadcasting associations, such as AVRO, KRO, VARA, and
NCRV. The PTT also constructed and maintained NOZEMA's facilities and
shared facilities with the Dr. Neher Laboratories (DNL) and PTT Telematics
Laboratories (PTL) for technical research and development.
Services
Telephone line density increased more than sixfold in twenty years, from 1
million in 1960 to 6.4 million in 1980. This growth rate slowed in the 1980s
The Netherlands 171
from 13 percent to 4 percent and lower, since most households had been reached.
Long waiting lists of potential subscribers had accumulated during the 1970s,
peaking in 1977 at 200,000. By 1982, this backlog had been reduced to 1
percent of main stations.
By 1988, telephone density in the Netherlands reached sixty-six sets per 100
population. According to the OECD, the 1990 tariffs in the Netherlands are
Europe's lowest for business services, $430.21 yearly, and the third lowest for
residential customers ($224.48 yearly), behind Iceland and Sweden. The den-
sity of the Dutch network allows the Netherlands to offer low rates (Montgo-
mery, et al., 1990).'
Despite the country's small size, PTT Telecom was the world's eighth larg-
est international carrier in 1988.
In 1962, the Netherlands became the second country in the world, after Swit-
zerland, to achieve a fully automated national telephone service. Electronic
switching started in 1972, with the Ericsson AKE-13, a stored-program-con-
trolled (SPC) exchange. The Philips PRX-205 local switch followed in 1974.
In 1980, Ericsson's AXE exchange was also put into service in order to provide
a second supplier. In that year, 30 percent of subscribers were served by SPC
switches, the highest such density in the world at the time according to ITT.
Since 1984, fully electronic exchanges have gradually replaced electromechan-
ical exchanges as part of a fifteen-year investment program to digitalize the
network, using AT&T technology after 1983. PTT Telecom inaugurated a pilot
test of international ISDN service in October 1989, linking Rotterdam with
Diisseldorf.
The public packet switching service Datanet-1 operated from 1982 on a trial
basis, and since 1984 commercially, with transmission rates up to 48 kbps. It
provides international access. Telex penetration in the Netherlands is among
the highest in Europe, but falling. Since 1984, an electronic mail service was
offered over Datanet-1. Multisat provides international broadband data com-
munications via satellites.2
Most post offices in the country carry Faxpost, a public facsimile service.
Other services include cellular telephone and radio paging through the Benelux
system. Videotex, under the name of Viditel, was introduced in 1980. The
broadcast text service teletext was started in 1980 by the broadcast associations
and NOS, their umbrella state organization. The PTT and several Dutch cable
operators formed a joint company in 1989 to develop a videotex service. How-
ever, PTT cooperation with cable television was controversial (Tutt, 1990).
Protracted debate occurred on privacy legislation. Following a commission's
reports of 1974 and 1976, a bill was introduced in 1981, but it was complex
and was withdrawn. The simplified substitute was debated for several years (de
Pous, 1988). After twenty years of discussion, a Data Protection Act was adopted
in 1988 and came into effect the next year. The Act required self-regulation in
the collection of personal data and established the right of individuals to access
and correct information files for a minimal fee.
172 The Netherlands
Equipment
In procuring equipment, the PTT has traditionally been close to the Dutch com-
pany Philips. Prior to the digitalization of the exchanges, Philips supplied about
70 percent of the PTT's SPC switches, with Ericsson's Dutch subsidiary sup-
plying the rest. In 1985, the PTT awarded a major contract for its 5ESS-PRX
digital switching system to the joint APT venture of Philips and AT&T. The
government also required the PTT to contract with a third and junior supplier
for digital switching, in addition to Ericsson and APT, to satisfy CGE-ITT
(Alcatel) with its subsidiary Nederlandsche Standard Electric Maatschappi
(NSEM) and its System 12. Subsequently, Alcatel, the second largest telecom-
munications firm in the world, established its formal headquarters in the Neth-
erlands, partly to soften its image as French dominated and to give it instead a
pan-European image. Much of Alcatel's actual management remained in France.
In the recent past, the PTT has had an extraordinary hold over customer
equipment. However, there has always been a distinction between the relative
severity of the letter of telecommunications regulation and the more flexible
reality. For example, large users understood well that the prohibition of sharing
leased lines was not an enforcement priority of the PTT. Similarly, although
the PTT had a total monopoly on all regular telephone sets until 1989, in prac-
tice, much customer equipment was purchased from private vendors (Hins and
Hugenholtz, 1986). The PTT also enjoyed a monopoly on the installation and
maintenance of PBXs and of the telephone sets attached to them. The PBX
remained the property of the PTT and was only leased to the user. In theory at
least, the PTT had the right to extend its monopoly into answering machines,
telefax, modems, and other data terminals, though it refrained from asserting
such broad authority.
Philips
Philips is the largest European electronics manufacturer and by far the largest
Dutch high-technology company (Tagliabue, 1987). Founded in 1894 by Anton
Frederik Philips and his brother Gerard to make electrical lamps, the company
keeps the name Gloewlampenfabrieken (incandescent lamp factory) in its title,
even though, beginning with the period after World War I, it has vastly ex-
panded its product line. Gerard, a mechanical engineer and chemist, initiated a
scientific tradition in the company that Gilles Hoist continued after World War
I. The company pioneered the electric dry shaver and was active in X-ray tubes
and hot gas engines. Gerard was childless, and the succession of control moved
to his brother's children (Philips, 1978).
Philips was hard hit by the Depression and later by the German occupation.
In anticipation of the occupation, Philips moved its legal headquarters to Cur-
acao, a Dutch possession safely located in the Western Hemisphere. Its exten-
sive American and British interests were placed in two separate trusts that were
The Netherlands 173
legally separate from the Dutch company. This maneuver guarded these parts
of the firm from take-over by the Allies after the German occupation of Hol-
land.
World War II put the company into the difficult situation of complying with
German-appointed supervisors while simultaneously trying to avoid attracting
Allied bombardment. The German supervisors were not interested in merely
running the company. In one instance, Philips was pressured to give up two
paintings by Lucas Cranach the Elder and coveted by Field Marshall Hermann
Goering, lest its raw material supply be cut off (Philips, 1978).
The Germans ordered Philips to reduce prices for its light bulbs. Frederik
Philips' comments on this incident in his corporate history provides a glimpse
into the protected Dutch electrical equipment market:
The former manager of our sales organization for the Netherlands and her colonies
had firmly believed in high prices for our lamps and would never budge. The result
was that, at the beginning of the occupation, there were in Holland no less than
twenty-three lamp factories, large and small, which survived under this heightened
price umbrella. Possibly the manufacturers loved us for this, but our lamps were
more expensive in Holland than abroada fact that the public knew. I decided that
we would lower our prices after the war by 50 percent. However, the Germans
were ahead of us: they suddenly ordered a substantial cut in retail prices [Philips,
1978].
After liberation, the company expanded substantially into other European
countries, the United States, and the Third World. It acquired Pye in Britain
and part of a major cable manufacturer in Germany. Philips became a world
pioneer in consumer electronicsfor example, in compact audio cassettes and
in video cassette recorders, where it introduced the world's first consumer VCR
(Video 2000) in 1972. In 1986, it adopted the Japanese VHS standards under
license for its VCRs, and withdrew its own system. However, by 1990 Philips
had only a small share of the European VCR market and a much smaller pro-
portion globally because of vigorous Asian competition. Together with the Jap-
anese firm, Sony, Philips established standards for optical laser compact disc
systems, but here, too, it was left behind by Japanese price competition. Sim-
ilar pressure on its other product lines led Philips to change its strategy from
in-house development to one of joint ventures with American, Japanese, and
other European firms. In 1984, only 20 percent of Philips' products were man-
ufactured by one of its own companies; many of the others were produced by
other companies and sold under the Philips label.
Philips spent over $1 billion a year on research and development, the third
largest such budget after IBM and AT&T (Tagliabue, 1984). It was Holland's
most important firm in the development of microprocessors and other semicon-
ductors, with its subsidiaries Valvo (Germany), Mullard (Britain), Complelec
(France), and Signetics (America) (Malerba, 1985). Philips also took over the
German firm Grundig, which, before faltering financially, had been Europe's
second-largest consumer electronics company. Together with Siemens, Philips
worked on developing advanced memory "megachip" components.
In telecommunications, Philips concentrated on cable transmission, tele-
174 The Netherlands
phone sets, PBX terminals, and smaller switches rather than on large central
office switches. Because it had little advanced digital technology, it entered
into a joint venture with AT&T in 1982. APT (AT&T-Philips Telecommuni-
cations) was carefully structured as a European venture; chartered in the Neth-
erlands, its subsidiaries are entirely European. The venture led Philips to stop
developing its own PRXD digital switch and adopt the AT&T 5ESS-PRX switch.
APT competed unsuccessfully in Britain and Italy; it received a Dutch con-
tract, some Saudi Arabian business, and a larger Indonesian award, and com-
bined with Alcatel to modernize the Venezuelan telephone system.
The joint venture was not as successful as originally hoped. Furthermore,
Philips was under financial pressure. In 1989, AT&T increased its share from
60 percent to 85 percent in the joint venture. Later that year, Italy's Itatel and
AT&T formed a joint venture (the former took a 20 percent stake in the proj-
ect). In 1990, AT&T purchased Philips' remaining 15 percent share and sold 6
percent to Spain's Telefonica (Wilde, 1990).
In 1990, Philips took a huge $1.4 billion charge against earnings and elimi-
nated 10,000 jobs from its workforce of almost 300,000 worldwide, mostly
because of its faltering computer and electronics divisions.
Policy Debate and Reform Legislation
An early attempt at policy discussion took place in the early 1960s, when the
so-called Goedhart Committee studied the possibility of PTT autonomy in in-
vestments and borrowing. But the committee's 1963 recommendation for greater
flexibility was never acted upon (Wieland, 1986).
In 1981, widespread dissatisfaction among users and equipment suppliers led
the government to appoint a new commission, chaired by Frans Swarttouw,
chairman of the Fokker aircraft manufacturing firm, to investigate the present
and future of the PTT and to recommend necessary reforms. In 1982, this
commission proposed to grant the PTT more autonomy in order to enter finan-
cial markets independently and to set rates and services with greater flexibility.
At the same time, it recommended that private competitors be permitted to
enter the terminal equipment market, while leaving intact the PTT's monopoly
for traditional services. The commission advised that the PTT be transformed
into a limited-liability company owned by the government, rather than remain
a part of the Ministry of Transport and Public Works. It further urged that the
new "enhanced" serviceslinking computers and communications to poten-
tially include videotex, video conferencing, and electronic mailshould be opened
to private service providers using the PTT network. Telecommunications op-
erations would be separate from postal operations, and PTT financial activities
would be placed into a separate "postbank."
Shortly after the report's release, the previous center-right government was
displaced by a left-of-center coalition. This enabled Philips (the major equip-
ment supplier) and the trade unions to block the recommended changes. In
1985, however, the Christian Democrats and Liberals regained power and de-
The Netherlands 175
clared their support of the Swarttouw recommendations. Another commission

was appointed to investigate the possibility of integrating telecommunications
with the numerous cable television networks, a move in response to the slow
but steady trend toward liberalization that had been going on in Dutch broad-
casting since 1969. T. H. Steenbergen, an economic liberal, chaired the new
commission, which consisted of two other members, one an industrialist with
Social Democratic leanings and the other a professor of engineering whose
Danish origin may have encouraged hopes of domestic neutrality. With the
assistance of the large American consulting firm McKinsey, and without the
cumbersome membership by representatives of every interest group and party,
the commission provided a report within the year, in July 1985.
The Steenbergen Commission report proposed a twofold reorganization: first,
a functional separation of the PTT's telecommunications activities into three
areas. "Public utility services"to include public networks, basic services,
and basic terminal equipmentwere distinguished from "competitive ser-
vices," which included user group networks, value-added networks, and com-
plex terminal equipment. Although both functions would remain under the aus-
pices of the PTT, "competitive services," unlike "public utility services,"
would lose monopolistic exclusivity. The third function of regulation, equip-
ment approval, and standardization, would be taken away from the PTT alto-
gether and lodged in a newly created department for telecommunications regu-
lation and approval under the jurisdiction of the Ministry of Transport and
Public Works (RUPTT).
The minister would regulate the PTT's "public utility" rates and quality,
approve equipment and operating licenses for value-added and regular net-
works, and allocate frequencies. The two limited companies would share a
board of directors, top management, and some central administrative functions,
although financially they would be independent of each other. In addition, no
hidden funds would flow between social and competitive tasks in telecommu-
nications. Competitive services would have to pay the regular value-added tax,
and PTT employees would lose their civil service status by 1989. The commis-
sion suggested that private value-added services would be a possible way of
dealing with competitive and network services without establishing alternative
networks, while at the same time solving the resale problem. Licenses for the
value-added network services would be sought from the approvals office, and
the services would have to be specific enough so that the system would be
protected from pure resale.
In addition to this functional separation, the report recommended that the
legal status of the PTT itself be transformed into a private holding company
(Royal PTT Nederland N.V.), entirely owned by the state, which in turn would
own three separate limited companies: for postal services, for basic public tele-
communications, and for competitive telecommunications services and prod-
ucts. However, the separation of telecommunications into two segments was
not adopted by the government. The basic and competitive tasks were left within
PTT Telecom Nederland B.V. after PTT resistance and were separated for the
time being only by undefined accounting procedures. The government-stated
176 The Netherlands
intention was ultimately to separate the two by legal means. It insisted on

continuing to receive from the holding company about the same contribution to
its general budget as before.
Both the Swarttouw and the Steenbergen reports proposed to end the PTT
monopoly over terminal equipment. But the actual loss to the PTT from liber-
alization was not great because privately supplied terminal equipment was al-
ready de facto widely available. The commission recommended greater com-
petition for PBXs to reduce their price. Partly because of high costs due to the
existence of only two suppliers, Philips and Ericsson, the Netherlands had only
half as many PBXs as West Germany. According to the commission, only four
models of small PBXs (for up to 100 lines) were available in Holland, com-
pared to seven such models available in Germany and twenty-five in the United
States. For medium-sized PBXs (100 to 400 lines), only one model was avail-
able in Holland, seven in Germany, and sixty in the United States (Wieland,
1986).
The government's official response to the Steenbergen report cautiously ac-
knowledged its conclusions: "Let there be no doubt that this government is in
favor of deregulation and privatization/denationalization, but it must be done
in a responsible fashion" (Netherlands PTT, 1985, p. 10). The PTT took a
critical stance in reaction to the report and presented itself as the last line of
defense for order. Hinting at a bias of the commission toward IBM, the PTT
claimed, "It is hard to avoid the conclusion that the Committee has chosen in
favor of computer suppliers who never cease to distance themselves from tele-
communications companies in Europe, mainly PTTs."
But despite several rearguard arguments, the PTT eventually reconciled itself
to change and concluded that reform was, in fact, in its own self-interest. The
Steenbergen recommendations were, after all, hardly radical. A German Bun-
despost study of the Dutch situation surmised that the change of heart in the
Dutch PTT may have reflected a generational shift in attitude, whereby the
younger management in leadership positions preferred a greater managerial in-
dependence outside of a government operation. Meanwhile, union members,
the majority of whom actively followed the deliberations, concluded that their
wages, salaries, and especially pensions would be improved by a switch from
a civil service to independent corporate status (Wieland, 1986).
Under the old system, PTT employees were paid as civil servants and were
tied to the entire bureaucracy's general pay scale. The many technically quali-
fied personnel needed by the PTT were in great demand by other employers,
and the maximum government salary was often not sufficient to draw them into
civil service. At the Dr. Neher Laboratories, for example, a large number of
all positions remained unfilled for that reason. In many instances, the PTT was
able to hire only recent graduates who sought to gain experience before moving
on to the private sector.
The PTT's new stance was strengthened by the fact that the Dutch equipment
industry that opposed the reform, fearing a loss in market share, was less pow-
erful than those in other countries. Since Philips is a greatly diversified global
The Netherlands 177
corporation, it was not as dependent on the PTT's status quo as some manufac-
turers in other European countries.
The government cabinet approved the Steenbergen proposal in 1985 and es-
tablished a phase-in period lasting until January 1, 1989. The proposal was
approved by the Dutch parliament in 1986 and included further provisions to
privatize much of PTT Telecom by selling its shares to the public by the mid-
1990s. A new department of telecommunications was created in the Ministry
of Transport and Public Works. The ministry oversees operator licenses, equip-
ment type approvals, and public service requirements.
Following the 1989 restructuring, PTT Telecom Nederland had exclusive
responsibilities for basic service, transmission, telex and telegraph service and
data transport services. Valued-added services and customer premises equip-
ment were liberalized.3 But these changes still left many large users, VANs,
and equipment suppliers dissatisfied. PTT Telecom faced complaints from users
over this role as equipment supplier and service provider. It did not allow
interconnection with the public network. Value-added services may be provided
only on the condition that the public infrastucture is used.
The $800 million equipment market was liberalized in 1989 but exclusive
distribution contracts between the PTT and Ericsson and Philips limited com-
petition and raised prices. The PTT still supplied a majority of basic phone
sets. However, a significant portion of the PTT's PBX market now belongs to
Ericsson and Philips. A Philips PBX costs 25 percent to 33 percent more in the
Netherlands than in the United States or West Germany. Responsibility for
testing was given to two independent institutes, KEMA and NTK.
A blue-ribbon panel, the Zegveld Commission, was appointed to investigate
the newly organized PTT's possible relationships with the municipal, indepen-
dent, and private cable networks. In 1986, it advised the gradual integration,
over a period of twenty years, of the cable network with the PTT-owned infra-
structure.
11
Belgium
With the erosion of its traditional economic base of heavy industry, Belgium's
economic priority is to become Europe's administrative center. Brussels already
serves as the headquarters for both the European Community and NATO. The
country also seeks to draw nongovernmental organizations; its capital city is
home to the European headquarters of many overseas companies and associa-
tions. Telecommunications links should therefore be an especially important
factor in the country's economic development. However, Belgium's traditional
monopolist Regie des Telegraphes et des Telephones (RTT), financially con-
strained in its goal to be at the forefront of service provision and development,
has been one of Europe's strongest defenders of traditionalism. In this position
it was supported by its major equipment supplier, Alcatel's BTM, and by the
labor unions. As a consequence, Belgium has not become a center of European
telecommunications, a role that would be otherwise natural. Partly in order to
achieve that goal, a 1991 reform reorganized RTT into the semi-independent
Belgacom.
Telephony and the RTT
The telephone was first demonstrated in Belgium in 1876. When the govern-
ment showed no interest in the new invention, a private telephone company
was formed in 1879. Other companies soon followed and intense competition
developed in several cities. Encouraged by national and local governments, the
private firms merged in 1881 into the Compagnie Beige des Telephones Bell.
In 1883, telephone service was made subject to concessions granted in the
context of the public telegraph monopoly. Concessions were granted for twenty-
five years, at the end of which the physical plant was to be turned over to the
government at no cost to the state; the government also had the option of choosing
to purchase the plant after ten years. Concessionaires had to compensate the
government at the rate of 5 francs per telephone station.
In 1884, the government began construction of interexchange long distance
lines. Two years later, it started to provide local exchange service to small and
medium-sized cities that were left out by the private concessionaires and it
stopped issuing new licenses. When the ten-year option periods began to run
178
Belgium 179
out, the government took over the private Bell telephone systems. By 1896, no
private telephone exchanges remained.
The first central battery network in Europe was opened in Brussels in 1902,
with 4800 subscribers. Much of the telephone plant was destroyed during World
War I, and it took until the end of 1920 to reestablish regular service. In 1922,
the first automatic central office was inaugurated. The establishment of tele-
phony and its improvement were hampered, however, by the weak financial
situation of the government. A submarine cable to England was laid in 1926
and established Belgium as an important transit point for international traffic.
A 1930 parliamentary act removed the telephone monopoly from direct state
control and placed it under the authority of the newly created RTT, an inde-
pendent state agency that has since been the nation's sole provider of telecom-
munications services. The act provided parliamentary controls, and RTT was
subject to the Ministry of PTT. It gave RTT some accounting flexibility, which
a pure state service would not have been granted. In an effort to unify the
status of the various semipublic organizations, however, parliament changed
the law in 1954, making RTT also report to the ministers of the budget, fi-
nance, and civil service, all of whom had some authority over aspects of RTT
operations. Subsequently, RTT lost even more autonomy. This was the reverse
of trends elsewhere, where PTT organization became more independent over
time. As a result, prior approval from one or several of the ministers in ques-
tion was needed before RTT could take action. Decision-making speed was
thus considerably reduced just when the opposite was needed in light of tech-
nological and business developments. These restrictions were supplemented by
additional ones, including controls on equipment procurement implemented in
1976.
During the early 1980s, RTT found itself in dire financial straits and had
difficulties meeting its debt burden. In 1985, its total outstanding debt was
about Bfr 180 billion ($3 billion), the service of which required nearly 30
percent of the agency's revenues. This situation created particular problems in
light of RTT's intention to enter in a significant way into digital switching.
The debt crisis created an impetus for rethinking the role of RTT, a process
that divided both the government and the PTT Ministry itself. In 1985, the
minister for communications, Herman de Croo, advocated a loosening of the
RTT monopoly. In contrast, RTT itself, the labor unions, and his parliamentary
state secretary, who was a member of another party, were opposed to such a
step, because they viewed these measures as steps toward privatization. Only a
limited liberalization of RTT activities took place; in 1985, the RTT was given
permission to invest in private telecommunications companies and to establish
"joint economy companies." Within RTT, there is a generational transition.
Older managers defended the monopoly, based on their notion of its social and
engineering function; younger managers tended to be more willing to experi-
ment with new structures.'
In 1986, the government established a commission of four university profes-
sors headed by F. DeBount to look into the question of RTT autonomy. This
commission of "four sages" delivered a very cautious report, staying within
180 Belgium
the narrow confines of what was deemed to be politically feasible. It advocated

a liberalization of terminal equipment, more autonomy to RTT managers, and
transformation of RTT into a public limited company, but only within several
years.
In the center-right government coalition, the powerful faction of Christian
Democrats was reluctant to liberalize. The conservatives' minister for the bud-
get, Verhoffsdt, was most strongly in favor of private initiatives but had little
support within the cabinet. The Social Democrats, the main opposition party of
the time, saw no need for new structures. The parties were also divided among
themselves according to language regions. Social Democrats from Flanders were
more flexible on telecommunications issues than some members of the Liberal
party from French-speaking Wallonia, since the protective attitude toward em-
ployment was much stronger in lagging Wallonia than in Flanders.
A force for change in Belgium has been the European Community's Com-
mission, which resides in Brussels and whose officials thus have daily firsthand
experience with RTT service and equipment options. Thus, RTT faced pressure
from the EC to end the embarrassing delays of up to eighty days for telephone
service for firms establishing offices there (Kellaway, 1990, p. IV). In 1986,
for example, the commission issued a directive to RTT to liberalize equipment
approval for terminals supplied by other countries.
After over thirty drafts and three years of debate, a reform bill finally passed
the Belgian Parliament in 1991. The law created a category of public autono-
mous enterprises for state-owned firms, and separated RTT into an operator
Belgacom and a regulatory body, the Belgian Institute for Telecommunications.
The Institute plays only an advisory role to the powerful Minister of Infrastruc-
ture of Economic Affairs, who is responsible for telecommunications matters.
Belgacom negotiated by 1992 a management contract outlining its budget, ser-
vice offerings, and investment priorities. A Competition Council was also es-
tablished to oversee public enterprises (R. Queck, 1991, communication). Bel-
gacom provides basic transmission as a monopoly, and enhanced services in
competition. However, authorized VAN licenses may be denied in the public
interest, or if the license would cause financial losses for Belgacom. Conditions
for the use of leased lines were also established. Belgacom also kept substantial
power over equipment, and its employees kept civil-servant status. On the other
hand, Bessel Kook, the former chief of the major user network SWIFT was
appointed as Belgacom's head.
In 1992, the privatization of Belgacom was proposed, partly in order to bal-
ance the national budget.
The Equipment Industry

RTT had a monopoly on some PBXs, which it procured mainly from its main
supplier BTM, and on first telephone sets. Liberalization emerged in the mar-
kets for more advanced equipment such as facsimile, teletext terminals, and
Belgium 181
cordless telephones. Policy debates focused on liberalizing all interconnection

equipment.
The Belgian telecommunications industry grew at a rate of more than 15
percent during the decade 1975-1985. More than half of its equipment produc-
tion was exported, and the country was in seventh place worldwide in net
telecommunications exports in 1983 (Belgian National Committee, 1983).
Among the manufacturing firms, ATEA predominantly supplies private users,
providing equipment such as computer terminals, telephones with magnetic card
readers, and PBXs. In the United States, the firm's equipment for credit card
checks has a major share of the market. ATEA was owned by the American
firm GTE until 1986, when GTE sold control to Siemens. Siemens employs
more than 2000 people in Belgian telecommunications and has engineering fa-
cilities for telecommunications and data-processing systems, as well as produc-
tion facilities for telex terminals and for radio, transmission, and switching
equipment.
The Dutch electric giant Philips has a Belgium telecommunications subsidi-
ary with about 1000 employees. The subsidiary produces transmission equip-
ment, multiplexers, integrated circuits, fiber-optic systems, PBXs, and radio
communications equipment.
ACEC, a large company employing 1500 in telecommunications, produces
teleprinters and satellite and fiber transmission systems. ACEC is owned by the
Societe Generale de Belgique and has links with Ericsson.
By far the most important Belgian equipment firm is Alcatel Bell Telephone,
formerly Bell Telephone Manufacturing Company (BTMC or BTM). Despite
its name, for many years BTM was an ITT firm, having been sold by AT&T
to ITT. It became a pillar of ITT's European operations until the 1987 ITT-
CGE deal transferred ITT's manufacturing operations into Alcatel, controlled
by France's CGE. Partly because of the importance of Alcatel Bell Telephone
in Belgian high technology, there is a special Belgian financial participation in
Alcatel via the holdings of the large conglomerate Societe Generale de Bel-
gique. In 1985, BTM employed more than 10,000 people, 85 percent of which
were in telecommunications. In 1990, Alcatel Bell Telephone employed 7455
people, all of them in telecommunications. The company exports to more than
eighty countries, and its digital switches are sold, for example, in Denmark,
South Korea, Israel, Taiwan, Switzerland, Mexico, Nepal, China, and Turkey
(G. Buyck, 1987, communication). 2
In 1985, BTM won a contract to supply China with System 12 digital ex-
changes for several cities, including Shanghai, and entered a joint manufactur-
ing venture with a production capacity of 300,000 lines annually. The United
States government objected to a transfer of such technology to China.
Historically, RTT had a relatively cozy relationship with BTM and ATEA
in the procurement of switches, and it maintained an informal "three-to-one"
order ratio between the two. In view of its location in the heart of European
integration, RTT cannot be blatantly protectionist. Nevertheless, RTT's 1987
digital switching contractdubbed "the sale of the century"was granted to
182 Belgium
BTM (66 percent) and ATEA (33 percent) (Miiller, 1988). Technically, both
companies are controlled by firms in other EC countries. The procurement pro-
cess is also complicated by the traditional rivalry between Flanders and Wal-
lonia. During price negotiations, the firms asked for $514 per line, whereas the
RTT refused to offer more than $420 per line. The RTT then decided it would
place the extra $92 per line into a R&D budget independent of the individual
suppliers but benefiting them (Euro-Telecom News, 1987). In the United States,
digital exchange for 1987 delivery was priced between $180 and $213 per line,
according to the U.S. Justice Department (U.S. Department of Justice, 1987).
In 1987, RTT filed a complaint against GB-Inno, a supermarket chain that
was selling nonapproved second telephone equipment. GB replied with a coun-
terclaim against RTT's equipment approval methods. European court proceed-
ings revealed that RTT's charges for equipment testing were often inflated.
Further inquiries demonstrated that the RTT was more likely to approve Bel-
gian equipment: 45 percent of overseas terminals applications failed compared
with 11 percent of Belgian terminals 3 (Tutt, 1989).
In 1989, the government reluctantly agreed to follow EC liberalization poli-
cies; but the actual process fell behind other European countries because of the
government's fear for the impact on jobs. It similarly opposed the EC's policies
on tariff principles and Open Network Provision.
Services
In 1988, telephone set density was 49.8 per 100 population (RTT, 1987). There
were 3.7 million telephone main lines in 1990. Rates were sixth lowest for
business and eighth lowest for residential service among the twenty-four OECD
countries. Residential and business users faced up to a six-month wait for ser-
vice.
An important first occurred in 1956, when RTT inaugurated, together with
France, international direct-dial telephone service, the first such service in
Europe. The installation of semielectronic space division stored program-con-
trol switches began in the early 1970s. The first fully electronic digital ex-
change became operational in 1983. In 1989, digital exchanges comprised 20
percent of the total network.
Data transmission began to grow strongly in the early 1980s. The RTT had
a monopoly on low-speed modems. It was strict in preventing resale. Shared
use was authorized under certain conditions, provided that volume tariffs were
applied. RTT was particularly concerned with the leakage into the public switched
network of calls carried on international leased lines (Pichault, 1985). Alcatel
entered the data services market in 1990 through a joint venture with RTT to
operate private networks and VANs.
Telex is also well developed in Belgium, though declining. (ITU, 1990a).
Mailbox service and video conferencing were made available experimentally in
1986.
Belgium 183
In 1982, a packet-switched network was put into operation under the name
of Data Communication Service (DCS-NET), with transmission rates up to 48
kbps. In 1990, RTT agreed to lift restrictions on international leased lines,
including mandatory routing through DCS-NET, after a lawsuit by GE Infor-
mation Services charged RTT with violating the EC competition policy.
In 1984, RTT initiated prototype ISDN trials, using the BTM System 12 and
ATEA switches. Belgium became the fourth European country to launch an
ISDN service, under the name of Aline.
Mobile telephone service was first provided in 1977, with equipment provi-
sion under the monopoly of the RTT and produced by BTM. A cellular system
went into operation in 1987 and is integrated within a Benelux system. In 1989,
RTT chose Siemens and Philips as the suppliers of a new digital mobile net-
work.
Videotex was launched by the RTT in 1986. Databases are mainly supplied
by private parties. One important and delicate task in language-divided Bel-
gium is how to allow connection to the surrounding countries that follow dif-
ferent protocols. Therefore, the system follows the British Prestel standard used
in the Netherlands, but it also allows the French Antiope standard, with access
to host computers in France. Connection to the German Btx was also provided.
In 1990, the RTT began testing the service over cable networks, since over
half a million Belgian television sets had teletext decoders suitable for some
videotex service.
12
Luxembourg
Luxembourg's tiny size and population seem to provide little incentive for new
entrants. The country is active in international financial services, incorpora-
tions, television, satellites, and even shipping registration. Telecommunications
is an important input for each of them. The PTT, under the supervision of the
minister for transport, communications, and information, held the national ser-
vice monopoly. It followed a policy of providing services compatible with those
of neighboring countries. In contrast to Luxembourg's minor role in telecom-
munications, its position in mass media transmission was far more significant.
Here both the television company CLT and the broadcast satellite provider SES
were at the forefront of Europe-wide change.
Between 1862 and 1880, the Luxembourg government installed a telegraph
network connecting Luxembourg with its neighbors to promote the country's
economic expansion. Telephone service was introduced in 1881, initially only
for verbal transmission of telegraph messages. In 1883, the administrations for
posts and for telegraphs were united to become the Administration for Posts
and Telecommunications (APT), and the Ministry of Finances proposed the
first telephone legislation. Under the provisions of the law of 1884, which is
still in effect, the government was authorized to construct or operate the tele-
phone network for its own use or to concede construction or operation to a
private firm. The government decided to undertake the project and built a net-
work in the capital city, which was eventually expanded (Bode, 1985). This
network opened in 1885.
To deal with low rural penetration, public phone booths, called auxiliary
telephone agencies, were installed, managed by private individuals but financed
by the state. Communities without such service could connect a line to the
general network for an annual fee and the cost of line construction.
User tariffs were flat-rate (unlimited communications) until 1920, when a
volume-based tariff replaced it. But this led volume to fall by 66 percent in the
first year of its implementation.
Luxembourg's geographical position fostered the early development of an
international network, with connections to Belgium in 1898 and to France,
Germany, and Switzerland by 1904. During World War I, Germany controlled
the network and suspended civilian communication until the 1918 armistice.
Following the war, Luxembourg made an effort at modernization. In 1922
184
Luxembourg 185
the first electromechanical automatic switches were installed and transmission

quality was improved by converting all lines to double wire circuits.
The German invasion in 1940 severely damaged the communications infra-
structure, especially in the north. Reconstruction was completed in 1948, and
automation of the domestic service was finished in 1963.
Telex service, introduced in 1951, became fully automated in 1958. That
year also marked the first stage of automation for international connections
when Luxembourg interconnected with West Germany.
During the 1970s, the APT set out to improve network penetration. The
Budgetary Law of 1973 created investment funds specifically for telecommu-
nications. Telephone density increased from 34.4 per 100 population in 1971
to 48.0 in 1990 (ITU, 1990). In 1974 a public airbill network opened. The first
digital connections were installed in 1976. Digital switching arrived in 1979,
and fiber trunks were installed in 1983. In the same year, the packet-switched
network Luxpac was inaugurated. Cellular service began in 1985 with a con-
nection to the Dutch MTX network. Teletex (e-mail) and videotex services
have been offered since 1986. By 1990, teletex had forty subscribers, videotex
had forty-four information providers and 360 subscribers (L. Kirch, 1990, com-
munication). Advanced data services, however, were slow in being offered.
Users had to wait six weeks for international leased lines and six months for
64-kbps data lines (Schenker, 1990). There were also waiting lists of three
months for regular service. Digital lines to France, Belgium and Germany were
established in 1989.
Luxembourg has no telecommunications equipment manufacturing industry,
and procurement is based on international tenders. The APT provides ordinary
telephone sets and detached moderns.
The APT is liberal-minded when it comes to upper level services and resale
but strict on basic infrastructure. The notion of rival transmission for a small
country like Luxembourg is not entirely far-fetched. A potential rival transmis-
sion system, at least for international service, already existed through the Astra
television satellite system of the Societe Europeenne des Satellites (SES), which
could potentially also be used for data and voice. APT opposed any loosening
of the infrastructure monopoly. But resale was permitted, and VAN and resale
services did not require licenses.
APT was hampered by its lack of independence over its budget and hiring
levels, since these were determined as part of the national budget. It therefore
sought and obtained in 1990 legislative approval to convert into a public, state-
owned corporation, effective in 1992. Regulatory functions were left with the
Minister of Communications, and APT was divided into posts and telecom-
munications branches, with a monopoly over public and infrastructure services.
Employee rights were maintained. The state retained the right to 20 percent of
income, and ordered APT to invest LuF 2 billion, the equivalent of 2 percent
of the national budget, in network digitalization.
13
Switzerland
The Swiss PTT is one of Europe's last "classic" PTTs: It is an exclusive and
monopolistic government administration where telecommunications subsidize
the postal service with which they are fully integrated. In its equipment inter-
connection policy and procurement practices, the Swiss PTT is among the more
rigid in Europe. At the same time, the quality of its servicein terms of both
equipment and manpoweris good, and the density of its coverage is among
the highest in the world. Thus, the traditional arrangement has worked reason-
ably well when the priority was to spread service to every household and when
technology and large-user demands were relatively stable.
Switzerland is a traditional center for financial and other international ser-
vices. However, rural interests are also very strong. In telecommunications
policy, they are allied with the labor movement and the advanced but high-cost
electronics sector, which in turn has close ties with the banking industry that
owns much of it. Nevertheless, it seems inevitable that Switzerland will have
to reconcile telecommunications with its traditional economic function as an
international cross-roads. The increasingly electronic form of financial transac-
tions and their resultant distance-insensitivity puts Swiss banks into direct com-
petition with such centers as London, New York, Singapore, Tokyo, and Hong
Kong. It is therefore likely that the country will move toward a more liberal
system that has characterized the rest of its economy, while finding new ways
of safeguarding its traditional concerns.
History
Historically, the postal service has been one of the relatively few central gov-
ernmental functions of the Swiss Confederacy. With the advent of the telegraph
in 1851, telegraph service was similarly put under national control of the Tele-
graph Directorate. The Swiss telegraph network began operation in 1852 and
international telegram service was started in the following year (Swiss PTT,
1983). When the telephone was introduced, the Swiss Federal Council passed
restrictive licensing regulations, distinguishing between telephone traffic in areas
where telegraph service existed and areas where it did not. Where the govern-
ment telegraph monopoly was affected, a license fee had to be paid that amounted
186
Switzerland 187
to at least two to three times the costs that had already been spent on telegraph
service in the area. Protection of the state monopoly was a major policy goal
from the beginning of the development of Swiss telecommunications. An early
telephone pioneer, Wilhelm Ehrenberg, challenged these restrictions, but his
proposal was rejected. The Telegraph Directorate was slow in providing service
itself, however. After a series of discussions, a license was extended to a com-
pany led by J. Ryf, a member of parliament, and T. F. Wild, a partner in a
major publishing and printing firm. As it later transpired, Wild also represented
the International Bell Telephone Company.
In 1880, the private Zurich Telephone Company was granted a license. Ser-
vice began almost immediately. In the first telephone directory, more than 10
percent of the subscribers were affiliated with Wild's publishing firm (Hof-
rnann, 1980). The status of the new company, however, soon became uncer-
tain. Following a series of disputes, the International Bell Telephone Company
withdrew, leaving the Zurich Telephone Company as a purely Swiss enterprise.
But this did not help to alleviate the political problems concerning the scope of
the Zurich company's license. The term of its franchise was reduced from twenty-
five years to five years. Recognizing the profitability of telephone service, the
Swiss government then decided to cease granting further private licenses and
instead to establish a state-operated service wherever feasible. In December
1880, this decision was ratified by the parliament, which wrote into law the
telephone monopoly of the federal government.
The telegraph administration opened its first urban telephone system in Basel
in 1881 and then in Berne at the bankers' association initiative. In 1883, the
first local networks were interconnected. In 1884, the operating license of the
private Zurich Telephone Company expired and was taken over by the govern-
mental system, with compensation paid. Today the Swiss PTT is more gracious
than other PTTs in acknowledging the role of early private operations. Its of-
ficial publication notes that "with hindsight, one must concede that the Zurich
Telephone Company performed a valuable service in acting as a driving force
and saving the Federal Government from paying dearly for their experience"
(Kobelt, 1980). By 1892, nearly all local systems were interconnected. Trunks
were initially run on overhead lines along railroad tracks and were later buried.
Because of Switzerland's status as an international financial center, interna-
tional telephone communications were implemented almost immediately. Ex-
periments were undertaken as early as 1878 with a connection to the Milan
telegraph office using telegraph lines. Cross-border telephone service on a reg-
ular basis was started in 1886 from Basel to Alsace and then rapidly expanded
to other countries. The first transatlantic radio telephone service opened in 1928
via London. During World War II, direct shortwave service to the United States
was introduced. In 1956, the PTT offered transatlantic calls over submarine
telephone cables, and after 1965, via satellite.
Semiautomatic switching was begun in 1917, using equipment from the Bel-
gian Bell Telephone Manufacturing Company (BTM). BTM, then owned by
AT&T, set up the Swiss subsidiary Standard Telephone and Radio (STR) and
became a major supplier. As an ITT company, STR introduced the crossbar
188 Switzerland
switch into Switzerland in 1965. Much of the automation of the network was
done with Strowger switches in cooperation with Siemens through a Swiss sub-
sidiary that is now named Siemens-Albis. Automatic long-distance service be-
gan in 1950 and was completed by 1959. Automated international service was
introduced in 1964. Some international telecommunications services were also
provided by the government-owned Radio Suisse.
Services
Switzerland has one of the highest telephone densities in the world. In 1988,
density was fifty-four telephone lines and eighty-eight telephone units per 100
population. Like Sweden, Switzerland's neutrality during the world wars spared
it from the destruction and resource drain that other countries suffered (Ducom-
mun and Keller, 1980).
Expansion in international traffic was especially great. More international
calls are made per capita from Switzerland than from any other country, aver-
aging twenty-eight per person yearly. Calls to West Germany, France and Italy
account for over 60 percent of that total (ITU, 1989, p. 764; Staple, 1990, p.
35). With Germany, Switzerland has the highest telex service density in the
world, with 530 telex connections per 100,000 population in 1988. Switzer-
land's extensive telex service is partly due to its importance in the financial
sector.
Not surprisingly for the world's center of watch making, Switzerland's first
special service was the transmission of the time signal: In 1935, a "speaking
clock" was introduced. Also provided today are regularly updated weather
forecasts, reports on avalanches and on ski and road conditions; sports results
and betting information.
Additional services include car radio paging, introduced in 1958, and mobile
telephony, available regionally since 1952 and nationwide since 1980 under the
name of Natel. Cellular telephony was introduced in 1978, and digitally after
1991. In 1982 the public packet-switched network Telepac became available.
Other PTT services include teletex (electronic mail), telefax, as well as a pop-
ular call forwarding service, Omnitel. Leased lines are available, under restric-
tions, and the need for private networks must be demonstrated. According to
the PTT, customer requirements should generally be met through the public
network. In 1988, the PTT began a Communes Modeles Suisses pour la Com-
munication (CMC) program. This introduced advanced servicessuch as vi-
deoconferencing, remote database, access and distance learningto twelve mu-
nicipalities to assess their communications needs. Videotex operations began in
1983 as a field test, and in 1987 as a regular service, with a faster Supertex
version offered in 1989. Switzerland has been and remains among Europe's
leaders in cashless public telephony, with its prepaid "taxcards" (Purton, 1990).
The PTT has been in charge of radio transmissions for the official Swiss
Broadcasting Corporation (SBC) since 1931. It offered, almost from the begin-
ning of radio, transmission service over wire lines in areas of poor reception,
Switzerland 189
an important consideration in a mountainous country such as Switzerland. In

1933, one foreign and three national radio channels were available over wire.
By 1956, there were six channels. The number of wire radio subscribers reached
its peak in 1969 with 440,000 subscribers. The service continues and is used
particularly in such institutions as hospitals, hostels, and hotels. In 1952, wire
radio broadcasting moved from the PTT to the SBC in terms of program con-
tent. The success of Swiss cable radio, preceding that of cable television by
decades, was remarkable, though it did not receive much attention.
A small additional provider of Swiss telecommunications service is Radio-
Suisse, a mixed-economy service organization 95 percent owned by the gov-
ernment. Until 1986, Radio-Suisse provided international telex, land-to-ships
and land-to-aircraft communications, and such services as voice storage and
retrieval, paging, consulting, and databases. It also provided civil air naviga-
tion for Switzerland. Radio-Suisse's service with ships at sea had a good rep-
utation, and ships in the Atlantic or the Mediterranean often interconnected via
Radio-Suisse into the European public telephone or telex networks, despite
Switzerland's being a land-locked country. Throughout the 1980s, however,
Radio-Suisse had difficulties, partly because the PTT occupied profitable lines
of telecommunication service in international telex and data transmission. It
was also in a legally complicated situation insofar as it provided services both
in monopolistic and competitive fields. This led to a 1984 report by the Federal
Ministry of Justice raising the question of the legality of the nominally private
Radio-Suisse providing monopolized governmental services in Switzerland (even
though the government owned 95 percent of the company).
Between 1986 and 1988, much of Radio-Suisse's telecommunications facil-
ities were transferred to the PTT. Radio-Suisse was thereafter left with a con-
sulting business and information service activities (i.e., largely competitive ser-
vices).
Switzerland participated in various European efforts on transborder data flow
protection, but a national privacy law was slow to be enacted, because of the
problem with Swiss banking and financial secrecy.
The PTT and Reform
The postal, telegraph, and telephone administration was formed in 1922 by

combining existing functions. It is the largest Swiss employer, with 61,000
employees in 1989, about one-third of which are in telecommunications (Wiihr-
mann, 1989).
The monopoly of the Swiss PTT is based on Article 36 of the Swiss Federal
Constitution. The primary legislation is the Federal Telegraph and Telephone
Law of 1922, which controls the activities of the PTT itself. The Federal De-
partment of Transport, Communications, and Energy provides political super-
vision. It is run by a governing board and by three directors-general, one of
whom is in charge of the telecommunications department. Labor union repre-
sentatives occupy three of the board seats. As stated in the PTT's master plan,
190 Switzerland
its service principles are to work for the benefit of the community, to provide
service for the entire country under the same conditions and economic princi-
ples, to provide unrestricted access, to protect privacy, and to be aware that
"not everything technically achievable and economically viable is necessarily
beneficial for society as a whole." In interpreting the specific applications of
these principles, the PTT has considerable discretion.
Following the takeover of the private Zurich Telephone Company in 1890,
the government had promulgated an installation monopoly for the telegraph
administration. This was changed in 1921, after pressures from the Association
of Swiss Electrical Installation Firms, when the government ceded interior in-
stallations to officially licensed private firms. Despite agitation in the 1930s,
only equipment supplied and maintained by the PTT could be used for all
telephone sets and telex machines. Approved data and fax terminals could be
privately supplied. PBXs were supplied by the PTT, and only PTT modems
were permitted for data transmission over the switched network.
Because the Swiss telecommunications law dates back to 1922, many new
telecommunications services have had to be governed by regulations and de-
crees. The Swiss political system makes it particularly difficult to modify the
telecommunications structure. Any legislation must be approved by the Swiss
upper chamber, in which the rural cantons, with their small populations, have
a disproportionate influence. They resist any modification that might lead to an
increase in telephone rates or postal service in rural areas.1 During the 1980s,
questions began to be raised about the scope of the PTT's monopoly. In 1980,
Zurich groups sought a referendum initiative to reduce PTT power, but they
did not succeed in getting enough signatures (Knieps, 1985a).
These signs of public dissatisfaction started a legislative reform process in
1981, when the PTT directorate general decided to draft provisions to revise
the old Telecommunications Law. In 1983, the council of the PTT approved
an early version of such a bill. However, other parts of the government and
various interest groups claimed that the draft was overly protective of the PTT
and demanded a draft legislation to be designed by a body independent from
the PTT. In 1984, the minister of transportation and energy appointed a study
commission for a new telecommunications law. The commission, however,
was chaired by the secretary general of the ministry itself. Other members
represented interest groups, such as PTT officials, the PTT union, the national
chamber of commerce, the consumer federation, the small business association,
the equipment industry, and large users. Independent members included Pro-
fessor Carl Christian von Weizsacker, an economist who had also been active
in German telecommunications commissions. In 1985, the commission submit-
ted to the department a preliminary report that was endorsed by the governing
PTT boards.
The study group's recommendations and the 1985 draft legislation were cau-
tious. Even in terminal equipment, some areas would be left entirely to the
PTT, with governmentally drawn lines based on vague criteria such as "interest
of the country" and "security of transmission." Main stations terminal equip-
ment and telex machines would remain under the control of the PTT. Further-
Switzerland 191
more, approval was left to the PIT and thus subject to its time schedule. The
proposal also prohibited resale of transmission capacity (Blankart and Knieps,
1985).
The partial liberalization of terminal equipment was largely a confirmation
of reality. PTT exclusivity for simple terminal equipment had been undercut by
the illegal attachment of widely available telephone sets, marked "for export
only." Article 28 of the draft law provided that terminal equipment can be
advertised, sold, leased, or rented only if it is technically approved, that the
PTT has the right to control all terminal equipment, and that the owner of a
terminal must provide physical access to it. In accordance with Article 3, the
cabinet would decide which types of terminal equipment would be exclusively
provided by the PTT.
The ministry describes such restrictive equipment rules as a "step-wise lib-
eralization," which is an overstatement unless the government exercises its
line-drawing powers in a liberalizing fashion. The government's report alleged
that a competitive system would put into question the ability of some customers
to reach others, and that it would create disparities within the country. It also
stated that a small country such as Switzerland could provide an efficient tele-
communications network only under a single organization, and according to
unified norms. Despite these assertions, interconnection and compatibility could
be legally mandated; even if additional or specialized Swiss networks were not
economically feasible, there could still be general or specialized Europe-wide
networks, which would have the economies of scale that Switzerland might be
unable to provide. It would be pure happenstance if the optimal size of the
network coincided with the size of Switzerland.
The PTT also argued that the Swiss telecommunications industry cannot
presently compete with foreign imports:
A spontaneous and unrestricted opening up of Switzerland to the telecommunica-

tions markets would pose a serious threat to the Swiss telecommunications industry
and consequently affect the social and economic life of our country. At present the
Swiss telecommunications industry cannot offer products at competitive prices in
comparison with cheap products from the Far East or countries with a large do-
mestic market. As long as foreign markets remain closed to products from the
Swiss telecommunications industry, the existence of the latter can only be assured
by supplying equipment to the PTT. The PTT monopoly may thus, to some extent,
serve as a defensive measure against the protectionist attitude of foreign markets
with regards to indigenous products [Butikofer, 1987, p. 7].
To alleviate the problem of foreign competition, the government's position

was that it needed to ensure an adequate domestic supply by granting its indus-
try enough time to adjust to liberalization. The government acknowledged the
industry's interest in a PTT monopoly arrangement more honestly did than the
government of most other countries. The argument that protectionism abroad
has harmed Swiss manufacturing, however, is unpersuasive. It is true that many
countries have such protectionist rules on the first terminal set (though these
rules are fading in favor of EC norms), but if it were competitive in price,
192 Switzerland
features, and performance, Swiss equipment could be sold more widely across
Europe, America, and the developing world.
In 1989, a parliamentary draft of the new law proposed that services be
divided into basic and extended categories, with the former, such as transmis-
sion and switching, remaining the sole province of the PTT. Extended services
"which complement, modify, store, or otherwise manipulate messages" would
be opened to private competition. This would include second telephone sets,
telefax, modems, PBXs, and teletype terminals. Regulatory and operational
functions of the PTT would be separated.
The Swiss telecommunications industry is hampered by the relatively small size

of its home market (the PTT procures equipment costing about 1 billion Swiss
francs a year). Import orders are about 30 percent of sales, which is relatively
low for Swiss industry. The PTT is closely linked to the well-organized do-
mestic equipment industry. According to a survey by the Swiss Industry As-
sociation, the share of the PTT in the total of Swiss telecommunications man-
ufacturing was 80 percent in switching equipment, 60 percent in customer terminal
equipment, 42 percent in transmission equipment, and 40 percent in cable.
Upon the initiative of the major suppliers and a high PTT official, a "Pro
Telephone" trade association was formed in 1927 by representatives of ten
firms, renamed later Pro Telecom (Piquet, 1980). Its objective is to promote
telephone usage and to organize collaboration with the PTT in many areas. The
PTT is represented on two advertising committees, and Pro Telecom's manag-
ing director is a member of the PTT's marketing board. Several working groups
of Pro Telecom's R&D committee played a role in joint development by three
suppliers of a new electronic set telephone, TS85. Similarly, the semiconductor
components group promoted the use of Swiss components in electronic PBXs
and telephone sets. The manufacturers also cooperate in export drives through
Swisscom, Teleconseil, and Telesuisse.
Total telecommunications manufacturing employment in Switzerland num-
bered approximately 20,000. The largest Swiss telecommunications companies
by market share in 1990 were the following: (1) Ascom, which controls 51.6
percent of the market and makes mobile, standard, and cordless phones; PBXs;
and banking systems. (2) Siemens-AIbis, with 18.5 percent of the market, pro-
viding cables, wires, exchanges, modems, and multiplexers. (3) STR (Alcatel),
with 19.0 percent of the market. For those three large companies, telecommu-
nications equipment was by far the predominant part of their business (Blankart
and Schneider, 1984). Other firms in the telecommunications market include
BBC, Alteau, Landis & Gyr, and Gretag (Northern Business Information, 1990).
The narrowness of the market has led to concentration. In 1984, the Swiss
firms Hasler, Zellweger, Ustcr, Autophon, and Gfeller (the last of which is
controlled by Autophon) joined to form a cooperative arrangement under the
name of Ascom, for development, export, production, and purchases. This led
Switzerland 193
to the merging of the top two firms, Hasler and Autophon, in 1987 as part of
the Ascom holding company, which in turn owned 50 percent of Zellweger and
has a connection to the German firm Bosch. Ascom, through a holding com-
pany, became the dominant Swiss equipment firm, strengthening the position
of the purely Swiss companies versus Siemens-Albis and Alcatel's STR. An-
other cooperative venture is the Tritel telephone production agreement involv-
ing Gfeller, Autophon, and Zellweger (Union Bank, 1986).
After 1970, the PTT cooperated with Siemens-Albis, STR, and Hasler to
create the Integrated Telecommunications System, or ITS. The primary aim
was to develop a digital system to replace about 1000 analog switches. In 1983,
140 million Swiss francs later, the PTT decided instead to opt for a foreign-
developed system and was criticized that it had not established an effective
project management organization for such a large-scale project and that the
scope of the project had been underestimated (ASUT, 1983) .
The PTT decision did not, however, eliminate the three companies as poten-
tial suppliers of a digital switch. After the failure of the initial IPS concept, it
introduced in 1985 the concept of a new IFS-Swissnet, with foreign-developed
switches. Both Siemens-Albis and STR provided their parent companies' elec-
tronic switches, EWSD and System 12. Hasler, a Swiss company, offered the
AXE 10 switch of its traditional technology source L. M. Ericsson. However,
digitalization of switching was slow, with the first full digital exchanges being
put into operation in 1987 (Ducommun, 1987). The PTT predicted the connec-
tion of 90 percent of subscribers to SWISSNET by 1995 (Trachsel, 1987).
In the supply of customer equipment, the cumbersome system of approval
and adoption of PTT-owned equipment led to a substantal gray market. A con-
sumer movement magazine reported that the annual rental price for the PTT
telephone model TS85 with twenty-number storage was SFr 192, whereas a
corresponding foreign model could be purchased illegally in a store for only
SFr 125. Cordless telephones were officially introduced only in 1985, although
an active market in illegal equipment already existed.
When the official equipment was introduced, the monthly rental charges were
SFr 26 (then about $12). In the United States, these rates would pay for pur-
chasing the equipment in less than half a year. This also corresponds to a study
that found that the cost of some equipment available for purchase in the United
States was equal to its rental fee for three to four months in Switzerland (Blan-
kart and Schneider, 1984).
In 1983, the PTT still believed that the cordless equipment would cost SFr
1500-2000 (about $1000) per station. After much criticism, this was reduced
to a still high SFr 1000. At the same time, a letter to the editor of the major
newspaper Neue Ziircher Zeitung pointed out that they were available in the
United States for only $59. After a lengthy development process, the PTT in
1984 and 1985 issued a fifty-five page technical specifications manual that hap-
pened to match the system developed by the largest Swiss telecommunications
manufacturer. For example, the PTT required that each telephone have forty
duplex channels and an automatic frequency search scanner. Furthermore, the
timing of procurement for distribution was a few months hence, which was
194 Switzerland
almost impossible for foreign competitors to meet (letter to Neue Ziircher Zeit-
ung, January 6, 1984). The PTT's response (Neue Ziircher Zeitung, January
20, 1984) was that the cheaper equipment was probably an inferior Far Eastern
product and that it permitted others to eavesdrop or to illegally use someone
else's main stations.
Switzerland was not under a liberalization pressure from Brussels like most
of its neighbors since it was not a member of the EC. But in time, it could not
remain apart from the broader trends, with EFT A moving close to the EC. In
1990, even the PTT's telecommunications head admitted that
Switzerland has too long under-estimated the significance of developments outside
its borders. Given the worldwide expansion of markets and the European Com-
munity's internal market, our insular solutions could no longer be tolerated, espe-
cially for telecommunications. If our country is to profit from the economic liber-
alization and deregulation of Europe and remain competitive, liberalization is un-
avoidable (Wuhrmann, 1990).
Recognizing the significance of telecommunications to the Swiss role in inter-

national transactions, the PTT unveiled in 1991 an ambitious $10 billion three-
year plan to establish Switzerland as a European telecommunications center.
Telecommunications in Liechtenstein
Although the principality of Liechtenstein is independent, its telephone service
is largely handled by the Swiss PTT. Telephone service in Liechtenstein dates
back to the turn of the century, when it was operated by the Austrian govern-
ment telegraph administration and was linked to the Austrian system. With the
outbreak of World War I, Liechtenstein telephone service was totally isolated
from the rest of the world for a year, until it was connected to Switzerland.
Since 1921, the Swiss PTT has operated Liechtenstein telephony as part of
Swiss domestic service.
14
Austria
Austria's telecommunication system is a classic postal-industrial complex: a

powerful PTT monopoly closely allied with labor unions and with a virtual
cartel of private equipment firms known as the "Four Sisters." Because large
users are frequently state-owned firms or international organizations, little counter-
pressure exists.
History
Telecommunications began in Austria in 1846, when a railroad company con-

structed a telegraph line from Vienna to a suburb and onward to the city of
Brunn (Brno) for its evolving operational needs (Holcombe, 1911). Emperor
Ferdinand quickly declared telegraphy a governmental monopoly, the Telegra-
phenregal, assigning the new technology solely for the service of the railroads
and the state. His successor, Emperor Franz Joseph, cast the future of the
telecommunications monopoly by his 1849 decree, which, among other things,
permitted public access to the telegraph system.
Telephony was introduced in 1879 within the Austrian military. Civilian ser-
vice started in 1881 by the Vienna Privat-Telegraphen-Gesellschaft. Two years
later, the London-based Consolidated Telephone Construction and Maintenance
Company began service in the outlying cities of the Austrian empire, including
Prague, Trieste, Pilsen, and Brno. Once the various private companies had
demonstrated the profitability of the new service, the government, despite the
emperor's pronounced skepticism, put telephony within its domain in 1887 and
introduced a long-distance service connecting Vienna to Brno and then to other
cities. In 1888, the private and state networks were interconnected. The forty-
seven local networks that were linked in 1892 rapidly swelled to 198 by the
turn of the century. Full state monopoly was achieved in 1895, when the last
of the private systems was taken over. Communications became part of the
Commerce Ministry, were later moved into the Ministry of Transportation, and
were reunited with the postal service in 1923.
At the beginning of World War I, Austria was a sprawling empire with one
of the most extensive long-distance telephone systems in Europe. The first reg-
ular automated local exchange opened in 1910 in Graz. After the war, local
195
196 Austria
exchange service was considerably expanded. In 1938, Austrian telecommuni-

cations became part of the German Reichspost. During World War II the trunk
lines, exchanges, and equipment were substantially damaged, and afterward,
the Soviet Union dismantled some of the remaining equipment in its occupation
zone. With funds provided by the Marshall Plan during the reconstruction, the
second Austrian republic recovered rapidly to become a leader in the introduc-
tion of coaxial cable and microwave transmission (PTV, 1984).
The Structure of the PTV

Operating under the Telecommunications Law of 1949, the Osterreichische Post
und Telegraphenverwaltung (PTV) has a monopoly on telecommunications ser-
vices in Austria. The PTV, which employs about 58,000 people, 18,400 of
which are in the telecommunications sector (ITU, 1990), is a branch of the
Federal Ministry of Public Economy and Transport, with the PTV's director
general appointed by the minister (Bauer, 1986). Telecommunications (the "gray
post") is organizationally linked with the mail service (the "yellow post"). A
central office for telecommunications technology (FZA) oversees specifications
and type approval.
The PTV is subject to the parliament for its budget and investments, and for
the rates of inland tariffs. Parliament shifted rate-setting responsibility for in-
ternational tariffs to the ministries of Finance and Public Economy and Trans-
portation. Domestic rate changes, five-year investment plans, and various reg-
ulations must be approved in the form of legislation (Eward, 1984). There have
been minor modifications and some additions to the law, such as the 1977 law
on private telecommunications installations. The legal framework has been crit-
icized as outmoded and inconsistent, although there have been no serious at-
tempts to rewrite it. Most of the parliament's discussion of telecommunications
law is centered on tariffs. In Austria, an informal body known as the "Social
Partnership," which combines the major economic interest groups, considers
and resolves many of the country's significant issues, including those concern-
ing telecommunications. Telecommunications reform was not deemed neces-
sary.
Strictly speaking, the PTV's monopoly applies only to the public network. 1
It can provide other services itself or license others. The 1977 law on private
telecommunications installations (Privatfernmeldeanlagengesetz)specifies that
license applications can be rejected by the PTV if it deems the services tech-
nically inadequate, if the needs "can be served with the required reliability and
speed by existing services or facilities of the PTV," or if it is "against the
business interests of the PTV" (quoted in Bauer and Latzer, 1987, p. 8). In
effect, the PTV has a right of first refusal on new services. In one instance,
the PTV rejected Philips' interest in providing a videotex service on the grounds
of its "right of priority," and proceeded to provide Btx itself. Only the state-
owned Radio Austria, close to the PTV, received a license to provide transcon-
tinental telex and data transmission.
Austria 197
Services
In response to criticism throughout the 1960s that the PTV's restricted budget
and bureaucratic organization prevented modernization, the government granted
it a five-year investment plan. As a result, the organization could expand the
reach of the network.
Between 1970 and 1988, telephone density almost tripled from 13.1 to 39.5
per 100 population. At the end of 1989, there were 3.1 million main telephone
stations and about 21,500 telex connections in a population of about 7.6 mil-
lion. Telex density was a very high 0.34 percent of the population, but telex
traffic dropped 25 percent from 1988 to 1989 alone. It should be noted that
there is a marked difference between the number of main line installations and
actual telephone sets (3 million lines to 4.1 million sets in 1988) due to the
frequency of party lines. After 1970 there were periods when the waiting list
contained more than 200,000 applicants for telephone service and when delays
were over two months. But by 1989, the number of applicants on the waiting
list had greatly decreased to 35,000 (ITU, 1990).
Tariffs are high. Calls placed from Austria to other European countries tend
to cost twice as much as calls traveling in the opposite direction (TDR, 1988,
p. 9). Although the PTV claims to be adapting tariffs to real costs, Austria's
rates for business ($1408 per year) and residential ($529 per year) users are
among Europe's highest (Montgomery et al., 1990).
The Austrian telephone network is technically complicated, because it em-
ploys two sets of technologies, one of which uses nonlinear hybrid circuits at
the junction of two- and four-wire circuits, and thus suffers from performance
problems in data transmission. ITT electromechanical (HK8) and semielec-
tronic (11E) exchanges were installed in the south, west, and northeast of the
country. Elsewhere, Siemens ESK-A3, A5 and F semielectronic switches were
primarily used.
In 1978, the PTV decided in favor of digitalization. A concept was devel-
oped by the consortium organization OFEG, which brought together the Re-
public of Austria (51 percent), represented by the PTV and the Social Partners,
and the telecommunications manufacturers Siemens, ITT (now Alcatel), Kapsch,
and Schrack, which are collectively known as the Four Sisters. Not surpris-
ingly, the consortium decided to "carefully adapt" a foreign system to Aus-
trian specifications under control of the Four Sisters and the PTV, thereby
foreclosing the market to others.
Two supplier groups were chosen. The first group was comprised of the two
domestic Austrian firms Kapsch and Schrack, and offered a version of the Ca-
nadian Northern Telecom DMS-100 switch as the OES-E system. The second
group, Siemens and ITT, offered a Siemens EWSD system. There was some
competition between the two groups whereby the lower bidder receives an extra
10 percent of orders while the higher bidder adjusts its prices to the lower bid.
In 1986, digital operation commenced in Vienna. By 1990, there were more
than 500,000 digital subscriber lines linked to seventy switching centers. PTV
198 Austria
planned to increase that number to 2.3 million, or about two-thirds of all phones,
by 1995 (Sindelka, 1990, p. 50; CWI, 1990, p. 4). Since 1989, the PTV in-
vested in its broadband fiber trunk network ("6-Netz") (Purton, 1990, p. 13).
ISDN trials began in 1991.
In 1982, two new data servicescircuit-switched, Datex-L and packet-switched
Datex-Pwere introduced. Together with the existing telex network, they form
the integrated telex and data network (IFSD). In 1988, Datex-L had about 4000
subscribers and Datex-P had 2500. All Datex services had 10,300 total sub-
scribers in 1989.
Videotex, known in Austria as Btx, was introduced to the public in 1985.
Btx telephone rates are the same as those of local calls, even though the trans-
mission distance may be considerably greater. However, indicative of user ap-
athy, there were only 9717 registered subscribers in 1989.2 The Austrian Btx
terminal MUPID failed to gain acceptance after nine years of state support.
The market for videotex terminals was subsequently liberalized in 1989 (M.
Latzer, 1990, communication).
Cellular mobile telephony (Network C) was introduced in 1984 with Moto-
rola equipment, joined shortly thereafter by Ericsson (with Schrack) and Sie-
mens (with Kapsch). In 1990, the PTV awarded a contract for digital mobile
radio to Motorola, in cooperation with Austrian firms.
Private leased lines are available. Resale of lines is not permitted, but be-
cause of the difficulty of enforcement it exists informally. Since the introduc-
tion of its data networks, the PTV has discouraged, through rate setting, the
use of leased private circuits, and even aimed for regulatory measures to con-
vert all private leased circuits into circuits that can be used by public data
networks (Bruckner, 1982).
Because the Austrian telecommunications law has not been updated, the le-
gal status of value-added services by alternative providers is a complex matter.
The law does not differentiate between equipment and service. Approval by the
PTV is required for both the technical specifications of the connected equip-
ment and the corresponding software of value-added network services. The
extent of scrutiny varies between the various regions and depends partly on the
use of public networks versus leased lines and the applicability of public versus
private laws and contractual arrangements (J. Bauer, 1987, communication).
Equipment
Subscribers' equipment is supplied privately or by the PTV. The basic tele-

phone set, however, must be rented or bought from the PTV. The PTV is both
supplier and regulator of equipment. After 1987, the PTV followed a more
liberal policy for consumer premises equipment. In 1988 a triple socket for
connection of privately owned telephone terminal devices was introduced to
foster greater liberalization of equipment markets in practical terms (Sindelka,
1988).
Telex terminals can be obtained privately, but this policy is less open than it
Austria 199
may appear. In 1987, the PTV alone maintained all telex terminals and it had
certified only four telex models. The Four SistersSiemens, Alcatel 3 , and the
domestic firms of Kapsch and Schracksupply from 90 percent to 95 percent
of telecommunications equipment to the PTV (Bauer and Latzer, 1988;
CWI, 1990). They employ 20,000 people, with 11,000 in their telecommuni-
cations divisions. A comment in a PTV report reveals its attitude to foreign
equipment: "Since only a small share of the computer market is held by Aus-
trian equipment manufacturers, no restrictions are imposed on foreign [com-
puter] equipment" (Bruckner, 1982, p. 12). 4
Telecommunications in Scandinavia
and the North Atlantic
15
Sweden
Televerket, Sweden's telecommunications operator, has long pursued an effec-

tive technological and social agenda while creating a vertical integration that
reached into equipment manufacture. No statutory monopoly exists and the
introduction of competition was accomplished smoothly. The market reality is
one of Televerket dominance, given its managerial effectiveness and its politi-
cal support. Aggressive rivalry from a well-connected small rival, Comvik, and
a move toward partial privatization of Televerket promise to make Swedish
telecommunications one of the more interesting scenes of network evolution.
History
From 1853, Swedish telegraph service was mainly run by the government, but
no legal state monopoly on telecommunications existed when the telephone was
introduced (Holcombe, 1911). Initially, the government refrained from involve-
ment in telephone service, but by the 1890s its role had changed dramatically.
The first telephone connections were established in Stockholm in 1877. One
of the earliest users, H. T. Cedergren, later founded the General Telephone
Company (Gustafsson, 1987). In 1880, the International Bell Telephone Com-
pany opened local systems in Stockholm, Gothenburg, and Malmo. Almost
immediately, several domestic firms were attracted to this market by Bell's
high prices. Although Bell would not sell equipment to its competitors, L. M.
Ericsson entered the field in 1878 and proved a viable alternative in the absence
of telephone patents. The first non-Bell network, established in 1881, charged
lower rates than Bell. In 1883, Cedergren started the General Telephone Com-
pany, Allmanna Telefonaktiebolag, and within one year served three times as
many subscribers as Bell. The two firms engaged in a lively head-on competi-
tion in price and service, particularly in Stockholm. In 1887, General Tele-
phone opened the world's largest exchanges (7000 lines). By 1888, Bell suc-
cumbed and relinquished control to its Swedish competitor, which merged the
two systems.
Still, the demise of Bell did not signal the end of competition, although
Sweden's second phase of competition was less dramatic than its first. Small
stock companies, mutual associations, cooperative societies, and municipal sys-
203
204 Sweden
tems rapidly established themselves in rural areas and began to coexist through
long-distance trunk lines. The government Telegraph Company (Telegrafver-
ket) also began to construct long-distance lines, and in some instances acquired
or built local networks. In 1881, it started a local network interconnecting gov-
ernment offices. In 1883, a Royal Ordinance required the state's permission for
interconnection, and it has survived to the present day.
In 1889, General Telephone applied for rights-of-way to establish trunk lines
to connect its three municipal systems. Thus prompted to review its role in
long-distance telephony, the government decided to be the main builder of in-
terurban lines. Significantly, long-distance telephone service would only be readily
available to subscribers of state local exchanges. In 1891, General Telephone
reached an agreement with the state. It was licensed to operate and use other
rights-of-way for fifty years and allowed it access to the state's long-distance
lines. Outside a 70-kilometer radius from Stockholm, the telegraph authority
imposed access fees and technical requirements on private systems that were
so burdensome that many sold out to the government. Within three years, the
government had taken control of three-fourths of all the local systems.
General Telephone, however, did not give in. The government thereupon
expanded its small system in Stockholm, which had previously operated for
official administrative use only, into a general public network. In 1905, an act
on wireless telegraphy strengthened the state's centralized authority in granting
transmission rights (Heimburger, 1931), and by 1907, it renounced its willing-
ness to interconnect. At that time, General Telephone had 37,000 subscribers
in Stockholm, and the government had 13,000, but in the competition between
the two that ensued, the government, with its resources and long-distance in-
terconnection, held the upper hand.
Within Stockholm the vigorous rivalry led to reduced rates, high technical
performance, and experimentation by the government with new types of service
and billing (including usage-measured tariffs), making Stockholm's telephone
system the most advanced in the world. Whereas the original flat rate of the
Bell company had been about $40 in the inner city of Stockholm and $60-70
elsewhere, after a few years of this unique competition, rates fell to about $12
for residential customers and $15 for businessesthe lowest in Sweden (Hoi-
combe, 1911).
The government, with its superior resources and power, ultimately took con-
trol of General Telephone in 1918, and integrated it by 1923. General Tele-
phone's management moved to L. M. Ericsson, which became General
Telephone Company-L. M. Ericsson. The Swedish telecommunications ad-
ministration, subsequently named Televerket, thus assumed its role as sole ac-
tor in Swedish public telecommunications. The exceptions to its monopoly are
several other state administrations that can provide for their own communica-
tionsthe state railway company, the state power board, and the Swedish de-
fense forces, to name a few. The only real private encroachment to the monop-
oly since General Telephone has recently emerged. Technically, however,
Televerket never held a legal monopoly. To satisfy the distinction, in the fol-
lowing the word monopoly should be read as de facto monopoly.
Sweden 205
Televerket
Telecommunications in Sweden became more vertically integrated than almost
anywhere else in the world after the United States broke up much of AT&T's
integrated structure. Televerket (often referred to internationally as Swedish
Telecom) had a de facto service monopoly over local, long-distance, and inter-
national communications; in 1988, however, Comvik AB received permission
to begin an international satellite communication network. Televerket has also
long held a connection monopoly over much of the customer equipment in
voice telephony and telex, though that was stepwisely relaxed. In 1989, ter-
minal markets were fully liberalized for PBXs and high-speed modems (Thorn-
gren, 1990). Televerket still produces much equipment itself. Not even AT&T
in its heyday had the same end-to-end control over production, services, and
equipment connections; AT&T had no monopoly on international service, was
excluded from telegraphy, coexisted with about 2000 other local telephone
companies, and had to share its equipment patents with others.
Part of Televerket's reason for success has been that it has been long outside
the postal administration. It is a public service corporation and reports to the
Ministry of Transport and Communications. But whereas Televerket has over
40,000 employees, of whom about 30,000 serve basic telephony, the entire
ministry has seventy employees, of whom only six deal with telecommunica-
tions issues (Televerket, 1989, p. 7). In 1989, Sweden's pool of telecommu-
nications policy expertise was severly depleted by the deaths of the finance
director of Televerket and several members of the Ministry of Communication
in a plane crash.
Since 1975, Televerket has been organized into twenty local areas, each
operating with substantial independence. Televerket's board, appointed by the
government, represents the political parties, the business and scientific com-
munities, and several of Sweden's labor unions. A director general manages
day-to-day operations. In the past, the annual budget was proposed by the gov-
ernment and passed through the budgetary process of the parliament. Profits
had to be returned to the Treasury. Since 1980, the budget became independent
from the public sector budget, and profits are now returned to the Treasury
only at a predetermined rate-of-return on the capital that the government sup-
plies to Televerket. Televerket has also been allowed to borrow on the open
market and is thus unencumbered from general state budget difficulties, nor
does it need to subsidize other governmental functions, such as the postal ser-
vice. Since 1984, Televerket's supervision has followed three-year programs,
which it submits to the government for parliamentary approval. This plan in-
cludes an investment and finance schedule, production goals, and a rate-of-
return proposal. Once this plan is approved, Televerket is substantially free to
operate within its parameters. In 1989, Televerket proposed the expansion of
the existing de facto price cap system (Krzywicki, 1990), and in 1990, it sought
to be privatized.
Sweden's telephone penetration is high. Like Switzerland, Sweden was lucky
206 Sweden
to escape the drain and destruction of two world wars, and it experienced the
Great Depression in a relatively mild form only. Between 1930 and 1940, the
number of telephones increased by 70 percent (H. Bergendorff, 1987, com-
munication). The country enjoys high standards of public services and signifi-
cant income equality. In 1988, telephone density was 66.2 main lines per 100
inhabitants (ITU, 1990). (For a discussion of Televerket's data, see Chapter
32.)
Vertical Integration
The Televerket monopoly, for a time, extended beyond services to all customer
hand sets; after the 1970s it was liberalized for secondary sets, telex, and low-
speed modems. It maintained PBXs and high-speed modems until 1989.' Vid-
eotex terminals and cellular telephone equipment were never restricted (Brown,
1985).
With respect to equipment provision, Sweden is therefore similar to the more
liberal parts of Europe. The difference to most other countries is that Televerket
is also involved in equipment manufacturing and R&D in two major ways: first,
through its wholly owned manufacturing arm, TELI, which dates back to 1891;
second, through its partnership with L. M. Ericsson in the development firm
ELLEMTEL,
TELI has 3500 employees and produces much of Televerket's domestic
equipment. Ninety percent of TELI production is destined for Televerket. Tel-
everket also buys some equipment from Ericsson, but Ericsson relies primarily
on export markets for its revenues. TELI produces digital AXE switches, PBXs
(including Northern Telecom-licensed products), hand sets, and a variety of
other transmission and data-processing equipment, as well as a limited number
of electromechanical switches. It has also been involved in the production of
low-cost computers used especially for school instructional purposes.
Ericsson and TELI established ELLEMTEL in 1970 after earlier disappoint-
ing experiences for each in the development of electronic SPC switches. They
developed independently the prototypes (ATE 12 in 1968 and A210 in 1970,
respectively). However, both Ericsson's chairman Wallenberg and TELI feared
the large costs of such endeavors and decided to pool expertise and costs. As
part of this decision, the two partners divided the market for ELLEMTEL prod-
ucts, especially its AXE digital switch, with TELI getting its traditional Swed-
ish (i.e., Televerket) market and Ericsson getting the rest of the world. Losses
are shared by Televerket (27 percent) and Ericsson (73 percent) according to
expected sales (OECD, 1987). In 1990, Televerket purchased Ericsson's do-
mestic telecommunications sales interests.
Televerket claims that vertical integration into manufacturing enabled it to
purchase equipment at favorable prices and pass these savings on to its custom-
ers. It also sought to avoid being captive of one national equipment firm. Be-
fore liberalization, however, subscriber terminal hand sets were available in
competitive markets at one quarter of the Televerket price (Brown, 1985). Lib-
Sweden 207
eralization of hand sets, largely ratifying the reality of customers' self-help,

legalized competition and forced prices down. Televerket also argued that, through
its manufacturing arm, it gains useful information about manufacturing costs,
which helps it in its other procurements. An example was its successful effort
to convert Northern Telecom equipment to European standards in the late 1970s.
Only 10-20 percent of Televerket's telecommunications equipment was im-
ported,2 although it has no formal "buy Swedish" procurement policy. After
TELI ceased headset production in the face of Asian competition, their imports
rose to fifty percent.
Televerket's vertical integration is not without critics. In 1980, the Swedish
parliament passed a telecommunications law with broad support to specify its
role. While emphasizing Televerket's position in modern information technol-
ogy, the law encouraged a further market orientation and required that Tele-
verket's competitive activities be undertaken in separate subsidaries. Telein-
vest, the resulting company, was set up with several affiliates, including Tele-
Larm, for security systems; Nerion, a Norwegian electronics manufacturer;
Swedtel, an international consulting company; Swedcom, for telecommunica-
tions systems development abroad; Telefinans, a finance company for access to
the international commercial financial markets; and Telelogicab, a software de-
sign subsidiary. The 1980 law limited Televerket's connection monopoly to
voice communications equipment, medium- and high-speed modems, and telex
equipment, but its provision for Telefinans significantly enhanced Televerket's
competitiveness by allowing it to borrow on the open market. In 1987, TELI
was integrated into the Teleinvest subsidiary Teleindustrier and given corporate
status. Teleindustrier dwarfs all other subsidiaries.
A 1982 law made certification of privately supplied equipment subject to a
liberalized procedure. Other equipment required type approval, for which a
department of Televerket itself was in charge. In other words, potential entrants
had to submit technical drawings and detailed descriptions of their equipment
to their major manufacturing and marketing rival for approval. Since 1984,
suppliers were allowed to have their equipment tested by any independent lab-
oratory, private or public, and even, in some instances, to be self-certified. But
approval was still lodged in the Televerket authority (Gleiss, 1983). A 1985
law scheduled further liberalizations in several steps. A 1988 parliamentary
decision mandated open markets for payphones, PBXs, and modems of any
speed, partly in response to trade friction with the United States, where Com-
merce Secretary Malcolm Baldridge had threatened to bar Ericsson from the
United States if no reciprocity was established. In 1989, the technical approval
functions of Televerket were moved to a separate regulatory entity, the Na-
tional Telecommunications Council (Statens Telenamnd, or STN) (Thorngren,
1990). Control over monopolistic activity or competitive policy are performed
by Sweden's Fair Trading Ombudsman (SPK). By 1990, Televerket's only for-
mal and disappearing regulatory power was related to spectrum management
(Whitehouse, 1989, pp. 3-6), which still afforded it a level of power against
its competitors in the mobile telephony field.
Televerket has a close relationship with its 48,000 employees through its
208 Sweden
unions. The three unions (for blue-collar, white-collar and college-trained staff)
are involved in all major decisions and must be consulted on matters affecting
workers. Most Swedish firms operate similarly. The unions have opposed
equipment liberalization, relaxation of monopoly, and the transformation of
TELI into a corporation.
Ericsson
Sweden's major telecommunications equipment firm, Telefonaktiebolaget

L. M. Ericsson, was founded by Lars Magnus Ericsson in 1876. Its two largest
stockholders are the Wallenberg Group and the financial institution Handels-
banken. The firm successfully offered stock (in the form of deposit receipts) on
the American market.
L. M. Ericsson began to manufacture telephone equipment in 1878, aided
by the absence of patent protection. Its first major customers were the telephone
companies to which Bell-affiliated manufacturers would not sell equipment.
Later, in 1918, when the state Telegraph Administration forced the General
Telephone Company out of network operations, the latter's management joined
Ericsson.
Internationally, Ericsson is a medium-sized but very active corporation. Its
net sales in 1989 were $6.3 billion and grew to $7.4 billion in 1991. Of these,
public telecommunications accounted for 44 percent of sales. Ericsson's suc-
cess in the export business (80 percent of its sales) is linked, paradoxically, to
the absence of a captive home market. Televerket's TELI is its parent's prime
supplier, and Ericsson is only a secondary supplier in Sweden. This is not to
say that Ericsson is a small firm in the Swedish context. With 70,000 employ-
ees worldwide, it is one of the country's ten largest industrial firms. Its success
in both industrialized and developing countries has proven that it is not neces-
sary to be large or backed by a large domestic market to be a technologically
competitive developer of sophisticated telecommunications equipment.
To reduce its labor costs and improve its international competitiveness, Er-
icsson built manufacturing plants in dozens of countries; half of its employees
work abroad. It is organized in seven business areas, including public tele-
phony, radio communications, information systems (including its acquisition,
Datasaab), defense systems and components (Rifa), and a large cable enter-
prise. Research and development for its AXE digital switches and other equip-
ment are conducted by ELLEMTEL, in joint operation with Televerket's TELI.
In 1987, Ericsson sold its office equipment operations to the Norwegian firm
Design Funktion.
Ericsson's technical successes include a photonic switch matrix which rep-
resents an important step toward building an optical switch in which photonic
signals can be routed without having to be converted into electronic signals for
switching purposes. The digital AXE switch was installed or ordered in many
countries and increased the company's share in the world market for digital
switches to 13 percent by 1984. In the United States, the AXE switches were
Sweden 209
first used outside the Bell System by special common carriers and indepen-
dents. Customers included MCI, Western Union, ITT, and US Sprint. The
opening of the Bell market through the AT&T divestiture increased Ericsson's
potential customers dramatically. But in 1979, the U.S. markets comprised
still a minor share of Ericsson's business, representing 5 percent of sales,
mostly in traditional technology. By boosting its AXE system with a new cen-
tral processor and increasing its capacity, Ericsson made it a viable choice for
American local exchanges. Although its critics complain that it is based on a
central rather than on a distributed processor, the AXE system's modular ar-
chitecture, for which it is noted, has advantages over more customized designs.
Ericsson's hopes of penetrating the Bell companies' central office market, how-
ever, were slow to materialize.
In the United Kingdom, Ericsson's AXE switch successfully won acceptance
in the public switching market, the first such entry by a foreign firm in decades.
This was accomplished when British Telecom, in order to be less dependent on
the delay-plagued System-X switch developed by Plessey and GEC, sought a
second supplier, dubbed System Y. It picked Ericsson's venture with Thorn
EMI, greatly disappointing the French, in particular, who argued that the choice
of a non-EC firm was a step in the wrong direction.
Ericsson operates in Italy through its subsidiary Fabbrica Apparechiature
Telefoniche Materiale Elettrico (FATME) and several smaller operations which
together comprise 19 percent of the Italian market, employ about 5000 persons,
and have received export orders for AXE exchanges from Cyprus, Swaziland,
Ethiopia, Guatemala, and Mozambique, among others.
In Latin America, Ericsson holds a strong position, particularly in Mexico,
Brazil, Venezuela, and Colombia. Ericsson is also strong in Australia, the
Netherlands, Denmark, and Spain, where its subsidiary Spain Intelsa has al-
most 50 percent of the digital public exchange market. In 1990, more than
2000 digital exchanges, including cellular exchanges, were in service in some
seventy-five countries, with 27 million lines installed or on order. In 1987, it
received a $46 million contract from the People's Republic of China for AXE-
10 digital switches, and thirty-six were installed by 1990. Similarly successful
was Ericsson's mobile cellular telephone technology, which was selected by
numerous countries.
By 1992, Ericsson found itself under financial pressures and a potential merger
partner.
Services
Formally, Televerket does not have a legal monopoly over network services.
De facto, however, it was the only supplier of public or private network ser-
vices, with the minor exception of networks of a few governmental departments
too powerful to overcome.
To make competition fair, the government resolved in 1986 to permit a move
210 Sweden
toward cost-based pricing by Televerket (Ministry of Transport and Commu-

nications, 1987). In 1987, the government took legislative action to rebalance
prices by reducing rates for long-distance calls and gradually increasing charges
for local calls (OECD, 1989).
The first significant entry in services occurred with the licensing of the com-
pany Comvik as a rival cellular telephone service provider. Liberalization of
service was introduced in 1989, when third-party traffic and international ser-
vice were opened to new producers (B. Thorngren, 1990, communication).
Comvik's Skyport service opened leased circuits for competition and resale.
Subsequently, Sweden's national railway also aimed to resell capacity on its
fiber-optic network.
Behind Comvik is Jan Stenbeck, a major Swedish industrialist engaged in
steel, coal, and other businesses. Stenbeck's strategy is to invest in telecom-
munications services with an eye toward their deregulation. He owns Kinnevik,
a major media company, and its subsidiary Comvik. Comvik was the first to
compete with Televerket with mobile telephone service and international satel-
lite connections. Stenbeck's TV3 channel, Sweden's first commercial cable
channel, was transmitted from London via the Astra satellite, in which he was
part owner. He also owned the pay service TV 1000 and cable television
systems (Finvik), and was a partner in an American DBS venture and in Mil-
licom, the holder of interests in cellular licenses in the developing world. Through
his diversified communications holdings, and subsequent partnership with Brit-
ain's far-flung Cable & Wireless, Stenbeck positioned himself to be the Nordic
equivalent of Berlusconi or Murdoch, but with a much stronger telecommuni-
cations presence. Forty percent of Comvik's satellite operations were held by
Cable & Wireless.
Cellular telephony is a special Swedish success story, primarily through the
joint Nordic cellular telephone system (NMT). In 1981, the initial year of its
operation, Televerket predicted it would have 50,000 subscribers by 1991. But
in 1987, there were already 120,000 subscribers, and by 1990 there were 450,000,
or 9 percent of all automobiles (Boan, 1990, p. 3). Other mobile services in-
clude Mobitex, which permits mobile transmission of data as well as voice
developed especially for intraorganizational use. National and local paging (Mbs
and Minicall) are also available.
Comvik holds about 7 percent of the cellular market and has the right to
interconnect with its rival's public network. In 1990 Comvik submitted accu-
sations of unfair practices by Televerket to the Ministry of Communications
and to the Fair Trading Ombudsman. One inquiry concerned the charge that
Televerket influenced Ericsson to refuse Comvik's order for mobile equipment
in an attempt to squeeze Comvik out of the market for mobile telephony. Also,
Comvik applied for spectrum capacity to expand its proposed GSM network
across all of Sweden by 1995. Televerket allegedly limited Comvik's original
frequency allocation. Televerket also responded with a 300 percent increase in
the fee on calls made between its system and Comvik's. On the other hand,
Televerket's view was that Comvik's difficulties were partly due to its decision
to support only its own equipment standard. Another potential mobile compel-
Sweden 211
itor, Nordictel, was formed by Scandinavian Airlines (SAS), Volvo, Custodia,

Pharos, and Racal. It was licensed in 1991.
Several public data networks exist. Datel service has long been offered for
moderate transmission rates. Datex-L is a circuit-switched network linked with
similar Scandinavian networks. Tariff rates per time unit were proportional to
the transmission speed, making the cost per bit relatively constant but reducing
the incentive to use data lines more intensively. The third public data network
is Datapak (formerly Telepak), a packet-switched network that has operated
since 1984. Rates were structured to make public data networks more advan-
tageous than private lines, particularly for small users.3 In 1990, Televerket
served 1 million data terminals (Boan, 1990).
The first competing public network started service in 1991. Tele2 AB was
owned 40 percent by Cable & Wireless and 60 percent by Kinnevik. It leased
fiber lines from the Swedish railroad. Thus, competition in transmission had
come to Sweden.
A major digitalization program for the public network was initiated in 1985,
following basic decisions of the 1970s. Digital upgrading proceeded rapidly.
In 1990 almost 50 percent of exchange lines were digitally switched. However,
Televerket was not giving ISDN the same priority as many other European
telephone administrations did.
Televerket's broadcast activities include more than 1200 radio and TV trans-
mitters. Televerket was also in charge of national frequency management and
runs coastal, maritime, and aviation communications. It also provides cable
television and SMATV services, with over 50 percent market share. In 1990,
the market for VSATs and TVRO dishes was liberalized, subject to frequency
coordination.
Videotex began in Sweden in 1982 under the name Datavision. The involve-
ment of Televerket is substantially smaller than that in many other European
countries. It limits itself to providing network services and does not manage
information provision. This is done by a private company, Telebild (in which
Televerket held 30 percent until it sold it off) and by the postal administration's
Postal. Televerket also introduced "071" service in 1989, a type of dial-it
service offering news, weather, jokes, sports scores, and other information.
Sweden's strong social welfare tradition has led it to the forefront of the
innovative Nordic "telecottages" program, which it began in 1985. Telecot-
tages bring telecommuting, distance learning, telefax, videotex, and some video
services to remote rural locations. In 1989, Sweden had seven existing and
twenty-one planned telecottages (Qvorturp, 1989).
In 1990, Televerket proposed to have its status changed from a Crown Cor-
poration and to sell a 45 percent stake to private investors. Regulatory powers
would be lodged in the National Telecommunications Council. The plan met
some opposition from socialists, greens, and communists; its passage would
give the company much greater autonomy, but would also lead to increased
competitiveness in the market. Evidently, Televerket was confident that it would
hold its own, though it also began to forge alliances with other medium-sized
PTOs, such as the Dutch PTT.
16
Finland
The Finnish telephone system is one of the most interesting in Europe. Instead
of a national monopoly, there are about fifty companies, either subscriber co-
operatives, municipal enterprises, or private firms, that provide local telephone
service and are directly accountable to subscribers. There is also a national PTT
(the P&T), which is the largest operator of local telephone service (covering
about one-third of all subscribers and three-quarters of the land area). The P&T
also offers domestic and international long-distance service, mobile telephony,
and other services.
Thus, the Finnish telecommunications system has forgone the economies of
scale and the unity of control that are sanctified elsewhere. Yet its telephone
system is among the most advanced in the world, moved along by a healthy
rivalry among the various participants, and especially by the responsiveness of
local companies to their subscribers/shareholders. The number of equipment
options is high. Despite the country's small domestic market, Finland's elec-
tronics and telecommunications industries are quite successful both in domestic
and export markets.
History
The Finnish state postal system dates back to 1638, when Queen Christina of
Sweden issued an edict that led to the establishment of five post offices. This
system first operated as part of the Swedish and later as part of the Russian
post. In 1812, the system became semiautonomous, and it received a monopoly
status in 1874. Savings bank services were added in 1886, which led to a rapid
increase in the number of offices. After national independence, a postal admin-
istration was set up in 1918 and was combined in 1927 with the Telegraph
administration into the P&T (Finland PTT, 1983).
The first telephone sets arrived in Finland as early as 1877. Within five
years, a number of local telephone companies began operation, the first started
by a Helsinki telegraph mechanic, Daniel Johannes Waden. Since Finland was
still a Russian province, it was subject to the Russian telegraph bureaucracy.
By forming local cooperatives, the Finns became less dependent on the Russian
authorities. After 1917, the new Finnish government inherited the imperial Russian
212
Finland 213
role in telegraphy and some long distance service, but independent local tele-
phony was already in place in the cities.
In 1894, the Interurban Telephone Company of Southern Finland was started.
The P&T, which gradually established telephone service in the northern part of
the country, acquired this company in 1935. Since then, the P&T has operated
long-distance service as a monopoly. Service with Russia began in 1909, ex-
tending to Sweden by 1919. Automatic telephone exchange was first introduced
in Helsinki in 1922 (AT&T, 1975).
The P&T acquired a private long-distance company, and expanded it until it
included all long-distance and international service, plus the new offering of
telex. It also provided local service in those areas where local organizations did
not exist (Bruce et al., 1986).
The basic laws of Finnish telecommunications were set by the Czarist Tele-
phone Manifest of 1886, the Telegraphy Law of 1919, and the Radio Law of
1927. Efforts were made to update the laws, but for a long time there was no
political consensus to accomplish more than minor changes (OECD, 1982, p.
10). Eventually, a new telecommunications law was passed and became effec-
tive in 1987. It provided a legislative framework to promote development of
the telecommunications industry. The traditional operational division of respon-
sibilities between the telecommunications bodies remained the same, but the
Ministry of Transport and Communications assumed the supervisory functions
previously under the jurisdiction of the P&T.
The P&T and the Local Companies
Telephone service in Finland is unique in that it is provided by both the gov-

ernment through its P&T and by fifty-one local telephone companies that op-
erate with a state concession. These local companies serve more than 72 per-
cent of the telephones in Finland, with the P&T serving the others. By population,
the local telephone companies' service covers 68 percent of households, and
the P&T service covers 32 percent. In terms of area, however, the P&T serves
75 percent (i.e., the rural parts of the country). The Ministry of Communica-
tions has overall responsibility for telecommunications and also sets the P&T
budget.
The local companies provide data transmission, but the P&T provides tele-
graph and telex services. Transmission of TV and radio broadcasts is under-
taken by the Finnish Broadcasting Company YLE and by private local radio
broadcasters rather than by the P&T. A data network (circuit-switched Datex)
is operated jointly by the P&T and the local telephone companies. Nationwide
cellular telephone (the Nordic Mobile Telephone Service) is also provided by
the P&T.
Organizationally, the P&T is headed by a director general and a deputy di-
rector general for telecommunications, and is divided into departments of gen-
eral telecommunications and radio. The radio department provides mobile radio
and paging services.
214 Finland
The local companies are either private firms or cooperatives owned by their
subscribers or municipal companies. In the 1930s, over 800 private telephone
companies operated (Williamson, 1986, p. 52). The smallest company supplies
approximately 1900 connections; the largest, the Helsinki Telephone Company,
supplies almost 650,000. By their charters, the cooperatives' pricing is sup-
posed to be cost-based and not for profit. Subscribers elect management and
directors, either at a general meeting or by mail. This process is taken seri-
ously. At the board election of the Helsinki Telephone Company, 50 percent
of the 350,000 owners participated in the mail balloting. Subscribers, when
joining the system, provide a payment and receive a negotiable share certifi-
cate. Nearly three-quarters of subscribers are shareholders. No single share-
holder can have more than twenty votes, however, and each shareholder must
pay an annual telephone subscription, whether a telephone is used or not, which
discourages the holding of multiple shares and of speculation. This system af-
fords about 1 million telephone subscribers a direct role in telecommunications
matters (Myllo, 1984). A wide variety of organizational structures exists among
the companies, including in several cases outright municipal ownership. Merg-
ers have reduced the numbers of companies to fifty-one. In matters of service
quality and tariffs, the independent telephone companies are bound by general
conditions of their concessions and are otherwise free to set their own policies.
No direct popular involvement exists for the P&T, but it has public advisory
committees and is subject to oversight by the government.
The state-owned P&T is Finland's largest employer, with 45,000 employees.
Of those, more than 20,000 worked in telecommunications in 1988. It is a
profitable organization: in 1988, its surplus was 13 percent of revenues (which
was contributed to the state budget) and its return was 20 percent of fixed assets
(ITU, 1990). Before 1989, it was headed by Pekka Tarjanne, who became the
secretary general of the ITU in Geneva.
In 1988, P&T's profits were about FIM 800 million ($200 million). The
1989 telecommunications act requires the P&T to contribute a fixed amount of
profits, about $100 million, to the general budget each year. The P&T contri-
bution is more than that of all other state industries combined. Sixty percent of
P&T revenues come from domestic service, with the rest split among mobile
equipment leasing and international and telex service (Whitehouse, 1989b).
The relation between P&T and the local companies is characterized by both
cooperation and conflict. The local operators are collectively represented by the
Association of Telephone companies of Finland. Known also as toimilupalai-
tokset, or "licensed ones," they are a strong voice against the P&T. The P&T
has long claimed that to ensure the goal of social equity, a full national mo-
nopoly is required; but the local companies have stressed, just as adamantly,
their century-old tradition of technical progressivity and political legitimacy
through the participation of owners/subscribers. Friction arises over the extent
of the P&T monopoly, control of advanced services, and revenue and cost
allocations.
Finland 215
Legislation and Change

The Finnish system bears a certain resemblance to that of the United States
after the AT&T divestiture, in that it is composed of independent local com-
panies interconnected by a powerful long-distance carrier. However, the Fin-
nish P&T, at least before a 1987 law, both acted as regulatory and franchising
authority over the local companies and approved equipment that the local com-
panies supplied in partial competition with it. It also held a monopoly over
long-distance and international services. The P&T's dual role of service pro-
vider and regulator, as well as its claim of monopoly power over the ambiguous
data services, was a source of criticism and growing discord between the P&T
and local companies.
The 1987 telecommunications law was an effort to define jurisdictions and
thus end the disputes. Regulatory powers moved to the Ministry of Commu-
nications. Although the new law allowed the local telephone companies to enter
parts of the market for data services, various overlapping service issues re-
mained unresolved. An example for such a dilemma is the long-distance data
communications services offered by OY Datatie, owned by a number of local
telephone companies. The P&T maintained that Datatie should first have a li-
cense, granted by the P&T, to enter this market. Similarly, overlap problems
exist for electronic mail and videotex which are now regarded as competitive
services.
The independent companies must also pay for carriage on the P&T's long-
distance network; these payments help keep the P&T's local telephone rates
low. Sixty percent of the independent companies' income goes to the P&T for
long-distance charges, whereas 80 percent of all calls are local. Not surpris-
ingly, the independent companies are interested in long-distance competition
that would provide them with alternatives. (In the United States, long-distance
carriers support, through access payments, the local exchange companies, not
vice versa.)
In 1980, the local telephone companies terminated an agreement on data
transmission. Two court decisions in 1982 restricted the P&T's ability to limit
the activities of private telephone companies.
The local telephone companies feared that the P&T would limit them to basic
telecommunications services and even compete with them, while reserving for
itself the potentially lucrative advanced commercial services. The P&T, could,
for example, operate in the areas of the local companies by using their lines to
link the PBXs of large users directly to the distant P&T exchange, in effect
"bypassing" the local companies.
To deal with these and other problems an early governmental proposal was
presented in 1982, but it found no approval. The fight was on between the
P&T and the companies and led to the appointment in 1983 of the so-called
Rekola Commission. This panel proposed independent equipment approval,
simplified attachment by users (a bow to reality), liberalized resale of services,
and introduced the possibility of "special" private networks without common
216 Finland
carrier obligations. The commission, seeking to fashion a characteristically Fin-

nish consensus, proposed more control for the P&T over its own activities and
financing, and greater institutional distance from the government. But it also
recommended ending the situation in which the P&T was both a licensing au-
thority and a competitive equipment provider.
However, the P&T opposed the proposed regulatory entity within the Min-
istry of Communications. The 1987 law followed most recommendations but
provided for type approval authority to be lodged in the ministry. The ministry
can, however, delegate this authority to a control body that can employ outside
laboratories, including those of the P&T and the local companies. In general,
the new law's regulatory structure distinguished between services provided as
"official duties" and "commercial activities." Private companies can be li-
censed by the government for general telecommunications. For separate net-
works, they must seek a license from the Ministry of Communications, assisted
by an expert advisory committee. Carriers being granted a general telecommu-
nications license are subject to a variety of duties and rights. Among the former
is a universal service requirement, and the obligation to lease connections, in-
cluding those for further uses such as value-added services.
The 1987 law gave the Ministry of Communications authority to set P&T
rates and to decide the basic guidelines for the tariffs of the private companies
or to issue unfavorable equipment approval. The local companies tried to pre-
vent P&T exclusivity for long-distance and international service to be written
into law. In 1991 they applied again for permission to offer such services.
In addition to defining the role of the P&T, the 1987 legislation set require-
ments on its efficiency and productivity. Although the P&T remained a state-
owned enterprise, there were plans to enact further legislation separating its
corporate and government budgets.
In 1989, the Finnish parliament passed two telecommunications laws (Law
748 and Decree 928) to make the P&T a public company on January 1, 1990.
However, the sale of P&T stock was delayed well past that date, as was the
split of postal and telecommunications activities.
Services
Telephone density and quality of service are high in Finland. In 1988, density
was 49.86 subscribers and 68.5 sets per 100 people (ITU, 1990). This is par-
ticularly remarkable in light of the low population density. Although Finland is
Western Europe's fourth largest country, its total population is less than 5 mil-
lion. Also, despite its rural character, Finland maintains rates well below the
OECD average for residential and business lines (OECD, 1990).
Local exchange automation dates back to 1922 and was completed in the
1960s. Long-distance automation began in 1958, with Ericsson ARM equip-
ment. Electronic Analog SPC switches were introduced in the 1970s, and dig-
ital trunk exchanges were introduced in 1978. The P&T introduced digital
switching in the trunk network as well as in the local exchange in 1983.
Finland 217
The decentralized Finnish system permits substantial flexibility. Because of

Finland's multiplicity of standards and operators, it is an ideal testing ground
for new products and services. The Helsinki Telephone Company set up, among
others, a packet-switched network and a special data network, a subsidiary to
sell subscriber equipment throughout Finland. It also began construction of a
cable network. According to its managing director, "We didn't ask anybody,
we just did it" (Williamson, 1986). The "Dataway" network was introduced
in 1985 by the Helsinki Telephone Company and Finland's biggest bank, K.O.P.,
to operate an intercity data network. It aimed to bypass the P&T's trunk lines
with fiber-optic transmission.
In 1985, thirty-eight local phone companies formed a long-distance operator,
Datatie, to compete with the P&T. Datatie carries voice and data traffic and
leased line services but no switching. Its shareholders and customers are major
banks and insurance companies. Its leased line prices tend to be 25 percent
below the P&T's tariffs. In response, P&T invested in Business Networks'
(Yritysverkot) bypass network for high-speed leased lines.
In 1990 local companies began fiber-distributed data interface (FDDI) service
through Datatie. Helsinki Telephone offered its own service over a 100-mbps
backbone, and P&T planned two FDDI rings for Helsinki, one for government
and one for commercial uses.
Since 1981, a circuit-switched data network (Datex) has been in operation,
in cooperation with other Scandinavian countries. The P&T and the local com-
panies also provide a packet-switched network, which became fully operative
in 1984. This packet-switching system was developed separately by the local
companies and the P&T.
In 1992, the rivalry between P&T and the local companies intensified, when
they sought permission to enter each other's core markets of long-distance and
local services.
ISDN trials began in 1987. An intermediate digital service, Diginet, has been
available since 1986, offering one 64-kbps channel. Full service began in Hel-
sinki in 1990 under the cooperation of P&T, Helsinki Telephone, and the man-
ufacturer Nokia, and also used Ericsson and Siemens ISDN exchanges (T. Salo,
1990, communication). In 1989, the independent telephone organizations formed
the joint company Teleryhama, which offers data transmission services.
In 1978, Finnish local telephone companies were the first in Europe to pro-
vide public facsimile. In the same year, Finland became the second country in
the world to offer videotex. Fiber links were introduced into the standard tele-
phone network after 1979. Electronic mail and teletext services were introduced
after 1983.
P&T's cellular service operates at 450 MHz and 900 MHz and is part of the
Nordic system NMT. Demand for mobile telephony has far exceeded predic-
tions, with some 100,000 Finns using the NMT 450 system and 25,000 using
NMT 900 in 1989. A digital GSM service and telepoint CT-2 are being intro-
duced. There will be two separate networks for GSM: Radio Linja's and the
P&T's.
An interesting indigenous technical development is the Wilderness and Island
218 Finland
Telephone Network, developed by the P&T and introduced in 1977 as an emer-

gency service (Karvonen, 1984). These are unmanned telephone booths, lo-
cated in distant locations, and operating by radio with a special battery that
lasts two to four years. The conditions of the telephone and the battery can be
monitored from a central facility. The telephones are connected to a base sta-
tion powered by generators, windmills, or solar cells.
Equipment
For a country with a small population, Finland has a well-developed electronics
industry. In principle, the equipment market is open to foreign firms, but most
equipment is supplied by domestic companies, such as Nokia or the Finnish
subsidiaries of Ericsson, Siemens, and Alcatel. Telephone companies attempt
to buy from several sources to maintain competition among suppliers. There
are four digital exchange suppliers: Telenokia, with 40 percent of the market;
Ericsson, with 34 percent; Siemens, with 25 percent; and Standard Electric
Puhelinteollisuus (formerly ITT, subsequently Alcatel), with 1 percent.
Telex terminals are provided solely by the P&T; cellular telephones and pag-
ing equipment generally are privately supplied, rather than provided by tele-
phone companies. This is due to the liberalization of interconnection rules. In
the past, the telephone companies and the P&T provided most equipment,
including telephone sets, modems, and data terminals. Interconnection of
subscriber-owned equipment has subsequently become permissible.
Other internationally successful Finnish Telephone equipment manufacturers
are VISTACOM (videophones), BENEFON (mobile telephone) and TECNO-
MEN (paging, automatic voice-messaging, encryption). The largest Finnish
equipment manufacturer is Telenokia, part of the electronics division of Nokia,
Finland's largest private company. In 1989, it controlled 19 percent of the
Finnish telecommunications market (Northern Business Information, 1990). The
Nokia group is also involved in forest products, machinery, plastics, and chem-
icals. A sister organization, Nokia Data Systems, produces PBXs, LANs, and
other equipment and terminals. Nokia Cables is an important European cable
manufacturer. Still another sister company, Mobira, is Scandinavia's largest
producer of mobile telephones.'
In 1989, Nokia had approximately 37,000 employees and its turnover was
about $5.6 billion (Nokia Annual Report, 1989, p. 3). Although in the early
1970s Nokia was not very active in the electronics field, by 1990, electronics
and electronic products comprised 90 percent of the group's business, mainly
in the consumer area. Telecommunications represented 9 percent of sales, or
$570 million.
Nokia is active in exports. Only 32 percent of its sales in 1989 were in
Finland, with other Scandinavian countries accounting for another 16 percent
(Nokia, 1989). It holds 35 percent of the digital transmission market in Swe-
den, 95 percent in Qatar, and supplies the U.K., France, China, and the United
Arab Emirates with its DX-200. In 1990, Nokia had 37 percent of the mobile
Finland 219
phone market in Thailand, 35 percent in Hong Kong, and 10 percent in the

United Kingdom. In 1990, it formed a joint venture with Moscow Telephone
Network to provide the Soviet Union's first mobile telephone network. Its suc-
cess contradicts the conventional wisdom that it takes a large company and a
protected home market to survive internationally, because of economies of scale.
Nokia has demonstrated that there are niche markets, and it has found one in
small rural exchanges with less than 1000 lines.
17
Norway
History
The Norwegian state telegraph company was established in 1855, four years
after the first telegraph line was installed. The telephone was introduced in
1880 by the International Bell Telephone Company. Within a year, Bell estab-
lished local franchises in Oslo and Dramrnen. A second private system was
launched in the capital to compete with the Bell System, but the two were
combined in 1886 into the Christiana Telephone Company when the vigorous
competition between them yielded what municipal officials deemed were
counterproductive results. The city government thus committed itself to a role
in the development of Oslo's communications system and retained a share in
the new company (Holcombe, 1911).
Norway soon faced the question of who would provide the long-distance link
between the Oslo and Drammen exchanges and what role the government should
play in telephony. In 1881, the parliament established state control over tele-
phony and telegraphy and provided for private telephone concessions. A private
project for the Oslo-Drammen line collapsed when the state telegraph authori-
ties demanded full compensation for all revenue losses incurred as a result of
intercity telephone service. The project was also impeded by the merchants of
Drammen, who feared that a readily available means of communicating with
the capital would hurt their businesses. Similar to other European states' short-
sighted protection of the telegraph, the state telegraph office decided not to
provide long-distance service on its own to guard the health of the telegraph.
In an attempt to block all intercity telephone service, the telegraph administra-
tion forced local exchange systems of adjoining communities to maintain a 2-
kilometer buffer area between their respective service areas. However, as de-
mand in rural areas increased and companies responded with expanded service,
interconnection was permitted, although still only where no equivalent tele-
graph connection existed. This restriction was abolished as it became increas-
ingly clear that the telegraph could not substitute for the telephone. In 1885,
the first interurban telephone line parallel to a telegraph route was constructed
and some telegraph offices were converted to public telephone offices. But
where telephone companies interfered with its revenue sources, the telegraph
administration still required compensation for the resulting falloff in its busi-
ness.
220
Norway 221
Because of the dispersed nature of settlements in Norway, the telephone was

well-received. Within a decade of its introduction most villages in the country
had their own systems. This extraordinarily rapid expansion of telephone ser-
vice was neither supervised nor operated by any central state organization. Most
of the local telephone systems were built by small joint stock associations,
cooperatives, or public ventures. The majority were small efforts with neigh-
bors helping each other and contributing rights~of-way, materials, and labor.
The systems initially used different equipment and standards. This lack of tech-
nical coordination impeded later attempts to link systems through trunk lines.
By 1908, however, local systems began to cooperate to create a homogeneous
national network.
In 1892, a royal commission was appointed that recommended that long-
distance service be privately provided too. Several of the larger local compa-
nies attempted to become involved with the plan. But the telegraph administra-
tion began to buy local exchanges when their concessions expired and entered
the long-distance service field. By 1906, provision of both local and long-
distance service was about half private and half public, with the government
serving primarily the larger cities and the private companies serving the rural
areas. This is the reverse of situations ordinarily prevailing in countries with
mixed public-private systems. Eventually, the state system absorbed all private
and cooperative networks. In 1899 the agency that would later become the
NTA was established and its authority remained largely unchallenged until the
1980s (Nyheim, 1987).
The Norwegian Telecommunications Authority (NTA)
The Norwegian Telecommunications Administration (NTA or Televerket) is the

country's de facto telecommunications monopolist. It operates under the 1899
Monopoly Act, which delegated to the state exclusive control over telecom-
munications in all forms except broadcasting but which does allow private par-
ties to provide telecommunications services under government license. The NTA
is supervised by both the minister of communications and an Advisory Council
on Telecommunications. In 1988, it was the country's largest non-defense em-
ployer, with a staff of 6400.
The NTA's development proceeded in accordance with a planning document
whose focus led to significant increases in telephone penetration and productiv-
ity in the 1980s (Foreman-Peck and Manning, 1988). In 1980, the government
appointed a commission to recommend long-range telecommunications policy.
Its report, issued in 1983, was supportive of a strong state role. By then, how-
ever, the government was led by a right-of-center coalition which did not par-
ticularly care for the recommendations and instead established the Stette Com-
mission to re-examine the telecommunications organization in light of deregulatory
trends.
That panel recommended dividing Televerket into three separate organiza-
222 Norway
tions: one for basic monopoly services, another for competitive activities in
equipment markets, and a third for technical control functions (Nyheim, 1984).
In 1986, the parliament adopted a resolution to reorganize the NTA begin-
ning in 1988. The parliamentary plan allowed the NTA to keep its basic net-
work monopoly but required that its unregulated businesses be consolidated
under a separate subsidiary, Televerkets Konkurranseorganisasjan (TBK).
Structural safeguards have been implemented against cross-subsidization of TBK
by NTA's monopoly. After the defeat of the social democratic government, the
new coalition government began to consider further deregulation. The value-
added network services were opened for competition. When the market was
opened, there were no competitors. With some reluctance, NTA participation
in this market was permitted. Applications for a second cellular service were
received from Netcom, owned by Norway's Nova Industrier and Orkla Borre-
gaard and Sweden's Comvik. Aside from telecommunications services such as
VANs, TBK markets, sells, installs, and services user equipment and offers
cable TV systems.
The reorganization also included the establishment of the Norwegian Tele-
communications Regulatory Authority (NTRA), overseen by the Ministry of
Communications, which represents Norwegian interests before bodies deter-
mining international standards and handles licensing, spectrum allocation, and
equipment type-approval.
The privatization of 49 percent of NTA and divestiture of TBK promised by
the conservative and Progress parties was averted by a narrow Labor/Left vic-
tory in 1989. But competition was assured for mobile telephony.
Services
Penetration of main lines reached 47.7 per 100 persons in 1988 (ITU, 1990).
These investments are financed by rates, identical for business and residential
customers, which are among the highest in Europe, despite the 1981 elimina-
tion of connection charges. (Foreman-Peck and Manning, 1988). In a contro-
versial move, NTA in 1989 decreased the price of international leased lines
while increasing domestic leased line prices by 16 percent.
Telex service began operating in 1946. Data transmission, using circuit-
switched Datex-L service, is available. A public packet-switched network, Nor-
pak, operating since 1980, offers international connections. In 1990, Norway
ordered X.400 and X.500 systems from the British firm ICL. The 1983 gov-
ernment commission recommended the development of ISDN, and in 1986 the
NTA started an ISDN research laboratory and a pilot ISDN project. Videotex
trials began in 1981 (Nyheim, 1984).
A mobile telephony network began operating in 1966 and was automated
and integrated into the Nordic Mobile System in 1981. In 1989, Norway had
the highest rate of mobile communications penetration in the world, thirty-two
subscribers per 1000 persons. ROGALAND radio, established in 1927 and ex-
panded in 1960, also provides maritime services. A GSM network in Oslo was
Norway 223
planned for 1991, with a consortium of Norwegian equipment suppliers, in-

cluding EB Telecom, Siomonsen, Elektro, and ELAB (Green, 1989).
Equipment
One peculiarity of Norwegian network and equipment has been that since the
introduction of automatic dialing in the early 1920s, equipment in the Oslo
area has operated with a reversed dial (i.e., the digit 1 gives nine pulses, etc).
In contrast, the rest of the country operates on the regular system. Interconnec-
tions required dial connectors in the public network and in the leased private
networks, and equipment had to be available for both systems. This problem is
gradually being solved by the introduction of electronic-tone equipment.
Most telecommunications equipmentincluding data transmission terminals,
paging receivers, and private handsetsis privately available with NTRA type
approval. Telex terminals and PBXs were once supplied only through the tele-
communications authority, though that did not appear to be lasting. Domestic
public switch manufacturers were Standard Telefon of Kabelfabrik (STK) (for-
merly the Norwegian subsidiary of ITT and subsequently of Alcatel) and the
Elektrisk Bureau (EB) Group. EB supplied electromechanical switches. But in
1983, the NTA began to digitalize and adopted STK's System 12 switch. In
1989, STK had monopoly control over Norway's public switching market. By
1991, 40 percent of Norway was digitalized, with STK providing 1.2 million
lines.
All this affected EB negatively. Previously owned by L. M. Ericsson, EB
was one of Norway's strongest private industry groups. In 1987, EB and Erics-
son established EB Ericsson Information Systems to manufacture terminal
equipment in which each company has a 50 percent interest. In the same year,
Asea of Sweden acquired a majority interest in EB, which it soon merged with
its two Norwegian subsidiaries. Asea, in turn, joined the Swiss company Brown
Boveri to form the Asea Brown Boveri group. EB became Norway's second
largest privately owned industry group and employed 15,000 people, of whom
2800 were in its telecommunications division. However, in 1989, EB traded
its public switch, PBX, mobile telephone, and other telecommunications oper-
ations to L.M. Ericsson in return for the latter's road and railway signalling
interests, and concentrated on satellite and power line communications. Erics-
son thus acquired a direct role in Norwegian markets.
There are several manufacturers of PBXs. Northern Telecom acquired an
interest in one of the larger ones, G. A. Ring. Other domestic telecommuni-
cations manufacturers include Universal Communications, for modems; Tand-
berg Telecom, for videocodecs; Stentor, for advanced intercoms; SysScan, for
optical scanning devices; Simonsen Electro, for mobile telephones; and Scan-
vest Ring Communications, for telemetry.
18
Denmark
Denmark's telephone system is unusual in Europe insofar as it is a shared

arrangement of several state-dominated organizations. An 1897 law established
a government monopoly for Danish telephony, but concessions granted since
then have transferred actual operations to several telephone companies. This
structure was until 1990 remarkably similar to that of the United States after
its AT&T divestiture: several regional companies offer local service, and a
national carrier (the PTT, subsequently Telecom Denmark) provides national
and international long-distance service. A separate body serves regulatory func-
tions. This system contradicted traditional FIT tenets of unified structure, end-
to-end technical responsibility, and economies of scale.
But for all of its unusual structure, the Danish telephone system cannot be
described as either private or competitive. The local companies were dominated
by the government and did not compete head-on. But they were rivals insofar
as the public was offered an opportunity, almost nonexistent in Europe, to
compare local service and to pressure for improvement in lagging performance.
Furthermore, a healthy rivalry used to exist between the regional companies
and the national PTT, particularly in the areas of new data and mobile tele-
phone services. In 1986, Denmark even had its own mini-divestiture when the
PTT was required to reorganize its small local telephone operations. The 'tele-
phone organizations, in turn, have to face an influential and aggressive constit-
uencythe housing associations with their television community antenna as-
sociationsespecially on issues involving broadband services. However, in 1990
the rivalries were reduced by the government's putting the regional companies
under the PTT holding company, with the aim of reducing internal competition
in order to prepare for the expected European competition.
Structure
The government holds the legal telegraph and telephone monopoly concession
in accordance with a 1897 law; it franchised three regional telephone compa-
niesfor Copenhagen (KTAS), Jutland (JTAS), and Funen (FKT)to provide
service. A fourth concessionary company (Tele Sonderjylland), under the
224
Denmark 225
Danish Post and Telegraph office (PTT), was added to provide service to the
South Jutland region after it was returned to Danish rule in 1920. The individ-
ual companies operated trunk lines within the areas of their individual fran-
chises, and the government played no role in local operations for a long time.
Thus, telecommunications services were supplied by two sources, the PTT
and four local telephone companies. These local companies were not private.
The government held a majority control of the Copenhagen and Jutland tele-
phone companies as well as 100 percent of the South-Jutland telephone com-
pany. The Funen Company was a cooperative of local councils, and the gov-
ernment owned the remaining 45 percent. KTAS operates on the islands of
Zealand, Lolland-Falster, and Bornholm, and serves 45 percent of the popula-
tion and 49 percent of all telephones. JTAS has 39 percent and FKT has 8
percent. In 1990, all regional companies were placed under a PTT holding
organization, with partial privatization involving investors close to the govern-
ment (such as pension funds).
The need for long-distance interconnection and approval of rates suggested a
government presence. In 1950 an agreement was reached between the state and
regional companies (the concordate) in which the state acted as arbitrator for
issues of standards, pricing and cost-sharing, and traffic planning (Olsen, 1988).
But this agreement proved unstable in the 1980s when new technologies were
being introduced. In 1982, the minister of public works, who holds overall
responsibility for telecommunications policy, established and appointed a Tele-
communications Council comprised of thirteen members, representing the min-
istries, the PTT, the independent telephone companies, subscribers, industry,
and employees.
In 1986, the Telecommunications Council was reduced in significance when
the general directorate for Telecommunications was carved out of the PTT.
This reorganization came in the wake of the so-called Bernstein Committee
report. The regional structure was strengthened by adding a fourth concession-
ary state enterprise for South Jutland, which was divested from the PTT, though
it was still owned by it. The PTT, which in 1978 had still called for a unified
system under its control, also lost its operating divisions for data transmission,
mobile telephone terminals, telex, telefax, and other customer services to the
four concessionary companies. In addition, the PTT was divided into a general
directorate for political, administrative, and regulatory functions (including rate
setting, spectrum management, equipment certification, and licensing) and two
operating divisions for postal service and telecommunications. The latter was
named Telecom Denmark, and was in charge of connections between conces-
sionary areas, international long-distance service, the national trunk network,
nationwide cellular mobile telephones, and broadcasting. The Telecommunica-
tions Council was reduced to an advisory capacity, and regulatory matters were
more independently lodged in the ministry and the general directorate (Peder-
sen, 1987).
The creation of the general directorate was significant in unification and con-
solidation of regulating authority. The weak arbitrator's approach of the former
226 Denmark
agency was replaced by a politically strong "caretaker of the public interest".

The directorate outlined a timetable of liberalization consistent with EC reforms
(Olsen, 1988).
Yet this liberalization was not without opposition. In 1990, the government,
moving against the European trend, decided to purchase the Copenhagen and
Jutland telephone companies by 1992, arguing that replication of costly ser-
vices such as ISDN and intelligent networks was inefficient in a small country
and that it reduced service quality. Like Italy, it sought to protect monopoly
by consolidating it instead of using institutional diversity as an invigorating
factor.
In July 1990, the conservative government reached an agreement with the
Social Democrats to place the four telephone companies under a new holding
firm, teleDenmark, which has 17,000 employees and revenues of $3 billion.
The government kept a 51 percent stake and retained control of the tele-
Denmark Board, selling the rest to private investors. Much of the private in-
vestment, however, is from pension funds with ties to the government. The
General Assembly appoints eight out of twelve members of teleDenmark's Board,
including the chairman. The employees of teleDenmark, who retain their civil
servant status, appoint the other four Board members. Telex, satellite and basic
telephone services were reserved to teleDenmark. A second cellular operator
(Dansk Mobil telefon, owned by Great Northern, Bell South, and Nordic Tel)
was franchised.
Equipment
In 1875, F. C. Tietgen, founder of the Great Northern Telegraph Company

(1869), created the first telegraph manufacturing company. Domestic telephone
set production began in 1892, followed by a number of foreign subsidiaries.
By 1987, the three major companies were GN-Telematic, Alcatel Kirk, and
NKT (Nordisk Kabel og Trad) (Pedersen, 1988).
Telecommunications equipment represents a quarter of the country's total
output of electronic products. In 1987, the telecommunications industry em-
ployed 32,000 people and exported 60 percent of its products (NKT, 1987).
The largest export markets are Scandinavian nations and the United Kingdom.
Although Danish companies generally have a small share of foreign markets,
mobile cellular firms established major shares in several countries.
Microwave transmission was patented in Denmark in 1902 but did not ex-
perience major growth until after World War II. In 1947, the Great Northern
Company established a terrestrial and maritime radio communications manufac-
turing subsidiary, Stomo, which became the dominant producer of mobile radio
equipment after opening a public mobile telephone network in 1963. Storno is
no longer Danish owned; it was acquired by General Electric in 1976 and sub-
sequently sold to Motorola in 1985. The other major radio manufacturer, A.P.
Radiotelefon, was created in 1953 and acquired by foreign interests in 1978.
The only major Danish-owned transmission equipment company left was Dan-
Denmark 227
call (originally Dancom maritime radio manufacturer), which was established

in 1980.
All the proceeding companies dominate the Danish equipment industry in
certain product lines. NKT manufactures optical fibers. Storno, A.P., and Dan-
call all produce cellular mobile telephone systems. Alcatel Kirk manufactures
PBXs and manufactures telephones and small telephone systems. GN-
Communications produces payphones, and GN-Datacom makes data transmis-
sion equipment. Yet another company, Regnecentralen (RC), supplies directory
systems and the packet-switched network. Danish Telecom International (DTI),
owned by the concessionary telephone companies, distributes telephone sets,
digital switching systems, PBXs, and the Dikon ISDN switch to international
markets (Pedersen, 1988).
In 1979, when optical fibers were first used in the Danish network, Siemens
was the supplier. NKT entered the market in 1983 after five years of R&D. In
1987, NKT was joined by AT&T in creating Lycom, an optical fiber produc-
tion subsidiary.
Alcatel Kirk (SEK) and GN Telematic (both established in 1892) held sig-
nificant shares of the telephone system market until 1990, when GN Telematic
was reorganized. Alcatel Kirk, originally Emil Moller's Telephone Company,
adopted its present name after the acquisition of ITT's telecommunications
business by France's Alcatel.
GN-Telematic was the only Danish-owned telephone set manufacturer until
Bang & Olufsen (B&O) began producing telephones in 1986. Bang & Olufsen
was established in 1925 and is the largest Danish-owned audio and video sys-
tems producer. The Dikon Systems Division was created in 1979 with the sup-
port of JTAS to bring B&O into the telecommunications industry with the pro-
duction of an advanced subscribers network switch. The Dikon switch had
advantages of maintenance and installation, and its success encouraged B&O
to expand further into telecommunications (Pedersen, 1988). In 1990 Ericsson
acquired 50 percent of Dikon and renamed it Diax Telecommunications.
Regnecentralen Computer (RC) was founded in 1955 and was jointly owned
by ICL, the telephone companies, pension funds, and the government. Its greatest
success was its 1975 advanced minicomputer (RC 8000), which enabled RC to
create a telephone directory system for the local companies. This system is
currently used by other countries, including the United States. RC also devel-
oped the packet-switched public data network, Paxnet, used by the Danish ROCs
since 1983. In 3988, the British firm ICL, subsequently controlled by Fujitsu,
bought a 51 percent stake in RC, changed its name to RC International and
shifted to UNIX-based systems.
Originally, the PTT and the concessionary companies held an equipment
supply monopoly in their respective areas. The four telephone administrations
still provide telephone sets, modems, PBXs, and telex, teletex, and data com-
munications equipment. All telephone equipment attached to the network either
must be supplied by the telephone administration or must meet the type-
approval set by the government Telecommunications Inspectorate.
Reform legislation in 1986 provided for successive liberalization of equip-
228 Denmark
ment, first of telephone extensions, then of PBXs, modems, telex terminals,

and main telephone sets. In 1989, all but first telephone sets had been liberal-
ized (TDR, 1989). By 1990, the entire terminal equipment market had been
liberalized, in line with EC directives to ensure Danish manufacturers' recip-
rocal access to the export market (M.K. Pedersen, 1990, communication).
Liberalization was slowed by telephone company ownership of the equip-
ment installed in homes and businesses. The electricians association claimed
that its Telepunkt stores were unable to compete with the existing de facto
equipment monopoly.
Services
Denmark's telephone companies offer a variety of data transmission services.

Paxnet is a packet-switched data network. Datex is a public circuit-switched
data network. The three local telephone organizations also provide slow-speed
Datel service. Until the end of 1986, the PTT provided and maintained all
modems, except the lower-speed (300 baud) modems, which were supplied by
the local companies. Subsequently, local operators provided all modems above
2400 baud, and the market for lower-speed modems was liberalized. In 1988,
all modems were liberalized.
Since 1985, extensive investments in the future of ISDN have been under-
taken by both the PTT and the local companies. Field trials were begun in
1989 using Ericsson AXE and Siemens EWSD switches. The local companies
also introduced Meganet, a 2-Mbps data transmission service using Dikon
switches.
The Nordic Mobile Telephone system (NMT) has offered mobile cellular
telephone service since 1982. NMT was provided by the PTT and not by the
regional companies. Demand exceeded expectations. The combined subscriber
base for the NMT450 and NMT900 systems is almost equivalent to those of
France or West Germany.
Next to the United Kingdom, Denmark has the world's second-lowest telex
charges (Barnes, 1987). Electronic mail teletext service was introduced in 1984
and operates at a speed of 2400 baud. TeleDenmark also provides a videotex
service named Teledata, which operates under the Prestel standard. Prestel de-
velopment was stopped, however, while the PTT and the regional companies
developed a new system based on the CEPT standard. Telex and teletex users
have access to the electronic mailbox service, Databoks. In 1988, Denmark
liberalized value-added services, subject to registration. Transmission capacity
could be resold to data customers on a leased-line basis, although a 5 percent
limit was placed on resold traffic (P&T Directorate General, 1988). Anticipat-
ing liberalization, KTAS established joint ventures with IBM (danNet) for data
processing, teletext, and EDI.
In 1979, the cable manufacturer NKT proposed ambitious plans to connect
with local telephone companies to provide an integrated broadband network
(IBN). This followed growing concern with the emergence of cable television
Denmark 229
and direct broadcast satellites. For example, households could have access to
their own or community satellite antennas to the exclusion of the established
telephone carriers.
In 1983, a Royal Commission on Mass Media proposed that a national
broadband network be established that would provide both television and tele-
phone linkages. The independent telephone companies were in favor of a na-
tionwide broadband system because it would free them from prior agreements
with the FIT (Qvortrup, 1984). The PTT, however, preferred to separate tele-
phone from broadband services, with a narrowband ISDN network to link tele-
phone connections and a broadband network to interconnect the national cable
television and master antenna systems. The latter would use a nationwide PTT
microwave system and PTT-operated satellite earth stations. These issues were
resolved in the Bernstein Committee's 1988 recommendations, which left
broadband services such as data transmission and video service providers to the
local companies.
Much of the country was cabled for the Digital Optical CATV Trunk (DO-
CAT) system, which allows transmission of eight television and twelve FM
channels on each broadband optical fiber.
The PTT and the telephone companies were given a monopoly for the recep-
tion of television programs from communications satellites, thus forcing house-
holds to partly finance the advanced national cable system. This led to much
unhappiness among cable firms and the numerous housing associations that
operated master antenna satellite systems. Provocative violations of the law
ensued. In consequence, the parliament changed the law in 1987 and permitted
a system by which private parties could receive satellite-delivered programs and
compete with the traditional telephone carriers, which themselves can transmit
these satellite programs to master antenna associations as "hybrid network con-
nection" (hybrid both in terms of analog-digital and fiber-coax combinations).
The notion of the unified national broadband network was thus significantly
affected, since the new law permitted broadband reception and limited distri-
bution outside of it.
19
Iceland
Iceland is a prime example of a country suffering from the diseconomies of

small scale. Its telecommunications are constrained by the country's basic real-
ities: Its rugged terrain is sparsely populated by only a quarter of a million
people; its remote geographic location requires advanced international commu-
nications; and it lacks an electronics industry. It is difficult even to imagine
competitive telecommunications in Iceland. Yet by the same logic of scale, the
integration of parts of Icelandic telecommunications with the larger systems of
its frequent Nordic partners would be conceivable.
Telephone service began in 1889, when a businessman in Isafjordur, in
northwestern Iceland, connected his various offices by telephone lines. In 1890,
a private line was opened between Reykjavik and nearby Hafnarfjordhur. Gov-
ernment operations began in 1906 with the trunk connection of Reykjavik to
the east coast and the creation of telephone and telegraph exchanges. By 1929,
service reached around the country. A large automatic telephone exchange be-
gan operating in Reykjavik and Hafnarfjordhur in 1932. In 1935, shortwave
radio telephony was opened to England and Denmark. In 1947, service to the
United States was implemented.
The Icelandic PIT has a monopoly over telecommunications. It is state owned
and reports to the minister of transport. In 1988, telephone density was about
46.6 main lines per 100 persons and there were approximately 117,000 sub-
scribers. Quarterly subscription for residential phone service is 585 Kr (about
$16), and a local call costs 1.32 Kr ($0.04) per six minutes in daytime. Long-
distance calls up to 100 kilometers cost 1.32 Kr ($0.04) per eighteen seconds.
A one-minute call to the United States costs 85 Kr (about $2). These rates were
the lowest among all OECD countries for both residential and business service
(OECD, 1990, p. 30).
Between 1965 and 1975, all the urban exchanges were automated, and full
automation of the nation's system was achieved by the end of 1986. In 1981,
the PTT decided to replace the older analog switches with digital switches. The
first digital exchanges employed Ericsson AXE-10 switches and began operat-
ing in 1984 in Reykjavik and Keflavik. Switching equipment by Elektrisk Bu-
reau (EB) of Norway was also used. At the end of 1989, seven AXE central
exchanges and twenty-two secondary exchanges served 35 percent of the sys-
tem's subscribers (Tomasson, 1990). Plans to increase digitalization and deploy
fiber-optic cables were slowed by a government austerity program.
230
Iceland 231
Iceland has no telecommunications equipment industry. Siemens, Ericsson,

Philips, Telic/Alcatel, and SEK all offer PBXs, most of which are supplied and
serviced, but not rented, by the PTT. Also offered are key systems from a wide
variety of suppliers, including Japan. All terminal equipment, including mo-
dems, is supplied through the PTT. Equipment must not disturb or cause in-
convenience for the PTT or its subscribers. Telex for data transmission, tele-
graph, and telephone circuits can be leased from the PTT.
Mobile telephony was introduced in 1986 with the implementation of the
NMT 450 system. In 1987, an AXE exchange was added to the system and the
number of mobile subscribers increased to 4400. Such traffic is especially heavy
between fishing boats and the shore.
After planning a public data network since 1983, the PTT implemented the
Eripax X.25 system ICEPAC for domestic use in 1986. The network is cen-
tered in Reykjavik with six remote nodes. ICEPAC was first linked to Copen-
hagen and London and is now linked to most Western countries. Since 1986,
digitalization of the network has been pursued through the AXE exchanges,
digital radio systems, and fiber-optic links.
A microwave ring around the country was completed in the early 1980s.
Since 1980, the Skyggnir earth station has permitted international direct dial-
ing. By 1987, 500 kilometers of fiber-optic cable had been laid.
20
Ireland
History
The United Telephone Company opened its first exchange (with five subscrib-
ers) in 1880.' United was taken over in 1882 by the Telephone Company of
Ireland, which oversaw a slow growth to 500 lines in Dublin by 1988. The
operation was acquired in 1893 by the National Telephone Company, which
rapidly developed the network to encompass fifty-six exchanges by 1900. The
Post Office, which operated the telegraph and feared revenue losses, began in
1893 slowly to invest in trunk lines and submarine cables. By 1908 the Post
owned thirty-three exchanges to National's eighty-five (Litton, 1961).
The threat of Post Office take-over of National Telephone became a reality
in 1905; the postmaster general forced a sell-out at the end of National's license
period in 1911, paying only 12.4 million where the company had claimed its
value was 20.9 million. Military involvement in network construction began
in 1909, completely replacing civilians until the outbreak of World War I. By
1918, there were 12,500 lines, half of them around Dublin; three counties were
still without exchanges. The network was greatly damaged in the Anglo-Irish
and civil wars of the 1920s.
In 1924, the Post Office launched a vigorous program to rebuild and expand
the crippled network. By 1930, only western pockets of Mayo and Donegal
counties were without exchanges. Dublin received the country's first automatic
exchange system in 1927, but economic hardship in the 1930s stalled the net-
work's growth.
In 1937, there were still fewer than 40,000 lines in service (Department of
Industry and Commerce of Ireland, 1944). Except for the emergency installa-
tion of telephone lines to coastal lookout posts, progress halted altogether dur-
ing World War II. In 1945 the government earmarked 10 million for a pro-
gram to reach 100,000 subscribers within fifteen years, which was achieved
ahead of schedule. The country's first transatlantic cable was laid in 1956, and
Limerick began operating Ireland's first crossbar switch in 1957 (Telecom Ei-
reann, 1987). By 1960, however, Ireland still had only 145,000 telephones in
service, fewer than any European nation except Greece, and line density in
Dublin was a mere 9.25 percent (Department of Industry and Commerce of
Ireland, 1969).
By the late 1960s, connections were accelerated. By 1979, a total of 436,000
232
Ireland 233
lines were in operation, but the waiting list was still lengthy (Central Statistics
Office, 1979, p. 322).
Telecom Eireann
For Irish telecommunications, the 1980s was an era of catching up. In 1978,
the inferior quality of services provided by the government's Department of
Posts and Telegraphs forced the newly elected Fianna Fail government to com-
mission an external and independent group to review Irish telecommunications
(Raggett, 1984). Upon completion of its task, the commission issued a report
that urged immediate action. The inadequacy of telecommunications was ac-
knowledged. The government established an interim Telecoms Board, and as
part of a broad infrastructure program it approved accelerated development.
Soon after, under the P&T minister Albert Reynolds, the government launched
a five-year telecommunications spending program of over $1 billion for the
early 1980s. The program aimed to modernize and upgrade the network to the
quality of other European countries and to increase its availability throughout
the country (Ergas and Okavana, 1984). The implementation of this program
helped to change the government's previous tendency to consider telecommu-
nications as a money maker for its other services.
The government published legislative proposals based on the review group's
report, but enactment of the legislation was impeded by five changes of gov-
ernment in the three years of the interim Telecom Board. The proposals as
outlined, however, encountered little political opposition (Keenan, 1985; Rag-
gett, 1984).
Eventually, the Postal and Telecommunications Services Act of 1983 was
passed. It created two state-sponsored enterprises, the Bord Telecom Eireann
(TE), for telecommunications, and An Post, for the national postal service.
Carved from the Department of Post and Telegraphs, a Department of Com-
munications was created to supervise general policy on telecommunications is-
sues (NTIA, 1985). Actual telecommunications operations are provided by
Telecom Eireann (TE). Tom Byrnes, an Irish-American who has been the man-
aging director of IBM Ireland, served as Chief Executive of TE. TE was re-
quired to operate on a profit basis, with near-monopoly rights of service, and
embarked on an ambitious modernization program. In 1987, foreign loans helped
fund 50 percent of TE's investment program (Garnett, 1987).
After separating from the PTT, Bord Telecom Eireann was still a state-owned
enterprise but operated with greater autonomy, under an exclusive government
license. It has an "exclusive privilege," which is somewhat less than a mo-
nopoly, up to and including the connection point in the customers' premises.
Where TE decides not to provide service, another network operator can be
licensed, as in the case of cable service. Such a license can be issued by the
PTT minister or by TE itself. Where services are uneconomical, the govern-
ment can force TE to provide them. In 1985, the cable television committee,
established by Communications Minister James Mitchell, recommended the es-
234 Ireland
tablishment of a national distribution network linking local cable networks through

TE.
Compared to other European countries, Ireland has a relatively low telephone
density. In 1990, telephone line density was 26 per 100 people, or 64 percent
of households. Total number of lines by 1990 was 950,000, of which 34 per-
cent were business and 64 percent were residential. Annual line charges in
1990 were about $201 for residential service and $222 for business.
TE's main problem has been to improve service in Dublin to be comparable
to the service existing outside of the capital, a reversal of the usual quality
differential (Keenan, 1985). Partly because Dublin was the fastest-growing ma-
jor city in Europe, telephone installations have lagged far behind. Much of the
investment funds, therefore, have been used simply to catch up in the mass
provision of basic telephony. TE has also had to grapple with the rate structure,
which had been time insensitive for local calls. In addition, it inherited an
inflexible, graded employment structure that created problems in the rapidly
changing environment.
At first, TE depended on funds from the Ministry of Finance. But as the
coalition government of the mid-1980s became increasingly mired in debt, a
new strategy was devised that was influenced by the French model. It estab-
lished a subsidiary, Irish Telecommunications Investments (ITI), which invests
funds in the private financial markets and in the telecommunications network.
ITI technically owns these assets and charges TE for their use.
Unfilled service orders dropped from 20 percent of total customers in 1980
to 2.9 percent in 1986, and the waiting period fell to six weeks. Operating
revenues rose to IR784 million, a 20 percent increase from 1985 to 1986
(Dillon, 1987). By 1990, the waiting list for main lines had dropped to 6000,
less than one percent of total subscribers (ITU, 1990). In real terms, TE's rates
were reduced by almost 30 percent between TE's establishment in 1984 and
1990 (F. McGovern, 1990, communication).
In 1985, Byrnes left TE at the end of his five-year contract; Ireland's Min-
ister for Communications had let the contract lapse without signing the seven-
year extension on the terms approved by TE's board. In departing, Byrnes
minced no words: "the real tragedy, in my view, is that shortsighted, very
shortsighted, politicians combined with arrogant power-hungry bureaucrats, have
reinstated the same practices which were disastrous for this country prior to
1979" (Telephony, 1985, p. 83). As an example, Byrnes cites a 1985 attempt
by the government to borrow up to $200 million from private sources through
TE's Irish Telecommunications Investments (ITI) subsidiary. ITI was estab-
lished to help fund TE construction projects, but the Department of Finance
used it to relieve Ireland's federal deficit.
Byrnes was succeeded by Fergus McGovern, former head of operations of
TE. Some financial help came from the European Community. Ireland was
eligible for 64 percent of the budget of the first phase of the EC's $865 million
STAR development program. Its dependence on EC funds was likely to nudge
Telecom Eireann to comply with Brussels' liberalization policy.
Ireland 235
Equipment
The supply of subscriber equipment, including first sets, is largely liberalized.

For both equipment and other supplies a private provider must obtain type
approval from the government and consent from TE for each connection to the
public network. The minister of communications grants equipment approval
and TE and the Institute of Industrial Research and Standards administers
equipment tests. Each private supplier must also demonstrate technical ability
and financial viability.
TE supplies equipment through a wholly owned subsidiary, Telecom Eire-
ann Information System (TEIS), which was established to avoid the cross-
subsidization and unfair tax advantages of TE (C. D. Rafferty, 1987, commu-
nication). However in 1990, the Telecommunications Association (TA), a group
of some twenty Irish equipment suppliers, charged TE and TEIS with unfair
competition before the European Commission. TA accused TE of eighty vio-
lations, including undercharging TEIS for services, releasing advance infor-
mation on orders, and operating as a single source despite professed indepen-
dence (Evagora, 1990a). The TA urged the establishment of an independent
watchdog group modeled on the U.K.'s Oftel. TE denied the charges and TA's
lawsuits were unable to stimulate EC injunctive action.
In 1979, a decision was made in favor of digital technology. Two switching
systems were introduced, Alcatel's E10B and Ericsson's AXE. Both are pro-
duced or assembled in Ireland. These modernization plans largely eliminated
manual switching, which in 1978 still accounted for 50 percent of exchanges
serving 10 percent of subscribers.
For a quarter century, the Swedish L.M. Ericsson supplied exchanges to
Ireland from a manufacturing plant in Athlone. For the new digital switching
system, a joint venture of the French Alcatel and Telectron was chosen. Active
in transmission, Telectron is the main Irish telecommunications firm, with sig-
nificant exports to the Middle East.
In 1981, AT&T acquired 45 percent of Telectron. This acquisition forced
the relationship between Telectron and Alcatel to dissolve. Alcatel then formed
Alcatel Ireland, a local affiliate that is 75 percent French and 25 percent Irish.
Services
For a long time, Irish international traffic used to be routed through Britain. In
1984, however, Ireland established an Intelsat earth station near Cork. Since
then, it has vied to become an international gateway for European traffic, but
this goal clashed with tariff reality. Calls placed from Ireland to a number of
other European nations tend to cost considerably more than calls traveling in
the opposite direction (Transnational Data and Communications Report, 1988).
Exceptions are calls to generally high-priced countries (Belgum, Greece, Italy,
236 Ireland
Spain, and Portugal). In 1990, however, rates for digital lines were reduced
significantly.
Since 1981, international packet switching service has been available, and in
1984, TE introduced the packet-switched network Eirpac. It allows access to
bibliographic and full-text databases. Eirpac also connects to national and in-
ternational videotex services.
A major VAN service is Cognotec, established by the Confederation of Irish
Industry in 1984 and restructured in 1987 with strong insurance industry partic-
ipation. Its Corporate Treasury service provides company controllers with ac-
cess to financial information. In partnership with Istel (a subsidiary of AT&T),
Cognotec launched its Corporate Treasury service in the U.K. market in 1990.
Cognotec's other major service is Clientbank, which gives insurance brokers
access to host computers of leading insurance companies.
Agriline, an agricultural videotex database, was launched in 1986 following
a two-year EC-supported trial. It offers weather, market prices, a calendar of
events, and farm business news as well as information on crops and livestock
management.
Eircell mobile communication is provided by TE. The system uses the TAGS
technical standard and had 19,000 subscribers in 1990, the EC's third-highest
penetration per working population, at comparatively low rates for Europe (F.
McGovern, 1990, communication). Eirpage is a joint venture with Motorola
Ireland in which TE has 51 percent. It holds the monopoly on national paging
services in Ireland.
Data and Special Services Network (DASSNET), launched in 1990, is a
significant improvement of the digital leased-line network, providing 64-kbit-
2-Mbit connections. The Government Telecommunications Network, linking
government departments in Dublin and five provincial centers, planned to use
the advanced DASSNET.
Telecommunications in the
Mediterranean Countries and
Eastern Europe
21
Italy
No European telecommunications system has been institutionally more complex

than Italy's. Through the 1980s, it was best described as a shared monopoly of
five organizations. Two of these organizations are government administrations;
the other three are nominally private groups. The state directly runs the State
Agency for Telephone Services (ASST), which handles long-distance and Eu-
ropean telephony, and the Post and Telecommunications Administration (PT),
which handles telegraph and telex networks. The three other groups are Societa
Italiana per 1'Esercizio Telefonico (SIP), for local exchanges and some long-
distance service; Italcable, for intercontinental service, and Telespazio, for sat-
ellite service, all of which are part of the state-dominated Societa Finanziaria
Telefonica (STET) holding company. In theory, the system separates segments
of network control as the AT&T divestiture did in the United States, and it
reduces monopoly power. In practice, however, the system in its aggregate is
an inflexible bureaucracy, with service problems for small and large users.
There are few performance rivalries, but frequent jurisdictional and political
disputes. These revolve around financial transfers, particularly over dominance
in new fields, such as data transmission and ISDN. Added to these problems
are the vertical ties the STET companies maintain with sister firms in manufac-
turing, notably Italtel, the country's dominant telecommunications equipment
maker. The protected market shares make Italy's electronics industry less dy-
namic than several other segments of its economy. One exception is Olivetti,
which is outside this arrangement. On the positive side, the existing complex
institutional system would allow for relatively easy modifications. For instance,
a privatization of the STET network firms and their separation from equipment
suppliers is gradually taking place, although not as part of a planned telecom-
munications policy. Additionally, equipment liberalization has progressed, partly
because users' self-help could not be contained. As the STET firms become
more entrepreneurial, the shared service monopoly could break down in the
future. But this is counteracted by the government, whose chief initiative in
telecommunications seeks to centralize the various institutions into one super-
provider.
239
240 Italy
History
Italy's postal service originated in medieval times when the major cities and
trading companies established Europe's first courier services. In Venice, a
guildlike messenger organization existed and operated in a monopolistic fashion
after 1305. Milan's system dates to 1385, and Naples created a runner course
for southern Italy in 1444 (Dallmeyer, 1977). The Tassis family, which became
a major presence in the postal system of much of Europe, also established
service in Italy. The Italian PTT was established in 1870, during the unification
of Italy.
Italian telephony started as a private business. The first large company, So-
cieta Generate di Telefonia, was established in 1881 (Holcombe, 1911). Lim-
ited telephone service began in Rome in 1878; the American Bell Company
opened exchanges in Rome and Milan in 1881. Also in 1881, the first inter-
urban service was provided between Rome and Tivoli on an experimental basis.
Rival companies and competitive service emerged in some of the larger cities.
Pressure to consolidate soon mounted. Within a short time, only two private
telephone companies remained. In 1883, a royal decree established uniform
obligations for concessionary firms, and burdensome requirements were im-
posed to protect the telegraph authority. Concessions were not awarded exclu-
sively, and ran for only three years. The telegraph authority approved all public
call offices as well as the private telephone rates. It levied a heavy concession
fee on local exchanges and permitted municipalities to purchase the private
telephone systems after twelve years. However, only one municipality wound
up owning a local telephone ownership. In 1907, there were 141 local networks
with 43,000 subscribers. In that year, the Italian parliament voted to purchase
eighteen long-distance lines and twenty-seven local exchanges, then operated
by the two major private telephone companies (Societa Generale Italiana de
Telefonia e Applicazione Elettriche and Societa Telefonica dell'Alta Italiana,
accounting for 75 percent of Italy's telephones). Under the plan, compensation
to the owners of these companies was to be paid over eleven years out of future
telephone profits.
Between 1907 and 1925, telecommunications were jointly provided by the
state and by sixty-three regional private concessionaires. Local telephony was
legally franchised to private firms, but long-distance communications remained
under state control. Subsequently, under Mussolini, five telephone regions were
established and assigned to different concessionaires: STIPEL, TELVE, TIMO,
TETI, and SET. The PT organization ASST was established to provide long-
distance interconnection between the five regions and international services with
European and Mediterranean nations. This interexchange network was com-
pleted in 1928.
Forces for further centralization were strong. In October 1933, three of the
concessionary firms were absorbed into the government holding company STET,
which in turn was part of the government's industrial reconstruction institute,
IRI. In 1958, the two remaining regional companies, TETI and SET, became
Italy 241
part of STET. In 1965, these five concessionaires were among nine firms merged
into the Societa Italiana per 1'Esercizio Telefonica (SIP), a company that had
started as a northern Italian electrical utility. About 70,000 shareholders owned
approximately 45 percent of SIP's capital; STET controlled the remainder. In
turn, about 50,000 shareholders controlled 42.3 percent of STET, with the
remaining 57.7 percent held by IRI.
In the meantime, international communications also consolidated. Interna-
tional telegraph service was initially divided between the radio provider, Italia
Radio, and the submarine cable provider, Italcable, and was centralized in 1941
when Italcable acquired its rival. It was further consolidated in 1965, when
Italcable became a subsidiary of STET. Telespazio, Italy's satellite commu-
nications provider, was established as part of STET in 1963.
Structure
The complexity of the Italian system of telecommunications is largely hidden

from the user's view. Virtually all user transactions except telegraph and telex
are through SIP, which provides customers with a single telephone bill. The
complexity of the system is upstream in the transmission path, where various
organizations carved out segments of control. Despite the multiplicity of actors,
at no point could users choose a provider, and they faced a monopoly, or more
accurately, a shared monopoly. By tradition, politics, or influence, five orga-
nizations were able to control geographical or service areas within the monop-
oly. Virtually all these organizations were in one way or another government
controlled, but by different parts of the state bureaucracy. In theory, this sys-
tem could permit a comparison of performance and diffuse otherwise consid-
erable power. In practice, the organizations were political rivals and not market
competitors. Italy's telecommunications structure reflects the more general per-
vasiveness of politicization. By law, all telecommunications services are the
province of the state. The complexity starts because the government operated
telecommunications directly, through two of its own administrations, and it
also granted operational authority to three outside concessionaires, notably SIP.
These concessionaires were controlled by parts of the government other than
the PTT ministry. These entities divided domestic and international long-dis-
tance service among themselves in no obvious pattern. The lead agency respon-
sible for telecommunications is the Ministry of Post and Telecommunications
(MPT). The MPT provides mail service and the vast postal bank transfer sys-
tem. Two bodies, the Administrative Council and the High Technical Council
for Telecommunications, assisted the ministry in its overall control of telecom-
munications. The Administrative Council included a wide array of governmen-
tal interests and is consulted on tariffs, proposed regulations, and concession-
aires. The High Technical Council for Telecommunications consisted of experts
from government, universities, and the concessionaires. It advised the ministry
on technical and economic matters, development plans, and operations.
Subordinate to the MPT through the 1980s were two autonomous administra-
242 Italy
tions: the Post and Telecommunications Administration (FT) and the State Agency
for Telephone Services (ASST). The PT conducted postal telegraph, telex, te-
lefax, and radio electric services. ASST, an autonomous public corporation,
operated the primary long-distance transmission network that connected the
twenty-one telecommunication "compartments" of the country. ASST also op-
erated international telephony within Europe and the Mediterranean basin.
Through the PT, the MPT provided a coastal station network (DCR) for com-
munication with ships, minor islands, and Albania. It also provided broadcast-
ing transmission for government operations such as embassies abroad. Two
other companies, Sirm and Telemar, operated radio equipment aboard Italian
ships, and potentially other services where no private concessioniaries can be
secured. The national network's layout consisted of twenty-one regional com-
partments, subdivided into 231 districts, each with its own area code. These
were, in turn, divided into about 1400 sectors and 10,000 local exchanges, of
which more than half have less than 500 subscribers.
Perhaps the best way to conceptualize the allocation of services in Italy is to
imagine four concentric rings, in which the responsibilities alternated between
the government-controlled STET holding company and direct government pro-
vision. At the core, STET's subsidiary company SIP operated local distribu-
tion. SIP also provided domestic packet-switched service. The next ring around
it, corresponding to a greater distance of communication, was domestic long-
distance transmission, operated by ASST. (For a long time, SIP also ran long-
distance service in many rural districts.) The next ring, international service
within Europe and many countries in North Africa, was also in the domain of
ASST. The third ring was international transmission outside of Europe and
North Africa, which was operated by the STET companies Italcable (for terres-
trial and submarine transmission) and Telespazio (for satellites). Both also pro-
vided the PT administration with international record traffic.
SIP is by far the largest Italian service operator, with a turnover of 14,900
billion lire and 77,000 employees in 1989; the next largest, ASST, had reve-
nues of 2,400 billion lire and under 13,000 employees.
An interesting side-effect of this system was that the separation of firms by
different functions makes internal cross-subsidization more difficult (Benedetti,
1983). The creation of a cross-subsidization fund in 1980, the Cassa Conquag-
lio per II Services Telefonico, did not clarify this complex accounting process.
SIP's local service was separate from ASST's long-distance service. Thus, where
contributions existed, they were by payment to another entity or by contribu-
tions to a parent company, and were thus more transparent than internal subsi-
dies would be within a unified PTT system.
In 1985, the PTT ministry (MPT) initiated a study of the feasibility of sep-
arating postal and telecommunications activities. The study proposed separate
agencies within the ministry to administer each of these activities throughout
the country. The recommendations, however, were not instituted. In 1987, a
bill was introduced that would change the PTT role from operating services to
planning and coordination and that would divide service provision so that Ital-
Staly 243
cable would provide the entire international service and SIP would provide the
domestic one. Additionally, the bill proposed that Telespazio operate indepen-
dently, providing satellite service for SIP, Italcable, and RAI. ASST's opera-
tions would be assumed by SIP, Italcable, and the ministry. The bill's oppo-
nents focused on job relocation and preserving the state control.
Subsequently, the Christian Democrats offered a plan in 1988 to consolidate
the telecommunications agencies into a "Super-STET"including SIP, ASST,
Italcable, and Telespazioaggregated as "Italia Telecom." It would have rel-
egated the ministry to a watchdog role (Rosenbaum, 1988). But this project
was blocked by the PTT minister, trade unions, and southern Italian Christian
Democrats, who traditionally were influential in MPT matters and were con-
cerned with loss of telecommunications and postal patronage. An example for
the politicization of the decision process: ASST was headed by the leader of
the Christian Democratic Italian Union Workers' Confederation. After the pro-
posal failed, an alternate plan to create a "Super-SIP" was introduced by So-
cialists, but this was also blocked by the Christian Democrats as well as by top
management. In 1990, two bills were introduced to break the logjam. One bill
proposed a change in the union contract for ASST, to protect those members
opposed a move to the private-sector IRI-STET group where they would lose
the right to retire after twenty years and receive a smaller pension. The other
bill proposed separation of post and telecommunications activities of the PTT.
In 1990, the telecommunications ministry recommended a "Super-STET"
under which the government's ASST would be transferred to STET. But the
push for structural reform lost some of its steam when a government crisis led
to the replacement of Minister Oscar Mammi, a member of the economically
liberal Republican Party.
Italy was slow to focus on the role of telecommunications in the development
of its economy. Although it was advanced in completing a universal subscriber
trunk dialing, the system stagnated in the late 1970s, when, for political rea-
sons, rate increases lagged behind rapid inflation. The major problem in up-
grading Italian telecommunications has been reconciling the cost of moderniz-
ing the network and of trade union demands for high wages and flexible working
conditions with political pressures to keep telephone rates from rising. This
dilemma became particularly acute in the late 1970s, when local rates were
frozen even with inflation raging at 20 percent. In consequence, no funds re-
mained for investment. The ratio of SIP's self-financing (the contribution of a
company's own earnings to capital formation) plummeted to 10 percent in 1980,
when the firm reported losses of 538 million lire (Benedetti, 1983). SIP had to
resort to indirect ways to increase rates, such as requiring subscribers to in-
crease their deposits. It also had to borrow at the prevailing high interest rates.
Subsequently, interest payments at times ate up 30 percent of its total income.
The investment slowdown, in turn, adversely affected equipment suppliers and
penetration. Italy's telephone density was about half of West Germany's; there
were only half as many telex subscribers as in France and one-third as many
as in the United Kingdom. Telephone switches were largely electromechanical
244 Italy
rather than electronic, and the system was chronically congested. Italy was also
late in introducing national data networks, which were implemented eleven
years after Germany and seven years behind France's Transpac (Pozzi, 1987).
Tariffs were often set by politics, resulting in a complex split of local, long-
distance, and international revenues. A parliamentary committee finally rec-
ommended streamlining the cumbersome and restrictive rate-setting process and
permitting rates to reflect inflation. The government responded in 1986 by in-
creasing rates, giving SIP more pricing flexibility and fine-tuning the rate struc-
ture with time-measured local calls and peak load pricing. It also used a com-
pensation fund, the Cassa Conquaglio, to transfer carrier profits from ASST
and Italcable to SIP. It even temporarily reduced SIP's concession fee from 4.5
percent to 0.5 percent of its revenues, invested in its shares, and guaranteed
purchase of the remainder. SIP's self-financing ratio increased to almost 50
percent by 1983 as a result of these actionsstill low by comparison. France's
DOT had a ratio of approximately 65 percent, that of British Telecom was 60
percent, and the Bundespost's was 100 percent (Benedetti, 1983). By 1988,
SIP's self-financing had reached 90 percent of its capital expenditures.
In 1982, the PTT ministry also formulated a ten-year plan for the telecom-
munications sector. The targets for 1990 were a density of thirty-eight subscrib-
ers and fifty-seven telephones per 100 population (up from twenty-four and
thirty-eight in 1981), and a digitalization, by the year 2000, of 50 percent local
and 80 percent long-distance switching. In 1989, digitalization predictions were
revised upward for 55 percent of all switches by 1993 and 100 percent by 2000.
Trunk lines would be 85 percent digital by 1993 (Benzoni, 1990).
Even after streamlining, the rate-setting process was extremely complex. Users
making fewer than fourty calls per month paid 40 lire per billing unit, and
others paid 127 lire. Italy's 3.9 million business users paid rates well above
the OECD average, and telex service was Europe's most expensive. Interna-
tional rates were also the highest in Europe in 1988, with a three-minute call
from Milan to London costing 33 percent more than a call in the other direction
(OECD, 1990, p. 30; Schenker, 1990). Perhaps because of its high rates, Italy's
international traffic was half that of France's and one-third of Germany's, ac-
counting for less than 1 percent of network usage. Even Switzerland generated
more total minutes of international calling. Those three nations accounted for
50 percent of Italy's traffic, with an additional 20 percent going to the United
States and the United Kingdom (Staple, 1990; Smau, 1990).
SIP greatly reduced the waiting list for basic service, from 750,000 in 1981
to 118,900 in 1988, though penetration rates were relatively low (34.9 main
lines and fifty-one telephones per 100 persons). A great disparity persisted be-
tween telephone penetration in the industrial north/central region (fifty-seven
per 100 persons) and the rural south (thirty-seven per 100) (ITU, 1990).
In 1989, SIP launched a four-year investment and expansion plan called Piano
Europa to invest $8 biliion above the already budgeted $18 billion to connect
4 million new basic subscribers by the end of 1992, 43 percent of which reside
in southern Italy.
Italy 245
Most equipment procurement is from Italian companies, or Italian operations

of foreign companies. Italtel, known originally as SIT-Siemens, is Italy's major
telecommunications equipment manufacturer, with 1986 sales of around $1 bil-
lion and 17,745 employees. It supplies more than half of SIP's switching
equipment needs and 40 percent of its transmission equipment. Under Marisa
Bellisario, Italy's only woman CEO of a major firm, Italtel emerged from near
bankruptcy in 1981 to financial solvency, mostly by cutting employment from
30,000 in 1981 to 19,000 three years later.
In 1985, SIP bought 75 percent of Italtel's production. Italtel's exports are
relatively anemic. In 1985, they totaled about $30 million, or 5 percent of total
sales, which consisted primarily of government-subsidized aid to developing
countries, including Zambia and Guatemala.
In 1986, STET and Fiat attempted to merge Italtel and Telettra against in-
creasing international competition. The Italtel/Telettra merger would have joined
the leading public switch manufacturer with the top transmission and radio
communications supplier. Plans for the new company, to be known as Telit,
were approved by parliament, but later suspended because of bickering. The
fundamental logic behind the Italtel/Telettra merger was not enough to over-
come the struggles for managerial control. Fiat's chairman, Giovanni Agnelli,
disapproved of allowing Bellesario to run Telit, fearing the influence of bureau-
crats at IRT/STET, Italtel's parent firm. Tragically, Bellisario died at a rela-
tively young age in 1988.
Another manufacturer is Ericsson's subsidiary, FATME, which employs 4500
workers in nine factories around the country. FATME had about 20 percent of
the market. Alcatel, controlled by the French CGE with minority participation
by ITT, employs about 10,000 at its Italian subsidiary, FACE. GTE had a
substantial presence in Italy for many years. In 1986, as part of a global agree-
ment with Siemens, it sold its manufacturing facilities and contracts to the
German firm. Hewlett-Packard joined forces with Telettra, the telecommuni-
cations arm of the Fiat group, for advanced private communications and data-
processing systems and equipment.
Conversion to digital switches in Italy was launched in the early 1970s, when
SIP began a program for the introduction of PCM systems. Following the ex-
ample of Alcatel in France, Italtel bypassed the semielectronic stage of switch-
ing and installed digital switches, except in some smaller exchanges. In 1980,
the first digital exchanges were introduced, Italtel Proteo's TN-16, and AXE of
Ericsson's Italian company FATME.
In 1982, the governmental interdepartmental committee for economic plan-
ning decided that telecommunications modernization was to be based on elec-
tronic switching equipment provided by a competition between two systems,
one of which had to be Italian. The primary system, known as the first "pole,"
was awarded to the Italian joint venture based on the Proteo technology.
246 Italy
SIP planned to procure the majority of its exchanges from the National Pole
and the remainder from either FACE, then an ITT subsidiary, or FATME,
Ericsson's subsidiary, both of which were domiciled in Italy. However, since
the employees of the loser among FACE and FATME would likely have to lay
off employees, both companies were chosen as suppliers with the vague expec-
tation that their systems would be modified into a uniform switch. In 1985, the
National Pole (Italtel, GTE, Telettra) received two-thirds of the $900 million
contract, FACE had 14.2 percent, and FATME had 20.4 percent. This was a
continuation of their already existing and quasi-established market shares. Those
steady market shares for switching equipment had been Italtel, 52 percent; Te-
lettra, 1 percent; FATME, 18 percent; FACE, 17 percent; GTE (later Siemens),
12 percent (R. Lauro, 1987, communication).
In developing the Proteo system, Italtel first entered into a joint venture co-
operation agreement with the American-owned GTE Telecomunicazioni and
Telettra, the Fiat subsidiary. It soon required another major partner. This led
to a major joint venture between AT&T and Italtel, which won a large portion
of the $28 billion that Italy planned to spend on network equipment by 1992
(Colby and Hudson, 1989, p. 27). The Italtel-AT&T deal followed some heavy-
handed lobbying for the partnership position, including personal intervention
by Francois Mitterrand for Alcatel and Ronald Reagan for AT&T. Alcatel
chairman Pierre Suard played the European card, stating that the "goal of Eu-
ropean telecommunications policy must be to maintain and improve the lead-
ership of European industry" (Roussel, 1990). Germany, on behalf of Sie-
mens, threatened to block $4 billion in subsidies for Italy's steel industry. Former
Prime Minister Bettino Craxi, a socialist, supported AT&T (Hayes, 1988). Ul-
timately, AT&T paid the government $130 million, the difference in value of
the stock swap of its 20 percent share of Italtel and Italtel's 20 percent stake in
AT&T's Network Systems International.
In response to the AT&T-Italtel partnership, Alcatel acquired Telettra in
1990, thereby gaining Teletra's thirty-three percent market share in transmis-
sion equipment in Italy and forty-five percent in Spain. But the European Com-
mission first required Spain's Telefonica telephone monopoly to divest itself of
its shares in Alcatel (21 percent in the Spanish subsidiary) and Telettra (10
percent), as well as to open its purchasing policy.
The STET Group

Societa Finanziaria Telefonica (STET) is the mainstay of Italian telecommuni-
cations. With its array of local, domestic long-distance, and international tele-
phone services, as well as manufacturing, it is a formidable presence in Italian
communications. In 1990, STET's revenues were 17,200 billion lire, including
exports of $1 billion; investments were 5000 billion lire; employment numbered
130,000 people, almost 6000 in R&D (ITU, 1990). STET is controlled in turn
by the Institute for Industrial Reconstruction (IRI), the major national holding
company for all government-controlled economic enterprises. IRI, a vast com-
pany, was created in the 1930s by Mussolini to control the government's op-
Italy 247
erations in telecommunications, transportation, utilities, and several other fields.

STET was formed by IRI in the same year as the umbrella organization for its
telecommunications activities. In 1986, IRI controlled 84.6 percent of STET,
with the rest widely held, and there was a continuing discussion of permitting
greater participation by outsiders. In 1985 STET began to sell 30 percent of its
subsidiary SIP to the general public as part of a quiet trend toward privatiza-
tion. Whereas in 1984, private ownership in SIP had been 8.17 percent, two
years later, it had grown to 35.69 percent and was expected to reach 46 percent
in 1989, almost half of a privatization and not much less than that for British
Telecom (La Repubblica, 1987). The move was inspired more by IRI's efforts
to reduce its substantial losses than by telecommunications policy.
Customer premises equipment (CPE) markets were severely restricted, largely
because services were vertically integrated through STET, which favored its
equipment subsidiary, Italtel. In 1984, SIP provided 97.6 percent of all addi-
tional telephone sets and 45.4 percent of PBXs (R. Lauro, 1987, communica-
tion). SIP held an 88 percent market share in key telephone systems and a 60
percent share in facsimile markets. Maintenance of CPE, even that purchased
elsewhere, was a SIP or PTT responsibility; this situation changed after 1988,
except for first sets. Applicazioni Elettro Telefoniche (AET), an outside plant
installer whose core business was cable installation, was acquired by STET in
1989 (G. Pozzi, 1990, communication).
Liberalization began when it became clear that existing regulations were being
circumvented and were ultimately counterproductive. Equipment can be pur-
chased from outside vendors, but it has to meet Ministry specifications for type
approval, and SIP had to provide the interconnections. The 1987 national ten-
year plan liberalized modems and provided for the liberalization of the remain-
ing telephone and telex. Modem liberalization was carried out in 1989 only
after the European Commission filed an Article 90 complaint against the Italian
government.
SIP is the largest of the STET firms, with about 77,000 employees, 10,000
exchanges, and $4.8 billion in revenues. It had 20 million main lines and 29.2
million telephones at the end of 1988. For a long time, SIP was perceived as
an organization barely keeping pace with demand, and not as an innovator in
service development.
To improve this situation, SIP initiated a 500 billion-lire ($350 million) pro-
gram in 1987. As network modernization became a priority, SIP allocated 36,000
billion lire ($26 billion) over four years (1989-1992), 33 percent of which
would go to southern Italy (Benzoni, 1990).
Italcable, created in 1921 as an operator of submarine cables, has been part
of STET since 1965 and provides transoceanic services outside of Europe. Its
1989 turnover was 659 billion lire, and it employed 3500. From 1968 to 1985,
Italcable's telephone traffic grew from 3.7 to 226 million minutes, and telex
traffic grew from 3.0 to 71 million minutes. Italcable operates the message and
data packet switching network IRICON and provides access services to inter-
national databases.
Telespazio is Italy's satellite carrier and its signatory in Intelsat, Inmarsat,
248 Italy
and Eutelsat. Like the American Comsat, it is a carriers' carrier. It also pro-
vides remote sensing services. Its revenues in 1985 were $43 million, and it
employed about 640. One-third of Telespazio is held by the broadcast authority
RAI, which uses its services.
The Selenia-Elsag group was another major STET subsidiary. It is exten-
sively involved in manufacturing civilian and military electronic systems and
industrial automation and employs 13,000. Elsag won a major contract award
from the U.S. Postal Service to automate its system, an export achievement
that received much attention because of its size and because of the implicit
affirmation of Italy's high-technology capability. In 1989, STET sold its con-
trol in Selenia and Elsag.
SGS, until 1989, was a STET company in the microelectronics field and is
the major Italian manufacturer of integrated circuits. In 1985, its revenues were
$300 million and it employed almost 10,000. In 1987, it entered into a joint
venture for component manufacture with the French electronic giant Thomson.
But the venture did not flourish and STET sold its share in 1989. Other major
STET subsidiaries include Seat, a publishing company for telephone directories
and electronic yellow page directory service, and Consultel, a telecommunica-
tions consulting firm active in the developing world. STET's research & devel-
opment organization is CSELT, established in 1984.
Olivetti
Olivetti (formally Ing. C. Olivetti & Co.), headquartered in the northern city
of Ivrea, is perhaps Italy's most noted entrepreneurial firm in advanced elec-
tronics. For a time it was renowned for its equipment's design rather than its
marketplace success. This changed when Carlo De Benedetti assumed its lead-
ership. De Benedetti is to Italy's electronic sector what Silvio Berlusconi is to
its broadcasting: an empire builder of the first order. He managed the automo-
bile firm Fiat for a short while, until he had a falling-out with the Agnelli
family, which controls it. In 1978, De Benedetti acquired 15 percent of the
financially ailing Olivetti.
Within a few years Olivetti became Europe's largest office equipment maker
and the world's second largest producer of personal computers. In 1982, AT&T
acquired 25 percent of the company for $230 million, with an option to raise
its share to 40 percent in 1987. With its AT&T connection in the United States,
Olivetti aimed to become the world's number 2 company in professional desk-
top computers. But in 1988, AT&T, frustrated at its absence of control, de-
clined to increase its financial involvement and cut its computer order by 75
percent (Guyon and Colloy, 1988).
In 1988, the West German firm Volkswagen acquired a 5 percent stake in
Olivetti, for which Olivetti received the large German office equipment maker,
Adler-Triumph. Olivetti also bought interests in dozens of international high-
tech firms, many of them in the United States. In the United Kingdom, Olivetti
bought most of the small computer maker Acorn. It also joined with GeDa, a
computer services company, to form Olinet, providing database management,
Italy 249
data processing, and so forth. Thus, Olivetti had a highly international set of
owners, subsidiaries, and markets, all part of De Benedetti's strategy of making
it a global electronics company.
In the process, De Benedetti's own shares multiplied in value. He had ac-
quired 15 percent of Olivetti for $17 million. By 1986, this was worth half a
billion dollars. He also embarked on building a personal business empire dis
tinct from Olivetti. He acquired for a time Italy's largest food company, Bui-
toni, and several smaller food companies, an automobile component maker, a
tobacco equipment manufacturer, a share in the publishing house Mondadori
(where he fought with Silvio Berlusconi for control), and an investment bank.
He also established close relations and an ownership share with Pirelli, the
large tire manufacturer. In 1987, he embarked on a take-over bid for Belgium's
huge but stodgy conglomerate Societe Generale de Belgique (SGB). Though
established Belgian and French interests succeeded in beating him back, the
effort proved profitable for De Benedetti. In the early 1990s, Olivetti experi-
enced serious deficits again, raising questions about its long-term prospects as
an independent company.
Services
In data transmission, the structure of the system reflects the complexity of the
underlying carriers. The PT administration (not to be confused with the State
Administration for Telephone Services, ASST) provides slow-speed data trans-
mission services. Italcable provides intercontinental data transmission lines. SIP
supplies slow data transmission over the switched network, and higher rates
over leased circuits.
Telex emerged under restrictive regulation. A message could only be sent by
its originator; only PTT-supplied devices could be attached to the equipment;
and a deposit of $2,000 to $3000 plus annual fees of $2000 made use expen-
sive. Poor service and installation waits that exceeded two years compounded
the problems. Users circumvented some of these restrictions by forming asso-
ciations enabling members to use a community-owned telex.
Since i983, SIP has operated a circuit-switched data network, Rete Fonia-
Dati (RFD), using a technology by Telettra. In 1986, a new digital circuit-
switched network, CDN, went on line. Itapac, the packet-switched network
operated jointly by SIP and MPT, opened in 1986 for general use. In the first
year of full operation, it expected 5000 subscribers but only had 2700. Itapac's
main problem was operational, since each of the two partners controls separate
parts of the network. In addition, high tariffs and low-speed connections pre-
sented problems. Switches were Italtel-modifted Siemens equipment. After 1989,
Itapac was run by SIP alone.
Dedicated private network facilities are permitted only for services not pro-
vided by the public network and exist in a gray zone of tacit agreement among
large corporations. Resale is technically prohibited, but appears to exist.
Because of the administrative complexity of the Italian telephone system,
250 Italy
moves toward ISDN required more coordination than in most countries. In

1984, the PTT Ministry announced a national plan for telecommunications for
the next decade, addressing in particular network digitalization (MPT, 1984).
After 1989, only digital equipment was cut over. SIP also introduced the CQTT
Common Channel Signalling System 7 into the national network. In Milan, a
fiber local loop trial project capable of videotex data, TV, and teleconferencing
transmission was started in 1987.
Videotex began experimentally in 1980 under the name Videotel and has
operated on a regular basis since 1986. Users can access the system center in
Milan from anywhere in Italy with a local call. Videotel is based on the Italtel
and CEPT standards, and is concentrated on business subscribers.1 In 1991,
SIP had an installed base of 100,000 terminals and planned to become the
second largest videotex system after France.
SIP controlled cellular telephony as the only service provider of its kind in
Italy. SIP introduced a 900-MHz system and aimed at a digital system. But
increased demand brought pressures to end SIP's monopoly in mobile service.
To apply for licenses for digital mobile services, Media czar Berlusconi's Fin-
invest joined Fiat and the U.K.'s Racal; Olivetti formed the Omnitel consor-
tium with Bell Atlantic and Sweden's Televerket; and Pacific Telesis was part
of a third applicant group.
Telecommunications in Malta
Telephone service in Malta, an island nation between Italy and Tunisia, dates
back to 1882. Service is operated by the private Malta Telephone Company,
which was taken over by the government in 1933. The telephone network was
largely destroyed during World War II. In 1957, the manual system was up-
graded by STC to a stronger automated system. Later, Siemens exchanges were
added. Overseas telephone service to London was established in 1947, and
service to Rome began in 1952. A submarine cable to Italy soon became the
main connection.
22
Spain
Spain's telecommunications system is unusual. It is dominated by one semipri-

vate firm, Telefonica de Espana, which is strongly integrated vertically into
equipment manufacturing and horizontally into Latin American telecommuni-
cations. Telefonica has been assigned the role of a locomotive in Spain's rapid
industrialization, and its powerful and entrepeneurial presence is likely to in-
crease still further, as long as its primary mission of providing basic service
does not lag.
History
In 1884, a royal decree established a government monopoly over telephony.

Two years later, service concessions were awarded to private firms through an
auction of monopoly franchises for each city. Contracts were awarded to appli-
cants who promised to provide the state with the highest percentage of gross
receipts of at least 10 percent; bids for some cities far exceeded that percentage.
Thirty-five local exchanges were established under the system, but problems
soon became apparent; rate reductions could not be instituted, and the private
companies were unwilling to extend service to rural and isolated areas. As a
result, the government decided in 1890 to drop the monopoly system and to
allow any interested parties to establish exchanges, while absolving the existing
licensees of their revenue-sharing obligations. Commercial long-distance lines
were also left to private firms. No competition emerged, however, since com-
panies were unwilling or unable financially to enter into each others' territories.
Thus, de facto unrestricted monopolies remained in existence and were insti-
tutionalized in 1903. However, service remained unsatisfactory, and in 1924,
the government granted a major concession to operate all of Spanish telephony
to the Companfa Telefonica National de Espana S.A. (CTNE), a firm that had
only recently been formed by the American firm International Telephone and
Telegraph (ITT) and private Spanish investors. Later that year, ITT acquired
all the company's shares. This purchase was the first major step taken by ITT
in its efforts to become a major international presence in telecommunications.
ITT's founders, the Behn brothers, had little capital and no manufacturing sup-
port behind them. Despite its grandiose name, the company was a midget in
251
252 Spain
comparison to AT&T. The Behns now sought a manufacturing base; at the

same time, AT&T happened to be looking to unload its European manufactur-
ing operations because of U.S. domestic pressure. ITT bought them, and ac-
quired a major presence in international telecommunications. ITT rapidly be-
came Spain's dominant service and equipment firm. In 1945, however, it was
forced to relinquish ownership of its network operation CTNE to the Spanish
government, which then directly or indirectly controlled about half of its shares.
In 1986, ITT's equipment subsidiaries became part of the French-dominated
Alcatel. Also in 1986, CTNE's name was formally shortened into Telefonica
de Espana.
Organization
Overall regulation of the Spanish telecommunications sector lies with the Junta
Nacional de Telecomunicaciones (National Telecommunications Board), an in-
terministerial commission answering to the Ministry of Transport, Tourism,
and Communications. The board is responsible for assigning radio frequencies,
setting investment guidelines, and maintaining relations with foreign adminis-
trations.
Within the Ministry of Transport, Tourism, and Communications, two sep-
arate agencies have responsibility for telecommunications matters: the Direc-
cion General de Correos y Telecomunicaciones (Directorate General of Post
and Telecommunications, or DGCyT), which provides postal service as well
as telegraph, telex, facsimile, electronic mail, and message switching; and Di-
reccion General de Electronica y Informatica (Directorate General for Electron-
ics and Informatics, or DGEI), which is in charge of promoting the develop-
ment of a domestic electronics industry.
An independent communications network is run by the broadcast authority
RTVE, which operates its own transmitters and relay stations (Lopez-Escobar,
1985). Most of telephone operations, however, are the domain of Telefonica,
which was given monopoly status by government decree in 1970. Television
and limited voice and data services have also been offered by Retevision since
1989. Holding about 36 percent of shares in Telefonica, the Spanish govern-
ment acts as the majority holder of the company and, importantly, appoints its
chairman. The remainder of the firm's stock is held by more than 750,000
private shareholders, including a number of foreign interestsmostly Ameri-
can, British, and German institutional investors. American investors held about
17 percent of the company's stock in 1990.
The company's stock was a favorite of small investors; in 1985, CTNE ac-
counted for 17.2 percent of the entire Spanish stock market's capitalization! In
comparison, the ten largest stocks on the New York Stock Exchange account
for 15 percent (The Economist, 1985). Because Telefonica's capital needs are
so great relative to the size of the Spanish economy as a whole, it had problems
in raising adequate funds to finance its investments, and in 1985 it began to
offer its stock on foreign exchanges. As a "strategic sector" firm, the Spanish
Spain 253
government refused to raise the ceiling on the amount of Telefonica stock that
could be held by foreign interests from 25 percent. A large portion of the
state's holdings in Telefonica are administered by the central bank of Spain and
the Institute Nacional de Industria (INI), a public holding authority that was
established in 1941 modeled after Mussolini's IRI. By 1982, INI firms had
more than 200,000 employees, about 7 percent of the country's labor force. Its
losses, however, amounted to $1 billion. Telefonica's semiprivate, semigovern-
mental structure has made the privatization versus nationalization debate less
pressing than that in most European countries.
Telefonica's presence in the Spanish economy is colossal and is reinforced
by its vertical integration. With 71,000 employees, Telefonica is Spain's larg-
est company. It holds stock in numerous firms, including Amper-Elasa, SIN-
TEL, TEFISA, TIDSA, and Cetesa (de Moragas et al., 1987). It accounts for
nearly 4 percent of gross national capital formation in Spain and 2.7 percent of
the gross added value in the service industry. Telefonica has one of Europe's
highest rates of telecommunications investment as a share of gross fixed capital
formation (3.9 percent) (ITU, 1990).
In 1985, the government proposed legislation to reorganize and centralize
the telecommunications sector, with the specific objectives of creating an inte-
grated network, defining telecommunications services in a consistent manner,
and coordinating CTNE, RTVE (broadcast transmissions), and DGTyT (tele-
graph, telex). But the proposed expansion of governmental influence was crit-
icized by large users as running counter to the trend of reducing the state's role
in telecommunications.
In 1986, three narrower bills were substituted for the more comprehensive
effort, providing for the reorganization of telecommunications, private tele-
vision, and the postal service. The telecommunications bill provided for keep-
ing CTNE's monopoly as a telephone carrier intact and for the liberalization of
the terminal equipment and VAN markets as well as for the transfer of respon-
sibility for equipment approval from the telephone authority to the Ministry of
Industry (White, 1986). The liberalization, intended to proceed gradually from
1988 to 1992, began with three corporations requesting Ministry of Industry
approval. Under this plan, Telefonica's carrier monopoly was extended for an-
other thirty years.
Services
The penetration of telephone service in Spain has lagged behind that in other
European countries. In 1990, there were 30.4 telephone lines per 100 people
(ITU, 1990). In the same year, however, applications for basic service in-
creased 51 percent over the previous year, and the growth of usage of existing
lines increased 5 percent. The wait for a new telephone connection was typi-
cally still over half a year, although this figure varied, depending on the region.
In metropolitan areas service was faster, but in rural areas there were longer
waits and greater costs of connection.
254 Spain
Luis Solana, a former banker, took over management of Telefonica in 1982

and led to a major expansion into worldwide markets. Eventually, this led to
criticism that the globalization of Telefonica's activities was at the expense of
the country. In 1989, there were 1.4 million applications for new lines, 535,000
of which went unfilled. This led business groups like the Madrid Chamber of
Commerce and the Catalan Industrialists Association to complain that telephone
failures and high costs hurt Spanish business (Purton, 1990). Public dissatisfac-
tion with Telefonica's failure to keep pace with demand for new connections
then forced Solona to resign in 1988. He was replaced by Candido Velazquez-
Gaztelu. Telefonica increased network investment by 60 percent from 1988 to
1989, reaching $5.8 billion in 1990, and it allocated another $35 billion through
1994. As a result, Telefonica reduced in 1989 the length of its waiting list for
the first time in four years. In 1990 Telefonica lost its monopoly over extension
telephone terminals, and by 1992 also over the first set, in accordance with EC
rules.
The telegraph administration, DGCyT, originally provided the bulk of data
services but in time collided with the telephone carrier. DGCyT operates a telex
network and leased telegraph circuits for slow transmission rates. But in 1970,
the government granted CTNE the authority to provide data transmission ser-
vices, electronic mail, facsimile, and videotex. As a result, it began to infringe
on its rival's business, for example, through the successful public packet-switched
data transmission network Iberpac, which originated as the first European packet-
switched network. Data transmissions comprise 9 percent of Telefonica's total
income.
Limited "Ibertex" videotex operations began during the 1982 World Cup
soccer competition. Attempts were made to involve the Spanish-speaking coun-
tries of Latin America in the project.
Telefonica is remarkably active internationally. Latin America became the
site of Telefonica expansion, particularly in Argentina and Chile. In 1988, the
Argentine government announced that, after forty-two years of state ownership,
Argentina Empresa Nacional de Telefonica (Entel) would be privatized. In 1990,
Telefonica outbid six European and American competitors in 1990 for a con-
trolling interest of Entel's southern operations. The winning package offered
$114 million in cash, $2.18 billion in debt, and $540 million in forgiven inter-
est for a 60 percent share of the new company, to be called Telcosur. The
Telefonica consortium consisted of an investment company including Citicorp
(20 percent), Telefonica (10 percent), and the Argentina's Techint Group (15
percent). Telefonica, the operator, agreed to install 610,000 phone lines and
12,900 public phones by 1996, at a cost of $1 billion The new entity was
named Telefonica Argentina, and it planned to invest $400 million a year.
When Chile's telephone system was privatized, Telefonica initially lost to
Australian financier Alan Bond in the bidding for a portion of Compania de
Telefonos de Chile (CTC). But in 1990, Bond's financial difficulties forced
him to sell his 43.7 percent stake to Telefonica for $388.5 million. However,
soon after the purchase, Chile's antimonopoly commission ruled that Telefon-
Spain 255
ica could not own part of both CTC and ENTEL, Chile's long-distance carrier,
in which it also held a 20 percent stake. Telefonica appealed the decision to
Chile's Comision Resolutiva.
Telefonica was also a bidder for a part of Mexico's telephone system; it also
purchased a stake in the U.S. firm Infonet (5 percent), along with other Euro-
pean telecommunications organizations, as well as in Geostar (3 percent) and
in Mercury's PCN subsidiary (10 percent). It also gained a 40 percent share in
1991 in Venezuela's privatized system, in a consortium led by GTE.
Equipment
Telefonica has a direct stake in determining which firms are allowed access to
the equipment market, since it owns a majority of the stock of a dozen equip-
ment firms and holds minority interests in several others. Telefonica controlled
21 percent of Alcatel Standard Electrica; 20 percent of Citesa, another Alcatel
subsidiary; 10 percent of Telettra, the Spanish subsidiary of the Italian telecom-
munications manufacturer of the same name, then owned by Fiat; and 7.6 per-
cent of Amper, a Spanish telephone set manufacturer (Telefonica, 1989). Thus,
until the 1986 reform giving equipment approval power to the Ministry of In-
dustry, the manufacturer not only had to meet both CCITT and CTNE technical
standards, but also often needed to cultivate a good relationship with Telefon-
ica, their own competitor, including the disclosure of information, in order to
secure approval for the sale of their products.
In the past, CTNE gave clearance only to its own or to Spanish-made equip-
ment. In 1981, foreign suppliers provided only 13 percent of the company's
equipment (MarTech, 1983). However, the 1987 regulatory statute ley de or-
denacion de las telecomunicaciones (LOT) partially opened CPE markets. But
it was challenged by the European Commission since it still enabled control
over terminal equipment by the vertically integrated Telefonica. Telefonica's
practices were also subject to two 1990 investigations, one by the Spanish anti-
monopolies board over bundling of connection to the network with purchase of
more expensive equipment, and one by the Spanish consumer's advocate over
a tariff structure that decreased the advantage of off-peak rates.
The Socialist government that came to power in 1982 increased Telefonica's
importance in the equipment field by giving it a central role in its ambitious
high-technology plans. Telefonica, RTVE, and the Ministry of Defense were
each required to establish four-year plans of investment and development. When
Standard Electric threatened to eliminate 6000 manufacturing jobs in 1985, the
government pressured Telefonica to increase its purchases of that company's
equipment.
In 1991, Telettra was acquired by France's Alcatel. This lead the European
Commission to require that Telefonica divest itself of its shares in Alcatel and
Telettra, as well as to open its purchasing policy.
Luis Solana, Telefonica's president until 1988, actively sought to spur the
256 Spain
development of Spanish telecommunications by entering into still more joint

ventures with Spanish and foreign companies. Pacific Telesis, the American
regional bell holding company, was invited to design a Telefonica telecommu-
nications research center in Madrid. The main partner of Spain's Intelsa, Erics-
son, also financed the establishment of a large research center in Madrid, and
in 1990 won 43 percent of a major equipment contract through Intelsa. In 1990,
half of the digital exchanges in the Spanish public network were Ericsson's
AXE exchanges (Martinez, 1990). Telefonica entered a large semiconductor
venture with AT&T in 1984. Together, the two firms built a plant in Spain that
produces microcomponents, largely for export. Of the project's $65 million
direct investment, AT&T put up 80 percent and Telefonica provided the re-
mainder. In addition, the Spanish government provided subsidies totaling $60
million and credits worth another $75 million. The transaction made Spain the
major European manufacturing base for AT&T. In 1987, this collaboration was
expanded as Telefonica contracted with AT&T/Philips for 20,000 lines of an
"intelligent" overlay for the Spanish network (Williamson, 1987). In 1990,
Telefonica joined AT&T's Europe-wide equipment efforts by obtaining a 6
percent share in AT&T's network systems (Philips had 15 percent; Italtel, 20
percent; and AT&T, 53 percent), in return for its shares in the joint microelec-
tronics venture with AT&T in Spain.
Telefonica is involved in other joint ventures: In cooperation with Corning
Glass, it entered the market for optical fiber; it also engaged in a $200 million
computer production venture with Fujitsu. Given its history of close ITT con-
nections, Telefonica also tried, unsuccessfully, to obtain a 10 percent stake in
the CGE/ITT Alcatel venture.
In 1990, jointly with Amper, Telefonica won a contract to produce between
500,000 and 1 million telephones in the Soviet Union.
Piher Semiconductor, a firm headquartered in Barcelona, is the primary Spanish
component producer, outside of Telefonica's ventures. The firm, part of the
state-owned INI group, was fined $1 million in 1985 by an American court,
given five years probation, and denied all American export privileges for ten
years after it pleaded guilty to illegally diverting more than $2 million worth
of semiconductor manufacturing and testing equipment to Cuba and the Soviet
Union.
Telecommunications in Gibraltar
In Gibraltar, the British possession at the tip of Spain, telephone service first
operated in 1892 for civilian service, with a parallel military system. In 1926,
the city council established an automatic Strowger exchange service. In 1969,
the operation of the telephone service was taken over by the government of
Gibraltar, and in 1973 a crossbar exchange was opened.
International service was opened to Spain in 1927. This service was broken
off in 1969 for several years because of the dispute over control of Gibraltar.
Spain 257
In 1989, the Gibraltar government chose NYNEX to enter into a fifty-fifty

partnership to run its domestic telecommunications network. NYNEX will in-
vest $10 million over five years to deploy an advanced intelligent network,
with a fiber backbone, to serve Gibraltar's needs as a burgeoning financial
center.
23
Portugal
The telecommunications sector in Portugal reflects the country's relative pov-

erty and its struggle to catch up with the rest of Europe in industry and ad-
vanced services during a period of domestic turmoil. Although it has made
much progress, it has not yet reached a level where the policy issues that affect
most other Western European countries are part of the Portuguese agenda. Por-
tugal is, however, affected by these issues in its procurement decisions.
Portuguese telecommunications are the charge of three organizations, two of
which, Correios e Telecommunicacoes de Portugal (CTT) (renamed Telecom
Portugal) and Telefones de Lisboa e Porto (TLP), operate according to the
directives of a common General Board; the third, Compania Portuguesa Radio
Marconi (CPRM), derives from the Anglo-Portuguese telephone company. Tele-
com Portugal's service area covers most of the country; it is also responsible
for international telephony within Europe and for the Portuguese telex network.
TLP operates telephone service and networks within a 30 kilometer radius of
Lisbon and a 15 kilometer radius of Porto, the two largest Portuguese cities.
The nominally private CPRM provides satellite links and submarine cable ser-
vice, including connections to the Azores and Madeira. Its franchise covers all
public overseas international telecommunications, as well as the concession for
maritime mobile service in the whole Portuguese territory.
Telephone and telex service to the European countries are operated by Tele-
com Portugal in agreement with CPRM, using the latter's transmission facili-
ties. Although it is technically an independent operation with its own budget,
51 percent of CPRM shares are held by IPE, a holding company for state
participation in private firms. The chairman of the CPRM board and two of its
members are government appointees. For years, relations between the two ex-
change companies and ITT's Portuguese manufacturing subsidiary Standard
Electrica were extraordinarily close.
All three telephone enterprises report to the Ministry of Public Works, Trans-
port and Communications, which also supervises the postal service. Govern-
ment control over telecommunications is exercised through appointments to the
General Board, as well as through legislation on tariffs, the budget, borrowing,
and other financial matters. A merger of the three organizations was expected.
The Portuguese telecommunications sector, although relatively underdevel-
oped by European standards, began to receive increasing attention from the
government in the 1980s, as reflected by raises in budgetary allocations. From
258
Portugal 259
1978 to 1987, the investment per line rate was the third lowest among OECD
nations. However, 1988 figures show a significant shift in priorities, placing
Portugal first in the European Community in terms of telecommunications in-
vestment as a percentage of fixed capital formation (ITU, 1990).
In addition, the separation of the telecommunications and postal authorities
and the merger of the three telephone service companies became topics of dis-
cussion. Telecom Portugal and TLP agreed in principle to a statute that pro-
vided for their merger; they share administrative and fiscal counsel, as well as
the same chairman as part of the General Board. Legislation passed in 1989
brought Portugal more in line with E.G. directives, (de la Cal, 1990, p. 13).
The law also created a Communications Institute of Portugal to coordinate de-
velopment and liberalization.
To increase capital resources and to prepare for the single European market
in 1992, the center-right government which came to power in 1987 announced
a process of "partial privatization" to begin after 1990. The government would
keep 51 percent of the three companies, with the remaining shares sold to
private investors. Foreign holdings were limited to 10 percent, and 20 percent
are owned by small investors and employees.
A 1990 law created a holding company for the telecommunications opera-
tors, named Telecom Portugal. The state retained control over 51 percent of
TLP. The law also created the Institute das Communicacoes de Portugal to
coordinate development and liberalization; regulatory powers remained with the
Ministry of Communications.
Services
Telephone density in Portugal stood at eighteen lines per 100 persons in 1988,
the lowest in Western Europe. The government's goal is to have 32 connections
per 100 population by 1993. In 1988, the average wait for a telephone was 9.4
months, down from almost three years. The waiting list was still 15 percent of
the total of main lines (ITU, 1990).
Data transmission service is available on analog leased lines over the public
switched telephone network at rates of up to 2.4 kbps. Portugal restricted the
creation of private networks based on the use of leased circuits, and customers
were consequently not allowed to install their own data multiplexers (Eurodata
Foundation Yearbook, 1983). This provision was lifted in 1986. Telpac, a packet-
switched network featuring Northern Telecom equipment, was opened in 1984
under the control of Transdata, an organization set up by CTT and TLP. The
expansion of data usage was curtailed by the limitations of the Portuguese net-
work. A 1990 survey by the European Association of Information Services
(EUSIDIC) found that 64 percent of international data test calls made in 1990
from Portugal failed at three times the average rate for the fourteen countries
studied, which were themselves low in performance (EUSIDIC, 1990).
Portuguese international rates are high; calls placed from Portugal to other
260 Portugal
European nations tend to cost twice as much as calls traveling in the opposite
direction (Transnational Data and Communications Report, 1988).
Given the priority of developing basic services, Portugal has been slower
than other European countries to introduce cellular mobile telephony, teletext,
and videotex. In 1984, public telefax service was introduced through the post
office. In 1987, Portugal selected Germany's C-Netz standard for cellular radio
and contracted Siemens to supply the infrastructure (Purton, 1987). By 1990,
cellular phone service was available to 90 percent of Portugal's population in
30 percent of the country, but subscription levels were low.
Equipment
The two local operating organizations held a monopoly over the first telephone
sets. Since 1983, it has been legal to obtain second sets from other sources.
There is no government monopoly over the PBX or modem market, although
they are also supplied by operating companies.
There are two main telecommunications manufacturers in Portugal: Standard
Electrica, a subsidiary of ITT and subsequently of Alcatel, which has had a
presence in the country since 1932, and Centrel, a wholly Portuguese company
(ITT, 1983). Centrel goes back to the British Plessey, at one time the second
leading telecommunications equipment supplier in Portugal, which left the country
in 1979 and sold its Portuguese factory for 1 to the then unknown local firm,
Centrel.
Standard Electrica is the country's largest communications equipment and
electronics manufacturer. According to its own estimates, the company has
trained about 60 percent of Portugal's telecommunications engineers, a fact that
reflects the extremely close relationship that it maintains with the operating
companies. Standard Electrica supplies 50 percent of Portugal's telecommuni-
cations equipment, producing two crossbar exchange switches (one of which is
Portuguese developed) as well as System 12 switches. It also provides PBXs
(holding a 90 percent market share), telephone sets, television sets (60 percent
of all black-and-white sets sold in the country), and transmission equipment.
Its share of Portuguese telex switch sales rose to 35 percent after Siemens's
thirty-year hold over supply was ended.
Products by Centrel, the other manufacturer, include systems, repeaters, public
switches of the crossbar and Strowger types, handsets, coin box telephones,
and transmission equipment. Centrel's public switch division, using Siemens
technology, accounted for more than half of the company's total revenue. Al-
though the bulk of the firm's output is purchased within Portugal, it also ex-
ports.
The allocation of the larger Portuguese procurement orders for central elec-
tronic switches was a multiyear story of intrigue on many levels. Throughout
much of the organization's history, the traditional Portuguese supplier to CTT/
TLP had been ITT's Standard Electrica. In March of 1985, however, a coali-
tion government opened up the switch contract bidding to seven different com-
Portugal 261
panics to demonstrate openness to the European Economic Community. These

six additional firms were Ericsson, which had been a supplier to Portugal's
former African colonies, AT&T/Phillips, Northern Telecom, Siemens, PSessey,
and Alcatel/Thomson. At the time of the awarding of the contract, Centrel had
just entered into a joint venture with Siemens for the production of switches
developed by the German firm, with some production to be undertaken in Por-
tugal. This gave Siemens an inside track in the competition for the switch
contract because it was a European company and could assure Portuguese jobs.
The lobbying for the contract became highly political. Alcatel's interests were
personally pushed by French President Mitterrand, who was a friend of Portu-
guese Prime Minister Mario Scares. German politicians involved themselves in
similar lobbying for Siemens, and ITT's interests were advanced by the tele-
phone operating companies CTT/TLP. The government took away the tele-
phone authorities' power to allocate the order, and instead gave it to a more
politically oriented interministerial commission. Ericsson, Northern Telecom,
AT&T/Phillips, and Plessey were next excluded from bidding; surprisingly, so
was ITT. The commission allocated 55 percent of the order to Siemens and 35
percent to Alcatel, despite the latter's lack of Portuguese production facilities,
a disadvantage that had been stressed throughout the allocation process. The
remaining 10 percent was allocated in a complex fashion among the two vic-
torious companies, which were forced to compete for the last slice by offering
technology transfer, guarantees of local production content, and software pro-
duction (Smith, 1986).
The selection of Alcatel, made just before the election of 1987, raised a
major controversy because, in addition to having no Portuguese production fa-
cilities at the time, the company had no history of working with the country's
telephone organizations. To deal with this problem, as well as with the Stan-
dard Electrica employees who felt abandoned because their firm had been left
out of the running entirely, the government ordered the ITT subsidiary to turn
its physical plant and 20 percent of its equity over to Alcatel so that the French
company could use it for the production of switches. This was considered a
blatantly political act, and allegations of impropriety followed. After the elec-
tion, the Socialists were replaced by a more conservative Social Democratic
government, which immediately reopened the case. The Siemens/Centrel con-
tract was not controversial and was allowed to stand, but the Alcatel share was
opened for bidding, with the operating companies regaining their traditional
leverage in the decision process. In the end, Standard Electrica received a 45
percent share. Alcatel was out in the cold, or so it seemed. But within a short
time, it pulled a major coup when its parent company, CGE, acquired control-
ling interests in most of ITT's international telecommunications operations, al-
lowing Alcatel to crack the Portuguese market after all.
24
Israel
Israel is closely related to Western Europe in its telecommunications system.

Its dynamics of institutional change are similar, and are therefore included in
this study.
For a long time, Israel's telecommunications system was that of the classic
postal-industrial complex, with the state PTT nurturing the electronics industry
in close collaboration with the trade unions. The government, meanwhile, re-
garded telecommunications as a source of revenue. This system led to dissat-
isfaction because the expansion and quality of the network were inadequate for
the rapidly modernizing society. In 1982, it was decided to spin off telecom-
munications from the Ministry of Communications and to establish the inde-
pendent operating company Bezeq. Bezeq (also often spelled Bezek) is still a
state monopoly, albeit a regulated and part-private one, but its service orienta-
tion and performance have greatly improved.
The telephone system dates back to the years of the Turkish and British
governmental rule over the country. After the founding of Israel in 1948, the
telephone network greatly increased in penetration and technical capability and
was enhanced by the development of a domestic electronics industry. In 1948,
there were only 11,000 lines and usage was 45 million meter pulses; in 1987,
there were 1.3 million lines, and usage was almost 9 billion meter pulses. In
the same year, 90 percent of households in the major cities had a telephone.
This is despite the fact that Israeli telephone service is among the most expen-
sive in the world because of the very high up-front connection charge and the
significant government levies on telephone services as part of attempts to raise
revenues.
An important change was the transfer, in 1984, of the telecommunications
operations from the Ministry of Communications to the separate but state-owned
Bezeq Israel Telecommunications Corporation Ltd.
The governmental service had been unable to keep up with the pace of the
emerging sophisticated business environment in Israel, and this had led to sub-
stantial unhappiness among the public and in the business community. The
backlog could have been reduced, of course, by substantial investment pro-
grams, but this was constrained by the need to coordinate investments within
general governmental priorities. Under governmental control, the telephone system
was not able to move as rapidly as the public wished. Greater managerial in-
262
Israel 263
dependence was one reason that the telephone organization wanted to reduce
direct governmental operational supervision. Another factor was that many of
the telephone administration employees sought to extricate themselves from
civil service salaries, particularly when they had technical skills and prestigious
engineering degrees. Telecommunications were also subsidizing postal ser-
vices, a bottomless pit since postal rates were kept low for political reasons.
In the 1960s, the Dinstein Committee was formed to examine the status of
public utilities. It recommended splitting these operations from government
control, but nothing came of it. In 1970-1972, the Herzog Commission (chaired
by the future president of the state, Chaim Herzog) was appointed by the min-
ister of communications (and later Prime Minister) Shimon Peres, to report on
the status of telecommunications.
In 1978, under a newly created right-of-center government, an expert con-
sulting body was created that consisted primarily of academics from the fields
of business administration, economics, law, and engineering. Their task was to
prepare a concrete proposal to transfer the governmental function to a new
company. The proposal was completed in 1979, with Yitzhak Modai, the then
minister of telecommunications providing strong support. The Bezeq law was
prepared and passed by the Knesset in 1982. A prime bone of contention was
the status of employees. Without union agreement the plan was politically in-
feasible. In 1982 and 1983, intensive negotiations took place. The labor unions
did not actively oppose the reform, once they received a virtual veto right over
certain aspects of the organization and had them written into the law. Within
the Labor party, Gad Yaacobi, a future minister of communications, was also
supportive. The strongest opposition came from the bureaucracy, particularly
the Ministry of Finance, which feared the loss of substantial revenue contribu-
tions from the telephone service. This reflected the main problem of the past:
the perception of telecommunications as a cash cow rather than as a public
service or infrastructure.
Two collective bargaining agreements preserved the various rights and se-
niorities that telecommunication employees had accumulated. Another agree-
ment with the government settled the transfer of the assets and service obliga-
tions. On January 31, 1984, the reorganization became effective, including the
transfer of 8000 employees from government service to company service. The
government changed from a service provider to a service receiver, and the
Communications Ministry was transformed from an operating department to a
policymaking, regulatory, and supervisory department.
The reorganization of the telephone system was undertaken under Minister
of Telecommunications Mordechai Zippori. After the elections of 1984, Pro-
fessor Amnon Rubinstein, a prominent reformist member of parliament, be-
came minister and established early agency direction for policy planning, reg-
ulation, development, licensing, spectrum management, and equipment approval.
The Ministry of Communications regulates Bezeq's tariffs by approving rate
increases and general changes in rate structure, in consultation with the finance
minister and the economic committee of the Knesset. Rubinstein considered the
264 Israel
feasibility of a partial privatization of Bezeq, and advocated full equipment

liberalization, including the first telephone set. These proposals, however, were
not instituted.
The significance of Bezeq's status outside the government must be viewed
in its historical context. In Israel, more than in most European countries, a
large number of companies, including those providing infrastructure services in
all parts of the economy, are either controlled directly through the government
or indirectly through the Histadrut trade union federation and other semipublic
bodies. Public ownership of the key means of production in Israelby govern-
ment or the trade unionshad been for a long time an article of faith in the
ideology of the state, even more than in Western Europe. In its purest reflec-
tion, this belief led to the collectivist kibbutz settlement. The partial loosening
of governmental control over the telephone system was therefore a major step.
Through the change, the relationship between the company and the public
was transformed to some degree. Bezeq attempted to be more responsive to the
public, although it is still a monopoly service provider. Bezeq organized itself
into six divisions: Finance; Economics and Logistics; Human Resources; En-
gineering; Marketing and Consumer Services; and Systems and Computers. It
divided Israel into twenty-four telecommunications districts within four regions
to foster local service (TRC, 1990). The new company's primary goal was to
deal with an abominable waiting list for service, as long as 260,000 in a coun-
try of less than 1.5 million households. Where in the last year of the old min-
istry arrangement, $110 million was invested, Bezeq spent $250 million in
1985, $300 million in 1987 (Louisson, 1987), and $412 million in 1989 (Hai,
1990).
In 1986-1987, 22,000 trunk lines were added, compared to about 5000 in
1983-1984. Bezeq established a five-year plan during which time it planned to
install 600,000 new lines, thereby increasing the system by 50 percent and
doubling the annual installations. Given Israel's population, 550,000 new lines
in five years is the equivalent of about 25 million new telephone lines installed
in a five-year period in the United States. Bezeq was active in the Arab villages
of Israel, where the demand for telephones had increased rapidly. By 1988,
almost 1.5 million main lines were installed. The number of new public tele-
phones increased by 65 percent during four years, and the directory information
service was fully computerized (Hai, 1990).
By 1988, Bezeq claimed a productivity increase of 100 percent for its 9000
employees; workers per 100 lines were reduced from 7.6 in 1984 to 6.3 in
1988. The backlog of installation requests was reduced to 75,000 in 1989 (TRC,
1990). The annual growth rate of telephone lines was 10 to 12 percent, almost
seven times higher than that in the United States. In terms of telephone density,
Israel became comparable to Belgium or Italy, with thirty-two main lines per
100 population in 1989 (ITU, 1990).
Private leased lines rose from 1350 in 1981 to 4440 in 1984 and 10,500 in
1988. Bezeq also implemented the ISRANET public packet-switching network,
operating at speeds of up to 1.2 kbps, with modems supplied by either Bezeq
or the customer and approved by the Ministry of Communications. The net-
Israel 263
work connects to similar networks in other countries, but at a higher usage cost
than similar networks. Also, the initial cost of connection is high. The high-
speed data network SIFRANET, designed to serve large users, began service
in 1988, with the government signing on as the first customer. Demand for
data services consistently outpaced supply. One study found that the 16 to 18
percent annual growth rate for leased data circuits would not keep pace with
the 45 percent annual growth for VANS (Bainerman, 1989). A government-
appointed panel, the Fogel Commission, recommended that rates be reduced.
This was done in 1990, as part of a tariff rebalancing.
Bezeq also provides radio paging, alarm systems, and, in a joint venture
with Motorola and Tadiran, cellular telephone service, where it holds 35 per-
cent of the market. Competitors in paging services are Iturit, Page-call, and
Beeper.
By law, Bezeq is the only company to receive a "general permit" for tele-
communications services, although there are no restrictions on the minister's
granting special licenses to others for particular services. A wide-ranging li-
cense was granted to the Postal Authority in 1990. Other organizations provide
telex service, data transmission, and remote data processing. The leading com-
pany is Aurec, which supplies electronic mail services as a node of Dialcom in
addition to other value-added services in partnership with Bezeq. Other com-
panies provide wireless communications and wireless telex or licenses in the
international field (Israel Statutes, 1982). The license fee is 11 percent of rev-
enues.
Rival international telex service is provided by the Postal Authority, Ram
Telex, and Cosmic Telex, all of which use Bezeq's telephone and data trans-
mission lines to capture a combined market share of about 25 percent. Bezeq
competes with these firms, subject to its own tariffs. The large labor union-
controlled industrial group Koor owns a domestic value-added network under
the name of Koornet, offering telefax, telex, and data transmission services.
Koornet also competes with Aurec and Kav Manche in electronic data inter-
change (EDI) and VANS. Two popular services are telefax and BITNET, a
heavily used standard mode of communications both domestically and internation-
ally. Israel participates in three Mediterranean submarine cables: Tel Aviv--
Marseilles, which has operated since 1968; also Tel Aviv-Rome-Marseilles,
in operation since 1975; and the Eastern Mediterranean Fiber Optic Cable Sys-
tem, in operation since 1990.
In 1986-1987, Bezeq's contribution to the government was $169 million,
out of a total revenue of $738 million for royalties, interest, and taxes. In the
same year, Bezeq's rate of return was only 2.5 percent because of a general
governmental price freeze in 1985 (S. Hai, 1988, communication). This made
even a partial privatization unattractive to potential investors. In 1988, the Min-
istry of Communications approved a tariff increase of 8 percent. That year
Bezeq's revenues grew to $932 million, but it returned only $6.5 million in
profits (ITU, 1990). Bezeq was also required to transfer $190 million to the
government. In 1990, proposals to privatize Bezeq and sell 20 percent of the
company to a foreign owner were made. Fifteen percent of Bezeq's stock
266 Israel
was offered to the general public. A ceiling of 25 percent was set, but soon
raised, with an anticipated additional nine percent going to foreign shareholders
and fourteen percent to strategic investors. A government commission also rec-
ommended partial opening to competition.
The major manufacturer of telecommunications equipment in Israel is Tel-
rad, wholly owned by Israel's largest industrial group, Koor Industries, which
is in turn owned by the Histadrut labor union federation. Telrad was established
in 1951 and became a major developer of the civilian and defense communica-
tions network. By 1990, the company had installed more than 165,000 civilian
telephone lines per year and had supplied more than 750,000 lines of electronic
digital equipment for public systems. Telrad also supplies secure communica-
tion systems to the Israel Defense Forces (R. Sitter, 1990, communication).
Telrad's products include electronic switches, key telephones, terminal
equipment, modems, transmission equipment multiplexers, and PBXs. Of the
telephone handsets in the country, more than 90 percent are manufactured by
Telrad. In 1990, the company employed 1200 people and sales totaled $150
million, about a fifth of which were from exports.
Another Israeli telecommunications producer is Tadiran, which has a broad-
based product line and is one of the world's leading manufacturers of lithium
batteries. Tadiran has 12,000 employees and is owned by the labor unions'
Koor. Like Telrad, Tadiran and its subsidiary Tadiran Elisra supply systems
and equipment to both defense and commercial markets. Among its products
are tactical radios and data terminals, communication devices and radio tele-
phones. Tadiran also manufactures and installs public telephone exchanges (Al-
catel's System 12) and PBXs. Among industrial firms, Tadiran has Israel's
largest research and development effort.
Elron Electronics, a large electronics concern, owns several subsidiaries, among
them Elbit Computers, Elscint (for computer-based diagnostics), and Fibronics
(for fiber-optic systems). Other Israeli telecommunications firms are ECI Tele-
com, Electric Cable, Zion Cables, and Paliadent. Because of the rapid growth
of demand for data services, numerous small firms arose to challenge the "Big
Three" Telrad, Tadiran, and ECI. These firmsAdacom, Rad Data, Bitcom,
Efrat Future Technologies, Keren, Teledata, IAI, Fiobronics, Cvalim, Motor-
olahave filled niche markets, often teaming up with American firms.
In 1990, Raphael Pinhasi of the ultra-religious Shas party became commu-
nications minister, leading to a replacement in Bezeq's leadership. Chairman
Yoram Alster was succeeded by Akiva Atoun, a founder of Ofek Computers
and Software Research and Development. Atoun's father was a member of the
Council of Torah Sages, which oversees the Shas party, and this raised ques-
tions over the minister's ability to exercise regulatory functions over Bezeq.
In 1991, the so-called Boaz Committee recommended further liberalization
in equipment approval, international telephony, data communications, VANs,
and mobile service.
25
Turkey
Telephone service began in Turkey in 1881 with lines connecting various post
offices, government buildings, and banking branches. In 1909, a more com-
plete network was created in Istanbul. In 1913, three government exchanges
were established, followed by an exchange operated by a foreign company.
International telephony was established during World War I, with service to
Germany, Turkey's ally. After the war, international service was interrupted
until 1931, when service to Sofia, Bulgaria, was reopened.
In 1926, the first automatic exchange was opened by the state in Ankara. In
Istanbul, foreign companies played an increasingly important role. In 1935 and
1937, however, the state acquired the private exchanges in Istanbul and Ismir.
In the 1940s, new exchanges used equipment made by Ericsson and the LMT
Company.
The initial development of telecommunications was part of the modernization
of Turkey after World War II. However, the pace of development was mark-
edly slow. In 1954, Ericsson was contracted for switches to serve thirty cities,
but it was not until the mid-1960s that telecommunications again became a
priority. Fully automatic domestic telephone service was finally established in
1976, followed by similar international service in 1979.
The Turkish PTT holds a monopoly control over telecommunications and is
supervised by the Ministry of Transportation and Communications. In the mid-
1980s, the government considered a separation of telecommunications from
postal services and the privatization of telecommunications services.
In 1982, the average waiting time for a telephone was seven years (OECD,
1982)! More than 70 percent of villages had no telephones. In 1983, the PTT
announced its plans to give high priority to an expansion of service to all parts
of the country and to an increase in penetration. At the time, telephone line
density was only 3.5 per 100 population (Altay, 1987). The goals of the plan
included the elimination of the waiting list for telephone service by the end of
1995, modernization, and new services.
From 1985 to 1988, investment quadrupled to more than $365 million an-
nually (ITU, 1990). Financing was accomplished through the PTT's own sources
and through domestic and international borrowing. By 1986, the PTT had
achieved a number of improvements. The number of subscribers to telephone
service increased by 67 percent, and the ratio of the number of demands for
telephone service to the number of subscribers had fallen from a huge 1.09 in
267
268 Turkey
1985 to a still considerable 0.27 in 1988. Penetration went from 4.3 phones
per 100 population in 1985 to 12.2 in 1989. The waiting list dropped from 2.1
million to 1.3 million in 1988, but it increased again to 1.9 million in 1990
(ITU, 1990; General Directorate of the PTT, 1990). The percentage of villages
without telephones decreased in a few years from 72 percent to 28 percent in
1989. Although the wait for service had dropped to two to three days in Istan-
bul and Ankara by 1990, in more remote areas the wait was still two to three
months (Anik, 1990). But there was no longer a waiting list for telex service.
In 1990, Turkey had a total of 7 million phones.
After 1984, the PTT also introduced new services. These include digital
exchanges, fiber-optic cables, digital radio relay, multiaccess radio relay in
rural areas, paging, dial-up modems, telecard payphones, cellular and cord-
less telephones, teletext, and public telefax. The circuit-switched DATEX-1
data network and the packet-switched data service Turkpak were also intro-
duced.
The government embarked in particular upon an ambitious plan to improve
the infrastructure of Istanbul, a city of 6 million inhabitants that experienced
especially rapid growth during the 1980s. Istanbul was attempting to position
itself as an international business center linking the Middle East and Europe,
but it woefully lacked advanced telecommunications. In 1986, booking over-
seas calls sometimes required several attempts. In 1988, the country's two in-
ternational gateway switches were replaced with DMS 300 digital exchanges in
Istanbul and Ankara. Turkey's first communications satellite was scheduled for
launch in 1993.
After 1988, telephone sets, telefax machines, and mobile telephones were
available from private sources, as well as from the PTT. Equipment not sup-
plied by the PTT is subject to its approval. For terminal equipment, Turkish-
made equipment that meets established standards is generally used. Procure-
ment by the PTT is in principle by public tender, but with the major loophole
that tenders need not be undertaken where the equipment is produced by com-
panies in which the PTT holds an interest.
In 1967, the PTT embarked on a course to upgrade Turkish equipment man-
ufacturing: to design and construct transmission equipment, partly under license
from the Belgian ITT subsidiary BTM. It established the firm Netas in collab-
oration with Northern Telecom for the production of crossbar exchanges. It also
decided to spin off its Aria research laboratories, which had become quite large,
into the new firm Teletas, a joint venture with Alcatel-BTM, and expanded it
to serve as a second source for digital switches with its System 12. In 1987,
Siemens became Turkey's third digital switch supplier with 100,000 digital
EWSD lines, and it has since added 300,000 more. Yugoslavia's Iskra was
brought in to provide small rural switches.
Though the PTT privatized Netas and Teletas, it maintains a strong influence
through its research labs. Netas also manufactures telephone sets and the lo-
cally developed rural switch ELIF, as well as PBX equipment. Teletas also
manufactures transmission equipment, teleprinters, and telephone sets, and the
Turkey 269
firm TTE produces rural switches. According to the PTT, about 85 percent of
network equipment is produced in Turkey (Raggett, 1986). Supported by this
high level of domestic PTT procurement, Netas and Teletas also sought export
markets in northern Africa and the Middle East, and in 1991 Netas won a
contract in the Soviet Republic of Azerbaijan.
26
Greece
By European standards, Greek telephone service is backward. This situation is

partly due to the poverty of the country relative to Western Europe and the
turmoil and polarization of its politics, including military dictatorships, foreign
occupation, and divisive elections. It is also attributable to the problems of its
telecommunications monopoly, the Hellenic Telecommunications Organization
(OTE).
Compounding the problem is the rugged landscape of the interior areas of
Greece, the need to reach the many islands, and the politicization of OTE.
Given this scenario, it will be difficult for Greece to become the crossroads
rather than the weak link of communications between Europe and Asia.
Greek telephony started in 1895 with two local systems. For a long time,
the system was rudimentary. In 1930, the private Anonymous Eteria Telepi-
kononion Ellados (AETE) was established to operate and develop local tele-
phone service. In 1931, the AETE installed three automatic telephone ex-
changes in Athens and later expanded service into the provinces. The system
used equipment from Siemens, which dominated Greek telecommunications be-
fore World War II. After the war, the system was nationalized and OTE was
created. Siemens is still the main supplier, and its close ties to OTE continue.
By 1949, there were fifteen automatic local exchanges in Athens and twenty-
three in the provinces, with a total capacity of 6500 lines; 364 manual ex-
changes provided another 4230 lines. Telephone penetration at this time was
only one per 100 population. During the same year, the government established
OTE to take over telephony. Its sole share is held by the government. OTE
operates under the supervision of the PTT, which is part of the Ministry of
Transportation and Communications, and has a monopoly over most forms of
communications, including broadcast transmission. One exception is local mo-
bile telephone service, which may also be provided by private operators as long
as it does not connect to the public network.
Greek telecommunications are lagging. The conservative Karamanlis govern-
ment assigned a fairly low development priority to telecommunications. It also
entered into a barter deal with Romania and East Germany for switches in
return for agriculture products, which led to the introduction of low-quality
equipment. The socialist Papandreou government that followed emphasized
telecommunications and undertook a program of digitalization, but implemen-
tation of these plans was slow. From 1978 to 1987, Greece's yearly investment
270
Greece 271
per line was Western Europe's lowest by far, a paltry $21.71, compared with
the OECD average of $572.13 (OECD, 1990).
In 1984, telephone density in some districts was only seven to fifteen per
100 (OTE, 1985). The telephone failure frequency per 100 inhabitants was
57.5. About one-half of switches were still of the rotary type, with the rest
being crossbar and other varieties. In 1983, the time required to satisfy the last
application for main connection was 4.04 years. The mean waiting time to
establish an operator-assisted international call was sixteen minutes in the main
cities and in Athens (OTE, 1985, p. 30).
In 1985, the waiting list for main connections was over 970,000, about a
third of total main lines (3 million, half of which were in the greater Athens
region). By 1988, telephone density had grown to approximately thirty-six per
100 and there were about 3.6 million main lines. The wait for a private tele-
phone installation was typically three to five years, and the waiting list had
grown to 1.05 million requests (ITU, 1990). Even the highest-priority orders,
those for hospital and business users, required at least several months and as
much as one year to fill. Digital exchanges began to be introduced in 1986 and
1987, but in 1990, OTE was unable to supply digital lines to private VAN
operators (Schenker, 1990, p. C5).
Equipment used by OTE is supplied primarily by Ericsson and the Greek
subsidiaries of ITT and subsequently Alcatel, Philips, and Siemens. OTE pro-
vides Siemens PBX equipment, but users are free to order from other vendors.
Telephone handsets, however, are supplied and maintained by the PTT. OTE's
approval policy for private terminal equipment is relatively liberal, but the sub-
mission of a detailed technical description in Greek is required, and the proce-
dure usually necessitates a representative in Greece. Modems for leased circuits
are privately supplied, although OTE will install them. Telefax equipment is
also available from private suppliers, but attachment requires OTE permission.
Telex equipment can be obtained from either the OTE or private firms and must
have both Latin and Greek characters.
Relatively little data transmission takes place over the public switched net-
work because of its poor quality. Though rates of more than 1200 band are
available, error rates are relatively high. As a result, only 5 percent of all Greek
modems were connected to the PSTN in 1985. OTE regulations specified that
each leased line may be used to transmit only one form of information. For
example, despite the backlog in request for lines, a line could not be used for
both voice and data. In 1987, the public-packet switched data network Hellas-
pac began operation, with transmission rates of up to 40 kbps. Its largest cus-
tomer in 1980 was the on-line documentation center of the Greek Ministry of
Industry, Trade, and Research.
In 1984, OTE decided to replace 400,000 telephone sets. It rejected several
international supply offers and instead chose an offer from a joint venture of
Eommex and the Greek firms Intracom, Elinda, Azinko, and Hourdakis. Eorn-
rnex is an organization of small and medium-sized businesses headed, in 1984,
by Vaso Papandreou, no relation to then Prime Minister Andreas Papandreou,
but a close associate of his. Eventually, Intracom handled most of the order.
272 Greece
The primary argument for using the Greek suppliers was the promise of sub-
stantial domestic manufacturing. It soon was charged, however, that most parts
were imported from abroad and merely assembled in Greece. An outside inves-
tigation revealed a 47 percent Greek value added instead of the guaranteed 73
percent.
At that time, OTE was also embroiled in charges of improprieties involving
its new headquarters building. Construction of the building proceeded at a very
slow pace for ten years, and the building stood empty for a number of years
thereafter.
In the early 1980s OTE decided to introduce digital technology into its net-
work and exchanges. A 1988 agreement with Greek assembly chose Ericsson's
AXE and Siemens's EWSD switches among four bidders. Northern Telecom
supplied crossbar equipment. A special committee headed by a high-ranking
official of the Industrial Development Bank was established to negotiate their
purchase. The negotiating team included two economists, two university-based
technologists, and the deputy managing director of the OTE. This structure,
over which OTE did not predominate, led to constant clashes with the manag-
ing director of OTE, Theofanis Tombras, who was accused of continuous in-
> Tvention in the negotiations.
Influential in OTE personnel matters is the FIT union, which plays an un-
official but important role in managerial appointments. It is about one-third
conservative, one-third socialist, and one-third communist.
In 1987, the minister of transport and communications expressed pessimism
about the future of OTE: "OTE desperately tries to identify capital sources to
satisfy over one million pending petitions for new service . . . but the required
amount is presently a dream." And he pledged to "bring order to OTE's chaos."
Shortly thereafter, the minister was removed. The real power in Greek telecom-
munications was OTE's director general Tombras, who was appointed by Prime
Minister Papandreou in 1981. Tombras had been a military officer who was
part of a leftist army group; he was accused of a plot, went into exile, returned,
and was close to the prime minister. In that position, he was useful to the
government. Tombras bragged to the press about wiretapping opposition lead-
ers by a special department of OTE. He also cut communication links to broad-
cast stations of the opposition during the election campaign.
Telephone investments were made based on political considerations (where
loyal voters were), and it was alleged that to receive priority on the waiting
line required political patronage.
As the government's point man, Tombras became well known in Greece,
and the opposition announced its intention to bring him to trial if it were to
win the election.
The elections were held under a cloud of scandal in government, when a
former Crete banker confessed to massive improprieties in the handling of state-
owned organizations' funds, which benefited government officials and the rul-
ing Pasok party. After several inconclusive elections, the conservatives came
to power in 1990.
Not surprisingly, this climate was not conducive to the development of tele-
Greece 273
communications infrastructure. (Tombras found enough time, however, to un-

successfully intervene with the president of Columbia University, asking him
to censure my observations about Greek telecommunications, which I had re-
peatedly sent to OTE for corrections and comments. The draft did not include
the political discussion of the last several paragraphs, since this was not known
to me at the time. When I flew to Athens, with an appointment to meet mem-
bers of Tombras's staff to receive corrections, I was refused the meeting.)
When a conservative government came to power, the hope of many Greeks
for better telephone service rested with the possibility of its winning the bid for
the 1996 Summer Olympic Games, an event that would have established a firm
deadline for improvement of the country's infrastructure. But that bid failed,
leaving Greek telecommunications still a weak link in Greece's infrastructure.
In 1990, the new government announced an ambitious three-year plan to add
1.2 million digital lines to the OTE network. It also proposed the introduction
of cellular service and the liberalization of VANs.
Since Greece was the only EC country left without a cellular telephone sys-
tem, the government invited in 1991 bids for two competitive cellular telephone
systems, which could be foreign owned. Although OTE would own one third
of each, thus first opening in the service monopoly, the change augured that
Greece, too, was following the broader European trend.
Problems persisted. Procurement contracts were awarded to Greek firms which
largely reassembled parts from the Far East. After OTE Director General Kyr-
iakos Kioulafas refused to halt negotiations with Siemens Greece over an un-
derseas cable project (when Denmark's NKT had proposed a similar deal for
$5.5 million less), he and others in OTE's top management were dismissed.
OTE's digital exchange procurement led to bitter disputes over the suppliers
and politicized the decisions process. These procurement scandals encouraged
calls for reform which included proposals for the sale of 49 percent of OTE to
private investors to raise capital needed for massive modernization programs
(Kotsonis, 1990). Due to the less-developed state of the Greek network, the
EC proposed in 1991 to apply a special set of telecommunications policies for
Greece.
Cyprus
In 1955, the Cyprus Telecommunications Authority (CTA) was formed to re-

place Cable and Wireless as the telephone provider for the newly independent
country. CTA telephone service reaches 35 out of every 100 persons, with one-
third of traffic digitally switched. CTA holds a monopoly over telex, facsimile,
and data services. Cellular service based on the NMT 450 standard was started
in 1988. The equipment market in Cyprus was liberalized in 1988 after CTA
could no longer keep pace with demand. CTA did, however, retain type ap-
proval powers as well as its complete monopoly on services.
27
Telecommunications in Eastern Europe
Introduction
Telecommunications in pre-reform Eastern Europe resembled that of the West

in one important aspect: monopoly control over all aspects, under the aegis of
a state PTT. Technical capability, performance quality, and service availability,
however, were far behind. Telecommunications had long been neglected. Ac-
cording to a World Bank study, networks in Eastern Europe had growth rates
of only 3 percent over the last thirty years, and what little investment was made
went to new lines, not maintenance or infrastructure (Nulty, 1990). Penetration
figures therefore did not reflect poor service quality and outdated equipment.
Disparities in telephone penetration existed between neighboring countries
and between urban and rural areas. Throughout Eastern Europe a typical appli-
cant for new telephone service had to wait a decade or more for service. A
post-reform report in Poland noted that a thirty-three year wait for telephone
installation was not unheard of. Some of the equipment dated back to World
War II or before; international service was often limited to less than 100 in-
bound and outbound lines, and advanced applications such as cellular mobile
phone service were unavailable. Given that the poor quality of transmission
prevented facsimile and data services, telex service was heavily used.
This neglect partly reflects the view that telecommunications is a nonpro-
ductive consumption sector of the economy, as opposed to agriculture and in-
dustry. Telecommunications profits were typically used for other projects. In
Czechoslovakia, for example, 87 percent of PTT profits were diverted to the
general budget. Another problem was state control: international direct dialing
was often prohibited and telephone directories were made classified documents.
Once economic and political reforms were introduced, the upgrading of the
telecommunications infrastructure received high priority as central to overall
economic development. East European governments sought to almost quadru-
ple growth to 10-11 percent annually. Change was introduced in several areas:
improvement of existing networks and customer service, accommodating new
technologies such as digital switching and VANs, and institutional and legal
reform.
East European governments had to balance competing pressures. The up-
grade of public networks requires huge investments, as much as $140 billion
in the next decade (Pearce, 1990). There is pressure to target investment for
274
Telecommunications in Eastern Europe 273
special user groups and to allow private providers of telecommunications ser-

vices to enter the market.
East European countries traditionally exported telecommunications equip-
ment to the Soviet Union. These industries existed without competitive pres-
sures to develop or modernize and were in fact protected by the West's so-
called CoCom restrictions that kept advanced technology from the East. With
the easing of restrictions, more efficient Western firms could compete to supply
equipment to fill these countries' huge needs. However, all the Eastern Bloc
countries shared a similar obstacle: They lacked the hard currency to import
Western equipment. Most countries either had to strike barter agreements or
had to trade an equity stake as a means of acquiring badly needed technology.1
The following sections provide capsule descriptions about change in the var-
ious Eastern European countries. Because of the pace of transition, these sur-
veys are brief.
Hungary
Despite its relatively high per capita income, Hungary has one of the least
developed telecommunications networks among East European countries (EESTR,
1990a). At the same time, Hungary earned a reputation as a leader: It was the
first Eastern Bloc country to invest in Western digital switching technology,
first to introduce cellular service, and first to apply for membership in the Con-
ference of European Postal and Telecommunications Administrations (CEPT),
the organization of (West) European PTTs.
Hungary's first telephone exchange was installed in 1881. Development slowed
during and after World War II, when priorities shifted toward economic cen-
tralization and the military. Existing facilities deteriorated for lack of invest-
ment funds; annual growth in telephones from 1967 to 1987 was 4.3 percent,
the lowest among countries of similar economic development (Datapro, 1990).
Demand for telephone lines grew with the economic reforms of 1968, when
economic planners allowed greater horizontal relations between companies. In
1983 the PTT Magyar Posta was turned into an organization separate from the
Ministry of Transport, Communications and Construction. In 1989, Magyar
Posta was split into three bodies, for telecommunications, broadcasting, and
traditional postal service. In anticipation of a privatization, the telecommuni-
cations group was renamed the Hungarian Telecommunications Company
(HTC). The Telecommunications Act was revised in 1989 to allow limited
private investments in Hungarian service providers (EESTR, 1990a). Shares of
HTC were offered to Hungarian firms in 1990 and to foreign investors in 1991,
with total foreign ownership limited to 25 percent. HTC's investments were still
directed by the Ministry for Transport, and rates are approved by the State
Price Control Office. Further legislation proposed competition in value-added
services and the establishment of private local telephone companies, leaving
the PTT with a monopoly over long-distance services.
In 1984 Magyar Posta announced plans for fully automatic switching by
276 Telecommunications in Eastern Europe
1995. Equipment production was converted from electromechanical to elec-

tronic switches in the same year (Budapress, 1984). In 1985 Hungary finally
expanded direct access for international calls to seventy-three countries from
major cities. A $70 million loan from the World Bank in 1987 was used by
the PTT to purchase cable and switching equipment from West Germany, Tur-
key, and France.
The five-year plan for 1986-1990 called for 420,000 new lines to increase
telephone penetration to seventeen main lines per 100 people by 1990 and to
40 per 100 by the year 2000. Estimates of telephone penetration for 1990 fall
between 7.6 and 9.1 main lines per 100 people. Residential telephone service
is inexpensive. The monthly line charge is $1.50 and 3 cents for each local call
(Rocks, 1990). The wait for phone service averages ten to twelve years, and
550,000 people are on the waiting list. The PTT's goal is to cut the waiting
time to one year by the year 2000.
Two-thirds of calls in 1988 could not be completed in the main cities, with
service even worse in the outlying areas. The government estimated that a 1
percent loss in national income could be attributed to poor telephone service.
Since over 80 percent of local exchanges are manual, many subscribers cannot
obtain service after the post office closes at 4 P.M. Urban areas and the trunk
network have crossbar switches, and a few digital switches are in place. Hun-
gary's service problems are worsened by the fact that over half of its 1.4 mil-
lion lines serve the 20 percent of the population that lives in major cities.
Modernization plans call for replacement of all rotary switches in the local
network by the early 1990s and in the toll network by 2000. Investments of $5
billion were projected to satisfy demand and deploy digital switching equip-
ment. HTC expected to generate 50 percent of investment from revenues and
10 percent from government, with the remainder coming from outside sources.
To that end, Hungary sought additional loans of $100 to $150 million from the
World Bank and the European Investment Bank (Pronay, 1990).
Hungary manufactures telecommunications equipment through its main
equipment firms Budavox and Beloyanis Telecommunications Factory (BHG).
It exports to the ex-Soviet Union and Eastern Bloc nations as well as to Greece,
Italy, and England. Joint ventures with Western firms were crucial in Hunga-
ry's post-reform policy. Siemens, which was blocked under CoCom restrictions
from exporting its switches to Hungary, in 1984 entered a joint venture with
the Hungarian firm Videoton to manufacture equipment in Hungary. Similarly,
Alcatel Austria entered a joint venture with HT Hirdastechnika Szovetkezet to
manufacture the Alcatel 5200 PBX in Hungary. Northern Telecom formed a
joint venture to build digital switches for HTC and private business customers.
A contract for cellular service was awarded in 1990 to a joint venture be-
tween HTC and US West for a system that would ultimately interconnect with
the pan-European GSM digital mobile system. Critics argued that Hungary should
instead invest in the more basic aspects of the existing telephone system. How-
ever, rival cellular providers were likely. In 1989, Bond Hungaria Telecom
a joint venture between Coopinvest, Australia's Bond Corporation, and oth-
ersannounced plans to start a separate 900MHz system. The old government
had awarded a cellular license to the American firm Contel, but the new one
switched the franchise to US West and HTC.
Poland
The Polish public telecommunications system suffers from forty years of ne-
glect on top of a devastating war. It was not a priority to allow citizens access
to a well-developed communications network, and there was consequently little
investment or R&D in telecommunications. As a result, the system suffered
from inadequate technology and bottlenecks at every level. With political and
economic reform, Poland's financial limitations led to an approach toward in-
stitutional reform that went further than that of Hungary, embracing full de-
monopolization of the PTT and allowing foreign interests to compete directly
with it.
Communications in Poland dates back to 1558, when King Sigismund Au-
gustus established the Polish mail between Cracow and Venice. The Polish Post
Office, Telegraph and Telephones (PPTT) was formed between the two world
wars and built a telecommunications network with foreign financial assistance.
The system was greatly damaged in World War II. By 1944, the Soviet Union,
Britain, Sweden, and the United States were supporting efforts to rebuild it.
The number of telephones more than doubled between 1960 and 1972, reach-
ing a penetration of 3.62 per 100 (Poland: A Handbook, 1974). In 1990, there
were 3.28 million subscribers, or 8.3 phones per 100 people. The average wait
for service was thirteen years, and delays as long as thirty-three years were
reported. The 2.2 million people waiting for a phone may be only a portion of
those actually interested in telephone service. Poland's network employs local
exchanges dating from the 1930s and the 1960s. Fifteen percent of switches
are over thirty years old, and the 1990 call completion rate was only 30 percent
(EESTR, 1990b).
The PPTT was removed from the control of the Ministry of Communications
in 1987, after the ministry was merged into the Ministry of Transport and
Shipping, which gained control over rates and technical policies. In 1990 a
new ministry was created to oversee the PPTT, the Ministry for Posts, Broad-
casting and Telecommunications.
The PPTT operates the national network and ran postal and broadcast ser-
vices (Datapro, 1990). The PPTT is financially self-sufficient but must contrib-
ute 40 percent of its income to the government and is obligated to subsidize
the post office. In 1991, postal, broadcast, and telecommunications services
were separated into individual operating units as a step toward privatization.
Telecommunications were provided by Telekomunikacja Polska SA. Long dis-
tance services were also separated from local ones, which were to be offered
by independent regional companies, potentially private and with minority for-
eign participation.
The 1991 Telecommunications Act permitted competition with the TPSA in
many services, such as local, cellular, and domestic long-distance. Regulation
was by the Ministry and its subagencies PIT and PAR. The government's re-
form plans were ambitious. TPSA was to be privatized in the future, and per-
haps divided into a national long-distance organization and several regional
companies. Foreign applicants could not own a majority of shares, except for
local service companies.
The post-reform Polish government made upgrading the telecommunications
system one of its two economic development priorities and targeted Warsaw
and the Silesia region. PPTT goals were to quadruple telephone penetration by
2000, setting a goal of 10-12 million subscribers by that time. In addition, the
PPTT estimated that as much as 60 percent of the embedded network will also
have to be replaced or significantly improved. The PPTT sought to upgrade
international capacity and to modernize the network through AT&T. To im-
prove the domestic network the PPTT bought eight transit exchanges from
Alcatel-Spain on generous terms arranged for by the Spanish government.
In 1990, the PPTT installed Kommertel, a separate overlay business network
in Warsaw. Installation fees and line charges were five times the normal, but
interconnection was immediate. Poland had Eastern Europe's highest telex ser-
vice penetration. Satellite services were offered through Intersputnik. Poland
joined Eutelsat in 1990.
Before the reforms of the late 1980s, manufacturing had been handled by
United Telekom (UT), a state monopoly with 20,000 employees. In 1989, UT
was split into three separate entities that are to compete with one another: ZWUT,
Teletra, and the Polish Transmission Works. The three firms were expected to
produce 1.8 million lines; of those, the PPTT orders came to 1 million lines
per year, and the remainder was aimed for export (Datapro, 1990).
The country was open to joint ventures with Western equipment providers.
Domestic manufacture was directed by the Ministry of Industry in partnerships
with Western firms. ZWUT produces the EWSD switch with Siemens, and
Alcatel produces its E10 and System 12 switches with Telettra and the Polish
Transmission Works (PZT). In 1990 Siemens bought a 49 percent stake of
ZWUT for DM 50 million. Ericsson was involved with Telecom Telfa. AT&T,
Italtel, and Samsung were also active. Customer premises equipment was lib-
eralized subject to type approvals of the PPTT's Institute of Telecommunica-
tions.
Twenty-five companies offered bids for cellular overlay networks. Sweden's
Televerket, Finland's P&T, and British Telecom formed the Baltic Mobile
Telephone System (BMTS) consortium, which bid to expand the NMT 450
cellular system now operating in the Scandinavian countries to the Baltic-rim
countries, including Poland. A franchise was won by a consortium of Ameri-
tech and France Telecom.
Czecho-Slovakia
Czecho-Slovakia's telecommunications network is controlled by the Federal
Ministry of Posts and Telecommunications (PTT), which was separated from
the Ministry of Transport in 1990. All equipment is produced under the super-
vision of the Federal Ministry of Metallurgy, Engineering and Electrotechnical

Industries. In 1989, legislation was passed that required the PTT to earn a
profit, but the amount of those profits that would be transferred to the state was
left unspecified.
The PTT's structure is federal. The two republic-level operators are SPT
Praha and SPT Bratislava. And the struggle over the extent of centralism led
to creation of a Slovak PTT Ministry. Despite limited investments and growth
rates under 3 percent from 1978 to 1989, the infrastructure is above average
for East European nations. Telephone penetration led the Eastern Bloc, with
twenty lines per 100 households in 1978 and twenty-seven in 1990, and per
capita penetration at fourteen main lines per 100 persons. Nevertheless, the
1990 waiting list was 370,000 names long with a ten-year wait. The figure
rises to over 700,000 when those who have dropped off the list are included.
The network is mostly analog, with 3 percent of lines served by Alcatel E10
digital exchanges. It also serves the broadcast needs of radio and television
(Antono, 1990). Two international exchanges are located in Prague. The SPTs
announced ambitious plans for expanded digital switching, packet-switched ser-
vice, paging, videotex, e-mail, and private networks.
After taking bids from British Telecom, McCaw, and a West German con-
sortium, the PTT awarded a twenty-year license for an analog cellular network
to a joint venture of the PTT and US West and Bell Atlantic. Bell Atlantic and
US West won a related contract to build a public packet-switched data network
(Eurotel).
The Czechoslovak equipment industry was dominated by the state-controlled
manufacturer Tesla Karlin, part of the Tesla Electronics Group. After the gov-
ernment decided to privatize industry and allow joint ventures, Tesla Karlin
formed a joint venture with Siemens to manufacture switches. Alcatel SEL
formed a joint venture with Tesla-Liptovsky. Modems are manufactured by
ZVTf Banska Bystrica. There were about 5000 modems connected in 1989
(Datapro, 1990, p. 264). The PTT also sought to improve services by purchas-
ing Ericsson, Alcatel, Philips, and Telettra equipment for switching and data
transmission.
In 1992, a new telecommunications law provided for eventual competition,
except in basic voice service, regulated nationally.
YugoslaviaCroatiaSlovenia
The Yugoslav postal and telephone system was highly decentralized, with six
republic PTTs under a weak federal administration. Although each of these
individual organizations ostensibly determined the prices for its services, the
government oversaw rates and manages revenues and expenses within the state
budget (Lakicevic, 1970).
The number of telephones and automatic exchanges doubled from 1939 to
1950 (Byrnes, 1957), but growth was slow until the mid-1970s. Since 1979,
Yugoslavia encouraged foreign investment, which led to 9 percent yearly growth

and the deployment of a packet-switched network. In 1988, telephone penetra-
tion was 13 main lines per 100 population. Digital service began in 1986 with
the purchase of twelve ITT System 12 switches. In 1987, satellite communica-
tion began under an agreement between Mitsui, NEC, and the PTT to construct
a satellite station linking Yugoslavia with Western Europe via Eutelsat (Busi-
ness Eastern Europe, 1987).
Yugoslavia also imported foreign technology and expertise through its equip-
ment manufacturers. A partnership of GTE and Elektronska Industrija of Nis,
known as GTE-Pupin, produces PBXs and exchanges. In 1990 it installed its
first digital phone exchange near Ljubliana, as part of a joint venture between
Siemens of Germany and Iskratel of Yugoslavia (U.S. Department of State,
1990). Iskratel also has technology transfer agreements with ITT and American
Microsystems. Other joint ventures included Alcatel with Elektronska Indus-
trija, and Ericsson with Tesla.
Despite these moves to integrate modern technology into its telecommuni-
cations infrastructure and a 50 percent increase in investments, Yugoslavia's
telephone penetration was only slightly above the Eastern European average.
The subsequent secession of Croatia and Slovenia, the ensuing civil war, and
the establishment of Bosnia-Herzegovina and Macedonia disrupted and dam-
aged the telecommunications infrastructure and institutions.
Bulgaria
Telecommunications in Bulgaria began with government ownership under the

General Directorate of Post, Telegraph and Telephone in 1878. Telegraph lines
were completed between Sofia and Plovdiv in 1892. Services continued to ex-
pand, and in 1938 there were 8600 km of telegraph lines with 750 stations and
21,086 km of telephone lines with 584 exchanges and over 27,000 phones,
processing 54 million calls annually (Dellin, 1957).
When the Communists seized power in 1944, they made no legal changes to
the PTT. By 1950, over 57,000 lines were reported to be in service, compared
to 27,500 in 1939. Revenues and expenses of this system were included in the
state budget. Government reports at the time claimed a fourfold increase in
land-based wire line facilities, compared to prewar levels.
According to OECD figures, Bulgaria had the greatest number of main lines,
18.4 per 100 population in Eastern Europe by 1988 (Kelly, 1990). This relative
strength came from Bulgaria's investments in telecommunications in the 1980s,
which gave it a reputation as Eastern Europe's best developed system (Nulty,
1990).
Bulgaria's plan to upgrade telecommunications, which involved installing
electronic and semiautomatic exchanges to enable 155,000 new subscribers to
join the system, anticipated a foreign equity participation in a privatized BPT.
The main equipment manufacturer, Sofia Telecommunications Plant, exported
16 percent of its output to the Soviet Union, including over 50 percent of

handsets. BPT aimed to liberalize the equipment market.
Romania
Romania's telephone network has been operated with some rotary switch equip-
ment since 1933. The equipment in the network's central offices is estimated
to be thirty years old on average (U.S. Department of State, 1990). Following
the overthrow of President Ceaucescu, however, the new government moved
to significantly upgrade all its basic services and to establish some value-added
services. It also separated the Ministry of Post and Telecommunications from
the Ministry of Transportation. As in most Eastern Bloc nations, telecommu-
nications heavily subsidizes the postal service.
Romania's first automatic telephone exchange, a rotary switch built by ITT,
began operation in 1927 in Bucharest. The network continues to rely on such
rotary equipment, although in 1968 a licensing agreement was signed with ITT's
Belgian BTM that allowed domestic production of more advanced Pen-
taconta automatic switches for local and trunk traffic (Popescu, 1990). In 1990,
Romania contracted Alcatel FACE to provide System 12 switches for Bucha-
rest.
Estimates of Romanian telephone penetration vary from 6.7 to 7.4 main Sines
per 100 in 1990 (Fidler, 1990). The government's own official estimate is ten
lines (including party lines) per 100 persons. Service is unevenly distributed
between urban and rural areas. Penetration in Bucharest runs at thirty lines per
100 persons, with other urban areas averaging fifteen per 100. The rural aver-
age is 2.5 lines, and 3300 villages have yet to receive service (Popescu, 1990).
The number of subscribers tripled from 1965 to 1975, and an aggressive five-
year plan was proposed for 1976-1980 that called for a minimum expansion in
facilities of 30-40 percent to accommodate international traffic, with special
emphasis on data (Avramescou and Celac, 1980). By 1990, 89 percent of tele-
phone subscribers were served by some form of automatic switch, but data
services had not improved. For example, there were only 200 facsimile ma-
chines in the country in 1990.
A backlog of 800,000 orders for telephone service in urban areas alone has
resulted in a ten-year wait for service. The government forecast an additional
2.5 million lines to meet rural needs and has set a goal of 50 percent penetra-
tion in the next fifteen to twenty years. To meet these needs, the aim was
400,000 new lines per year, assisted by a joint venture to manufacture digital
equipment at Romania's Electromagnetic Plant.
The government also planned to upgrade the network by completing the ex-
isting analog trunk network and adding electronic switches, digital microwave
lines, fiber-optic cable, digital multiplexing systems, and a cellular network. In
addition, there were plans to develop a high-speed, packet-switched data trans-
mission network to upgrade the previous ROMPAK network.
In 1990, all international traffic was handled by a single crossbar switch with
a 340-line capacity that had been in place since 1974. To modernize interna-
tional access, the PTT purchased a Siemens EWSD digital switch. Siemens
also acquired 49 percent of the manufacturer EMCOM. Romania joined Eutel-
sat and became the first Eastern Bloc country to gain membership in Intelsat in
1990.
For much of this technology, the new Romanian government sought joint
ventures with Western companies, and the PTT indicated that it will privatize
mobile cellular systems, packet switching equipment, and terminal equipment
for value-added services, but not for basic services. Rom Telecom was sepa-
rated from the PTT and assisted by France.
Albania
The close government control over telecommunications in Albania allowed lit-

tle information about its network to be disseminated. In 1939, the country had
fifty-six hand-operated exchanges serving fewer than 900 lines. Only one au-
tomatic switch existed, serving 150 lines in Tirana. Telegraph connections were
in place with both Greece and Yugoslavia (Skendi, 1956, p. 251). Albania
joined the ITU in 1922 but has not participated in any working groups or as-
semblies, or reported penetration statistics.
The Italian occupation of Albania from 1939 to 1943 first brought some
expansion for the country's system, in establishing international connections
with Yugoslavia, Bulgaria, and Italy. The domestic telecommunications infra-
structure suffered extensive damage in the war.
Albania's telecommunications systems are owned and operated by the state
through the Directorate of Post, Telephone, and Telegraph, within the Ministry
of Communications. Network expansion was never a priority. In 1956, only
fifty-six telegraph and seventy-five telephone exchanges were in service, the
largest of which served 500 lines. Residential telephones were almost unheard
of except for officials and doctors. In 1990, direct-dial long distance access
started in several cities, though all traffic is routed through a single switch in
Tirana.
With a democratic government elected in 1992, major changes were antici-
pated.
Russia2 and Commonwealth of Independent States
Telecommunications were long neglected in the former Soviet Union because

they were viewed as a consumptive rather than a productive activity. The So-
viet telecommunications system was probably not simply a failure in terms of
management, but also a fairly accurate reflection of the low priority held by
the state for individualized and decentralized civilian communications. The
country's centralized economy, compartmentalized bureaucracy, and wariness
about citizen access to communications flows were reflected in the rigid hier-
archical structure and poor condition of the public network. The result was a
communications system that was inadequate for the economy of a superpower.
By comparison, the United States had more than seven times the number of
telephone lines and nine times the number of residential lines. However, the
Soviet Union under Mikhail Gorbachev put telecommunications development
more into the forefront. Whereas 1 million new lines were installed annually
from 1980 to 1985 (already a major increase over the 1970s), twice as many
were added in the second half of the decade, and residential service became a
greater priority. At the same time, the process of disintegration into constituent
republics led to an increasingly decentralized telecommunications system.
Organizational Structure
In the former Soviet Union, telecommunications facilities, services, and man-
ufacturing were state monopolies, long under the Federal Ministry of Posts and
Telecommunications (MPT). The Ministry of Communications, known as
Minsviaz, operated national and international networks and services, as well as
postal services and TV and radio broadcasting. In addition to national Mins-
viaz, each of the fifteen republics had its own Minsviaz, which was subordi-
nated to the national ministry as well as to the government of the respective
republic. (The Russian Minsviaz was created fairly late, its functions having
been undertaken before by the national ministry. Even then, telecommunica-
tions in the city of Moscow were still run by the national ministry.)
The national Minsviaz' responsibility was for overall system coordination
and operation of overarching activities, such as national long-distance service,
communication satellites, and national broadcast services. Minsviaz seemed to
have no oversight responsibility over important intraorganizational networks,
such as those of the defense establishment, the KGB, and the ministry of In-
ternal Affairs (Campbell, 1988).
Subordinated to Minsviaz were several chief administrations, including those
for industrial enterprises (GUPP), science and technology (GNTU), satellite
and radio broadcasting (GKRU), postal services (GUPS), long-distance trans-
mission (GUMTS), telegraphy (GTU), and urban and rural services (GUTS).
The republic ministries were responsible for services in their territory, provided
through regional organizations (PTUS) and production enterprises (RUS). Most
of Minsviaz' 7000 such enterprises had some financial independence, but finan-
cial subsidies and redistribution were also necessary, especially since more than
1000 lost money. Overall development plans were formulated by Minsviaz and
were part of the national five-year plans. But actual investment funds had to be
provided by the republics, regional and local governments, and even industrial
and agricultural users. Thus, network construction was to some extent at the
mercy of the investment priorities and red tape of numerous agencies. With the
increasing independence of the republics, telecommunications became increas-
ingly controlled by the various republics' ministries, and their coordinating
committees' became critical.
Services
The Soviet telecommunications network was organized to serve official and
administrative rather than household needs. Military and industrial communica-
tions were especially well developed for utilities departments such as railroads,
pipelines, and power, which have separate networks comprising about 6.5 mil-
lion lines, 40 percent of which are connected to the public network. Frequently,
those systems utilize Minsviaz leased lines. The department networks occupied
20 percent of Soviet exchanges (Campbell, 1988). The 1966-1970 Five Year
Plan (FYP) mandated a Unified Automated System of Telecommunications
(EASS), conceptualized as an integrated network with the capacity to carry
voice, video, and data traffic in anticipation of an ISDN-type facility. How-
ever, little network integration actually took place and common facilities were
used only to a limited extent.
Once telecommunications became more of a priority, penetration increased
from 29.1 million phones in 1980 (of which 23.7 million were connected to
the public network) to 37.2 million in 1985, with 31.1 million connected to
the public network. Of all lines in Russia, 55 percent are residential, compared
with 84 percent in the United States.
In 1985, the central government initiated a program for telecommunications
growth and converted some military communications manufacturing to civilian
production. The twelfth five-year plan, covering 1986 to 1990, sought to add
12.1 million new lines to the system, shift to semielectronic exchanges, and
introduce fiber-optic and digital technology. Telecommunications budgets grew
from 1.2 billion rubles annually in the eleventh five-year plan to 2 billion in
the twelfth. Keeping with the emphasis on residential service, 75 percent of the
new installations were to be residential. As part of the ongoing effort to de-
velop the EASS, 2Mbps and SMbps digital lines were put in some major cities
(Datapro, 1990).
Structurally, the network consisted of a fairly rigid hierarchy of local single
or interlinked exchanges; zonal systems (about 175-200) that were assigned
area code numbers; and a third-level national trunk network with fifteen interre-
gional transit offices linked through Moscow, using foreign equipment such as
the French MT-20. Moscow was also the prime link with the Intersputnik and
Intelsat systems.
Domestic long-distance voice communications are mostly terrestrial. But ex-
tensive satellite facilities are available, as befitting a space power. The Orbita-
Molniia and Raduga satellite systems are used primarily for broadcasting and
play only a limited role in telecommunications. The next generation of satel-
lites, named Mayak, aimed to have high elliptical orbits. Also planned was a
giant 18-ton, 30-meter diameter communications satellite platform for launch
in 1993.
In 1989 there were 40 million telephones or 13 per 100 population inhabi-
tants. But regional disparities in penetration are great; 25 percent of urban
households have phones, compared with 10 percent of rural households. Rela-
tively high penetration (from 30-50 lines per 100 households) was reported for
Moscow, St. Petersburg, Kiev, Armenia, and the Baltic republics, whereas
Siberia and the Central Asian regions have less than one line per 100 homes.
Moscow accounts for 11 percent of all telephones; 20 percent of all residential
subscribers use party lines.
Tariffs were set favorably for users. Although handsets and installations were
costly, subscribers paid only 2.5 rubles per month for unlimited local service
(Campbell, 1988), among the least expensive rates in the world. Local mea-
sured billing was introduced experimentally in a number of locations but was
unreliable. Installation costs rose 500 percent, but they were below the actual
cost of 250-500 rubles (Frankl, 1989). Rural service was heavily subsidized;
in 1980 its loss was 74 percent of cost. Partly in response, a 1983 price reform
moved prices closer to cost. All state and collective farms have been connected
to the network. The allocation of telephone service was not by price, but fre-
quently as a privilege for well-connected individuals. The wait in 1989 aver-
aged four years and the waiting list included twelve million households. The
availability of pay phones was limited. Nationally, in 1987, there were only
35,000 pay phones that could reach out of the local exchange (in the United
States, the number was sixty times as high), and one survey found that half of
pay phones were out of service. In 1988, GEC Plessey Telecommunications
(GPT) formed a joint venture called Comstar with Minsviaz to provide and
operate approximately 100 credit card pay phones, principally for use by for-
eigners in airports and hotels (Dixon, 1988a). Comstar failed to show signifi-
cant results after two years.
Two joint ventures with the United States for international data transmission
were also planned. Sovam Teleport, which began operations in 1990, connects
to public data networks and electronic mail service around the world. Sprint
Networks U.S.S.R. was a 1990 joint activity of US Sprint and the Moscow
telephone organization.
In 1985, one-third of all long-distance calls still required operators. Where
operators are necessary, it can often take a very long time to establish a con-
nection. Even for automatic trunks blockage rates are 3 percent or more on
almost one half of the automatic trunks. At the same time, the rigid hierarchy
of network architecture leads to underutilization of many circuits. All interna-
tional calls were routed through one gateway switch in Moscow with 1500 lines
and approximately twenty operators to serve the entire country. In 1989, 11
million international calls were made from the Soviet Union.
The goal for the year 2000to reach 90-100 percent urban penetration and
50 percent in rural areaswill require 60 million lines in the 1990s (i.e., 6
million lines per year as opposed to 2 million in the late 1980s and 1 million
earlier in the decade). To meet these targets, the republics, like other East
European administrations, must increase spending to more than three times their
already accelerated investment rate of the late 1980s (Nulty, 1990). It was not
clear where the funds would come from. In one effort to develop innovative
financing, Minsviaz formed a commercial bank for deposits from the public
and as a financial clearinghouse for its various operations. Another strategy was
for Minsviaz to retain more of its earnings for reinvestment. Its profits in 1988
totaled 5.3 billion rubles, but it claimed that 75 percent of this was returned to
the central government budget (Campbell, 1990).
The analog network supported data transmission at 9.6 kbps on some dedi-
cated lines (which are very expensive), but the number of users was low be-
cause of the lack of transmission devices (i.e., modems) and minicomputers.

There was one X.25 packet-switched network, Academnet. There was limited
64 kbps ISDN service in Moscow to leased line business subscribers through a
joint venture with IDB (U.S.). Facsimile technology was introduced for news-
papers in 1990, but was not available to the public.
Telex and telegraph services developed as reliable alternatives to telephone
service; as a result, they enjoyed high traffic volume. All telex and telegraph
traffic passed through a gateway in Moscow. The telex network had 110,000
subscribers in 1990, growing by 5000 subscribers per year. Soviet telegraph
volume was greater than that of all other industrialized countries together, with
450 million telegrams sent yearly, and there were no public facsimile, video-
tex, or e-mail services. A more general state system for data transfer (OGSPD)
was proposed in 1972 for speeds above telex's anemic 50 baud. Such a system,
PD-200, was online in the late 1970s. It did not evolve further, partly because
of lack of terminal equipment, and partly because Minsviaz permitted data ac-
cess only to organizations, not to individuals (Campbell, 1988). Even autho-
rized users could use the system only for several minutes each hour or had to
pay very high rates for access.
Nokia of Finland proposed a cellular network for Moscow. A group of West-
ern telecommunications firmsincluding Televerket of Sweden, British Tele-
com, Finnish Telecom International, and NordicTeljoined to provide a cel-
lular network, called the Baltic Mobile Telephone System (BMTS), for the
Baltic region (Estonia, Lithuania, and Latvia). The republics of Armenia and
Kazakhstan reached agreement with AT&T for international switching, and, in
Ukraine, for network operations, with Dutch and German PTOs participation.
Equipment
A key reason for the backwardness of the Soviet telephone system was the
country's weakness in the design and production of advanced equipment. And
this had structural reasons. Whereas in the telecommunications industry Mins-
viaz controlled the operations of the network, it had only weak powers when it
came to the guidance of the development and production of the equipment that
in the aggregate constituted the network. Although Minsviaz had some produc-
tion plants under its own control, the equipment area was the province of sev-
eral ministriesMinradioprom (radio equipment), Minelektronprom (electronic
components and transmission systems), Minelektrotekhprom (electrical equip-
ment), and especially Minpromsviaz (Communications)that were part of the
defense industry and were controlled by the military industrial commission,
VPK. Within the Soviet bureaucracy, Minsviaz was no match for the powerful
VPK, which allocated resources to defense and space and away from civilian
electronics uses. The telecommunications equipment ministry did not even have
a civilian branch. In one year, 30 percent of the telegraph equipment received
by Minsviaz from Minpromsviaz was defective. One of the important changes
of the twelfth five-year plan in 1985 was to shift some control over the elec-
tronics to Minsviaz and to civilian goals. To further the separation of military
and civilian institutions, the government in 1989 abolished altogether the Min-
istry of the Industry of Communications Equipment (Minpromsviaz, or MPSS).
Most of MPSS's responsibilities for manufacturing and service provision and
12,000 workers were transferred to Minsviaz. But the former MPSS minister,
Erlen Pervyshin, became Minsviaz' new head.
Much of the Soviet telecommunications system relied on outdated equip-
ment, some of which was installed before the 1917 revolution! Soviet authori-
ties acknowledged that copper wire laid by Ericsson in 1907 was still carrying
traffic (Gulyaev, 1990). Most of the switches were analog and electromechan-
ical. Forecasts indicate that local industries cannot produce more than a fraction
of the equipment needed for the ambitious expansion plans (Frankl, 1989).
Since the 1970s, terminal equipment was owned by subscribers. PBXs, al-
though rare, could also be owned by the user. In the past, much of Soviet
switching equipment, handsets, and cables had been imported under a deliber-
ate program of support for East European countries. Over 600,000 handsets
were imported from Bulgaria in 1986, mostly from the Sofia Telecommunica-
tions Plant, and constituted more than half of that country's production. Czech-
oslovakia had also provided 2 million handsets to the Soviet Union by 1976
and several million more since. East Germany (NEK) was a major supplier of
exchanges, and Hungary served the Soviet Union with the some advanced
equipment, under Western license. Poland and Yugoslavia were also equipment
sources, as were a number of Western countries, most notably Finland. The
Finnish firm Nokia supplied exchanges and cables, and France's Thomson CSF
also provided switching equipment. With the abolition of Minpromsviaz, Mins-
viaz wrested control of the main Soviet telecommunications manufacturers from
the military industrial commission VPK. Even before, Minsviaz also controlled
a number of relatively small and undersupported manufacturing plants.3 But
the more significant development efforts were those of the VPK ministries,
which were traditionally shrouded in secrecy. Among these enterprises, of par-
ticular importance is VET in Riga, Latvia, a company originating before the
Soviet annexation. Other major plants are in Peraa, Kaunas, Vilnius, and other
locations (providing a nucleus for electronic industries in the newly indepen-
dent Baltic States). VEF developed the semi-electronic Kvant PBX. Kvant was
plagued by design and production problems. It did not interconnect easily with
the Soviet network. It was used only within bureaucratic administrations' pri-
vate networks and developed outside of Minsviaz' own priorities. It was being
adapted as a rural exchange. After eight years the switch could only handle 62
percent of its designed capacity. The first electronic automatic exchange was
not introduced until 1986, and domestic manufacturers were still producing
step-by-step switches as late as 1988.
Because of the variety of foreign suppliers and the increase in transmission
capacity without modern switching equipment, the Soviet network was poorly
interconnected. Among the major switches used were Alcatel's MT-20 (im-
ported or built under license), the Finnish Nokia's EATS-200, the Czech ATS-
K crossbar, the East German ATS-K, the Yugoslav Metaconta or IOC (licensed
from old ITT technology), and Soviet-made Kvarts and Istok switches. The
Istok switch, a joint U.S.S.R.-G.D.R. effort, exemplified the problems of So-

viet research and development efforts. The switch took ten years to develop
and was stalled by software problems, lack of skilled technicians, and design
flaws. One member of the Istok design team claimed the switch would never
have been produced without the East German pressure to complete the project.
In the end, the Soviet Union entirely depended on East German production by
Robotron and Nachrichtentechnik. Kvarts, a semi-electronic switch designed
for local or transit exchange, was also a joint design of VEF with the East
German's Robotron and the Soviet Union's VEF and TsNIIS.
In 1990, Belgium's BTM Alcatel won a contract to supply almost $1 billion
worth of System 12 digital switches, produced both in Belgium and through
joint production with local companies. Siemens and GPT will also produce
switching systems under joint venture agreements. In 1990, Minsviaz signed a
memorandum of understanding with AT&T to explore cooperation on telecom-
munications services, equipment manufacturing, and R&D.
The Soviet Union helped develop fiber optics in the late 1960s, and it was
used primarily for military applications. The network had only one long-dis-
tance fiber-optic line, running between Leningrad and Minsk and a few local
fiber trunk lines (Gulyaev, 1990). One ambitious part of modernization was a
$500 million trans-Siberian fiber optic line to be built by a Minsviaz and for-
eign consortium of US West, Japan's KDD, DBF Telekom, British Telecom,
Italy's Societa Finanziaria Telefonica, Australia's OTC, and others. The goal,
to provide domestic service and establish a high-capacity land-based link from
Europe to Japan, was stalled by CoCom restrictions against the export of sen-
sitive technology. Behind the American government's opposition was the intel-
ligence community, which did not wish its eavesdropping ability curtailed by a
shift of Soviet communications from radio-based transmission to fiber.
There were two major Minsviaz scientific research institutes, NIIR for radio
and NsNIIS for network technology. R&D efforts also included design bureaus,
research laboratories and higher educational institutes. The overall total of
Minsviaz R&D personnel was 10,000. Other research institutions were outside
Minsviaz and part of the Academy of Sciences of the Soviet Union (e.g., the
institutes for information transfer, microelectronics, cybernetics, and tele-
vision). Many of the research organizations were linked with each other and to
Eastern Europe by the packet-switched network Academnet.
Research and development was the weak spot for Soviet telecommunica-
tions. Bureaucracy, aging administrators, an exodus of talented personnel, poor
coordination, lack of direction, and low scientific productivity of several insti-
tutions have stalled innovation. These points of criticism were made public in
1987 by the deputy head of the Central Committee's department of transport
and communications, I. F. Trofimov. The problems with R&D increased reli-
ance on Eastern European suppliers. Given the changes in Eastern Europe, this
source of supply will not be accessible in the future without hard-currency
transactions.
Reforms
Under proposed legislation, the Russian Ministry of Communications would

become more of a regulator and less of an operator, with responsibility over
frequency allocations, equipment standards, and licensing service providers.
Minsviaz would own the local, trunk, and international network through a hold-
ing company and license competitors for local service, maintaining a monopoly
over international voice service. Coordination among the republics would be
through a committee.
The R&D institutes are being reorganized, put on a self-financing basis, and
given greater independence in their projects.
Joint ventures with foreign firms are encouraged. Research institutes may
also operate as commercial enterprises and compete in services such as packet-
switched networks with Minsviaz. The manufacturing firms formed a cartel
named Telekom.
Given the magnitude of the investments and upgrade tasks, and the reorga-
nization of the former Soviet Union into a new substructure of republics, no
European telecommunications system has a more formidable challenge ahead
in serving the transition to a high-technology economic system.
Ill
BATTLEFRONTS IN
TELECOMMUNICATIONS POLICY
28
The international Organizations of
Telecommunications
PTT Institutions
From the inception of early postal systems, communications were a highly in-
ternational affair. The checkered map of central Europe usually permitted alter-
native postal routes and thus made intergovernmental agreements desirable. As
postal administrations evolved throughout Europe, they prioritized the mainte-
nance of stable international arrangements as a central policy concern. Rooted
in this tradition, "integration" and "harmonization" achieved importance in
the value system of PTTs beyond purely technical needs. They led to interna-
tional organizations, one of whose important function was to shore up interna-
tionally the domestic arrangements and to anchor national monopolies in an
international cartel arrangement. These international organizations worked very
successfully to coordinate the old order. Not surprisingly, they were not partic-
ularly well-suited to be vehicles for reform. Given their traditional role, as well
as their control by traditional PTTs, change was not initiated by the interna-
tional organizations of telecommunications, but essentially outside and even
despite them. Of course, the international organizations served useful functions
as fora for discussion and as structures to channel change. But the main reform-
ing countries were quite prepared to ignore the restrictiveness of the traditional
international regime if necessary. They also began to activate international or-
ganizations associated with interests other than the PTTs, such as the Organi-
zation for Economic Cooperation and Development (OECD), the European
Community (E.G.), and the General Agreement on Tariffs and Trade (GATT).
The International Telecommunication Union (ITU)
The oldest and most venerable of institutions among international organizations

for telecommunications is the International Telecommunication Union (ITU),
founded in 1865 by several European telegraph administrations. It is both an
organization and a model for international collaboration in telecommunications.
Its history traces back to 1850, when several German states created, through
293
294 The International Organizations of Telecommunications
the Dresden Convention, the German-Austrian telegraph union as an extension

of their earlier postal cooperation, in order to coordinate telegraph operations.
A few years later, in 1855, several West European states set up their own
telegraph collaboration. In 1865, a Europe-wide conference of these two group-
ings convened in Paris and established the International Telegraph Union (ITU).
European members were joined by Egypt, Persia, British India, and the Neth-
erlands' East Asian possessions.
The ITU was established as a purely intergovernmental organization, and
despite its claim to primarily coordinate the technical aspects of telegraphy, it
took implicitly political positions by not encouraging the participation of pri-
vate telegraph firms. Thus, Britain's role was limited until the 1870s, when its
telegraphs came under government control; similarly, some American telegraph
firms or the private submarine cable firms were outside the ITU. Britain lobbied
strongly to be included in the ITU meetings and for a while participated through
its Indian colonial administration, which operated a governmental telegraph
system. British interests became concerned that intergovernmental agreements
that would be decided by the ITU without their participation would threaten
Britain's dominant position in submarine cables. This was one argument ad-
vanced in favor of the nationalization of the domestic British telegraph system
in 1870.
The Paris conference that established the ITU resulted in an international
convention and a set of telegraph regulations. The ITU is arguably the first
specialized international organization in any field, although that depends on the
definition. In 1868 it was agreed to establish a permanent ITU secretariat, the
International Bureau. Under the initiative of Heinrich von Stephan, the archi-
tect of Germany's Reichspost, the ITU adopted principles of international rates,
first developed in the 1850 Dresden Convention and modeled after even earlier
postal agreements.
The major European powers controlled the ITU. Not only were these coun-
tries at the forefront of telegraph technology and usage, but they had provided
themselves voting membership through the colonial telegraph administrations
of their overseas colonies. In 1925, France, Great Britain, Italy, and Portugal
each had seven votes in the ITU.
Technical coordination was only one aspect of the ITU activities, and it
turned out, in the early phase, not to be a particularly difficult one (Codding
and Rutkowski, 1982). It was only later that technical standards issues became
increasingly complex, leading to nonbinding standards known as "recommen-
dations." Issues of international rate making (i.e., economic collaboration) were
important from the beginning, and many ITU meetings were spent establishing
rates, arguing over uniform pricing, and agreeing upon the charges for coded
messages. An American observer to the 1875 International Telegraph Confer-
ence in St. Petersburg wrote critically, "The interest of the public who use the
telegraph seemed to be entirely subordinated to the interest of the states and to
the administrations: that is, to a fear lest any improvement (in the rate structure)
might produce less revenue than is got at present, and lest it might throw more
work on the telegraph bureau" (Codding and Rutkowski, 1982, p. 8).
The International Organizations of Telecommunications 295
The Telegraph Union gave little attention to the coordination of the emerging
telephony. Until 1925, it was not even obligatory for the various telephone
administrations to interconnect with other countries. Eventually, national tele-
phone administrations led by the French FIT created a new body, the Interna-
tional Consultative Committee for Long Distance Telephony, initially outside
of the ITU, to address telephone coordination issues.
Radio communications were also organizationally separate from the ITU;
coordination began in 1903 at a conference convened by Germany. Here too
technical issues coexisted with economic and political ones. Various countries
tried to counter the market power in wireless communication by the Marconi
Wireless Telegraph Company, which had instructed its radio operators to com-
municate only with those using Marconi equipment and to refuse all others.
This was an economic and strategic threat to other countries' emerging wireless
operations, manufacturing, and shipping. Germany had embarked on an ambi-
tious naval build-up in rivalry with Britain, and, given the importance of radio
communications for fleet operations, sought to break Marconi's power. Great
Britain and Italy, both of which had Marconi's interests at heart, consequently
boycotted the 1903 conference. The meeting established a protocol that pro-
vided rules of noninterference and interconnection. Another international con-
vention was passed in 1906, again in Berlin, establishing the framework (still
in existence today) of allocating radio bands to particular services to avoid
intermixing. Frequencies could be used on a first-come basis, but notification
had to be given. In 1912, the Marconi Company agreed to stop refusing com-
munications with rival equipment. World War I put an end to collaborative
arrangements. After the war, a 1920 conference drafted many of the principles
upon which today's ITU operates. The most difficult issues, not surprisingly,
dealt with a mechanism for frequency allocation and allotment; no agreement
was reached. Parallel to these efforts, the emergence of private radio broad-
casting led to a meeting in 1925, convened by the BBC, that created an inter-
national radio conference to deal with problems of interference. Eventually, at
a joint meeting in Madrid in 1932, the International Telegraph Union and the
loose International Radio Telegraph Conference were merged in 1934 and re-
named the International Telecommunication Union.
Initially, the Swiss government staffed, financed, and managed the ITU's
office in Berne. But as the ITU became a specialized diplomatic agency to the
League of Nations (and later part of the United Nations), its secretariat was
moved to Geneva and both its staff and financial support were international-
ized. To induce the United States to join, the new ITU convention was kept
general and flexible. A major meeting was held in Washington in 1927. The
U.S.S.R. was precluded from attendance, because of opposition to its govern-
ment. Thus, the alleged "nonpolitical" history of the ITU and U.S. adherence
to such principles should not be exaggerated.
Until World War II, the United States maintained an attitude of benign ne-
glect toward the ITU and its policies. It did not send government delegates or
observers to the ITU (leaving this to private firms) and did not participate in
the international consultative committees when they were formed in the 1920s.
296 The international Organizations of Telecommunications
It opposed the creation of the unified telecommunication ITU since this ex-
tended the potential for an international cartel (Rutkowski, 1982, p. 33). The
United States participated more actively, but still reluctantly, in the interna-
tional radio telegraph conferences, though it made clear its suspicions that they
served to retard the development of radio. The exceptions in the U.S. attitude
were the periods following the world wars, when it was more closely interested
in creating stable international arrangements. For the United States to sign the
telegraph regulations took until 1949, eighty-four years after the first set had
been approved and only after major changes.
The postwar interest of the United States led to Atlantic City conferences in
S947 that reshaped international communications into structures that have en-
dured until today. At the end of the meeting, the ITU framework was in place.
In 1952, an International Frequency Registration Board (IFRB) was created as
a body for the record-keeping of international frequency use and for dispute
resolution. In subsequent years, membership increased dramatically with de-
colonization.
The ITU's system is complex. It includes four secretariats, a multinational
Administrative Council, Plenipotentiary Conferences, World Administrative Radio
Conferences, World Administrative Telegraph and Telephone Conferences, the
IFRB, Regional Administrative Radio Conferences, Consultative Committees
on Radio (CCIR) and on Telephone and Telegraph (CCITT), and droves of
subcommittees and study groups. The ITU currently performs clearing house
activities, coordinates standards and spectrum allocations, and provides tech-
nical assistance to developing countries. Despite the elaborate structure, or per-
haps because of it, the pace of discussions, particularly in comparison with the
rapid advance of technology, has been torpid for a long time, though there
have been exceptions.
Private sector experts, particularly in the equipment industry, can participate
in the Consultative Committees of the ITU. In the American case, a majority
of committees were attended only by private-sector representatives.
The ITU was run in the 1980s by Richard Butler of Australia. Butler ably
managed an unprecedented expansion of membership and scope of responsibil-
ities and initiated reforms to move the organization from a federal to a more
centralized structure. He was succeeded as secretary general by Pekka Tar-
janne, a former physics professor, leader of Finland's Liberal party, and direc-
tor general of the Finnish PTT. A merger of CCITT and CCIR was proposed
but postponed. An independent Bureau for Telecommunications Development
(BDT) was created with status equal to CCITT and CCIR, subsuming the ex-
isting Center for Telecommunications Development, strengthening recommen-
dations made in 1982 in the so-called Maitland Report, "The Missing Link"
(Butler, 1989). Since assuming his post, Tarjanne has advocated greater re-
gional cooperation and a lessening of state control over telecommunications for
economic and administrative reasons.
A majority of ITU members in the 1980s were against any form of liberali-
zation. In this matter, the European PTTs until recently saw eye to eye with
virtually all developing countries and with the Soviet bloc. Hence, for the lib-
eralization-minded countries, the ITU was a basically uncongenial body in which

they are automatically a tiny minority. Nevertheless, the victories of the tradi-
tionalist majority are hollow if the minority, consisting of important high-
technology countries such as the United States, the United Kingdom, Japan,
and increasingly continental Europe, go their own way and are joined by some
developing countries. Partly in consequence, the ITU began to slowly change
its organization in the 1980s (Drake, 1989). For example, it strengthened the
participation of private telecommunications providers, and scheduled reform-
oriented conferences.
The ITU and its sister FIT organization, the Universal Postal Union (UPU),
conduct most of their affairs in closed meetings, in contrast with most interna-
tional organizations, including the United Nations and the Security Council
(Codding, 1984). Between 1947 and 1965 the ITU was unenthusiastically open
to the press. At the 1958 meeting in Geneva, for example, the media were
excluded with the argument that the press, as a customer of telecommunica-
tions, should not witness decisions that affect their business. This argument
was maintained in the 1980s. Its weakness is evident if one considers that in
most CCITT and CCIR committee meetings, private manufacturers, who often
have an enormous stake in the recommendations, participate officially and ac-
tively.
The CCITT
Of particular importance in the telecommunications field is the Consultative
Committee on International Telegraph and Telephone. The CCITT's predeces-
sor was established at the initiative of Britain and France subsequent to a Paris
meeting in 1923. In its first years it mainly concerned itself with the establish-
ment of desirable characteristics for long-distance international telephone lines.
In 1924, the organization was given the name of International Consultative
Committee on Long-Distance Telephone Communications, and it established a
permanent secretariat in Paris. Later it was renamed CCIT (International Con-
sultative Committee on Telephone), which became in 1956 the CCITT. It was
largely separate from the ITU for almost a quarter century. The organizations,
with the International Consultative Committee on Radio (CCIR), were brought
under the umbrella of the ITU in 1949 subsequent to its 1947 Atlantic City
meeting. The technical committees of the CCIR and the International Broad-
casting Union were united at that time.
The CCITT operates through administrations, operating agencies, non-
common carrier enterprises, and international organizations (A. Rutkowski, 1990,
communication). Working groups include representatives of governments and
private operating firms. After 1927 it also operated a laboratory under the name
of the Telephonometric Laboratory. The CCITT issues only recommendations
and has no enforcement power. It does not function as a treaty organization
with binding resolutions. Though developing countries and new electronics firms
have become more active, developed countries and established manufacturers
still dominate the discussion. Government representatives as well as represen-
tatives of fifty-five recognized private companies, several regional telecommu-

nications organizations, and other institutions also participate in the study groups.
Private companies with committee representation must contribute financially to
the cost of the meetings; members of scientific institutions also often participate
(Scherer, 1985).
The role of the CCITT is to develop operational and technical standards and
to harmonize international telecommunications tariff issues. Its expert groups
draft recommendations that are submitted for adoption by the CCITT Plenary
Assemblies held every four years. At the Assembly held in Spain in 1984,
almost 650 recommendations or amendments were submitted for approval. With
growing numbers of equipment firms, applications, and standards committees,
the CCITT's Book of Recommendations grew from 5,000 to 21,000 pages in
the past decade. This forced organizational change in 1988: an accelerated ap-
proval procedure, consolidated study groups, and new internal organization.
This allowed CCITT to deal more expeditiously with standards issues where its
role was being reduced by a regionalization, taking place through Bellcore and
American standards bodies, ETSI in Europe, the Technical Telecommunica-
tions Committee (TTC) in Japan, and other standards bodies, including the
International Organization for Standardization (ISO).
A major conference (WATTC-88) was held in Melbourne, Australia, to deal
with new telecommunications services such as enhanced or value-added ser-
vices, with the notion of protecting them from excessive regulation. It reached
an agreement on integrating telephone and telegraph service under the Interna-
tional Telecommunication Regulations (Drake, 1988; Codacovi, 1989). How-
ever, difficult problems of accounting rates and tariffs were left unresolved.
The CCITT's policy recommendations can clash not only with liberalization
policies of individual countries, but also with other standards bodies, and with
other international bodies and agreements. The European Community's Treaty
of Rome provides for the elimination of restrictions in trade of goods and ser-
vices among European countries. In the 1970s, so-called telex bureaus emerged
in Britain that would route telex messages from continental Europe to North
America through London at a considerably cheaper rate than that charged for
direct service by the European countries' PTTs. CCITT recommendations,
however, prohibit such third-country traffic, and the PTTs attempted to elimi-
nate such telex bureaus by invoking the rules. But as the chapter on U.K.
telecommunications described, the telex bureaus fought back and sued under
the EEC's Treaty of Rome, and eventually won a resounding victory before the
European Commission and the European High Court (Official Journal, 1982).
In 1992, CCITT's head Theodor Irmer proposed to transform it into a private
non-governmental organization.
Intelsat
Intelsat is the international organization providing international civilian satellite
communications. By the terms of the Intelsat Agreement entered into by mem-
ber states, it has exclusivity over these services, though this is under dispute.
The international Organizations of Telecommunications 299
Intelsat was formed in 1964 largely at the initiative of the United States, and
is headquartered in Washington, D.C. At the time, the United States was still
firmly in the camp of monopoly; Intelsat is a reflection of the U.S. desire to
embed its own technical lead in satellites and launchers in an international
regime of coordinated monopoly, subject to a weighted voting that benefits
industrialized countries. Intelsat operates through an Assembly of Parties (rep-
resentatives of government members to the agreement), which usually meets
every two years to discuss long-term issues. There are also Signatories' Meet-
ings in which the designated telecommunications entities are represented. A
board of governors, whose members are representatives of the larger users and
regional groups, deals with budget matters, procurement, and policy issues.
Intelsat business itself is conducted by its executive organ, which is headed by
a director general and employs about 600 people.
Intelsat satellites have evolved seven full and two half generations. The first
Intelsat satellite was Early Bird in 1965, with Hughes as the private contractor.
Its capacity could accommodate up to 240 telephone circuits or one TV chan-
nel. Design life was a brief eighteen months, though it operated satisfactorily
twice as long. In 1967 and 1968 the Intelsat II and III generations were launched,
produced by Hughes and TRW. Intelsat III had a capacity of 1500 circuits or
four TV channels. Multipoint communications were possible, and design life-
time was now five years. In 1971, the Intelsat IV generation was put aloft,
produced again by Hughes, and providing 4000 circuits. The satellite antenna
system was much more advanced and permitted spot beams and other features.
In 1975, the Intelsat IV-A satellite generation expanded capacity to 6000 two-
way circuits. Five years later, the first Intelsat V was launched with Ford Aero-
space as the prime contractor. Capacity doubled to 12,000 two-way telephone
circuits and two TV channels. The system was further improved in 1985 with
Intelsat V-A/B to 15,000 circuits and a variety of antennas. Hughes' Intelsat
VI was launched in 1986, with a capacity of 30,000 circuits and three TV
channels, using a variety of polarizations that make multiple usage possible.
The next generation, Intelsat VII, is ready for the 90s. Most of these satellites
are physically quite large. Intelsat VI measures 6.4 meters high and 3.6 meters
wide. These are very heavy pieces of equipment, densely packed with high-
performance electronics, and they are correspondingly expensive.
Critics contend that Intelsat created an overcapacity in order to reduce incen-
tives for others to enter. It also operates on very-low-power systems, which
require an expensive earth station segment that makes usage by smaller parties
other than the national PTT's less desirable. This also makes it expensive for
smaller countries with low traffic density to participate (Cowhey and Aronson,
1985).
Intelsat's director general in the beginning of the 1980s was Richard Colino,
an American lawyer. In 1986, it was discovered that Colino and a close asso-
ciate had arranged a real estate transaction in which they personally benefitted.
Colino was dismissed from his position and later convicted by an American
criminal court and sentenced to six years in prison. He was succeeded by Dean
Burch, former chairman of the American FCC and an insider in Republican
politics since 1964, when he was the campaign manager for Barry Goldwater's
conservative candidacy. Burch died in 1991. In 1990, Intelsat was criticized by
competing operators, who claimed it had set monopoly prices and refused to
use the most cost-effective technology. PanAmSat filed a $1.5 billion lawsuit
against Comsat, charging anticompetitive behavior.
Faced with added capacity and dropping prices, Intelsat's revenues stagnated
in 1989 and dropped almost by 20 percent in 1990. Jolted, this organization of
(largely) government organizations began to consider selling 49 percent of its
shares to the public, and its new head Irving Goldstein sought renegotiation on
launch contracts.
Policy issues involving Intelsat are discussed in greater detail in the chapter
on international communications and new entrants, both satellites and fiber cables.
CEPT
On the European level, the Conference of European Postal and Telecommuni-

cations Administrations (CEPT) provides PTT coordination. CEPT was estab-
lished in 1959 as the organization representing the interests of European PTTs
and included the telecommunications administrations from the twenty-five West
European countries, Hungary, and Yugoslavia. Other Eastern European coun-
tries were to join after 1990. It operates by annual plenary meetings and com-
mittees that are responsible for specific issues, such as transatlantic telecom-
munication, harmonization of telecommunications, commercial action, data
transmission, and satellites (Quander, 1982). Since it synthesized the views of
the powerful European PTTs that subsequently dominate worldwide CCITT
proceedings, CEPT wielded substantial power. CEPT was characterized by an
insider modus operandi that reinforced its image as a very conservative orga-
nization, concerned with enforcing a European telecommunications cartel. In
theory, CEPT made only recommendations. In practice, however, these pro-
posals, which are often quite detailed in the areas of technology, were mostly
followed (Le Boucher, 1984).
CEPT meetings too were closed to the press, and its working papers unavail-
able to the public. It argued that public meetings would politicize expert bod-
ies, reduce the free exchange of opinions, and lead to distorted reporting (Scherer,
1985). Even industry participation in CEPT deliberations was exclusive. For
example, the Bundespost coordinated its position on issues pending in CEPT
within a German committee, whose meetings until 1983 included only the clas-
sic telecommunications equipment industry. It took repeated attempts for the
Office and Information Technology group of the German Industry Association
to be included in the body also.
In 1984, CEPT adopted some measures to set technical norms and to open
the national European markets. These standards, established by a commission,
would gradually become obligatory. But the U.S. complained that PTT control
of CEPT foreclosed industry involvement, especially by foreign firms, in the
standards-setting process (Dougan, 1987).
In 1990, under pressure from the European Commission and telecommuni-
cations users, CEPT's Telecommunication Commission accepted the need to

separate network regulation from network operation. A new CEPT regulatory
committee sets policies relating to all activities apart from the network agree-
ments among its member PTOs. CEPT also agreed to shed some of its secrecy.
CEPT was also challenged in its economic coordination role. When it recom-
mended a tariff structure for leased lines for large users that opt for digital
technology that would reaquire a 30 percent access charge for connection to
the public network or carriage of third-party traffic, the European Commission
attacked the arrangement.
In 1989, the E.C.'s directorate for competition policy (DG IV) began an
inquiry into the legality of CEPT tariff recommendations. It found the CEPT
tariff plans to be cartel actions under Article 85 of the Treaty of Rome. Con-
sequently, CEPT withdrew its recommendations in 1990.
Following the tariff debate and the E.C.'s creation of the European standards
body ETSI that was outside CEPT, CEPT instituted a set of reforms, trans-
ferred most of its standards activity to ETSI, created a Commercial Action
Committee to deal with user demands, and reorganized its spectrum manage-
ment division. Most CEPT member organizations formed the European
Telecommunication Network Operators group (ETNO) as a carrier associa-
tion to represent their interests. They also formed the European Institute for Re-
search and Strategy Studies in Communication (Eurescon). This institute,
located in Heidelberg, Germany, has similar functions as Bellcore in the United
States.
Eutelsat
Eutelsat was created in 1977 with an objective of creating and operating a
European system of satellite telecommunications. The organization is head-
quartered in Paris. It includes an Assembly of Signatories, which sets general
policy and objectives for the organization and addresses questions related to
external organizations and the division of markets. Another agency, the Euro-
pean Space Agency, deals mainly with the coordination of satellite develop-
ment. It places satellites in orbit for Eutelsat.
Although established for telecommunications, Eutelsat's main business be-
came the transmission of television programs to cable networks and national
broadcasters. In 1990, telephony and business telecommunications accounted
for a quarter of revenues, and television broadcasting for the rest.
The newer satellite generation, Eutelsat II, in operation after 1990, has 60
percent more communications capacity than the Eutelsat I series. It has sixteen
television transponders at 50 watts of power. Eutelsat II is built by a consor-
tium of Aerospatiale and Alcatel Espace, of France; Aeritalia, of Italy; CASA,
of Spain; Germany's MBB; and Sweden's Ericsson. Eutelsat was shaken in
1991 by the B.C. Commission's intention to seek an "open skies" policy that
would free users to interconnect into other satellite systems and deal with Eu-
telsat directly rather than through a PTT.
International Institutions Outside the PTTs
The emergence of telecommunications policy discussion in other international

organizations such as the European Community and the OECD is a reaction to
the often narrow perspective of the PTTs in their own international bodies,
ITU, CCITT, CEPT, Intelsat, and Eutelsat. It is an extension to the interna-
tional level of domestic conflicts that exist in many advanced countries between
the PTTs, on the one hand, and the ministries of economics or industry and of
antimonopoly agencies, on the other. Generally speaking, the ministries of eco-
nomics or industry see telecommunications issues in the broader context of
high-technology development and national competitiveness of the information
sector.
The Organization for Economic Cooperation and Development (OECD) has
displayed particular initiative on matters of information policy, where it took a
role in identifying and discussing the issue of transborder data flows, privacy,
and national sovereignty. It has also issued several influential reports discussing
the problems of national compartmentalization of telecommunications, and
comparing services, rates, and performance in various countries. The OECD's
important role is referred to in other parts of this book.
Another international body into which telecommunications issues have spilled
is the General Agreement on Tariffs and Trade (GATT). GATT goes back to
1947 attempts to re-energize world trade after World War II and avoid protec-
tionism and bilateralism. The new framework of fairly open trade was remark-
ably successful. However, it applied only to commodities, and not to services.
In 1986 GATT had only two full-time professionals dealing with services. Starting
in the 1970s, several major U.S. companies such as insurance and tourism
firms began to advocate inclusion, and to ally themselves with the weakening
free-trade coalition that included, among others, retailers and farmers. Tele-
communications and information firms joined, too, but construction and ship-
ping companies were opposed to foreign competition. Other countries resisted,
but did not wish to provoke retaliatory restrictions of their nonservice exports
into the large American market. In 1980, the U.S. demanded GATT treatment
for services. This was rebuffed in 1982, primarily due to Third World opposi-
tion. In time, Great Britain, Japan, Canada, and Sweden supported a services
trade negotiation but continental European countries were mostly opposed. France
eventually began to support the negotiations, and GATT decided in its 1986
Punta del Este meeting to proceed. When telecommunications were introduced
into the topics under negotiations in the "Uruguay round," traditional interna-
tional organizations such as the CCITT were unenthusiastic, viewing it as either
a trespass by free-marketeers and American business interests, or at least, the
application of trade policy to what should be sovereign regulatory policy. Is-
sues for negotiation were reciprocal market access; national treatment; most
favored nation treatment; right of establishment; access to local distribution;
competition and transaction with government monopolies; free flows of infor-
mation; and nondiscriminatory technical standards and pricing policies (Aron-

son and Cowhey, 1988). Perhaps the most important issue was the interconnec-
tion of networks across frontiers (Transnational Data and Communications
Report, May 1988, pp. 18-22; Drake, 1988).
Four years of negotiations followed. A broad but vague consensus was reached
on VAN services, which were to be granted access to the public network,
though with considerable national discretion, in a cost-based, transparent, fair,
and unrestricted fashion. But no similar consensus was reached for basic ser-
vices and facilities. Here, the original U.S. position had been pro-liberaliza-
tion. However, U.S. international carriers, large user organizations, and even
the federation of consumer organizations began to oppose that position and the
policy of giving countries the ability to exempt certain services. The groups
lobbied instead for bilateral agreements, fearing asymmetric access opportuni-
ties due to the more deregulated U.S. market, as well as concessions in tele-
communications in favor of other industries. AT&T in particular concluded that
"national treatment" in a restrictive country would not give it any opportuni-
ties, in contrast to foreign countries which would be able to operate freely in
the United States.
This view complicated the American bargaining position, which had always
been hampered by vague linkages to financial and human services that had not
been carefully thought through. The U.S. position, in effect, switched to a
priority of regulatory policy in telecommunications (i.e., national liberaliza-
tions) in preference to trade policy and bilateral trade agreements.
As tortured as the progress in the telecommunications services basket was,
it was actually the most developed among the GATT service trade negotiations.
Agreements on tourism, professional services, construction, finance, and bank-
ing and transportation were far behind telecommunications.
Telecommunications was part of a much larger international trade package,
which included textiles, patents and copyrights, and agriculture, among others.
The package collapsed in 1990, largely due to disputes over European agricul-
tural subsidies.
Other Standards Setting Bodies

The International Organization for Standardization (ISO), whose Committee 79
sets standards for data-processing systems, computers, and office automation,
is a particularly important organization. Another international standards body
is the International Electrotechnical Commission (IEC), joined with the ISO,
which also deals with information technology equipment. European standards
coordination also takes place in the Comite Europeen de Normalisation (CEN)
and the European Committee for Electronic Technology Norms (CENELEC).
CEN/CENELEC is mostly geared toward development of OSI protocols. Two
other OSI standards groups are SPAG (Standards Promotion and Applications
Group), a consortium of major European electronics firms, and EWOS, the
European Workshop for Open Systems. Another standards group, the European
Computer Manufacturers Association (ECMA), was founded in 1961 with the
participation of US and European firms. ECMA, which specializes in private
networks, works more quickly than its bureaucratic cousins and frequently sub-
mits its reports to CCITT and ISO for adoption (Gibbons, 1989a).
29
Brussels Takes On the
Traditional System
Until recently, telecommunications matters were outside the jurisdiction of the

European Community and its commission. The Commission's first attempt in
the 1960s to play a role in the field was unsuccessful. In 1978, Commissioner
Count Davignon introduced a telematics initiative at the EC summit meeting in
Dublin. There was little activity on this issue until 1983, when the commission
accepted recommendations that it play a major role in the high-technology and
information industries. It concluded that the nationally protected environment
that fragmented European telecommunications was an important factor in Eu-
rope's weak competitive position relative to the United States and Japan. This
led to decisions that initiated the ESPRIT high-tech technology development
program. Although ESPRIT excluded hard-core telecommunications, it laid the
foundations for the RACE program that the commission established in 1985
specifically for telecommunications. The EUREKA program, another Europe-
wide support program, also grew from these beginnings. Davignon wanted to
go considerably further and advocated a European telecommunications agency
to enhance equipment production and service supply.
The European Commission's growing role in telecommunications followed
two paths that can easily be at tension with each other: antimonopoly policy
and industrial policy. The antimonopoly actions, in addition to the repeatedly
mentioned British telex case, which involved the abuse of BT's dominant po-
sition by imposing restrictions on telex message-forwarding agencies, also soon
included challenges to the following: the Bundespost's intent to extend its mo-
nopoly to cordless telephones; the extension of the Luxembourg broadcasters
to telemarketing into neighboring markets; the Bundespost's exclusive market-
ing and ownership of computer modems; the extension of the Belgian RTT
monopoly to include medium-sized PBXs; excessively long-term leasing con-
tracts for telephone equipment; and IBM's extension of its Systems Network
Architecture for interconnecting computers and other equipment (requiring it to
publish formats and protocols) (CEC, 1986), the PTTs' joint Managed Data
Network Services project, and the price cartel in international telephone rates
and in leased line tariff principles.
An example of the E.C.'s successful action involved the pricing of interna-
tional lines leased to SWIFT, the international interbank electronic transfer net-
305
306 Brussels Takes On the Traditional System
work. CEPT had repeatedly changed the rules applied to SWIFT with the aim
of recouping revenue lost to telex service. SWIFT, not a meek organization
itself, lodged a complaint with the commission charging a price cartel and
abuse of individual and collective dominance. Negotiations then took place that
ultimately resulted in the lowering of its rates.
Another case was a 1986 complaint by the French DOT against the German
Bundespost. The DOT charged the Bundespost with restraint of marketing when
it did not accept the French Minitel videotex terminal, preferring to keep
the distribution of more integrated models for itself (Schulte-Braucks,
1986).
In 1984, the member states of the European Community agreed on the details
of a European telecommunications technology policy. The B.C. proposals in-
cluded the following: definition of medium- and long-term European policy
objectives; a new forum for European telecommunications issues; collaborative
European technological projects; common equipment interconnection standards;
a common European front towards outsiders; opening of national European
markets to unbiased procurement from other European countries; and collabora-
tion in the development of new transnational services, such as ISDN, broad-
band networks, and satellites.
France was initially active in pushing for a joint European telecommuni-
cations policy. Given its own high standard of telecommunications technology
within Europe, it saw itself as a potential beneficiary. The French PTT minister
Louis Mexandeau argued for a reciprocal and open European Community mar-
ket, which could also include the non-European Community members of CEPT.
His plan specified a delay in the reciprocal opening of the market with the
United States, Canada, and Japan.
In 1987, eleven out of the twelve E.G. countries (with the exception of
Greece) were separately and actively reassessing their telecommunications pol-
icies. Assuming a responsibility for preventing diverging paths and seeking to
accelerate the process, the commission issued a Green Paper aimed at "achiev-
ing maximum synergy between current developments and debates within the
Member States" and "to launch a debate" (CEC, 1987, p. 2). The paper's
recommendations included a phased opening of competition in the terminal
equipment market. This was, in part, merely an acknowledgment of gray-mar-
ket liberalization. Accordingly, the paper sought mutual recognition of type
approvals for terminal equipment and mutual opening of procurement contracts.
Most importantly, the paper asserted unequivocally that the (liberal) competi-
tion rules of the Treaty of Rome apply to telecommunications, in particular
where PTTs engage in commercial practices rather than basic 'reserved' ser-
vices. It called for the creation of a European telecommunications standard
body to set common standards.
The paper's second recommendation aimed at the unrestricted provision of
competitive (in particular nonvoice) services both within and among member
states, including an obligation for PTTs to provide access to transfrontier value-
added networks. This "open network provision" would permit such a network,
once licensed in one country, to operate throughout Europe. Such liberalization
Brussels Takes On the Traditional System 307
would require agreement on definitions, termination points, tariff principles,

and so on.
The paper further recommended a clear separation of the regulatory and op-
erational functions of telecommunications administrations, which "cannot con-
tinue to be both regulator and market participant (i.e., referee and player)"
(CEC, 1987, p. 17). It also warned of cross-subsidization of competitive activ-
ities and cautioned against excessive restrictiveness toward the use of private
lines. Regulatory "methods must be limited to a legitimate level of protection
of financial viability and must not represent the abuse of a dominant position."
It observed that charges for leased lines among member countries "show in
some cases wide and unexplained divergences."
The PTTs' dominance over network infrastructure and over basic services
was accepted in the commission's Green Paper. However, the document con-
veyed the message that the PTTs' efforts at protecting their turf were counter-
productive to European high-technology capability and economic competitive-
ness and that change ought to be encouraged. Instead of letting inter-European
harmonization be used defensively as in the past, the commission now wanted
to initiate the reverse and let supranational coordination create a force for change.
Thus, the Green Paper was an important step, despite the modest scope of its
proposals. It marked a signal that the days in which PTTs defined the public
interest in telecommunications were drawing to a close.
As the European Community approached 1992, the date for widespread har-
monization, the role of the commission in setting telecommunications policy
grew increasingly contentious. The commission repeatedly used Article 90 of
the Treaty of Rome, which permits the bypassing of national governments (i.e.,
of PTT authority). In 1988, Italy, Germany, and Belgium joined France in
unsuccessfully filing suit against the commission in the European Court of Jus-
tice over the use of this article in telecommunications. But an unrepentant com-
mission kept using Article 90 to extend its guidelines for open competition in
terminal equipment markets to telecommunications service provision (CEC, 1988).
It also presented plans for open network provision, including harmonization of
tariffs, network interconnection, technical standards, quality targets, common
numbering, and a common ISDN policy. It was increasingly evolving into a
European regulatory agency.
In 1988, the commission, in the 'bloodless coup' of its Equipment Directive,
took standards setting authority from CEPT's notoriously slow Technical Rec-
ommendations Applications Committee (TRAC). ETSI, the European Telecom-
munications Standards Institute, was established in Sophia Antipolis, near Cannes
in France. PTTs' users and manufacturers are included in ETSI's General and
Technical Assemblies. Over 135 organizations had joined ETSI by 1990 (Be-
sen, 1990).
The commission also protested the activities of non-E.C. organizations. It
won the elimination of two CEPT tariff plans that allowed a 30 percent access
charge on international leased lines, on the grounds that it amounted to price-
fixing. That, on top of the creation of ETSI, significantly reduced the regula-
tory and cartel powers of CEPT.
The commission's approach to liberalization of terminal equipment by liti-

gation was supplemented by gradual Directives for equipment and services. In
1988 it required first telephone sets to be liberalized by June 30, 1990. But
when that date arrived, monopoly control over modems and telex terminals still
existed in Spain and Belgium, which also joined Ireland, Greece, and Italy in
retaining PTT monopoly rights to first sets. Denmark's PBX market was closed,
and France was accused of delaying type approvals for foreign companies
(Schenker, 1990). The commission challenged Belgium's and Spain's equip-
ment rules. It also challenged cartel-like coordination of PTTs for private net-
work services.
In 1989, the European Community reached agreement over data services,
after France along with other southern E.G. countries had sought to block data
service competition by requiring national licensing. National licensing condi-
tions were retained, but they became subject to approval by the commission.
To assist VAN access, the commission passed open network provision (ONP)
guidelines but retained the power to set mandatory standards (ITU, 1990).'
In 1992, the commission even edged toward establishing general tariff prin-
ciples, including price caps, acceptance of discounts to large users and the
needy, and a common stance on international settlements reform. And it pro-
posed a European licensing and regulatory agencythe Community Telecom-
munications Committeebased on national authorities.
Development Programs: ESPRIT, RACE, EUREKA
The second type of policy pursued by the European Community was to en-
courage technology development. Several joint projects were initiated. In ad-
dressing domestic protections, the commission argued that the Common Market
represents 30 percent of the world market for information technologies but that
it produces only 15 percent of such products. The European Strategic Program
of Research and Development in Information Technology (ESPRIT) whose goal
was to triple the E.C.'s share in the world market by 1990, was initiated to
alleviate this imbalance.
ESPRIT was established in 1984, seeking to promote cooperation between
European enterprises, research institutes, and universities. The total budget for
the first four years was $1.2 billion. A report of the French Senate delegation
for the European Communities noted, somewhat breathlessly, that, "The pro-
gram is a chanceperhaps the lastfor the countries of Europe to rejoin the
pack at the head of the industrial nations in new technologies, and to therefore
maintain technological autonomy . . ." (TDR, 1985, p. 49). Major European
manufacturing companies, including Bull, CGE, Thomson, ICL, Plessey, GEC,
Siemens, Nixdorf, AEG, Philips, and Olivetti supported ESPRIT, hoping for
its largesse.
In its first year, ESPRIT selected 90 projects out of 441 proposals, matching
industry financing with equal subsidies for projects joining firms from at least
two E.G. countries. Although small companies and universities gained, the
prime beneficiaries were the established large electronics firms.
ESPRIT has five main research themes: microelectronics, software, advanced
information processing, office electronics, and integrated computerized design
and production. Biotechnology, nuclear energy, and processing of radioactive
wastes were added later. Because of various turf-battles, telecommunications
policy was conspicuously missing, except indirectly through components and
applications.
To deal with this absence, the European Commission established the Re-
search and Development in Advanced Communications Technologies for Eu-
rope (RACE) program to target telecommunications issues. The program's re-
search priorities are very-high-speed integrated circuits; high-complexity inte-
grated circuits; broadband switching; fiber-optic components; components for
long-distance, high-power links; specialized communications software; and large
format, flat-screen display technology (Telediffusion de France, 1985).
A major goal for RACE is to develop a European integrated broadband com-
munications network (IBN) by 1995. Given its planned and periodically uncer-
tain budget, only a relatively minor amount could be spent at first on develop-
ment. In its initial phase, RACE defined a European model of IBN and provided
the necessary equipment and cooperation methods among different firms and
countries. The main phase, from 1986 onward, includes field trials (systems),
technology development (technology), and standardization (integration) under-
taken in integrated broadband technology to provide a "framework for subse-
quent competitive product development" (CEC, 1987). The commission con-
tributed $500 million.
In 1990, the European Commission considered a plan to accelerate RACE
and achieve pan-European broadband communications as early as 1993. Under
the three-phase plan, broadband "islands" would be linked by existing 140
Mbps fiber optic lines. By 1994, the system would connect about 100 compa-
nies. Starting in 1996, fiber-optic cable would bring residential subscribers into
the broadband network.
EUREKA, the European Research Coordination Agency, is a third joint Eu-
ropean development program and is based on an initiative proposed by French
President Mitterrand. EUREKA concentrates on six broad scientific areas: op-
Ironies, high-speed microelectronics, large computers, artificial intelligence, high-
power laser and particle beams, and new materials.
For each of these areas, a management committee is made up of members
from governments, industry, and research institutes. Financing is shared by the
national governments and the firms. EUREKA's goal was defined as the devel-
opment and potential manufacture of marketable products, with an eye toward
the perceived major European weaknessnot research, but its translation into
successful products. EUREKA is not specifically a project of the European
Community. Non-E.C. nations also participate in the program.
EUREKA was stressed as the European and civilian alternative to the Amer-
ican militarized research effort. Skeptics like The Economist described EU-
REKA as "[promising] to take more money from taxpayers to bribe rich Eu-
ropean companies to do the sort of R&D they should be doing anyway if they
want to stay in business."
Another program, the Special Telecommunications Action for Regional De-
velopment (STAR), began in 1988 to assist development of telecommunica-
tions services in less-developed E.G. nations, primarily Portugal, Greece, Ire-
land, Spain, and southern Italy (Lalor, 1987). The JESSI program (Joint European
Submicron Silicon Initiative) was developed to provide $4 billion in financing
for chip makers.
Data Base Development

The distinctions between organizations that coordinate policies and those estab-
lishing joint operations is a fluid one. Euronet and DIANE are in the latter
category. Euronet, a packet-switched data transmission network, was run pri-
marily for E.G. countries by the European PTTs. Its main purpose is to provide
international on-line database access, specifically to DIANE (Direct Access In-
formation Network for Europe).
Euronet was decided upon in 1971, and it was set up by 1979. To remove
locational advantages, the Euronet tariff structure was distance-independent.
Total cost to the European Community was approximately $25 million, about
two-thirds of the required budget. The PTT administrations contributed a smaller
amount (CEC, 1982). Once the system was operational, the European Com-
munity terminated its financial involvement in the project.
The Euronet-DIANE system's role was multifold. First, the projects aimed
to establish a European Community-based public data network specifically for
on-line information retrieval. Not only would this make intra-European flows
of information easier, but it would encourage the use of information by Euro-
pean scientists and business firms and government institutions. Second, it sought
to establish standards that would permit the interaction of users and multiple
databases within Europe by creating similar database standards, switching prin-
ciples, and a unified command language. In this respect, the network was a
success. Third, Euronet aimed to establish and promote the concept of data-
bases, to create a Europe-wide market for them, and to encourage the setting
up of host computers for their use.
The European Community's statistical office also established the CRONOS
data bank, which includes economic data from various countries. Another inter-
European information project is INSIS (Inter-Institutional Integrated Services
Information Systems), which aimed to provide information links between insti-
tutions of the European community and its member states (Van Rosenthal, 1983).
The E.G. also ventured into the electronic publishing field, subsidizing a num-
ber of experiments by a consortium of thirty-five publishers, software manufac-
turers, and data processing firms. And a $130 million program to develop in-
formation services was launched, as well as a program to fund personal computer
acquisitions in less developed regions of Europe.
Space Development
In 1975, eleven European countries created the European Space Agency (ESA).
Its origins began with European efforts to catch up with the American and
Soviet space programs. In 1960, the European Preparatory Commission for
Space (COPERS) was established, and in 1964, the conventions for the Euro-
pean Space Vehicle Launcher Development Organization (ELDO) and the Eu-
ropean Space Research Organization (ESRO) were signed. These organizations
were forerunners of ESA (ESA, 1984).2 ELDO developed a European booster
using British, French, and German rockets as stages and launched them from
Woomera, Australia. Three European scientific satellites were developed and
launchedby American rocketsin 1968.
ESA began operations in 1975, when eleven European countries signed its
convention. It moved into the development of weather forecasting and telecom-
munications satellites, and in 1985 launched a mission to explore Halley's comet.
Meteorological satellites were first launched in 1981 and operated by the Eu-
ropean Meteorological Satellite Organization (EUMETSAT), established in 1983.
In 1983, the American space shuttle launched Spacelab, the European space
laboratory, which carried the first ESA astronaut into space.
The Ariane space launcher program was started in 1973. In 1979, the first
successful test flight was completed, followed by consecutively more powerful
Ariane generations. Ariane IV can lift 4 tons into a stationary orbit. Technical
responsibility for the development of the launcher lies with CNEE, the Centre
National d'Etudes Espace. The firm Aerospatiale coordinates its systems inte-
gration. Arianspace operates commercial launches, and competes with NASA
for customers in the launching of telecommunications satellites. NASA's space
shuttle requires launch to a low orbit in a manned vehicle, from where a second
launch is undertaken to reach the geostationary orbit. In that second boost, a
number of problems occurred. Both organizations charge each other with unfair
competition by subsidized launches, and it is difficult to determine what the
"economic" cost is, particularly in a multipurpose mission such as a manned
space shuttle flight.
In 1983, the Ariane rocket launched the first European Communications Sat-
ellite (ECS). The ECS system, in turn, is operated by Eutelsat, the satellite
organization of the European PTTs. A new generation of European satellites
was approved in 1984 under the name of Apollo.
Within Europe, some of the larger countries felt reluctant to share their tech-
nology and its application with other countries. For example, France and Ger-
many did not develop broadcast-strength satellites through ESA, but instead
collaborated and produced them (joined later by Sweden).
30
Telecommunications Policy as
Industrial Policy
Convergence?
It is possible to view international telecommunications as merely a hodgepodge

of national systems, each reflecting its society's history and economics, and
each happily self-contained except for collaboration on technical issues. It is
equally possible to see national developments as variations on a single technol-
ogy theme, inexorably driven by an underlying technological determinism, such
as the convergence of telecommunications and computers. Or one can argue,
as this book does, that common and often non-technical dynamics are chang-
ing the institutions of telecommunications: On the demand side, the sharing
coalition of network users is breaking down, whereas on the supply side, the
coalition in support of the monopoly system is eroding. For a long time the
traditional arrangement was remarkably stable, successful, and undisputed. For
a number of years, however, it has been subject to forces of disintegration.
Although it was at first possible to dismiss change as the policy initiatives of
an economically conservative America, later events in Britain, Japan, the Neth-
erlands, and developing countries such as Pakistan and Malaysia, and even a
mainstay of the traditional system such as Germany, suggest a broader trend.
The question of convergence in economic systems has been extensively dis-
cussed in reference to the development paths of Western and socialist countries.
One argument for predicting convergence was that similarities in basic eco-
nomic processes, in particular the prerequisites of large scale manufacturing,
must lead to fundamentally similar structures: large production organizations,
centralized planning, and large social programs (Wiles, 1960). The culture,
ideology, and technology of industrialized societies may differ in early stages
of their development. But industrialization, it is asserted, is a standardizing
force that creates common institutional arrangements.
This view goes back to Marx and Engels, who interpreted society and tech-
nology as different aspects of the same process of social production, or with
certain social and institutional inevitabilities. In the United States, Thorsten
Veblen saw as one of the central tensions in modern economic society the
interaction of the market system with dominant engineering and science. Sim-
312
Telecommunications Policy as Industrial Policy 313
ilarly, many dominant European intellectuals believe that "technique has be-
come autonomous; it has fashioned an omnivorous world which obeys its own
laws and which has renounced all tradition" (Ellul, 1964, p. 14).
Western Europe and the United States are of course diverse in their economic
institutions, as are European countries among themselves. But if one looks at
the first derivativethe changerather than at the absolute, the United States
and Europe exhibited roughly similar trends for more than half a century. After
a relatively free-wheeling 1920s, the interventionist role of the state increased
considerably in the 1930s on both sides of the Atlantic under the effect of the
Great Depression. World War II led to a still increased governmental role. In
the 1950s, Europeans and Americans pursued an essentially conservative pro-
business policy in economic development, coupled with adherence to the wel-
fare state programs of the 1930s and late 1940s. Central to the economic sys-
tem of the postwar years was reliance on the large business corporation. In the
1960s, more liberal or social democratic trends emerged on both sides of the
Atlantic, followed by a tumultuous period of internal unrest.
In the mid-1970s, however, the economic theories of the Chicago School
a noninterventionist philosophybecame influential among economists and
provided the economic legitimization of a policy that called for a reduced gov-
ernmental role. During that period and not unrelated, American economic de-
velopment shifted from the traditional centers of production, primarily in the
northeast, to the west and south. The new firms responsible for the shift were
smaller or medium-sized, frequently run by entrepreneurs rather than by cor-
porate bureaucracies, and hostile to unionization. Meanwhile, the American
government had been substantially discredited by the disasters of Vietnam, Wa-
tergate, inflation, and social problems despite the reform programs of the 1960s.
These interacting forces contributed to America's move toward an increasing
withdrawal of the state from economic control.
In Europe, the state had not squandered public trust to the same extent.
European intellectuals were not moving generally to the right as in America,
though there were increasingly exceptions. European societies were comfort-
able with large bureaucratized firms; a pro-competitive policy had genuine sup-
port only from small liberal parties of the center. A further stabilizing force
was environmentalism. "Green" parties called for strict governmental regula-
tion and specifically advocated control over technology-driven change.
In continental Europe, corporatism as an ideology is palatable both to the
political right and left. Accordingly, in Europe the state has frequently acted to
encourage consolidation in order to make national industry viable in world
competition. In the United States however, antitrust policy, rooted in anti-big-
ness populism, was important for a century, even though its enforcement had
its ups and downs.
Thus, the 1980s witnessed a marked divergence of the paths of European
and American high technology development: In Europe, large firms were more
important than ever; in the United States, the giants were attacked by small
competitors and challenged by various take-over financiers, corporate empire-
314 Telecommunications Policy as Industrial Policy
builders, and government. The telecommunications and information sector pro-

vides a good illustration for this divergence. In the same decade that AT&T's
power declined, Silicon Valley was in ascendancy.
In Europe, economic growth became a major concern to governments and
began to affect telecommunications policy. In the decade from 1963 to 1973,
B.C. economies grew at a healthy average annual rate of 4.6 percent. In the
next ten years, annual growth slowed to 2 percent, and in the early 1980s it
practically ceased, as did the previously steady increase in individual income
and social services. For the first time in a generation, unemployment rose and
living standards stagnated and sometimes even fell. The slowdown of economic
expansion meant a move toward what Lester Thurow called the "zero sum
society," in which social tensions are sharpened by one group's gain being
another group's loss.
Though growth rates rose again, the concern of the early 1980s led to indus-
trial development policy receiving priority in most European countries. In mar-
kets for simple, mass-produced goods, heavy manufacturing products, and other
industrial mainstays, the European and North American economies were slip-
ping behind low-cost, high-volume producers, particularly those of the Far East.
There was no reason to assume that this was a temporary situation to be recti-
fied as soon as these countries' wage levels increased. Other low labor-cost
countries were waiting in the wings, eager to assume a role similar to that
played by Hong Kong, South Korea, Taiwan, and Singapore.
The main competitive advantage held by European and North American
countries is superior knowledge and experience in technical matters, which al-
lows them to develop products whose manufacture requires a substantial degree
of manufacturing know-how and sophistication. Rather than "brawny" goods,
such as steel, "brainy" or science-based production is better suited to main-
taining competitiveness. Also, such production better serves the domestic pref-
erences of countries that have developed more middle-class and more environ-
mentally conscious societies. Electronics and information are therefore an ideal
sector, since it is sophisticated in terms of design, production, and application
and can provide employment for the large number of university graduates gen-
erated by the industrial countries. Over time, success in this field has become
an indicator of international standing. Although it is acceptable to buy steel,
grain, or watches from abroad, the importation of electronic brains seems to
touch a sensitive nerve, especially since success or failure are not based on
locational or geological coincidences.
The path of technological development is littered with lofty but failed na-
tional prestige projects. One example is the World Center for Computer Sci-
ence and Human Resources (Centre Mondial de 1'Informatique et Resource
Humaine), which opened in Paris in 1982. President Francois Mitterrand per-
sonally supported the venture, proclaiming as its goal an understanding of the
computerization of society and the development of computer applications for
social and economic growth, particularly in the Third World. A noted journalist
and politician, Jean-Jacques Servan-Schreiber, was the primary French organi-
zer, and several eminent American computer scientists from the Massachusetts
Institute of Technology (MIT) were chosen to direct research. The project would
fuse American technical know-how with European social concern and human-
ism in the service of society and the Third World. It was generously financed
by the French government and was lavishly presented to the public (Etheridge,
1983).
Yet within one year the center was paralyzed and the Americans had left.
What had gone wrong? Among other reasons, the American scientists and the
French politicians had different conceptions of the center. The French wanted
to demonstrate France's national commitment to technological leadership. The
MIT technocrats took the humanist mission of the center more seriously and
favored a democratic style of management, which clashed with the style of the
French government officials who controlled it. Instead of becoming a world
forum with an international outlook, the center became a backwater catering to
domestic manufacturing interests.
This episode illuminates the dilemma facing European leaders. They are aware
of the importance of this sector, they want to attain rapid results, and are will-
ing to commit money and prestige. In the end, however, their efforts are ham-
pered by fundamental constraints: the self-interests of bureaucracies, the narrow
perspective of domestic manufacturers, and technological nationalism, even in
the European context.
The History of the Electronics Gap
Despite Europe's proud scientific and technological traditions, well-functioning

R&D infrastructures, sophisticated users, and large financial markets, European
views of a role in electronics were often colored by a surprisingly Spenglerian
pessimism.
This had not always been the case. Even after World War II, the United
States and Europe were at similar levels of technological development, with
Japan trailing. The war had accelerated major electronics R&D projects in
countries such as Britain and Germany, and although many German production
facilities were destroyed or dismantled, the technical know-how was there. Eu-
ropean firms were as advanced in tube technology as their American counter-
parts, and they were performing research in solid-state technology, including
research that led to the development of semiconductor diodes.
In late 1947, the transistor was invented at Bell Labs, and its superiority
over traditional receiving equipment was soon apparent. On both sides of the
Atlantic, large established tube manufacturers quickly expanded into transistor
manufacturing. And although Americans had a head start and a larger and richer
market, European companies managed to keep pace with the new develop-
ments. Philips, with its various European subsidiaries; Siemens; AEG-Telefunken;
Plessey; Ferranti; GEC; and Lucas were all doing quite well technologically,
and many had close development links with American manufacturers or with
each other. By the late 1950s, Philips and Siemens were particularly strong,
innovative, and internationally competitive (Malerba, 1985).
Much more significant than war and devastation in creating a disparity in

high technology was the different evolution of market structures on the two
sides of the Atlantic. In the United States, the needs of military, space, and the
emerging computer sectors were met not so much by the traditional tube man-
ufacturers as by new firms. Furthermore, computer manufacturers themselves
moved into component production. In Europe, meanwhile, much of the devel-
opment of the integrated circuit technology was undertaken by the traditional
large tube manufacturers, which were slower to innovate than their American
and later Japanese rivals. American producers gained an advantage that trans-
lated into larger market shares in Europe. European manufacturers were left
with the development of less innovative discrete devices and linear integrated
circuits for consumer and industrial markets. Among the tube manufacturers,
competition between new and existing technologies was frequently determined
by in-house bureaucratic rivalries instead of by competing companies.
The key development in the acceleration of high technology electronics in
the United States was the switch from the earlier discrete devices to the newer
integrated circuits that had been introduced by Texas Instruments in 1959 and
Fairchild in 1960. Instead of being based on germanium and the alloy-and-
mesa process, the new technology was based on silicon and on planar fabrica-
tion. These innovations made mass production easier and facilitated substantial
component integration within one chip. Young companies that were wedded
neither intellectually nor financially to the older ways moved into the new tech-
nology. These firms left the traditional, vertically integrated European and
American tube manufacturers far behind. For the most part, European firms
also made incorrect choices of American partners for technology transfers by
choosing their traditionalist counterparts across the Atlantic, such as RCA and
General Electric, both of which soon dropped out of integrated circuit produc-
tion.
The integrated circuit period lasted a decade, until the introduction of large-
scale integration (LSI) in 1971 with the development of the microprocessor.
The LSI period in turn lasted a decade and was followed by the very large-
scale integration (VLSI) stage, beginning in the early 1980s. At the beginning
of the LSI period, American firms were dominant in high technology and in-
creased their lead over their European competitors, while the Japanese also
made serious advances. American firms also took a lead in several VLSI prod-
ucts. Meanwhile, European public policy began to focus on microelectronics.
Finished goods producers such as telecommunications and consumer electronics
firms started to integrate vertically into component manufacture, and govern-
ment projects provided investment funds. On the whole, however, none of
these efforts were successful in significantly challenging the overall American
and Japanese strength, though achievements were made in niche markets. For
example, French firms were strong in developing digital exchanges, and the
"Smart Card" had a great number of future possibilities. The Prestel videotex
system in Britain established a novel information medium. Several Italian com-
panies were creative in software applications and in assembling robots. Sie-
mens was a leader in many products.
Still, European semiconductor sales declined from the 30 percent share in

the early 1970s to 18 percent in 1980, less than half the volume of the United
States alone. By 1984, the share of European firms in the world market had
decreased further to 8.5 percent. During that same period, Europe's share of
total global chip consumption dropped from 27 percent to 18 percent (Wiegner,
1985). By 1990, European chipmakers' market share was slightly up (7.5 per-
cent worldwide, 36.5 percent in Europe). Philips alone among European firms
ranked among the top ten suppliers of semiconductors in 1990, but much of its
worldwide semiconductor sales were attributable to its Silicon Valley subsidi-
ary Signetics. Japan and the United States were far ahead in semiconductor
production: 50 percent and 40 percent respectively in 1991.
In the 1980s, there were twice as many data processing terminals in the
United States as in all of Western Europe, and the ratio of European to Amer-
ican data processing terminals fell from 62 percent to 54 percent from 1978 to
1988.
American computer companies dominated the European market, holding for
a long time a collective share of more than 80 percent. IBM was the leader,
particularly in mainframes, and was strong in the medium-sized market, along
with DEC and Hewlett-Packard. Apple and Tandy were large suppliers of mi-
crocomputers.
European efforts to strengthen their industries were substantial.' Even so,
the European computer industry had only two of the world's top ten manufac-
turers in 1989. Olivetti was eighth, with $7.26 billion in revenues, and Bull
was ninth, with $6.47 billion. IBM had revenues greater than the combined
earnings of its top six rivals in 1989 (de Jonquieres, 1990). In contrast, Euro-
pean firms in telecommunications constituted three of the top six firms. In that
sector, which is technically similar to that of computers, government procure-
ments rather than markets govern.
It is often claimed that there is not enough R&D in Europe, an assumption
accompanying calls for more public subsidies. But the countries of the Euro-
pean Community spent only slightly less than the United States on high tech-
nology R&D, and they spent nearly ten times as much as Japan.2 A related
misconception is that European firms are not large enough to carry out R&D.
This belief frequently leads to recommendations that firms be consolidated to
establish larger units. Such observations confuse cause and effect. Size is not
a precondition to success, but its result (Ergas, 1985).
Eventually, European governments began to encourage the formation of new
firms. France granted six months of unemployment benefits in a lump sum if
the benefits were used for investment in a new business. In Britain, unemploy-
ment benefits were paid in some instances while a person's new business took
hold. The main purpose of these programs, however, was to help potentially
unemployed people, not high-technology entrepreneurs.
Perhaps the greatest impedance to new firms was the low availability of
venture capital in Western Europe. Banks tended to be conservative in their
lending policies. They viewed themselves as lenders rather than underwriters,
even though most European countries do not separate commercial and invest-
ment banking by regulation as in the United States.3 In France, the use of

venture capital to finance business was not widespread. In 1983, an attempt
was made to change this through the reform of the French over-the-counter
stock market.4
The failing of the private sector financing set the stage for the significance
of governmental institutions in high-technology ventures with much greater
success.
Development Projects and Joint Ventures
Every European country has some government-sponsored development program

in semiconductor and computer technology. In Germany, three successive data
processing programs were funded in the 1960s and 1970s. In France, two Plan
Calculs were instituted amid much publicity in 1966 and 1971.5
In 1978, the French government launched the Plan Circuits Integres, provid-
ing subsidies and preferential access to the French market in which the govern-
ment is the dominant purchaser, and buying 40 percent of all chips sold.6 The
costly French efforts in the computer and consumer electronics fields are de-
tailed in the chapter on that country's telecommunications.
In Britain, development by the main computer firm ICL was subsidized by
the government. Other support programs were also set up, but until 1978 this
assistance was relatively limited. 7
Despite its missionary stress on private initiative, the Thatcher government
strongly supported government-backed development of the information indus-
try. In 1982, the Alvey Report recommended substantial government invest-
ments in VLSI software engineering and artificial intelligence research and de-
velopment, with the goal of developing a fifth generation computer. A 120
million government semiconductor development program called Inmos was then
implemented. Inmos was later privatized and sold to Thorn EMI.
In Italy, an electronics industry development plan similar to the British one
was part of a large industrialization program formulated in 1978, most of which
remained largely a blueprint.
The strengthening of the electronics industry in Europe became a subject of
particular interest for the European Community. As discussed, a number of
supranational subsidy programs exist with the aim of helping multinationals'
efforts, such as EUREKA, RACE, and ESPRIT.
A large number of private European joint ventures were also established
(Aronson and Cowhey, 1987). Olivetti owns a major part of the British firm
Acorn Computers. Philips, in addition to its link with Siemens and AT&T,
entered into joint ventures with Ericsson in Sweden, and with Bull, Alcatel,
and Thomson in France. In 1983, Philips took control of the German consumer
electronics firm Grundig when that company had financial problems. Similarly,
Thomson acquired the consumer electronics division of AEG-Telefunken in
Germany when it was near bankruptcy, as well as the German television pro-
ducers Nordmende and Saba, and the stereo equipment maker Dual. CGE col-
laborates with the Belgian holding company ACEC. The Belgian SGB, part
Telecommunications Policy as industrial Policy 319
owned by Olivetti's De Benedetti, is also a part owner of Alcatel. Alcatel,

Siemens, Plessey, and Italtel have a joint research effort for telecommunica-
tions switches and transmission. Siemens and GEC own Plessey.8
Interfirm collaboration is not without limits. Multinational joint ventures can
be hampered by some countries' promotion of high technology companies as
"national champions," making it difficult to have these firms as junior partners
in a collaborative effort.
Those wishing to establish intra-European joint ventures must also consider
the European antitrust laws of the Treaty of Rome. Additionally, they must
take into account the various national antitrust legislation, particularly the Brit-
ish and German laws. The German Cartel Office, for example, vetoed a joint
project for fiber-optic cables that would have brought together Siemens, Phil-
ips, and other smaller companies.9
Industrial Policy: PTTs to the Rescue

As the 1990s began, Europe's and America's microelectronics industries had
serious problems. Philips, Groupe Bull, SGS, and Thomson incurred huge losses
and were forced to lay off thousands of employees. No European firm held
more than 4 percent of the world market, and the Japanese dominance in ad-
vanced components was increasingly worrisome.
The difficulties in establishing a strong, self-sustaining, and self-financing
electronics industry outside the telecommunications sector and of breakthrough
government projects strengthened the role of the PTTs. They are best able to
supply resources, field trials, standardization, and procurement markets. They
accomplish their activities while maintaining a stress on civilian rather than on
military projects, which have fallen increasingly into political disfavor.
PTTs were able to provide nontariff protection and export advantages to the
companies with which they work. They accomplished this through their buy-
domestic and preferential standards-setting policies, which generally favored
the purchase of domestic-manufactured equipment. This is particularly impor-
tant because tariff barriers within the European Community largely disappeared
and made industry more vulnerable to competition from European imports, while
the international free trade rules of GATT reduced the scope of protectionism
for non-European manufactured goods.
An assured domestic base was important as development costs were rising.
One compilation of R&D costs of digital exchanges for the major manufactur-
ers is listed in Table 30.!. The cost ranges from half a billion to three times as
much.
(These large numbers, however, must be used with caution. It is always
more expensive to develop an entirely new technological generation than to
upgrade it.)
The PTTs served as large-scale purchasers of almost every kind of electronic
equipment. For many manufacturers, PIT orders meant the difference between
survival and bankruptcy. The PTTs underwrote part of the development and
320 Telecommunications Policy as industrial Policy
Table 30.1 R&D Expenditure on Development of Digital

Exchanges, 1987
Types of Switches Cost (billions)

AXE (Ericsson) $0.5
EWS-D (Siemens) 0.7
DMS (Northern Telecom) 0.7
ESS-5 (AT&T) 0.75
E1O (Alcatel) 1.0
System 12 (ITT) 1.0
System X (GEC-Plessey) 1.4
Source: Palmer and Tunstall, 1990, p. 22.
trial costs of a new product by tending to be its first major user and assuming
some of the costs of the early shake-outs and of the early production runs.
PTTs therefore subsidized the development of products that were then offered
in the world market, often at costs below the domestic ones. Domestic tele-
phone users, in effect, subsidize the export activities of industrial firms.
For example, deviations from competitive world market prices of up to 120
percent existed for PTT central office equipment procurement in 1985. In Bel-
gium, this 120 percent markup allegedly supported company R&D. Germany
and Italy paid prices 100 percent above cost. France's payments for equipment
were only 50 percent above world prices, but the DGT funded R&D directly,
not through equipment purchases (Miiller, 1987, p. 12). The attempts to use
the telecommunications network as a motor for more general industrial policy
is not uniquely European. Japan, Korea, Singapore and Taiwan, for example,
have pursued similar policies. In the United States, too, there have been many
voices recommending such a direction in light of the decline of U.S. strength
in electronics. Their weight is partly dependent on the success of the European
efforts and its demonstration effects.
Toward Vertical Integration?
The weakness of some equipment firms once the protective bond of the postal-
industrial alliance is loosened means that, at least in the short term, the tradi-
tional network operator may gain rather than lose strength in a liberalized envi-
ronment. It would be naive to expect that reorganizing a PTT as an independent
corporationwhatever its ownership statuswould lead it automatically to strive
for improved efficiency. If the exhortations to act like a business are taken seri-
ously by a PTT and are acted upon, they result in an expanded set of activities.
In Europe, the network operators, now often permitted to do so, have inte-
grated horizontally and vertically and extended their reach. It is likely that we
are merely at the beginning of a process in which effective telecommunications
carriers move into adjacent lines of business.
In the United Kingdom, British Telecom began to pursue several avenues of
vertical integration. In the equipment field, it bought the Canadian PBX man-
ufacturer Mitel. The regulatory body, Oftel, and almost the entire British
equipment industry then argued that the acquisition would not be in the public
interest because it would strengthen BT's power in terminal equipment and
threaten British PBX manufacturers. Although it acknowledged the problem,
the Monopolies and Mergers Commission accepted the merger, with some con-
ditions attached. But even those conditions were waived by the government,
which approved the acquisition. In the end, the acquisition was not successful
economically, but that does not mean that future vertical integration is pre-
cluded.
BT is not unique in seeking to expand its market power vertically. In Spain,
the semi-private telecommunications monopoly, Telefonica, has a strong in-
volvement in manufacturing. It held a large interest in the Spanish ITT-Alcatel
subsdiary Standard Electrica, which was by far the largest electronics manufac-
turer in Spain, and it had major stock interests in a dozen other equipment
firms and minority interests in seven more. The aggregate output of these firms
accounted for about a third of total Spanish telecommunications production.
Vertical integration also existed for a long time in Italy. The predominant
telephone carrier, SIP, is part of the holding company Societa Finanziaria Tele-
fonico, which in turn is partly private, and mostly controlled by the government
holding organization Institut per la Reconstruzione Industriale (IRI). STET also
owns several major manufacturing firms, including Italtel, the country's largest
telecommunications equipment firm, and several leading components and high
technology firms.
Similarly, the Swedish national telephone administration, Televerket, has
substantial manufacturing operations in its TELI and other subsidiaries. Japan's
NTT formed within one year of privatization almost seventy subsidiaries or
new ventures that were active in new products, services, and marketing.
These instances of vertical integration by the network operators indicate that
liberalization transforms the relationship between PTTs and equipment firms
from one of partnership into one of potential rivalry. And while this has com-
petitive advantages, at least in the lengthy transition phase, the potential for
vertical integration increases rather than decreases the role of the network op-
erator.
The conclusion is that in their role as the state-sanctioned engine of domestic
technological developments, and through their ability to extend their market
power vertically, the PTTs, now transformed into semi-independent PTOs with
greater flexibility, will be even more important and powerful than in the past.
It is also conceivable that a similar role might be assigned in the U.S. to tele-
phone companies, as alarm grows over the Japanese lead in high technology.
Several congressional bills point in that direction. Thus, convergence may be
resurrected, after all.
31
Transatlantic Trade Friction
European Interpretation of U.S. Telecommunications Policy
Developments in the United States challenged the traditional status quo in the
European telecommunications field, threatening its traditional postal-industrial
coalition. For a long time, European interpretation of U.S. developments was
colored by the prevailing views of telecommunications experts, most of whom
were closely affiliated with the traditional system. The AT&T divestiture itself
was generally described as advantageous to AT&T: It was dropping the costly
baggage of the regulated operating companies and could now take on IBM.
European PTTs had a great amount of respect and sympathy for the old
AT&T. Although the American telephone operating company was privately
owned, its dominance was similar to that of European administrations. Inter-
national cooperation in such areas as transatlantic communications resulted in
close links as partners rather than competitors. The dismantling of AT&T be-
wildered the PTTs. The decision seemed arbitrary, inefficient, and political. It
should be noted that the belief that the competent and successful AT&T had
been needlessly dismembered by the government is at odds with the view that
the divestiture was a great success for AT&T.
The notion that the United States, with its advanced technology and success-
ful telecommunications monopoly, would choose voluntarily to dismember such
a system was unsettling. Whereas in the past, the development and adoption of
new technologies provided security to the PTTs, it now seemed to undermine
them. This perception resulted in strong defensive reactions. The American
circumstances were generally portrayed as inherently different from those in
European countries and hence not applicable to their situation, and little effort
was made to isolate those elements that could be seen as part of a generalizable
network evolution.
In some ways, the PTTs' views closely resembled the instant nostalgia that
occurred in the United States after the announcement of the AT&T divestiture
agreement. AT&T, which had been a favorite object of criticism, was suddenly
seen as an efficient and benevolent organization torn apart by economic zealots.
This ignored decades of major complaints about AT&T and its vast power in
the American economy. The lawsuit against AT&T by the Antitrust Division
of the Justice Department was the result of long-standing problems and was
pursued by both Democratic and Republican administrations.
322
Transatlantic Trade Friction 323
Generally the negative distributional impact of U.S. developments was stressed

by PTTs. It is true that the AT&T divestiture threatened the system of cross-
subsidies from business customers to residential customers, from long-distance
service to local service, and from urban and suburban users to rural users. But
these cross-subsidies did not disappear. Congressional reaction and state regu-
latory actions have shown that the protection of affordable universal service is
a high political priority. In New York State, for example, the Public Service
Commission, in a 1987 plan designed by the author, set rates for poor people
at a low $1 per month for a basic subscription, plus two discount packages for
usage, and a reduced installation charge for new subscribers of $2 per month
over a year. In many areas of the United States, similar basic "lifeline" tele-
phone service was instituted. In rural areas, low-priced telephone service is
maintained, among other means, through direct subsidies to providers of such
services, for example, through low interest loans from the Rural Electrification
Administration. Telephone rates for rural America are also often lower than for
urban users, because charges vary by exchange size. Small exchanges offer
services at markedly lower rates.
In the United States, the decision process for telecommunications is not con-
trolled by any political party. There is a multitude of decision pointsthe FCC;
the Departments of Justice, Commerce, and State; U.S. District Court Judge
Harold Greene; Congress; and the state regulatory commissions.1 This system
of decision making, resembling a war with a hundred local battles, is different
from the process in Europe. But for all its untidiness, this decentralized process
can accommodate change. Changes in European telecommunications policy may
seem more orderly, better planned, and less disruptive. But much of the order-
liness of the policy debate derives from the fact that for a long time there was
no real public debate on telephone issues in most countries. European telecom-
munications policy discussions were carried on by government insiders. When
regulation and operation coexist within the same institution, little information
flows to the public. Although there is often heated criticism of the quality of
service, as took place in France in the early 1970s, this rarely touches on the
nature of the system itself, only on the level of investment or the competence
of its managers.
Of course, policymakers within European governments try to accommodate
the interests of a variety of groups that have a stake in telecommunications, but
this is done by emphasizing the participation of insiders rather than by launch-
ing disruptive challenges to the status quo. These participants are typically
technologists and government officials. For a long time, the PTTs transformed
domestic policy questions into issues of electrical engineering or international
diplomacy that must be dealt with at a governmental level.
The Containment of American Equipment Manufacturers in Europe
Although the traditional postal-industrial coalitions in European countries were

sharply critical of U.S. liberalization, they were not above benefiting from it
324 Transatlantic Trade Friction
in terms of trade. This created problems of trade reciprocity that spilled into
the political arena.
The AT&T divestiture led to the emergence of AT&T as a competitor in
international markets, a sharp break with the past. This development received
much attention and led to fears of an American telecommunications equipment
offensive into Europe. But the opposite happened in the 1980s. American tele-
communications equipment makers were repulsed and almost expelled from the
European market, with ITT, GTE, and Honeywell largely departing. AT&T
and the Bell regional holding companies were only partly successful after large
scale efforts.
For more than fifty years, AT&T had stayed out of international equipment
activities, despite its position as the largest such manufacturer in the world.
But this had not always been the case. In the early years of the telephone, the
Bell System had licensed several European equipment manufacturers, acquired
others, built its own facilities in Europe, and had a substantial manufacturing
and distribution presence in several major countries. But in the 1920s, the com-
pany was under much U.S. pressure to sell its international operations, as
American critics of AT&T charged that American ratepayers were subsidizing
AT&T's international operations. For that and other reasons, the company in
1925 decided to sell its European operations to the then relatively insignifi-
cant firm ITT, run by the Virgin Islands entrepreneur Sosthenes Behn and his
brother.
This event marked ITT's entry into the big league of telecommunications.
ITT's major European operations included Standard Telephone Company (STC)
in the United Kingdom; Standard Lorenz Elektrik (SEL) in Germany; FACE in
Italy, Bell Telephone Manufacturing Company (BTM) in Belgium, Standard
Electrica in Spain, and LMT and CGCT in France. Other subsidiaries existed
in Norway, Denmark, Portugal, Austria, Ireland, and so on. Given the nation-
alistic nature of the telecommunications equipment market, it was to ITT's
advantage to have a physical presence in each country, because this allowed it
to present itself as a domestic rather than an American company; local ITT
companies often downplayed their U.S. connections.
The ITT companies included, among others, Standard Radio and Telefone
AB, Sweden; the Bell Telephone Manufacturing Company, Belgium; FACE,
Italy; ITT, Austria; ITT, Netherlands; Standard Electric Kirk A/S, Denmark;
Standard Electrica, Spain; Standard Electrica, Portugal; Standard Elektrik Lor-
enz (SEL), Germany; Standard Telephone and Radio, Switzerland; Standard
Telefone og Kabelfabrik A/S, Norway; Standard Telephone and Cable (30 per-
cent ITT owned), Britain; and an Australian subsidiary. ITT's sales were not
normally broken down by country, though one study reported the following
1976 sales figures (Northern Telecom, 1980):
West Germany $516 million

France $456 million
Spain $439 million
United Kingdom $330 million

Italy $259 million
Belgium $255 million
Brazil $130 million
Australia $ 96 million
Norway $ 88 million
Switzerland $ 71 million
Netherlands $ 65 million
Austria $ 44 million
ITT was a huge conglomerate, employing over 350,000 in 1980. At some

point, it own^d telephone companieslater nationalizedin Peru, Chile, Puerto
Rico, and Spain. Its original holding in the Virgin Islands Telephone Company
was still intact, and it owned an international record carrier (ITT Worldcom)
and a U.S. domestic long-distance carrier (USTT). In equipment, where about
three-quarters of its sales were in Europe, ITT had an impressive array of foot-
holds with aggregated economies of scale, as in the case of its advanced Sys-
tem 12 digital exchange, while at the same time assuming a certain national
stance by downplaying its American connections.
But the company, which had become a far-flung conglomerate in the 1960s
and 1970s, was unable to sustain the costs of extending its System 12 digital
switch to the U.S. market or financing development of the next generation of
technology. In 1986, control in ITT was acquired by France's CGE, then a
state company. For details, see the country chapter on France.
GTE was another American telecommunications company with a presence in
the European equipment market. It had substantial manufacturing involvements
in Italy and Belgium. But it could not make headway with its products, and in
1986 it sold 80 percent of its equipment interests to Siemens.
Following divestiture, the international market became interesting to AT&T.
Since the domestic equipment market had been opened to all comers, its U.S.
market share had nowhere to go but down. Hence, the rest of the world was
its field of growth. AT&T's strategy was to align itself with domestic interests,
thus overcoming the barriers an American company would face. The first such
toehold alliance was a joint venture agreement with the Dutch electronics giant
Philips, with the latter contributing access to the politicized European procure-
ment field. Philips eventually decreased its investment.
Another significant alliance was with Olivetti, the Italian office equipment
and small computer manufacturer. AT&T purchased 25 percent ownership of
Olivetti for $260 million, with the option to acquire another 15 percent after
four years.
Olivetti's ambition was to become a major European player in the world
computer markets. But the collaboration did not work out well. In 1988, AT&T
chose not to increase its financial and managerial involvement in the joint ven-
ture and cut its computer orders by 75 percent.
Perhaps the greatest AT&T breakthrough was with Italy's Italtel, a sister
company of the main operating company SIP. Italy had slated its lagging tele-
communications network for a major upgrade and expansion. But Italtel, the
main domestic supplier, was in need of a foreign technology source. A struggle
between Siemens, Alcatel, and AT&T for the huge contract ensued and led to
lobbying efforts by Presidents Mitterrand and Reagan. In 1989, AT&T was
selected. Italtel, in return, received a share of AT&T's European operations.
AT&T also entered into other international involvements in Korea, Spain, Ire-
land, and Taiwan, all primarily off-shore manufacturing efforts that provided
some local presence. In 1990, Spain's Telefonica upgraded its involvement
with AT&T. Telefonica became a participant in AT&T Network Systems In-
ternational, with 6 percent of the equity. AT&T had 59 percent, Italtel 20
percent, and Philips 15 percent, until the latter ran into financial difficulties and
sold out to AT&T. These partnerships opened the prospect for AT&T's entry
into Spain's expanding equipment market. Next, it established itself in Ukraine's
network operations, and in Armenia and Kazakhstan. Thus, AT&T seemed to
have finally established a European presence.
Another major move by AT&T was to try, unsuccessfully, for an agreement
with the dominant French firm of CGE, which unleashed a round of diplomatic
and economic intrigue. In particular, the German government began to lobby
on behalf of Siemens, pressuring the French government at the highest levels
to substitute Siemens for AT&T-Philips in the spirit of European solidarity, as
well as in reciprocity for the newly acquired German Alcatel subsidiary SEL.
The tug of war grew acrimonious. FCC Chairman Mark Fowler sent pointed
inquiries to major American telephone companies regarding their use of equip-
ment from countries where U.S. firms were being discriminated against. Within
the French government, the PTT preferred APT, but other ministries did not
wish to antagonize Germany. In the end, the French government chose the
Swedish Ericsson as a neutral compromise, together with the French defense
firm Matra, which thus gained a foothold in its country's telecommunications
market. Further details on AT&T and ITT in France are provided in the chapter
on that country.
Overall, the European equipment market was $18 billion in 1987 and had
grown to $28 billion by 1990, with huge modernization and digitalization
programs in Spain, France, Italy, and other countries; the European Commis-
sion predicted a 9 percent yearly growth rate for equipment and services. Within
the European market, Alcatel, Siemens, and Ericsson were the dominant sup-
pliers. In 1987, Alcatel was the largest supplier of public switching equipment
(with 40 percent of the market), private switches (20 percent), terminal equip-
ment (19 percent), and transmission systems. Siemens was second in public
switches (20 percent), terminal equipment (16 percent), and transmission
equipment (15 percent), followed by Ericsson in each category. Northern Te-
lecom was the only non-European company to break into the top three, with
an 18 percent share of private switching equipment. But with the saturation of
the domestic market for basic service expansion, exports became more impor-
tant to equipment manufacturers.
European Exports to America
The U.S. market is not only the largest domestic market in the world for tele-
communications products, but it is also relatively free, and it has a large num-
ber of potential customers in the 1200 local exchange companies, the various
new carriers and telephone companies. There are more networks as potential
customers in the United States than in the rest of the world put together. (Most
of the companies are, of course, quite small.)
The U.S. market for local exchange and transmission equipment was char-
acterized before the AT&T divestiture by competition only in the procurement
of equipment by non-AT&T telephone companies. Even GTE's local exchange
companies were tied to their parent company's manufacturing units. AT&T was
largely precluded from the independent market by the terms of legal agreements
with the Justice Department, butperhaps as a resultmany other companies
were active in it, including foreign suppliers such as Ericsson and Northern
Telecom. On the other hand, the vast Bell system and all of its customers
comprising 80 percent of the local marketwere largely inaccessible to other
suppliers because of AT&T's ties to its manufacturing subsidiary Western
Electric.
The U.S. liberalization provided non-U.S. manufacturers with opportunities.
The Bell operating companies, which prior to divestiture of AT&T had relied
mostly on Western Electric, were now free to buy equipment from other sup-
pliers, and have indeed actively done so.
Interconnection of customer premises equipment was also significantly lib-
eralized in the United States. Whereas it once was more restrictive than in
many European countries, it became much freer in the wake of regulatory and
court decisions. The relaxed rules ("Part 68") followed the 1968Carterfone
case. Now equipment sellers need only to register their products with the FCC
as complying with standards before marketing them. Registration requires the
disclosure of a unit's technical specifications, so that the FCC's staff can, if it
wishes, identify possible system degradation prior to installation of the equip-
ment. But there is no approval process. (There is a national security exception
to the registration requirement. If a federal agency certifies that compliance
with registration procedures would jeopardize security interests, equipment may
be connected to the network without publication of technical data.)
Although most analysts expected the BOCs to cling to AT&T as their equip-
ment supplier, they in fact rapidly embraced a wide variety of non-AT&T
equipment. In one instance involving equipment allegedly affecting defense
communications, the Defense Department reportedly used pressure to influence
a carrier not to buy Japanese equipment. But this was a widely noted exception.
Generally, the opening of the U.S. market to non-AT&T and foreign equip-
ment was rapid.
Network standards are coordinated for the BOCs by their joint organization,
Bell Communication Research (Bellcore). Although Bellcore's information re-
quests to vendors of central office switches are formidable, there is no eco-

nomic reason for it to use its role to favor AT&T or other U.S.-owned manu-
facturers. The rapid growth and large market share of the Canadian-owned
Northern Telecom would tend to indicate that neither the FCC nor the state
commissions have shown a desire to set standards.
Procurement of network equipment by local telephone companies is gov-
erned by their obligation to state regulators to pay the lowest possible prices
for qualifying equipment. The loss of subsidies from long-distance service cre-
ated pressures in keeping local rates low. The ability to compare costs for the
twenty-two companies also forces them to seek low-cost equipment. The "gold
plating" (overcapitalization) of the past is unlikely to persist in the new envi-
ronment. Because of the divestiture, the BOCs no longer have any incentive to
increase Western Electric's profits, because none of those profits benefit the
BOCs or their managers. Furthermore, the partial substitution of price cap reg-
ulation, or negotiated rates, for rate-of-return regulation provides an incentive
for cost cutting. Thus, the potential for new entry by suppliers existed.
Aiming at the U.S. market, George Pebereau, then president of the French
CGE, declared:
It is obvious that no European company, French or not, can remain a world com-
pany if it does not have a significant position in the American market, which rep-
resents 40 percent of the world market and, in addition, is from the point of view
of technology the best testing grounds one can imagine. Happily, we have a
historic opportunity to develop ourselves in the U.S., with the deregulation of
AT&T. . . . If, sadly, CGE's presence in the U.S. failed, we would need more
than a decade to regain the confidence of our American customers [L'Expansion,
1985].
But Alcatel, which bought ITT's switching systems division worldwide (in-
cluding the United States), was not successful at introducing ITT's switch into
America. It closed its U.S. switching division and cut back its efforts to sell
cable and transmission equipment. Northern Telecom, however, performed very
well in the U.S. market, securing a leadership position as a second source.
Three contenders for the number three slot were Siemens, Ericsson, and Fu-
jitsu. None of them emerged as a clear leader. Siemens, with the strongest
presence, had over 1 million installed lines as of 1990. Ericsson's success has
been mostly in rural applications of its AXE switch, and Fujitsu had various
trials in place throughout the United States.
The major problem, from the U.S. perspective, is that the opening of the
U.S. telecommunications equipment market to foreign suppliers was not matched
by a reciprocal opening of their markets to U.S. producers.
In 1981, the United States had a $6.04 billion trade surplus in electronics-
based products, which included telecommunications and computer goods. By
1986, the U.S. deficit in this area was $16.06 billion (NTIA, as cited by Rob-
inson, 1987).
In the subcategory of telecommunications products, the United States had a
$817 million trade surplus in 1981, and a $2.5 billion deficit in 1987. In the
same period, exports of telecommunications products increased by 41 percent,

whereas imports increased by 320 percent (Robinson, 1987). By 1989, the
deficit had dropped to $1.9 billion (Department of Commerce, 1990, p. 227).
On the other hand, the B.C. imported 21 percent of its equipment from Japan
and 11 percent from the United States, whereas exports to the United States
amounted to only 4.3 percent of B.C. exports. The United States had a positive
trade balance with B.C. nations of $417 million in 1988 and $800 million in
1989 (Department of Commerce, 1990, p. 229).2 Nevertheless, U.S. compa-
nies felt increasingly higher walls around a "Fortress Europe." It was much
easier for Siemens, Alcatel, or Ericsson to do business in the U.S. than it was
for AT&T to have a presence in Germany, France, or Sweden.
It was highly unlikely that the U.S. government would stand by as foreign
firms sold freely in America while U.S. manufacturers were shut out. The FCC
initiated a Notice of Inquiry in 1987 in which it invited comments on whether
there should be restrictions on the approval of equipment exported from coun-
tries that discriminated against American equipment. In 1988, the FCC ordered
large U.S. firms to report purchases of switching equipment from foreign sources.
This action came in retaliation to French and German resistance to AT&T op-
erations in Europe (Davis, 1988).
Another response to these developments was the introduction of proposed
federal legislation to require reciprocity. The United States also exerted pres-
sure on Japan and Europe to lower its nontariff barriers in equipment procure-
ment. For example, the U.S. International Trade Commission ruled that a num-
ber of Japanese manufacturers had "dumped" (i.e., sold below cost) cellular
car telephones in the United States. The decision allowed the U.S. Customs
Service to increase duties on these manufacturers' products.
Thus, the opportunity to enter the U.S. market is a double-edged sword,
threatening to bring about a reduction of European and Japanese firms' own
protected positions. This tends to split the telecommunications industries of
other countries: strong and advanced manufacturers who can compete success-
fully in the world market based on the merits of their products can accept
imports to their home base, but weak firms in need of protection have little to
gain and much to lose from lowering the barriers.
A round of accusations began in 1989 with a U.S. Trade Representative
(USTR) report finding that the European Community had "needlessly burden-
some" type approvals and discriminated against U.S. suppliers in procurement
and services. Two days later, the E.C.'s Directorate General for External Af-
fairs issued its own counterreport, entitled "U.S. Barriers to Trade," which
claimed that the cost of adapting equipment to the U.S. market and lengthy
Bellcore approvals effectively cut E.G. firms out of U.S. contracts (TDR, 1989).
This followed a 1986 report by the Commission of the European Communi-
ties criticizing the slowness and complexity in the American approval of equip-
ment and the oligopolistic features of the market. But this applies to central
office equipment; for terminal equipment, the charge is largely incorrect. It is
true, however, that the costs for foreign switch manufacturers to enter the United
States can be substantial. Not only must they pay $5 to $10 million to Bellcore
for testing the system (the so-called Phase E review), but it also costs $100
million or more to modify the system to North American standards and soft-
ware practices. There are technical differences that create barriers to entry by
European suppliers. But there is no evidence that these technical standards,
which evolved over decades some time ago, were set or are applied with a
protectionist purpose.3
In 1990, the European Commission issued a directive on public procurement
that included a "buy Europe" clause, stipulating that E.G. bids must be ac-
cepted if they are less than 3 percent higher than non-E.C. offers (for orders
over 600,000 ECU). Only bids from firms with more than 50 percent B.C.
ownership would be given this preference. The law was scheduled to take ef-
fect in 1992, except in Spain, Portugal, and Greece. Shortly after the directive
was passed, the French Ministry of European Affairs called for resistance to
the "invasion" by U.S. and Japanese firms. It claimed that Japan had a 95
percent share of the facsimile machine market, that Motorola had 25 percent of
cellular equipment, and that AT&T had increased from a 2.5 percent market
share in switching in 1985 to 13 percent in 1990 (Roussel, 1990).
The following are unconfirmed (and quite likely non-comparable) numbers
for digital switches that were provided by several market participants about
competitors' and suppliers' bids and contracts. The figures should be treated
with great caution, because each deal may include different packages of hard-
ware, software, and support. But the numbers, nevertheless, serve to illustrate
the order of magnitude of the high premiums charged domestically over the
lowest bid made by the same firm internationally in the early 1990s. The fig-
ures show a very significant domestic premium, indicating that much of R&D
and other initial costs are paid for by PTTs and borne by domestic rate payers
in those countries where foreign entry is disfavored over domestic "industrial
champions." The U.S. seems to have competitive pricing, with a small pre-
mium only. Japan, despite the four major vendors of the "NTT family," has
prices above world market prices. And European premiums are substantial.
Market Share Home Price Int'l Lowest % Home

Company in Home Market ($/Line) Bid($/Line) Market
Alcatel (France) 84 335 110 204
AT&T (U.S.) 42 110 100 10
Ericsson (Sweden) 100 325 130 150
Fujitsu (Japan) 24 290 110 164
NEC (Japan) 25 290 140 107
Northern Telecom
(Canada & U.S.) 80* 250 100 150
Siemens (Germany) 83 450 100 350
Sources: figures provided by various market participants (unverified).
* Canada
32
International Telecommunications
Services
The clash between the different policy approaches taken on the two sides of
the Atlantic was particularly acute in international communications, partly be-
cause of its great profitability. Historically, U.S. policy on international tele-
communications had carved up the market into distinct segments, assigning
each segment to different kinds of carriers. In the 1970s and 1980s, however,
the United States radically restructured its own rules of the game and forced
the European countries to respond to a new situation. This led to frequent
disputes. American changes destabilized the traditional European system, as
analyzed later with the theoretical model of Chapter 40.
The volume of international telecommunications traffic increased much faster
than international trade in general. From 1970 to 1981, for example, interna-
tional calls originating in the United States increased by a factor of 11.3, whereas
American international trade grew, in real terms, by a factor of 3 (Antonelli,
1984).
One part of the impetus behind this rise in international traffic was the dra-
matic decrease in investment cost for a transatlantic circuit, from $1.4 billion
per voice circuit on TAT-1 in 1956 down to $44,356 per circuit for the fiber-
optic TAT-8 cable in 1988 (Stanley, 1988). An FCC study found that the cost
per minute on transatlantic cable dropped from $2.53 in 1956 to $0.04 in 1988
and was expected to fall to $0.02 in 1992. In the same period, the number of
available voice circuits grew from 89 to 37,800. Satellite circuit costs similarly
fell from $32,000 each on the Early Bird satellite in 1968 to $4,680 for the
Intelsat-VI satellite generation in 1982.
However, this drop in costs was not matched by an equal drop in prices;
consequently, the profit margin on international service remains very high. Ac-
cording to one study, British Telecom charged $750,000 for a direct-broadcast-
grade connection between London and New York in 1981, whose cost to BT
was only $53,000, which in turn contained an Intelsat charge that already was
well above actual economic cost (Stapley, 1981).
Closely related to these high prices is their asymmetry. An FCC study showed
that the weighted average for foreign tariffs was almost 95 percent higher than
the American tariff (Kwerel, 1984).
Lower rates in the United States are partly the result of a long struggle among
331
332 Internationa! Telecommunications Services
various market segments and participants. In 1964, clear boundaries existed,

and the FCC prohibited AT&T from entering the international record market
(i.e., telegraph and data transmission) to protect the so-called International Re-
cord Carriers (IRCs). Authorization of the transatlantic TAT-4 cable was con-
tingent upon AT&T's exclusion from such services (GAO, 1983).
Among record services, the FCC made a further distinction between domes-
tic services, from which Western Union was restricted, and international ser-
vices, which were provided by the IRCs, including Western Union Interna-
tional, which had been divested from Western Union to become a wholly
independent and unaffiliated entity. IRCs could only operate in the United States
from certain limited and approved "gateways." A telegram from Cleveland to
Paris, for example, would be routed by Western Union to an IRC gateway,
transmitted by an IRC to Europe, and then passed on to the French PTT. Price
competition among the IRCs was restrained.
The market segmentation led to a lack of competition, as well as to substan-
tial earnings margins. Partly because of the high profitability, the situation be-
came unstable and cracks began to appear. The artificial nature of the market
segmentation then became evident and led to a policy response within a rela-
tively short time.
The FCC set maximum rates for international telecommunications services
in theory on the basis of rate-of-return regulation. In practice, however, these
rates were not closely monitored because AT&T's international department was
not examined separately from its overall operations. Figures for 1979, the first
year that AT&T was required to provide separate reporting, show that the net
return of overseas voice service represented a very high 36.5 percent on its
total investment.
Similarly, the FCC did not investigate the rate of return for any IRC between
1958 and 1976. A 1979 audit report found that telex service was subsidizing
telegraph and private services. The IRCs' rate of return for telex services ranged
from 34.4 to 58.3 percent for the most profitable carrier and from 18.6 to 25.4
percent for the least profitable carrier, with the variation in the percentages
depending on methodology (GAO, 1983).
High profits and differential pricing encouraged the emergence of arbitrage.
In 1981, a telex message from Germany directly to the United States cost $2.58
per minute, but it cost only $1.76 if routed via the United Kingdom. This led
to substantial transatlantic traffic through London telex bureaus. The European
PTTs tried to stamp out this arbitrage, citing CCITT rules they themselves had
authored, but they were rebuffed by the European Commission and the Euro-
pean High Court of Justice, as detailed in the chapter on U.K. telecommuni-
cations.
Not surprisingly, as the FCC's liberal domestic policies took shape, its re-
strictive entry and service policies for international telecommunications ap-
peared to make less and less sense, at least from the U.S. perspective. In 1976,
the FCC allowed competitive entry into international telecommunications, and
thereafter routinely approved applications by MCI, US Sprint, and others to
provide international service.
International Telecommunications Services 333
In a series of rulings in 1979 and 1980, the FCC also largely removed the
dichotomy of voice and record carriage, eliminated the rules prohibiting AT&T
and the IRCs from entering each other's markets, and expanded the number of
gateway cities from which international traffic could be sent.'
The FCC also eliminated rate-of-return regulation and tariffing. Only domi-
nant carriers (i.e., AT&T and GTE's Hawaiian Telephone Company) needed
to file international tariffs. Other carriers had merely to report their activities.
The FITs observed all this with some misgivings, for these rulings chal-
lenged long-established partnership arrangements and rate structures. But once
their initial distaste for the increased complexity in the international telecom-
munications regime subsided, they recognized the potential advantages. As the
only address within their countries for AT&T, MCI, Sprint, and others, the
PTTs were in a position to profit by forcing rival American carriers to compete
against each other for operating agreements.
To prevent the IRCs from being thus "whipsawed," the FCC in 1977 en-
forced a Uniform Settlements Policy requiring all U.S. carriers to have uniform
settlement rates with all other carriers for the same routes. When the Benelux
PTTs and Nordtel (the Inter-Scandinavian telecommunications body) invited all
potential suppliers of data communication services to submit bids that included
the division of accounting (i.e., an element of price bids), the American reac-
tion was swift. Despite normally championing liberalization, the FCC ironically
requested that U.S. carriers collectively defer negotiations with Nordtel. Nord-
tel backed off and notified the carriers that it did not plan to use its monopoly
power for exclusive bids.
When different entities provide international telecommunications service at
each end of a circuit, they agree upon a division of the revenues between them.
The entities create an "accounting rate" or "settlement rate" to be paid to one
carrier by the other carrier collecting from a customer. The accounting rate
may bear little or no relationship to the actual customer charge or "collection"
rate. As a hypothetical example, the accounting rate for the first three minutes
of a telephone call between New York and Paris might be $3.00; the charge
for the call in the United States, $4.50; and the charge in France, $6.00. When
U.S. customers call, they pay $4.50 to AT&T, which credits $3.00 to the
French PTT. When French customers call, they pay $6.00 to the French PTT,
which credits $3.00 to AT&T. The Uniform Settlements Policy does not regu-
late U.S. carriers' rates on the U.S. end, but attempts to protect U.S. compa-
nies from whipsawing by foreign PTTs by requiring that all U.S. carriers pay
them a uniform rate.
But given the profit margins it must be difficult to prevent non-price conces-
sions. In 1985, an example of whipsawing occurred when RCA filed a com-
plaint with the FCC, charging TRT and FTCC, two other international record
carriers, with using so-called special drawing rights instead of the established
gold franc settlements in their international telex accounts with the PTTs of
Finland, France, Norway, and Spain. RCA charged that this arrangement re-
duced the accounting rate they would receive from $1.38 to $1.14. FTCC de-
fended itself, arguing that it would actually receive $1.21 under the special
334 International Telecommunications Services
drawing right settlement, but it admitted that the figure was still lower than the
gold franc rate.
In 1984, the European PTTs affirmed their policy on the limitation of entry
by American competitors. The PTT organization CEPT recommended that its
members not open their markets to any new American carriers unless they
would provide better technical service at a lower cost (to PTTs) than at present.
New carriers were permitted for new types of communications services such as
videotex, teletext, facsimile, and packet switching, but the CEPT guidelines
restricted each to providing only one type of new service.
In an attempt to reduce the barriers to entry created by the PTTs' negotiation
requirements, MCI bought an existing IRC, Western Union International (re-
named MCI International) from Xerox. MCI International created a convenient
international outlet for MCI's American involvement in electronic mail and also
provided MCI with an already established relationship with the PTTs. The
company concluded agreements with several countries and established London
and Hong Kong as international hubs for its traffic to other countries. It also
complied with a host of burdensome requirements and procedures that made
service to some countries unprofitable.
A related question was the way in which European PTTs pick American
long-distance carriers for communication originating in Europe. For European
customers calling American cities, the PTT chooses which U.S. long-distance
carrier will transmit the call within the U.S. and realize the subsequent reve-
nue.
Of course, it would be possible to permit the European users to indicate
which American long-distance carrier they prefer. This could be accomplished
through the use of not one but several country codes for the United States (or
North America), with a different numeric access code assigned to each U.S.
international carrier.2 But such a choice of services, together with the possibil-
ity of advertising campaigns by various carriers directed at European custom-
ers, would have visibly demonstrated that network competition was feasible,
and this type of consumer choice was not granted to most European users.
Instead, negotiations centered on the ways in which the PTTs might allocate
their U.S.-bound traffic between AT&T and its competitors. One possibility
was to negotiate market shares in advance; another was to use a fixed share
allocation formula. The easiest was to allocate America-bound traffic to Amer-
ican carriers in the same proportion that those carriers supply traffic to Europe.
In addition to extending its pro-competitive and deregulatory policies to in-
ternational services, the FCC sought to increase competition between types of
transmission media and service providers.3
An important distinction is made in international communications between
transmission by submarine cable and transmission by satellite. The several sub-
marine cables linking North America and Europe are owned and operated by
consortia of European and North American telecommunications administrations
and firms. In contrast to their part-ownership in the submarine cable operations,
AT&T and the other American international carriers as well as domestic satel-
lite operators were specifically excluded from international satellite transmis-
sion, which was reserved for Comsat, the American designated carrier of the
international satellite organization Intelsat. Created in 1964 at the instigation of
the United States, Intelsat is a cartel-like organization with a considerable mo-
nopoly over satellite transmission of international public telecommunications.
Each member country designates a carrier to manage outgoing and incoming
Intelsat communications traffic. For most countries, this carrier is the govern-
mental PTT authority. Following intense domestic debate in the United States,
however, Congress denied AT&T this role in an attempt to limit its power. The
role was instead given to Comsat, which was created through the Communica-
tions Satellite Act of 1962 as a publicly chartered, privately owned company.
Under the 1962 legislation, Comsat was solely a "carrier's carrier"; neither
AT&T nor the IRCs were permitted direct access to Intelsat, and Comsat could
not connect directly with users. In 1965, Comsat had a 61 percent share in
Intelsat, reflecting its share of traffic. Its share later declined to approximately
25 percent.4
The FCC subsequently permitted Comsat to go beyond its role as a carrier's
carrier and to provide services to customers directly. The FCC made this con-
ditional upon a major restructuring of Comsat separating its unregulated com-
petitive activities from its regulated activities. Comsat sold its earth stations
and divested its manufacturing subsidiaries in 1988 and 1989.5
New International Carriers
Because some PTTs made almost one-quarter of their profits in international

services, it was not surprising that new entrants arrived, first by sky and then
by sea. In 1983, the FCC extended its domestic "Open Skies" satellite policy
and accepted an application for a license from a tiny entrepreneurial venture,
Orion Telecommunications, to build a private satellite system for North Atlan-
tic service. Orion planned to aim at customized business services and private
lines that were not well served by Intelsat. Just as MCI had done for domestic
services, Orion denied that it was trying to enter the market of the dominant
firm and instead argued that it would create a new market. Orion's application
was followed by similar filings from International Satellite, Inc. (backed by
TRT), Cygnus (backed by the earth station manufacturer MA/COM), RCA
Americom (for modification of an American domestic satellite), and Pan-
AmSat.
The applications caused a debate within the American government concern-
ing whether the United States should endorse or permit international systems
to "bypass" Intelsat. This kept the applications pending at the FCC and cul-
minated in 1985 with the issuance of a policy paper intended to provide guid-
ance to the FCC. It cautiously approved the concept of separate private sys-
tems, as long as they did not interconnect with public switched networks. The
FCC eventually agreed. Not surprisingly, Comsat opposed the private satellite
systems vehemently, and sought legislation that would preclude such systems
or restrict their operations.
According to a provision of the Intelsat agreement (Article 14D), no satellite

competition is permitted that would cause economic harm to Intelsat operations
and profits. Intelsat uses vague criteria in making such assessments, however.
It did not find that the PTTs' Eutelsat system was causing "significant harm"
because the European PTTs asserted, with logic more political than economic,
that they would use no satellite system other than one that they would operate.
Several regional and intercontinental satellite systems were established out-
side of the Intelsat organization. They included Arabsat, Eutelsat, a project run
by a Scandinavian consortium, and the French system (which is "domestic"
but which stretches that term to encompass communications with French pos-
sessions in the Western Hemisphere). These projects arose partly because sev-
eral countries believed that they could more easily reach their telecommunica-
tions goals if they had greater control over satellite communications, and partly
because they pursued various industrial policy goals in electronics. The aggre-
gate result was to weaken the argument that for reasons of economic and tech-
nical efficiency, international satellite telecommunications must be controlled
by a single organization.
As both users and shareholders of the Intelsat consortium, Intelsat's constit-
uent organizations did not want to see their profits whittled down by competi-
tion. To that end, they enlisted the traditional cross-subsidy argument. In inter-
national terms, the argument stated that the profits from the high-density
transatlantic and North Pacific routes were needed to provide a subsidy for low-
density traffic to and among Third World countries. But the monopoly profits
are far larger than those subsidies, and PTTs could assist less developed coun-
tries also by more direct contributions in the form of equipment, expertise,
financial subsidies, lower communications tariffs for calls to those countries,
or more advantageous settlement rates.
The conflict is not simply between Intelsat and its potential rivals, but just
as much between the PTTs and the new carriers. Consequently, various defen-
sive strategies were pursued against potential rival satellite carriers. An "up-
link" strategy was aimed at preventing the FCC from granting licenses to both
American and foreign applicants. This was supported by the argument that the
member states of the Intelsat agreement gave Intelsat the monopoly over com-
mercial international satellite telecommunications. The American applicants
contended that the agreement covered only public switched communications
and not private line leasing, and that the terms of the Intelsat agreement pro-
hibited only those rival systems that would cause "significant economic harm"
to Intelsat.
A preemptive strategy by Intelsat sought to deter potential entry by offering
new satellite service options at reduced rates. Similarly, submarine cable ca-
pacity was increased by the traditional service providers. In 1991, two addi-
tional trans-atlantic fiber cables were agreed upon by European and North
American carriers.
A "down-link" strategy tried to prevent new satellite carriers from connect-
ing into national networks. But this required a unified front of all PTT countries
in a region against the establishment of a beachhead or, if such were estab-
lished, against its use as a transfer point to other countries. As with other
cartel-like agreements, this was only as strong as its weakest link. Given its
general evolution toward liberalization of telecommunications and its privati-
zation of British Telecom, the United Kingdom did not remain agreeable to the
plan. Similarly, as in the case of tax havens, one could expect other European
countries to find it advantageous to become international transmissions hubs by
permitting down-links from non-Intelsat carriers. Moreover, limitations against
retransmission might not be supported by the European antitrust laws, as the
case of the British telex bureaus demonstrates.
Still, the delaying tactics took their toll. After a while, PanAmSat was the
only project that could afford to pursue its goals actively. In 1988, the Pan-
AmSat, with twenty-four C-Band transponders, was launched. Its combative
chairman, Rene Anselmo, promised to crack the monopoly of Intelsat with
service to Central and South America, the continental United States, the Carib-
bean, and, significantly, Western Europe. In 1990, PanAmSat filed a $1.5 mil-
lion lawsuit against Intelsat and won an easing of restrictions. In 1991, Intelsat,
having in 1985 been forced to agree to coordinate with separate systems, agreed
for PanAmSat to be interconnected with public switched networks in the Carib-
bean and Eastern Europe, with up to 100 circuits. Orion restructured itself into
a limited partnership including British Aerospace, General Dynamics, and Ma-
tra, with a 1994 launch date.
Although a single global system may be desirable because of its economies
of scale, a distance- and border-insensitive technology such as satellite trans-
mission cannot be successfully restricted for long. Sooner or later, companies
larger than Orion and PanAmSat will establish themselves in the market. Do-
mestic or regional PTT satellites with spare capacity will play a similar role.
Intersputnik, the Soviet-dominated Eastern European system was another pos-
sibility, and it began to offer its services to Western countries as its East Eu-
ropean customer base declined. Partly to preempt competitive entry, several
European operators began VSAT (very small aperture terminals) service, aimed
at the private networks of large users, particularly those with pan-European
communications needs such as automobile makers. Both British Telecom and
Mercury made plans for VSAT service. In 1990, only in Germany and poten-
tially in Britain could such service be offered outside the traditional carriers. In
1991, the European Commission drafted a directive to create a competitive
European satellite market. Soon, various PTOs began to coordinate their VSAT
plans, partly out of concern that VSAT service could be used to bypass national
networks.
Intelsat's revenues stagnated in 1989 and dropped by almost 20 percent in
1990. Jolted, it considered privatizing itself by selling 49 percent of its shares
to the public.
Even in the absence of competing satellites, Intelsat arrangements are threat-
ened by rivalry from already emerging competitors in private submarine cable
facilities. Two companies, Tel-Optik and Submarine Lightwave Cable Com-
pany (SLCC), applied for licenses to operate an international submarine cable
(PTAT) in the United States. The submarine cable applications did not raise
338 international Telecommunications Services
issues under the Intelsat agreements. Moreover, AT&T, the major American
owner of submarine cable systems, did not file any substantial objections. The
FCC thus moved expeditiously to grant the Tel-Optik application in 1985. Cable
& Wireless and E. F. Hutton participated in that venture. Soon one Bell re-
gional holding company, NYNEX, acquired an option, thereby raising the
question of the permissibility of Bell companies' international involvement in
this form, which was eventually denied. The plan was for two cables to be
operated in conjunction with Cable & Wireless in the United Kingdom, with
the first cable completed in 1989 and the second in 1992. Similar applications
were made and approved for Pacific routes.
Liberalization of entry led to the emergence of international carriers in Brit-
ain, Japan, and Sweden. Cable & Wireless (C&W) was the prototype for the
new generation of international carriers. As discussed in the chapter on U.K.
telecommunications, C&W's goal was to become the first global telephone car-
rier, and its strategy targeted the world's financial centers: London, New York,
Tokyo, Hong Kong, and Bahrain. It was already a dominant presence in Hong
Kong, where it owns the local telephone company. In Britain, C&W became
the sole owner of Mercury, which provided it with domestic long-distance ca-
pability within Britain and access to several European countries. C&W was a
major partner in the PTAT transatlantic fiber-optic cable to New York and held
transcontinental rights in the United States through its ownership of TDX, an
American long-distance carrier.
In Japan, the liberalization of long-distance communications also reached
international service. Two consortia applied for a license to provide such ser-
vice in competition with the previous monopolist KDD. The first was Interna-
tional Telecom Japan (ITJ), owned by fifty-three large users, including Mitsu-
bishi, Somitomo, Mitsui, Matsushita, and the Bank of Tokyo. The second con-
sortium was International Digital Communications (IDC), in which C. Itoh &
Co. and Cable & Wireless were the largest partners from among thirty-five
companies, including Toyota. The Ministry of Posts and Telecommunications
tried to convince the two ventures to merge. Part of the agreement would have
been to reduce C&W's share to 3 percent for reasons of "national security"
and to exclude it from a role in management. The British and American gov-
ernments opposed these restrictions as nontariff barriers. Both services were
launched in 1989. By 1990, they had captured 35 percent of all international
traffic between Japan and the U.S.
Challenges to the Traditional Rate System
In time the distortions of the traditional system reached the attention of the
public. Hugo Dixon of The Financial Times, in particular, argued that users
were overcharged by $10 billion because of cartel-like tariff arrangements. Costs
for international calls were estimated at $0.25 to $0,50, but rates averaged $1
per minute. It was estimated that $30 billion in revenues would generate $20
billion in profits in 1990 (Dixon, 1990a). BT reported 60 percent profit rates
internationally (Malik, 1990, p. 5). In 1990, the European Commission began

an investigation into artificially high international rates, and CCITT admitted
that its tariff recommendations needed to be revised. In defense, it was argued
that these profits subsidized residential local rates.
A related issue was the asymmetry in incoming and outgoing traffic to the
United States, which paradoxically created a major American deficit because
of its lower rates. A study by the International Institute of Communications
found that the United States originated 5.3 billion minutes of calls and received
only 3.1 billion minutes of traffic (Staple, 1990).
International telephone service in the United States grew more rapidly than
that of other countries, causing a rise in the deficit from $1.4 billion in 1987,
to $2.2 billion in 1989 (Stanley, 1988), and almost $3 billion in 1990. In 1990,
the FCC instituted a proceeding on this matter. The FCC required U.S. inter-
national carriers to negotiate a reduction in accounting rates in order to alleviate
the U.S. telecommunications deficit. And it announced that it could take uni-
lateral steps if necessary. Proposals were also made to reform the entire system
of settlements (Ergas and Patterson, 1989). In 1991, the OECD proposed re-
placing the traditional accounting rate system with access charges. The new
system would be cost-based and non-discriminatory against low-cost carriers.
But it would require international regulatory action.
Comparison of International Performance
We will now turn briefly to some comparisons of carriers. As was stated in the
introduction to this book, this is not a comparative study in the sense of mea-
suring the performance of various countries' PTTs and issuing report cards.
The study is concerned more with vertical changes over time than with hori-
zontal cross-country analysis. However, some comparative data will be pro-
vided in the following section, largely from OECD figures. But first, it is nec-
essary to discuss the difficulties of cross-country analysis.
The difficulties inherent in any comparison task can be shown with the ex-
ample of Sweden's Televerket reporting of its own performance.
One Televerket study shows the number of working hours required for an
average industrial laborer to pay for annual fixed telecommunication service
basket (Roos and Loenqvist, 1984). In Sweden, the basket required thirty hours
of work. For Great Britain, in contrast, it was eighty hours; for France, sixty-
five; for Germany, fifty-five; and for the United States, fifty. But nowhere in
the report is the telecommunications "basket" defined, and attempts to obtain
that information from Televerket were unsuccessful. Mitchell (1983) comes to
very different conclusions in that forty-two hours of work purchased one year
of residential service in the United States. In Sweden, the same service requires
fifty-five hours; in the United Kingdom, ninety-eight; in Italy, 111; in Ger-
many, 126; and in France, 165. Clearly, every country has different prices and
usage patterns for different components, permitting arbitrary comparisons. In
the United States, there are very different rates among customer types and
340 international Telecommunications Services
geographic regions; there are alternative competing carriers and substantial off-
peak discounts. For example, while Televerket assumes that U.S. terminal
equipment is rented by users, most Americans buy it, since that is much cheaper.
The same study uses the average industrial wage as a measuring rod, which
creates a bias toward richer countries, and within these countries toward those
with strongly unionized economies, where industrial wages are relatively high.
Televerket uses for local rates New York City as a typical representative. Un-
fortunately, for a variety of reasons, New York is at the high-cost end.
It is also difficult to define services. In the United States, operator assistance
and itemized bills are included. Swedish network quality, as measured by the
percentage of unsuccessful calls due to overload or technical faults, was at 2.4
percent of trunk calls in 1985 (Televerket, 1986, p. 9). In the United States,
the percentage of unsuccessful trunk calls was 1 percent for only twenty peak
hours per year, with the other times being lower.
It appears that in some comparisons virtually every judgment call ends up
with unfavorable assumptions, or noninclusion of favorable factors. Although
some simplifications are unavoidable, an analysis should not consistently err to
one side if it is to shed light.
In the decade between 1972 and 1982 alone, at least fourteen international
comparative studies of residential telephone rates were undertaken (see Mitch-
ell, 1983). Subsequent comparisons include Siemens (1988); Logica (1989);
McDowall (1987); and Morton and Donovan (1987). The results vary widely
but are consistent insofar as they usually favor the sponsoring administration.
Normally, the definition of the basket (local versus long-distance) can strongly
affect results, depending on the extent of subsidization of local calls. Given the
large number of variables to be considered and judgments to be made, one
could conduct defensible statistical studies that would show probably several
countries as the cheapest telephone country.
Perhaps the most thorough comparative study of rates and quality is a lengthy
OECD report issued in 1990. But it too makes numerous assumptions that are
problematic for the U.S. system, which has a structure considerably different
from European ones.
The OECD methodology uses an average ratio between fixed and usage-
sensitive charges (2:3 for residential subscribers 1:4 for business users). In ap-
plying these ratios to the U.S. situation, the OECD study apparently does not
take into account the fact that most U.S. monthly residential fixed service charges
include provisions for unlimited local calling. It is hard to understand how the
study could find residential fixed charges of $175.10, higher than those of busi-
ness charges, which were calculated at $174.67 (OECD 1990, p. 46). This
misconception skews the subsequent analysis. Other assumptions are similarly
unfavorable, such as the use of New York City as the comparison city; the
absence of quality factors; the lack of credit for operator availability and item-
ized billing; the assumption of AT&T as long-distance carrier; and the use of
only partial off-peak discounts.
The study itself concedes that: "on balance, the model works best for the
countries of Western Europe which tend to have similar tariff structures and
similar geographies" (p. 57). In consequence, it is best to use the following

figures as a comparison of traditional or semi-traditional systems, and to be
wary of applying them to the U.S., Canada, and perhaps Japan.
OECD calculations show that the lowest rates in Western Europe for a basket
of business telephone charges, including fixed and usage costs, are found in
Iceland ($365) and the Netherlands ($430). Swedish rates are $600 while the
highest charges are found in Austria ($1,409), Ireland ($1,320) and Germany
($1,326). The OECD average is $930 (OECD, 1990, p. 52). When purchasing
power parities are held constant (Figure 32.1), Iceland, Denmark, the Nether-
lands and Sweden are least expensive for telephone charges (Figure 32.la);
Portugal, Italy, and Ireland are the most costly. Germany is also above average
(OECD, 1990, p. 52).
The OECD's comparison of a basket of residential services shows similarly
that consumers in Iceland, Sweden, Denmark, and the Netherlands enjoy the
lowest rates, whereas rates in Portugal and Ireland are highest (Figure 32.2a).
Iceland's rates are only $191, representing 34 percent of the OECD average of
$354. Austrian consumers, on the other hand, pay $52964 percent above the
OECD average (OECD, 1990, p. 53).
Europe's lowest international charges are found in Scandinavia, with Den-
mark ($76) being the cheapest of all OECD countries for business service. On
the other end of the scale, Spain, Greece, and Portugal have the least favorable
business rates. Turkey and Portugal, at $165 and $131, have the highest rates
in Europe. They are followed by Greece ($124) and Spain ($123).
The residential basket of international charges (Figure 32.3) shows that among
the five least expensive countries in the OECD survey, Sweden (at $78) is the
cheapest European country. Turkey, Greece, Portugal, and Spain have the highest
charges in Europe.
The highest European charges for mobile services are found in Luxembourg,
France, and Germany. Germany, the highest at $2944, reflects a ratio of usage-
to-fixed charges of almost 3:1. France, which follows with $2630, has higher
fixed charges than Germany; Luxembourg ($2573) has the highest fixed costs
of the three.
The OECD's basket of mobile telephone charges calculated in purchasing
power parities (Figure 32.4) shows Iceland as the least expensive nation, with
the second cheapest, Denmark, over three and a half times more expensive
($189 versus $687). At the other end of the scale, Luxembourg had the highest
charges for mobile service ($2405) followed closely by Germany ($2358) and
France ($2316). The OECD average for mobile service is $1681 ($1116 for
usage and $565 for fixed charges).
The OECD comparison of charges for leased data lines with 1.5-2.0 Mbps
of capacity (Figure 32.5) shows a relatively even distribution for all OECD
countries except Germany. Here charges were about three times higher than the
OECD average. Another study by Logica showed that leased line rates varied
widely. It found for 64 kilobits per second, 100-kilometer leased lines, a range
from $545 (Mercury) to $4,904 (DBP Telekom). France ($3,403), Belgium
($3,457), and Spain ($4,332) were also on the high end. Comparable service
Figure 32.1 and 32.1 a. OECD Basket of Business Telephone Charges, in US$
Exchange Rates and PPPs, November 1989. (Source: OECD, 1990, p. 52.)
342
Figure 32.2 and 32.2a. OECD Basket of Residential Telephone Charges, in US$
Exchange Rates and PPPs, November 1989. (Source: OECD, 1990, p. 53.)
343
Source: OECD Tariff Comparison Model.
Figure 32.3 and 32.3a. OECD Basket of International Telephone Charges,
Ranked by Country, November 1989. (Source: OECD, 1990, p. 61.)
344
Source: OECD Tariff Comparison Model.
Figure 32.4 and 32.4a. OECD Basket of Mobile Telephone Charges, Ranked by
Country, November 1989. (Source: OECD, 1990, p. 64.)
345
Figure 32.5. Basket of Charges for 1.5/2.0 Mbit/s Leased Lines. (Source: OECD, 1990, p. 70.)
Figure 32.6. Waiting Time for Telephone Installation, Selected OECD Member Countries, 1979-1987. (Source: OECD, 1990, p. 126.)
Table 32.1 Fault Reporting Rates in Selected OECD

Countries
Faults per 100

Country Year main lines Comments
Belgium 1985 41.3
Denmark 1988 31.7 Jutland Telephone
Finland 1987 23.0
France 1986 21.7
Greece 1987 62.5
Ireland 1988/89 52.0
Japan 1988 2.2
Norway 1988 30.9
Portugal 1985 59.0
Spain 1986 46.3/31.0 Complaints/Faults
Sweden 1986 16.0
Turkey 1989 19.4 Jan. -Oct. only
United Kingdom 1988 22.0
United States 1990 July-Dec.
Consumer trouble,
New York Telephone 4.7 all types
SW Bell 2.4
Unweighted aver-
US West 1.8 age
Sources: FCC, Industry Analysis Division (J. Kraushaar, 1991, communication);
New York State Public Service Commission, 1991, communication; OECD, 1990,
p. 131.
costs in the U.S. $267 using MCI. For 2 megabits per second capacity, prices
in Europe ranged from $3,415 (Mercury) to $49,044 (DBF Telekom). Other
major European countries charged more than $10,000. In the United States,
this service is offered by U.S. Sprint for $3,100 (upwardly prorated T-l line)
(Hey wood, 1991).
Price is not the only performance dimension of significance. OECD figures
adapted from the ITU statistics reveal great variation in the amount of time it
took to obtain telephone service (Figure 32.6). The OECD average shows a
significant drop in the amount of waiting time between 1979 and 1987 to around
ten months, and the installed base of main lines grew 37 percent. Potential
subscribers in Greece and Turkey had to wait six and seven years, respectively,
for installation. This stood in stark contrast to the situation in Denmark and
Finland, where waiting time was negligible. Greece was the only country among
those shown where the wait actually increased between 1979 and 1987, while
at the same time its installed base of main lines increased 71.1 percent. Portu-
gal and Ireland both cut waiting periods significantly between 1979 and 1987,
but still ranked weak on this dimension.
The growth rates of installed lines in OECD countries for the same eight-
year period were remarkable (see also Figure 32.6), with Turkey (211 percent),
France (106 percent), and Ireland (93 percent) leading the way. Growth was
Table 32.2 Public Payphones in OECD Countries
Payphones per 1,000

Country No. of Payphones Population
Australia1 34,135 2.19
Austria 28,018 3.71
Belgium 11,373 1.15
Canada' 143,682 5.86
Denmark 5,792 1.13
Finland 17,739 3.61
France 213,126 3.86
Germany 162,458 2.66
Greece 21,815 2.20
Japan 910,000 7.54
Luxembourg 640 1.75
Netherlands 8,020 0.55
Norway 13,353 3.22
Portugal 18,126 1.77
Spain 1 40.7202 1.06"
Sweden 18,700s 2.10
Switzerland 57,526 8.90
Turkey 35,500 1.55
United Kingdom 76,500. 1.35
USA 1,714,055 7.18
Notes: 1. 1985 data.
2. Call box telephonesi.e. excluding public call offices.
3. 1988 data including privately operated payphones.
Source: OECD, 1990, p. 129.
less dramatic in percentage terms in more mature telephone systems (e.g., 26

percent in Switzerland).
OECD statistics reveal the range of faults, apart from network congestion,
reported per 100 main lines. Only fourteen countries compiled such statistics
nationally (Table 32.1). Japan's ratein 1988, 2.2 faults per 100 linesis far
below that of any European country participating in the survey. Greece (62)
has the highest rate, followed by Portugal (59) and Ireland (52). Some of Eu-
rope's most reliable networks include Sweden at 16 and France at 22 (both for
1986). In the United States, corresponding figures for two Bell regional com-
panies for which figures are available are 2.4 for Southwestern Bell and ap-
proximately 1.8 for US West (Table 32.1). The OECD estimates a 40 percent
likelihood of line faults occurring in OECD countries (OECD, 1990, p. 131).
OECD figures also show that the availability of pay phones per 1000 popu-
lation in European countries (1985) ranged from a low in the Netherlands of
0.55 to a high in Switzerland of 8.9 (Table 32.2). Between these extremes lay
the United Kingdom (1.35), Germany (2.66), and France (3.86). In the United
States, the number was 7.18. But this did not include the many privately op-
erated pay phones. The bulk of the world's payphones are in Japan and the
United States.
33
The Economics of ISDN Integration
The past chapters have discussed the evolution taking place in various coun-
tries. The next sections will deal with more general issues of European com-
munications. The first of them is ISDN.
The Setting
The integrated Services Digital Network (ISDN) is a major development prior-

ity of many European PTTs and equipment manufacturers. It is also one of
those subjects at once complex, obscure, and important. Deterred by the dense
technological jargon that surrounds ISDN, only a few social scientists, let alone
commentators or politicians, have ventured near the subject (Schiller, 1984).
Similarly, there has been no public discussion. Instead, decisions on this major
infrastructure investment were made almost entirely outside the public view by
engineering bureaucracies in government and established equipment manufac-
turing firms. This is not to say that the basic concept of ISDN is flawed as a
matter of technology or policy. (The author, while serving as a member of the
New York State Public Service Commission, was supportive of the concept
and initiated an ISDN trial that linked five different carriers and multiple users.)
But it is remarkable how the process of decision making had been transformed
into expert discussions of technical specifications. A book by a participant in
the international ISDN negotiations in the Consultative Committee on Interna-
tional Telephone and Telegraph (CCITT) and its Study Group XVII provides
detailed accounting of the discussion that lasted several years (Rutkowski, 1985).
Remarkable in these negotiations is the near total absence of references to con-
siderations of the cost or economic issues involved.
The Concept of ISDN

The term ISDN encompasses several subconcepts; thus, some confusion exists
about its primary rationale. It is, first, a movement toward end-to-end digitali-
zation. As such, it continues a development of several decades, accelerated by
350
The Economics of ISDN Integration 351
advances in computers, from analog to digital electronics. Digitalization has

been moving from data processing to telecommunication transmission and
switching, similar to its spread to consumer electronics, and is moving to
broadcasting and motion picture technology. The evolution of the network toward
digitalization precedes ISDN, and the development of ISDN lies squarely within
that trend.
The second element of ISDN is that of upgrading user access to the telecom-
munication network to a higher data transmission rate. In past decades, the
increased merging of computers and communications, together with the greater
information needs of users, have led to a vast increase in data communications
traffic. Data communications rose during the period from 1970 to 1985 almost
fortyfold, and communications and data processing increased about fivefold
(NTIA, 1983). This raised the importance for data communications links of a
high capacity, greater than could be provided on analog networks. Digital
transmission is also a more efficient means for the transmission of voice traffic
at points in the network where it is highly concentrated.
Whereas good-quality, regular analog voice-grade switched communications
links support a transmission rate of about 9.6 kbps, international ISDN rec-
ommendations provide a much higher total transmission of 144 kbps for two
switched channels of 64 kbps (B channels) chosen for voice quality, and one
"signalling" channel for housekeeping functions of 16 kbps (the D channel).
It is this CCITT-recommended narrowband system that will be referred to in the
following as "ISDN." Although analog technology does not stand still (newer
modems can upgrade the data capacity of analog lines to over 20 kbps), the
data transmission superiority of ISDN over the existing public network is clear.
The third element of ISDN is integration and is much weaker in its rationale.
ISDN claims to put together separate communications networks into one unified
superpipe. From the technologist's perspective, this is a more elegant solution
than duplication and multiplication.
Almost all ISDN discussions start out invoking the wasteful existence of
several parallel telecommunications networksif they bother to articulate a ra-
tionale at all. Some historic perspective is necessary: Parallel networks existed
from the moment that the telephone emerged more than a century ago as a
voice network separated from the already existing telegraphthe data network
of its time. Eventually, telex networks replaced much of the telegraph; still
later, the advent of computer communications led to specialized data transmis-
sion, and in particular to packet-switched networks, which move certain types
of digital data better than the standard telephone networks.
Digitalization often led to the establishment of integrated digital networks
(IDN) for various data services. But it did not yet reach the subscriber loop
and subscriber equipment itself (though it came in many instances to within a
few hundred feet of them). Thus, subscribers in need of digital data commu-
nications of more than slow speeds had to operate over special lines outside
the public switched analog network. ISDN represents the completion of net-
work digitalization by going the last mile and reaching subscribers and their
352 The Economics of ISDN Integration
equipment. It thus provides the capability to dispense with parallel lines for
analog voice and for digital services and instead puts them together in one
integrated bit stream of digitalized voice and data.
As long as ISDN simply implies no more, it is hard to find fault with this
development, which is part of a broad technological trend. ISDN, however, is
more than such upgrading, in Europe and elsewhere; it is also part of a business
and political strategy of telephone administrations. "Integration" is thus not
simply an issue of technology.
Economies of Scope
The classical arguments in favor of integration are the benefits of "economies

of scale" and the joint production benefits of "economies of scope." The latter
occurs since the duplication of equipment and personnel is eliminated; simi-
larly, less spare capacity is necessary to handle peak demand loads in integrated
production than is required to provide for them separately. If one denotes C as
cost, and assumes two services, voice service (V) and data service (D), econ-
omies of scope exist if the joint cost is less than the two separate costs; that is,
if the following inequality holds:
(1)
That this relation exists is generally asserted by ISDN proponents as a matter
of a priori reasoning. The same logicthat substitute and complementary prod-
ucts are cheaper if jointly providedapplies similarly to such product pairs as
orange juice and beer, beer and aluminum cans, aluminum and aircraft, and
aircraft and communications equipment. In the extreme, if one truly wants to
eliminate all duplication, the entire economy should consist of one giant inte-
grated enterprise. Thus, the significance of economies of scope is not necessar-
ily as clear-cut a matter as it may appear at first.
Furthermore, the question is not how best to structure a new network, but
how best to upgrade an already existing network. In the short term, virtually
all capital in the networks V and D is sunk, whereas a modification for the
integrated network (V + D) may require new investments /, as well as the
premature retirement R of some equipment used for the separate networks.
Hence, the test for integration should be expanded to require
(2)
The adjustment costs C(I) and C(R) can offset the pure economies of scope. In
the longer term, their significance diminishes as equipment gets replaced natu-
rally. But there are costs to a very gradual integration too, because it requires
parallel technologies of different generations to coexist for a long time.
Another problem with the argument of economies of scope is its implicit
assumption of static cost functions. That is, costs are defined by the relations
C(V), C(D), and C(V + D) without a provision for change. However, separate
networks are partial substitutes. Hence, a rivalry between separate specialized
networks can provide a competitive environment that can lead to cost reduc-
tions and technical innovation. In other words, the cost curves in a separate
environment can be dynamic. What starts out as the cost relation C(V) and
C(D) can become, under rivalry, C'(V) and C'(D), with the assumption that
(3)
The integrated network should then be more accurately required to meet the
condition
(4)
which is much less obvious than relation (1).

It is, of course, true that institutionally almost no country's telecommunica-
tions systems permit competitive networks under separate control. But this is
not to say that no rivalry exists at all. Most studies of organizational behavior
within government bureaucracies would lead one to predict that different net-
work services will have their own advocates, and constituencies, both inside
and outside of the organizations. Hence, at least a bit of rivalry between dif-
ferent networks could exist even in a monopolistic situation.
Another problem with economies of scope is their lack of generality when
the number of services is large, as demonstrated with the earlier orange juice
example. For example, suppose that instead of two networks V and D there
exist n different types of data networksfor example, for different transmis-
sion rates, different error rates, and so on, denoted by D\ . . . Dn. Similarly,
let there be a third type of telecommunications network, used for transmission
of television programs and denoted by TV. Then we have the separate cost
functions
(3)
It is far from clear where the economies of scope exist. Even if we assume that
each service has economies of scope with its immediate neighbors, this does
not prove that a total integration across the entire spectrum of networks would
produce economies of scope over stand-alone provision. To prove total integra-
tion, one would have to show economies of scope between every pair of ser-
vices (e.g., between voice and television). Furthermore, each service may have
its own and different economies of scale, and a bundling may lead to subopti-
mal pairing.
Thus, the question of integration then becomes one of choosing clusters of
services to integrate, such as [V, D\, D3], or [V, D2, TV]. It is far from obvious
which groupings of service networks are optimal. Should high-speed data be
integrated into broadband cable systems, as is the case in some American cable
systems (Noam, 1986)? Or should it be part of an intermediate integration of
data networks? Should it be integrated with voice service, even at a high cost?
Should it stand on its own? Or should it simply not be provided? None of these
questions is answered by the simple reasoning that avoidance of duplication is
cost efficient. As the various telecommunications services are bundled into
packages of various integration, invariably some services will be dropped. In-
tegration is a standardization process, which is always a trade-off between the

cost reduction of compatibility and the benefits of diversity. A process of inte-
gration is hence usually a process of reduction of options.
Furthermore, even where total integration may be cheaper than stand-alone
services, the integration may not be stable"sustainable"if some of the
services could drop out of the integrated package. To understand this, it is
necessary to distinguish between three cost concepts: first, the stand-alone cost
of service of type /, denoted by C,-. Second, the cost of operating / as part of a
service package, which is C(/,). Third, the incremental total cost /C, of adding
it to the service package.
1C may be lower than the cost of integration, positive externalities of / on

the services, if it acts, for example, as a backup. Suppose, for example, that
the cost of providing / i s 10 as a stand-alone service and 20 as an integrated
service because of technical problems of integration. This would suggest that
integration is uneconomical. However, total incremental cost of the integrated
service may be only 5, because of the positive externalities on the other ser-
vices. Thus, integration would be economically efficient, but it would not be a
stable solution. If each network could act independently, it would drop out of
an integrated network, since this would reduce its cost from 20 to 10. Its pos-
itive network externalities to the overall service would then be lost to the re-
maining network. Thus, the integrated system, if it is voluntary, is unstable, or
"unsustainable."
The sustainability argument for integration, presented here in very simple
terms, is one not made specifically by the telecommunications administrations.
It is based on a more complete general theory of industrial organization (Bau-
mol et al., 1983), with which PTTs are gradually becoming familiar (Weiz-
sacker, 1987). In any event, the sustainability argument is a theoretical one and
depends on the underlying cost characteristics, which may or may not exist for
telecommunications networks.
Economies of Scale
We now widen the discussion to economies of scale, the intellectual ancestor

of economies of scope. This concept, probably the single favorite idea in the
intellectual armory of telecommunications administrations, states that the cost
of providing additional users continuously decreases, and that is thus cheaper
to service all users by one large entity. The concept, closely related to the
vague notion of "natural monopoly," has been questioned on empirical grounds
as to its actual existence beyond a certain size. (For a review of the empirical
literature, see Meyer, 1980.) It can also be challenged along the lines used
above, namely, that the inefficiency of monopoly more than offsets its econo-
mies of scale. In the American context, for example, the system of rate-of-
return regulation created built-in incentives toward an overcapitalization, known
as the Averch-Johnson effect. As discussed earlier, competition can move cost

curves downward, which may result in greater efficiencies than through relying
on a movement along a downward-sloping static cost function.
In the context of integrated service, economies of scale are used to argue
that the integration of services creates cost efficiencies by bringing various new
communications services to the population at large. Whereas it would be un-
economic for a residential or small business user to have digital data links, this
would become affordable in an integrated universal network. By making such
services prevalent, their cost per user drops, benefitting also the earlier large
users.
If strong economies of scale exist in ISDN, one policy implication could be
to subsidize the early stages of its growth. Even if its cost cannot be recovered
for a while, eventually costs come down and benefits rise. Early users deserve
a subsidy, because their participation lowers the cost for previous and subse-
quent users; otherwise, they may well never sign up and start the chain of
economies of scale down the cost curve. (For a discussion of network exter-
nalities, see Katz and Shapiro, 1985.) This argument is a plausible one for a
subsidy, though it does not prove a case for integration or for governmental
control. But it is true for almost all start-up operations, in any line of busi-
ness, whether public or private. To judge the justification for subsidies in the
ISDN context one needs to have information about the size of the economies of
scale.
The need for a start-up subsidy can be used to justify a variety of restrictive
policies:
1. Prohibition against alternative network optionscertainly private, but also

publicin order to gain the economies of scale and avoid a loss in reve-
nue, which is termed cream skimming.
2. A tariff structure that reduces the required subsidy as much as possible
by increasing revenues through a rate system that charges according to
the value of the service to different users (i.e., their elasticity of demand),
rather than on the basis of cost. Value to userwhich is not a pricing
criterion for economic efficiencymeans discrimination between differ-
ent users on the basis of the service's value to them. Charges are made
on the basis of actual usage rather than on the basis of cost. To be effec-
tive over time, price discrimination requires market power, although such
power need not be governmental. Introducing such usage-sensitive pric-
ing was part of FIT policy; the German Bundespost introduced it to the
displeasure of large users of data communications and of equipment firms
such as IBM. (A volume-based pricing system was employed by IBM
itself years ago when it required the use of IBM-produced punched cards
to go with rented IBM computers, thus creating a usage metering and
pricing system, and by Xerox Corp., which rented its equipment and
charged according to the metered number of copies.) One optimal dis-
criminatory policy is known as Ramsey pricing (i.e., charging according
to the inverse of the demand elasticities of different customers).
Table 33.1 Cost of Providing Voice and Data Service to

Firms and Residents (Schematic)
V D (V+D)
F 8 9 16
R 11 12 21
(F + R) 16 20 35
The two policies are interrelated. The establishment of price discrimination

that is not based on cost leads to incentives for arbitrage and for competitive
alternatives. Hence, it becomes more important than ever to stem these threats
of cream skimming by prohibitions against competitive networks of reselling.
In other words, the start-up subsidy necessary for the service becomes an
argument for protection against competition and for maintenance of monopoly.
Indeed, a highly perverse incentive is built in, because the greater the required
subsidy, the greater the political support will be for a monopoly status that
reduces its cost to the taxpayers. An expensive project helps raise not only the
economic but also the political barriers to entry. This encourages the introduc-
tion of successive large projects that remain in deficit long enough to justify
barriers against cream skimming.
Scale and Scope
Economies of scale are used as an argument for the expansion of network

integration to all classes of participants in the public network. As with most of
the pro-integration arguments, this is true only under certain conditions. The
theoretical problem with the argument can be demonstrated with the following
table. Assume again two types of networks, voice (V) and digital data (D), and
two types of users, firms F and residents R. Service can be provided separately
or jointly across services (V + D), across user classes (F + R), or across both
services and user classes. Table 33.1 demonstrates schematically a situation in
which economies of scale and scope exist yet the integration is uneconomical.
Each number corresponds to the cost of providing a service of type / to user
class of type j.
In Table 33.1, economies of scale exist in the vertical columns, where inte-
gration across customers lowers costs for voice service from 8 + 11 = 19 to
16 and for data service D from 9 + 12 = 21 to 20.
Similarly, economies of scope exist horizontally in the rows of Table 33.1.
For firms they reduce cost from 8 + 9 = 17 to 16, and for residents from 11
4- 12 = 23 to 21. Furthermore, full integrationboth across services and
customerscombines economies of scale and scope and lowers total costs to
35, where they would be 40 in total separation and 36 and 37 in partial inte-
grations. This would argue for total integration. The economies of integration
(scale and scope) are greater than each of the subeconomies. Both horizontal
or vertical integration are cost-reducing.
Table 33.2 Benefits of Voice and Data Service to Firms and

Residents (Schematic)
V D (V + D)
F 9 10 19
R 11 6 17
(F + R) 20 16 36
This analysis, however, is based purely on cost considerations and is devoid

of any discussion of revenues, benefits, and demand. If the price for each of
the four subcategories is arrived at by simply dividing total cost equally, the
price for each subservice would be 35/4 = 8.75. This means that firms are
now charged for the integrated voice service 2 x 8.75 = 17.5, rather than the
16 of partial integration or the 17 of full separation. Thus, they pay more than
before.
Of course, it would be possible to structure a pricing scheme that would
distribute the cost savings to all types of usages. This would mean price dis-
crimination according to usage type or user category, or both. Yet such price
discrimination would establish incentives for arbitrage. Under certain circum-
stances no price vector could exist that would not make it preferable for some
service to be provided separately (nonsustainability).
A second problem exists if one considers, realistically, that the different
service types are of varying benefit to their users. For example, the benefit of
data services users may actually be quite small to residential users. Let us
assume for illustration that the benefits of the four categories are given by Table
33.2. These benefits also establish the maximum willingness to pay.
In this situation it will increase welfaredefined as benefits minus coststo
integrate voice and data on the firm level, since benefits are 19, against costs,
which are 16 (from Table 33.1). But it would not cover cost to integrate the
residential service, the cost of which is 21, whereas maximum revenue would
be only 11 + 6 = 17. Similarly, it would make sense to integrate voice
services for firms and residents, but not for data. Any inclusion of residential
data service would have a deficit of 16 - 21 = -5. This could be done
through an outside subsidy, as by using the total cost gains of integration (8 +
9 + 11 + 12 - 35 = 5) to offset the deficit, or by charging some or all
users and services above cost, up to the limit of their willingness to pay.
Of course, this means that three of the service types do not pay less, because
of integration; to the contrary, they are likely to be called upon to pay more
than before, since earlier they paid less than their benefit (i.e., had a consumer
surplus). On the other hand, the fourth service benefits. An otherwise lost ben-
efit of 5 is gained, though at a cost of a subsidy of 5.
Another pricing policy would be to eschew a subsidy from one user type to
another and to limit it to a subsidy within a category. This could be done by
requiring the residential user to obtain a package of the two services, or none
at all. Maximum total payment would be 21, from Table 33.1. This would not
be sufficient, and hence an additional subsidy of 4 would be necessary, either

from outside governmental services or, more likely, from the other user cate-
gory. This then creates the incentive to go beyond partial to full integration, to
find more services or user surplus for purposes of subsidy.
But in terms of welfare, is this the optimizing policy? Suppose that instead
of integrating all four services, one integrates only three and forgets about data
service for residents. This would generate costs of 28 (for the three separate
services) and benefits of 30, for a total net welfare of 2 (30-28), whereas in
the fully integrated system net benefits are only 36-35 = 1. In other words,
aggregate societal benefit may not be improved by a cross-subsidy to a user
whose benefits from the service are small, even where one can show economies
of both scale and scope. Thus, technical efficiencies by themselves do not prove
societal benefits of integration in the absence of consideration of user benefits
(i.e., of demand conditions). And it is not enough to show that there are some
benefits to residential users from data service, which is what ISDN boosters do
with anecdotal evidence. Instead, it is necessary to illustrate the magnitude of
these benefits relative to their costs.
This discussion can be extended beyond a 2 x 2 matrix to additional types
of telecommunications services and additional user categories. What is intended
here is only a schematic sketch of economic arguments that could be made
more rigorously. Economic reasoning suggests that a trade-off exists between
variety and specialization and that an optimal product diversity exists (Lancas-
ter, 1975; Salop, 1979; Hemenway, 1975; Spence, 1975). There can easily be
too much standardization or inefficient standardization (Farrell and Saloner,
1985, 1986a; Besen and Johnson, 1986). '
An important cost of systems standardization, such as that sought for ISDN,
is that it tends to lock equipment manufacturing, systems development, and
user applications into a pattern revolving around the one standard that may soon
be obsolete technologically. However, for any one agent to move to another
standard may prove prohibitively expensive. Hence, standards include an ele-
ment of discontinuity and retardation of innovation. For example, the keyboard
of typewriters and other input terminals has remained with the traditional
"QWERTY" key system, even though studies have shown that a different
configuration (Dvorak) of keys increases typing speed and reduces errors (David,
1984).
Such discontinuities exist in particular when, as in a network, positive exter-
nalities are generated by each participant. To leave a standardized integrated
system is costly, because those departing lose the benefits conveyed by the
other members. Only large agents could then be expected to move on their own
toward a new technical specification. However, because their departure im-
poses a cost (lost positive externalities) on the remaining adherents to the net-
work, the decision to move to another standard may be blocked by administra-
tive fiat as imposing a social cost.
Similarly, the early announcements of new standards, such as those for ISDN,
is not necessarily the efficient information-enhancing act claimed by its propo-
nents. Instead, it can be the act of a monopolist or of oligopolists aimed at
preventing users from moving toward superior technology that they would oth-
erwise choose, by signalling to them that they will lose the benefits of leaving
the coalition around the major standard that the monopolist controls (Farrell
and Saloner, 1986b).
The investment in a major upgrading of the network has the effect of raising
barriers to entry in the following ways: First, it increases the required initial
investment that a potential rival needs to match the upgraded technical capabil-
ities of the existing network. Second and similarly, where there is a trade-off
between fixed costs and marginal costs, as there often is, the latter are lowered
by the investment, making it more difficult for a rival to enter and match mar-
ginal cost pricing. Third, by raising the initial investment, one can stretch the
range of economies of scale (declining average cost) and thus of "natural mo-
nopoly." The trough of a U-shaped cost curve is shifted to a higher level of
production by an increase in fixed costs.
The theoretical analysis of the past sections demonstrated that the economic
case for ISDN, on purely analytical grounds, has not been made persuasively.
Empirical figures have not been presented in support of ISDN which would
prove the advantages of integration in light of the issues discussed earlier. This
is not to say that such a case cannot be made. However, the battles over ISDN
have never been fought on the grounds of economics. To understand why, one
has to go back to the history of ISDN. This will be done in the next chapter.
34
The Political Economy of ISDN
Computer Industry versus Telecommunications Providers

Debates over narrowband ISDN mirror the key conflicts in the information
sector. In the 1960s and early 1970s, the PTTs had failed to appreciate the
potential of data communications, and much of equipment, packet switching,
and so on, was developed outside the telephone industry and its traditional
suppliers. For the PTTs, the threat was that the computer industry would reach
standards that would be incompatible with the existing public or specialized
networks and would instead accelerate the creation of private networks. Com-
puter standards were the purview of the International Standards Organization,
ISO, whereas telecommunications standards were those of the ITU's CCITT,
the Consultative Committee on International Telephony and Telegraphy. Though
significant coordination between ISO and CCITT exists through Joint Technical
Committee 1 (JTC1), there is also a built-in rivalry between administrations
and private network operators. This is partly due to "forum shopping" by
interested parties, with the computer community preferring ISO. Joint devel-
opment of standards was undertaken through the JTC1. Given the complexity
and contiguity of technology, the dividing line between standards set by net-
work authorities and those of computer manufacturers is a meandering one.
Standards control varies, depending on the type of service (e.g., telephones,
telex, teletext, electronic mail) and depending on the layer in the OSI reference
model.
After years of preliminary research, the CCITT decided to study ISDN in
1980 and to issue recommendations. It established Study Group XVIII headed
by Theodor Inner of the German Bundespost, which proceeded very actively
to issue recommendations (Irmer later became head of CCITT, having won
respect for his ISDN work). While these efforts were significantly escalated in
1980, the origin of CCITT-defined ISDN (narrowband with two 64 kbps chan-
nels and a signaling channel of 16 kbps) goes back to a 1970s view of tech-
nology, markets, and strategic vision (A. Rutkowski, communication, 1990).
For a detailed chronology of these proceedings, see Rutkowski (1985). ISDN
was left in the domain of the technical study groups (H. Marks, 1987, com-
munication).
The ISDN concept was, in part, a defensive response by the telephone ad-
ministrations to assert control over the network. Technologists and inside ex-
360
The Political Economy of ISDN 361
perts may see ISDN merely as a technical issue, but massive investment pro-
grams are rarely free of other considerations. ISDN is no exception, unless one
entirely ignores the historic struggles over exclusivity in which most major
PTTs are engaged.
A similar implicit assertiveness to expand the range of the PTTs' activities
also lies in the potential of future integrated broadband networks (B-ISDN, or
IBN). This is likely to become a major issue. The cost of glass fiber terminal
equipment will soon permit the use of fiber to a point near the home or even
into it. The development of broadband switching has also progressed enor-
mously. Given these favorable technological and economic trends, telephone
carriers will be able in the foreseeable future to provide video program services
(even switched ones), and become rivals of traditional cable television carriers
and over-the-air transmitters. Conversely, cable carriers could engage in alter-
native local telecommunications transmission, challenging the FTT exclusivity
over the local loop (Noam, 1986). The British government's duopoly review
of 1990 recognized and encouraged such evolution. 1
The conflict of telecommunications providers with the computer industry
centered on IBM. For a long time, many Europeans endowed IBM with near-
mythical abilities, which justified drastic government actions. The preoccupa-
tion with IBM led to recommendations to use telecommunications, which gov-
ernments control, as a lever over the computer industry, now that the two
sectors are overlapping. This strategy was spelled out in the influential 1978
Nora-Mine report to the French president, which was discussed earlier, and
which, with its combination of brilliance, gloom, and nationalism, deeply in-
fluenced French and European policymakers and intellectuals.
Controlling the network system is thus an essential objective. This requires that its
framework be designed to serve the public. But it is also necessary for the state to
define access standards; otherwise the manufacturers will, utilizing the available
routes but subjecting them to their own protocols. The level of standardization will
thus shift the boundary between the manufacturers and the telecommunications or-
ganizations; it will be a bitter struggle, since it will develop out of a reciprocal
play for influence. But the objective of public control indicates the strategy to
follow: increase the pressure in favor of standardization [Nora and Mine, 1980, pp.
74-75].
ISDN and its standardization thus promised to be a tool of the kind advocated
by Nora and Mine at the time when ISDN planning moved into its serious
stage.
United States versus the PTTs

Given this perception, it is not surprising that there were conflicts between the
United States and the major PTT countries. The concept of ISDN, by itself,
does not require exclusivity and power. Strictly speaking, all it means is that
the same communication link is able to provide a range of digital telecommu-
362 The Political Economy of ISDN
nications services. It could be provided by multiple and competing ISDNs or

by private customized ISDNs. However, the usage has, at least within the PTTs,
implied a de facto exclusivity. The abolition of duplicative networks is stress-
ed as a main goal of integration. Economies of scale and scope are the argu-
ments. With such justification, the idea of permitting rival networks seems self-
defeating, and deeply at odds with the motivations for ISDN. In the United
States, on the other hand, an ISDN cannot be one of exclusivity, neither geo-
graphically nor functionally, but must involve the interconnectivity of multiple
networks, several of which could be ISDN types.
For United States policymakers, the ISDN creates several problems. Much
of the early ISDN discussion was technical, and not necessarily in tune with
the broader concerns of U.S. telecommunications policy (Marks, 1984). It was
advanced by the still undivested AT&T system, for reasons similar to those
motivating the PTTs.
Because ISDN calls for standardization, integration, and international coor-
dination, centralized telecommunications systems such as those in Europe had
an easier time formulating their ISDN goals than in the United States, where
such central decision making does not exist in telecommunications. Particularly
after the AT&T divestiture, American industry was fragmented, and coordina-
tion was difficult to achieve. The Bell operating companies formed a central
technical organization, Bellcore, to provide some of the services that the old
AT&T had provided; but by the nature of decentralization, it could move only
slowly. Furthermore, independent telephone companies, including ones as large
as GTE, are outside of Bellcore, as are the computer and component industries
and telephone uses. It was a matter of concern in America that international
standards could be agreed upon by entities whose operative philosophy was
very different from the American one.
As the chief of the FCC's Technical Analysis Division wrote:
[SJince only two other countries out of 159 members of the International Telecom-
munication Union (ITU) have any aspect of competition in their carrier industry,
technical decisions made in this forum have a tendency to be biased toward mo-
nopoly. This is not to imply that other countries seek to impose their industry
structure on the United States intentionally; rather they are focusing on their own
situations and honestly do not understand the implications of the US industry struc-
ture [Marcus, 1985, p. 33].
The FCC declared that customer provision of the network termination device
(NT1) should be a national option.2 It described as fundamental that CCITT
recommendations must be flexible enough for national options and that the
American distinction between basic and enhanced services must be maintained.
It turned out to be complicated to work out the proper interface point be-
tween the domain of the network and that of the users. Essentially, three de-
marcation points are possible, known, respectively, as the S, T, and U points
in CCITT terminology. The S point is closest to the end user, and hence carries
the network-controlled portion furthest upstream.3 The PTTs wanted point 5 as
the interface between user and network because they could control equipment
standards up to that point. Users and independent equipment manufacturers, on

the other hand, did not want to see PTT control reach that far.4 The FCC
supported U as the demarcation point in order to encourage the development of
more advanced and versatile terminal equipment.
Hegemonic ISDN
ISDN has always meant different things to different people, from a simple and
partial upgrading of digitalization to more ambitious undertakings encompass-
ing everything up to video transmission. However, recent international stan-
dardization efforts have narrowed the term for the present. But these different
definitions revolve around technology. A different classification of ISDN is
based on its purpose: it distinguishes "hegemonic" from "upgrade" ISDN.
The latter is a step in the technical evolution of telecommunications. The for-
mer, on the other hand, is part of a general attempt at maintaining control over
networks.
The key element of the coalition between bureaucracy and established equip-
ment manufacturers is strongly evident in the discussions on ISDN, and for
good reason. In most European countries, the decades following World War II
were periods of very active expansion in telecommunications. The combination
of war-damage repair and the expansion of telephone penetration from the busi-
ness and upper classes to universal use has kept telephone authorities busy and
manufacturers profitable. But this expansion reached its natural plateau when
most households were served. Whereas in 1960, only some 20 percent of the
German households had telephones, by 1982, over 80 percent did. In France,
after a major development push, penetration by 1984 was over 80 percent. The
implications for equipment manufacturers were clear; the domestic market was
close to saturation in terms of standard equipment and would decline; this would
leave them only with the export market, of which, according to an OECD
study, only 15 percent was free from protectionism. Furthermore, the absence
of a strong domestic market is also likely to make exports more costly, since
they would not benefit from economies of scale in production. One way to
activate the sagging domestic market was therefore to launch an ambitious pro-
gram of upgrading, and ISDN was just such a project. Figure 34.1 illustrates,
for Germany, the role of ISDN digital upgrade for both transmission and switching
(Schon, 1985).
The PTTs, seeking support for their domestic position, courted their tradi-
tional allies by dangling ISDN before them. Germany's Bundespost, for ex-
ample, argued that ISDN was important for the export success of domestic
industry and that it required the ability of the Bundespost to play a role in the
equipment supply field, a role that was under attack by Germany's Monopoly
Commission and the Ministry of Economics.
The PTT that takes on the leading role internationally when a new service is stan-
dardized gives the communications industry in their country a big head start in this
Figure 34.1. Annual Investments in Transmission and Switching Technology in

Germany. (Source: Schon, 1985.)
service. . . . The Deutsche Bundespost has so far played a prominent part in this
context. . . . Anyone who blocks this influence in his own country damages the
innovative force of a future technology and ultimately the entire economy [Schon,
1984, p. 22; emphasis in originall. 5
Network Segmentation and the Open Network Architecture Concept
In America, the idea of all telecommunications under one organization was

never palatable. Even in AT&T's heyday, it shared the field geographically
with almost 2000 other independent telephone companies (covering more than
half the country and 20 percent of subscribers) and functionally with the do-
mestic and international record carriers. Deregulation and divestiture acceler-
ated the segmentation of networks. These developments led to yet another and
still more radical approach that that begun to emerge, known as open network
architecture (ONA). This approach is not incompatible with ISDN in terms of
technology; and in the United States they are likely to coexist. But in terms of
underlying philosophy, ON A derived from a diametrically different concept of
the future environment of the telecommunications networks, and of the differ-
ent role of the major carriers in it than those held by ISDN's major international
supporters.
Open network architecture expanded the concepts of service alternatives
and network fragmentation into the very core of the networks, and lowered
barriers to entry for rival and varied communications services. ISDN, in con-
trast, raised entry barriers by providing a highly integrated network. ONA un-
bundles, whereas ISDN consolidates.
The concept of open network architecture must be distinguished from the
similarly named open systems interconnection (OS I) of the International Orga-
nization for Standardization which provides a definitional framework of seven
broad layers of the entire network process. ONA takes this further by not only
going into more detailed subfunctions of several of these layers, but also pro-
posing their functional separation, together with a business and regulatory pol-
icy concept. It must also be distinguished from the E.C.'s Open Network Pro-
vision, a concept that established access for value-added services, especially
across European frontiers, and which is much less far-reaching in its unbun-
dling.
A preliminary variant of ONA is also known as comparably efficient inter-
connection, or CEI.6 The FCC required the major carriers to submit ONA plans
in 1988 as a step in its evolution. The states (most notably New York and
California) also began to evolve ONA regulatory concepts (Noam, 1988a).
ONA is a framework of network components that disaggregate in such a way
as to permit open access; it operates on the concept that all central office func-
tions consist of fundamental components or Basic Service Elements (BSEs) and
that these components can and should be unbundled. Different communications
services use different BSEs, or different configurations of them, sequenced in
various ways. The open network architecture permits the use by outside parties
(users or third-party service providers) of the building blocks of their choice.
Where any of the blocks could be provided cheaper or better from another
supplier, it could be substituted and combined with BSEs or equipment of the
local exchange company. In other words, competition would exist for the var-
ious functions of the central exchange switch by unbundling its multiple func-
tions. To make such a system work, service providers conceivably could in
some instances collocate their own nodes on the physical premises of the local
exchange company. The third-party service providers are partly a form of value-
added networks, competing physical networks on the local level, and simple
resellers. In all of these functions they would compete head on with the local
exchange companies, who act both as retailers and as wholesalers of these
services (Noam, 1989).
Through ONA, local exchange carriers in effect permit the resale of separate
parts of their services, down to separate functions of the local exchange. This
is a radical reversal of past practice, where the monopolistic telephone compa-
nies tried to prevent resale. Now they have begun to recognize that their net-
work is their prime asset and that they should sell its capabilities as much as
possible, to the point of encouraging use by outsiders. In this fashion, the
network can be utilized more and a telephone company can still profit.
Once the building blocks are separated from each other, a pricing mechanism
need be put into place that can establish transparent and nondiscriminatory pric-
ing. For those building blocks where competition exists, deregulation can be
instituted. Elsewhere, tariffed rates would be set. Clearly, a wide array of com-
plex regulatory questions needs to be resolved in the process of introducing
ONA (Marks, 1986; Noam, 1988b, c). For example, the FCC announced the
possibility of extending ONA in the future into basic service access arrange-
ments (BSAs) of the local carriers. This step was taken by the New York State
Public Service Commission, which established ONA rights for basic transport
and created a type of access (collocation) in which alternative transport provid-
ers can physically locate their equipment in the central offices of the traditional
carriers, and vice versa (or do so in a functionally equivalent, i.e., "virtual"
form). Collocation also permits alternative carriers to provide network monitor-
ing and control of their own.
In technical terms, the ONA approach is not contradictory to ISDN, since
an ISDN operator could similarly provide for the subdivision of its functions,
selling them separately and permitting various reconfigurations and resale to
third parties. This is likely to happen in the United States in the future. But
attitudinally, the ISDN concept, as presently held by its champions abroad, is
widely different. Whereas the open network architecture is another step in the
segmentation of American networks, the PTTs' purpose for ISDN is another
step in its centralization.
ONA is not the only move toward segmentation in central office switching.
Already mentioned were shared tenant services, by which multiple users share
a common (nonpartitional) PBX. This concept has already begun to move into
the clustering of office buildings and provides a nucleus of alternative central
switches (Noam, 1986). Another manifestation is the emergence of interorgan-
izational networks (lONs) or enterprise networks.
Europe's ONP seeks to harmonize conditions of access, usage, interopera-
bility, and data protection among the networks of E.G. member states by elim-
inating the need for special adaptation of services between countries; it thus
provides a framework for a pan-European liberalized market for value-added
services. In 1988, the B.C. commission outlined its goals for ONP, including
the gradual establishment of technical standards, rules of interconnection, and
equal access to value-added services (Ungerer, 1989). Decisions concerning
technical standards were referred to the European Telecommunications Stan-
dards Institute (ETSI). Parameters were developed concerning service quality,
shared use, and resale of network capacity. Also, the commission intends to
implement a mutual recognition process of authorization for services, whereby
one member state can certify a supplier to serve the entire European Commu-
nity.
ONP has none of the unbundling elements of ONA and its establishment of
central office subfunctions, nor its related collocation and transport competi-
tion. (Avoiding the latter two is also the position of the American local ex-
change telephone companies.) Yet ONP, while a modest step, nevertheless cre-
ated a European rift.
Several PTTs (those of France, Italy, Belgium, Greece, and Portugal) were
reluctant to accept the liberal ONP guidelines. A compromise was reached. The
French proposed the licensing of data service providers, but the B.C. retained
approval of licensing procedures. Less-developed European countries were given
extensions to 1996 to implement ONP, and harmonization of technical stan-

dards was made voluntary.
In 1990, an B.C. directive was issued adopting technical interface and ser-
vice feature standards. Although compliance is voluntary, the directive gives
the European Community the option to make it mandatory in the interest of
interoperability. The directive's ONP framework provides the following: poli-
cies for leased lines and voice telephony to follow; standards for ISDN and
packet-switched services by 1991; recommendations for ISDN and packet-
switched services by 1992; and an examination by the European Community
and further recommendations after 1992 (European Commission, 1990, pp. 294-
95).
Technical Capability: Toward the Integrated Broadband Network

The two main narrowband ISDN channels operate on a speed of 64 kbps, but
user needs vary widely. Some services need only a very slow bit rate (e.g.,
telemetry, such as meter reading and alarm services). Inquiry response appli-
cations, such as airline flight reservations, electronic banking, or electronic
shopping, require rates of less than 4.8 kbps. A 64 kbps rate supports trans-
mission of high-quality facsimile as well as simple fixed graphic images and
slow video scan. Voice transmission in the United States is principally encoded
at 56 kbps, but 32 kbps seems to be emerging as the conventional rate. By
various methods of compression, voice transmission of sufficient quality has
been accomplished with as little as 9.6 kbps. Thus, one need not expect the 64
kbps requirement for digitalized voice to endure. It is a fairly high standard for
digital voice encoding, and it had been chosen because of the need for voice
quality in the context of an analog network with frequent demodulation.
On the other hand, the transfer of bulk data, such as large data files or
communications among mainframe computers, uses much higher rates. In the
United States, they are often in the Tl rate range of 1.544 Mbps (i.e., more
than twenty times faster than rates for the ISDN); and in Europe, they are at
the 2.048 Mbps rate.
Full-motion broadcast-quality video requires, depending on the protocols, an
analog signal of about 4.5 MHz, or about 100 Mbps as a digital signal. Using
signal-compression techniques, or a smaller number of lines, the signal can be
compressed down to 1.5 Mbps (i.e., the Tl rate).
Its name notwithstanding, ISDN, as defined by the CCITT, does not fully
integrate all communications services. The transmission of broadband ser-
vicesmost important, of video programscannot be accommodated, except
through bundling of several lines. The bandwidth of the ISDN network is too
narrow for it; four-wire transmission of one video signal can be accomplished
with some technical effort. In comparison, a standard coaxial cable television
link can easily carry seventy or more simultaneous video channels. At present,
digital switches are not yet well suited for large-scale switching of broadband
services. ISDN, as defined by the CCITT, thus results in a compromise rate:

faster than necessary for most small and medium-sized data users but too slow
for video.
The next technological step after integration and digitalization is integrated
broadband service (IBN). Such a broadband service could be more fully inte-
grated, carrying also video programming and mass media entertainment as well
as high-speed data streams.
The underlying technological development is optical fiber with its large
transmission capability. Experimentally, transmission rates of dozens of giga-
bits/sec over a single fiber strand have been achieved. At present, fiber is largely
used for high-capacity trunk transmission. Its cost (supporting equipment, in-
stallation, etc.) is still too high for regular subscriber loops, though some large
end users have already been reached (Egan, 1990; Elton, 1991). But it is likely
that fiber will be used in the future for small users too. At the same time,
development work is taking place for broadband electronic and optical switches.
The latter can process an optical flow without requiring its transformation into
electrons for the switching functions. It will take some time for these develop-
ments to be completed, but their eventual arrival is clear. Therefore, one may
ask whether it makes sense to move to the narrowband, electronic, copper-
based (for subscriber loops) ISDN if it is merely a transitional technology.
Should one postpone the next generation of communications technology to wait
for the one after it? This depends on the speed of innovation and the cost of
replacement, questions that call for analysis along technical and economical
lines. For some theoretical considerations, see Wilson (1984) and Rohlfs (1974).
It does not appear that fiber will rapidly replace already embedded copper
loops. And where separate coaxial cable television lines exist, the most de-
manded broadband services are already provided to subscribers. Thus, the time-
frame for the deployment of basic narrowband digitalization has a duration
longer than its stay in the technical labs, giving it a role in a transition to
broadband-ISDN.
Outlook
The past two chapters have asserted that the issue of integration stands for
much more fundamental questions of control over the telecommunications net-
work.
In a wider sense, it is part of a contest over where the intelligence in the
network resides (i.e., at the center or the periphery), who controls it (users or
network operators), who builds it (the telecommunications or the computer in-
dustry), and who runs the network (public providers or private ones).
Whether "segmentation" or "integration" are optimal solutions for net-
works is a matter that cannot be determined a priori. It is a trade-off of four
economic principles in two combinations: the efficiency of specialized produc-
tion and of a competitive environment versus the productivity contributions of
economies of scale and scope and the reduction of uncertainty. One cannot
generalize on which might work best.
35
Value-Added Networks and Services
Value-added networks receive greater attention in European policy discussions

than in the United States, and it is important to understand from the outset why
this is so. The key policy issue with respect to value-added networks is their
potential to resell basic transmission to third parties. Arbitrage by a service
reseller leads to loss of control by the basic network provider and to revenue
reduction, at least in the short run. In the United States, such resale is possible
and widely practiced; lessees of circuits can do almost anything they want.
Regulatory constraints exist only to prevent basic carriers from using internal
subsidies to extend their market power into the applications stages. In contrast,
European PTTs generally prohibit resale, though it exists unofficially in some
instances. But in time, European countries realized that the use of leased lines
can give rise to sophisticated applications and that such services should be
encouraged. They have consequently moved toward permitting the provision of
services that "add value" to basic transmission, and the sale of such VAN
services to third parties. This technical addition legally transforms what would
otherwise be an illegal resale into a legal sale. Another alternative to containing
resale is to establish usage-sensitive pricing for leased circuits. But this creates
problems of efficient pricing.
However, as with any kind of attempt at price discrimination that is not cost
based, one can never underestimate the ingenuity of arbitrageurs. Those who
still wish to resell basic transmission but can only sell "value-added" service
may try to add a trivial amount of value or an entirely unnecessary amount,
solely to cross the line into legality. To prevent this it becomes necessary for
PTTs or governments to require formal licensing of these value-added net-
works, which would involve scrutinizing the nature of their "value added" and
of their pricing. A formal approval process, or at least some form of ongoing
monitoring, is necessary to protect the system of price discrimination.
In the United States, as mentioned, such procedures do not exist. As a result,
VANs are merely a functional description and not a regulatory category. Being
not officially defined, they have different meanings for different people, and
often simply refer to packet-switching networks. It is true that in the United
States there is a regulatory distinction between "basic" and "enhanced" ser-
vices affecting VANs. But this distinction serves an entirely different purpose.
Whereas PTTs seek regulation of VANs to prevent the resale of leased capacity
(i.e., to protect the PTT service monopoly), the U.S. regulations serve to pre-
369
370 Value-Added Networks and Services
vent the cross-subsidization by a dominant carrier of its value-added services

through revenue gained in those dominant activities. In other words, the Amer-
ican "basic/enhanced" dichotomy is established to protect the newcomers from
the established carriers, and the PTTs' licensing of VANs aims at protecting
themselves.
A Conceptual Framework for the VANs

The VAN system is dynamic, but also organizationally complex and interre-
lated, to the despair of tidy minds. Terminology on VANs is vague, and some
attempt at clarification is in order. Conceptually, it is useful to think of VANs
as a system of multiple levels of sale and transformation. The first level con-
sists of the basic transport carriers such as the PTTs, AT&T, or the local Bell
companies. These carriers all provide lines to basic value-added networks (in
the United States, private firms such as Telenet, Tymnet, or AT&T Accunet;
in Europe, PTT entities and sometimes subsidiaries such as Transpac), which
constitute the second level. These essentially provide basic packet-switched
transmission service and some basic protocol conversions as common or private
carriers.
The third layer consists of "generic service providers." Such firms, also
called VANs, provide the additional software and organizational features that
make for added value beyond transmission. The third-level providers can be
identical to the second-level providers, and the operations can be functionally
integrated. Conceptually, however, the distinction is necessary. In the case of
AT&T and the BOCs, in fact, any third-level service must be provided by an
entity that keeps strict accounting separations from the rest of the firm (i.e.,
from its levels 1 and 2). AT&T offered level 3 service under the name of Net
1000, for which it leased time from its level 2 Accunet Packet Service, which
in turn used AT&T network dedicated lines (level 1).
However, things are never neat in this field. Level 2 VANs can also provide
capacity directly to end users, who can then provide their own level 3 enhance-
ments in private networks for their own operations. In effect, these users create
intrafirm "private" VANs. They can also make these private VANs available
to their customers, suppliers, or business partners, and thus create a private
"closed-user group" VAN. Level 2 VANs can also provide their transmission
capacity to other level 2 VANs, where legally possible, when these have not
established access of their own to a particular region. (This adds another layer
to the resale hierarchy.)
The level 3 VANs, in turn, sell their services. Buyers can be end users of
the generic applications, but they can also be providers of specific applications
who add value to the generic services as level 4 VANs. For example, a level
4 VAN can be a network of electronic store-and-retrieval service (a generic
offering) that is made to fit the specific needs of the insurance industry, to-
gether with other generics, such as a bulletin board, or a level 4 insurance
network. A combination of several generic services may be involved in such
Value-Added Networks and Services 371
an application. Another example is that of networks that connect automatic

teller machines. Providers of these network services often retail them to a num-
ber of banks, which thus need not establish their own systems. The banks, in
turn, give their customers access to their fourth-level VAN. Through open or
hidden service charges, they in effect resell electronic banking services to the
public. Thus, when a customer uses a bank teller machine to withdraw money,
the communications may easily involve four layers of communications services
and several firms, all contributing to the end service in a distinct way.
Because of the peculiarities of the American divestiture of AT&T, one can
often add to the U.S. scenario a long-distance carrier (which may in turn lease
its capacity from another), and two local exchange companies at each end that
have several sets of VAN relationships. The communications process could
thus involve almost a dozen firms.
The system is incestuous because competitors can be at the same time each
others' suppliers and customers of software, hardware, and transmission capac-
ity. In the United States, they can compete on level 1, collaborate on level 2,
and compete again on level 3. These relationships are unavoidably compli-
cated, as are the production and distribution of almost any sophisticated prod-
uct. Rarely are all stages of production of a complex product vertically inte-
grated within one company.
Level 2 VANS: Basic Packet Transmission
Packet transmission originated with a Pentagon effort, whose Defense Ad-

vanced Research Projects Agency (DARPA) had the firm of Bolt, Beranek and
Newman (BBN) develop in the early 1970s the "Arpanet" nationwide network
to link researchers. Arpanet was a major success and induced BBN to start the
commercial network Telenet, which has been in operation since 1975 as the
precursor to packet-switched (so-called X.25 protocol) networks around the world.
Basic packet-switching transmission has two main advantages. The first,
technical in nature, is error detection and correction, which is enormously use-
ful for data transmission. The second advantage, economic/regulatory in na-
ture, is the ability to slice transmission time into minute quantities, which can
be resold to provide a profit where there is a differential between retail and
wholesale prices. Data transmission rates, however, are not high enough for
high-traffic users of data. That, together with the tendency of packet networks
to charge by volume, means that leasing private lines may make more sense
for large users than using a "public" VAN. (However, packet switches are
getting fasterfrom 1500 packets/sec to 5000 packets/sec, and even to 60,000,
anticipated through parallel processing).
Telenet was not profitable and it was eventually sold to GTE. In the period
1978-1983 Telenet grew at rates of up to 40 percent annually. After this pe-
riod, growth slowed to a still high 30 percent. But it reached critical mass and
broke even only after 1983, with revenues of about $100 million. At that time
it connected about 2000 host computers, and in 1984, it averaged 200,000
sessions a day (Link Resources, 1984). In 1986, GTE Telenet was combined
into GTE's joint venture with United Telecommunications, soon dominated by
the latter, which in turn contributed its own Uninet (levels 2 and 3) and a
substantially fiber-optic physical network (level 1).
Tymnet was originally an internal operation of Tymshare, but its initial ad-
vantage of having a customer base of time-sharing computer users turned out
to be a problem later on, as time-sharing went into a steep decline with the
advent of inexpensive mini- and microcomputers. Users typically accessed the
packet-switched networks' nodes either through leased lines (digital or analog)
or through a regular public dial-up line. Tymnet and its parent Tymshare were
acquired by the aircraft manufacturer McDonnell Douglas, which resold it in
1989 to British Telecom.1
Other U.S. entrants were Graphic Scanning (Graphnet), PCI, CompuServe
Network, Autonet, MarkNet, Cylix, IBM Information Network, and MCI Data
Transport and EDS.2
In 1986, the FCC's Computer III ruling established the important policy
concept of open network architecture (ONA), discussed in the previous chapter
of this book. Under ONA, VANs can obtain special unbundled service ele-
ments from the telephone carriers, add network building blocks of their own
choice, and add their own blocks to them.
The PTTs and Value-Added Networks in Europe
All European countries have some forms of upper-level value-added services;

but in the past, such services were at the discretion of the PTTs, which would
set service conditions and technical specifications that were often influenced by
the desire to protect their own services and equipment offerings. For indepen-
dent VANs to become vigorous required strengthening their rights and indepen-
dence vis-a-vis the PTTs.
Liberalization took place first in Britain. The Telecommunications Act of
1981 provided for the secretary of state for industry to grant licenses for value-
added networks. To operate a VAN, a general license was required. Because
of the complexity of enforcing these separations, VAN licensing rules have
been constantly modified.
The Dutch PTT reform of 1987 opened that country's network to competition
in VAN services. Germany's 1988 legislation restructuring the Bundespost per-
mitted considerably liberalized VAN services, including pure resale of data (not
voice) services. In all instances, however, the PTTs secured their own right to
participate in VAN offerings, and did so vigorously. This led to the E.C.'s
policy on open network provision, discussed in Chapter 34.
The market for European VANs reached about $4.7 billion in 1991. The
U.K. market was the largest ($918 million), followed by those of France ($665
million), Germany ($428 million), and Italy ($230 million) (Whitehouse, 1990;
Boult, 1990; Blau, 1990).
The top VANs firms in Europe were Telenet ($500 million), IBM ($500
million), and GE Information Services ($450 million). Tymnet (BT after 1989)
and Istel (owned by AT&T after 1990) were the fourth and sixth largest pro-
viders, with revenues of $430 million and $170 million in 1989.
In some instances, technical problems slowed the spread of VANs. In 1988,
a survey by an end-user association found a failure rate of 25 percent in 5700
attempted data calls (Lill, 1988). Similar large-scale surveys found a 24 percent
failure rate in 1989 and 25 percent in 1990 (18.7 percent failing for telecom-
munications reasons) (Roussel, 1989; Schenker, 1990a).3
Another barrier to VAN services was the high cost of leased lines. A 1989
report noted that the average cost in Europe of 9.6 kbps leased-lines was, on
average, $1,500 for national circuits and $2,400 for international circuits. But
for national service, Spain charged eleven times that amount ($16,000) and
Germany's tariffs were six times costs. On international circuits, Italy charged
a 1700 percent markup, to $42,500, followed by Spain ($31,000) and Sweden
($30,000) (Woollacott, 1989a). Another study claimed that most European PTTs
priced their public data networks too low and lost money on them, while leased
lines were priced very high (Purton, 1989).
In 1987, after many large European users expressed dissatisfaction with the
CCITT X.25 public packet-switched network structure, the European PTTs,
through CEPT, promoted a joint Managed Data Network Services project. But
the European Commission, concerned about cartel behavior, limited MDNS to
offering three pan-European services and restricted it from advanced gateways
and network management. The project was abandoned after a year, and the
PTTs focused on their own projects, such as Eucom, Tymnet, Infonet, and
Syncordia.
Infonet was the fifth largest European VAN. In 1989, it was acquired by a
powerful cast of owners: France Telecom, DBF Telekom, Telefonica, Belga-
com, PTT Netherlands, and the Swiss PTT. MCI held the largest share, 25
percent. The European Commission investigated France Telecom, the Bundes-
post, and others for conflicts of interest in their investments in Infonet.
In 1988, the DBF and France Telecom also formed Eunetcorn, a joint ven-
ture for industry-specific VANs. In 1991, several major carriersBritish Te-
lecom, DBP Telekom, and NTTannounced their intention to set up a global
facilities management service named Syncordia, possibly with France Telecom.
Level 3 VANS: Generic Services
The following sections describe the more important upper-level VAN services.
Voice Mail
Voice messaging (also known as VSR, or voice storage and retrieval, not to be
confused with electronic mail) is a service that permits a computer to store
digitalized voice messages, like an answering machine. Different configurations
are possible: voice mail can be part of PBXs; it can be resold by service bur-
eaus; and it can be embedded in the central telephone switch. Typical level 4
users' applications of voice mail (level 3) are purchase-order-taking systems,
ticket reservations, scheduling of work crews, hospital paging, and hotel res-
ervations and guest messaging.
Voice Retrieval (Audiotex)

Voice retrieval (audiotex) is related to voice mail, but with the emphasis on
retrieval rather than on input and storage. A computer typically stores a large
variety of information in digital voice form that can then be recalled from afar
by calling in. One application is detailed weather forecasts, which pilots can
access and select according to region. Similarly, a theater reservation system
can have the ticket availability of each play stored separately, along with a
brief description of the play and the cast. Other applications are train schedules,
ordering of merchandise, airline ticket reservations, and mass announcements
such as dial-a-joke or music juke-box. These services can be lodged in cus-
tomer equipment or in the network itself. These applications will be consider-
ably boosted as synthetic voice technology develops.
Electronic Mail
In Europe, CEPT has set standards for "e-mail" service known as teletext (not
to be confused with broadcast videotex known in many countries as teletext
too). Particularly active in the development was the German Bundespost and
the firm of Siemens. In the United States, electronic mail, much more hetero-
geneous, was offered by a number of firms. ITT's ailing Dialcom was acquired
by British Telecom in 1986.
Data Retrieval and Search

The large number of personal computers and office desk-top terminals, many
with built-in modems, greatly increased the number of parties with the ability
and interest to access data banks. This led to the emergence of an electronic
publishing industry and on-line databases. There are specialized services as
well as more general "information utilities" that include a portfolio of data
services. In some ways these are similar to videotex, though usually without
the graphic capability, but with more data, faster response rates, and "smart"
computer terminals that can process the called-in information rather than TV
screens or dumb terminals. European PTTs have chosen the videotex approach,
which is discussed in a separate chapter. Videotex has not been a success in
the United States, partly because on-line data services have preempted much of
its role in the commercial field.
Telephone (Audio) Conferencing

In addition to the standard teleconference bridges, more advanced systems per-
mit dial-up (operator-free) conferencing, for both voice and data/text, and with-
out subscription. A "meet me" option permits conference participants to call

in to join the conference, without having to wait to be contacted.
Video Conferencing
Video conferencing is an active but commercially not particularly successful
area. Many large firms have video-conferencing facilities but do not use them
to capacity. Several U.S. hotel chains have nationwide interconnected video-
conferencing facilities.
Part of video conferencing's slow acceptance derives from the need for ex-
pensive dedicated studio-like conference rooms and wide-band transmission.
Both of these problems are being reduced through equipment capable of slow-
scan transmission and data compression and enhancement, which permit video
use in regular offices and require only one regular telephone line.
Telemetry
Alarm systems can be offered based on a new "derived channel transport" that
overlays the regular voice channel with a second narrow channel, creating an
independent transmission path for low-rate data. In addition to alarms, it can
be used for utility meter reading and for pay-per-view cable television control.
Alarm service can also be carried by cable television networks.
Computer Bulletin Board Systems

Computer bulletin board systems (BBSs) have proliferated in recent years with
the increase in personal computers. In the United States, there were an esti-
mated 25,000 in 1988. They are run by a "sysop" (systems operator), mostly
amateur enthusiasts, and often include a wide menu of subgroups and services,
including personal mailboxes. Specialized BBSs include professional confer-
ences and matchmaking services.
Call Forwarding
For some years, telephone service providers have offered enhanced services
such as call waiting, automatic call forwarding to other numbers, speed dialing,
and three-way calling.
Call Identification
New common channel signaling systems permit an identification of the incom-
ing call and make possible several features: call screening (blocking of unde-
sired callers), a selective call forwarding, identification of incoming call num-
bers, call-back of last number(s) that had called in but were not connected, and
special rings for preselected incoming calls, to permit, for example, separation
of incoming personal and business calls. These services are important in a
broader sense, since they give some measure of choice over the telecommuni-
cations process to the party being called, who in the past has had to guess at
the nature of the incoming call. They also permit verified billing arrangements.
They also raise important privacy issues.
Electronic Data Interchange (EDI)

Generic form transmission can replace traditional practices of purchase orders,
invoices, bills of lading, and so on, that require separately processed docu-
ments. Instead, the documentation of an entire transaction is electronic, inte-
grated, and nearly instantaneous. An important level 4 application for EDI is
for "just-in-time" production, such as for automobiles. Some automobile man-
ufacturers established an EDI system with their suppliers. Purchase orders are
entered entirely automatically, according to programmed instructions, and sent
to suppliers, who confirm, process, ship, bill, advise, get paid, and so on, all
within the same set of documentation. The system provides some of the advan-
tages of vertical integration and single sourcing without the cost. EDI systems
can be provided by the private network of a firm, an industry group, or a public
VAN. The EDI market is potentially quite large.
Level 4 VANS: Specific Applications
Telekurs and the Society for Worldwide Interbank Financial Telecommunica-

tions (SWIFT) are two examples of level 4 application VANs. Telekurs AG
was an early entrant in the VAN market. An independent Swiss firm, jointly
held by 350 Swiss banks and the Swiss stock exchanges, it began operations
in 1930 as Ticker AG by transmitting stock market information via telegraph
lines. Its business evolved to include financial information services, interbank
clearing, and management of Eurocard, a travel and entertainment card issued
by Swiss banks. Telekurs AG's most important financial information service,
Investdata, provides strong competition to Reuters, offering direct access to
eighty stock exchanges worldwide. Its interbank clearing service produces 30
percent of Telekurs' income. The Eurocard network serves over 850,000 card
holders, who can conduct their transactions at any of about 500 Bancomat
machines throughout Switzerland. In 1985, Telekurs employed 850 people and
had sales totaling SFr 150 million (von Weizsacker, 1987).
SWIFT is a particularly important participant in the financial VAN market.
A consortium of European and American banks created SWIFT in 1973, and it
went on line in 1977. All SWIFT participants are also shareholders. In 1977,
240 banks in fifteen countries were members, and SWIFT processed 27,000
transactions daily. In 1985, the ranks had grown to include 1275 banks in forty-
six countries, generating 192 million transactions. By 1989, the system was
handling up to 1.3 million transactions a day. SWIFT's main communications
activities are customer transfers, accounting for 30 percent of total volume;
bank transfers, another 30 percent; and statements and conformations, with 25
percent. Participating banks derive benefits from the service because of reduced
transactions and better control of costs. A SWIFT message costs BFr 18 ($0.48)
independent of distance.
To start operating, SWIFT had to establish standards for international trans-
actions; these have since been instituted worldwide. The nature of its business
requires strict security measures. SWIFT accepts liability for all losses due to
errors or cheating, underscoring its confidence in the system. It has never had
to fulfill this guarantee. SWIFT achieves this level of security with a hierarchi-
cal network structure. Individual banks connect to national nodes (regional pro-
cessors or RPs) via leased or dialed lines. The RPs encrypt, store, and pass the
messages on to the System Control Centers (SCCs) in Washington, D.C. or
Leyden, Holland.4 In 1989, European, African, and Middle Eastern banks gen-
erated 68 percent of SWIFT's transactions. About 13 percent of messages came
from the Pacific Rim nations, and North America accounted for 18 percent
(Hayes, 1989).
To obtain its leased lines, SWIFT had to negotiate special arrangements with
each country. It pays a set amount per message plus a volume-based tariff; the
volume-based pricing often causes SWIFT to pay more than four times the
ordinary leased-line tariffs. The PTTs thus extract a premium from SWIFT for
leased lines, but this was also an advantage, since SWIFT obtained special
licenses when there was no similar offering. It is therefore more difficult for
new entrants to emerge.
In time, SWIFT encountered competition from IBM, which set up a global
banking network of its own, and from Reuters. Furthermore, some of SWIFT's
bigger members set up their own networks to counter what they see as SWIFT's
subsidization of smaller banks. Thus, the exit mechanism discussed in Chapter
4 applies also to subnetworks.
The SWIFT network grew from 27,000 messages daily in 1977 to 1 million
in 1989, and rapidly neared its capacity.5 A more advanced network, SWIFT
II, was authorized, and SWIFT chose Northern Telecom for network equip-
ment. It will be a distributed network, unlike the more centralized SWIFT I,
with more value-added services, including messaging and EDI. Fast-packet
switching was introduced in 1991.
The following services are part of the variety of specific applications of level
4 VAN-type services that have emerged. Several other applications have al-
ready been mentioned.
1. Accelerated international trade payments. This service, its providers claim,

accelerates international payments often by two weeks.
2. International trade shipment data service. Helps trade shipment trans-
actions, documentation, billings, insurance, and so on.
3. Company dealer networks. Used for orders, product information, service
problems, billing, and customer information. Also referred to as Elec-
tronic Order Exchange (EOE).
4. Health care providers and insurance networks. Permits transactions be-
tween hospitals and insurers.
5. Credit card verification and processing. Merchants have terminals that

can read credit card magnetic stripes and transmit the information to a
central location for approval and processing. Some transactions are han-
dled by local banks or bank associations, such as VISA for its members.
Others are offered by public level 4 VANs. Credit authorization is being
integrated into electronic record keeping and transaction accounting.
6. Point-of-sale services. These retail services permit merchants to transfer
payments, send bills, verify credit, and reorder inventory. Because of
the high cost involved to set up such a system, several point-of-sale
switching networks exist. Some are affiliated with automated teller ma-
chine networks and often are owned by several banks or by more general
service providers.
1. Manufacturing design. Computer-aided design and manufacturing (CAD/
CAM) has led to private VAN applications. GM's blueprints are elec-
tronically accessible by its suppliers (which are thus forced into the elec-
tronic mode of design themselves). The high cost of a CAD/CAM ter-
minal (plus computer, software, and know-how) has been a problem for
small suppliers. In response, reselling has emerged in this segment.
8. Factory production. Automation produces and requires constant data flows.
One development priority for the near future is to permit equipment to
interconnect better with each other and with support services. Electronic
data interchange (EDI) systems offered by several VANs provide some
such integration and permit a "just-in-time" production process, with
interconnection with suppliers and programmed purchase orders. GM
adopted a manufacturing automation protocol (MAP). It acquired the
major data-processing firm EDS, which set up sophisticated networks
with dealers and suppliers.
9. Spare-parts service. Industrial database/transaction systems provide in-
formation on and transactions for millions of products, parts, and sup-
plies.
10. Service dispatch. A combination radio beeper/access terminal permits
input and communications with service personnel in the field.
11. Electronic banking and brokering. A number of financial institutions
enable customers to use their computers to reach their account informa-
tion, obtain investment data and place stock orders.
12. Electronic Fund Transfers (EFT). This is one of the earliest network
transaction uses. Various clearinghouse arrangements exist domestically
and internationally, and were discussed earlier with the examples of Te-
lekurs and SWIFT.
13. Automated teller machine (ATM) networks. The popularity and low cost
of the automated bank presence of ATMs spread enormously. In the
United States, a 1986 Supreme Court decision permitted ATM place-
ment out of state, and thus enabled banks to move across state lines,
which they could do before only to a limited extent.
Large banks can offer these services on private networks. Smaller and
medium-sized banks depend on VAN intermediaries. Several ATM

switching networks exist, often owned by a consortium of banks. Large
data processors also provide such services. Increasingly, the ATM of
one bank accepts transactions for others.
14. Commodity trading. Trading in commodities and precious metals, where
time is of the essence, has been enhanced by networks with brokers in
some instances simultaneously using regular voice call, viewing market
data, transacting trades, retrieving customer information, and entering
notes about the call.
15. Insurance industry networking. Because many independent insurance
agents deal with many underwriter firms, it was important to provide
them with a network for communications. Such a network had to fit
many firms' modus operand! and business forms and had to be compati-
ble with thousands of different equipment systems in agents' offices.
16. Medical communications. An information and transaction network exists
to connect doctors and drug companies as well as reference services and
advanced medical education databases.
17. Moving CAT-SCAN images. Medical CAT-scanning procedures are ex-
pensive and very data intensive. One VAN transmits image data from
smaller hospitals to larger data facilities for processing and storage.
18. Consumer information. VANs provide a variety of consumer informa-
tion and transaction services.
19. Job searches. There are data banks for employment, particularly for
data-processing professionals.
20. On-line databases. These are discussed in Chapter 37 of this book.
21. Teleshopping. Several varieties of teleshopping exist: on computer on-
line services, on cable television, and by automated phone-in orders.
One supermarket chain permits call-in orders with automated reception
from a 4000-item catalog, with the teleshopper picking up the order
soon thereafter.
22. Agricultural networks. Information and transaction systems exist for
commodity trading, weather, help, and advice.
23. Hotel in-room services. Services permit guests to access information and
electronic mail from their rooms and to receive information about city
or hotel activities, airlines, and so on.
24. Grocery networking. VANs can provide an electronic data interchange
(EDI) for a grocery industry group, which permits them to pool their
purchases and realize bulk discounts.
25. Up- and downloading with personal computers. PCs can be used as data
input and output terminals for a mainframe. Data can be exchanged,
software can be shared, and so on.
26. Yellow Page service. On-line service permits nationwide and interna-
tional compilations and searches of businesses.
27. Automobile collision estimation. This service permits garages and insur-
ance companies to estimate repair costs.
28. Animal breeder services. Permits matching of livestock.

29. Library shared cataloguing. Permits interlibrary searches, exchanges,
acquisitions, and automation in cataloging.
30. Credit history. Several commercial systems permit lenders to check on
the credit history of borrowers. This application, more than any other,
has been controversial and has led to laws protecting privacy and accu-
racy.
31. Telemarketing. Automatic dialing machines call potential customers ran-
domly or from prescreened lists. The sales message is taped; responses
are given either to a person who comes on the line or to a voice mail-
box.
Outlook
The list of VAN services and applications is not indicative of their commercial
or technical success. There is no reason to assume that today's mix of offerings
is more than temporary. It would not be surprising if half of today's offerings
and services would be gone in a few years and replaced by other services and
offerers. Given the rapid developments in hardware, software, and user orga-
nizations, the main attribute of a VAN system is not predictability of success
but flexibility of process. It is hard to see how restrictive rules on VANs could
be effective in the long run. If VAN services are successfuland it is impor-
tant to a competitive economy that they arethey will dance electronic circles
around the restrictions.
When it comes to policy liberalization, VANs ride on the coattails of equip-
ment. A liberalization of equipment has meaning primarily when the equipment
can be used in varied ways. Conversely, one cannot expect dynamic VAN
development if users are limited in their choices of equipment to a few slowly
approved models. VANs can be important to manufacturers because much of
the pure equipment can become a merchant market, with East Asian countries
serving as the low-cost producers. Hence, a link of equipment to services, of
hardware to software, and of hardware elements to each other by networks can
be critical to competitiveness.
This raises a final question. Are VANs another instance of "supply-side
telecommunications," with appeal to computer enthusiasts, equipment manu-
facturers, and telecommunications carriers, but not to regular users? There is
no question that many VAN services have been excessively hyped, and the
reality invariably fails in comparison. In the aggregate, the VAN service indus-
try is not yet particularly large in dollar terms. In the United States, level 2
VANs (packet switching) were in 1988 estimated at $850 million; for the higher-
level application levels 3 and 4 they were about $1.5 billion. In comparison
with basic voice telephony, these amounts are trivial. However, the market is
very innovative and is developing expertise in fashioning configurations of users,
equipment, and services. A market cannot move much faster than the users,
which had to absorb entirely new systems, work procedures, and organizational
patterns. These things take time. But users are steadily becoming more appli-
cations-conscious; several industries have already reached the stage of depen-
dence on them. Thus, the VAN services sector is riding a favorable trend.
Today VANs are "discretionary" services. Over time, however, some of
them may become first essential, and then basic services, as happened to voice
telephony.
36
Videotex
The New Technology
Videotex emerged in 1970, when the British Post Office started work on its
viewdata system Prestel, with the aim of integrating the telephones, computers,
and the television sets. France followed suit in 1972 with Antiope. In Ger-
many, the Bundespost bought and built upon British rights in the development
of its Btx system. In 1979, Canada developed Telidon, which, unlike the oth-
ers, is based on a geometric rather than an alpha-mosaic display.
All systems contended before the CCITT for recognition as the single inter-
national standard. The American position held that it was too early to imple-
ment standardization because of the rapid development of technologies. Euro-
peans suspected that this view was based on the Americans' absence from the
forefront of videotex development. In 1980, the CCITT recognized three sys-
temsAntiope, Prestel, and Telidonand agreement on a common protocol
seemed impossible. In 1981, AT&T established its Presentation Level Protocol
(PLP), closely related to the Canadian standard. In 1982, the American Na-
tional Standards Institute (ANSI) agreed upon a more mature version of PLP,
the North American Presentation Level Protocol Standard (NAPLPS), and sub-
mitted it to the CCITT. The European PTTs agreed on a common overarching
CEPT standard in an attempt to harmonize their systems and make them com-
patible. They also induced other European countries, including Italy and Spain,
to join them in those CEPT standards, and launched an export offensive.
In the meantime, a variety of nonvideotex computer information host sys-
tems emerged in the United States. These host systems are accessed over tele-
phone lines by a microcomputer as a terminal and a modem, usually via a
packet-switched network such as Telenet or Tymnet. Other database providers
are accessed directly, without an intermediary host. The on-line service provi-
sion has become known as electronic publishing, and traditional and new pub-
lishers have entered the field. Compared to videotex, these on-line services
require more intelligent terminals, but they can also perform more tasks. With
the rapid decline in the cost of microcomputers and monitors, there has been a
great increase in the number of households with microcomputers. For these
users, access to databases such as computer and mail-box service becomes rel-
atively simple. These data services do not package the information in multico-
382
Videotex 383
lored graphics, as videotex often does. In 1990, there were about twenty-seven
million microcomputers in U.S. businesses and homesa much larger number
than Minitel terminals in France, and with greater intelligence of equipment.
Videotex offers primarily five basic types of services: information (such as
news, weather, financial data, and regional movie listings), commercial trans-
actions (such as airline ticket orders), closed user-group interactions, electronic
mail and mailboxes, and information processing. For the latter, an intercon-
nected computer system (such as the one implemented by the German Bundes-
post) facilitates customer use. In effect, a PTT's videotex system can operate
as a computer utility. With interconnection, it can also interact with computers
of other service vendors.
By 1990, videotex had not become a success in any European country, with
the exception of France, where over 5 million Minitel terminals had been dis-
tributed. Germany had 200,000 subscribers; Great Britain, 95,000; Italy, 100,000;
and the Netherlands, 28,000. All other countries had less than 5000. There
were only 5000 subscribers in Japan, and 10,000 in Canada, mostly for farm
reports.
The differing philosophies on telecommunications policy are reflected in four
countries' approach to videotex. Britain's service developed only slowly, caught
between the government's prevalent indecision between deregulatory tele-
communications policy and pro-active technology policy. In Germany, the Bun-
despost set up a huge centralized system and promoted it heavily, but demand
was low. In France, in contrast, the government was massively engaged in
direct subsidization, buying millions of terminals from domestic manufacturers
and literally giving them away. It financially supported the usage, information
production, and equipment, in the hope that in time the subsidized egg would
hatch a chicken. Usage has been brisk, but the costs of keeping the system
going have been high. In the United States, videotex was left to the market,
and the local Bell telephone companies were restricted to protect competition.
There is little videotex in America, though much usage of its on-line cousins.
Videotex in the United Kingdom
Using the Prestel protocol, interactive videotex is offered by British Telecom

(BT) under the name of Viewtel. Prestel was the world's first public interactive
videotex system, developed in the early 1970s at the Post Office Research Cen-
ter by researcher Sam Fedeida from work involving a "viewphone." The sys-
tem was demonstrated in 1975, and a pilot project was initiated the following
year with the participation of the main manufacturersGEC, ITT, Philips, Pye,
Rank, and Thorn-EMI. About 400 organizations provided information, includ-
ing the Financial Times, the London Stock Exchange, and several government
enterprises. Testing with a sample of domestic and business users began in
1976, and regular service was initiated in 1979 (Tydeman, 1982). In 1990,
about 95,000 terminals were attached to the Prestel network and more than
384 Videotex
2000 main providers offered information. As a premium service at an extra

charge, information pages could be stored for closed user groups.
The success of the service was important to the British government as a
highly visible indicator of British scientific and engineering knowhow. Refer-
ring to videotex and broadcast teletext, the former minister for information
technology, Kenneth Baker, exclaimed:
First, we see them as two technological developments in that Britain now leads the
world and which must be promoted, because of the commercial and industrial ben-
efits which can accrue to the nation. Second, we see teletext and viewdata as swing
doors through which the country can gain an entry into what some people are
calling the "second microelectronics revolution." 1 mean, of course, information
technology. Third, it provides employment for set manufacturers, component man-
ufacturers, and software companies. The government is convinced that information
technology is going to be the key growth sector in this country's economy [British
Business, 1981, p. 1480].
Prestel service was enthusiastically welcomed not only by the government

and BT, but also by electronic firms and television set manufacturers, who saw
a potentially large consumer market for new types of equipment. Furthermore,
mail-order houses, travel agencies, and banks were interested in using Prestel
as a new marketing tool. Because of the belief that videotex could fulfill many
of the functions currently handled by newspapers, the service was viewed war-
ily by newspaper publishers, who feared the potential loss of advertising reve-
nue and newsstand sales. Other publishers, however, recognized that their long-
range role was not merely to publish news on paper but to collect, edit, and
distribute information in various forms; they moved to become videotex infor-
mation providers. Although Prestel was originally conceived as a residential
customer service, businesses were the most eager customers during the first
few years that the service was offered. Of the 7605 subscriptions in January
1981, 87 percent were for businesses, and only 13 percent were from residen-
tial locations (British Business, 1981). Both figures fell far below BT's original
forecast of 50,000. By 1988, the proportion of residential users had increased
to 39 percent of Prestel's 80,000 subscribers (British Telecom, 1988, commu-
nication).
In 1986, Prestel began to provide more specialized services, such as home
banking, teleshopping, and educational offerings. British Telecom entered the
information provider market itself by purchasing a majority in Farmlink, a spe-
cialized service aimed at the rural population. BT also acquired a substantial
share of Telemap, the operator of Micronet 800 service, which provides com-
puter news, advice, and games for computer users.
The Prestel protocol was adopted by other countries, including Italy, Bel-
gium, the Netherlands, Switzerland, Hong Kong, and Australia. Although Prestel
is an impressive display of advanced British engineering, development of other
systems did not stand still. Partly as a consequence, BT developed Photovideo-
tex, a substantially upgraded standard that can transmit picture-quality repro-
ductions for use in picture databases.
Videotex 385
Videotex in Germany
Initially, the German telecommunications equipment industry was not particu-

larly interested in videotex. The Bundespost bought British rights and expertise
and built upon Prestel in the development of its Bildschirmtext, known as Btx.
Following recommendations by the blue-ribbon commission KtK and an affir-
mative government decision, Btx pilot projects were initiated in 1980 in Dus-
seldorf and Berlin. Interconnected financial institutions and mail-order firms
were soon active. One travel organizer made forty to fifty bookings per week.
The government institutions decided to advance the development of the new
technology, and in 1981, the Bundespost awarded the major contract to IBM
Germany.
The advantage of Btx is its uniform technical standards, which allow a wide
range of applications in an integrated and commonly used system. Thus, even
rarely used services share the benefits of a mass service. The concept is based
on uniformity and centralization and is concerned with economies of scale. The
entire national Btx system was centralized in one giant IBM computer located
in Ulm and known as the "Mother." This central location, which allegedly
can be infinitely expanded, stores participant data, all personally addressed in-
formation, and all supplied information for the entire Federal Republic. Users
are linked through regular telephone lines with regional Btx offices, which are
in turn linked to the central computer by a packet-switched network. The re-
gional offices consist of two data bank computers and up to six communications
computers for storage of regionally relevant information. Btx's design innova-
tion over Prestel is its use of private computers connected to the central PTT
computer.
Most of the Btx's information is provided by travel agencies, insurance com-
panies, banks, publishers, and mail order houses. However, most retailers have
discounted Btx's efficacy for transactional purchases unless the user also has a
catalogue. Home banking requires a special alphanumeric keyboard. One bank
reported that the turnover on their Btx accounts was higher than that for their
general accounts (Tonnemacher, 1982).
One problem has been the high cost of decoders. Although new television
sets of high quality have built-in decoders, most other television sets must be
supported by a separate box. In a 1985 analysis by the Bundespost of its vid-
eotex strategy, it was found that the terminal equipment supply by the equip-
ment sector was problematic in that the necessary diversity at acceptable prices
did not exist. The system was also not fulfilling the desired expectations of
information seekers. One early survey indicated that only 27 percent of infor-
mation providers had expected Btx to move as slowly as it had (Neue Medien,
1985). Only 18 percent of service information providers considered their Btx
investment to have paid off. When information providers were asked for their
opinions about the causes of the slow demand for Btx, virtually everyone crit-
icized the Bundespost: 46 percent for the flawed advertising concept, 19 per-
cent for the incorrect forecasts that had led to false expectations, 12 percent for
386 Videotex
overly complicated or expensive standards, 10 percent for high DBF telephone

rates, and 2 percent for the system of search trees. The Bundespost was recep-
tive to the criticism. It took the long view, expecting slow going until take-off.
In 1990, the DBF had 200,000 subscribers and 3,000 providers.
Videotex in France
During the 1970s, the French telecommunications administration DGT (later

France Telecom) and its research labs CNET investigated the use of television
sets for novel telecommunications data and text services. In 1977, the Nora-
Mine report galvanized the public with concern about the "informatization" of
society. Prodded by the British videotex advances, the DGT adopted a new
initiative in the area of videotex in 1977. In 1978, it gave videotex the green
light, with the economic rationale that human directory assistance service and
paper telephone directories would be replaced by the cheaper electronic service.
This concept neatly side-stepped opposition to the new and competitive me-
dium by the newspaper publishers. In 1979, field trials took place in Brittany,
and the DGT then decided to distribute, for free, 30 million terminals over
fifteen years (Vedel, 1987).
The DGT's technocratic moves aroused opposition by the press, civil liber-
tarians, and some economists, and led to the establishment of a permanent
monitoring commission in 1981 and to stronger attempts by the DGT to place
its technological goals within a broader social policy rhetoric. Following the
change in government in 1981, these principles were incorporated into a large
trial in Velizy, near Paris. The DGT began to subsidize press applications in
order to encourage development of their Teletel videotex service and set rates
in an advantageous "Kiosk" system. By 1983, mass distribution of Minitel
terminals was begun, and by the end of 1984, half a million terminals had been
given away. Usage was encouraged by a generous charging system, no sub-
scription cost, free terminals, and incentives to service providers.
The electronic directory service was officially launched in the Ile-et-Vilaine
department around Rennes in February 1983. After some experience, the policy
on electronic directories shifted from forced to voluntary usage. This shift un-
dermined some of the economic logic of substituting electronic for paper direc-
tories.
Initially, each household used the electronic directories only about twice a
week. After a while, surveys showed that residents used Teletel about one-half
hour per week, calling four or five services in a single fifteen-minute access
session. The surveys also showed that teenagers between twelve and fourteen
years of age were the most frequent and most proficient users of the service,
and people over sixty years old largely ignored it (Sichel, 1983).
In 1991, there were nearly 12,000 private information providers and three
levels of private information service: Teletel 1, which is mainly for commercial
usage, costs $0.08 per 20 minutes. The user must subscribe to each service
individually and pay local telephone rates, while the provider pays the rest of
Videotex 387
the telecommunication fees. Teletel 2, which is for both private and commer-
cial usage and for which the user pays all communication costs, costs $0.08
per two minutes. And Teletel 3, or "Kiosk," which is also for private and
commercial usage, requires the user to pay the communication costs of $0.10
per minute plus a time-sensitive service fee. On the Kiosk system, there are
five levels of service fees, with the private providers and the telecommunica-
tions authority receiving varying proportions.
Penetration has been large because the terminals have been distributed at no
charge. Government support was also given to information providers to en-
courage them to use the medium.
France Telecom's own 1987 figures for total hours of usage divided by the
number of households results in an average monthly usage of sixty-two minutes
per household (Telematique News, 1987). Disregarding the revenue loss be-
cause of lower long-distance calling, and making the generous assumptions that
two-thirds of the network usage is newly generated traffic rather than replace-
ment of calls to airlines, for example, and that two-thirds of new revenue is
pure France Telecom profit, the added revenue should come to an average of
FFr 90 million (about $15 million) per month. At that rate, a terminal, which
costs about FFr 2000, is paid for in about five years.
For the total actual and planned Minitel investment of FFr 17 billion from
1984 to 1995, 54 percent goes toward the development of Minitel, 18 percent
to the packet-switched network Transpac, 15 percent for additional lines, and
13 percent for the DOT network. Annual investment costs are FFr 1.5 billion
($240 million). At the present utilization and revenue rate given earlier, annual
system investments per subscriber are FFr 750 ($150), while revenues are only
FFr 550 ($90) (T. Vedel, 1988, communication).
The financial performance of the project is not readily apparent, sin'ce several
public subsidies are involved. In 1984, Teletel expenses were FFr 2 billion
($320 million) and brought in FFr 407 million ($68 million). According to
French critic Jacques Darmon, there will be a permanent deficit (Darmon, 1985,
p. 108). France Telecom disputes this, claiming that Teletel increases usage of
the network.
On the revenue side of France Telecom performance, there are a number of
empirical and conceptual problems to overcome. France Telecom discloses rev-
enue and minutes of usage on Teletel, but it does not provide net figures. A
Teletel call for travel information provides revenues, but at the same time it may
replace a regular telephone call. Indeed, the distance-insensitive (local) Teletel
call may replace toll calls, thus causing a net revenue loss. Similarly, an elec-
tronic directory assistance call replaces the revenues of a similar call to a human
operator.
In addition, not all revenues of new types of calls, such as those of the chat-
lines, should be allocated to Teletel per se. Such conversations could have oc-
curred on privately provided bulletin boards or computer conference systems. In
the Unites States, there are about 25,000 such systems, many of which are ar-
ranged by teenagers and computer hackers. The resources of the centralized state
are not required to accomplish communications among the people. Thus, even
388 Videotex
without videotex, new and revenue-producing calls would be generated by com-

municating terminals.
Extensive marketing campaigns were used to increase Teletel usage. As a
result, the number of calls to information services rose from 104 million in
1985 to 513 million in 1987, and the hours of connection time rose from 11.7
to 52.4 million over the same period (Booker, 1989). Growth in the average
number of yearly connection hours per terminal leveled off in 1987, dropping
from twenty hours to eighteen in 1988 and seventeen in 1989 (France Telecom
News, 1990).
Despite the marketing push, the state accounting agency for state-owned
companies, the Cour des Comptes, reported that the Teletel system lost $742
million in 1988, and projected 1995 losses at $574 million. To bolster reve-
nues, a monthly charge was suggested. However, the analysis did not include
Transpac revenues.
France Telecom denounced the results of the Cour des Comptes report. It
commissioned an analysis by Coopers & Lybrand showing a breakeven starting
in 1991, and a rate-of-return of 11.3 percent. The 5 millionth Minitel was
delivered in 1990. But this growth was primarily in business use (plus 26 per-
cent), whereas there was an 8 percent decrease in residential use (France Te-
lecom News, 1990, p. 3; Roussel, 1989b, p. 52). The average household made
twelve calls and used less than one hour of service per month (The Economist,
1989, p. 55).
Among the Teletel services, the messagerie service has proven popular among
users and lucrative for the telecommunications administration. The messaging
service was allegedly invented by a newspaper in Strasbourg that encouraged
users to talk to each other. The service provides an on-line dialogue capability
between users, similar to the electronic mailboxes and bulletin boards of com-
puter hackers. It also offers various bulletin board services, such as classified
and personal ads and electronic messages. Cost is about FFr 260 per hour
(about $10 per hour) of connection time. Users often adopt pseudonyms or
"handles." In 1989, a best-selling book by Denis Perier chronicled the story
of a Paris call girl whose brutal murder was arranged over the Minitel. The
case led to a round of protests from French family and church organizations
about the very character of messaging services (Kramer, 1990).
A majority of the users of the service are young men, and many of the
dialogues are flirtatious in nature. When the service was first offered, peak
usage occurred at the end of the lunchtime break, indicating usage by employ-
ees at the workplace, and employers began to block Teletel usage. Because of
the newspaper involvement and the human interest angle, the Kiosk offerings
received lavish media attention in France and abroad.
Partly because of their classified ads, the Kiosk offerings could be offered
until late 1985 only by newspapers. The major presence was Le Parisien Lib-
ere, with about 1 million monthly connections in 1985. The newspaper Liber-
ation had 300,000 calls per month.
Teletel reaped its highest yield from "messaging," which accounts for 8
percent of calls and 22 percent of revenues. The electronic directory, on the
Videotex 389
other hand, was a loss-leader, providing 18 percent of revenue from 33 percent

of the calls (Thomas, 1988).
As applications and familiarities grow, per capita usage of Teletel could well
increase. On the other hand, as the novelty of the chatlines, the main revenue
source, wears down, their per capita usage may decline. Furthermore, usage
among the present users may not indicate future usage patterns in the general
populace. Although Minitels were distributed around the country, they were
given first to the heavy telephone users within each district. Light users may
be less comfortable with Teletel or have less of a need for its services.
Minitel terminals under the Antiope system are provided by a number of
suppliers, including Alcatel, Matra, and TRT. In addition to the simple termi-
nals, French manufacturers have developed several more expensive models with
added features. The Minitel 12, for example, is geared toward professional
users. For other foreign markets, particularly in Arab countries, these firms
developed videotex terminals with more than one character set, but have re-
ceived relatively few export orders. An upgraded M10B terminal has a separate
screen and memory telephone; an intelligent Minitel costs between FFr 10,000
and FFr 15,000. CNET developed a Minitel model with the capability of run-
ning various peripheral functions as a microcomputer. However, according to
the associate director of industrial and international affairs at the DGT, there
are several obstacles to upgrading the Minitel:
Changing a screw on a Minitel takes a year! Then there is the price. Our ambition
is to see the price of the Minitel reduced. To give it the advantage of added func-
tions would condemn us to pay more, and just imagine the FIT having to manage
a market of several million microcomputers which would very quickly become
obsolete. That would be untenable. And who would rapidly develop software for
these micros? Not the PTT! We are not a service company [Clavaud, 1984, p. 37].
At France Telecom, there are two contrasting strategies for the future devel-
opment of the Minitel. One strategy advocates the rapid development of an
intelligent second-generation Minitel in order to forestall the flood of cheap
microcomputers. The other stresses the expansion of the market for the first-
generation Minitel by freely distributing the terminals and allowing the private
market to supply more advanced attachments.
France has exported Teletel development worldwide. In the United States,
the Minitel Services Company has made French services available over Bell-
South, US West, and Southwestern Bell gateways. In 1987, Intelmatique estab-
lished Minitelnet, an international interconnection service with Transpac.
Agreements with Germany, Belgium, and Luxembourg to permit access to its
services were reached in 1988, and in 1989 Intelmatique made services avail-
able to Japan's Mitsui Knowledge Industry Co., complete with a Japanese lan-
guage interface. Switzerland and Spain also allowed interconnection with the
Minitel network, as did Italy in 1990. Also, in 1989 Mercury considered adopt-
ing the French standard for its service instead of the Prestel system. However,
despite the number of countries reached, only 30,000 hours of connection time
out of 85 million hours of total traffic came from outside France in 1989, or
390 Videotex
0.05 percent of total traffic (France Telecom News, 1990; Roussel and Rock-
well, 1989).
For all its birth-pains that were described above, French engineering in the
videotex field is indisputably outstanding. As an example of forward looking
government strategy, Teletel may well provide a triumph in the future. This
high visibility project can succeed in technical, social, and economic terms,
and prove the viability of videotexand more generally of PTT supportin
establishing a critical mass for a new information service.
Videotex in Other Countries
Many other European nations have experimented with videotex and similar
electronic information services. None have matched the penetration of the French
system.
In Finland, databases and data processing industries have developed at a fast
pace, and the country's dynamic publishing industry has been a leader in de-
veloping electronic information services. The remote access data processing
industry is fairly open to outside suppliers. Finland was the second country to
offer videotex, known as Teleset; it was started experimentally in 1978 by the
Helsinki Telephone Company, together with Nokia Electronics (equipment) and
the Sanoma Corporation (information), using a Prestel-like system. Teleset
standards conform to European standards. American DEC computers are used;
decoders are supplied by the two Finnish TV set manufacturers. The informa-
tion provided is primarily business and consumer information, news, statistical
data, and local municipal announcements. No P&T license is necessary for
videotex, but initially the P&T unsuccessfully argued that videotex, as a text
transmission, was part of its monopoly on "telegraphy." The P&T also oper-
ates its own videotex operation, Telset, which had some paying subscribers in
its Helsinki system (Howkins, 1982). Other local telephone companies in Turku
and Tampere also followed the model, entering cooperative ventures with
newspaper publishers.
In Italy, videotex services are provided by a number of STET subsidiaries
SARIN, SIP, and SEAT. In 1980, these services began experimental operation
under the name Videotel, and the telephone company, SIP, has operated vid-
eotex on a regular basis since 1986. Users can access the host computer, lo-
cated in Milan, from throughout the country with a local call. Videotel is base
on Italtel and CEPT standards, and concentrates on business subscribers. SARIN
handles the engineering and market research aspects of telematics. It develops
the mainframe and system software to support Italy's Mediatel terminals and
provides the X.25 and Italpac packet-switched services. On the marketing side,
SARIN offers telemarketing, advertising services and user profiles, all geared
toward encouraging on-line usage.1
In the United States there have been several unsuccessful pilot videotex proj-
ects in the period from 1981 to 1985, including such companies as the New
York Times and AT&T, in Coral Gables, Florida; Time, Inc., and AT&T, in
Videotex 391
Ridgefield, New Jersey; Times Mirror, in Los Angeles; and Knight-Ridder, in

Lake Forest, 111. Though these experiments were targeted to wealthy suburbs,
only few subscribers were willing to pay connection fees of up to $20 a month
along with one-time charges for terminals of $400 or more. A second wave of
videotex ventures in the late 1980s involved Prodigy, a joint venture of Sears
and IBM, which has a user-friendly system and aims at the personal computer
user. Its billing is simple too, with a usage insensitive flat monthly charge of
$12.95 for most services. Prodigy offered transactional and travel services,
news, sports, and an on-line encyclopedia. Its membership in 1991 was about
1 million, although many were inactive. Sears' and IBM's investment was es-
timated to be between one half and one billion dollars, but it was losing money
and was reliant on continued financial support.
Another joint venture of Citicorp, NYNEX, and RCA, CNR Partners, folded
in 1988 after two years. Likewise, the Covidea project, involving Chemical
Bank and CBS, folded in 1987. The government and telephone companies have
so far played only minimal roles, although Judge Greene's 1987 relaxation of
restrictions has encouraged the Bell companies to enter this field. They estab-
lished the concept of "gateways" to provide easy connections to any database
or enhanced service provider by offering user-friendly menu services. These
data and information providers can usually also be accessed directly rather than
through the gateway, although in that case other billing arrangements are nec-
essary. Information service gateways were begun by several regional Bell com-
panies in 1988 and 1989, but most failed soon after their introduction. In 1991,
the Bell companies succeeded in court and won the right to offer information
services. This made it likely that they would enter videotex markets more ac-
tively than by providing gateways.
The absence of success of U.S. videotex stands in contrast to more decen-
tralized services of on-line databases and information hosts. By 1990, the lead-
ing on-line service was CompuServe, with 620,000 subscribers, followed by
Dow Jones News Service, with 400,000; and GENIE, with 150,000. An esti-
mated 25,000 bulletin board services were operating in the United States, and
6.5 million out of 27 million microcomputers were equipped with modems.
37
Transborder Data Flows
Impediments to the free flow of information predate the advent of electronic

communications by centuries. From the earliest years of book printing, some
published materials were prohibited from importation to other countries or ju-
risdictions for reasons of politics, religion, or economics. Written correspon-
dence was often censored, although this practice had some limits because of its
expense. The advent of wire-line communications gave individuals new means
to communicate across borders, but it also provided states with a smaller num-
ber of entry channels that needed to be controlled. In the case of the telegraph,
state employees handled each word of the messages, and often kept copies.
Broadcasting proved harder to contain, since over-the-air messages could cross
political boundaries with impunity. Some states deterred individuals from lis-
tening to unwelcome news from outside sources by making the activity a crime
and by interfering with foreign broadcasting frequencies through electronic
jamming. On the whole, however, a strong tradition of unimpeded information
flow during peacetime emerged in Western democracies, including the freedom
to speak, compose, and distribute messages nationally and internationally. Even
as European governments continued to control the transmission channels for
electronic communications between individuals, they did not control or tamper
with the content of the communicationsat least in the democratic countries
and limited themselves to indirect restrictions governing tariffs, technical stan-
dards, and attached equipment.
Given this strong tradition, the advent of new technologies that generate,
process, and transmit information might be expected to reduce or even elimi-
nate restrictions. Yet the proliferation of electronic transmission in Europe cre-
ated new types of issues that raised the free-flow question again.
Data Protection and Transborder Flows
Computers and information storage led to fears about the use and abuse of
electronic data.' The tremendous ability of computers to store vast amounts of
information, to centralize individual data from a large number of sources, and
to recall and disseminate it rapidly makes this new technology an effective tool
for government surveillance and business power. For example, government in-
392
Transborder Data Flows 393
vestigations can rapidly correlate an individual's medical history, financial

transactions, consumption patterns, travel, reading habits, and so on. Many
fears expressed in the 1970s were based on the notion of computers as vast
centralized mainframes, which roughly corresponded to the nature of technol-
ogy of the 1960s. Since then, technology has moved toward decentralized sys-
tems, with millions of interconnected computers. This has reduced the need for
centralized storage and has shifted the problem of protecting individual privacy
from one of data storage to one of data flow.
Protective legislation on electronic data collection and storage began in the
early 1970s with a law in the German state of Hesse (1970) and a national law
in Sweden (1973). From there, the movement spread to most of Western Eu-
rope. Most European countries used similar approaches to data privacy. They
include a uniform national law (in contrast to the piecemeal American ap-
proach), an independent specialized agency for enforcement and regulation,
registration or licensing of data files that include personal information, access
and correction rights for individuals, use and disclosure limitations, and a struc-
ture for the internal management of databases (Pipe, 1984b; Turn, 1979).
Though the origin of concern over privacy was the potential abuse of data
by government agencies, the focus of remedial action shifted to control of the
data collection activities by private organizations. Rules against the govern-
ment's collection of data were also set, but with less severity. At the same time
that Germany promulgated the first data protection laws against private data
abuse, its federal and state governments took a quantum leap in the use of data-
processing technology for the surveillance of its citizenry. During the 1970s,
political violence by the so-called Baader-Meinhof gang prompted the German
police to institute a chillingly efficient system of border checks, citizen regis-
tration, and domestic road blocks, all of which were interconnected by data
banks and communications links. Although the terrorism was mostly stopped,
many control mechanisms were not.
In the meantime, it was recognized that privacy laws had a loophole: inter-
national data transfers permitted the evasion of data protection laws. Under the
Swedish Data Protection Law of 1973, for example, a data file on any Swedish
employee is subject to certain protections from disclosure to third persons.
However, if a Swede works for a foreign firm, it is possible that the data would
be routinely transmitted to the headquarters of the firm, where it would be less
protected. Conceivably, given this loophole, some countries could set them-
selves up as "data havens" in order to attract businesses determined to circum-
vent privacy laws. Although these threats were more theoretical than real, they
led to a movement to harmonize data protection practices or to restrict the flow
of sensitive data in the absence of such harmonization.
The Organization of Economic Cooperation and Development (OECD), through
the initiative of Hans-Peter Gassmann, was instrumental in formulating the
principles. It created a Committee on Information, Computer and Information
Policy (ICCP). In 1979, the OECD drafted a first set of guidelines for its mem-
ber states: Data collection should be limited to necessary information obtained
lawfully, and, where appropriate, with consent; data should be accurate, com-
394 Transborder Data Flows
plete, up-to-date, and relevant to the needs of the collector; use of the data
ought to be specified at the time of collection, and its disclosure should be in
conformity with the purpose of collection; assurances must be made against
unauthorized access, use, and disclosure; and data should be open to inspection
and correction by the individual to whom it refers (OECD, 1979).
The Council of Europe had considered privacy issues since 1969. In 1980,
it incorporated the OECD guidelines in a convention on the Protection of In-
dividuals with Regard to Automatic Processing of Personal Data. The proposal
achieved binding status after ratification in 1985 by a majority of Council of
Europe members. The convention affected all transborder data flow among Eu-
ropean countries and with other countries, such as the United States. Multina-
tional firms became nervous about all this, since the convention provided that
any country could restrict the transmission of data to another country that did
not have data protection legislation comparable to its own or that had not rati-
fied the convention. This provided a means of affecting transnational opera-
tions. Theoretically, any country could restrict the flow of data to other coun-
tries by the simple device of raising the level of its own privacy protection.
Since firms conducting international transactions generally prefer to have uni-
form procedures for transactions in various countries, procedures were likely
to conform to the strictest of the national rules.
In 1985, the OECD's ICCP adopted a Data Declaration on transborder data
flows as a more specific framework for data flow policy, or "rules of the road
on TDF" (Robinson, 1990). TDF issues were also examined by the U.N.'s
Intergovernmental Bureau for Informatics (IBI) a Rome-based Third World-
oriented communications body that eventually folded for financial reasons.
UNESCO also showed involvement (Roche, 1984). The United Nations' Cen-
ter on Transnational Corporations similarly initiated a TDF project to increase
developed and developing countries' sensitivity on the issue (Sauvant, 1986).
Even the Vatican took a position: "There is important work to be done to
harmonize [national regulations] at the international level and finalize them by
using suggestions coming from different social contexts and by employing con-
tinuing technological developments" (in Pipe, 1984a).
Europe-wide efforts at ensuring privacy continued with a 1989 Council of
Europe recommendation on protection of personal data used for employment
purposes (TDR, 1989). In 1991, the E.C. passed a directive covering also man-
ual files, and forcing legislation by member states. Yet, at the same time,
European police data collaboration was substantially upgraded (Masden, 1992).
Data Protection in a National Context
France
In 1974, a Commission on Information Policy and Civil Liberties was estab-
lished within the Ministry of Justice to regulate databases and protect individual
privacy. Subsequently, in 1978, a law on Information Processing, Files, and

Civil Liberties was passed that became effective two years later and affected
domestic databases and cross-border data flows. The law attempts to regulate
biased and unfair methods of data collection and prohibits the recording of
information on race, politics, religion, union membership, and so on. The law
also acknowledges the individual's right of access to the information, imposes
the obligation to correct false or incomplete information, and sets rules for its
distribution.
All data systems must be registered with a special agency, the National Com-
mission on Informatics and Liberties (Commission National d'Informatique et
Liberte, or CNIL). The CNIL is an independent administrative authority, sim-
ilar in some ways to the independent American regulatory commissions, and as
such is a rare body in French administration. A majority of its seventeen mem-
bers are chosen by the judicial and parliamentary bodies. The commission can
issue advisory opinions about data protection and data processing, which are
then subject to judicial review by a higher administrative court, the Conseil
d'Etat.
A database must be registered with CNIL, so that the commission will be
able to analyze it. A declaration must state the purpose of the database, the
nature of the personal data collected, and its sources as well as destinations. In
1985 and 1986, the commission received almost 10,000 filings of data banks
(TDR, 1986). In 1987, the number rose to some 45,000. Simplified procedures
are available for routine personal data such as payroll, banking accounts, utility
billings, and mailing lists.
The first case in which the CNIL recommended criminal prosecution in-
volved the Swedish ball bearing company SKF and its French subsidiary. Dur-
ing a labor dispute, the company was occupied by trade unionists, who discov-
ered a notebook with 600 entries on employment applicants for the years 1971-
1982. Although the data collections had been discontinued and were not elec-
tronic, the CNIL recommended prosecution. Another case that nearly led to
criminal prosecution involved an infraction by the Communist trade union CGT,
which had used the automated payroll list of the electric and gas company for
mailing election materials on the eve of workers' council elections (Kuiten-
brouwer, 1985).
In 1986, the CNIL brought a criminal case against a director of an informa-
tion bureau in Nantes. The director had not registered his data bank and refused
to cooperate with the CNIL. Found guilty of illegally collecting personal data
and obstructing a CNIL investigation, he received a FFr 20,000 fine and a
suspended prison sentence of two months.
In 1986, the CNIL required that caller identification services in the Biarritz
fiber-optic trial offer a blocking function to all subscribers. Call-by-call block-
ing was also included in the French ISDN service Numeris in 1988, and pro-
tections were instituted in the Mestel and Minicom electronic messaging ser-
vices (Gebhardt, 1990). Another case limited collection of genealogical records
by Mormons.
Germany
As mentioned, the state of Hesse enacted the first data protection law in 1970;
in 1977, a national law was passed. The discussion of data privacy in Germany
in the 1970s was colored by a fierce debate over how to contain a small group
of political extremists who engaged in terrorism. Data issues were described as
a choice between Datenschutz versus Tatenschutz (protection of data versus
protection of criminal deeds). Social Democrats were concerned with protection
of employees' privacy. Law-and-order conservatives, on the other hand, pushed
for elaborate police matching of various databases in order to identify potential
criminal suspects. They also supported a national machine-readable personal
identification card that would permit the tracking of individual movement; this
led to an anti-data-collection backlash. The 1983 national census was opposed
by many groups, and in an important decision, the Federal Constitutional Court
invalidated the holding of the census on the grounds of "informational self-
determination." In the meantime, several social democratic states pushed ahead
with their own legislation. In the state of Hessen, for example, the automated
processing of medical and psychological data is prohibited where information
is judgmental rather than factual. The use of data laws was not confined to
matters of privacy. It also provided a lever in other disputes. A 1984 decision
of the German High Court for Industrial Relations required the express consent
of a workers' council for the introduction of a computer system that would be
used to collect information on the performance of data technicians. The deci-
sion gave unions some influence over computerization where it affects employ-
ment security.
In 1988, telecommunications regulations were instituted that provided some
protection in calling identification and collection of subscriber data through
ISDN. Data compiled on Btx videotex and Temex may be used only for billing
and must be erased six months after collection (Gebhardt, 1990). Unification
extended privacy laws into the Eastern Lander.
United Kingdom
In the United Kingdom, the Advisory Council on Applied Research and Devel-
opment (ACARD), a consultative body to the prime minister and the cabinet,
recommended that one government agency coordinate all policies concerning
information technology and its applications, development, and operations. In
November 1982, the government adopted ACARD's recommendation by appoint-
ing a minister of state for industry and information technology to serve under
the secretary of state for industry. It was the first such appointment anywhere.
In response to another ACARD recommendation and spurred by the Council
of Europe's Convention on Data Protection, which the United Kingdom signed
in 1981, the government introduced a bill that eventually became the Data
Protection Act 1984. The Act requires the registration of data banks containing
personal information and sets other limitations on data collection and access
(Niblett, 1984). The home secretary designated as the first data protection re-
gistrar a computer expert, Eric Howe. Initially critical of the legislation, the
data-processing industry was partly mollified by Howe's appointment. The du-
ties of the registrar include the establishment of public files of data users and
data-processing bureaus. He is also an ombudsman for complaints of data abuse
and inaccessibility of records. Data subjects have the right to a copy of the data
held about them, and if such information is not provided, they may complain
to the registrar or to a court, both of which are empowered to order access.
There is also a right to compensation for damages resulting from inadequate
data security precautions. By 1990, approximately 153,000 companies and or-
ganizations with computerized lists of information on individuals had registered
with the Home Office representing 130,000 data users (Sixth Report of the Data
Distribution Registrar, 1990, p. 22).2
In 1987, the United Kingdom adopted a Model Code of Practice for telecom-
munications service operators that forbids unauthorized dissemination of data
on users (Gebhardt, 1990).
Other European Countries

In Sweden, government involvement in a large number of areas has led to the
evolution of many data banks in different government agencies. It was esti-
mated that the average Swede is registered in about 150 governmental computer
databases. This situation led to a special government commission and subse-
quent legislation in 1973, the first nationwide comprehensive privacy law any-
where. The law set up a Data Inspection Board for administration and policy
development that has the right to grant permits for the maintenance of personal
data files. A Data Policy Commission (subsequently the Information Policy
Commission) was created to include representatives of federal agencies, politi-
cal parties, the private sector, unions, employers, and local authorities.
Denmark passed two of the most complex and specific TDF laws in Europe,
based in part on Swedish and British privacy law (Feldman and Garcia, 1982).
A Data Surveillance Authority must provide a license for any database to be
collected or transmitted abroad. The authority also reviews proposals for inter-
national communication links to assure that data flows will not lower Den-
mark's standard of protection, and registers sensitive files.
In Austria, privacy policy illustrates how data protection laws can potentially
operate as a nontariff barrier. To transmit personal data abroad, a license must
be obtained from the Austrian Data Protection Commission. For countries where
similar data protection standards exist, no license is required (Wigand, 1985).
Switzerland provided privacy protection in its 1987 draft law on telecom-
munications. Article 16 of the law prevented data on telephone users from
appearing in statements and extended protections to videotex (Gebhardt, 1990).
In 1988, privacy legislation was passed in Ireland and the Netherlands. In
the Netherlands, the legislation followed twenty years of parliamentary debate.
The Dutch law permits any subject of a database file to inspect and correct
collected information for a small fee (de Pous, 1989; Gassman, 1989; Fauret,
1989). Italy, Spain, Portugal, and Belgium were slow to pass privacy laws.
A Comparison of European and American Approaches to Data Privacy
Legal and historical traditions, as well as economic motivations, explain the

disparity between European and American conceptions of TDF, particularly in
the realm of privacy protection.
In the United States, a generally more pragmatic approach to legislation has
led to the tackling of specific data abuses rather than to comprehensive laws.
Numerous laws exist, some of which are quite far-reaching (Noam, 1990)3.
The so-called Buckley Amendment, for example, permits students access to
files that are kept on them by public or private schools and universities, even
including letters of evaluation. Similarly, the Freedom of Information Act gives
the public access to government documents, excluding only information that is
deemed vital to national security, that is needed for the protection of confiden-
tial sources or the conduct of an ongoing criminal investigation, or that contains
trade secrets of other firms.
Nevertheless, U.S. privacy legislation remains considerably less strict than
European law in the regulation of private databases, and the coverage of Amer-
ican governmental organizations by privacy law is not comprehensive. Al-
though the Privacy Act of 1974 restricts collection and disclosure by the federal
government, only a few states and local governments have passed similar fair
information practices laws. The Privacy Act requires each federal government
agency to issue a public notice on its record-keeping activities. The Office of
Management and Budget (OMB) coordinates the government's efforts in this
area, and the protections cover all data files, electronic and conventional. The
Privacy Act explicitly protects only U.S. citizens and permanent residents, thus
excluding foreign nationals, whose personal data is exported from their national
place of employment to U.S. headquarters. Furthermore, the U.S. has no gov-
ernment agency specifically charged with data protection similar to the central-
ized data protection commissions or authorities established in European coun-
tries, though proposals have been advanced in Congress.
U.S. federal data protection requirements cover only consumer credit and
reporting agencies, and educational and financial institutions. Several of the
states have similar provisions. Under the Fair Credit Reporting Act of 1970,
individuals have the right to access their credit-rating file, to have corrections
included in the file, and to know the sources of information in credit files
(Noam, 1990).4
Whereas the European approach is to protect data by making its collection
and security requirements explicit, in the United States the abuse of information
rather than its collection is the target, and harm must be shown for laws to
apply (Feldman and Garcia, 1982). In many European countries, electronic
record-keeping systems must be registered with or licensed by a national data
protection board and may be subjected to inspection for compliance with the
law. Where private recordkeeping involves a large number of individuals, it is
permissible only if the individuals have a clear relationship with the record-
keeping organization (OTA, 1983).
Privacy and Free Speech
Privacy is not an absolute value. As the traditional clash between the privacy
rights of the individual and the free speech rights of the press illustrates, dif-
ferent societal objectives must be balanced against each other. It is by no means
clear that a maximum of privacy is desirable. If files that included personal
information could not be maintained, it would be difficult for the press to de-
velop an investigative story or for banks to extend credit.
Free speech is perhaps the main consideration that must be balanced against
the data privacy principle. Western democracies guarantee freedom of speech
in their constitutions. Labeling some types of speech "information flows" and
subjecting them to restrictive rules of a tradelike nature does not remove the
constitutional protection. Thus, the U.S. government may not enter into agree-
ments restricting the free speech of its citizens without subjecting itself to court
challenges.
Privacy protection can conflict with other societal interests. Law enforcement
is a consideration, as is technological advancement. Other countervailing fac-
tors to consider are protection of consumers (who may be preyed upon by third
parties whose identity is being protected), operational ease for networks, and
conflicting privacy interests. Although these factors do not obviate the need for
privacy protection, they suggest that a balancing of interests needs to take place.
Privacy, broadly defined, consists of two distinguishable but related aspects:
1. The protection against intrusion by unwanted information. This is some-
times termed "the right to be left alone" (Warren and Brandeis, 1890),
analogous to the constitutional protection to be secure in one's home against
intrusion by government.
2. The ability to control information about oneself and one's activities; this
is related in some ways to proprietary protection accorded to other forms
of information through copyright laws.5 A related aspect is the protection
of information about oneself from being tampered with by others.
The common aspect of both these elements is that they establish a barrier to
information flows between the individual and society at large. In the first case,
it is a barrier against informational outflows.
The concept of privacy is not without its detractors. There are three major
criticisms:
1. "Only the guilty need privacy." To the contrary, privacy is one of the
touchstones of a civilized and free society.6 Authoritarian or backward
societies do not value a private sphere, since they rarely respect individ-
uality and subordinate it to the demands of rulers or social groups.7
2. "Privacy is a drag on the economy." Privacy protections raise the cost
of information search. Potential employers and buyers, for example, would
have to spend more effort and money to find out who they are dealing
with if access to personal information is restricted. Deception becomes
easier, and transaction costs rise (Posner, 1981).
But there are also good economic arguments for privacy. It ensures
the ability of companies and organizations to hold on to their trade secrets
and details of their operations and to protect themselves from leaks of
insider information and against governmental intrusion. Information often
has actual value, and since much of it has no protection through property
rights, it must be protected through confidentiality or secrecy.8 To permit
its easy breach9 would lead to a lesser production of such information. It
has been shown in a theorem by Noam and Greenawalt (1979) that under
normal conditions "information of value, once released to one person (or
very few persons at most) will spreadin the absence of collusion
to all participants." Hence, the absence of privacy protection to stem out-
flow of information will lead to suboptimal production of such informa-
tion.
Similarly, anonymity may increase economic risk taking (though in-
crease risk for the partners to a transaction); certain investments may be
curtailed if the identity of their investors were disclosed. In that sense,
privacy protection acts as a spur to investment, as does the protection of
limited liability offered to corporations. (Of course, illegal activities are
also made easier.)
The loss of privacy also leads to inefficiency in information flows, just
as excessive privacy protection may. In the absence of privacy, people
use all kinds of hints or codes to reduce the outflow of information. Or
they may meet face to face instead of using the telephone.
Partly in response to economic and social needs, many transactions
have been specifically accorded special informational protection, known
as privileges (e.g., between attorney and client, penitent and clergy, pa-
tient and doctor, citizen and census taker). The idea in each case is that
the protection of information leads to a socially superior result even if it
is inconvenient in an individual instance to others.
3. "Privacy is of interest to a small elite only." To the contrary, attention
to privacy is widely shared. For example, according to information from
the New York Telephone Co., 34 percent of all residential households in
Manhattan and 24 percent of all its residential households in the state
have unpublished telephone numbers at subscribers' request. Most police-
men, doctors, or judges, to name but a few professionals, have unlisted
numbers. Elsewhere, it appears that the spread of unlisting is still further
advanced, reaching 55 percent in California.10
Pacific Bell planned in 1986 to sell subscriber information such as new
phone orders; more than 75,000 complaints came in, and the company
backed off.''
Will competition take care of privacy problems? Not necessarily. A compet-

itive environment may resolve some privacy issues, especially if it is possible
for a user to select a service provider that offers the desired level of privacy
protection. Carriers would lose business if customers felt unsure about privacy
of usage. But in many other instances, the greater openness of a competitive

system and the greater complexities of its multiple networks may also mean a
greater openness of information. It is probably easier to restrict the dissemina-
tion of information in a monopoly setting. By its nature, a network is a sharing
arrangement. In the past, this sharing encompassed mostly physical assets such
as trunks and switches. But as the "intelligence" of networks increases, and
as enhanced service networks and physical networks that particpate in com-
munications services evolve, the sharing also reaches data and other informa-
tional resources.
In telecommunications, privacy can be protected by either hardware or soft-
ware options. It makes sense to provide a guaranteed level of privacy protection
as a basic right and at no extra charge, and to offer enhanced privacy measures
on a for-pay basis for those that need them. It makes sense to define property
rights and transaction-generated information as belonging to both sides of the
transaction; to require the disclosure of privacy jeopardies; and to limit moni-
toring of employee calls. These principles, and others, were based on the pro-
ceeding initiated by the author (NYPSC, 1991).
Data protection laws can have unintended and paradoxical results for politi-
cal action. The case of the French section of the human rights organization
Amnesty International presents an extreme example. The organization main-
tains personal records on "prisoners of conscience" and victims of capital pun-
ishment and torture. French data protection law prohibits personal files from
including information on prison sentences or convictions. Such information,
however, is fundamental to AFs activities. The law also prohibits the recording
of information about the racial, political, philosophical, religious, or union af-
filiation of individuals. Again, such information is essential to Amnesty Inter-
national's activities. Under the French law, personal information collected can
be transmitted only to the few countries that have similar data protection laws.
But the essence of AFs work is to disseminate information widely about pris-
oners of conscience, and this would be impossible under a literal reading of
the law. In proposing a solution to these problems, a study group of interna-
tional data protection commissioners recommended the creation of a committee
to deal with "bona-fide international organizations pursuing humanitarian goals
or defending human rights on an international basis, such as Amnesty Interna-
tional or the International Red Cross" (TDR, 1983). Yet the selection and cer-
tification of such organizations enables government authorities to certify some
organizations as worthy and humanitarian and others as not.
Privacy protection has become especially important in the telecommunica-
tions field, as new technology such as automatic number identification began
to be offered.
In 1990, the European Commission issued a draft directive establishing basic
telecommunications privacy rights across the twelve member states. The direc-
tive includes restrictions on unsolicited calls, calling number indentification,
and use and storage of data collected by telephone carriers for electronic pro-
files.
Sovereignty
The fact that foreign institutions and governments have access to domestic data
can be troublesome to many countries. A 1979 Canadian study found that nearly
50 percent of U.S.-based information bureaus storing data on Canadian trans-
actions would have provided data access to U.S. organizations such as credit
bureaus, law enforcement agencies, employers insurance firms, and govern-
ment agencies (CCITCS, 1979).
Similarly, Sweden became alarmed when the fire department of the city of
Malmo accessed data on chemical storage and truck routing on a computer of
the American firm TRW, located in Cleveland, Ohio. Swedes feared that the
U.S. government could interfere with Malmo's fire-fighting efforts. In conse-
quence, the fire department began using a data file located, apparently more
safely, in Holland (Feketekuty and Aronson, 1983).
The French government was concerned with the use of American-based
econometrics models in the projection of French economic trends. It feared that
the U.S. government could access those models and gain insight into the French
economy.
To be fair, the United States has given Europeans some cause for concern.
In one often retold episode, it tried unsuccessfully to pressure European coun-
tries to cancel a major gas pipeline deal with the Soviet Union. When that
failed, the American government ordered the producer of the essential com-
pressors for the project, Dresser Industries, to lock out French engineers from
the design database (Botein and Noam, 1986). Another example involved
Lockheed's Dialog on-line database. Dialog contracted with the International
Institute for Advanced Systems Analysis (IIASA) in Vienna to provide access
to its publicly available database. Since the Soviets had access to the IIASA
data gateway, the U.S. government became concerned that they could establish
easy access to U.S. data systems, and it restricted Dialog access through IIASA
(Pipe, 1984b).
The United States is not consistent in its advocacy of a free flow of infor-
mation (Cass and Noam, 1990). In fairness, however, these departures are ex-
ceptions. Furthermore, to argue legitimately for the principles of free flow of
information, one need not be virtuous beyond reproach.
Industrial Policy and Data Flows
The United States enjoyed a headstart in on-line databases from which it still
benefits. In 1983, worldwide revenue from on-line services was about $2 bil-
lion, of which the United States accounted for more than three-quarters (An-
derla, 1983). In 1983 figures, an OECD report found sales of on-line data
services in Western Europe to be less than 10 percent of sales in the United
States.
As with most information produced initially for a domestic market, the mar-
ginal cost for the export of these databases is low. With a head start and a
substantial domestic customer base, U.S. databases provide tough competition
for European systems. Although the availability of low-cost and well-tested
foreign databases would be in some cases welcome, since domestic replication
is made unnecessary, in most cases, other countries fear informational depen-
dency.
Given this database and data-processing discrepancy, some American firms
believe that what TDF is really about is economic protectionism in favor of
local data processing firms. A Canadian study found that, in 1985, the value
for imported computing services was about $1.5 billion, which could fund an
estimated 23,000 jobs in the Canadian data-processing industry. With greater
frankness than is normally offered, the study conceded that issues of privacy
and protectionism are closely intertwined. The report found that the major problem
inherent in flows of Canadian data to the United States is "not one of privacy
of Canadian data subjects being invaded by data about them being stored in the
United States. It is rather that data processing in the communications business
may be lost to Canadians as a result of this foreign flow" (Turn, 1979, p. 6).
To combat this trend, the Canadian Banking Act of 1980 required that customer
data be processed and stored in Canada, thereby forcing U.S. banks to dupli-
cate their hardware and software instead of relying on existing data-processing
facilities across the border.
Among Third World countries, Brazil has been particularly active in data
and telematics issues, and its policies have received wide publicity. Instituted
during the military dictatorship, the thrust of Brazil's policy is evident in the
statement of its top information officer, in both senses of that title:
[T]he administration [i.e., restriction] of TDF appears to be an effective govern-

ment instrument for the creation of an environment that makes the emergence of
an internationally viable national data-service industry possible. By itself, such an
industry would have had great difficulties in overcoming the obstacles of a com-
pletely "laissez-faire" environment. The country's TDF policy altered that situa-
tion [Pipe, 1984b].
In Brazil, a license must be obtained from the special Secretariat for Informat-
ics before establishing international data links. Applications for foreign pro-
cessing, software import, and database access are rejected if domestic capabil-
ity exists. The rules are an attempt to strengthen and develop the domestic
industry. The policy was strongly embraced by the Brazilian military dictator-
ship and its business and industrial allies, and it was admired by many observ-
ers who otherwise feel no kindness toward right-wing juntas.
Beyond the issues of privacy, sovereignty, and protectionism, data flows also
affect the future of PTT monopolies. The 1980s witnessed an acceleration in
the creation of private data networks, and technology offered many possibilities
for data flows bypassing PTTs. Privacy protection provides an argument for
service monopoly, given that private systems, especially if they are shared by
several users, cannot be trusted to enforce data protection provisions.Thus, in
Germany, the media policy spokesman and secretary general of the Social
Democratic party, Peter Glotz, argued that the network monopoly of the Ger-
man Bundespost was necessary for data protection.
Another German official, Spiro Simitis, the pioneering data protection com-
missioner for the state of Hesse, argued that it was necessary to enact special
hardware standards:
[T]he rudiments of a possible set of regulations are emerging. Until now legislation
has been chiefly concerned with storage; the conditions governing processing hav-
ing been abstract and controlled general procedures. Now we must turn our atten-
tion directly to hardware, in order to find out whether and how far technical devices
conforming to data protection laws can be built into all the equipment in question.
A decisive number of statutory requirements could then be met through construc-
tion standards [Simitis, 1985].
At present, the possible effect of TDF restrictions on trade has been largely
theoretical. But since the United States will probably not adopt data protection
legislation to match West European strictness, data flows across the Atlantic
could be impeded. This would affect trade, because the amount of electronic
data flow that accompanies business transactions has increased dramatically.
Can one truly expect a leak-proof system of data protection? Not surpris-
ingly, "data havens" soon began to emerge, including Monaco and Liechten-
stein, and Andorra.
In the friction that ensued on TDF issues and in the traditional manner of
international compromise, which acknowledges that there is merit to the claims
of both sides, the formula of "regulated free flow" emerged. But no elegant
formula could easily cover the differing attitudes toward transborder data flow
on the two sides of the Atlantic.
IV
THE FUTURE OF
TELECOMMUNICATIONS
38
Networks in the Future
The scenario of the future of telecommunications that is held by traditionalists

is well-known: increased integration, broader bandwidth, wider mobility, greater
ubiquity. All these tendencies are anticipated correctly, yet they represent only
one side of the coin. The discussion of earlier sections of this book aimed at
demonstrating the dynamics of disaggregation in networks. If one gives indi-
viduals the freedom of association and if the economics are right, in time they
will form new types of interlinkages that we call networks. What are some of
the long-term implications for networks?
7. Networks will become transnational. As the cost of transmission continues
to drop, networks will not be territorially organized. Territoriality was based
on the need for a network architecture that primarily minimized cost by mini-
mizing transmission distance. It led to the creation of the "German network,"
the "French network," and so on. This technological and economic territori-
ality suited governments everywhere, because they too were based on territo-
riality of jurisdiction and could thus conveniently exercise control and even
ownership over "their" network. But that is changing. Networks are increas-
ingly becoming pluralistic group affairs. Groups break off parts of their com-
munications needs from the public network and aggregate them in their own
associations, usually on the facilities of a public carrier, but under different
rules and prices.
Territoriality becomes secondary. Many of these communities of interest
transcend national frontiers. Their interests are continental and global, and so
are their networks. When the computers of brokers and investment banks in
New York are interconnected by a continuous network and interact with those
in Tokyo and London to trade and clear transactions, one can no longer say
that there is a New York or Tokyo market. There is no longer a physical locus
for transactions; the network itself becomes the market.
2. New electronic neighborhoods will emerge. A few years ago, it became
fashionable to speak of telecommunications creating the "global village." There
was something inspiring in this image, communal and peaceful. But there is
nothing villagelike in the unfolding reality. Instead, groups with shared eco-
nomic interests are extending national group pluralism into the international
sphere through the opportunity to create global interconnections with each other.
Indeed, communications make international pluralism easier because it is easier
to reach a critical mass for subnetworks if one aggregates users across several
407
408 Networks in the Future
countries. The new group networks do not create a global village; they create
the world as a series of electronic neighborhoods.
In the past, neighborhoods had economic and social functions. In New York,
for example, there are Chinatown, Little Italy, the Garment District, Wall Street,
Madison Avenue, and the Theater District. All have defined social and eco-
nomic functions. There are regions with specialized productionSolingen and
Sheffield for cutlery; Lyon for silk; Hollywood for films; Silicon Valley for
microelectronics (Piore and Sabel, 1984). Production clusters create economies
of aggregation that substitute for the economies of scale and scope of the giant
multiproduct firm. Physical proximity was a key. But now group networks can
serve many of the functions of physical proximity. They interconnect special-
ized producers, suppliers, buyers, experts, and markets. They create new ways
of clustering, spread around the world.
Some of these electronic neighborhoods will be nicer than others. They will
perform better, faster, and often even cheaper. In developing countries, the
networks of those transacting with the world are already becoming better than
those of local people. In places like China or Egypt, a two-tier communications
system has emerged.
Networks might also be stratified along socio-demographic dimensions. Al-
ready, some long-distance resellers in the United States offer bonuses to churches
if they sign up their members. Such marketing efforts can lead over time to
identification of some network with particular ethnic, religious, or political groups.
Similarly, some networks may be shunned by labor union members if they have
a history of labor problems.
3. Personal networks will emerge. But why stop at networks for groups? If
the trend is from national public networks covering the entire population to a
pluralist system, why not expect still further disaggregation? This additional
step means individualized networks. Before dismissing the notion of personal
networks (PNs) as extravagant, let us remember that twenty years ago nobody
expected computers on everybody's lap either.
What does a personal network mean? It means an individually tailored net-
work arrangement. It does not mean a separate physical system, except for
inside wiring and maybe the last mile of circuits, plus some radio-mobile links
and terminal equipment. The rest could consist of virtual networks, provided
by a whole range of service providers and carriers, and packaged together to
provide easy access to an individual's primary communications needs: friends
and family; work colleagues; frequent business contacts, both domestic and
foreign; data sources; transaction programs; frequently accessed video publish-
ers; telemetry services, such as alarm companies; bulletin boards, and so on.
Contact to and from these destinations would move with the individuals, whether
they are at home, at the office, or moving about.
4. Networks will assume political power as quasi-jurisdictions. Historically,
the nation-state was at tension with cross-border allegianceswhether proletar-
ian international solidarity, rebellious youth culture, international financial cap-
ital, or ethnic minorities. The new network environment weakens national
cohesion. It strengthens particularism and internationalizes it. It is difficult for
Networks in the Future 409
a state to extend its powers beyond traditional frontiers, but it is easy for the
new networks to do so.
Furthermore, these network associations possess and acquire powers of their
own. They already may link powerful entities and can bring their combined
powers to bear. For example, the combined weight of the members of the
SWIFT banking network got the powerful national PTT monopolies to cave in
on a number of crucial issues. And there is no reason to expect the power of
network combinations to be directed only at communications issues. Once groups
are in constant touch, they may as well get organized on other issues too. The
communications network becomes the political network.
They will coordinate in the economic sphere. When it comes to the role of
information, the line between competition and cartel coordination has always
been a fine one. In the 1920s, various American industries established so-called
fair-price bureaus that gave each member of the industry a convenient look at
what its competitors were charging. This practice was outlawed in a series of
antitrust cases. Imagine if one leaves instead information exchange to a series
of artificial intelligence programs communicating internationally. One has a
real problem of conceptualizing, detecting, and preventing international cartels.
One person's collusion is another person's programmed trading. The network
becomes the cartel.
The network associations are also likely to become quasi-jurisdictions them-
selves. They have to mediate the conflicting interests of their members. They
have to establish cost shares, sometimes creating their own de facto taxing
mechanism as well as redistribution. They have to determine major invest-
ments, to set standards, to decide whom to admit and whom to expel. As a
network becomes more important and complex, control over its management is
fought over. Elections may take place. Constitutions, bylaws, and regulations
are passed. Arbitration mechanisms are set up. Members are assessed finan-
cially. Networks become political entities.
Thus, we may be witnessing the creation of new and often extraterritorial
forms of new quasi-jurisdictions that are not clearly subordinated to others. In
response, governments might create forms of domestic and international regu-
latory mechanisms for specified sets of problems, possibly based on global
networks themselves that continuously collect and exchange information, track
activities, and coordinate enforcement.1
5. Networks will exercise power toward their members. A major question is
whether a network group can dominate its own members or be restrictive in
allowing others to join. The power of the network becomes most obvious when
it is operated by a dominant entity. For example:
The network of a university such as Columbia can legally exercise restric-
tions toward its members. It can and does limit terminal equipment and
options and charge monopolistic prices, and it could legally refuse to serve
political groups that are disfavored.
The major U.S. videotex service, Prodigy, prevents its users from dis-
cussing politics or Prodigy itself on the system. When Prodigy, which
410 Networks in the Future
provides an extensive messaging service, announced that it would raise

the rates for such messages, a group of subscribers posted notices in a
"public area" of the system encouraging other subscribers to protest. When
Prodigy removed these messages, the protesters turned to the private mes-
sage feature, and sought help from advertisers. Thereupon, Prodigy can-
celed the subscriptions of the protesters.2 It later permitted the discussion
of politics, only to become the subject of protests when such discussions
became nasty.
On the public networks, too, content control emerges. American tele-
phone companies recently sought to establish their right to the restrict
otherwise lawful "adult" communications if they were harmful to telco's
image. The three major U.S. long-distance carriers (AT&T, MCI, and
US Sprint) all instituted restrictions on their audiotex ("900") services.
Employers frequently block the ability of their employees to reach certain
numbers. While this is based on protections against running up telephone
bills generated by dial-it services, the principle could be extended to ex-
clude messages of a type undesirable to employers, such as those of labor
unions. A growing number of businesses maintain electronic mail systems
for internal communications, and some have asserted their right to moni-
tor employees' messages. In one instance, a mayor read the private mes-
sages that city council members had sent over the municipal system.
In so-called intelligent buildings, landlords provide communications to
occupants. These "shared-tenant services" are largely under the control
of the building owners, whose interconnection decisions determine which
networks tenants can reach.
Petty monopolies can thus emerge, largely unencumbered by the protections
built into the public network, at least in the past, by law, custom, and regula-
tion. The option is to exit, which may mean giving up a job or an apartment
usually not a realistic option.
Are there freedom-of-speech rights for users (in network terminology "com-
mon carriage obligations") in group networks? The scope of these rights is
undefined. Constitutional rights are unclear in the case of a private group net-
work. Regulatory impositions of such obligations are possible but are limited
by the rights of groups to substantially define their membership and the rules
under which they operate, especially where a major purpose of the groups is
communication, and thus the exercise of a fundamental right itself (i.e., free
speech). In such circumstances group activities have strong protection from
restrictive regulation.
Even where network groups are organized democratically, they may well be
restrictive. A major function of liberties, after all, is to protect minorities from
unsympathetic majorities. In the public sphere, guarantees of free speech against
governments are part of constitutions. In the network environment, the granting
of access and nondiscriminatory content-neutrality is required of the general
"public" networks by law or by common carriage regulation. But common
carriage does not necessarily apply to group networks. Groups may institute
Networks in the Future 411
restrictions on the exercise of speech over their network and assert that their
status is akin to publishers, with no rights for users. They can exclude certain
subjects from being discussed or certain speakers from having access to the
network. This could become particularly an issue when telecommunications
networks gain the ability to transmit video programs. It is true that individuals
could form alternative networks if they are being restricted. Thus, market forces
could help, but not if some of the networks control some segments of a chain
of communications, or where the ability of any link in such a chain to institute
content-based tests would impose transaction costs on the entire system.
In this context, the exercise of freedom of association may lead to group
formations that restrict speech. Hence, the evolving pluralistic structure of tele-
communications may bear the seeds for a new type of bottleneck to the free
flow of information that did not exist on the traditional public network and its
common carriage.
These conclusions point to bottlenecks, instability, conflict, and new forms
of restrictions, all within an environment rich with communications. What are
the long-term implications for policy? In the past decade, policy discussion was
correctly focused on creating openness by reducing barriers and permitting en-
try. In Europe, this process is only in its early stages. But as the diversification
of the network environment is proceeding, the next set of issues is to create
points and rules for integration that permit the continued interoperability of a
"network of networks." These issues are more difficult to deal with than the
policy questions of the past. They require the conceptualization of a post-
liberalization environment.
39
Toward a Modular Network
In the future it will become increasingly difficult to define what "the network"
is. The distinctions between private and public, basic and enhanced, terminal
and network equipment, national and international, will fade. It is best to think
of the network as a federation that must interact. The traditional arrangement
was based on the notion of sharing of resources in terms of technology, eco-
nomics, and politics. But the evolution of networks moves away from this
sharing arrangement, because of new entry by suppliers, and, perhaps even
more important, of exit of major users from the old sharing coalition. It may
be more efficient on some level to share, but by the same logic one should not
buy books but use the library, or one should not buy a car but hire a taxi. In
other words, the efficiency of capacity utilization is not the only economic
force driving an economic system. What is important for policy is not to ban
the private car, figuratively speaking, but to make sure that there are basic rules
of the road for interaction, based on the principles of compatibility, intercon-
nection, nondiscrimination, and reciprocity, that make it possible for the emerging
quarrelsome network family to live and work together. It is true that in the
short term, revenues may be lost to the public network operator and that addi-
tional costs may be incurred. But one must take the long view: Cost tends to
come down with competition; network revenues will go up as utilization goes
up, and new applications should increase utilization. Where there are still rev-
enue losses, regulators should help make them up elsewhere, such as through
access charges, rather than stop the evolution of the network. Contributions to
traditional social services could also be made by the newer members of the
network family.
The trends toward technical integration and toward institutional and business
diversity are, to some extent, substitutes for each other. To advance technolog-
ically, one can upgrade a telecommunications system by more powerful inte-
gration, such as through integrated narrowband and broadband networks (ISDN
and IBN), and benefit from their economies of scale and scope and from their
greater technical standardization and compatibility. Or one can choose a more
competitive diversity and benefit from its dynamism and cost consciousness.
Figure 39.1 maps the strategies of several countries over the past few years.
Generally speaking, the European PTTs stressed ISDN-style integration,
whereas the United States mostly followed the path of diversity, the compara-
tive advantage of its society. Japan's approach has been fairly balanced in com-
412
Toward a Modular Network 413
Technical integration
Figure 39.1. National Strategies for Network Evolution.
bining a major push for both in terms of diversity and integration. Diversity
can lead to innovation, but it can also retard innovation where there are many
independent parts of a system that must interact. Then change can become
much harder.
The implication is not that one should go to the other extreme and protect a
monopoly system, but rather that it is necessary to provide the system with
some tools of integration where they are not self-generating.
Mapping the Network
Analyzing the network environment requires a more systematic look. Ob-

viously it is possible to provide access and interconnection of networks to each
other in ways that are technically and economically inefficient.
It is helpful to think of a network as consisting of hardware and software
functions.1 In software the tendency is toward modularity.2 An example for
modular software hierarchy is the Open Systems Interconnection (OSI) model,
which was adopted in 1986 by the International Standards Organization (ISO).
OSI is based on a hierarchy of seven layers, each of which has defined func-
tional responsibilities. An upper level layer is reliant on the lower layers. But
they are, in principle, independent modules, and in theory one can rewrite the
software protocol for any layer, and replace it without having to change any of
the other layers. In actuality, some layers are integrated, but this need not
affect a conceptual map.3
The other dimension is hardware. Here it is helpful to think of a network
architecture as a sequence of physical segments. For example, the subscriber
terminal itself. Or the inside wiring from the terminal to the network termina-
tion point. Or the trunk between the local office and the tandem office higher
414 Toward a Modular Network
Figure 39.2. Mapping the Network and Interconnecting the Competitive

Islands.
up in the switching hierarchy. In the context of defining ISDN standards, the

international standards body CCITT defined the segments close to the user very
carefully and separated them with demarcation points known as the R, S, T, U,
and so on. One can use the same technique to define segments throughout the
network.
One can put together the software and the hardware presentations into a
system of coordinates. On the horizontal axis of Figure 39.2, we have the
physical segments, from the periphery of the end user (say a point Z) up through
the network hierarchy to point A, and back to another end user, Z. On the
vertical side, one has a software hierarchy.
The boxes of Figure 39.2 thus graph the networks schematically. Each part
of the network is defined by a set of coordinates for its software and hardware
location, and each service element can be graphed into this map. Element a,
for example, could be an interoffice transmission trunk. It is located between
two points, which we assume hypothetically to be the physical coordinates for
interoffice transmission, and it occupies the first three layers of the OSI hier-
archy. Element 6, similarly, is an applications module, located in the top layer
and physically in the central office port that lies between some horizontal points.
Element /3 is terminal equipment, such as a fax machine.
Almost all this territory used to be occupied by telecommunications monop-
olies such as AT&T or the PTTs. But one of the developments of the last two
decades has been that other suppliers are available too. The alternatives are
schematically graphed in Figure 39.2. In this case, there are a^ and 182 elements
that are offered by alternative vendors, in competition with the a\ and /3, of
the traditional monopoly carrier. But, and this is important, the alternative ser-
vice blocks that are offered usually lack the connecting physical and software
elements that are necessary for an end-to-end connection with users, which the
traditional carriers possess. This is why, if alternative service elements are to
exist and survive, one must provide a framework of interconnection with the
Toward a Modular Network 415
other elements of the network, in a way shown schematically by the winding

path in the graph, so that one could use the alternative a2 and /32 and still not
be left cut off from the rest of network functionalities. And this is the major
rationale for ON A and similar access and interconnection arrangements. Even-
tually, the islands will grow larger and fill the entire map. In the meantime,
however, one can establish islands of competition only if one assures the ferry
service to them.
A Network Grid
As these islands grow, they must operate together in a sensible manner in terms
of technical standards, protocols, and boundaries. This is why it is necessary
to establish a network blueprint. It would be, conceptually, a grid of defined
vertical and horizontal coordinates, and the technical standards of interconnec-
tion and interface between them. In this fashion it would set out a system of
modularity that would make possible an interconnecting modular network sys-
tem. Within the modules, providers could do more or less whatever they wanted.
And they could connect modules together. But one could replace one module
with another, and it could interact with the rest of the network.4
The transfer from one module to the next will not be free. The charges can
be structured to support traditional concerns, such as universal service, and
assure the viability of the core network.
Modularity enhances competition and will therefore be viewed negatively by
the traditional carriers. On the other hand, it would make them much less de-
pendent on any particular equipment manufacturers, since there is likely to be
more competition to supply any specialized module than to provide an entire
central office switch, which is a billion-dollar development effort. Thus, the
modularity of software will free carriers from switches with multimillion-line
programs.5
Today most computer hardware is designed to accommodate an operating
system such as DOS or UNIX with applications programs such as spreadsheets
and word processing. Telephone digital switches, though similar to computers,
in effect mix the operating system with the applications, so that it is difficult
for telephone companies or independent software companies (as opposed to
switch manufacturers) to write the new applications software either because its
millions of lines are impenetrable or because they cannot touch it legally. Mod-
ularity would also deal with the inevitably increasing competitive overlap be-
tween telecommunications and computer industries and would assure that intel-
ligence does not migrate into the terminal equipment periphery of the network
for purely regulatory reasons.
Market niches for small hardware suppliers would open. The carriers could
encourage the development of software applications by outside suppliers, just
as IBM did by opening software applications for its personal computers. This
would enhance the telephone carriers' flexibility. Right now, changing network
capability and services is a ponderous process.
416 Toward a Modular Network
It would be similarly possible for the VAN service providers to provide new
applications by placing them among the central office software functions them-
selves. In other words, software by outsiders would be put into the central
exchange. This could be called software collocation. PTTs will shudder today
at the notion. But it may be attractive commercially to them, as long as it
conforms with standards and protocols, does not displace their own functions
because of limited capacity, is limited to higher-layer applications rather than
network control functions, and of course yields revenues. This could open up
a scenario of exciting applications.
40
Telecommunications Liberalization:
An Expansionary Process?
A Model of Interaction in Telecommunications Policy

A look across countries and across different economic sectors shows a spread-
ing in the liberalization of previously strictly controlled economic activities. In
the United States, deregulation expanded functionally from one line of business
to adjoining ones (e.g., in transportation, from airlines to railroads and then to
trucks; in telecommunications, from equipment to long-distance service, local
transmission, and central office functions). Liberalization has also spread geo-
graphically. In telecommunications it has moved from the United States to the
United Kingdom, Japan, and to some extent continental Europe. In air trans-
portation, too, it proceeded from the United States to Britain and Europe. One
should note that in other historical periods the opposite trends have occurred,
and regulation has expanded. In the United States, for example, state railroad
regulation led to federal railroad regulation in the 1880s, which in turn spread
in the 1920s and 1930s to trucking, buses, and airlines. In financial services,
regulation of savings and loan banks expanded to commercial and investment
banks and brokerage services. In telecommunications, national telegraph regu-
lation of European countries was extended, in the 1850s, to international ar-
rangements, and later, to telephone service.
This leads to the question of why these long-term international trends are
taking place. Is it a change in ideology, in which the Chicago School of Eco-
nomics follows, for example, the Fabian Society? Or is it the political domi-
nance of one country, which is then reflected in international trends of policy?
Or is it a dialectic cycle, in which the inevitable shortcomings of any policy
lead in time to the adoption of another? Chapters 5 and 6 discussed this ques-
tion, largely in domestic terms. We shall now add the element of international
interaction.
Interaction creates instability. The more interrelated countries and economic
activities are, the less likely are stable solutions to separate policies. And where
instabilities exist, they ripple throughout the entire system. It becomes increas-
ingly difficult to control all the elements in a complex matrix of interrelations.
Ultimately, overarching control over many countries and many economic activ-
417
418 Telecommunications Liberalization: An Expansionary Process?
ities would be necessary to restore stability. And since this power does not
exist, or is usually not deemed desirable, regulatory strictness may unravel.
The following pages provide a simple analysis of regulation in its intersec-
toral and international dimensions. We start off by narrowly defining regulation
as the setting, by a regulatory agency, of a price vector R for a set of economic
activities. A total prohibition is an infinite price set by the state; a totally
laissez-faire regime means a market price. Most regulation, however, is some-
where in between and can be viewed as a way of making an economic activity
costlier (as in pollution control) or cheaper (as in residential telephone usage).
Various interest groups are affected by the setting of these prices, and they
seek favorable P's by exercising pressure through the political process.
Where will R be set? This depends on the optimizing function of the regu-
lator. For purposes of the argument, it is not necessary to specify this function
in detail, but to assume an interest group model of politics. Regulatory behav-
ior is affected by the groups according to their power and stake in the outcome.
Let us assume two interest groups, A and B, each with a "political weight" of
WA and WB. Each group is affected by the regulation R, with the effect E
described by , = gi(R). We assume for simplicity that the two groups are
impacted differently by regulation, in that one gains from a higher R, whereas
the other loses from it.
(1)
Each group asserts pressure P, according to the strength of impact ,, weighted
by the group's political weight Wt.
(2)
The various pressures are in equilibrium when
(3)
which is where the ratio of the benefit functions is equal to the inverse of the
political weights of the respective groups. Since the W[ and g, are given, we
can determine the expected regulation R in such a system as
(4)
We now introduce a second and related economic activity and denote it by

2. This activity affects both A and B in their activity 1. For example, activity
2 could be telephony, which affects telegraphy (activity 1) and two interests
affected by it, carriers and users. Similarly, activity 2 could be functionally the
same as 1 but exercised in a different political jurisdiction. The demand and
supply for one activity tend to be related to the other, either as substitutes or
as complements. Hence, in an interrelated world, the politically optimal regu-
lations may be different from those for a single activity in an isolated jurisdic-
tion.
Telecommunications Liberalization: An Expansionary Process? 419
Figure 40.1. Reaction Functions of Regulatory Policy by Two jurisdictions.
(5)
In most instances, we will encounter a cross-elasticity C that is positive:
(6)
But in some instances, cross-elasticity of regulation C would be negative. For

example, if banking laws are tightened in Italy, they may be lowered in Swit-
zerland, since its banking industry, which benefits from inflows from Italy,
would be worse off than before without counter-relaxation. Similarly, if Swit-
zerland lowers the strictness of its banking regulations, Italy may have to tighten
up its own to reduce outflow. This can lead to instability. As Italy successively
tightens up, Switzerland keeps liberalizing.
An example in telecommunications is transborder data flow protection laws.
The less protected data is in one country, the tighter the other country may
become in response.
Now for activity 2 the same holds true. Here too are two interest groups,
denoted by C and D, and an effect of /?/ with a cross-elasticity of C2. There-
fore, we can think of two "reaction functions"// and/2, that track the response
of one regulation to the other's given level.
A point such as R' would denote the two independently set regulatory poli-
cies (Figure 40.1).
But once we postulate reactions to each other, there would be a shift to R*.
(7)
Under moderately sized and positive cross-elasticities, there will be an equi-

librium point such that regulation will be lower for desirable activities and
higher for undesirable ones.
Examples are
(a) lower telephone rates, if one wants to attract business from the other
jurisdiction; and
(b) stricter local zoning laws, to prevent undesirable activities from spilling
in.
There is no need for coordination between 1 and 2; an equilibrium can be
reached by unilateral actions and reactions.
However, an equilibrium requires that the reaction functions 1 and 2 are such
that 2 is steeper than the inverse of 1 at the point of intersection (i.e., that 2
cuts 1 from below).
If the reverse is true, then there is no equilibrium, and the regulatory strict-
ness either moves successively higher or lower to corner solutions. Examples
are the "race to the bottom" in state corporation law in the United States, or
tax havens internationally. The reverse is the increasing exclusion by strict
regulation of chemical waste dump sites. These are instances of corner solu-
tions. For other configurations of the reaction functions, cyclical change is pos-
sible.
Instability raises questions of how to prevent it, and therefore leads to the
issue of policy coordination. But even in stable situations, R* may not be the
optimal result for either 1 or 2, or both, and policies may be sought to affect
moving R* back to R'.
There are several possibilities for such policy coordination and they are touched
upon in the following:
Supraregulation
Supraregulation is encompassing regulation, either across jurisdictions (e.g., a
supranational policy) or across functions (i.e., "supra-modal"). This expands
regulation to a higher level of institutions (e.g., to the European Commission)
or to a wider institution, such as the Interstate Commerce Commission in the
United States which regulates all modes of surface transportation.
How is Supraregulation set? By analogy to the single-jurisdiction case, the
two sets of interest groups are assumed to affect the joint jurisdiction in an
aggregated fashion, if the suprajurisdiction is answerable to the body politic.
(8)
This can be higher or lower than uncoordinated outcomes. Supraregulation
is not invariably stricter than particularist regulation, for the reasons discussed.
In telecommunications, for example, the regulatory principles of the European
Commission are less strict than those of most of the member states. In the
United States, the same holds true for the FCC vis-a-vis the state Public Utility
Commissions. But the reverse is also often the case (e.g., in the regulation of
securities).
One question is "Why would interest groups (or a whole country) consent
to supraregulation?" Would this not dilute the power of dominant groups? This
is indeed true. Therefore, a dominant group will normally consent to a shift to
supraregulation only where its favored policy would be enhanced (e.g., if the
balance of power of interest groups in the other sector is even more favorable
to its concerns). However, for symmetrical reasons, the dominant group in the
other sector would then oppose supraregulation lest it dilute its own influence.
Log rolling aside, this then leaves two primary scenarios (the second of which
is the more important) in which mutual joining of supraregulation will occur:
1. When the balance of power is essentially equal in the two sectors, so that
supraregulation does not make much difference. This is why policy co-
ordination is easier among Western European countries than, for ex-
ample, East Asian ones.
1. When supraregulation establishes a policy cartel to avoid separate regu-
lation to affect each other and to lead to results that are considered sub-
optimal by the dominant groups. This is more likely to be important where
the cross-elasticity of regulation is high, which is likely to be the case
with sectors or countries that interact strongly with each other.
In other words, the advantage of the elimination of uncontrolled interaction
must be greater than the value attributed to control and independence. Of course,
de facto independence already had been lost through the mechanism of inter-
action, and supraregulation reflects this.
A related issue is that of uniformity in regulation. Here the issue is not the
strictness per se, but the importance of being identical. There are situations
where efficiencies exist in uniformity or connectivity. (The width of railroad
gauge or protocols in telecommunications are examples.)
Technologists tend to favor standardization. Economists have more mixed
views, because uniformity has its costs. To have cars with identical pollution
controls in both Norway and Belgium may not necessarily be optimal for either
or for both jointly.
In terms of the model, uniformity is given by the 45 line, which is likely to
be off the equilibrium point /?*. Uniformity is the dominant policy throughout
if at least one reaction function is sloped at 45 (i.e., where at least one juris-
diction reacts to the other's change by an identical change so as to preserve
uniformity). But that is an exceptional situation.
There may be great incentives for one state to be nonuniform. Examples are
large countries for whom international interaction may be small in relative terms
(e.g., the United States, which for example affords a nonmetric system of mea-
surements). At the same time, many other examples are small jurisdictions:
Switzerland in banking; Delaware in coiporation law; Hong Kong in tariff du-
ties; Liechtenstein in taxes; Monaco in gambling; and Luxembourg in broad-
casting. These examples suggest that small countries have incentives to be non-
conforming, probably since the loss in revenue, control, and so on, from their
own relatively small domestic economies is more than offset by the inflow from
the larger countries resulting from nonconformity. To prevent such nonuni-
formity, the other states may have to impose substantial pressure on the mav-
erick jurisdictions or pay significant compensation.
Regulatory Treaties
Another possibility of coordination is to establish interjurisdictional or intersec-
toral treaties. Here there is no supraregulation by a suprabody, only an agree-
ment, and each side must be better off than before to enter into it. Agreement
will stop at the point where marginal benefits of marginal regulation will begin
to be negative for at least one jurisdiction. Thus, such a system is a convoy
traveling at the speed of its slowest ship. It can be extended further where
compensation to some participants is possible, which is one of the ways the
European Community operates in the agricultural field.
Regulatory Colonialism
An extreme example for regulatory coordination could be called regulatory co-
lonialism, when one jurisdiction can set regulations for another jurisdiction solely
to benefit itself. Britain's imposition of regulations on cotton spinning in India
or on opium trade in China are examples. Other illustrations are the use of the
Interstate Commerce Commission by American railroads to impose regulations
on trucking, or the successful pressure put on the FCC by American broadcast-
ers to restrict cable television for a number of years.
Instability
The problem of any coordinated regulatory structure is its instability. First, no
equilibrium may be possible, because the reaction functions do not meet the
stability criterion described earlier. Second, one jurisdiction's adherence to an
agreement provides the other with an opportunity of gain by seeking a non-
cooperative policy. In each jurisdiction there are pressures to seek one's own
ideal regulatory level, which is likely to be different from the agreed upon level
or from the interactive equilibrium. Going it alone can be due to shortsighted-
ness or lack of understanding of the interaction involved. But it can be based
on the rational desire to gain advantages over others by breaking joint policy,
at least in the short run.
In telecommunications, for example, communications "havens" are possible
and likely to emerge. The example of telex service is instructive. As discussed
repeatedly in this book, in the 1980s, London-based telex bureaus started to
retransmit traffic between North America and continental Europe in defiance of
CCITT cartel "recommendations" against such retransmission. It was profit-
able for U.K. firms to break these rules, since this generated more traffic and
made the United Kingdom more attractive as a business location. In time, the
cartel rules were held to be illegal.
It is important to recognize that domestic intersectoral instability is linked to

international instability in regulation. A matrix describes the set of intersectoral
(vertical) and international (horizontal) regulatory cross-elasticities.
Generally, a stable solution is less likely as:

1. The sectors increasingly overlap (higher vertical cross-elasticities), for ex-
ample, due to "merging" technology. (Example: telecommunications, mass
media, computers.)
2. Jurisdictions increasingly interrelate (higher horizontal cross-elasticities),
for example, because of lower transportation costs.
3. The relative weight of interest groups in different jurisdictions varies,
as their economies develop on different trajectories. This can be a self-
feeding mechanism, as some countries become specialized in certain sec-
tors.
4. The regulatory price set diverges from cost, generating incentives for
breaching the set.
These changes lead to unstable situations that affect the entire system. A
single inconsistency has multiple secondary effects, which in turn lead to fur-
ther inconsistencies. At the same time, collaborative regulatory adjustments
become more difficult, because they cannot be confined to subsectors.
Outlook
Is there anything to stop the process? At some point, the cost of instability and
diversity becomes high enough for a coordinated regime to be reestablished.
But this is not likely to be stable over time, especially since no real interna-
tional enforcement or compensation mechanisms exist, and their absence has
domestic reverberations. For example, if international airline cartels break down,
domestic ones are threatened. One can fly from Toronto to Vancouver through
the United States, if Canadian prices are too high.
Applied to telecommunications, one should therefore expect an overall trend
toward liberalization, though accompanied by efforts to stabilize its collabo-
rative aspects. As the matrix of interrelations becomes steadily more cross-
elastic, one could have some oscillations. But the overall tendency should lead
to reduced regulatory strictness internationally. In that sense, liberalization is
an expansionary process. It may not be just an ideological choice but a re-
sponse to an internal inability to structure a stable equilibrium that serves mul-
tiple domestic interests and goals. One has to predict that similar inconsisten-
cies will spread throughout the system, even where they are riot based on the
same domestic pressure to change that led to the original instability in the other
jurisdiction.
Traditional telecommunications operated through national monopolies pro-
tected internationally by a cartel arrangement. Now, a challenge to domestic
monopolies threatens the international cartel, and the breakdown in interna-
tional arrangements threatens the domestic stability of others. It is difficult to
see how the simplicity of the traditional system can be maintained or restored.
Most likely we are merely at the beginning of a lengthy, dynamic, and untidy
process, of which the presently asymmetric liberalization across the two sides
of the Atlantic is a manifestation.
41
Regulatory Tasks for the Future:
Interconnectivity in the
Pluralist Network
Two basic forces shape change in today's telecommunications networks in Eu-

rope and elsewhere: the integrative forces of technology that push towards ISDN
and integrated broadband networks, raising barriers to entry; and the social and
economic forces of pluralism, which move the network toward a decentralized
and segmented federation of subnetworks. The tension between these forces is
most pronounced on the front where they intersect: the rules of interconnection
of the multiple hardware and software subnetworks and their access into the
integrated whole. Such interconnection and access extend traditional common
carrier principles from users to networks. In coming years, policymakers must
structure ways in which network interconnection is granted, defined, policed,
priced, and harmonized. Many questions must be resolved involving technical
standards, national uniformity, and international collaboration.
Regulatory structure in telecommunications has paralleled the stages of the
industry itself. The monopoly stage of industry structure was accompanied by
state ownership and price and profit regulation. The breach of monopoly was
tracked and sometimes facilitated by regulation focusing on industry structure.
In the next stage, the network is being rearranged from a centralized starlike
structure into a matrix of interconnected but decentralized networks. This moves
the focus of regulation to encompass not only traditional consumer protection,
but also network protection where necessary, mediating the interaction of car-
riers, network operators, value-added service providers, end users, and equip-
ment manufacturers.
Where does a pluralist network leave future policy? It would be naive to
expect fewer regulatory tasks. Many disputes become less intraorganizational
within the single network and more public in nature. The main regulatory tasks
that the pluralistic network system raises for the future would deal especially
with interconnectivity issuestechnical, legal, content, financial, social, inter-
national, and otherwise. They are characterized by inter words: interoperabil-
ity, interconnection, integration, intermedia, international. These issues include
the following:
1. Technical interconnectivity: protection of a balance between standardiza-
425
426 Regulatory Tasks for the Future
tion and diversity. The advantages of uniformity derive from greater industry
and service stability, and increased compatibility and portability between dif-
ferent hardware and software. Its primary disadvantages involve reduced free-
dom to innovate and experiment, and the loss of flexibility to adopt to changing
conditions. What is needed is a process to weigh and balance the various needs,
and a hierarchy of uniformity.
Experience and the theoretical economic literature on standards setting and
game theory show that standards do not necessarily evolve optimally, smoothly,
or speedily in a purely voluntary setting. This would suggest a governmental
role in interoperability, as long as one can prevent their use for protectionist
purposes. Europe's ETSI is an institutional response to that need.
2. Legal interconnectivity: protection of interconnection and access. The
tension between the integrative and pluralistic forces is most pronounced on the
front where they intersect: the rules of interconnection of the multiple hardware
and software subnetworks and their access into the integrated whole. Policy
must structure ways in which network interconnection is granted, defined, priced,
and harmonized. To control access arrangements is to control the network. This
is detailed in Chapter 42, where an open network system is discussed (NYPSC,
1989; NYPSC, 1990c).' The liberalization of terminal equipment interconnec-
tion and open network provision in Europe are steps in that direction. But they
are only the first steps in a long journey.
3. Content interconnectivity: regulatory treatment of telephone carriers in
their capacity as mass media. The upgrading of the telephone network toward
broadband capability and its use for video, data, and text transmission brings
telephone transmission ever closer to mass media. Audio mass announcement
services are already widely offered, and video broadband is a likely future
scenario for telecommunications carriers.
In the United States, there have been claims by network operators to possess
the status of "broadcasters" or "publishers" of information, or at least the
arbiter of what fits their organization's image, with the right to select the infor-
mation carried over their network. In Europe, these issues are beginning to
arise, and the television field itself is in the turmoil of change, as discussed in
the companion volume to this book (Noam, 1991c).
In the common law tradition, carriers and other businesses "affected with
the public interest" had an obligation to provide service to all indiscriminately.
They provided transport or transmission function, with no influence over or
liability for the content of transmissions, as common carriers. Common car-
riage is among the most important long-term issues in telecommunications pol-
icy. Common carriage is a form of content interconnection in an intermedia
environment. As such, it is a cousin of the other forms of interconnection.
What does common carriage mean? That is a surprisingly difficult question
for such a well-worn term.2 It is often used imprecisely as a synonym with
other terms. For example, it does not mean universal service, that is, the re-
quirement to serve anyone equally, regardless of geography. Nor is it a syn-
onym for monopoly, public utility, or regulated firm. It is best defined for our
purposes as a "carrier of electronic information offering service on demand to
Regulatory Tasks for the Future 427
the general public, for hire." Its attributes are neutrality among similar users
and usages, content neutrality, and no liability for content and user actions.
This is nothing new. For centuries under English common law, transportation
carriers had special obligations when dealing with the public, primarily the
responsibility to treat customers equally.
Early forms of common carriage date back at least to the Roman Empire,
which developed law on the duties of shipowners, innkeepers, and stable keep-
ers. Early English common law established rules on businesses that were con-
sidered "public callings." These included bakers, brewers, cab drivers, ferry-
men, innkeepers, millers, smiths, surgeons, tailors, and wharfingers (Phillips,
1988, p. 83). Common, in this sense, meant "open to public service."3 By
force of common law enforced by the courts, common carriers were required
to serve, upon reasonable demand, any and all who sought their services (Aust,
1990).
The concept of common carriage crossed the Atlantic and became part of the
American legal system. Common carriage was applied to railroads and later to
other distribution mediums. In 1901, the U.S. Supreme Court affirmed that
under common law a telegraph company is a common carrier and owes a duty
of nondiscrimination.4
In today's legal terms, common carrier obligations translate into nondiscrim-
ination and noninterference with content. What are the benefits of such a sys-
tem? In economic terms, common carriage is one of those arrangements whose
purpose is to reduce transaction costs, like limited liability, legal tender, or
negotiable instruments.
As with other efforts to balance private and public interests, common car-
riage is at times burdensome to the carrier. Yet it is probably fair to say that
the common carrier system has served telecommunications participants well: It
has permitted society to entrust its vital highways of information to governmen-
tal administrations or for-profit companies, without the specter of discrimina-
tion and censorship by government or private monopolies; it helped establish a
free flow of information; it reduced administrative costs and the burden of lia-
bility because the network operator did not need to concern itself with content;
and it protected the carriers from pressure groups who wanted to exclude tele-
phone service to some other groups or activities they opposed. This system has
probably resulted in a broader, more useful, lower cost, and more profitable
network system than would have developed without it.
Although common carrier principles go back a long time, their applications
are in a constantly shifting terrain and require continuous updating, particularly
in an extraordinarily dynamic period in telecommunications.
Perhaps the most important change has been the emergence of the telephone
network as a carrier of mass media messages, provided mostly by other infor-
mation suppliers. Examples include videotex, dial-it, mass announcement ser-
vices, interactive services, and future telephone broadband services, including
video. This raises the question of the regulatory status of such a new mass
medium. Here it is helpful to understand what the options are. Mass media are
basically treated in one of three ways. The print media is virtually unregulated;
in the United States and Europe, free speech governs. The main restrictions are
the prohibition of libel, the protection of privacy, and some protection from
obscenity. Other countries have similar, though usually somewhat more restric-
tive, free-speech protection. Private broadcasters and cable operators are treated
somewhat differently. They require licenses and franchises to operate and are
subject to certain obligations, such as, in the United States, the fairness doc-
trine, the must-carry rule, and so on. They still enjoy significant free-speech
rights; in particular, they are able, to a large extent, to determine what will be
shown on their medium. The third type of treatment is common carriage. This
is the traditional way for telephone carriers to operate. That is, any user, and
any form of content, as long as it is lawful and as long as it pays the regular
price, will be carried over the telephone network.
The question now arises, how should the telephone carriers' increasing mass
media role be handled: Should it become something akin to a second cable
television system (i.e., with the powers of selecting and programming many of
its channels and excluding others)? This would give an unprecedented gate-
keeper role over mass media to private or governmental organizations with
considerable market power. Instead, telephone organizations' role as mass me-
dia carriers should be governed by the conduit and nondiscrimination principles
of common carriage, which is the traditional way that telephone carriers have
provided services in the past.5
It is important to distinguish conduit, or transmission, functions from content
functions involving information provision or modification. This is the most ba-
sic way to distinguish common carriage from publishing.
Whether or not telephone carriers offer program services does not affect the
validity of common carriage principles. Common carriage applies only to a
communication carrier's provision of conduit services. If a conduit provider is
also permitted to offer content, the basic principles of common carriage are not
reduced; On the contrary, they are even more important, because otherwise the
carrier could keep its content competitors out, or discriminate against them.
There is a simple way to summarize common carriage. All electrons and
photons are created equal. (In 1990, the New York State Public Service Com-
mission established the first U.S. rules to this effect, drafted by the author,
denying telephone carriers censorship rights for lawful messages, or the ability
to institute discriminatory access arrangements for information providers or lis-
teners.) In the United Kingdom, on the other hand, under traditionalist pres-
sure, the regulatory agency Oftel established a censorship board and other forms
of control. Most countries also permit or require various forms of content dis-
crimination.
4. Financial Interconnectivity: Protecting the Viability of the Core Network.
Clearly, an expanded, upgraded, and enhanced network utilizing the best tech-
nology requires significant capital investment. A competitive network system
may drive prices not only to costs but to marginal costs, making it impossible
to cover fixed costs, let alone to have monopoly profits to support upgrades.
Furthermore, as discussed, there may be a problem of investing in a critical
mass (i.e., in generating externalities) when the risk is borne by the initiating
network but the rewards are shared by others through the mechanism of inter-
connection. Finally, a technologically dynamic environment means shorter eco-
nomic lives for equipment, requiring faster amortization. It is therefore a chal-
lenge for future policy to structure arrangements such that adequate investments
in the network are available in a decentralized environment.
5. Social interconnectivity: establishing new mechanisms of redistribution.
The pluralistic network will make it increasingly difficult to maintain the tra-
ditional system of internal transfers from one class of users to another. But this
does not spell the end of transfers as such. There is still ample possibility and
opportunity to subsidize some categories of service for reasons of social policy
or regional development, or for the positive benefits that new subscribers pro-
vide to existing users. One type of support mechanism is a "lifeline" program
to insure access by the poor.6
A new mechanism to finance desired subsidies and end the primary reliance
on the PTTs and their customers would be to establish a "universal service
fund" with contributions by all network providers, instead of the present sys-
tem of internal transfers. Such funds could be raised through various access
charge mechanisms, or, more neutrally and efficiently, through a value-added
tax on communication services.
6. Informational interconnectivity: privacy protection. The new telecommu-
nications environment raises privacy problems. Privacy issues, in effect, are
the flow of information beyond the intended recipient. Privacy protection in
telecommunications is not a new issue. In the past, manual operators and party
lines all created their own problems. The first patent for a voice scrambling
device was issued in 1881, only five years after the invention of the telephone.7
Relatively strong expectations of privacy developed in time. Today a new gen-
eration of privacy problems has emerged. The reasons include:
a. More and more transactions are conducted electronically.8
b. It has become easier and cheaper to collect, store, match, and redistribute
information about transactions and individuals.9
c. The number of carriers and service providers has grown enormously, leading
to an increasingly open network system in which information about use
and user is exchanged across companies.
d. Transmission conduits increasingly include unsecured portionsfor ex-
ample, because of mobile communications.
Concern with electronic privacy has led to different policy approaches. In Western
Europe, attention to privacy issues has led to data protection laws and the
establishment of institutionalized boards and commissions that have often im-
posed fairly rigorous restrictions on data collection and data flows. This is
discussed in Chapter 37 on transborder data flows. In the United States, the
approach has instead been based on a variety of ad hoc federal and state legis-
lation.
The policy challenge is to structure a more flexible approach: a framework
in which the user would have several hardware and software "privacy options"
of service.
A competitive environment may resolve some privacy issues, especially if it

is possible for a user to select a service provider that offers the desired level of
privacy protection. But as mentioned before, the greater openness of a compet-
itive system and the greater complexities of its multiple and networks may also
mean a greater openness of information (NYPSC, 1990a, Noam 199la). In the
absence of protection it is likely that advocates of the traditional system
will argue that a monopoly is required for privacy protection.
7. Service quality interconnectivity: monitoring and maintaining service
quality. Quality issues are becoming important with price regulation. There are
more incentives to cut corners as a hidden way of increasing prices. As society
becomes ever more dependent on communications streams, it may reach a point
of great vulnerability because of a technical failure or intentional harm.
Furthermore, the network system is increasingly nontransparent to end users.
In a chain of several carriers that are strung together for a transmission path,
which one causes faulty transmission? This difficulty in identifying the culprit
could encourage "free riding" by some companies (Noam, 1991b). Also, there
may be selective deterioration (red-lining). As PTTs become entities that are
more driven by profit motives than in the past, these issues will become im-
portant.
8. Infrastructure interconnectivity: the role of telecommunications policy as
economic development policy. The competitiveness of an economy is affected
by the state of its communications. Many European countries are active in
using telecommunications as a strategic tool for technological development.
Planning horizons in telecommunications are very long and suggest the need
to facilitate coordination among carriers, where governments assume infrastruc-
ture responsibility. Depending on the success of such policies, the U.S. may
embark on a similar course, as its recent science network, NREN, demon-
strates.
9. Competitive interconnectivity: the prevention of oligopolistic behavior and
of cyclical instability. A decentralized and pluralistic network system is less
efficient in terms of minimizing resources, and there is likely to be excess
capacity. There is nothing unusual about this. Almost every industry has excess
productive capacity, and the competitive effect is usually beneficial for custom-
ers. In the telecommunications field, with its low marginal costs, we must
expect that the competition will cause periodic price instability. One of the
functions of future regulation will therefore be to moderate the worst effects of
price volatility and at the same time prevent industry efforts at collusion. In
the future, volatility may be most pronounced in international communications.
10. Interconnectivity of planning information. A decentralized network sys-
tem will have much less of a comprehensive planning role than in the past. To
planners and engineers this may seem heresy, but most markets for products
and services have no long-range, industry-wide plan. However, since telecom-
munications are characterized by interconnectivity issues and large-scale in-
vestments, it is useful to establish clearinghouses of information about network
deployments. That is, there may be a need for some form of disclosure.
11. International interconnectivity: establishment of global arrangements to
match the global scope of networks. The openness of the evolving network
system will not stop at the national frontiers. In the long run, telecommunica-
tions will move beyond the territorial concept, and the notion of each country
having full territorial control over electronic communications will become
anachronistic. Communications are becoming distance insensitive. Rerouting,
arbitrage, and the establishment of communications "havens" become easily
possible. This undermines attempts to set rules administratively for prices and
service conditions. No country can be a telecommunications island anymore.
The more interrelated countries and economic activities are, the less likely it is
that there are stable solutions to separate policies. Moreover, instabilities in
one country affect the entire system. It becomes increasingly difficult to control
all the elements in such a complex matrix of interrelations (See Chapter 40).
This means conceding a greater role to international bodies, while recognizing
their limitations, and similarly conceding a greater role to the coordination powers
of markets.
Conclusion
These issues will, no doubt, lead to significant regulatory controversies and

will occupy policy discussions in the United States, Europe, and elsewhere for
a long time. None of the tasks is beyond the grasp of policy makers in terms
of complexity or political feasibility. But it requires an end to the nostalgia for
the simplicity of the golden age, a vision of a very different network environ-
ment, and the willingness to engage in an analysis that goes beyond that of
competition versus monopoly, because most future issues cannot be analyzed
in such simple terms.
42
The Pluralist Network
The New Network Coalitions

In the first part of this book, we analyzed the incentives and forces for centrif-
ugalism in the previously all-encompassing network. We then turned to the
various countries and issues to observe these centrifugal tendencies. Now, we
close the circle and return to the nature of change. The incentives to group
formation, discussed in the model of Chapter 3, can lead, where they are per-
mitted to do so legally, to new "network associations." It is a process that we
called earlier the "tragedy of the common network," because it is not its fail-
ing but rather its very success which undermined the public network. As we
observed, the success of a communal network creates overexpansion, cost shift-
ing, and the forces for particularism. For example, the largest of users increase
their electronic communications to huge volumes, and their technical require-
ments are increasingly differentiated from those of average users. Because the
combined volume of large users has risen so much, they can account for much
of the cost savings of sharing just between themselves. They can form alter-
native network associations for large parts of their communications needs, first
in-house, then with their closest suppliers, customers, or market partners.
Examples are large private intraorganization networks, shared tenant ser-
vices, local area networks, wide area networks, and other specialized services.
These groupings of users need not be territorial. The notion of telecommu-
nications as consisting of interconnected national systems is likely to be tran-
scended in many instances, and specialized transnational networks will emerge.
This becomes possible with the drop in cost of international circuits.
For satellite transmission, in particular, the marginal cost with respect to
distance is close to zero. Fiber-optic links also have lower distance-sensitive
costs. The implications are that communication flows can be routed in indirect
ways to exit previous shared arrangements or to join new and more congenial
ones. Arbitrage becomes easily possible. This undermines attempts to set rules
administratively for prices and service conditions.
In the future it is likely that specialized global networks will emerge for a
variety of groups that communicate with each other intensely. Their relation to
each other is functional rather than territorial, and they can create global clus-
tering of economically interrelated activities in much the way that, in the past,
related activities clustered physically near each other.
432
The Pluralist Network 433
This has begun with the private networks of large users. These private net-
works, some of whose operation and administration require hundreds of skilled
technicians and managers, even where they usually rely on the infrastructure of
the regular carriers, have emerged in developed countries. These activities are
spearheaded by private firms, but are pursued also by nonprofit and public
organizations. Similarly, users with common interests and activities have begun
to interlink through private value-added networks such as SWIFT.
Examples for actual or potential group networks are those of (1) advertising
agencies linked to media firms and printers; (2) chemical manufacturers and
environmental protection agencies; (3) insurance firms, hospitals, record rooms,
police departments, and so on; or (4) of automobile manufacturers, dealers,
suppliers, and shippers.
In some instances these networks will have special performance features that
distinguish them from the general "public" network. In the first example just
cited, network bandwidth must be high to permit transfer of high-resolution
graphics. In the second example, the usefulness of the network may be in
supporting software and databases. In many instances, such as the third ex-
ample, it is probably the price of intercommunications that drives the arrange-
ment. In the fourth example, the standardization of transactions may be the
primary advantage.
Most users are likely to participate in several networks. Furthermore, such a
pluralist network arrangement does not mean separate transmission links for
each subnetwork at every point. It will often make sense to transport the traffic
of several low-volume users part of the way on the general network until the
point where there is enough aggregate traffic to branch off. The economics of
sharing are not abolished. But they must prove to be superior as a matter of
choice, not imposed by a legal requirement. Hybrid arrangements like virtual
networks are therefore likely to be popular.
The earlier observations suggest that no specific event undermines a Euro-
pean PTT's role; instead, the aggregate of challenges will lead to the "tipping
of the monopoly." Restrictions in one area are interconnected with others; con-
versely, liberalization in one field is hard to contain. Thus, equipment liberali-
zation leads to the demand for new services, such as value-added networks,
which in turn creates pressures for resale opportunities. It is unlikely, given the
nature of PTT organizations, rapid technological developments, and resultant
market pressures, that they will be able to put out every fire, man every ram-
part, and seal every dike successfully. At the same time, a partial liberalization
also provides opportunities for the traditional telecommunications authorities,
and they may seize them for the temporary advantage, though change will
undermine their exclusivity in the long term.
The conclusion is that European telecommunications will, in the short run,
experience a push by PTTs to extend and defend their monopoly position and
to expand in the direction of new services and equipment operations. However,
long-term forces are at work that are likely to lead in time to an unraveling of
the monopoly system. These factors, in combination, make it unlikely that the
PTTs will maintain the present extent of their monopoly control.
434 The Pluralist Network
Liberalization will first be granted for value-added networks and customer

terminal equipment because these are the least controversial. Once the genie is
out of its bottle and different service options are available, all kinds of new
incentives for entry will be created. In those areas where monopoly service is
not necessary or preferablethat is, where neither social policy reasons nor the
strong economies of scale of a natural monopoly existsome form of compe-
tition will emerge. This will leave for a while a natural monopoly core under
the control of the PIT and its affiliated system. Such a system may be termed
creeping liberalization.
Thus, the days of the PTTsin their present manifestation as PTOs, public
telecommunications operatorsare in no way numbered. Just as AT&T, even
after its massive divestiture, is still a major supplier of long-distance commu-
nications services and equipment, so will the PTTs maintain a formidable pres-
ence. Indeed, they will be able, for a time, to leverage their formidable re-
sources and market power in core markets. They will become more independent,
more powerful, less controllable, and less socially oriented. But in time they
will lose their exclusivity and a more pluralist network system will emerge.
This will require the development of alternative forms for the PTTs' cross-
subsidies to support local exchange traffic, residential users, and the rural pop-
ulation. But the system will no doubt be able to adapt. In the end, large PTT
organizations will still be linked with powerful groups in society, but without
the status of a monopoly. European countries will gradually discover that a
monopoly in telecommunications, controlled by PTTs, is not a prerequisite to
a technically efficient and socially equitable communications system.
The new system will considerably lower terminal and central office equip-
ment prices and reduce the profitability of the postal-industrial coalition. The
equipment market will become much more open to foreign manufacturers and
to members of the "second" electronics industry. Traditional telecommunica-
tions firms will accelerate their rate of innovation. Equipment itself will be-
come unbundled, modular, and specialized, so that multiple suppliers can seek
their niches. The conventional wisdom that there is room for only a handful of
switch manufacturers worldwide will prove as nearsighted as the "mainframe
thinking" in the computer industry of the 1960s. Over time, developments will
push telecommunications rates toward cost, particularly in the highly profitable
international services. These prices will be unstable, since the excess capacity
will lead to periodic price wars, and they will be differentiated among routes
of different traffic density and competitiveness.
As this process of normalization takes place, those identified with the tradi-
tional system, who are rightly proud of its technical and social accomplish-
ment, will defend it as best as they can. The transition will, therefore, be a
difficult one, but it is unavoidable.
We are merely at the beginning of a lengthy process of change in the net-
work: The centrifugal forces discussed earlier are encouraging the evolution of
a new network model of telecommunications characterized by a great deal of
openness, transforming its structure from a centralized star into an intercon-
necting matrix as noted above. The new structure has various forms of open-
ness: of entry (be it a carrier, specialized service provider, or equipment ven-

dor), of interconnection into other networks, of access to other networks, and
of standards. What are the main characteristics of this open network system?
Network of Networks
The network system of the future is one of great institutional, technical, and
legal complexity. The network environment will consist of an untidy patchwork
of many subnetworks, serving different geographical regions, customer types,
user groups, categories, and service, with no neat classification or compartmen-
talization possible. It includes a hodgepodge of participants, governmental and
private, national and regional, general and specialized, narrow and wideband,
terrestrial and satellite, tiny and vast, domestic and multinational. The Ameri-
can experience demonstrates the long-range instability of structural regulation,
such as compartmentalization of the telecommunications sector along different
functional and geographic dimensions and the assignment of market segments
to different carriers. This is due to the overlapping and ever-changing nature of
services, the inability to define clear boundary lines, and the incentives for
participants to breach restrictions. Hence, the future telecommunications net-
work environment will have carriers engaged in multiple functions, though there
will be no shortage of official attempts to establish order.
Substantial Reduction of Central Control

The primary characteristic of the open network model is a substantial reduction
of central control, though with a significant PIT role and some military plan-
ning functions. The network becomes a composite of numerous separate plan-
ning decisions. This notion is alien to the engineering world-view of tradition-
alists in telecommunications. To leave an integrated system to the vagaries of
hundreds of uncoordinated and selfish actors seems to invite disaster. Can it
work? Perhaps this is not the right way to frame the question. Can there be a
stable alternative in economies that otherwise favor a market mechanism and
that want to stay on the leading edge of technology and applications?
Nonsustainability of Most Regulation

Telecommunications is in the process of moving from one of the most regulated
industries to one of the least regulated. One reason is that the growing com-
plexity of the system makes it increasingly difficult to fashion consistent rules,
whether behavioral or structural. The American experience with the FCC's
Computer Inquiry decisions gives an early taste of this difficulty.
In addition, rules are not likely to be enforceable. The subject of the regu-
lationstreams of electrons and photons, and patterns of signals that constitute
informationis so elusive in physical or even conceptual terms, and at the
same time so fast and distance insensitive that a regulatory mechanism, to be
effective, must be draconian, and for that the traditional system has neither the
will nor the political support. Yet there is need for regulatory oversight of the
rules under which networks and users interrelate in the future, as detailed next.
To bridge this tension will be one of the central challenges for future policy.
The Public System as Core

The telecommunications system will evolve into a mixed public-private ar-
rangement. The public network will not cease to exist. It is likely to remain
for a long time the core of the system and its prime standard setter, with public
support for the same reasons that have existed before, but without the exclusiv-
ity that characterized it for over a century. This is comparable to the situation
prevailing in transportation. A state railroad system exists in most industrialized
countries, often subsidized directly and indirectly, but it is supplemented by a
mixture of trucking firms, airlines, barges, passenger automobiles, and small
railroads. No one advocates a transportation system that bans all private trucks
just because they reduce the scope and revenues of public railroads.
Gradualism
As the earlier discussion of interconnection of old and new networks demon-
strates, it is not necessary to cut one's links to other participants. Multiple
networks need not lead to multiple telephone sets on the desk. This is not more
necessary than having multiple currencies if there are several banks in a coun-
try. In the situation of multiple networks, the functioning of the network fed-
eration depends on rules of interconnection. Such rules are being structured in
regulatory approaches of open network architecture (ONA) in the United States
and of open network provision (ONP) in Europe. Such arrangements provide
additional impetus for new networks in that they permit the gradual establish-
ment of a new network. They also permit a wide range of options between a
full and separate physical network, and the use of an existing network in dif-
ferent and innovative ways, somewhat apart from the general "public" user-
ship.
The concept of gradualism inherent in open network architecture is impor-
tant. Many advocates of the traditional shared network system believe that the
demands of pluralism could be met by a flexibility for software network op-
tions, without altering the exclusivity of the physical network of the traditional
monopolist. This is wishful thinking. Granted, to permit software networks on
a transmission monopoly is a correct first response to the emerging pressures.
However, it is unlikely to be adequate in the long run. Soon users will want to
supplement transmission offerings with additions to satisfy their preferences in
terms of technology, control, and economics. The exclusive network cannot be
the superior solution in each instance, particularly if it has to follow political
mandates, or if it cannot bargain individually on prices.
Expansion of the Public System
The traditional public system may be losing its exclusivity, but it is gaining the
flexibility of moving into new activities. The experience of British Telecom,
Sweden's Televerket, the semi-independent public system of the STET group
in Italy, and of the Telefonica group in Spain indicates strong tendencies of
dominant carriers to expand vertically into equipment and manufacturing, even
internationally. These tendencies are similar to the developments in the United
States, where the scope of activity of the AT&T successor companies is stead-
ily increasing. Such new horizons are an attraction to the PTTs as they consent
to the loss of monopoly; for policymakers, they raise regulatory issues on how
to deal with the substantial economic power of independent PTTs now renamed
PTOs, but with market power intact.
Universal Service: Narrower but Deeper
The traditional public network system operated with the obligation of universal
service, that is, virtually any interested customer had to be served, regardless
of location. In a pluralist network system, the question is whether universal
service obligations apply to all participants. The answer is likely to be differ-
entiated. For some of the more specialized services, the obligation will not
exist. But for "basic" service, it will continue, and the definition of basic is
likely to expand. The boundary line is likely to be an ongoing issue of policy
discussion.
Internal Subsidies
In a pluralist network system it is unlikely that the traditional system of internal

transfers from one class of users to others can be maintained. But this does not
spell their end. There is still ample reason and opportunity to subsidize some
categories of service and of user classes, just as in the case of railroads. Ser-
vices for that subsidy can be raised and distributed in the normal way of taxa-
tion and budget allocation in which income redistribution takes place in society.
It is incorrect to consider a monopoly as a necessary condition for redistribu-
tion. Justification of the subsidy exists for reasons of general social policy or
regional development, and because of the positive externalities that new sub-
scribers provide to previous subscribers. Nevertheless, the extent of the subsi-
dies could be smaller once their scope is open to scrutiny. As a result, they are
likely to become more targeted toward the truly needy (i.e., the poor), and
away from the middle class. The remainder of residential users will be forced
to pay more than before and will also increase their usage of the telephone as
its functions grow, making telecommunications a larger budget item than in the
past.
Quality and Price Differentiation

There will be more choice but less equity. Whereas in the past all subscribers
had a fairly similar quality of telephone service and equipment, an open net-
work system will have much variation, depending on the preferences of cus-
tomers and their willingness and ability to pay. There will also be a much
greater differentiation in the cost of communications. Just as two adjoining
passengers in an airplane may have paid widely different prices for their ticket,
so will telecommunications users pay different rates for similar service. Those
with small usage and few alternatives will pay more per volume than large
users, and more than alert consumers. This reverses the traditional price rela-
tionship of business users subsidizing residential and rural service, yet it does
not necessarily point to the end of universal telephone penetration. Given the
low elasticity of demand, the increasing importance of telecommunications,
and new ways to raise and distribute subsidies, the drop-off of subscribers is
not likely to be large, despite some negative redistributory effects.
Transnationalism
The traditional centralized system was directly mirrored internationally. It held

together well because of a similarity in viewsthe values of engineering and
bureaucracyand because of a common interest in protecting the domestic ar-
rangements. For a long time, national PTT administrations almost joyfully par-
ticipated in the international sphere, because they could return home with an
international agreement that would buttress their domestic position. But in the
age of satellites, internationalism became a threat, because it was initiated by
users and new carriers. International communications are the most vunerable
part of the domestic service monopoly. In the long run, telecommunications
will go beyond the territorial concept. Communications are becoming too var-
ied, complex, and significant for one organization to do it all well. Similarly,
the notion that, in the age of information, all communications flows in societies
operating largely on the market principle would pass through one streamlined
super pipe controlled by a single organization is hard to entertain on technical,
economic, or political grounds, except by reference to the present balance of
power.
Once the notions of the centralized network are breached in some respects,
the process is hard to contain. This development is inevitable not because it
leads necessarily to a superior result, but because the traditional network is an
anomaly once it moves beyond its cost-sharing and redistributory stagesthough
it is an anomaly too familiar to notice. As long as the economic system of
Western industrialized democracies is based on markets and private firms, the
exclusion of major economic parties from a major field is an unstable affair. It
is hard to maintain a dichotomy between telecommunications and the rest of
the economy. To differentiate it as an infrastructure service is too vague to be
useful. The traditional public network was an appealing idea. It incorporated a
notion of sharing which reached every member of society. Yet these are also
the concepts of authoritarianism. The historical origin of the system, rooted as

it is in seventeenth-century absolutism, does not support those who view its
defense as a progressive act. The telecommunications field will more closely
resemble the rest of the economic system. It may be much more complex, and
in some aspects less resource-efficient than the old system, but it will be a truer
reflection of the underlying pluralist society.
Notes
Preface
1. There are, of course, semiofficial volumes in which various countries' PTTs report
about themselves, but these reports are neither analytical nor probing.
2. The Establishment of the PIT System
1. From Hansard's Parliamentary Debates, March, 1895, p. 214, in Meyer, 1907,

p. 117.
3. Network Tipping: The Rise and Fall of the Public Network Monopoly
1. GalbraithInterview with Francis Cairncross, The Observer, November 22, 1970.

2. One attempt was a U.S. Department of Justice report on the postdivestiture net-
work (Huber, 1987). Another approach is that of Koichiro Hayashi of NTT (1988). An
earlier version of the present model was Eli Noam, The Next Stage in Telecommunica-
tions Evolution: The Pluralistic Network, paper presented at the ITS Conference, June
1988.
3. Exceptions were CoCom restrictions to East bloc countries that were fading, to-
gether with the latter.
4. It should be noted that this model can be easily adopted for "standards coalitions"
rather than for "network coalitions." For the literature on standards, see David (1987).
5. I will follow the network analysis as developed in Noam (1988). Also see Heal
(1989).
6. The case of n assumes that users are homogeneous. Of course, some network
participants are much larger than others, but that poses no problem if we define a large
organization to consist of multiple members of type n (e.g., telephone lines or termi-
nals). Later, we will drop this assumption.
7. Strictly speaking, income is allocated to telephone and to other consumption
y = cp + C(n)ln.
A subscriber's utility is then given by
u {(yip - 1C (n)tn]p; n}
8. For convexity, assume u(c, P, 1) > u(c, P, 0) (i.e., the first user has positive
benefits even if no one else is on the network).
441
442 Notes
9. For simplicity, utility is expressed in monetary units. Then u(P,n) = P(n) + u(n)
10. Heal defines it, similarly, as
u <y'P~~[f + f(n)]/np: n) = u(y/p; 0)

See also Allen (1988).
11. This analysis should not suggest that a voting mechanism is governing in reality
(although it exists for telephone cooperatives in Finland and the United States); rather,
it helps us to understand the pressures and dynamics that are transmitted to the govern-
mental institutions that embody the different user interests.
12. For simplicity, we use in the following n/2 rather than (n/2) 1.
13. A very similar analysis can be undertaken for a political system in which network
outsiders have a vote, as in a governmentally directed system.
14. I owe the terminology of "tipping" to Schelling (1978).
15. This corresponds to reality; new networks such as those of MCI and Sprint have
lost money for a number of years until turning profitable with growth.
16. The results discussed would not hold if the marginal costs of new network par-
ticipants drop continuously more than their marginal benefit to an existing network user.
The latter is unlikely, since marginal cost, beyond a certain range, is either flat or very
slowly decreasing, or in fact increasing.
4. Forces of Centrifugal ism
1. The author served as a member of the Advisory Board for the selection.
2. NYNEX data, see Noam (forthcoming).
3. New York PSC data, see Noam (forthcoming).
4. Encompasses all U.S. carriers; translated for new telephones from data on access
line using 1975 ratio. Sources: Telecom Factbook, 1986; FCC, Statistics of Commu-
nications Common Carriers, 1945, 1955, 1965, 1975, 1985.
5. This classification extends those of Ergas (1989).
5. Defense of the Telecommunications Monopoly
1. Residential penetration is not the entire story. In public (i.e., coin) telephones of
several European countries, Switzerland had the highest density (8.90 per 1000 popu-
lation). Densities are substantially lower in the rest of Europe. In the United States, the
corresponding density is 7.18, to which figure one must add the many private coin
telephone networks that have mushroomed since liberalization of this sector. In New
York State in 1988, privately owned coin telephones amounted to an additional 20
percent of those operated by the telephone companies, suggesting a density of 8.6 per
1000 population. The OECD countries apart from Switzerland and Japan have substan-
tially lower densities. See Table 32.2.
7. Germany
1. In a letter to a colleague in both the party and the Reichspost, Ohnesorge rhapso-
dized in 1931 about Hitler, then still in the opposition:
Notes 443
Dear God, what luck that the country has been blessed with a man of such stature!
. . . He is the only, the last great German! And everything is right as he approaches it
and as he wishes it. ... Haase, dear Haase, when finally the marching steps of the
battalions will sound again through the streets of Berlin! Both of us are still young!
Then we will march, too, unyielding, unrelenting to wherever Adolf Hitler wants [Na-
gel, 1937, p. 24].
Such sentiments, expressed also publicly, had not slowed Ohnesorge's career in the
Reichspost during the Weimar Republic; he was promoted to considerable responsibil-
ity, including being in charge of the Reichspost's priority project of transition to auto-
matic exchanges.
2. In 1972, Germany's high-technology exports had been a respectable 26.3 percent
of the six major countries. However, by 1983, this share had fallen to 17 percent. The
French share in the same period declined from 11.1 percent to 8 percent, Britain's share
from 13.8 percent to 10 percent, and Switzerland's share from 3.5 percent to 3 percent.
In contrast, the non-European countries improved their situation significantly. Japan's
market share almost doubled from 13 percent to 25 percent, and that of the United States
rose from 32.2 percent to 37 percent, amounting to more than one-third of the total.
Furthermore, the market as a whole increased sixfold, from $8.4 billion in 1972 to $54
billion in 1983. In telecommunications, Germany's share fell from 22.8 percent to 14.5
percent: in electronic tubes and transistors, from 17.4 percent to 11 percent (Bundes-
bank, Annual Report, 1983).
3. It also invested extensively in America and acquired 20 percent of the American
company Advanced Micro Devices (AMD), a company that second-sourced Intel and
Zilog microprocessors. It also bought 100 percent of Microwave Semiconductor, 80
percent of Litronix, and 60 percent of Advanced Microcromputer (AMC). Through Corning
Glass it established the successful optical fiber venture Siecor and acquired the floppy
disk operations of Perkins-Elmer. Siemens also formed joint ventures with Intel and
Westinghouse, in 1988.
4. Today the Bundespost's record of service in the rural population is often used to
support the argument for its monopoly. According to the Bundespost, the Bell System
in the United States concentrated on the lucrative urban areas to the disadvantage of less
profitable and more sparsely populated areas: "A telephone connection in structurally
weak areas is either not offered at all, or even today is offered more expensively and
maintained or provided technically less accomplished than in urban regions" (DBF,
1981, para. 96a). In fact, rural telephony in the United States is often cheaper than an
urban one. In typical farm states such as Iowa and Kansas, the percentage of house-
holds having telephones in 1987 was even higher (95.1 percent and 95.2 percent,
respectively) than the national average of 92.4 percent (Federal-State Joint-Board,
1988), and a high 96 percent of all American farms had telephone service in the
same year (Rural Electrification Administration, 1988). This was despite the signifi-
cantly lower density of population, particularly in the American West, as compared to
that in Europe. Population density in the state of Wyoming is 12.4 inhabitants per km 2 .
In comparison, density in Germany is 246 inhabitants per km 2 ; in the United King-
dom it is 233 per km 2 ; and in France, 100 per km 2 . In some instances, the technical
quality of independent rural telephony has not been as high as that in urban areas,
but it is often better, too. In 1985, the average residential flat rate for unlimited local-
call service was $13.80 per month in urban U.S. areas and $10.15 per month in rural
areas.
5. The Bundespost planned for a natural evolution from ISDN to broadband ISDN,
444 Notes
with a structure suitable for higher-speed applications like television or video telephony
(Ohmann and Armbruster, 1986; Armbruster, 1986).
6. A study undertaken by the Deutsches Institut fiir Wirtschaftsforschung (DIW) in
1980 found that for each additional billion DM in investment in telecommunications
beyond the existing 10 billion, 21,000 to 22,000 jobs would be generated if such in-
vestment continued (Schnoring, 1982). However, another study, commissioned by the
Ministry of Economics in 1983 and carried out by the IFO Institute and looking at the
period 1983-1990, found that an additional billion DM a year in telecommunications
investment would result in only 4000 additional jobs per year. This is not an impressive
figure; it means that for each new job, a new and ongoing investment level of DM
250,000 ($134,000) per year would have to be maintained.
7. To ascertain the efficacy of this system, the FCC also asked that AT&T and eleven
public utility commissions determine whether network damage has occurred. They con-
cluded that there had been no negative effects on the network or physical impact on
personnel.
8. The interest of private households and business customers in various new features
of telephone service and equipment grew dramatically. There was a demand, approach-
ing two-thirds of all businesses and residences, for detailed specification of billings.
Such billings are not normally provided (Schulte, 1982, p. 322, Table 12B).
Another common U.S. service, call-forwarding, began to be introduced in Germany
under the name of GEDAN only in the 1980s. But the basic rate for this service was
between DM 133 and DM 160 ($71 and $85) per month. Cost for call forwarding in
New York is, for businesses, $6 per month after an initial installation fee of $54; for
residences it is $4 per month after an initial fee of $15.50. (Frequently, the initial
installation rate is cut or dropped to attract usage.)
9. It should be noted, however, that although Wagner advocated public ownership of
the railroads as essential, another eminent German institutionalist economist, E. Sax,
argued that a form of private enterprise could fulfill public tasks just as well (Sax,
1887). Wagner's views, however, won in the political sphere.
8. The United Kingdom

1. Carsberg was knighted in 1989 for his service, and promoted in 1992 to head the
Fair Trade Office (FTO).
2. In the first year after being granted its license, BT applied for price changes. The
inflation rate for the period had been 7 percent, making the average permissible price
increase 4 percent. BT increased the price of local calls by 6.4 percent after allowing
for changes in the categories of a calling unit. Coupled with increases for five types of
long distance calls and reductions for four, the weighted average price increase was 3.7
percent, slightly below the allowable rate. Among residential customers, rates increased
by 7.1 percent for low-volume users, 6.3 percent for moderate users, and 5.7 percent
for high-volume users. For business users on average, however, rate increases were
lower. To minimize rate increases for the lightest business users, BT instituted a low-
user rebate in 1983. For moderately high business users, rates went up by about 2
percent (Oftel, 1985a).
3. In 1988, Oftel granted six satellite licenses to Robert Maxwell; British Satellite
Broadcasting (BSB); EDS U.K.; British Aerospace; Satellite Information Services; and
Uplink Ltd. The firms were allowed to offer point-to-multipoint services, but were barred
from two-way communication (Hayes, 1988).
Notes 445
9. France
1. Although the DGT installs and maintains customer equipment, authorized inde-
pendent installers and their subcontractors can also provide such services. There are
about 500 primary installer authorizations in the country, employing about 15,000 peo-
ple (Voge, 1986).
2. The digitalization of central office switches permitted the introduction in 1983 of
itemized billing (at a separate charge), conference calls, call forwarding, free-phone
(800 service), and call waiting (Bruce et al., 1985).
3. When the agreement expired in 1990, Thomson won a France Telecom contract
for broadband switches and CGE's Alcatel made defense electronics one focus of a new
division (Boult, 1990, p. 10).
4. The reorganization was also believed to help French telecommunications exports
by avoiding head-on competition between two suppliers for international contracts. For
example, in the case of major Egyptian telecommunications contract, CIT-Alcatel bid
against Thomson and Siemens. The merger eliminated the potential for such intra-French
competition.
5. In the United States, Alcatel's sales efforts began in 1984 and resulted in orders
for several dozen of a U.S. switch version, the E-10-5, primarily for small independent
telephone companies, such as the Clay County Rural Telephone Cooperative in Indiana,
rather than to the Bell Operating Companies (Telephony, 1985, pp. 32-33). For equip-
ment other than local exchanges, Alcatel's customers in the United States were MCI
and AT&T for multiplexing systems. Having adapted ITT's System 12 switch, Alcatel
reentered an American central exchange market that ITT had abandoned in 1986 when
adapting the System 12 proved too expensive and time-consuming.
In 1990 Alcatel pulled many of its operations out of the United States except for
cable television and SONET transmission equipment development (Roussel, 1990d).
10. The Netherlands
1. Until 1987, when the zones were redrawn, the small call areaswith an average
of 31 square kilometers versus 2673 square kilometers in the United Kingdom (Dabbs
et al., 1982)allowed for untimed local calls. After 1987, only two charging zones
remained. Areas for local and long-distance calls were extended and local calls became
timed.
2. The PTT is required to meet performance standards in "reserved" services and
has lowered rates for several services since the reform,
3. To enhance its position as the dominant supplier in the newly liberalized value-
added services market, PTT Telecom acquired 45 percent of a Dutch networking firm,
Impact Automation, and began negotiations with France Telecom and the DBP Telekom
about an international consortium (Schenker, 1989, p. 8).
11. Belgium
I . Although it opposed the loosening of its own monopoly, the RTT was not above
benefiting from the opportunities presented by the liberalization of others. After an
agreement between the U.S. carrier MCI and the Belgian Post Office, the PTT became
the major entry point to Europe for the American alternative long-distance carrier.
446 Notes
2. In Turkey, BTMC has a particularly large contract, which includes the transfer of
technology and manufacturing expertise to the Turkish equipment producer Teletas in a
turnkey arrangement.
3. Also, under its guidelines for the licensing of VANs providers, RTT granted ap-
proval to five service providers: GEIS, Sabena, SITA, SWIFT, and Infonet. However,
the licensing process was criticized because the requirements for VANs providers were
unclear and because the RTT's 95 percent ownership of Infonet's Belgian operations
could result in preferential treatment (Tutt, 1990, p. 15).
13. Switzerland
1. Part of the telecommunications profits are used to subsidize the postal services
within the PTT. Swiss mail service is excellent, but partly at the expense of telecom-
munications users and network quality. Transfers in 1983 were more than SFr 600
million; in addition, about SFr 90 million, or 3 percent of value added, were transferred
to the general federal treasury (Knieps, 1985b, p. 19).
14. Austria
1. Unlike many PTTs, the Austrian PTV does not operate broadcast transmission
services, though it distributes broadcast programs to transmitters and cable networks.
2. In 1984, a reduced-cost phone service (Service 660) was introduced in which
recipients pay part of the tariffs for received calls. Starting in 1985, the PTV in con-
junction with Radio Austria initiated a text mail and data service "Telebox," which
attracted about 2100 subscribers by 1988.
3. ITT Austria was a successor to Czeija & Nissel, Austria's first telephone equip-
ment firm, which was at one time partly owned by AT&T and which exclusively sup-
plied large manual exchanges from 1896 to 1913.
4. There are claims that Austria was the first place in Europe with a transistorized
computer, the "Mailiipferl" (Smith, 1984, p. 14).
15. Sweden
1. In West Germany, in contrast, the Bundespost had a monopoly on low-speed

modems, although private supply of high-speed modems was permitted; in Sweden, the
opposite was the case. In each market, extensive arguments were made about why it
was in the public interest to maintain its role in a particular market segment, but the
glaring contradictions of the different approaches illustrates the weakness of the argu-
ments. These restrictions were subsequently dropped in both countries.
2. When total purchases, including such items as vehicles, are considered, 20-25
percent of equipment is imported (OECD, 1987).
3. Teletext electronic mail was started in 1983. "Telebox," an electronic mailbox,
has been offered since 1986. In 1987, "020" free-phone service was introduced.
16. Finland
I . In 1989, Motorola filed a patent suit against Nokia related to cellular handsets.
Nokia held 16 percent of the U.S. cellular equipment market (O'Dwyer, 1989b, p. 8).
Notes 447
The suit was resolved out-of-court, with Nokia and Korea's Tandy Corporation signing
a license agreement favorable to Motorola (O'Dwyer, 1989a).
20. Ireland
1. Its first operator was a boy who was dismissed for playing marbles while the
switchboard was idle.
21. Italy
1. Since 1984, broadcast teletext has been offered under the name of Televideo. The
FT administration also provides public bureaufax service with facsimile equipment in-
stalled at post offices; paging (Teledrin); electronic mail teletex (since 1985), and a slow
scan video system useful for traffic control and surveillance.
27. Telecommunications in Eastern Europe
1. Some of the joint East-West ventures would not have been possible had not the
Coordinating Committee for Multi-Lateral Export Control (CoCom) moved in June 1990
to relax certain trade barriers against exporting telecommunications technology to East-
ern Europe, especially those that have demonstrated that they have ended "intelligence
cooperation" with the Soviet Union (Datapro, 1990). As a result, high-speed fiber op-
tics as well as data equipment (up to 9600 bps), analog mobile interface equipment, and
small digital PABXs could be sold to Hungary, Poland, and Czechoslovakia, though
not to Romania, Bulgaria, and Soviet Union, which were identified as more problematic
from a strategic perspective (Woollacott, 1990).
2. This section owes much of its information to Campbell (1988).
3. In Minsk, Kiev, Barabinskii, Erevan, Akhtyrskii, Perm, Sverdlovsk, Taldom,
Tashkent, Ufa, Kaunas, and Navlinskii and in Estonia (under that republic's control)
(Campbell, 1988, pp. 83-84).
29. Brussels Takes On the Traditional System
1. The E.G. also faced disagreements among member states on satellite policy and
spectrum management. Tensions over service development especially in the area of
VSATs led to the postponement of the E.C.'s Blue Paper on satellite policy in 1990
(Roussel, 1990). Frequency coordination problems beyond the scope of CEPT's Radio-
communications Committee led the B.C. to call for a European Frequency Research
Institute to allocate spectrum for mobile services.
2. In order to satisfy the various national demands for participation, they were ex-
traordinarily complex in structure and acronyms. ESRO was headquartered in Paris,
while its two main technical centers were located in the Netherlands (ESTEC) and in
Germany (ESDAC, later ESOC). Other centers were ESLAB in the Netherlands, ESRIN
in Italy, and ESRANGE and ESTRACK in Sweden.
448 Notes
30. Telecommunications Policy as Industrial Policy
1. For example, Siemens produced computers during the 1960s under licenses ob-
tained from RCA, but was left in the lurch when RCA abandoned its computer business.
Siemens then joined Philips and CIT to form Unidata, a joint computer venture that
faltered in the mid-1970s. After the failure, Siemens entered into an arrangement with
Fujitsu for the manufacture of IBM plug-compatible machines, but the Japanese backed
out in 1985. In 1974, Siemens acquired an unsuccessful joint computer venture of AEG-
Telefunken, which also produced specialized mainframe computers under license from
RCA.
In Britain, the major computer manufacturer has been ICL, which has had some
success in mainframes, but less in the market for smaller units. Its overall market share
declined throughout the 1970s.
2. With the onset of integrated circuit technology, the role of the American govern-
ment as a purchaser of electronic components, until then a major factor, began to de-
cline. In 1960, public procurement of semiconductors was $258 million, 48 percent of
the total. By 1973, it had fallen to $201 million, 6 percent of the total U.S. sales (Levin,
1982). In 1987, because of the military buildup, government procurement (military and
civilian) had risen again to about $1.2 billion, or 12 percent of domestic sales (Semi-
conductor Industry Association, 1988, communication).
3. Capital markets in America, and to a lesser extent in Japan, were vastly more
active than their European counterparts, with the exception of London after the liberal-
ization of its trading practices. For the first ten months of 1985, for example, the turn-
over on U.S. security exchanges was $755.9 billion. In Japan, it was $267.1 billion. In
Britain, capital market turnover was only $48.8 billion, in Germany $29.7 billion, in
France $10.2 billion, and in Italy $3.8 billion (The Economist, 1985, p. 77). By 1990,
total turnover was greatest in Japan ($2,313 billion), followed by the U.S. ($1,542
billion), Germany ($661 billion), the U.K. ($379 billion), France ($65 billion) and Italy
($36 billion) (NYSE, 1990, communication).
4. In Germany, capital investment in private entrepreneurial operations also devel-
oped only slowly. An OECD report found that "Germany has long stood out by the
virtual absence of venture capitalism, in contrast to the dynamism that its small and
medium-sized firms display" (OECD, 1985, p. 31).
5. In absolute terms, support levels for the French plans were small: for the first
five-year plan, funding was $140 million; and for the second five-year period, $263
million.
6. A hierarchy of procurement preferences was established: top priority was given to
French companies, and lowest to imports to France. This system tended to exclude some
of the firms that produced the most advanced designs of chips, and thus restricted French
computer manufacturers at times from the use of the newest technology.
7. The Micro-Electronics Support Scheme, for example, provided only 12 million
of research and development money to British microelectronic manufacturers, mostly to
established firms such as GEC, Ferranti, and Plessey. Similarly, the Advanced Com-
puter Technology Project provided a tiny .8 million per year to British semiconductor
manufacturers between 1976 and 1979. During that same period, the Component Indus-
try Scheme distributed about 20 million (Malerba, 1985). Subsequently, the govern-
ment increased its support by contributing a quarter to a third of R&D expenditures and
a quarter of total necessary investments. The Micro-Electronics Industry Support Pro-
gram (MISP), the Micro-Electronics Application Project (MAP), and the Electronic
Component Industry Scheme (EC1S) were all launched in 1978. Also, the memory com-
Notes 449
ponent firm Inmos was created through the National Enterprise Board. Britain even
provided support to foreign companies that were domiciled in the United Kingdom.
8. International joint ventures often present difficult problems in practice. In addition
to the obstacle of incompatible products, problems include selection of the physical
location, the operational language, the composition of management, and labor sensitiv-
ities. The collaborative effort of Siemens, ICL, and Bull to develop fifth-generation
supercomputer technology was launched with a French director, English as the lan-
guage, and a shared R&D lab located in Germany. For the same reasons that firms like
to see duplication of efforts reduced by joint ventures, trade unions are suspicious of
such efforts. They are fearful of employment reduction and of the ability to deflect the
effects of strikes in one country.
9. One of the consequences of joint venturing is that it leads toward greater stress on
standardization. In March 1984, the twelve leading European computer and commu-
nication firms agreed to draft common standards for the interconnection of their prod-
ucts. In 1985, six European computer manufacturersSTC, Nixdorf, Siemens, Olivetti,
Philips, and Bulljointly determined to base their future computer systems on AT&T's
Unix operating system and to develop software for such uses. In 1986, they further
agreed to collaborate on OSI interconnection standards. Membership was open to other
European firms, but non-European companies (i.e., American and Japanese) were ex-
cluded.
31. Transatlantic Trade Friction
1. The indirect influence of technology must be sharply distinguished from a direct

role of technologists. American engineers, despite the widespread respect for technology
in the abstract, do not have much of a direct political role in society. In 1984, Congress
had only five engineers as members but 261 lawyers (Frey, 1984).
2. Even with the deficit to the United States, the B.C. had a positive trade balance
with the rest of the world of $730 million in 1987, with $2.7 billion in imports and
$3.43 billion in exports (Whitehouse, 1990).
3. Some modification is unavoidable, because European equipment does not match
North American standards. In digital multiplex systems, pulse code modulation (PCM)
is different. Different coding systems translate the amplitude of the analog voice signal
into digital code. Whereas Europe uses the so called "A" law, North America, Japan,
and some others use the "/a" ' aw - The number of voice channels and the signalling
and synchronization are also different. The standard high-volume rate in North America
provides for twenty-four voice channels at 1.544 megabits per second. In Europe, the
equivalent is thirty channels with a total bit rate of 2.048 megabits per second.
In microwave radio equipment, North American equipment interleaves the "go" and
the "return" path frequencies at 20 MHz intervals. On the other hand, the CCIR fre-
quency plan, to which Europeans and Japanese usually adhere, allocates the "go" fre-
quencies at one end of the band, and the "return" frequencies at the other, spaced at
29 MHz each. Furthermore, the frequency transmission band itself is slightly different.
There are also differences in the number of radio and voice channels on each radio
frequency. In North America, twelve are carried on each radio frequency channel, as
opposed to the CCIR standard of six. In contrast, there are 1200 voice channels per
radio channel, whereas for the CCIR, the number is 1800. Other differences exist for
signal level and pilot frequencies and in noise performance specifications. For cables,
the attenuation, wire diameters, cable construction, and color coding are different.
450 Notes
32. International Telecommunications Services
1. This meant that the majority of the large business centers could now be served
directly by an international record carrier, rendering Western Union unnecessary for the
domestic leg of international transmissions. At the same time, the IRCs were required
to unbundle their rates into separate charges for terminal equipment, transmission, and
local access and to interconnect so that customers could use one telex machine to access
all IRCs.
To permit Western Union to offer international communications services, the FCC
had to overcome the opposition of the IRCs, which feared that Western Union would
provide far stronger international competition than domestic competition because of public
recognition of its name and the large number of telex machines already operating in the
United States. Although the U.S. Court of Appeals, 2nd Circuit, overturned this deci-
sion in 1980, the International Record Carrier Competition Act subsequently amended
the Communications Act of 1934 to permit Western Union IRC service. At the same
time, the legislation permitted the IRCs to provide domestic record service.
2. Although there are a number of technical objections to this proposal, none of them
is particularly convincing, considering the sophistication of telecommunications tech-
nology. Another objection is that multiple codes would impose an extra cost on the
PTTs. This extra cost, however, could be compensated by payments from the carriers,
which would benefit from access.
3. Prior to the advent of communications satellites, the FCC focused on authorization
for and ownership of submarine cable facilities. It scrutinized applications for these
facilities to decide whether their need justified an increase in a carrier's rate base. The
commission's close review of the applications resulted partly from the fact that invest-
ments in international submarine cables were quite large in comparison to investments
in most domestic facility applications.
AT&T, the IRCs, and other carriers used these cables and were at least theoretically
subject to rate base regulations and thus sought to obtain ownership interests in these
facilities, in the form of "indefeasible rights of use" (IRU). The FCC concluded that it
was impossible to audit the IRCs and that no benefits would flow from rate regulation
of the industry. The carriers sought ownership interests in order to expand their rate
bases and to realize certain benefits under the U.S. tax code. The IRUs still exist, and
the ownership interests in addition to the PTTs' interests have added new parties to the
negotiating process.
4. Following enactment of the Comsat Act, the FCC developed various policies to
establish and protect Comsat's role as the U.S. signatory and monopoly U.S. provider
of international satellite service. A key component of this role was the construction and
operation of these stations. Comsat and the U.S. international service carriers, AT&T
and the IRCs, would own and operate stations jointly through a cooperative Earth Sta-
tion Ownership Committee (ESOC). This approach gave Comsat the major role in earth
station management as well as in investment decisions and allowed Comsat to bundle
earth station costs with space segment costs in setting rates.
Following pressure from various carriers and users, the FCC proposed a more liberal
international earth stations policy in 1982. Carriers and users wanted Comsat to separate
out charges for its space segment (satellite) and earth segment (earth station), and they
also wanted the option of building their own lower-cost earth stations at sites with
efficient access to Intelsat. In 1984, the FCC authorized international earth stations
(Frieden, 1983).
Not surprisingly, the competitive pressures that led to modifications of policies re-
Notes 451
garding earth station ownership and authorization of users necessitated an examination

of whether Comsat should continue to be the sole source of access to Intelsat.
5. The FCC also gradually decreased its role in the planning aspects of international
facilities in favor of reliance on the market. In the past, the FCC had been jointly
involved in the international facilities planning process with carriers and foreign author-
ities. The planning process had begun as simple negotiations between the FCC and the
PTTs for landing rights and terminal points. As technology became more varied, the
planning process also became increasingly complex. With the advent of satellite com-
munications, the FCC began to regulate the ratio of satellite to cable circuits for trans-
atlantic telecommunications services. For the TAT-5 cable, the FCC required AT&T to
use one cable circuit for every five satellite circuits activated. After considerable pres-
sure from PTTs and American carriers, this ratio was reduced to l:l in the 1972 ap-
proval of the TAT-6 cable, subsequently applied only to AT&T. The carriers tradition-
ally preferred submarine cables because of their proven technology and the carriers'
ownership of the IRUs. Moreover, carriers used satellites under leases, which could not
be included in a carrier's rate base. Following the PTTs initiative for comprehensive
planning, the FCC agreed in 1976 to substitute a long-range planning process for the
disruptive and time-consuming review process previously used. In 1988, the FCC ter-
minated AT&T's traffic balance restrictions, thus allowing Comsat and AT&T to reach
their own agreement on facilities usage.
Using the planning process, the FCC consulted with interested parties and in 1977
rejected the TAT-7 cable as being economically unjustified. The rejection caused strong
protests from many of the European PTTs, other governmental agencies, AT&T, and
the IRCs. In 1979, the FCC reversed its decision, claiming that the new cable would
result in increased service reliability. Although the FCC still believed that TAT-7 was
unnecessary to satisfy demand through 1985, the strong European pressure led the com-
mission to balance the economic efficiency arguments, which were supported by the
revenue requirements of the new facilities of about $68 million, with its desire to main-
tain good international relations.
33. The Economics of ISDN Integration
1. Other contributions to the standardization literature include Arthur, 1983; Kindle-

berger, 1983; Wilson, 1984; Besen and Johnson, 1986; and Katz and Shapiro, 1985.
34. The Political Economy of ISDN
1. A large research project on broadband-ISDN was conducted at Columbia Univer-

sity, funded by the Markle Foundation and led by Professor Martin Elton. It marshals
technological, economic, and policy issues of broadband fiber (Elton, 1991).
2. To deal with the ISDN issues, the FCC issued a Notice of Inquiry in 1983. Its
goals were both to generate comments on the FCC's role in ISDN and to stimulate
interest in the policy discussion on ISDN itself. The first report, issued in 1984, restated
the FCC's intention to take a limited role. It set, however, several policy principles for
ISDN design: a flexible numbering plan that permits user choice of carriers, domesti-
cally and internationally (this was possible for international telex, but not for interna-
tional telephony under the existing numbering plan); and no limitations of satellite hops
in international connections. On the numbering issue, the 1984 Brasilia meeting of the
452 Notes
CCITT adopted a numbering plan that permits a multivendor system in the United States,
and also accommodates in PTT countries alternative providers of databases and local
area networks (Schiller, 1984).
3. Between the transmission line and the user, two sets of network termination equip-
ment exist, known as NT1 and NT2. NT1 provides the subscriber functions such as test
loops, power feedings, timing, multiplexing, and synchronization. The network termi-
nation point T is located between NT1 and NT2. NT2 equipment is more sophisticated
and provides functions of switching, multiplexing, and protocol handling by equipment
such as PBXs, local area networks, and terminal controllers.
4. The original American telephone companies' position had been to go with T as
the demarcation point; this point had to be defined by the CCITT Study Group, which
had thought in terms of an integrated rather than separated NT1 and NT2. PTTs can
integrate them if they wish to do so; and the German Bundespost's elaborate graphic
representation of ISDN architecture omits even an acknowledgement of point T's exis-
tence.
5. Going one step further, the German telecommunications equipment industry com-
missioned a report on ISDN by another former head of the governmental telecommuni-
cations department, Franz Arnold. That study was, if anything, critical of the Bundes-
post's ISDN and cable television plans as too timid. Arnold called for a "national
effort," which would include private and public funds, in which ISDN was a first step
in the right direction. This would provide German export industry with future potential,
which the present traditional copper-wire-oriented technology does not offer. For such
support, public funds must be available to reduce private risk (Muller-Sachs, 1984, pp.
6-7). In pursuit of goals, the Bundespost took an administrative leadership role in the
CCITT Study Group XVIII, which was chaired by one of its officials, Theodore Irmer,
who subsequently became the head of the CCITT itself. The DBF was also the first
PTT administration to announce a tariff structure for ISDN.
6. An operational framework for a CEI was reached in March 1986 by three major
firms with diverse interests: Pacific Bell (carrier), IBM (equipment supplier), and Tym-
net (value-added provider).
35. Value-Added Networks and Services

1. AT&T's involvement in packet switching service was tumultuous and has so far
been unsuccessful. Regulation caused some of its problems, but others were a result of
its own operations, and indicate that economies of scale should not be overvalued in
this field.
Following the divestiture decree in 1982, it was unclear whether the Bell Operating
Companies could provide VAN service and whether the Computer II rules regarding
separation of enhanced from basic services applied. This was clarified by the FCC when
it declared basic packet switching (X.25-to-X.25) to be a "basic" service that BOCs
therefore could provide, subject to regulation.
2. In 1985, the FCC removed barriers from the BOCs and permitted a bundled pro-
vision of basic packet transmission with the "enhanced" protocol conversion asyn-
chronous X.25-to-X.25 and X.25-to-X.75. However, they were also required to provide
such services to their competitors at nondiscriminatory terms; they had to file an ac-
counting plan of separation, and they could not unfairly cross-subsidize their service.
Specific rules were established for cost allocation and pricing.
3. EUSIDIC later revised its 1990 findings to a 19 percent failure rate after discount-
Notes 453
ing a one-day Transpac failure. However, CEPT challenged the survey, claiming that
96 percent of calls were completed successfully in 1989. The CEPT figures measured
only X.25 and not dial-up access (Schenker, 1990b).
4. The SCC also stores the message and relays it to the addressee. The RPs and
SCCs interconnect with leased lines since both the sender and receiver are encrypted so
that it is impossible to control which banks communicate with each other. The RPs
cannot communicate directly with each other, but must go through the SCCs. The lines
between the RPs and SCCs are doubled as insurance against lines failure, and the SCCs
also have back-up installations.
5. SWIFT supplies users with the necessary hardware and software through its sub-
sidiary STS SWIFT terminal services, with equipment by IBM, DEC, NCR, Burroughs,
and others. About 70 percent of the member banks opt to purchase SWIFT equipment
over equipment supplied directly by computer products vendors.
36. Videotex
1. SEAT is the official "publisher" of on-line information with its main product
being the Pagine Gialle Elettroniche (Electronic Yellow Pages). SEAT has data from 1
million Italian firms classified into 1400 categories. The system allows companies to
include technical and commercial data on their products, yet had only 25,836 users in
1987. Of the user sessions, 39 percent were to access PGE, 11 percent made use of
teletourism services, and 42 percent were on Amadeus, a real-time credit rating and
financial facts service aimed at business users. The PGE system developed by SEAT
runs on two simple commands, cerca (find) and mostra (show). Other services offered
by SEAT include Mastermail (electronic mail), Banca dati Tributaria (home banking),
teletourism, and an Official Airlines Guide. These services are generally not used much
beyond corporate applications. In 1990, there were 100,000 terminals in use, and SIP
had invested only 100 billion lire ($1 million).
37. Transborder Data Flows
1. Technological advances in general facilitate invasions of privacy. Electronic data

banks can be abused to obtain vast amounts of personal data, as discussed in this chap-
ter. However, other privacy abuses exist as well. For example, cellular, mobile, and
satellite telephony increase the chance that conversations may be intercepted and moni-
tored. Beepers and intraorganizational networks can be used by corporations to track
employee activities. Call forwarding and facsimile machines make it easier to intrude
on an individual. Almost every new telecommunications service has raised new types
of privacy issues and concerns.
2. In 1990, over 100,000 registration entries were received by the Home Office for
renewal, and 37,000 amendments to register entries and over 19,000 new applications
were filed.
3. For a summary of these laws as they deal with telecommunications, see Noam
(1990 and forthcoming).
4. One initiative was taken by the author in 1990 when he served as a commissioner
in New York State. A draft set of privacy principles for telecommunications service was
adopted by the New York Public Service Commission (Noam, 1990; NYPSC, 1991).
5. The common law copyright protection provided primarily that if one had not pub-
454 Notes
lished information in one's possession, no one else could take it and publish it. This
was similar to a trespass and conversion action.
6. Justice Louis Brandeis, in a famous dissent, wrote of "the right to be left alone
the most comprehensive of rights and the right most valued by civilized men" (Olm-
steadv. U.S., 1927).
7. On the history of privacy, see Posner (1981); Simmel (1906); Westin (1967);
Seipp (1978). In the United States, privacy is a nonpartisan issue. The Privacy Act of
1974 was cosponsored by Senators Edward Kennedy and Barry Goldwater.
8. In the extreme, private information is so valuable to an individual as to make him
a target for blackmail. See also Brown and Gordon (1980) for an economic perspective
from the FCC.
9. In an information-based society and economy, the incentives to acquire informa-
tion are continuously increasing. See Posner (1981, pp. 231-347) for the most compre-
hensive discussion of the economics of privacy.
10. Another indication is provided by a survey conducted by American Express among
its card holders, which showed that 90 percent felt that mailing list practices were
inadequately disclosed, 80 percent felt that information should not be given to a third
party without permission, and more than 30 percent believed federal legislation was
needed to restrict the use of lists ["Privacy Study Reveals Lack of Consumer Confi-
dence," Direct Marketing, December 1988, p. 8, in McManus (1989)]. It should be
noted that American Express makes extensive use of the data it has accumulated on its
cardholders. According to Fortune, the company computers "maintain and update weekly
a profile of 450 attributessuch as age, sex, and purchasing patternson every card-
holder" (Newpert, 1989, p. 82).
A 1988 survey by the Massachusetts Executive Office of Consumer Affairs of the
main consumer complaints found them topped by telemarketing and promotional mail-
ings (Marchocki, 1989, p. 47; McManus, 1989, p. 82).
11. "Pac Bell backs off selling lists," Alameda Times Star, April 16, 1986, p. 16,
as cited in McManus (1989).
38. Networks in the Future
1. The optimal size of jurisdictions was always dependent on communications. French

departments were based on the distance that a horseback rider could cover in a day.
Transportation and communications technology changes the optimal size. It is hard to
imagine a voluntary European integration without telecommunications.
2. Professor Henry Niman, per Marc Rotenberg; communication.
39. Toward a Modular Network
1. These can partly substitute for each other.

2. First for data and in an ISDN network also for voice service.
3. One can similarly use software hierarchies other than OSI.
4. It is easiest to set the horizontal physical coordinate where segments are defined
in purely spatial terms, such as "inside wiring." But for central office functions the
dividing lines are more complicated and controversial. Central exchanges combine sev-
eral physical as well as software segments. (The switch generics of SPC central offices
combine functions of access, features, functions, and transport.) One may need some
Notes 455
underlying operating protocol across modular lines. (Such a system, if it is proprietary,

could give its owner-manufacturer an advantage. But this could be resolved by nonpro-
prietary standards.)
This does not mean modules and interface points everywhere, since this would be
technologically burdensome. Nor would there have to be more unbundling than before.
Nor does it mean that one would have to set all points at once. Setting the specifics of
interfaces is complex and need not be ahead of demand (Some embedded technology
today may straddle several modules and separation may be impractical). That is a reg-
ulatory policy decision that will have to take into account cost and performance impli-
cations. But it would require a more systematic approach to however much disaggre-
gation is decided upon.
5. At present, one cannot, for example, upgrade an analog SPC switch by a digital
adjunct or overlay; instead, one must replace the entire switch.
41. Regulatory Tasks for the Future: Interconnectivity in the Pluralist Network
1. The New York State Public Service Commission established, as the first state,
principles of ONA (Noam, 1990); it was also the first to establish collocation (NYPSC,
1989; NYPSC, 1990c).
2. In the United States, the fundamental 1934 Communications Act, for example,
defines it in a totally circular fashion. Section 153(h) states that "a common carrier
means any person engaged as a common carrier for hire. . . . "
3. The origins of common callings arose from the medieval custom of carrying on
business at stated periods in fairs within defined market areas. Fraudulent commercial
practices were prohibited and disputes were settled by the steward of the fair with the
assistance of the other merchants.
4. Western Union Telegraph Co. v. Call Publishing Co., 181 U.S. 92, 98 (1901).
5. These are part of the questions that the New York Public Service Commission
addressed. After public comment, the commission adopted common carriage principles
of content-neutrality-nondiscrimination and separation of content and conduit func-
tions. Issues include line-drawing problems. For example, are supplementary or subsid-
iary functions such as the signaling channel and network management functions part of
common carriage? Or is the central office affected, which is an issue in collocation? Or
billing and collection services, which are important to VANs?
Another kind of a question is whether common carriage and private carriage can
coexist in the same entity, or in the same service. For example, can a carrier be a
common carrier to users at one end of a communications link, but a private carrier to
the user at the other end?
Similarly, there is the question of how market power or monopoly should enter into
common carrier issues. Or the questions of what would qualify as legitimate distinctions
between classes of users, or what forms of prior subscription may be used for "adult"
programming.
6. New York State has a $1 per month lifeline plan, with installation spread over a
year to be $2 per month (NYPSC, 1987).
7. There is evidence for telephone wiretaps by private parties and individuals ten
years after Bell's patent (Westin, 1967). The New York Police Department tapped tele-
phones since at least 1895. In 1916, this led to a scandal about listening in on private
conversations of Catholic priests and to those of a law firm involved in competition to
J.P. Morgan & Co. for World War I munitions contracts (Seipp, 1978).
456 Notes
8. For example, in 1962, the U.S. federal government had 1030 computer central
processing units; in 1972, 6731; in 1982, 18,747; and in 1985, over 100,000 (Linowes,
1989).
9. In the past twenty years the cost of access to a name on a computer-based mailing
list has come down to about one thousandth of its earlier cost.
References
1. Public Telecommunications: A Concept in Transition
Dallmeyer, Martin. 1977. Quellen zur Geschichte Des Europdischen Postwesens 1501 -
1806. Bern: Verlag Michael Lassleben, Kallmiinz, vol. 1.
Ministry of Communications. 1990. Political Agreement on the Telecommunications
Structure. Copenhagen. June 22, p. 2.
Schulte, Josef. 1982. Endgeratekonzeption im Fernsprechdienst der Deutschen Bundes-
post. In Telekommunikation in der Bundesrepublik Deutschland, ed. Dietrich
Elias. Heidelberg: R. V. Decker's Verlag, G. Schenck, pp. 319-48.
Stephan, Heinrich. 1859. Geschichte der Preussischen Post von ihrem Ursprunge bis
auf die Gegenwart. Berlin: Verlag der Koniglichen Geheimen Ober-Hof buch-
druckerei, (R. Decker).
Wolf, Thomas. 1983. Akzeptanz und Kosten des Schnurlosen Telephons, Neue Ziiricher
Zeitung. October 26.
2. The Establishment of the PTT System
Brock, Gerald W. 1981. The Telecommunications Industry: The Dynamics of Market

Structure. Cambridge, MA: Harvard University Press.
Clear, C. R. 1940. The Birth of the Postal Service. In Penny Postage Centenary, ed.
Samuel Graveson. London: The Postal History Society, pp. 21-33.
Collis, Maurice. 1960. Marco Polo. New York: New Directions.
Dallmeyer, Martin. 1977. Quellen zur Geschichte des Europaischen Postwesenes 1501-
1806. Bern: Verlag Michael Lassleben Kallmunz, vol. 1.
Edwards, Eliezer. 1879. Sir Rowland Hill. London: Frederick Warne & Co.
Grosse, Oskar. 1931. Stephan: Von Postschreiber zum Minister. Berlin: Mittler und
Sohn.
Hemmeon, J. C. 1912. The History of the British Post Office. Cambridge, MA: Harvard
University Press.
Hill, Sir Rowland, and George Birbeck Hill. 1880. The Life of Sir Rowland Hill. Lon-
don: Thomas de la Rune & Co.
Kramer, Richard. 1991. A Faith That Divides: Competition Policy in European Tele-
communications. Master's Thesis. Philadelphia, PA: Annenberg School for
Communication.
Meyer, Hugo Richard. 1907. Public Ownership and the Telephone in Great Britain.
New York: Macmillan.
457
458 References
Organization for Economic Co-Operation and Development (OECD). 1983. Telecom-

munications: Pressures and Policies for Change. Paris: OECD.
. 1990. Committee for Information, Computer and Communications Policy. Per-
formance Indicators for Public Telecommunications Operators. DSTI/ICCP/TISP/
89.10. February. Paris: OECD.
Perry, Charles R. 1977. The British Experience 1876 to 1912: The Impact of the Tele-
phone During the Years of Delay. In The Social Impact of the Telephone, ed.
Ithiel de Sola Pool. Cambridge, MA: MIT Press, pp. 64-96.
Smyth, Eleanor C. 1907. Sir Rowland Hill: The Story of a Great Reform. London: T.
Fisher Unwin.
auf die Gegenwart. Berlin: Verlag der Koniglichen Geheimen Ober-Hofbuch-
druckerei (R. Decker).
von Beust, Christopher. 1748. Versuch einer aufiihrlichen Erklarung des Post-Regals.
Berlin: Jena, vol. II, sec V.
3. Network Tipping: The Rise and Fall of the Public Network Monopoly
Allen, David. 1988. New Telecommunications Services: Network Externalities and Critical
Mass. Telecommunications Policy 12(3): 257-71.
Aronson, Jonathan, and Peter F. Cowhey. 1988. When Countries Talk. Cambridge,
MA: Ballinger.
Barnes, J. A. 1954. Class and Committees in a Norwegian Island Parish. Human Re-
lations, 1.
Baumol, William J., John C. Panzar, and Robert D. Willig. 1982. Contestable Markets
and the Theory of Industry Structure. New York: Harcourt Brace Jovanovich.
Boissevain, J. 1979. Network Analysis: A Reappraisal. Current Anthropology, 20.
Bott, E., 1971. Family and Social Network, New York: Free Press.
Buchanan, James M. 1965. An Economic Theory of Clubs. Economica 32(125): 1-14.
David, Paul A. 1987. Some New Standards for the Economists of Standardization in
the Information Age. In Economic Policy and Technological Performance, ed.
P. Dasgupta and P. L. Stoneman. Cambridge: Cambridge University Press.
Dumey, Richard. 1983. Telecommunications and the BT Case Competition Rules. Un-
published Speech at Antitrust Luncheon. Commission of The European Com-
munities, Directorate General for Competition, Brussels.
Economides, Nicholas. 1989. Desirability of Compatibility in the Absence of Network
Externalities. American Economic Review 79(5): 1165-81.
Elmaghraby, Salah E. 1970. Some Network Models in Management Science. New York:
Springer.
Hardin, Garrett. 1968. The Tragedy of the Commons. Science 162 (December 13).
Hayashi, Koichiro. 1988. The Economies of NetworkingImplications for Telecommu-
nications Liberalization, paper presented at the IIC Conference, Washington,
D.C., September.
Heal, Geoffrey. 1989. The Economics of Networks. New York: Columbia Institute for
Tele-Information. Working Paper Series.
Huber, Peter. 1987. The Geodesic Network. Washington, D.C.: U.S. Government Printing
Office.
Kami, Shlomo. 1986. An Analysis of Electrical Networks. New York: Wiley, pp. 1-4.
References 459
Klingman, David J., and J. Mulvey eds. 1981. Network Models and Associated Appli-
cations. New York: Elsevier.
Knox, Robert. 1830. Elements of General Anatomy, Translated from Inst. edition of
Beclard's Anatomy by D. A. Beclard. Edinburgh: Maclachlan & Stewart, p.
214.
McGuire, Martin. 1972. Private Good Clubs and Public Good Clubs: Economic Models
of Group Formation. Swedish Journal of Economics 74(l):84-99.
Mueller, Milton. 1988. Interconnection Policy and Network Economics. Paper presented
at Telecommunications Policy Research Conference, Airlie, Virginia, October
31.
Noam, Eli M. 1986. The "New" Local Communications: Office Networks and Private
Cable. Computer Law Journal 6(2).
. 1988. The Next Stage in Telecommunications Evolution: The Pluralistic Net-
work. Paper presented to the International Telecommunications Society, New
York: Columbia Institute for Tele-Information Working Paper Series.
Richardson, Jeremy John, Gunnel, Gustafson, and Art Jordan. 1985. In R. A. W. Rhodes,
Power Dependence, Policy Communities and IntergovernmentalNetworks.
Colchester, Essex: Department of Government, University of Essex, Wivenhoe
Park, pp. 6-8.
Rothenberg, Jerome. 1976. Inadvertent Distributional Impacts in the Provision of Public
Services to Individuals. In Public and Urban Economics, Ronald Grieson, ed.
Lexington, MA: Lexington Books.
Schelling, Thomas. 1978. Micromotives and Macrobehavior. New York: W. W. Nor-
ton.
. 1969. Models of Segregation. Santa Monica, CA: RAND Corporation.
Shepherd, William. 1983. Concepts of Competition and Efficient Policy in the Telecom-
munications Sector. In Telecommunications Today and Tomorrow, ed. Eli M.
Noam. New York: Harcourt Brace Jovanovich.
Tiebout, Charles. 1956. A Pure Theory of Local Expenditures. Journal of Political
Economy 64(5): 414-24.
Tullock, Gordon. 1971. Public Decisions as Public Goods. Journal of Political Econ-
omy 179(4): 913-18.
Woroch, Glenn A. 1990. On The Stability of Efficient Networks. Waltham MA: GTE
Labs. Mimeo.
Zacharisen, W. M. 1932. The Atomic Arrangement in Glass. Journal of the American
Chemical Society 54(10): 38-42.
4. Forces of Centrifugalism
AT&T Communications. 1983. The World's Telephones. Morris Plains, NJ.

Beniger, James. 1986. The Control Revolution. Cambridge, MA: Harvard University
Press.
Cole, Barry, ed. 1990. After the Break-up: Assessing the New Post-AT&T Divestiture
Era. New York: Columbia University Press.
Crandall, Robert W. 1988. Fragmentation of the Telephone Network: Implications for
the Policymaker. Washington, DC: Brookings Institution.
Egan, Bruce. 1990. Information Superhighways: The Future of Broadband Communica-
tions. Dedham, MA: Artech House.
460 References
Elton, Martin, ed. 1991. Integrated Broadband Networks: The Public Policy Issues,
Amsterdam: Elsevier.
Ergas, Henry. 1989. International Telecommunications Accounting Arrangements: An
Unsustainable Inheritance? Paper presented at Colloque Villefranche-sur-Mer.
June 1-3.
FCC. 1945. Statistics of Common Carriers. Washington, D.C.: U.S. Government Print-
ing Office.
. 1955. Statistics of Common Carriers. Washington, D.C.: U.S. Government
Printing Office.
Printing Office.
Printing Office.
Printing Office.
Guerard, A., G. Lafarge, and C. Pautrat. 1979. Les Regions dans la Course au Tele-
phone. Economic et Statistique 117 (December):37-49.
Ito, Yoichi. Historical Review of Japanese Telecommunications Policies. In Pacific
Basin Telecommunications, ed. Eli Noam and Seisuke Komatsuzaki, forth-
coming.
Komatsuzaki, Seisuke. 1986. The Economic Impact of Informatization. In KEIO Com-
munication Review 7(March): 13-24. Tokyo: Institute for Communications Re-
search.
Nagai, Susuma. On the Deregulation Process in Japanese Telecommunications. In Pa-
cific Basin Telecommunications, eds. Eli Noam, and Seisuke Komatsuzaki, and
Douglas Conn, forthcoming.
Nambu, Tsuruhiko. A Comparison of Telecommunications Policies for Deregulation. In
Pacific Basin Telecommunications, eds. Eli Noam, Seisuke Komatsuzaki, and
Douglas Conn, forthcoming.
Noam, Eli M. 1986. The "New" Local Communications: Office Networks and Private
Cable. Computer Law Journal 6(2): 247-82.
. Divestiture Plus V and the Coming Regulatory Agenda. In New Regulatory
Concepts, issues, and Controversies, ed. Harry Trebing, forthcoming.
Nora, Simon, and Alain Mine. 1980. The Computerization of Society: Report to the
President of the French Republic. Cambridge, MA: MIT Press.
Pfeiffer, Giinter, and Bernhard Wieland. 1990. Telecommunications in Germany: An
Economic Perspective. New York: Springer-Verlag.
Piore, Michael J. and Charles F. Sabel. 1984. The Second Industrial Divide: Possibili-
ties for Prosperity. New York: Basic Books.
Pool, Ithiel de Sola. 1990. Technologies Without Boundaries: On Telecommunica-
tions in a Global Age, ed. Eli Noam. Cambridge, MA: Harvard University
Press.
Porat, Marc U. 1978. Communication Policy in an Information Society. In Communica-
tions for Tommorrow, Policy Perspectives for the 1980s, ed. Glen O. Robinson.
New York: Praeger.
Schiller, Dan. 1982. Business Users and the Telecommunications Network. Journal of
Communication 32(4): 8497.
Elias. Heidelberg: R. V. Decker's Verlag, G. Schenck, p. 321.
References 461
Steckel, Marie-Monique, and Marc Fossier. 1990. France Telecom: An Insider's Guide.
Chicago: Intertec.
Strassman, Paul. 1985. The Information Payoff: The Transformation of Work in the
Electronic Age. New York: Free Press.
Telecom Factbook. 1986. New York: Television Digest.
5. Defense of the Telecommunications Monopoly
AT&T. 1976. An Econometric Study of Returns to Scale in the Bell System. Bell
Exhibit 60, FCC Docket 20003 (Fifth Supplemental Response), August 20.
Baumol, William J., John C. Panzar, and Robert D. Willig. 1982. Contestable Markets
and the Theory of Industry Structure. New York: Harcourt Brace Jovanovich.
Chrust, Steven. 1985. MCI Communications Corporation. New York: Bernstein Re-
search.
Dobell, Rodney A., et al. 1972. Telephone Communications in Canada: Demand, Pro-
duction, and Investment Decisions. Bell Journal of Economics and Management
Science 3(1): 175-219.
Federal-State Joint Board. 1988. Monitoring Report of the Federal-State Joint Board.
CC Docket No. 80-286, 87-339.
Fuss, Melvyn, and Leonard Waverman. 1977. Multi-product Multi-input Cost Functions
for a Regulated Utility: The Case of Telecommunications in Canada. NBER
Conference on Public Regulation. Washington, D.C., February 15-17.
Mantell, L. 1974. An Econometric Study of Returns to Scale in the Bell System. Staff
Research Paper. Office of Telecommunications Policy, Executive Office of the
President. Washington, D.C., February.
Meyer, John R., Robert W. Wilson, M. Alan Baughcum, Ellen Burton, and Louise
Caoulette. 1980. The Economics of Competition in the Telecommunications In-
dustry. Cambridge, MA: Oelgeschlager, Gunn & Hain.
Nadiri, Ishaq and Mark Schankerman. 1981. Production, Technological Change and
Productivity in the Bell System. In Productivity Measurement in Regulated In-
dustries, ed. Thomas Cowing and Rodney Stevenson. New York: Academic Press,
pp. 219-48.
OECD. 1990. Committee for Information, Computer and Communications Policy. Per-
Elias. Heidelberg: R. V. Decker's Verlag, G. Schenck, pp. 319-48.
Shepherd, William. 1983. Concepts of Competition and Efficient Policy in the Telecom-
munications Sector. In Telecommunications Regulation Today and Tomorrow,
ed. Eli Noam. New York: Harcourt Brace Jovanovich, pp. 79-120.
6. Policy Directions
Le Boucher, Eric, and Jean-Michel Quatrepoint. 1984. La Guerre Mondiale de la Com-

munication (four parts). Le Monde (January 11): 1, 44.
Le Monde. 1984. La Guerre Mondiale de la Communication. January 11-14.
462 References
Nora, Simon, and Alain Mine. 1980. The Computerization of Society: Report to the
President of the French Republic. Cambridge, MA: MIT Press.
7. Germany
Alcatel Annual Report. 1989.

Allentier. 1973. Die Telegraphenstation Koln-Flittard, Eine Kleine Ceschichte der Na-
chrichtentechnik. Rheinisch-Westfaelisches Wirtschaftsarchiv, Koln.
Appel, Timothy. 1990. Siemens Unveils East German Investment Plan. Communica-
tions Week International, March 26, p. 6.
Armbruster, Heinrich. 1986. Breitband-ISDN erfiillt die wachsenden Telekommunika-
tionswilnsche. Telecom Report 9 (3): 168-75.
Bauer, H. 1982. Future Outlook for Communications in the Light of New Technologies.
Telecommunications Journal 49(9): 579.
Blau, John. 1989. Telecoms Show Biggest Profit. Communications Week International,
September 4, p. 17.
. 1990. Siemens May Shed Parts of Nixdorf. Communications Week Interna-
tional, January 29, p. 12.
Blau, John, and Karen Lynch. 1990. Germany Opens Satcoms. Communications Week
International, July 2, p. 3.
Bruce, Peter. 1986. Bundespost Reaches Deregulation Milestone. Financial Times, July
31.
Communications Week International. 1989. German Cellular Race Is On. May 15.
Deutsche Bundepost (DPB). 1981. "Stellungnahme der DBP zum Sondergutachten
der Monpolkommission 'Die Rolle der DBP in Fermeldewesen." Referat 603.
May.
. 1982a. Deutsche Bundespost Annual Report. Bad Windsheim: Georg Hei-
decker.
. 1982b. Harmonization of the Tariffs for Fixed Connections with Those for
Switched Connections in the Public Telecommunications Network. Bad Windes-
heim: Verlag fur Wissenschaft und LebenGeorge Heidecker.
. 1985. From the Analogue Telephone Network to the Integrated Broadband
Telecom Network. DBP.
. 1987. Statistisches Jahrbuch 1986. London/Pfalz: Pfalzische Verlagsanstalt.
Dodsworth, Terry. 1987. Manufacturing with Passion. Financial Times, June 29, p. 15.
Dorrenbacher, Christoph. 1988. Telecommunications in West Germany. Telecommuni-
cations Policy, 12(4): 344-52.
The Economist. 1985. March in the EEC Competition. April 6.
Elias, Dietrich (ed.). Telekommunikation in der Bundesrepublik Deutschland 1982.
Hamburg: R. V. Decker.
, Roland Klett, and Eberhard Witte. 1980. Neue Kommunikationsdienste der
Bundespost in der Wirtschaftsordnung und Gemeinwirtschaft. Heft 19. Baden-
Baden: Nomos.
Engals, W., A. Gutowski, W. Hamm, et al. 1988. More Competition in Telecommu-
nications. Frankfurt: Frankfurt Institute.
Federal Ministry of Posts and Telecommunications (PTT). 1990. Documents Concern-
ing the Licensing of Private Satellite Network in the FRG. Bonn. June 19.
Federal-State Joint Board. 1988. Monitoring Report in CC Docket No. 80-286, CC
Docket No. 87-339. March.
References 463
Fuhrman, Peter. 1990. Papa Siemens. Forbes, May 28, p. 100.

Gilhooly, Dennis, and Jennifer Schenker. 1989. Germans Rethink ISDN. Communica-
tions Week International, March 13, p. 1.
Gartner, Bernhard, and Georg Habenicht, eds. 1990. Doing Business with DBF Telekom
in the Center of Europe. Bonn: Deutsche Bundespost Telekom.
Glotz, Peter. 1983. Unpublished manuscript of speech.
Goetzler, Herbert. 1976. 100 Jahre Fernsprechtechnik. Sonderdruck aus der Nachri-
chtentechnischen Zeitschrift 29(2): 105-8.
Goodhart, David. 1990. Combining Resolution and Tradition. Financial Times, June
21, p. 20.
Hafner, Katie. 1990 German Phone System is Taxed By Unification. The New York
Times, December 10, p. D8.
Hayes, Dawn, and Candee Wilde, 1988. NYNEX Teams in German Bid. Communica-
tions Week International, December 19.
Henneman, Gerhard. 1984. Bin Gigant rekelt sich. Neue Medien l(June): 70-77.
Heuss, Theodor. 1946. Robert Bosch: Leben und Liestung. Stuttgart: Rainer Wunderlich
Verlag.
Hoffmann-Riem, Wolfgang. 1984. Policy Research on Telecommunications in West
Germany. In Policy Research in Telecommunications, ed. Vincent Mosco. Nor-
wood, NJ: Ablex.
Holcombe, A. N. 1911. Public Ownership of Telephones on the Continent of Europe.
Boston: Houghton Mifflin.
Inquiry Commission. 1984. Bericht der Enquete Kommission. "Neue Information und
Kommunikationstechniken." Deutscher Bundestag: D. Bopnn.
ITU. 1990. Annual Telephone Statistics. Geneva: International Telecommunications Union.
ITU. 1985. Telecommunications Journal: 191.
Kelly, Tim. 1990. Telecommunications and Eastern European Economies. Paris: OECD.
Knieps, Giinther, Jurgen Miiller, and Carl Christian von Weizsa'cker. 1981. Die Rolle
des Wettbewerbs im Fernmeldebereich. Baden-Baden: Nomos,
Korner, E. 1937. Die Deutsche ReichspostBine Hoheitsverwaltung. In Die Reichspost
im Staate Adolf Hitler, ed. Jakob Nagel. Berlin: Verlag Georg Konig.
Lange, Bernd-Peter. 1984. Dokument Einer Einseitigen Politik. Medium, May.
Malerba, Franco. 1985. The Semi-Conductor Business. Madison, WI: University of
Wisconsin Press.
Markoff, John. 1988. I.B.M. to Sell Rolm to Siemens. New York Times, December 14,
p. D l .
Mestma'cker, Ernst-Joachim. 1980. Fernmeldenmonopol und NachfragemachtWirt-
schaftsrechtliche und ordnungspolitische Probleme der hoheitlichen und unter-
nehmerischen Funktion der DBP. In Kommunikation ohne Monopole, ed. Ernst-
Joachim Mestma'cker. Baden-Baden: Nomos, pp. 185, 196.
Mestma'cker, Ernst-Joachim, ed. 1980. Kommunikation Ohne Monopole. Baden-Baden:
Nomos.
Moschel, Wernhard. 1988. (No) Restructuring of Telecommunications in Germany?
Unpublished paper.
Monopolkommission. 1981. Die Rolle der Deutschen Bundespost im Fernmeldewesen.
Sondergutachten 9. Baden-Baden: Nomos.
Nagel, Jakob, ed. 1937. Die Reichspost im Staat Adolf Hitler. Berlin: Verlag Georg
Konig.
Neumann, Karl-Heinz. 1986. Economic Policy Toward Telecommunications, Informa-
tion and the Media in West Germany. In Marketplace for Telecommunications:
464 References
Regulation and Deregulation in Industrialized Democracies, ed. Marcellus Snow.

White Plains, NY: Longman, pp. 131-52.
. 1990. The Unification of Telecommunications in Germany. Paper prepared for
the Brussels Meeting of the International Institute of Communications Telecom-
munications Forum, July 16-17.
Ohmann, Friederich, and Heinrich Armbruster. 1986. Beitrage zur Fernmeldepolitischen
Diskussion in der Bundesrepublik Deutschland. April.
Pfeiffer, Giinter, and Bernhard Wieland. 1990. Telecommunications in West Germany.
Berlin: Springer-Verlag.
Purton, Peter. 1990. Siemens to Go East. Communications International, February, pp.
7-28.
Roth, Terrence. 1990. West German Firm to Supply Phones for East Germany. Wall
Street Journal, March 12, p. A7.
Rural Electrification Administration. 1988. A Brief History of the Rural Electric and
Telephone Programs. Washington, D.C.: U.S. Government Printing Office.
Sax, E. 1887. Die Verkehrsmittel in Volks-und Staatswirtschaft. Vienna: Grundlegung
der theoretischen Staatswirtschaft.
Schares, Gail. 1989. Siemens: A Plodding Giant Starts to Pick Up Speed. Business
Week, February 20, p. 136-37.
Schenker, Jennifer. 1990. German Angst. Communications Week International, January
29, p. 1.
Scherer, Joachim. 1985. Telekommunikationsrecht und Telekommunikationspolitik. Baden-
Baden: Nomos.
Schnoring, M. 1982. Direkte und indirekte Investitionsaufwendungen der Deutschen
Bundespost fur Fernmeldeanlagen auf die Produktion und die Beschaftigung in
der Bundesrepublik Deutschland. Berlin: DIW.
Schulte, Josef. 1982. Endgeratekonzeption in Fernsprechdienst der Deutschen Bundes-
post. In Telekommunikation in der Bundesrepublik Deutschland 1982, ed. Die-
trich Elias. Heidelberg: R. V. Decker's Verlag, G. Schenck.
Schwarz-Schilling, Christian. 1990. Germany. Telephony, January 22, p. 58.
Scott, John Dick. 1958. Siemens' Brothers, 1858-1958: An Essay in the History of the
Industry. London: Weidenfeld and Nicholson.
Siemens, Georg. 1949. Geschichte des Hauses Siemens. 3 vols. Munich: Verlag Karl
Alber.
. 1957. History of the House of Siemens, English Translation. Munich: Verlag
Karl Alber.
Siemens, 1989. Keeping You Posted. Siemens Press Release. January 5.
Snow, Marcellus S. 1983. Comparative Policy Research and Policy Making in Tele-
communications: The Case of West Germany. Paper prepared for the Eleventh
Annual Telecommunications Policy Research Conference. Annapolis, MD.
Stuebing, Heinz Volkmar. 1982. Der Deutsche Kraftwagenpostpersonenverkehr in sei-
nem rechtlichen Aufbau. Weimar. Ph.D. dissertation.
Tagliabue, John. 1984. Nixdorf takes on the Americans. New York Times, September
16, p. F4.
Telekom. 1990. Telecommunications "Made in Germany" Annual Report. Deutsche
Bundespost Telekom.
Thiemeyer, Theo. 1983. Deregulation in the Perspective of the German Gemeinwirt-
schaftslehre. Journal of Institutional and Theoretical Economics 139:40518.
U.S. Department of State. 1990. Eastern Europe: Please Stand By. Washington, D.C.:
U.S. Department of State.
References 465
von Weizsacker, Carl Christian. 1980. Wirtschaftspolitische Begriindung und Abgren-

zung des Fernmeldemonopols. In Kommunikation ohne Monopole, ed. Ernst-
Joachim Mestmacker. Baden-Baden: Nomos.
Wagner, Adolf. 1887. Finanzwissenschaft und Staatssozialismus. Zeitschrift fur die
Gesamte Staatswissenschaft, 43: 37-122, 675-746.
Witte, Eberhard. 1987. Neuordung der Telekommunikation. Heidelberg: R. V. Decker's
Verlag, G. Schenck.
Wittiber. 1934. Die Entwicklung und Neuregelung des Fernsprechnebenstellenwesens
in Deutschland, Deutsche Verkehrszeitung 1934. Nv. 19. As quoted in Mest-
macker, 1980, p. 184.
Wolf, Roswitha. 1982. Telefonladen und TelefonmobilTeile des Marketing-Konzepts
der Deutschen Bundespost. In Telekommunikation in der Bundesrepublik
Deutschland, ed. Dietrich Elias. Heidelberg: R.V. Decker's Verlag, G. Schenck,
p. 26.
ZVEI. 1983. Telecommunications in the Federal Republic of Germany, Frankfurt: Fach-
verband Fermeldetechnik.
8. The United Kingdom
Abel, Glen. 1989. BT is Planting Global Seeds. Communications Week International,

May 29.
Arlandis, Jacques, and Laurent Gille. 1989. Performance Under Different Regulatory
Regimes: A Comparison of British Telecom and France Telecom. Montpellier:
IDATE.
Beesley, Michael E. 1981. Liberalization of the Use of the British Telecommunications
Network. Report to the Secretary of State. London: Her Majesty's Stationery
Office.
Beesley, Michael, and Bruce Laidlaw. 1990. The Future of Telecommunications. Lon-
don: Institute of International Affairs.
Bird June, and John Huxley. 1984. Tomorrow the World? The New York Times, Octo-
ber 7.
British Telecom. 1981. Further Considerations Relating to the British Telecommunica-
tions Network and Proposal to Permit Competition. London: British Telecom.
Bruce, Robert R., Jeffrey P. Cunard, and Mark D. Director. 1986. From Telecommu-
nications to Electronic Services: A Global Spectrum of Definitions, Boundary
Lines, and Structures. London: Butterworths.
Cable and Wireless. 1990. Report and Accounts, p. 8
Connections. 1985. British Telecom License Out of Reach? May 24, p. 32.
Court of Justice of the European Communities. 1985. Judgement of the Court of 20
March, 1985 in Case 41/83: Italian Republic of Commission of the European
Communities. Reference No. 85/c96/06, April 17.
Department of Industry. 1982. The Future of Telecommunications in Britain. White
Paper. London: Department of Industry.
Department of Trade and Industry. 1985. Government Proposals for the Future Licens-
ing of Value Added and Data Services. Unpublished document of the Depart-
ment of Trade and Industry.
. 1987. Registrations Under the VANs General License OSAT. Unpublished
document. February.
Dixon, Hugo. 1990a. Freeing the Phone Networks. Financial Times, February 5, p. 20.
466 References
. 1990b. U.K. Watchdog Launches Plan to Cut International Telephone Charges.

Financial Times, October 2, p. 24.
, and Alan Cane. 1990. STC to Use Fujitsu's 743m for Expansion in Telecoms.
Financial Times, July 31, pp. 1, 22.
Dumey, R. 1983. Telecommunications and the BT Case Competition Rules. Unpub-
lished speech given at an Antitrust Luncheon. Commission of the European
Communities, Directorate General for Competition. Brussels.
The Economist. 1985. A Hull of a Difference. November 16, p. 90.
. 1989. Telephones That Get Up and Go. September 16, p. 71-72.
. 1990a. Gnatwork. March 31, p. 56.
. 1990b. Duopoly Rhymes with Monopoly. July 7, p. 72.
Evagora, Andreas. 1990. Cable TV-Telecoms Showdown. Communications Week In-
ternational, May 21, p. 16.
Financial Times. 1990. British Telecom Pushes Up Prices. July 19, p. 6.
Foreman-Peck, James. 1985. Competition and Performance in the UK Telecommuni-
cation Industry. Telecommunications Policy, 9(3): 215-28.
Fortune. 1990. August 28, p. 15.
Fuller, Claire, and Pam Mitchell. 1986. A "Jolly Good" Start for Cellular. Telelocator,
March, pp. 42-46.
Garnham, Nicholas. 1985. Telecommunications Policy in the United Kingdom. Media,
Culture and Society. London: Sage, vol. 7, pp. 7-29.
Garrett, John. 1984. Tacs: The UK Standard for Cellular Radio. Telephony, July 23,
pp. 58, 62.
Gebhardt, H. P. 1989. Analysis of the Present Situation and Future Trends in Telecom-
munications Regulation in the Member States in the Light of EC Policy. Brus-
sels: EC.
Gist, Peter. 1990. The Role of Oftel. Telecommunications Policy, February, pp. 26-
51.
Green, Jeremy. 1989. BT Looks Westward. Communications International, September,
p. 8.
. 1990a. At Last, UK ISDN. Communications International, January, p. 10.
. 1990b. PCN: A Role for ETSI? Communications International, January, p. 14.
Hayes, Dawn. 1988. Carrington's Mobile Vision. Communications Week International,
September 12.
. 1989a. U.K. Spurs New Competition. Communications Week Interna-
tional, July 26, p. 3.
. 1989b. BT to Raise Leased Line Prices. Communications Week International,
October 30, p. 3.
. 1989c. BT to Run Video Trial Over Fiber. Communications Week Interna-
tional, November 27.
. 1990a. BT Reorganizes to Meet Competition. Communications Week Inter-
national, April 9, p. 2.
. 1990b. PCN Market Faces Sobering Reality. Communications Week Interna-
tional, July 16, pp. 10-11.
Heuermann, A. and K. H. Neumann. 1983. Die Neue Fernmeldepolitik Grossbritan-
niensEine oekonomische Darstellung. Bad Honnef: Wissenschaftliches Institut
Fur Kommunikationsdienste der Deutschen Bundespost.
Hudson, Richard L. 1987. British Telecom's Modernization Falters. Wall Street Jour-
nal, August 27.
ITU. 1990. Annual Telephone Statistics. Geneva: ITU.
References 467
Jason, Chris. 1985. Mercury Dials the Right Number. Financial Times, October 2,
p. 4.
Jones, Phil. 1989. UK Frees Leased Line Resale. Communications International, July,
p. 5.
Journal of Commerce. 1983.
Laws, Malcolm. 1989. U.K. Vendors Target New Network Niche. Communications
Week International, July 26, p. 1.
Littlechild, Stephen. 1983a. Deregulation of UK Telecommunications: Some Economic
Aspects. Economic Review 1(2): 29.
. 1983b. Regulations of British Telecommunications Profitability. London: De-
partment of Industry.
Locksley, Gareth. 1982. The EEC Telecommunications Industry: Competition, Concen-
tration and Competitiveness. Brussels: Commission of the European Communi-
ties.
. 1983. Europe and the Electronics Industry: Conflicting Strategies in Positive
Restructuring. West European Politics 6(2): 128-38.
Lynch, Karen, and Dawn Hayes. 1990. Telepoint Operators Refocusing. Communica-
tions Week International, July 16, p. 11.
Manning, D. 1988. Telecommunications in the United Kingdom. In European Telecom-
munications Organizations, ed. James Foreman-Peck and Jiirgen Miiller. Baden-
Baden: Nomos.
Maremont, Mark. 1989. British Telecom Is Getting Less British All the Time. Business
Week, August 14.
Monopolies and Mergers Commission, The General Electric Company, PLC, and The
Plessey Company. 1986. PLC: A Report on the Proposed Merger. London: Her
Majesty's Stationery Office.
Morgan, Kevin. 1987. Breaching the Monopoly: Telecommunications and the State in
Britain. Working Paper Series on Government-Industry Relations. No. 7. Uni-
versity of Sussex.
Noam, Eli M. 1982. Towards an Integrated Telecommunications Market: Overcoming
the Local Monopoly of Cable Television. Federal Communications Bar Journal,
34(2): 209-57.
. 1985. Video Media Competition. New York: Columbia University Press.
. 1986. The "New" Local Communications: Office Networks and Private Cable.
Computer Law Journal 6: 901-42. New York: Columbia University Press.
Official Journal L360. 1982. Congressional Record, U.K. December 21, p. 36
Oftel. 1985a. British Telcom's Price Changes. Oftel press notice. December 16.
. 1985b. Professor Carsberg Congratulates BT for Achieving Call Box Target.
Press release. April 20.
. 1988. Chatlines and Other Message Services. Statement from Professor Bryan
Carsberg. July 19.
. 1990a. Director General of Telecommunications Publishes Sixth Annual Re-
port. June 27.
. 1990b. PCN Licenses. Oftel News, March.
Purton, Peter. 1989. Slow Progress of VANS. Communications International, July, pp.
27-32.
. 1990. Old Player Returns to UK Telecom. Telephony, March 26, p. 14
Raggett, R. J. 1986. Racal Vodaphone Fights for Its Share. Telephony, April 28.
Redux, Huber. 1989. BT Ventures Proliferate as International Markets Complicate.
Telecommunications, September, pp. 57-58.
468 References
Reiter, Aian. 1989. Focus on Technology. Telocator, July, pp. 36-42.

Sims, Calvin. 1989. Licenses Awarded by Britain for 3 New Phone Networks. The New
York Times, December 12.
Staple, Greg. 1990. The Global Telecommunications Traffic Boom. London: Interna-
tional Institute of Communications.
Stapley, Barry. 1981. Managing CommunicationsThe Value of Choice. Telecommu-
nications Policy 5(2): 149-51.
Telecommunications Reports International. 1990. Seven UK Paging Operators Sign MOU
on ERMES Message System. February 16, p. 7.
Thomas, David. 1987. Mercury Communications: Making Good Headway. Financial
Times, October 19.
Williamson, John. 1986. Cellnet Intends to Keep Its Lead. Telephony, April 28, pp.
37-40.
. 1987. ISDN in the UK: Enthusiasm Overshadows Skepticism as Private Appli-
cations Emerge. Telephony, April 27.
Woollacott, Emma. 1990a. BT Slims and Prices ISDN. Communications International,
May, p. 7.
. 1990b. Small Screen Services. Communications International, February, pp.
41-43.
Young, Peter. 1983. Power of Speech: History of Standard Telephones and Cable 1883
to 1983. Winchester, MA: Allen and Unwin.
9. France
Alletier. 1973. Die Telegraphenstation Koeln-Flittard, Eine Kleine Geschichte der Na-
chrichtentechnik. Rheinisch-Westfaelisches: Wirtschaftsarchive zu Koeln.
Arlandis, Jacques and Laurent Gille. 1989. Performance Under Different Regulatory
Regimes: A Comparision of British Telecom and France Telecom. Montpellier:
IDATE.
Attali, Jacques, and Yves Stourdze. 1977. The Birth of the Telephone and Economic
Crisis: The Slow Death of Monologue in French Society. In The Social Impact
of the Telephone, ed. Ithiel de Sola Pool. Cambridge, MA: MIT Press, pp. 97-
111.
Aurelle, Bruno. 1986. Les Telecommunications. Paris: Editions de la Couverte.
Benedetti, Marino. 1983. Telecommunications Services: Policy in Europe. Paper pre-
sented at the European Institute of Public Administration, Maestricht/Valkerburt.
Bertho, Catherine. 1987. Origins of the Organization of Telecommunications in France.
New York: Columbia Institute for Tele-Information. Working Paper Series.
, et al., ed. 1984. Histoire des Telecommunications en France. Toulouse: Edi-
tions Eres.
, and Michelle Nouvion. 1986. In Les Telecommunications, Bruno Aurelle. Paris:
Editions de la Couverte.
Bienaim, Jean-Pierre, and Michel Picaud. 1984. Telecommunications et le IXeme (ninth)
Plan. Revue Fran^aise des Telecommunications, January 19, pp. 35-41.
Bonan, M., Dominique Roux, and Jean Michel Delbarade. 1985. L'importance de Tele-
communications sur la Croissance Economique. Paris: Universite de Paris-
Dauphine.
Boult, Raymond. 1989. Users Back PTT Split. Communications International, October,
pp. 13-14.
References 469
. 1990. Alcatel Takes Military Turn. Communications International, March, pp.

10-12.
Bransten, Thomas R., and Stanley H. Brown. 1964. Machines Bull's Computer Crisis.
Fortune, July, pp. 154-55, 242-44.
Bright, Julian. 1990. The Smart Card: An Application in Search of a Technology. Tele-
communications 24(March) 3.
Bruce, Robert R., Jeffrey P. Cunard, and Mark D. Director. 1985. Country Report:
France. In Study of Telecommunications Structures. Washington, D.C.: Debe-
voise & Plimpton, p. 34.
Business Week. 1982. Thomson: First Test of a Socialist Theory. May 5, p. 53.
Chamoux, Jean-Pierre. 1988. The Current-French Telecommunication Policy. Mission
a la Reglementation.
Connaughton, Bemadette. 1982. The French Telecommunications Industry. Unpub-
lished paper. University of Pennsylvania, Wharton School of Finance.
Crane, Rhonda J. 1979. The Politics of International Standards: France and the Color
TV War. Norwood, NJ: Ablex.
Curien, N., and J. de la Brunetiere. 1984. Les Transferts de Revenus Induits par la
Tarification Telephonique entre Categories d'Abonnes et entre Types de Presta-
tions. Telecommunications, vol. 39. pp. 11-12.
Darmon, Jacques. 1985. Le Grand Derangement: La Guerre du Telephone. Paris: Edi-
tions Jean Claude Lattes.
Dawkins, William. 1990a. France Makes a Point of Catching Up in the Mobile Phone
Race. Financial Times, March 1.
. !990b. Shackles to Be Removed. Financial Times, June 26, p. 11.
Delamaide, Darrell. 1982. Will the Socialist Experiment Work? Institutional Investor.
March, pp. 161-75.
The Economist. 1967. Knocking Microheads Together. December 23, p. 1245.
. 1989. Silicon Ballet. February 18.
Gleckman, Howard. 1990. CGE Gains a Chunk of Alcatel. Business Week, June 25, p.
34.
Guerard, Agnes, Guy Lafarge, and Charles Pautrat. 1979. Les Regions dans la Course
au Telephone. Economic et Statistique, 177(December): 37-49.
Hudson, Richard. 1990. Groupe Bull's Loss Widened Sharply During First Half. Wall
Street Journal, July 31.
Locksley, Gareth. 1983. Europe and the Electronics Industry, Conflicting Strategies in
Positive Restructuring. West European Politics 64(2): 128-38.
Logica. 1979. Structure and Performance Comparison, vol. IV (France). Oct. 12:5.
Malerba, Franco. 1985. The Semi-Conductor Business. Madison, WI: University of
Wisconsin Press.
Mclnnes, Neil. 1964. The Story of Machines Bull. Barron's, February 24, pp. 9-21.
Monsen, R. Joseph. 1984. French Socialists March to the Right. Challenge, September/
October, pp. 37-42.
Morley, Meg. 1986. French Telecoms Place Call. Connections, 59(June): 1.
Nouvion, Mireille. 1984. L'Automatisation du Reseau Telephonique Franais. Revue
Franqaise de Telecommunications, January, pp. 76-85.
Parry, John. 1989. Cashless Europe. International Management 44(September): 8.
470 References
Pautrat, Charles. 1984. La Nouvelle Structure Tarifaire. Revue Franfaise des Telecom-
munications, January, pp. 127-34.
Petersen, Thane. 1989. A New CEO Is Taking Bull by the Horns. Business Week, July
17, p. 80.
Pierrand, Jean. 1984. Thomson Voit Grand pour ses "Micro." Le Monde, April 23.
Quatrepoint, Jean-Michel. 1984. CIT-Alcatel n'a Fait qu'une Perce Limite sur les Marches
Europeens. Le Monde, May 15, pp. 20-21.
Rice, Valerie. 1989. It's Now or Never for Europe's Biggest Chip Makers. Electronic
Business 15(August 7): 16.
Roussel, Anne-Marie. 1989a. French Clarify VAN Regulation. Communications Week
International, May 15, p. 13.
. 1989b. French Revision. Communications Week International, May 29,
p. 22.
. 1989c. The Top 25. Communications Week International, October 24, p. CIO.
. 1989d. Quiles Offers Reform Plan. Communications Week International, No-
vember 13, p. 13.
. 1990a. French Reform Still Hotly Debated. Communications Week Interna-
tional, April 9, p. 3.
. 1990b. Reforms Gaining Ground. Communications Week International, May
21, p. 3.
. 1990c. French Senators Want More Reform. Communications Week Interna-
tional, June 18, p. 13.
. 1990d. Report Warns of Foreign "Invasion." Communications Week Interna-
tional, August 13.
Schenker, Jennifer. 1989a. Bull Reorganizes in Aftermath of Merger. Communications
Week International, February 6, p. 13.
. 1989b. New Minister on the Spot in France. Communications Week Interna-
tional, July, p. 24.
. 1989c. Users Combat French Initiative. Communications Week International,
November 27, p. 2.
. 1990. Transpac Ready to Branch Out. Communications Week International,
January 29, p. 2.
Scientific American. 1983. Special Issue on French Telecommunications. October.
Simon, Jean-Paul. 1990. Aftermath: Deregulation in France in the Eighties. Paper pre-
sented at the 18th Annual TPRC, Airlie VA. September.
Steckel, Marie-Monique, and Marc Fossier. 1990. France Telecom: An Insider's Guide.
Chicago: Telephony.
auf die Gegenwart. Berlin: Verlag der Koniglichen Geheimen Ober-Hofbuch-
druckerei (R. V. Decker).
Telecom France. 1990. Revue Franqaise des Telecommunications. International Edition.
April.
Telecommuncations Reports. 1990. National Assembly Approves Government Plan to
Separate France Telecom, Postal Service. May 28.
Telephony. 1985. A Partnership in Progress. February 11, pp. 32-33.
Tully, Shawn. 1989. Europe Goes on a Telephone Binge. Fortune, August 28.
U.S. Department of Commerce, Bureau of the Census. 1939. Cited in Telephone and
Telegraph Statistics of the World, January, 1938, American Telephone and
Telegraph Co. Comptroller's Department, Chief Statistician's Division. Wash-
ington, DC: Government Printing Office.
References 471
U.S. Department of Commerce, Domestic and International Business Administration.

1975. Communications Equipment and Systems-France. September.
Vasseur, Frederic. 1984. Bull: C'est Reparti pour un Tour. L'Expansion, April 20/May
10, pp. 39-40.
Voge, Jean. 1986. Survey of French Regulatory Policy. In The Marketplace for Tele-
communications: Regulation and Deregulation in Industrial Democracies, ed.
Marcellus Snow. White Plains, NY: Longman.
Western Electric. 1978. Study: Telecommunication Services in France.
10. The Netherlands
Dabbs, P., J. J. E. Swaffield, C. M. Aust, and I. Sarwar. 1982. Going Dutch! British
Telecom Journal. Spring.
de Pous, Victor. 1988. Dutch Privacy Bill Again Delayed. Transnational Data and
Communications Report. December, pp. 6-7.
. 1989. Dutch Data Protection Act in Force. Transnational Data & Communica-
tions Report. December, pp. 21-22.
Forden, Sara Gay. 1990. AT&T-Italian Joint Venture Targets and U.K., Greece, Por-
tugal, Spain Markets. The Wall Street Journal, July 28, pp. A3E, 3.
Hayes, Dawn. 1989. Suppliers Split Over GSM. Communications International, Sep-
tember 18, p. 1.
Hins, Wouter, and Bernt Hugenholtz. 1986. The Law and Economics of Transborder
Telecommunications: Report on the Netherlands. Hamburg: Max-Planck Institut.
Kingdom of the Netherlands. 1981. Facts and Figures: Transport and Communications.
Amsterdam: Ministry of Foreign Affairs.
Lablans, Peter. 1989. Privatization of the Dutch PIT: New Telecommunications Op-
portunities. Telecommunications Business 1(3): 61-66.
Malerba, Franco. 1985. The Semiconductor Business. Madison: University of Wisconsin
Press.
Montgomery, Page, Lee Selwyn, and Paul Keller. 1990. The Telecommunications In-
frastructure in Perspective. Economics & Technology, Inc. March.
Netherlands PTT. 1985. PTT in Motion. Amsterdam: Netherlands PIT.
Philips, Frederik. 1978. Forty-Five Years With Philips. Dorset, UK: Blandford Press.
Prokesch, Steven. 1990. Philips Sees $1 Billion Loss in Reorganizing 2 Divisions. New
York Times. July 3, pp. D1-D2.
PTT Telecommunicatie. 1983. The Dutch Telecommunications Service in Brief. Am-
sterdam: PTT Telecommunicatie.
Schenker, Jennifer. 1989. Dutch PTT Enters Datacoms Market. Communications Week
International. October 16, p. 87.
. 1990. Judging the PTTs. Communications Week International. July 16, pp.
Cll-12.
Staple, Gregory. 1990. The Global Telecommunications Traffic Boom. London: IIC.
Tagliabue, Paul. 1987. New York Times. January 15, p. F8.
Transnational Data and Communications Report. (TDR). 1989. Equipment Approval
Row Mars Start of Dutch PTT. February, p. 3.
472 References
Tutt, Nigel. 1990. Belgium Belies Status as Netherlands Reforms. Communications

International, July, pp. 15-19.
Van de Krol, Ronald. 1990. Philips Reorganises After Rationalisation Costs. Financial
Times. July 3, p. 1.
The Wall Street Journal. 1989. April 4, pp. A19, 6.
. 1989. March 9, pp. A13, 4.
Wieland, Bernhard. 1986. Die Neuordnung des Fernmeldewesens in den Niederlanden.
Bad Honnef: Deutsche Bundespost, Wissenschaftliches Institut fuer Kommuni-
kationsdienste.
. 1988. Telecommunications in the Netherlands. In European Telecommunica-
tions Organizations, ed. James Foreman-Peck and Jiirgen Miiller. Baden-Baden:
Nomos, pp. 203-19.
Wilde, Candee. 1990. Philips Pulls Out of AT&T Venture. Communication Week Inter-
national. October 1.
11. Belgium
Belgian National Committee for the World Year of Communications. 1983. Belgian
Communications Technology, November.
Cheeseright, Paul. 1985. Key Decisions Still Awaited. Financial Times, January 14,
p. 8.
Euro-Telecom News. 1987. Belgium. 1(10).
ITU. 1990a. Annual Telephone Statistics. Geneva: ITU.
. 1990b. European Digital Mobile Telephone Network: Belgium Chooses System
Suppliers. Telecommunications Journal 57(3): 191.
Kellaway, Lucy. 1990. Don't Ring Us. ... Financial Times, June 18, p. 4.
Miiller, Jiirgen. 1988. Telecommunications in Belgium. In European Telecommunica-
tions Organizations, ed. James Foreman-Peck and Jiirgen Miiller. Baden-Baden:
Nomos, pp. 87-102.
Pichault, Frangois. 1985. La Telematique dans le Cadre Reglementaire et Institutionnel
de la Belgique. Courier Hebdomadaire, CRISP, December 6.
Poullet, Yves. 1989. The Belgian Telecommunications Case. Presented at FNRS Inter-
national Conference on Telecommunications. Brussels. January 20.
RTT. 1985. Current Situation and Development of Telecommunications in Belgium.
Telecommunication Journal, pp. 52, 191.
. 1987. Belgium Telecommunications. Brussels: RTT.
Schenker, Jennifer. 1990. Judging the PTTs. Communications Week International, July
16, p. C13.
Tutt, Nigel. 1989. A Question of Balance. Communications International, August, pp.
21-26.
. 1990. Belgium Belies Status as Netherlands Reforms. Communications Inter-
national, July, pp. 15-19.
U.S. Department of Justice. 1987. The Geodesic Network: 1987 Report on Competition
in the Telephone Industry (The Huber Report). Washington, D.C.: Government
Printing Office. January.
References 473
12. Luxembourg
Bode, Leon. 1985. Les Origines et ['Extension du Telephone au Grand-Duche de Lux-

embourg (1884-1920). Luxembourg: Administration des Posts et Telecommu-
nications.
16, pp. C12-C13.
13. Switzerland
Association Suisse d'Usager de Telecommunications (ASUT). 1983. ASM Bulletin, No.

28, October, pp. 108-9, 114.
Blankart, Charles D., and Friedrich Schneider. 1984. Schweizer Fernmeldetarife in In-
ternationalen Vergleich. ASUT Bulletin, No. 32, December.
Blankart, Charles B., and Giinter Knieps. 1985. Die Herausforderungen des Techn-
ischen Fortschrittes an das Fernmelderecht. Neue Ziircher Zeitung, November 1,
p. 37.
Biitikofer, Jean-Fredy. 1987. The Swiss Response to a Changing Telecommunications
Environment. Speech given at the Annenberg Washington Forum. Berne, Octo-
ber 29.
Ducommun, Maurice. 1987. SWISSNET-the Swiss ISDN. Bulletin Technique PTT Beme:
General Directorate of the Swiss PTT.
, and Frangois Keller. 1980. Telecommunications Today and in the Near Future.
Bulletin Technique PTT, October.
Hofmann, Max. 1980. The Telephone and Its Significance to Switzerland's First Sub-
scriber, Bulletin Technique PTT, No. 10, pp. 390-92.
ITU. 1989. More and More Main Stations in the World Telephone Network. Telecom-
munications Journal 56(7): 764.
Knieps, Giinter. 1985a. Entstaatlichung itn Telekommunikationsbereich. Tubingen: Mohr,
pp. 167-81.
. 1985b. Die Notwendigkeit einer Reform des Telegrafen-und Telefonverkehrs-
gesetzes. Neue Ziircher Zeitung, April 9.
Kobelt, Christian. 1980. 100 years of Telephone Service in Switzerland. Bulletin Tech-
nique PTT, No. 10, pp. F344-63.
Northern Business Information. 1990. European Telecom Market. New York: McGraw
Hill.
Piquet, Jacques. 1980. Cooperation of Switzerland's Telecommunications Industry with
the PTT. Bulletin Technique PTT. October.
Purton, Peter. 1990. More than a Crossroads. Communications International, June, pp.
13-18.
Staple, Gregory. 1990. The Global Telecommunications Traffic Boom. London: Inter-
national Institute of Communications.
Swiss PTT. 1983. Telecommunications in Switzerland, Swiss Posts, Telephones, and
Telegraphs.
Trachsel, Rudolf. 1987. Pending Changes in Telecommunications Policy: The United
States, Europe, and Japan. Keynote Speech, Annenberg Washington Forum. Beme.
October 29.
474 References
Union Bank of Switzerland. 1986. Telecommunications Industry in Switzerland. Indus-

try Studies.
Wiihrmann, Karl. 1989. Swiss PTT's Approach to the European Market Unification.
Bulletin Technique PTT. Berne: General Directorate of the Swiss PTT.
. 1990. Switzerland. Telephony, January 22, p. 56.
14. Austria
Bauer, Johannes M. 1986. Telekommunikation in Osterreich: Innovationspolitik in ei-

nem dynamischen Infrastrukturbereich. In Stahl und Eisen Bericht: Industrie und
staatliche Politik in Ostereich, ed. M. Scherb and I. Morawetz. Vienna: Verlag
fiir Gesellschaftskritik.
, and Michael Latzer. 1987. Okonomische Analyse des Osterreichischen Tele-
kommunikationsektors. In Technikbewertung neuer Telekommunikationsdienste.
Vienna: Osterreichische Akademie de Wissenschaften.
, and Michael Latzer. 1988. Telecommunications in Austria: Past Developments
and Prospects for Change. In European Telecommunications Organizations, ed.
James Foreman-Peck and Jiirgen Miiller. 1988. Baden-Baden: Nomos, pp. 53-
82.
Bruckner, Georg. 1982. Background Report: Austria. Committee for Information, Com-
puter Communications Policy, Special Session on International Implications of
Changing Market Structures in Telecommunications Services. Paris: OECD.
Communications Week International (CWI). 1990. Austria Modernizing Telecoms. May
21, p. 4.
Eward, Ronald. 1984. The Competition for Markets in International Telecommunica-
tions. Dedham, MA: Artech.
Holcombe, A. N. 1911. Public Ownership of Telephone on the Continent of Europe.
Montgomery, Page, Lee Selwyn, and Paul Keller. 1990. The Telecommunications In-
frastructure in Perspective. Boston: Economics & Technology, Inc., March.
PTV. 1984. Geschaftsbericht 1984. Vienna: PTV.
Purton, Peter. 1990. More Than a Crossroads. Communications International, June, pp.
13-19.
Sindelka, Josef. 1988. Austria. Telephony, February 22, p. 46.
. 1990. Austria. Telephony, January 22, p. 50.
Smith, Gordon. 1984. Can Austria Meet the Challenge of the Electronics Industry?
Focus on Europe, June 25, pp. 12, 14.
Transnational Data and Communications Report (TDR). 1988. European Consumers
Voice Telephone Goals. October, p. 9.
15. Sweden
Boan, Kelley. 1990. The Future of Co-Location in Sweden: An "Unoffical" View.

Presented at Central Office Interconnection Conference, Washington DC.
Brown, David. 1985. Televerket's Monopoly Remains Impregnable. Financial Times,
14 (January).
Connections. 1984. Corporate Report. Ericsson. December 3.
References 475
Dabbs, P., J. J. E. Swaffiekl, C. M. Aust, and I. Sarwar. 1982. Going Dutch! British
Telecom Journal, Spring.
Evagora, Andreas. 1990. Privatization Proposed. Communications Week International,
October 1, p. 10.
Gilhooly, Denis. 1990. Swedes Opening Resale Market. Communications Week Inter-
national, June 4, p. 1.
Gleiss, Norman. 1983. Connection of Privately Owned Equipment. Tele, XXXVI(21).
English edition.
Gustafsson, Karl Erik. 1987. Televisioneren studie i branschutveckling. Kristianstad:
Raben & Sjogren.
Heimburger, Hans. 1931. The State Telephone System 1881-1902. Gothenburg.
ITU. 1990a. Annual Telephone Statistics. Geneva: International Telecommunications
Union.
. 1990b. Swedish Telecom Introduces Information and Entertainment Service.
Telecommunications Journal 57(4): 258.
Krzywicki, John. 1990. Japan and Sweden: Comparative Regulatory Frameworks. Cam-
bridge Strategic Management Group. Presented at the Telecommunications Re-
ports Price Caps Pre-Conference Seminar. July 13.
Ministry of Transport and Communications. 1987. Telecom in Change. Stockholm.
June.
Mitchell, Bridger M. 1983. The Cost of Telephone Service: An International Compari-
son of Rates in Major Countries. Telecommunications Policy. March.
OECD. 1985. Report to the Committee for Information, Computer and Communications
Policy. Special Session on Telecommunications Policy. Paris.
. 1987. The Swedish Telecom Market. Report No. 1437. Unpublished paper.
Paris. May.
. 1989. Telecommunication Network-Based Services: Policy Implications. Paris.
O'Dwyer, Gerauerd. 1990a. Nordic Free Market Looms. Communications Interna-
tional, April, pp. 5-6.
. 1990b. Televerket to Face Appeal. Communications International, June, pp.
6-7.
Qvortrup, Lars. 1989. The Nordic Telecottages. Telecommunications Policy, March,
pp. 59-68.
Satellite Communications. 1988. Comvik Receives Permission to Transmit to North
America. December, p. 8.
Siemens. 1985. Study on National Tariffs Worldwide.
Swedish Telecommunications Authority. 1987. Facts About Televerket: 1987. Farsta:
Televerket Information Unit.
Telephony. 1984. A Commitment to Quality. October 7, p. 7,
. 1987. Ericsson Divests Office Equipment Firm. October 26, p. 26.
Televerket. 1983. Organization and Operation of the Swedish Telecommunications Ad-
ministration. Stockholm: Televerket.
. 1984. Annual Report. Stockholm.
. 1989. Who Represents Europe? Paper presented at Assessment 1989. Luxem-
bourg. May 11.
476 References
Thorngren, Bertil. 1990. The Swedish Road to Liberalization. Telecommunications Pol-

icy, April, pp. 94-98.
Whitehouse, Bob. 1989. The Swedes Try Out Liberalization. Communication Interna-
tional, November (supplement): pp. 2-7.
16. Finland
Association of Telephone Companies in Finland. 1986. Facts About Telephony in Fin-

land, 1986. Helsinki: ATCF.
AT&T. 1975. Calling the World. New York: AT&T Long Lines.
Bruce, Robert R., Jeffrey P. Cunard, and Mark D. Director. 1986. From Telecommu-
nications to Electronic Services. Telecommunications Policy 10(1).
Communications Systems Worldwide. 1987. Firms Decide Soon on Open Networks.
September.
The Economist. 1985. Privatization: Everybody's Doing It Differently, December 21,
pp. 71-86.
Evagora, Andreas. 1990. Finland Serving FDDI. Communications Week International,
July 16, p. 4.
Finland PTT. 1983.Postal Service Today. Helsinki: PTT.
. 1985.Telecommunications. Helsinki: PTT.
Karvonen, Mati. 1984. Ah, Wilderness . . . Telephones!Telephony, May 28, pp. 52-
63.
Myllo, Heikki. 1984. Telecom Reform Legislation Nears Completion. Telephony, June
25, pp. 58-63.
Nokia. 1989. Nokia Annual Report, pp. 3-19.
Northern Business Information. 1990. New York, McGraw-Hill, p. 79.
O'Dwyer, Gerard. 1989a. PTT Rival. Communications International, December, pp.
8-11.
. 1989b. Finnish Ministers Back Nokia. Communications International, Novem-
ber, p. 8.
OECD. 1982. Committee for Information, Computer and Communications Policy. Spe-
cial Session on the International Implications of Changing Market Structures in
Telecommunications Services. Paris: OECD. December 13-15.
Qvortrup, Lars. 1989. The Nordic Telecottages. Telecommunications Policy, March,
pp. 59-68.
Tanhuanpa'a, Arno. 1990. Finland. Telephony, January 22, p. 50.
Telephony. 1990. News Bytes. July 30, p. 25.
Whitehouse, Bob. 1989a. "IT in the ITU" Says Doctor Tarjanne. Communications
International, October, pp. 20-23.
. 1989b. Finland Pushes Forward. Communications International, November,
pp. 21-22.
. 1989c. PTT Rival. Communications International, November, p. 21.
Williamson, John. 1986. The Finnish Telecom Industry: As Unusual as the County
Itself. Telephony, January 27, pp. 50-54.
References 477
17. Norway
Foreman-Peck, James, and Dorothy Manning. 1988. Telecommunications in Norway.

In European Telecommunications Organisations, ed. James Foreman-Peck and
Jiirgen Miiller. Baden-Baden: Nomos.
Green, Jeremy. 1989. Norway's Middle Path. Communications International, Novem-
ber (supplement), pp. 9-18.
Holcombe, A. N. 1911. Public Ownership of Telephone on the Continent of Europe.
ITU. 1990. Annual Telephone Statistics. Geneva: ITU,
Norwegian Telecommunications Administration (NTA). 1988. Annual Report 1987. Oslo.
Nyheim, Jan Henrik. 1984. Norwegian Broadcasting and Telecommunications. The
Norseman 5(November): 39-41.
. 1987. Norway: RegulationWho Needs It? Intermedia 15(4-5): 61-62.
18. Denmark
Barnes, Hillary. 1987. Low-Cost Calls. Financial Times, October 19, p. 29.
Feldman, Mark B., and David R. Garcia. 1982. National Regulation of Transborder
Data Flows. North Carolina Journal of International Law and Commercial Reg-
ulation 7(1): 1-25.
Green, Jeremy. 1989. The Unconventional Ways of Danish Comms. Communications
International, November supplement, pp. 29-32.
NKT. 1987. NKT at Telecom 87, October.
OECD. 1982. Committee for Information, Computer and Communications Policy.
Background Report: Denmark. Unpublished report for Special Session on the
International Implications on Changing Market Structures in Telecommunica-
tions. Paris.
Olsen, O. Jess. 1988. Deregulation and Reorganizationthe Case of Danish Telecom-
munications. Memo 4. Copenhagen: Institute of Economics, University of Co-
penhagen.
Pedersen, Mogens Kuhn. 1987. The State and Telecoms Policy: Convenient or Con-
trived? Unpublished paper.
. 1988. The Danish Telecommunications Equipment Industry. Paper presented
at European Association for Research in Industrial Economics, Rotterdam. Au-
gust.
P&T Directorate General. 1988. Liberalization of the Rules Governing Text and Data
Transmissions on the Danish Telecommunications Network. June 21.
Qvotrup, Lars. 1984. Cable Television: Public Infrastructure or Private Business, Cable
Politics in Denmark. Le Bulletin de I'Idate: Le Prix des Nouveaux Medias
17(October): 173-79.
TeleDenmark. 1987. Strongest Growth Is in International Telecom Traffic.
Transnational Data and Communications Report. (TDR) 1989. Europe Achieving
Telecom Reforms. October, pp. 6-8.
Williamson, John. 1990. Danes Buy Back Local Telcos. Telephony, April 2, p. 16.
478 References
19. Iceland
Eurodata Foundation Yearbook. 1987. London: Eurodata Foundation.
OECD. Committee for Information, Computer and Communications Policy. 1990. Per-
formance Indicators for Public Telecommunications Operators. February, DSTI/
ICCP/TISP/89.10. Paris: OECD.
Post and Telecommunications Administration. 1986. Telecommunications in Iceland.
Tomasson, Olafur. 1990. Iceland. Telephony, January 22, p. 54.
20. Ireland
Central Statistics Office. 1979. Statistical Abstract of Ireland. Vol. 43.

Department of Industry and Commerce of Ireland. 1944. Irish Trade Journal and Sta-
tistical Bulletin. Vol. 19
. 1969. Irish Trade Journal and Statistical Bulletin. Vol. 44.
Dillon, Liam. 1987. Ireland Boosts Its Operational Efficiency. Telephony, December
22.
Ergas, Henry, and Jun Okavana. 1984. Changing Market Structures in Telecommuni-
cation. Paris: OECD.
Evagora, Andreas. 1990a. Irish Suppliers Charge Favoritism. Communications Week
International, May 21, p. 12.
. 1990b. Terminal Equipment: Opening the Door. CWI, September 3.
Garnett, Nick. 1987. Substantial Progress. Financial Times, October 19.
Keenan, Brennan. 1985. Growing-up Pains for State Infant. Financial Times, January
14, p. 7.
Litton, A. J. 1961. The Growth and Development of the Irish Telephone System. Dub-
lin: Institute of Electrical Engineers.
NTIA. 1985. Telecommunications Policies in Ten Countries: Prospects for Future Com-
petitive Access. Washington, D.C.: U.S. Government Printing Office. March.
Raggett, R. J. 1984. The Fall and Rise of Irish Telecommunications. Telephony, Feb-
ruary, p. 31.
Schenker, Jennifer. 1990. Judging the PTT's. Communications Week International, July
16, pp. Cl-16.
Statistical Abstract of Ireland. 1979. Central Statistics Office. Vol. 43.
Telecom Eireann. 1987. Step by Step to the Digital Exchange.
Telephony. 1985. Irish Telecom Chief Quits. June 24, pp. 13, 83.
. 1987. Northern Ireland May Get FO Ring. February 23, p. 26.
Transnational Data and Communications Report. (TOR) 1988. European Consumers
Voice Telephone Goals. October.
. 1989. Europe Achieving Telecom Reforms. October, pp. 6-8.
21. Italy
Antonelli, Cristiano. 1988. The Emergence of the Network Firm. In New Information
Technology and Industrial Change: The Italian Case, ed. Cristiano Antonelli.
Dordrecht: Kluwer.
References 479
Benedetti, Marino. 1983. Telecommunications Services: Policy in Europe. Paper pre-

sented at the European Institute of Public Administration. Maastricht/Valkerburt.
Benzoni, Paolo. 1990. Italy. Telephony, January 22, p. 56.
Bonafield, Christine. 1988. AT&T May Redefine Partnership with Olivetti. Commu-
nications Week, May 30, p. 6.
Colby, Laura, and Richard Hudson. 1989. AT&T Learns Right Choices for Europe.
The Wall Street Journal, February 2, p. 27.
Communications Week International (CWl). 1990. Italians Consort. June 4, p. 3.
Dallmeyer, Martin. 1977. Quellen zur Geschichte des Europdischen Postwesens 1501-
1806. Kallmiinz: Verlag Michael Lassleben.
Del Terra, Laura. 1989. SIP's Four Year Plan. Communications International, Septem-
ber, p. 53.
The Economist. 1989. Meeeow. July 22, pp. 62-63.
Friedman, Alan. 1987. The War of Words that Surrounds Telit. Financial Times, Oc-
tober 19.
Graziosi, Guiliano. 1988. Telecommunications in Italy. Telecommunications Policy,
December, pp. 303-10.
Green, Jeremy. 1989. Italy in the Fast Lane. Communications International, December,
pp. 15-17.
Greenhouse, Steven. 1989. AT&T Deal with Italtel Advances. The New York Times,
February 10, p. D l .
Guyon, Janet, and Laura Colloy. 1988. Firm Declines to Boost Stake in Olivetti. The
Wall Street Journal, April 12, p. 4.
Hayes, Dawn. 1988. Italian Leaders at Odds Over AT&T-Italtel Deal. Communications
Week International, August 22, p. 21.
Italtel. 1989. Guatemala Telecommunications Development Project, Zambia Rural Ra-
diotelephony Project, Zimbabwe 140 Mbit/s Fiber Optic Link. Milano: Italtel
Societa Italiana Telecomunicazioni.
ITU. 1989. Supply of Modems: Liberalization in Italy. Telecommunications Journal
56(8): 533.
. 1990. Annual Telephone Statistics. Geneva: ITU.
La Repubblica, II Grande Salto della SIP. 1987. February 10.
Lynch, Karen. 1990. Italy Struggles with 1992. Communications Week International,
May 21, p. 6.
Ministero delle Poste e delle Telecomunicazioni (MPT). 1984. Piano Nazionale delle
Telecomunicazioni 1985-1994. Rome: Ministero delle Poste e delle Telecomun-
icazioni.
Pozzi, Giovanni. 1987. The Evolution of the Italian Telecommunications System. Milan:
Reseau.
Rosenbaum, Andrew. 1988. Country by Country, The Old Rules Change. Communica-
tions Week International, September 12, p. C4.
Rossant, John. 1988. Does AT&T Have a Real Shot at This Juicy Contract? Business
Week, November 21, p. 59.
Roussel, Anne-Marie. 1988. Italy Weighs Telecom Restructuring. Communications Week
International, November 22, p. 15.
480 References
. 1990. Italians, French Pushing Videotex. Communications Week International,

March 12, p. 17.
Sabel, Charles F. 1982. Work and Politics: The Division of Labor in Industry. New
York: Cambridge University Press.
16, pp. Cl-16.
Smau. 1990. Information Technology. Milan: L'osservatorio Smau, pp. 74-111.
Staple, Gregory. 1990. The Global Telecommunications Traffic Boom. London: IIC.
STET. 1985. Annual Report.
Symonds, Williams. 1987. A Failed Merger Blows Italy's Shot at the Big Time. Busi-
ness Week, November 23, p. 52.
Telephony. 1983. After Ten Years of Neglect, Italian Telecommunications Comes of
Age. September 25-26, pp. 32-37.
Wilson, Carol. 1987. BellSouth lands Italian net contract. Telephony, December 7.
22. Spain
Bradsher, Keith. 1990. Allure of Spain's Phone Company. The New York Times, March,
p. D12.
Burns, Tom. 1985. ITT's Redundancy Plans Unacceptable, Says Spain. Financial Times,
October 1, p. 2.
de Moragas, Miguel, Rosario de Mateo, and Emilio Prado. 1987. Spain. Electronic
Mass Media and Politics in Western Europe, ed. Hans J. Kleinsteuber, Denis
McQuail, and Karen Siune. Frankfurt: Campus Verlag, pp. 251-72.
The Economist. 1985. Hello World, Madrid Calling. May 25.
. 1990. Please Try Again. May 12, pp. 71-72.
Financial Times. 1989. Nynex Sees Gibraltar As Its Telecoms Showcase. October 5.
Lopez-Escobar, Esteban. 1985. Spain Telecom's Monopoly. Intermedia, July/Septem-
ber, pp. 3-4.
Manning, Dorothy, Diego Bader von Jagow, James Foreman-Peck, and Jiirgen Muller.
1988. In European Telecommunications Organisations, ed. James Foreman-Peck
and Jiirgen Muller. Baden-Baden: Nomos.
MarTech Strategies, Inc. 1983. Telecommunication Policies in Ten Countries: Prospects
for Future Competitive Access. Washington, D.C.: Government Printing Office.
Martinez, Miguel. 1990. A Digital Missing Link. Communications International, Feb-
ruary, pp. 18-20.
Morgan, Jeremy, and Charles Mason. 1990. Argentina's Entel Deal Postponed as Banks
Struggle to Meet Terms. Telephony, August 20, p. 8.
Purton, Peter. 1989. What the EC Is Trying to Achieve. Communications International,
July, pp. 28-29.
. 1990. Nineties May Be Fiesta for Spanish Telecoms. Communications Inter-
national, February, pp. 15-17.
Ryser, Jefferey. 1988. Getting South America's #!*% Phones to Work. Business Week,
April 18.
Taylor, Robert. 1990. Ericsson Offshoot Wins Spanish Telecoms Order. Financial Times,
July 6.
Telefonica. 1990. 1989 Annual Report Madrid: Telefonica.
References 481
Warden, Becky. 1990. Playing Safe. Communications International, May, p. 8.

White, David. 1986. Clarification of Game Plan After EEC Entry. Financial Times,
January 6.
Whitehouse, Bob. 1990. Telefonica's Waiting List. Communications International, March,
pp. 13-16.
Williamson, John. 1987. AT&T/Philips Inks Spanish Contract. Telephony, October.
Woolacott, Emma. 1990. Telefonica in Cartel Debate. Communications International,
June.
23. Portugal
Baptista, Jose Viana. 1988. Portugal. Telephony, February 22.

de la Cal, Martha. 1990. Portugal: A Radical Transformation of the Telecommunica-
tions Sector. 7992 Single Market Communications Review 1(4): 12-17.
EUSIDIC. 1990. 7990 PDN Study.
ITT. 1983. ITT in Portugal: The First 50 Years. February, No. 2.
Purton, Peter. 1987. Last European Country Chooses Cellular Standard. Telephony,
December.
Roussel, Anne-Marie. 1990. Portugal to Privatize. Communications Week International,
November 12.
Smith, Diana. 1986. Portugal Ends Battle Over Telephones. Financial Times, August
19.
Transnational Data and Communications Report (TDR). 1988. European Consumers
Voice Telephone Goals, pp. 6-8. October.
. 1989. Europe Achieving Telecom Reforms. October, pp. 6-8.
Warden, Becky. 1990a. Portugal Sets Out Reforms. Communications International, March,
p. 7.
. 1990b. Fresh Face for Privatisation. Communications International, July, p.
10.
24. Israel
Bainerman, Joel. 1988. Israeli Upstarts Target Niches. Communications Week Interna-
tional, November 7, p. 24.
. 1989. The Pragmatic Approach. Communications Systems Worldwide, Septem-
ber, p. 50.
Eres, Beth. 1989. Status and Trends in the use of Computers and Telecommunications
in Israel. Interdisciplinary Center for Technological Analysis and Forecasting,
TelAviv University.
Hai, Shaul. 1988. Israel. Telephony, January 22, p. 62.
Israel Statutes. 1982. 7060: The Telecommunication Law. Jerusalem. August 22.
Louisson, Simon. 1987. After Three Years Success Still on Hold for Bezeq. Jerusalem
Post.
TRC. 1990. Market Report. Bezeq-Israel. World Telecom Daily. February 26, pp. 15-
19.
482 References
25. Turkey
Altay, Sefik. 1987. The Development of Telecommunications in Turkey. October 30.

Anik, Gengiz. 1990. General Directorate of PTT, personal communication. August 29.
AT&T. 1975. Calling the World. New York: AT&T Long Lines.
General Directorate of the PTT. 1990. Brief Report on Telecommunications in Turkey.
August.
OECBt 1982. Committee for Information, Computer and Communications Policy. Spe-
cial Session on the International Implications of Changing Market Structures in
Telecommunications Services. Paris: OECD. December 13-15.
Raggett, R. J. 1986. Emir Outlines an Ambitious Program for the Turkish PTT. Te-
lephony, July 28, p. 42.
TRC. 1990. Market Report: Turkey. World Telecom Daily, April 30, pp. 6-7.
Williamson, John. 1988. Turkey Heads for the Next Decade with a 21st Century Tele-
com System. Telephony, December 26, pp. 21, 24-25.
26. Greece
Efstathiou, James. 1990. Greece Going Digital. Communications Week International,
September 17.
ITU. 1990. Annual Telephone Statistics. Geneva: ITV.
Kotsonis, Stefan. 1990. "Greek Telecoms Shake-up" Communications International
December, pp. 14, 16.
OTE. 1985. Annual Report 1984. Athens.
. 1985. Telecommunications Statistics in Greece '84. Athens.
Schenker, Jennifer. 1990. Judging the PTTs. Communications Week International, June
12, pp. Cl-16.
27. Telecommunications in Eastern Europe
Antono, Miroslav, and B. Kubin. 1990. Czechoslovakia. Bucharest: Federal Ministry

of Posts and Telecommunications.
Avramesou, Aristita, and Sergiu Celac, eds. 1980. Romania: An Encyclopaedic Survey.
Bucharest: Editura Sciintiflca si Enciclopedica, pp. 158-60.
Budapress. 1984. Telephone Network Expansion. XXIII(14). Budapest: Hungarian News
Agency. April 4, p. 8.
. 1988. 420,000 New Telephones. XXVII(32). Budapest: Hungarian News Agency.
August 12, pp. 5-6.
Business Eastern Europe. 1986. December 8, p. 392.
. 1987. October 26, p. 344.
Byrnes, Robert F., ed. 1957. East Central Europe Under the Communists: Yugoslavia.
New York: Frederick A. Praeger, pp. 386-87.
Campbell, Robert W. 1988. The Soviet Telecommunications System. Indianapolis: Hud-
son Institute.
References 483
. 1990. The Soviet Telecom Challenge. Eastern Europe and Soviet Telecom
Report 1(6).
Communications International. 1989. December, p. 12.
Datapro. 1990. Datapro Reports on International Telecommunications. New York:
McGraw-Hill.
Dellin, L. A. D., ed. 1957. East Central Europe Under the Communists: Bulgaria.
New York: Fredrick A. Praeger, pp. 369, 383-84.
Dixon, Hugo. 1988a. UK Group in Soviet Pay Phone Venture. Financial Times, July
5, p. 3.
. 1988b. UK Digital Telecoms Switch for Moscow. Financial Times, October
18, p. 6.
. 1990. The Paranoia Eases. Financial Times, April 19.
Eastern European and Soviet Telecom Report (EESTR). 1990a. Hungarian Telecom:
Catching Up With Europe, 1(3).
. 1990b. Polish Telecom: New Directions, 1(4).
The Economist. 1990. Sorry, Reformer, the Line Is Dead. July 21, p. 2.
Fidler, Stephen. 1990. Western Companies Hold Lifeline for Eastern Europe. Financial
Times, April 17, p. 4.
Financial Times. 1987. October 19, p. XXIV.
Frank!, Judy. 1989. The Telecoms Glasnost. Communications International, July, pp.
39-41.
Gilhooly, Denis. 1989. Hungary Heralds Telecoms Reform. Communications Week In-
ternational, October 30, p. 1.
Goodhart, David. 1989. Siemens to Modernize Soviet Telephones. Financial Times,
August 3, p. 3.
Gulyaev, Yuri V. 1990. Directions in Telecommunications Policy in the USSR. Paper
presented to the International Telecommunications Society. Venice, Italy, March.
Hayes, Dawn. 1990. Cellular Network Proposed for Baltic. Communications Week In-
ternational, June 18, p. 18.
Kelly, Tim. 1990. Telecommunications and Eastern European Economies. Paris: OECD.
Lakicevic, Ognjen, ed. 1970. A Handbook of Yugoslavia. Belgrade: Beogradski Graficki
Zavod.
Lewis, Patrick. 1976. Communications Output in the USSR: A Study of The Soviet
Telephone Systems. Soviet Studies, July, pp. 406-17.
Major, Iva. 1980. The Progress of the Hungarian Telephone. Hungarian Academy of
Sciences: Institute of Economics Series, 17.
Nulty, Timothy. 1990. Considerations in Telecom Investment in Eastern Europe. Wash-
ington, D.C.: World Bank.
Pearce, Alan. 1990. A Capital Question. Network World, April 30.
Poland: A Handbook. 1974. Warsaw: Interpress, pp. 276-77'.
Popescu, Virgil. 1990. Romania. Bucharest: Ministry of Communications.
Pronay, G. 1990. Freedom of Speech. IEEE Review, July/August.
Purton, Peter. 1990. Western Firms Plan Baltic Service. Telephony, June 11, pp. 16,
20.
Rocks, David. 1990. Hungary Grants Second Cellular Net. Communications Week In-
ternational, June 4, p. 7.
Satellite Communications. 1990. July, p. 9.
Schenker, Jennifer. 1990. Upgrade in Czechoslovakia. Communications Week Interna-
tional, July 2.
484 References
Selin, Ivan. 1986. Communications and Computers in the Soviet Union. Signal, De-
cember, pp. 91-95.
Skendi, Stavro. ed. 1956. East Central Europe Under the Communists: Albania. New
York: Frederick A. Praeger.
U.S. Department of State, Advisory Committee on International Communications and
Information Policy. 1990. Eastern Europe: Please Stand By. Washington, D.C.:
U.S. Department of State.
Woollacott, Emma. 1990. Relaxation on a Sliding Scale. Communications Interna-
tional, July, pp. 32-33.
28. The International Organizations of Telecommunications
Aronson, Jonathan, and Peter Cowhey. 1988. When Countries Talk. Cambridge, MA:
Ballinger.
Butler, Richard. 1989. ITU Plenipotentiary Conference: Closing Address. June 30.
Codacovi, Lawrence. 1989. WATTC: Impact on Services. Transnational Data and
Communications Report, June-July, pp. 18-20.
Codding, George A., Jr. 1984. Public Access to International Organizations: the ITU.
Intermedia 12(6).
. 1990. The Nice ITU Plenipotentiary Conference. Telecom Policy, April, pp.
139-50.
, and Anthony M. Rutkowski. 1982. The International Telecommunication Union
in a Changing World. Dedham, MA: Artech
Cowhey, Peter S., and Jonathan B. Aronson. 1985. The Great Satellite Shoot-Out.
Regulation, May-June, pp. 27-35.
Dougan, Diana Lady. 1987. The High Stakes Game of International Standards Setting.
Washington D.C.: U.S. Department of State.
Drake, William. 1988. WATTC-88: Restructuring the International Telecommunication
Regulations. Telecommunications Policy, September, pp. 217-23.
. 1989. The CCITT: Time for Reform? in Reforming the Global Network. Lon-
don: International Institute of Communications, pp. 28-43.
Gibbons, Roger. 1989a. Surveying Europe's Standardization Scene. Communications
International, November, pp. 25-31.
. 1989b. European Standards. Communications International, December, pp. 19-
22.
Gilhooly, Denis. 1989. Finland's Tarjanne Elected ITU Head. Communications Week
International, June 26, p. 41.
. 1990. CEPT Faces Up to Its Changing Role. Communications Week Interna-
tional, March 12, p. 13.
Le Boucher, Eric. 1984. Europe des Telecommunications S'Organise. Le Monde, Jan-
uary 21.
Lynch, Karen. 1990. U.S. Turnaround. Communications Week International, October
29, p. 4.
Northern Telecom. 1980. Nature of the Telecommunications Industry Throughout the
World. Submission to the Canadian Restrictive Trade Practices Commission. May.
Official Journal L360. 1982. Congressional Record, December 21, p. 36.
Quander, Peter. 1982. Internationale Zusammenarbeit? In Telekommunikation in der
Bundesrepublik Deutschland, ed. Dietrich Elias. Heidelberg: R.V. Decker's Verlag,
G. Schenck, pp. 375-90.
References 485
Rutkowski, Anthony M. 1982. The U.S.A. and the ITU: Many Attitudes, Few Policies.
Intermedia 10(4-5).
Satelllite News. 1990. Execs Spar at Philips Satellite IX Conference. March 19.
Schenker, Jennifer. 1989. CEPT Revamping Rates. Communications Week Interna-
tional, February 6, p. 1.
Baden: Nomos.
Schnurr, L. E. 1988. The Single European Act: Towards an Open Market in 1992.
Essex: Essex Institute of Higher Education.
Warden, Becky. 1990. The Curtain Rises on a New Satellite Generation. Communica-
tions International, July, p. 5.
Whitehouse, Bob. 1989. "IT in the ITU" Says Dr. Tarjanne. Communications Inter-
national, October, pp. 20-23.
. 1990. CEPT Accedes Under Siege. Communications International, April,
p. 6.
29. Brussels Takes On the Traditional System

Besen, Stanley M. 1990. The European Telecommunications Standards Institute. Tele-
communications Policy. December.
Commission of the European Communities (CEC). 1982. Euronet-DIANE: Towards a
Common Information Market.
. 1986. EC Competition in Telecommunications. Unpublished document IV/B/1.
March 4.
. 1987. Green Paper on the Development of the Common Market for Telecom-
munication Services and Equipment. COM (87) 290 Final. Brussels. June 30.
. 1988. Directive 88/301. Official Journal of the European Communities. L. 131.
European Space Agency. 1984. European Space Agency, 20 Years of European Coop-
eration in Space. Paris: ESA.
Gilhooly, Denis, and Malcolm Laws. 1990. New Course Set for RACE. Communica-
tions Week International, July 2, p. 1.
Green, Jeremy. 1989. Who Rules on Radio. Communications International, November,
p. 10.
ITU. 1990. A European Market for Telecommunications. ITU Telecommunications
Journal, 57(6): 294.
Lalor, Eamon. 1987. Action for Telecommunications Development. Telecommunica-
tions Policy, June, pp. 115-20.
Roussel, Anne-Marie. 1990. Satcoms Policy Up in the Air. Communications Week In-
ternational, June 4.
Schenker, Jennifer. 1989. Services Clash Could Derail Deregulation. Communications
Week International, May 1, p. 6.
. 1990. Putting off 1992. Communications Week International, July 2, p. 1.
Schulte-Braucks, Reinhard. 1986. European Telecommunications Law in the Light of
the British Telecom Judgment. Common Market Law Review, vol. 23.
Telediffusion de France. 1985. Europe: Vers une Europe de la (Haute) Technologic.
Actualite des Techniques de Communication dans le Monde. June.
Transnational Data and Communications Report (TDR). 1988. Europe Adopts New
Standards Approach. August-September, p. 6.
1989. Euro-Parliament Favors Network Monopoly. February, p. 8.
486 References
van Rosenthal, C. Jansen. 1983. The Implications of Telecommunications for a Euro-

pean Community Information Market Policy. Commission of the European Com-
munities, Directorate General for Telecommunications and Information Indus-
tries and Innovation. Luxembourg.
von Martin, Klaus and Walter S. Kirchen. 1981. Euronet: Growth of A Direct-Access
Information Network. ESTA Bulletin. January.
Woollacott, Emma. 1990. Leased Cost Solution. Communications International, April,
p. 5.
30. Telecommunications Policy as Industrial Policy
Aronson, Jonathan David, and Peter F. Cowhey. 1987. When Countries Talk. Cam-
bridge, MA: Ballinger.
Cane, Alan. 1985. Producers Pin Hopes on New Alliances. Financial Times, June 24,
p. 3.
Dang-Nguyen, Godefroy. 1986. A European Telecommunications PolicyWhich In-
struments for Which Prospects? ENST-Bretagne. August.
The Economist. 1985. Privatization: Everybody's Doing It Differently. December 21.
Ellul, Jacques. 1964. The Technological Society. New York: Knopf.
Ergas, Henry. 1985. Exploding the Myths About What's Wrong. Financial Times, June
26.
Etheridge, James, 1983. Center of Controversy. Datamation 240(June): 31-34.
de Jonquieres, Guy. 1990. Shadows Over the Sunrise Sector. Financial Times, July 25,
p. 12.
Knieps, Giinter, Jiirgen Miiller, and Carl Christian von Weizsacker. 1981. Die Rolle
des Wettbewerbs im Fernmeldebereich. Baden-Baden: Nomos, 1984.
Levin, R. C. 1982. The Semi-conductor Industry. In Government and Technical Prog-
ress: A Cross Industry Analysis, ed. R. R. Nelson. New York: Paramount Press.
Malerba, Franco. 1985. The Semi-conductor Business. Madison: University of Wiscon-
sin Press.
Miiller, Jiirgen. The Benefits of Completing the Internal Market for Telecommunications
Equipment in the Community. Research on the Cost of Non-Europe. Fontain-
bleau: INSEAD. vol. 1.
OECD. 1985. Committee for Information, Computer and Communications Policy. Ven-
ture Capital in Information Technology. Paris: OECD.
Palmer, Michael, and Jeremy Tunstall. 1990. Liberating Communications: Policymak-
ing in France and Britian, Oxford: Basil Blackwell.
Wiegner, Kathleen. 1985. Europe Fights Back. Forbes, August 12, p. 82.
Wiles, Peter. 1960. Will Capitalism and Communism Spontaneously Converge? For-
eign Affairs, February.
31. Transatlantic Trade Friction
Colby, Laura, and Richard Hudson. 1989. AT&T Learns Right Choices for Europe.
Wall Street Journal, February 2. p. 27.
Davis, Bob. 1988. FCC Orders Phone Companies to Report Yearly Purchases of For-
eign Equipment. Wall Street Journal, February 26.
References 487
Department of Commerce. 1990. U.S. Telecommunications in a Global Economy.

Washington, D.C.: Government Printing Office.
Frey, Carl. 1984. What is Due the Engineer? IEEE Spectrum, June.
Greenhouse, Steven. 1989. AT&T Deal with Italtel Advances. The New York Times,
February 10, p. Dl.
Guyon, Janet and Laura Colloy. 1988. Firm Declines to Boost Stakes in Olivetti. The
Wall Street Journal, April 12.
Hoffman-Riem, Wolfgang. 1984. Policy Research on Telecommunications in West Ger-
many. In Policy Research in Telecommunications, ed. Vincent Mosco. Nor-
wood, NJ: Ablex.
L'Expansion. 1985. Pebereau joue quitte ou double. June 7-20, pp. 67-81.
Northern Telecom. 1980. Nature of the Telecommunications Industry Throughout the
World. Revised May 1980, Submission to the Canadian Restrictive Trade Prac-
tices Commission, p. 74.
Robinson, Kenneth. 1987. The International Competitiveness of Judge Greene's Recent
Order. Paper presented at the 1987 USTA Convention, Orlando, Florida. Octo-
ber 13.
Roussel, Anne-Marie. 1990. Report Warns of Foreign Invasion. Communications Week
International, August 13, p. 7.
Transnational Data and Communications Report (TDK). 1989. Telecom Trade Barriers
Alleged by EC and US. May, p. 12.
Tully, Shawn. 1989. Europe Goes on a Telephone Binge. Fortune, August 28, p. 102-
08.
Whitehouse, Bob. 1990. Euroboom. Communications International, January, pp. 47-
48.
Woollacott, Emma. 1990. Transatlantic Battle Lines. Communications International, p.
22.
32. International Telecommunications Services

Antonelli, Cristiano. 1984. Multinational Firms, International Trade and International
Telecommunications. Information, Economics, and Policy, 1: 333-43.
Chase, Scott. 1990. An Interview with Comsat's Bruce Crockett. Satellite News. May
7, p. 4.
Cowhey, Peter S., and Jonathan B. Aronson. 1985. The Great Satellite Shoot-Out.
Regulation, May-June, pp. 27-35.
Dickinson, Tim, and Charles Leadbeater. 1990. Brussels to Probe World Telephone
Cartel Allegations. Financial Times, April.
Dixon, Hugo. 1990a. Phone Companies Overcharge Callers $10 Billion a Year. Finan-
cial Times, April 3, p. 1.
. I990b. Phone Club Days Are Numbered. Financial Times (international edi-
tion), May 15, p. 7.
Ergas, Henry, and Paul Paterson. 1989. International Telecommunications Accounting
Arrangements: An Unsustainable Inheritance. Presented at Colloque Ville-
franche-Sur-Mer, June 1-3.
Frieden, Robert. 1983. Getting Closer to the Source: New Policies for International
Satellite Access. 37 Fed. Comm. L.J. 293.
General Accounting Office (GAO). 1983. FCC Needs to Monitor a Changing Interna-
tional Telecommunications Market. March 14, GAO/RCED-865-892.
488 References
Horton, R. and J. Donovan. 1987. International Comparisons of Telecommunications

Charges. Telecommunications Policy. 11(3): 269-90.
International Telecommunications Deregulation Act of 1982. Hearings Before the Sub-
committee on Communications of the Committee on Science, Commerce, and
Transportation, U.S. Senate, 97th Congress, 2nd session on S.2469. June 14,
15, 17, 1982. Serial No. 96-126. Washington, D.C.: U.S. Government Printing
Office.
Kwerel, Evan. 1984. Promoting Piecemeal Competition in International Telecommuni-
cations. OPP Working Paper Series. Washington, D.C.: Federal Communica-
tions Commission.
Logica Consultancy Ltd. 1989. The TARIF1CA Annual Review. London: Logica.
Malik, Rex. 1990. A Press Commentator's Guide to the Complexities and Economics
of Telephone Call Charging. Intermedia 18(3): 4-7.
McDowall, M. E. 1987. International Comparisons of Telephone Charges. OFTEL
Working Paper No. 2. London.
Mitchell, B. M. 1983. The Cost of Telephone Service: An International Comparison of
Rates in Major Countries. Telecommunications Policy. March, pp. 5363.
Roos, Ingevar, and Ivar Loenqvist. 1984. Telephone Rates in Various Countries. Tele
(English edition), XXXVII(2): 1-37.
Siemens. 1988. Study on National Telephone Tariffs Worldwide: A Detailed Compari-
son. Munich: Siemens
Stanley, Kenneth. 1988. The Balance of Payments Deficit in International Telecommu-
nications Services. Washington DC: Industry Analysis Division, Federal Com-
munications Commission.
Staple, Gregory. 1990. The Global Telecommunications Traffic Boom. London: Inter-
national Institute of Communications.
Stapley, Barry. 1981. Managing Communications: The Value of Choice. Telecommu-
nications Policy, June.
Televerket. 1986. Facts about Televerket. Stockholm, p. 9.
Walter Hinchman Associates, Inc. 1984. The Economics of International Satellite Com-
munications. Unpublished report. May 18.
33. The Economics of ISDN Integration
Baumol, W. J., J. C. Panzar, and R. D. Willig. 1983. Contestable Markets and the
Theory of Industry Structure. New York: Harcourt Brace Jovanovich.
Besen, S., and L. Johnson. 1986. Compatibility Standards: Innovation and Competition
in the Broadcast Industry. Santa Monica, CA: RAND Corporation.
David, P. A. 1984. Understanding the Economics of QWERTY, or Is History Neces-
sary? Mimeo. Stanford University.
Digital NCTE Decision. FCC Docket No. 81-216.
Farrell, Joseph, and Garth Saloner. 1985. Standardization, Compatibility, and Innova-
tion. Rand Journal of Economics 16(1): 7083.
. 1986a. Standardization and Variety. Economics Letters, January.
. 1986b. Installed Base and Compatibility: Innovation, Product Preannounce-
ments and Predation. MIT Working Papers, No. 411, February.
References 489
Hemenway, D. 1975. Industrywide Voluntary Product Standards. Cambridge, MA:

Ballinger.
Katz, Michael, and Carl Shapiro. 1985. Network Externalities, Competition, and Com-
patibility. American Economic Review 75(3): 424-40.
Lancaster, K. 1975. Socially Optimal Product Differentiation. American Economic Re-
view 64: 567-85.
Meyer, J. R., Wilson, R. W., Baughcum, M. A., Burton, E., and Caouette, L. 1980.
The Economics of Competition in the Telecommunications Industry. Cambridge,
MA: Oelgeschlager, Quinn & Hain.
National Telecommunications and Information Administration (NTIA). 1983. Primer on
Integrated Services Digital Network (ISDN): Implications for Future Global
Communications. NTIA Report No. 1983-138. Washington, DC: U.S. Depart-
ment of Commerce.
Noam, Eli. 1986. The "New" Local Communications: Office Networks and Private
Rutkowski, Anthony. 1985. Integrated Services Digital Networks. Dedham, MA: Ar-
tech.
Salop, S. C. 1979. Monopolistic Competition and Outside Goods. Bell Journal of Eco-
nomics and Management Science 10(1): 141-56.
Schiller, Dan. 1984. The Emerging Global Grid: Planning for What? Paper presented
to the Final Plenary Session of the 14th IAMCR Conference and General Assem-
bly. Prague, Czechoslovakia. August 27-September 1.
Spence, A. M. 1975. Monopoly, Quality, and Regulation. Bell Journal of Economics
and Management Science, Autumn.
von Weizsacker, Carl Christian. 1987. The Economics of Value Added Network Ser-
vices. Unpublished study. Cologne.
34. The Political Economy of ISDN

Egan, Bruce L. 1990. Information Superhighways: The Economics of Advanced Public
Communications Networks. Dedham, MA: Artech.
Elton, Martin, ed. 1991. Integrated Broadband Networks: The Public Policy Issues.
Amsterdam: Elsevier.
European Commission. 1990. A European Market for Telecommunications. ITU Tele-
communications Journal 57: 294-95.
Frey, Carl. 1984. What is Due the Engineer? IEEE Spectrum. June.
Gilhooly, Denis. 1988a. EC Report: ISDN Rollout Is 2 Years Behind. Communications
Week International, November 21.
. 1988b. Dismal Report on European ISDN. Communications Week Interna-
tional, August.
Green, Jeremy. 1989a. A Deadline for ONP. Communications International, December.
. 1989b. High Noon for ONP. Communications International, December.
Marcus, Michael J. 1985. The Regulatory Point of View: ISDN in the United States.
Telephony, March.
Marks, Herbert E. 1984. ISDN in the United States. Paper delivered at the Twelfth
Annual Telecommunications Policy Research Conference. Airlie, Virginia. April.
. 1986. Comparably Efficient Interconnection: Equal Access Is Better. Unpub-
lished paper presented at EIA Roundtable.
490 References
Mueller-Sachse, Karl H. 1984. Noch fehlt di nationale Anstrengung. Medium. Febru-

ary.
Noam, Eli. 1986. The "New" Local Communications: Office Networks and Private
. 1988a. Filing and Review of Open Network Architecture Plans. CC Docket
No. 88-2, Phase I. Federal Communications Commission, Washington, D.C.
. 1988b. Implementing ONA: Federal-State Partnership Needed to Connect Net-
work of Networks. Communications Week, May 2.
. 1989a. Beyond ONA: Designing Modular Network as a Strategy for National
Competitiveness. New York: Columbia Institute for Tele-Information Working
Paper Series. May.
. 1989b. Network Pluralism and Regulatory Pluralism. Paula R. Newberg. ed.
New Directions in Telecommunications Policy, vol. 1. North Carolina: Duke
University Press, pp. 66-91.
. 1989c. ONA: The Need for Partnership. Hearing before the Subcommittee on
Commerce Science and Transportation. Senate Hearing 100-510. Part 2.
Nora, S. and A. Mine. 1980. The Computerization of Society. Report to the President
of the French Republic. Cambridge, MA: MIT Press.
Purton, Peter. 1989. What the EC Is Trying to Achieve. Communications International,
July, pp. 28-29.
Rohlfs, J. 1974. A Theory of Interdependent Demand for a Communications Service.
Bell Journal of Economics and Management Science, vol. 5.
Rutkowski, Anthony. 1985. Integrated Services Digital Networks. Dedham, MA: Ar-
tech.
Sapronov, Walt. 1985. Technical and Regulatory Issues Are Challenging ISDN's Prog-
ress. Data Communication, November.
Schenker, Jennifer. 1990. Europe Sets ONP Timeframe. Communications Week Inter-
national, February 2, p. 3.
Baden: Nomos.
Schon, Helmut. 1984. The Deutsche Bundespost on Its Way Towards the ISDN. Zeit-
schrift fur das Post- und Fernmeldwesen, 6(27). Joseph Keller Gimbolt & Co.
(English language version).
. 1985. Das ISDN im Investitions-, Industrie- und Fernmeldepolitischen Kon-
text. In Intergrierte Telekommunikation, ed. W. Kaiser Miinchner Kreis (No.
11). Berlin: Springer-Verlag.
Ungerer, Herbert. 1989. EC Position and Green Paper Implementation. Paper presented
at Conference on Telecommunications, Brussels. January.
U.S. Department of Commerce, NTIS. 1983. Primer on Integrated Service Digital Net-
work (ISDN): Implications for Future Global Communications.
Weiss, Ernest O. 1984. ISDN-User and Social Aspects. Paper presented to the World
Telecommunications Forum. Geneva, Switzerland.
Wilson, C. 1984. Games of Timing with Incomplete Information. Mimeo. New York
University.
References 491
35. Value-Added Networks and Services
Beesley, Michael. 1981. Liberalization of the Use of the British Telecommunications

Network. Report to the Secretary of State. London: Her Majesty's Stationery
Office. January.
Blau, John. 1990. Germany Lags in VANs. Communications Week International, June
4, p. 10.
Boult, Raymond. 1990. French Review of Services. Communications International. April,
p. 6.
Gilhooly, Denis, 1990. CEPT Revamps X-25. Communications Week International,
July 2, p. 1.
Hayes, Dawn. 1989. Swift UpgradesBut Slowly. Communications Week Interna-
tional, October 2, p. 3.
Lill, A. 1988. European Public Data Network Service Marginally Improved. Stamford,
CT: Gartner Group.
Link Resources. 1984. Competition in Value Added Networks. New York: Link.
Purton, Peter. 1989. European Public Data Networks Losing Money. Telephony, June
5.
Roussel, Anne-Marie. 1989. Public Data Nets Still Failing Users. Communications Week
International, September 4.
Schenker, Jennifer. 1990a. Data Networks Disappoint. Communications Week Interna-
. 1990b. CEPT Rebuts EUSIDIC Study. Communications Week International,
August 13, p. 16.
Sherrat, William. 1988. Important New Initiatives for European Packet Data Networks.
Telematics International.
Tully, Shawn, 1989. Europe Goes on a Telephone Binge. Fortune, August 28, pp.
102-08.
von Weizsacker, Carl-Christian. 1987. The Economics of Value Added Network Ser-
vices. Mimeo. Cologne.
Whitehouse, Bob. 1990. Adding Value. Communications International, July, p. 39.
Woollacott, Emma. 1989. Hiding the Real Cost. Communications International, July,
p. 43.
Yankee Group. 1986. VANs: Vertical Market Implementations. Boston. January.
36. Videotex
Booker, Ellis. 1989. Vive le Minitel. Telephony, August 8, pp. 23-25.
British Business. 1981. CCITT. 1980. Study Group VII. Recommendation S.100. Lon-
don: Crown Publishing. January.
Clavaud, Richard. 1984. Le Destin Incertain de Minitel. L'Expansion, April, p. 37.
Darmon, Jacques. 1985. Le Grand Derangement: La Guerre du Telephone. Paris: Edi-
tions Jean Claude Latles.
The Economist. 1989. Highwired Society. August 9, p. 55.
France Telecom. 1990. Teletel Meets with Worldwide Success. Revue Franc,aise des
Telecommunications, No. 12 (April).
France Telecom News. 1990. Minitel Penetrates French Business. Winter.
Howkins, John. 1985. Communications in Finland. Intermedia, 10(4-5): 48-49.
492 References
Kramer, Richard. 1991. Misapplying the Model: The French Minitel and U.S. Video-
tex. New York: Columbia Institute for Tele-Information Working Paper Senes.
Neue Medien. August 1985.
Roussel, Anne-Marie. 1989a. Italians, French, Purchasing Videotex. Communications
Week International, March 12, p. 6.
. 1989b. Minitel Gets New Terminal. Communications Week International, April
17, p. 52.
. 1989c. France's Minitel Shows Major Losses. Communications Week Interna-
. 1989d. France Telecom Rebuts Teletel Drain. Communications Week Interna-
tional, August 14, p. 8.
, and Mark Rockwell. 1989. Minitel Spreads Beyond France. Communications
Week International, June 12.
Sichel, Bertha. 1983. How the French Are Using Videotex. ITP Report, No. 1 (Spring-
Summer): 23.
Telematique News. 1987. 1987 Minitel Statistics: January to June. Fall.
Thomas, Hilary. 1988. Mintel USA. Mimeo. November.
Tonnemacher, Jan. 1982. The New Media in West Germany. Intermedia 10(2).
Tydeman, John, et al. 1982. Teletext and Videotex in the United States. New York:
McGraw Hill.
Vedel, Thierry. 1987. New Media Politics and Policies in France from 1981 to 1986:
What Is Left? New York: Columbia Institute for Tele-Information Working Pa-
per Series.
Woollacott, Emma. 1990. Small Screen Services. Communications International, Feb-
ruary, pp. 41-43.
37. Transborder Data Flows
Anderla, G. 1983. The International Data Market Revisited. Report to the Second Sym-
posium on TBDF. Paris: OECD. October.
Botein, Michael, and Eli M. Noam. 1986. U.S. Restrictions on International Data
Transmission by Telecommunications Networks. New York: Columbia Institute
for Tele-Information Working Paper Series.
Cass, Ron, and Eli Noam. 1990. Services Trade and Services Regulation in the United
States. Rules for Free International Trade in Services, eds. D. Friedman and
E. J. Mestmacker. Baden-Baden: Nomos.
Commission of the European Communities (CEC). 1990. Concerning the Protection of
Personal Data and Privacy in the Context of Public Digital Telecommunications
Networks. June 5.
Consultative Committee on the Implications of Telecommunications for Canadian Sov-
ereignty (CCITCS). 1979. Telecommunications in Canada. Ottawa: Information
Canada.
de Pous, Victor. 1989. Dutch Data Protection Act in Force. Transnational Data and
Communications Report, December, pp. 21-22.
Fauret, Jacques. 1989. Privacy in the New Europe. Transnational Data and Commu-
nications Report. November, pp. 17-18.
Feketekuty, Geza, and Jonathan David Aronson. 1983. The World Information Econ-
omy. Background Paper of the George Washington and McGill Universities'
References 493
Conference on Policy Issues in the Canadian-American Information Sector. No-

vember 17-18.
Feldman, Mark B., and David R. Garcia. 1982. National Regulation of Transborder
Data Flows. North Carolina Journal of International Law and Commercial Reg-
ulation 7(1): 1-25.
Gassman, Hans-Peter. 1989. Privacy Protection and Computer Networks. Transnational
Data and Communications Report, November, pp. 19-20.
Gebhardt, H. P. 1990. The Legal Basis of Data Protection and Data Protection in the
Field of Telecommunications in Seven Countries. ITU Telecommunications Journal
57(1): 40.
Gesellschaft fur Information und Dokumentation (GID). 1983. Analyse des Informa-
tionsmarkets. AIM Projekt. March.
Gilhooly, Denis. 1990. EC Tackles Privacy. Communications Week International, July
16. p. 1.
Kuitenbrouwer, Frank. 1985. CNIL Warns Against Breach of Privacy Act. Transna-
tional Data Report 8(5): 226-27.
Madsen, Wayne. 1992. Handbook of Personal Data Communications. New York: Stockton
Press.
McManus, Thomas E. 1989. Telephone Transaction-Generated Information: Rights and
Restrictions. Harvard Program on Information Resources Policy. August. Per-
mission to cite gratefully acknowledged.
Newpert, John Paul, Jr. 1989. American Express: Service That Sells. Fortune 120(12)
:82.
New York Public Service Commission (NYPSC). Case 90-C-0075, Proceeding on Mo-
tion of the Commission to Review Issues Concerning Privacy in Telecommuni-
cations: Statement of Policy. March 22, New York: NYPSC.
Niblett, Bryan. 1984. Data Protection Act 1984. London: Oyez Longman.
Noam, Eli. 1990. Telecom Privacy Policy Elements. Transnational Data and Commu-
nications Report, March, p. 9.
. Privacy in Telecommunication Services. In Critical Yearbook on Telecommu-
nications, vol. 1. Karlsruhe: C.S. Mueller, forthcoming.
, and Kent Greenawalt. 1979. Confidentiality Claims of Business Organizations.
In Business Disclosure: Government's Need to Know, ed. Harvey J. Gold-
schmidt. New York: McGraw-Hill, 1979, pp. 378-412.
OECD. 1979. Draft Guidelines Governing the Protection of Privacy and Transborder
Flows of Personal Data. Paris. June 22.
. 1983. Second Symposium on Transborder Data Flows. London. November 30-
December 2.
Office of Technology Assessment, Communications and Information Technologies Pro-
gram (OTA). 1983. Institutional Options for Addressing Information Policy Is-
sues: A Preliminary Framework for Analyzing the Choices. Staff Memorandum.
Washington, D.C. November 29.
Olmstead v. U.S., 277 U.S. 438, at 478 (1927).
Pipe, G. Russel. 1984a. Getting on the TDF Track. Datamation 30(1): 211-13.
. 1984b. International Information Policy: Evolution of Transborder Data Flow
Issues. Telematics and Informatics 1(4): 409-18.
Posner, Richard. 1981. The Economics of Justice. Cambridge, MA: Harvard University
Press.
Robinson, Peter. 1990. TDF Issues: Hard Issues for Governments. Telecommunications
Policy 14(1): 64-70.
494 References
Roche, Edward M. 1984. New Agency on Technology Takes Flight. The Independent
10(4).
Sauvant, Karl P. 1986. Trade and Foreign Direct Investment in Data Services. West-
view Special Studies in International Economics and Business. Boulder, CO:
Praeger, Westview Press.
Seipp, David John. 1978. The Right to Privacy in American History. Cambridge, MA:
Harvard Program on Information Resources Policy, pp. 78-83.
Simmel, Georg. 1906. The Sociology of Secrecy and of Secret Societies, American
Journal of Sociology 11: 441, 446, 450.
Simitis, Spiros. 1985. Data Protection: New Developments, New Challenges. Trans-
national Data and Communications Report 8(2): 95-96
Sixth Report of the Data Distribution Registrar, 1990. London: Her Majesty's Stationery
Office, June.
Transnational Data and Communications Report (TDR). 1983. European Data Base
Searches Chapter Than Using US. 6(8): 423.
. 1986. OECD Weighs Future Telcom Policy Role. 9(1).
. 1989. Protection of Personal Data Used for Employment Purposes. March, pp.
26-29.
Turn, Rein, ed. 1979. Transborder Data Flows: Concern in Privacy Protection and
Free Flow of Information. Washington, DC: American Federation of Information
Processing Societies, vol. 1.
Warren, Samuel D., and Louis D. Brandeis. 1890. The Right to Privacy. Harvard Law
Review 4: 193,196.
Westin, Alan. 1967. Privacy and Freedom. New York: Atheneum.
Wigand, Rolf T. 1985. Transborder Data Flow and Its Impact on Business and Govern-
ment. Information Management Review 1(2).
38. Networks in the Future
Piore, Michael, and Charles Sabel. 1984. The Second Industrial Divide: Possibilities
for Prosperity. New York: Basic Books.
41. Regulatory Tasks for the Future: Interconnectivity in the Pluralist Network
Adler, E.. 1914. Business Jurisprudence. Harvard Law Review 28(135).

Aust, Thomas. 1990. Principles in the New Telecommunications Environment. Attach-
ment to Case 89-C-009-Opinion No. 90-9. Opinion and Order Adopting Reg-
ulation Concerning Common Carriage, Appendix A. February 20, New York:
NYPSC.
Linowes, David E. 1989. Privacy in America: Is Your Private Life in the Public Eye.
Chicago: University of Illinois Press.
New York Public Service Commision (NYPSC). 1987. Case 28961, New York Tele-
phone CompanyRate Moratorium. August 21, New York: NYPSC.
. 1989. Case 29469, Proceeding on Motion of the Commission to Review Regu-
latory Policies for Segments of the Telecommunications Industry Subject to Com-
petition, Opinion No. 89-12. Issues and Effective May 16, New York: NYPSC.
. 1990a. Case 90-C-0075, Proceeding on Motion of the Commission to Review
References 495
Issues Concerning Privacy in Telecommunications. January 31, New York:

NYPSC.
. 1990b. Report on the ISDN Trial in New York State. March 14, New York:
NYPSC.
1990c. Case 88-C-004, Order Accepting in Part and Rejecting in Part Com-
parably Efficient Interconnection (CEI) Task Force Recommendations. June 1,
New York: NYPSC.
Noam, E. 1990. Beyond ONA: Beginning a Modular Network. New York: NYPSC.
. 1991a. Privacy in Telecommunication Services. In Critical Yearbook on Tele-
communications, vol. 1. Karlsruhe: C.S. Mueller, forthcoming.
. 1991b. The Quality of Regulation in Regulating Quality: A Proposal for an
Integrated Incentive Approach to Telephone Service. Price Caps and Incentive
Regulation in Telecommunications. Michael Einhorn, ed. Norwell, MA: Kluwer
Academic Publications.
-. 1991c. Television In Europe. New York: Oxford University Press.
Phillips, Jr., C. 1988. The Regulation of Public Utilities, 2nd ed. Arlington, VA.
Seipp, David John. 1978. The Right to Privacy in American History. Cambridge, MA:
Harvard Program on Information Resources Policy, Harvard University.
Westin, Alan F. 1967. Privacy and Freedom. New York: Atheneum.
Index
Academnet, 286, 288 Alteau, 192

Access and distance learning, 188 Alvey Report, 318
ACEC, 181, 318 American Express, 47
Acorn Computers, 248, 318 American National Standards Institute (ANSI),
Adacom, 266 382
Adler-Triumph, 248 Ameritech, 278
Administration for Posts and Amnesty International, 401
Telecommunications (Luxembourg), 184, Amper, 256
185 Amper-Elasa, 253
AEG (Allgemeine Electricitaets Gesellschaft), Anik, Gengiz, 268
72, 73 Animal breeder services, 380
AEG-Telefunken, 70, 72, 74, 76, 153, 308, Anonymous Eteria Telepikononion Ellados
315, 318 (AETE), 270
Aeritalia, 301 Anselmo, Rene, 337
Aerospatiale, 159, 301, 311 Answering machines, 52, 172
Agnelli, Giovanni, 245 ANT, 77
Agricultural networks, 379 Antimonopoly policy, 305-8
Agriline, 236 Antiope, 183, 382, 389
Airline deregulation, 23 Antitrust policy, 313
ARE-13 system, 17! Antonelli, Cristiano, 331
Alarm systems, 265, 375 Antono, Miroslav, 279
Albania, 282 A. P. Radiotelefon, 226
Alcatel, 75, 76, 82, 143, 145, 157-59, 163, Apollo, 311
164, 172, 174, 182, 197, 199, 218, 235, Appel, Timothy, 71
245, 246, 260, 271, 280, 281, 288, 318, Apple Computer, 317
319, 326, 328, 329, 389 Applicazioni Elettro Telefoniche (AET), 247
Alcatel Austria, 276 Arabsat, 336
Alcatel Bell Telephone, 181 Araskog, Rand, 157
Alcatel-BTM, 268 Argentina, 254
Alcatel Electronique, 153 Argentina Empresa Nacional de Telef6nica
Alcatel Espace, 301 (Entel), 254, 255
Alcatel Ireland, 235 Argo Communications Corporation, 330
Alcatel Kirk (SEK), 226, 227 Ariane space launcher program, 311
Alcatel SEL, 279, 326 Aria, 268
Alcatel-Spain, 278 Arlandis, Jacques, 113, 143, 145
Alcatel Standard Electrica, 255 Armenia, 286
Alcatel-Thomson, 149, 151, 155, 156, 261 ARM equipment, 216
Aline, 183 Aron, Raymond, 150
Allen, Ralph, 15-16 Aronson, Jonathan David, 299, 303, 318, 402
Allentier, 69, 135 Arpanet network, 371
Alster, Yoram, 266 Artificial intelligence, 318
Altay, Senk, 267 Ascom, 192-93
497
498 Index
Asea Brown Boveri group, 223 Azinko, 271

Asea of Sweden, 223 Azores, 259
Association des Ouvriers en Instruments de
Precision, L' (AOIP), 151-53 Baader-Meinhof group, 393
Association of German Electro-Technical Bahrain, 110
Engineers, 73 Bailey, Elizabeth, 57
Association of German Postal Users, 80 Bainerman, Joel, 265
Association of Swiss Electrical Installation Baker, Kenneth, 384
Firms, 190 Baldridge, Malcolm, 207
Astra satellite, 210 Baltic Mobile Telephone System (BMTS),
ATEA, 181, 183 278, 286
Atlantic City conferences, 123, 296 Bancomat, 376
Atlantic Telegraph Company, 19 Bang & Olufsen (B&O), 227
Atlas 400, 144 Bank of Tokyo, 338
Atoun, Akiva, 266 Barclay's Bank, 104, 110, 119
ATS-K, 287 Barnes, Hillary, 228
AT&T, 3, 4, 23, 28, 45, 57-59, 61, 89, 110, Barnes, J. A., 27
111, 113, 117, 126, 138, 151, 173, 181. Basic Service Elements (BSEs), 365, 366
205, 213, 227, 235, 246, 248, 252, 256, Bauer, Johannes M., 196, 198, 199
278, 286, 288, 303, 314, 318, 332-35, Baughcum, Alan, 57
340, 362, 364, 370, 371, 390, 410, 437 Baumol, William J., 27, 29, 57, 354
divestiture, 27, 47, 49, 62, 65, 112, 155, BBC (British Broadcasting Corporation), 192,
322-25, 327, 328, 434 295
AT&T-Philips (APT), 156-58, 172, 174, 256, Beesley, Michael, 113-15
261, 326 Beesley Report, 114-15
Attali, Jacques, 137, 147 Behn, Sosthenes, 324
Audiotex, 51, 374, 410 Behn brothers, 251-52
Auer, 73 Belgacom, 178, 180, 373
Aurec, 265 Belgian Institute for Telecommunications, 180
Aurelle, Bruno, 151 Belgium
Aust, C. M., 170 cellular telephony in, 183
Aust, Thomas, 427 equipment industry in, 180-82, 308, 320,
Australia, 209, 288, 384 324, 325
Austria international service in, 341, 344
cellular telephony in, 198 ISDN in, 183
data protection in, 397 Open Network Provision and, 366
equipment industry in, 198-99, 324 PBXs in, 180
international service in, 341, 344 telephony in, 25, 178-80, 182
postal service in, 195 telex in, 182, 308
telegraphy in, 195 value-added services in, 182
telephony in, 195-97 videotex in, 183, 384, 389
telex in, 198-99 Bell, Alexander Graham, 19, 69, 70
videotex in, 196, 198 Bell Atlantic, 250, 279
Automated teller machine (ATM) networks, Bell Canada, 57
371, 378-79 Bell Communications Research (Bellcore), 76,
Automatic Electric Company (AEC), 124, 140 301, 327-39, 362
Automobile collision estimates, 379 Bellisario, Marisa, 245
Autonet, 372 Bell Labs, 315
Autophon, 192, 193 Bell Operating Companies, 155
Averch-Johnson effect, 49, 58, 108, 355 Bell South, 76, 98, 145, 226, 389
Avramescu, Aristita, 281 Bell System, 90, 324
AXE system, 127, 152, 171, 206, 208-9, Bell Telephone Company, 14, 240
228, 230, 231, 235, 245, 256, 272, 328 Bell Telephone Manufacturing Company
Azerbaijan, 269 (BTM), 24, 181, 187, 268, 281, 288, 324
Index 499
Beloyanis Telecommunications Factory British Petroleum (BP), 104, 110

(BHG), 276 British Rail, 115, 132
Bendix Electronics, 76 British Standards Institution (BSI), 104
Benedetti, Marino, 142, 242-44 British Telecom (BT), 63, 98, 104-32, 145,
BENEFON, 218 163, 209, 244, 247, 278, 279, 286, 288,
Beniger, James, 44 305, 320-21, 330, 331, 337, 338, 372,
Benzoni, Paolo, 244, 247 373, 374, 437
Bergendorff, H., 206 British Telecommunications Act of 1981, 104,
Berliner, 83 372
Berlusconi, Silvio, 248, 249 British Waterways, 132
Bernstein Committee, 225, 229 Brock, Gerald W., 18, 56, 134
Bertho, Catherine, 135, 140, 151, 152 Brown, David, 206
Besen, Stanley ML, 307, 358 Brown, Stanley H., 159
Bezeq (Bezek) Israel Telecommunications Brown, Boveri, 223
Corporation Ltd., 262-66 Bruce, Peter, 91
Bienaim, Jean-Pierre, 146 Bruce, Robert R., 213
BIGFON, 75, 79 Bruckner, Georg, 198, 199
Bird, June, 116 Brunetiere, J. de la, 143
Birtwistle Committee, 125 Btx system, 91, 198, 382, 385, 396
Bismarck, Otto von, 69 Buchanan, James M., 31, 41
Bitcom, 266 Buckley Amendment, 398
BITNET, 265 Budavox, 276
Blake-Bell, 136 Bulgaria, 280-82, 287
Blankart, Charles B., 191 Bulk supply agreements (BSAs), 124
Blankart, Charles D., 192, 193 Bull, Fredrik, 159
Blau, John, 80, 372 Bundesbank, 75
Blaupunkt, 77 Burch, Dean, 299-300
Boan, Kelley, 210, 211 Bureaufax, 128
Bode, Leon, 184 Burton, Ellen, 57
Boissevain, J., 27 Business Networks' (Yritysverkot), 217
Bolt, Beranek and Newman (BBN), 371 Butikofer, Jean-Fredy, 191
Bonan, M., 147 Butler, Richard, 296
Bond, Alan, 254 Buyck, G., 181
Bond Corporation, 276 BYPSComms, 119
Bond Hungaria telecom, 276 Byrnes, Robert F., 280
Booker, Ellis, 388 Byrnes, Tom, 233, 234
Bosch, Robert, 72
Bosch Company, 72, 75, 77, 102, 193 C. Itoh & Company, 338
Botein, Michael, 402 Cable television, 45, 52, 361
Boult, Raymond, 372 in Denmark, 228
Brandeis, Louis D., 399 in France, 146, 153
Brandenburg, 10 in Germany, 91, 92
Bransten, Thomas R., 159 in Ireland, 233-34
Brazil, 209, 403 in Netherlands, 171, 175
Bredow, Hans, 74 in Norway, 222
Bridgeman Inquiry, 103 in Sweden, 210
Bristish Telecom Mobile Phone, 118 in Switzerland, 189
Britain. See United Kingdom in United Kingdom, 53, 120-21
British Aerospace, 119, 132, 163, 337 Cable & Wireless (C&W), 28, 104, 107, 109-
British and Irish Magnetic Company, 18 11, 119, 210, 211, 330, 338
British Approvals Board for CAE, 160
Telecommunications (BABT), 104 Caisse Nationale des Telecommunications
British Electrotechnical Approvals Board (CNT), 143
(BEAB), 104 Call forwarding, 188, 375
500 Index
Call identification, 375-76 Telecommunications (CNET), 143, 152,

Callies family, 159 155, 162, 386, 389
Callpoint, 119 Centre National d'Etudes Espace (CNEE), 311
Call waiting, 375 Centrex service, 76
Campbell, Robert W., 283-86 Centrifugalism, 28, 43-54, 432, 434
Canada, 382, 383, 403 CEPT. See Conference of European Postal and
Caoulette, Louise, 57 Telecommunications Authority (CEPT)
Carnot exchange, 151 Cetesa, 253
Carsberg, Sir Brian, 107, 112, 118, 129, 130 CGCT, 139, 148, 149, 151, 152, 154, 156-
Carter/one decision (1968), 89, 327 57, 324
Carter Report, 103 CGE. See Compagnie Generate d'Electricite
CASA, 301 (CGE)
CASEMA, 170 Chamoux, Jean-Pierre, 145, 164
Cass, Ron, 402 Chappe, Claude, 133
Cassa Conquaglio per II Services Telefonico, Chappe, Ignace, 133-34
242, 244 Charles I, (king of England), 15
CAT-SCAN images, 379 Charles V, Emperor, 9
CBS, 391 Chatline services, 129
CCITT. See Consultative Committee on Chemical Bank, 391
International Telegraph and Telephone Chevenement, Jean-Pierre, 147, 148
(CCITT) Chicago School of Economics, 313, 417
CDN, 249 Chile, 254-55, 325
Ceaucescu, Nicolai, 281 China, 181
Cedergren, H. T., 203 Chirac, Jacques, 157, 163
Cellnet, 107, 118-19 Christiana Telephone Company, 220
Cellular telephony, 5, 98, 218-19, 222, 226, Christian Democratic party
274 in Belgium, 180
in Austria, 198 in Germany, 95
in Belgium, 183 in Italy, 243
charges, 341, 345 in Netherlands, 174
in Cyprus, 273 Christiana (queen of Sweden), 212
in Denmark, 226, 228 CIT, 152
in Finland, 213, 217, 218 CIT-Alcatel, 148, 152, 153-55
in France, 145 Citerel, 152
in Greece, 273 Citesa, 255
in Hungary, 276-77 CIT-Honeywell-Bull (CIT-HB), 147, 148
in Iceland, 231 Citicorp, 44, 47, 254, 391
in Ireland, 236 Class License for Value Added and Data
in Israel, 265 Services (VADS), 116
in Italy, 250 Clavaud, Richard, 389
in Luxembourg, 185 Clear, C. R., 15
in Netherlands, 171 Clientbank, 236
in Norway, 222 Closed-use group networks, 51
in Poland, 277, 278 Clossets, Francois de, 150
in Portugal, 260 CLT, 184
in Romania, 281-82 C-Netz standard, 260
in Russia, 286, 289 CNR Partners, 391
in Sweden, 209, 210 Coaxial cable, 52, 140, 367, 368
in Switzerland, 188 CoCom, 100, 275, 276, 288
in Turkey, 268 Codacovi, Lawrence, 298
in United Kingdom, 118-20 Codding, George A., Jr., 294, 297
in United States, 329 Cognotec, 236
Centrel, 260 Coke, Lord, 15
Centre National d'Etudes des Colbert, Jean-Baptiste, 148
Index 501
Colby, Laura, 246 Computer GmbH, 75

Cole, Barry, 47 Computer industry, 24, 47, 64, 159-62, 317,
Colino, Richard, 299 318,351,360-61,434
Collins, Maurice, 7 Comsat, 248, 300, 335
Colloy, Laura, 248 Comstar, 285
Colombia, 209 Comvik, 203, 204, 210, 222
Comite Europeen de Normalisation (CEN), Concentrators, 46
303 Conduit services, 428
Commerce, U.S. Department of, 155, 323, Confederation of Irish Industry, 236
329 Conference of European Postal and
Commission for the Development of the Telecommunications Authority (CEPT),
Telecommunications System, 92 228, 250, 275, 300-301, 306, 307, 373,
Commission National d'Informatique et Liberte 382, 390
(CNIL), 395 Connaughton, Bernadette, 142, 146
Commission on Communication Technology, Conservative party
79, 92 in France, 163-64
Committee on Information, Computer and in United Kingdom, 103-5, 109, 110, 112
Information Policy (ICCP), 393, 394 Consolidated Telephone Construction and
Commodity trading, 379 Maintenance Company, 195
Common carriage, 15, 16, 426-28 Consultative Committee on International
Communes Modeles Suisses pour la Telegraph and Telephone (CCITT), 121,
Communication (CMC) program, 188 255, 296-98, 302, 304, 332, 339, 350,
Communications Institute of Portugal, 259 351, 360, 362, 367, 368, 373, 382, 414
Communications Satellite Act of 1962, 335 Consultative Committee on Radio (CCIR),
Communications Workers Union, 105 296-98
Compagnie Beige des Telephones Bell, 178 Consultative Committee on the Implications of
Compagnie des Telephones Thomson-Houston, Telecommunications for Canadian
139 Sovereignty (CCITCS), 402
Compagnie Franchise au Cable Telegraphique Consumer information, 379
(CFCT), 135 Contel Cellular, 277
Compagnie Generate de Communication Content interconnectivity, 426-28
(COGECOM), 142, 144, 145, 165 Control Data Corporation (CDC), 160
Compagnie Generate d'Electricite (CGE), 24, Convergent, 161
138, 140, 148, 150-58, 160, 161, 164, Cooke, Edward, 17, 18
181, 245, 261, 308, 318, 325, 326, 328 Coopinvest, 276
Compagnie Generate des Eaux, 145 Copper, 47
Compagnie G6nrale de Telegraphic et de Cordless equipment, 5, 91, 180, 193
Telephonic, 139 Corning Glass, 256
Compagnie Internationale pour ['Information Corporate Treasury service, 236
(CII), 75, 160, 161 Corporatism, as ideology, 313
Compagnie Sans Fils (CSF), 139, 160 Corporatization, 61, 63
Compania de Telefonos de Chile (CTC), 254 Correios e Telecommunicacoes de Portugal
Compania Portuguesa Radio Marconi (CPRM), (CTT), 258-61
258 Cosmic Telex, 265
Company dealer networks, 377 Cost-sharing networks, 31-35, 43
Comparably efficient interconnection (CEI), Council of Europe, 394
365 Countervailing powers, 29
Competitive interconnectivity, 430 Court of Justice of the European Communities,
Complelec, 173 123, 129, 307, 332
Comprehensive planning, 59-60 Covidea project, 391
CompuServe, 372, 391 Cowhey, Peter F., 299, 303, 318
Computer-aided design and manufacturing CP 400 system, 142
(CAD/CAM), 378 Crandall, Robert W., 46
Computer bulletin board systems (BBSs), 375 Crane, Rhonda J., 147
502 Index
Craxi, Bettino, 246 Datel service, 228

Cream skimming, 55, 57, 114, 355, 356 Datex, 128, 213
Credit card verification and processing, 378 DATEX-1, 268
Credit history, 380 Datex-L, 198, 211, 222
Credit Lyonnais, 157 Datex-P, 198
Creeping liberalization, 434 David, P. A., 358
Cresson, Edith, 149 Davignon, Count, 305
Critical mass, 32-33, 36-40, 43, 47, 428 Davis, Bob, 329
Croatia, 280 Dawkins, William, 145, 165
CRONOS data bank, 310 DBP Telekom, 86, 96-98, 101, 163, 288,
Croo, Herman de, 179 341, 348, 373
Cross-and-bye posts, 15 De Benedetti, Carlo, 248, 249, 319
Cross-subsidization, 8, 27, 55-57, 89, 107, DeBount, F., 179
307, 323, 336, 358 Defense, U.S. Department of, 327
CSEE, 148 Defense Advanced Research Projects Agency
CSELT, 248 (DARPA), 371
Cunard, Jeffrey P., 213 de Gaulle, Charles, 160
Curien, N., 143 de Jonquieres, Guy, 317
Custodia, 211 de la Cal, Martha, 259
Cygnus, 335 Delamaide, Darrell, 147, 148
Cylix, 372 Delbarade, Jean Michel, 147
Cyprus, 273 Dellin, L. A. D., 280
Cyprus Telecommunications Authority (CTA), Demand-side telecommunications analysis, 28
273 de Mateo, Rosario, 253
Czarist Telephone Manifest of 1886, 213 de Moragas, Miguel, 253
Czecho-Slovakia, 274, 278-79, 287 Denmark, 5-6, 62, 63, 209
cable television in, 228
Dabbs, P., 170 cellular telephony in, 226, 228
Daimler-Benz, 76, 98, 102 data protection in, 397
Dallmeyer, Martin, 3, 8, 240 equipment industry in, 226-28, 324
Dancall, 226-27 international service in, 341, 344
Danish Post and Telegraph Office (PTT), 224, ISDN in, 228, 229
225 PBXs in, 227, 228, 308
Danish Telecom International (DTI), 227 satellite communications in, 229
Dansk Mobil, 226 telephony in, 25, 224-26, 348
Darmon, Jacques, 147, 149, 153-55, 387 telex in, 227, 228
DASSNET (Data and Special Services value-added networks in, 228
Network), 236 videotex in, 228
Data base development, 310 Department of Trade and Industry (United
Databoks, 228 Kingdom), 106, 107, 115-17, 119, 126,
Data Communication Service (DCS-NET), 182 132
Datanet-1, 171 de Pous, Victor, 397
Datapak, 211 de Quester, 15
Datapro, 275, 277, 278, 279, 284 Deregulation, 23, 31, 62-65, 417
Data protection, 393-404, 419, 429 Deschamps, Louis, 138
Data Protection Act of 1988 (Netherlands), Design Funktion, 208
171 Deutsche Aerospace AG, 102
Data Protection Act of 1981 (United Deutsche Bank, 72
Kingdom), 396 Deutsche Bundespost, 10, 23, 59, 63, 75, 77,
Data retrieval and search, 374 79-94, 97, 102, 119, 244, 306, 355,
Datasaab, 208 363-64, 374, 382, 383, 385-86
Datatie, 215, 217 Deutsche Fernsprecher Gesellschaft, 91
Data transmission. See Value-added networks Deutsche Genossenschaftbank, 98
(VANs); Videotex Deutsche Telephone und Kabel Werke
Datavision, 211 (DeTeWe), 75, 82, 83, 85, 86, 102
Index 503
Development programs, 308-10, 318-19 Bulgaria, 280-81, 287

Devolution, 62, 63 Czechoslovakia, 274, 278-79, 287
DOT. See Direction General de Hungary, 275-77, 287
Telecommunications (DOT) Poland, 277-78, 287
Dialcom, 265, 374 Romania, 270, 281-82
DIANE (Direct Access Information Network Russia, 275, 281, 282-89
for Europe), 310 Yugoslavia, 279-80, 282, 287
Diax Telecommunications, 227 Eastern European and Soviet Telecom Report
Diginet, 217 (EESTR), 275
Digital Equipment Corporation (DEC), 317 East Germany, 270, 287, 288. See also
Digital European Cordless Telecommunications Germany
(DECT) standard, 119 EATS-200 switch, 287
Digital Optical CATV Trunk (DOCAT) EB Ericsson Information Systems, 223
system, 229 EB Telecom, 223
DigiVision, 77 ECI Telecom, 266
Dikon switch, 227 Economides, Nicholas, 27
Dikon Systems Division, 227 Economies of scale and scope, 27, 49, 57-59,
Dillon, Liam, 234 352-59, 362
Dinstein Committee, 263 Edinet, 131
Direction de la Rgglementation Generate Edison, Thomas, 19, 72
(DRG), 164 Edison Company, 136
Direction des Services Publiques (DSP), 165 EDS, 372
Direction General de Telecommunications Edward IV (king of England), 14
(DOT), 140, 142, 143, 146, 148, 149, Edwards, Eliezer, 17
152-58, 163-64, 244, 306, 320 E-ll system, 152
Director, Mark D., 213 Efrat Future Technologies, 266
Directorate for Electronic Industries and Data Egan, Bruce L., 47, 368
Processing (DIELI), 146 Ehmke, Horst, 92
Divestiture, 27, 31, 47, 49, 62, 65, 112, 155, Ehrenberg, Wilhelm, 187
322-25, 327, 328, 434 Eircell, 236
Dixon, Hugo, 109, 285, 338 Eirpac, 236
DMS-100 switch, 197 Eirpage, 236
Dobell, Rodney A., 57 ELAB, 223
Dockwa, William, 15 Elbit Computers, 266
Dr. Neher Laboratories (DNL), 170, 176 Electric and International Telegraph Company,
Dodsworth, Terry, 71, 76 18
Donovan, J., 340 Electric Cable, 266
Dougan, Lady Diana, 300 Electric Telegraph Company, 18
Dow Jones News Service, 391 Electronic Analog, 216
Drake, William, 297, 298 Electronic banking and brokering, 378
Dresden Convention of 1850, 294 Electronic data exchange (EDI), 376-79
Dresser Industries, 402 Electronic Funds Transfer (EFT), 378
Dreyfus, Pierre, 147, 148 Electronic mail, 52, 75, 79, 144, 171, 180,
D-10 system, 152 185, 188, 217, 228, 252, 254, 374, 410
Dual, 318 Electronic Order Exchange (EOE), 377
Ducommun, Maurice, 188 Electronic publishing, 52, 382
Dumey, R., 121, 122 Electronics industry development projects and
DX-200, 218 joint ventures, 318-19
Dyads, 27 in Finland, 218
in France, 145-47, 162-63, 318
E. F. Mutton, 338 in Germany, 318
Early Bird satellite, 299, 331 history of, 315-18
Eastern Europe, telecommunications in, 43, in Israel, 262
274-89 in Italy, 239, 248-49, 318
Albania, 282 in Netherlands, 172-74
504 Index
Electronics (continued) in Turkey, 268-69

in Spain, 252 in United Kingdom, 104, 115, 123-27,
in United Kingdom, 47-48, 104, 318 317, 320-21, 324
in United States, 315-17 in Yugoslavia, 280
Elektrisk Bureau (EB), 223, 230 Ergas, Henry, 233, 317, 339
Elektro, 223 Ericsson, Lars Magnus, 208
Elektronska Industrija of Nis, 280 Ericsson, L. M., Company, 24, 119, 127,
Elias, Dietrich, 80, 86 138, 139, 146, 151, 152, 154, 156, 158,
Elinda, 271 169, 171, 172, 176, 177, 181, 198, 203,
Elizabeth I (queen of England), 14 204, 206-10, 216-18, 223, 227, 228,
ELLEMTEL, 206, 208 230, 231, 235, 245, 256, 261, 267, 271,
Ellul, Jacques, 312 272, 278, 279, 280, 301, 318, 326-29
Elmaghraby, Salah E., 27 Estonia, 286, 289
Elron Electronics, 266 E10 system, 153, 156, 158, 235, 278, 279
Elscint, 266 Etheridge, James, 315
Elton, Martin, 47, 368 Eucom, 373
EMCOM, 282 Eurocard, 376
EMD switch, 83 Euronet, 310
Engels, Friedrich, 312 European Association of Information Services
Enterprise networks, 366 (EUSIDIC), 259
Entitlement growth, 34-35 European Commission, 6, 90, 91, 122, 180,
Entrepreneurial function, 50 255, 300, 301, 329-30, 332, 339, 401,
Entreprise GeneYale des Telecommunications 420
(EOT), 142 antimonopoly policy, 305-8
Entry barriers, 26, 329-30, 334 data base development, 310
Environmentalism, 313 Open Network Provision, 365-67, 372
Eommex, 271 space development, 311
Equipment industry, 4, 23-24, 434 technology development, 308-10
activist role by, 45-46 European Committee for Electronic
in Austria, 198-99, 324 Technology Norms (CENELEC), 303
in Belgium, 180-82, 308, 320, 324, 325 European Communications Satellite (ECS),
in Bulgaria, 280-81 163, 311
containment of American in Europe, 323-25 European Computer Manufacturers Association
in Cyprus, 273 (ECMA), 304
in Czecho-Slovakia, 279 European Conference on Post and
in Denmark, 226-28, 324 Telecommunications (CEPT), 121
in Finland, 218-19 European Institute for Research and Strategy
in France, 63, 317, 318, 320, 324, 326 Studies in Communication (Eurescon),
in Germany, 71-77, 88-89, 94, 320, 324 301
in Greece, 271-72 European Investment Bank, 276
in Hungary, 276 European Meterological Satellite Organization
in Ireland, 235 (EUMETSAT), 311
in Israel, 266 European Preparatory Commission for Space
in Italy, 245-49, 316, 320, 324-26 (COPERS), 311
in Japan, 48, 321 European Research Coordination Agency
in Netherlands, 172, 177, 324 (EUREKA), 305, 309-10, 318
in Norway, 223, 324 European Space Agency (ESA), 163, 301,
in Poland, 278 311
in Portugal, 260-61, 324 European Space Research Organization
reductions in cost and increases in (ESRO), 311
productivity, 46-47 European Space Vehicle Launcher
in Russia, 286-88 Development Organization (ELDO), 311
in Spain, 255-56, 308, 321, 324, 326 European Strategic Program of Research and
in Sweden, 206-9, 321, 324 Development in Information Technology
in Switzerland, 192-94, 324 (ESPRIT), 305, 308-9, 318
Index 505
European Telecommunication Network Fiber, price of, 46, 47

Operators group (ETNO), 301 Fiber optics, 52, 79, 110, 128, 230, 231, 288,
European Telecommunications Foundation, 432
165 Fibronics, 266
European Telecommunications Standards Fidler, Stephen, 281
Institute (ETSI), 306, 307, 366, 426 Financial interconnectivity, 428-29
European Workshop for Open Systems Fininvest, 250
(EWOS), 303-4 Finland, 128, 278, 287
Eurotechnique, 162 cellular telephony in, 213, 217, 218
Eutelsat, 163, 248, 278, 280, 282, 301, 311, equipment industry in, 218-19
336 international service in, 344
Evagora, Andreas, 235 ISDN in, 217
Eward, Ronald, 196 PBXs in, 215
EWSD system, 75, 85, 193, 197, 228, 268, postal service in, 212
272, 278, 282 telegraphy in, 212-13
EWS system, 152 telephony in, 212-18, 348
telex in, 218
videotex in, 215, 217
Fabbrica Apparechiature Telefoniche Materiale Finnish Telecom International, 286
Elettrico (FATME), 209, 245, 246 Finvik, 210
Fabian Society, 417 5-ESS Prx switch, 156, 172, 174
Fabius, Laurent, 148-49, 154, 155 Flanders, 180, 181
FACE, 245, 246, 281, 324 Fogel Commission, 265
Facsimile service, 79, 90, 100, 128, 171, 180, Ford Aerospace, 299
217, 247, 252, 254, 286, 330 Ford Motor Company, 130
Factory production, 378 Foreman-Peck, James, 103, 107, 124, 221,
Fairchild, 162, 316 222
Fair Credit Reporting Act of 1970, 398 Fossier, Marc, 45, 143, 144, 148, 163
Fairness doctrine, 428 Four Sisters (Austria), 195, 197, 199
Fair-price bureaus, 409 Fowler, Mark, 157, 326
Fair-way, 110 France, 63, 133-69, 306, 308
Farmlink, 384 banking in, 61
Farrell, Joseph, 358, 359 cable television in, 146, 153
Faulhaber, Gerald, 57 cellular telephony in, 145
Fault reporting rates, 348, 349 computer industry in, 159-62
Fauret, Jacques, 397 data protection in, 394-95, 401, 402
Fawcett, Henry, 20 early history of telecommunications in,
Faxpost, 171 133-35
FDDI service, 217 electronics industry in, 145-47, 162-63,
Fedeida, Sam, 383 318
Federal Associate of German Industry, 95 equipment industry in, 63, 317, 318, 320,
Federal Communications Commission (FCC), 324, 326
57, 89, 106, 113, 115, 323, 327-29, exports, 155-56
331-35, 338, 339, 362, 363, 365, 366, international service in, 339, 341, 344
372, 420, 422, 435 ISDN in, 144, 395
Federal Telegraph and Telephone Law of 1922 nationalization in, 136-41, 146-50, 153,
(Switzerland), 189, 190 154
Feldman, Mark B., 397, 398 Nora-Mine report, 64-65, 361, 386
Felten & Guilleaume, 74 Open Network Provision and, 366
Ferdinand I (emperor), 195 PBXs in, 140, 152-53
Ferdinand III (emperor), 9 postal service in, 133, 142
Fernmeldetechnisches Zentralamt (FTZ), 82 radio in, 139, 146
Ferranti, 119, 315 satellite communications in, 163, 164
Ferrier, Alexander, 134 service sector in, 48
Fiat, 245, 248, 250, 255 telegraphy in, 133-35, 137
506 Index
France (continued) General Telephone Company (Sweden), 203,

telephony in, 25, 45, 56, 135-44, 151-55, 204, 208
163-66, 348 Generic services, 370, 373-76
telex in, 140 GENIE, 391
value-added services in, 144-45, 164, 372 Geostar, 255
videotex in, 145, 382, 383, 386-90 German Edison Company, 72
France Cables de Lyon, 153 German Federal Accounting Office, 82-84
France Cables et Radio, 142, 165, 330 German Industry Association, 300
France Telecom, 61, 119, 142-45, 148, 149, German Monopoly Commission, 79-82, 86,
162-66, 278, 330, 373, 386-90 88, 89, 92-94, 102
Franking privilege, 16 German Telegraph Union, 69
Frankl, Judy, 287 Germany, 63, 69-102
Franz Joseph (emperor), 195 Bundespost as institution, 79-81, 86-87
Frederick (duke of Wiirttemberg), 9 cable television in, 91, 92
Frederick the Great, 12, 13 data protection in, 393, 396, 403-4
Free Democratic party (Germany), 95 electronics industry in, 318
Freedom of Information Act, 398 entry barriers in, 26
Free speech, 399-401, 410-11, 428 equipment industry in, 71-77, 88-89, 94,
Friedman, Milton, 30 320, 324
FTCC, 333 exports, 75
Fugger family, 9 international service in, 339, 341, 344
Fuhrman, Peter, 71 ISDN in, 79, 363-64
Fujitsu, 24, 75, 127, 227, 256, 328 newspapers in, 81
Fuller, Claire, 118 PBXs in, 85, 86, 91
Funen Company (FKT), 224, 225 postal system in, 8-9, 13-14, 69, 70, 81,
Fuss, Melvyn, 57 91-92, 94, 97
procurement practices in, 81-91
Galbraith, John Kenneth, 26 radio in, 74
Garcia, David R., 297, 398 reform, 91-98
Garnett, Nick, 233 satellite communications in, 163
Garnham, Nicholas, 128 telegraphy in, 14, 69-70, 81, 98, 100
Gartner, Bernhard, 100, 101 telephony in, 14, 25, 45, 56, 69-74, 77-
Gassmann, Hans-Peter, 393, 397 79, 81-101, 140, 141
GB-Inno, 182 telex in, 79
Gebhardt, H. P., 395-97 value-added services in, 372, 373
GEC Plessey Telecommunications (OPT), 285 videotex in, 79, 91, 382, 385-86
GeDa, 248 Gescheidle, Kurt, 85
Gelac, Sergiu, 281 Gfeller, 192, 193
Geneen, Harold, 157 Gibbons, Roger, 304
General Accounting Office (GAO), 332 Gibraltar, 256-57
General Agreement on Tariffs and Trade Gilhooly, Dennis, 70
(GATT), 293, 302, 303, 319 Gille, Laurent, 113, 143, 145
General Dynamics, 337 Giscard d'Estaing, Valery, 141, 146, 149
General Electric Company (GEC) (United Gist, Peter, 109
Kingdom), 24, 76, 107, 111, 124-27, Gleiss, Norman, 207
151, 209, 308, 315, 319, 383 Global village concept, 407, 408
General Electric Corporation (United States), Glotz, Peter, 95, 404
160, 316 Glucksmann, Andr, 150
General Electric Information Services, 117, GN-Communications, 227
373 GN-Datacom, 227
General International Telegraph and Telephone GN-Telematic, 226, 227
Company, 127 Gobelins tapestry, 148
General Post Office (United Kingdom), 18-22, Goedhart Committee, 174
103, 104, 117, 122, 123, 127, 132 Goetzler, Herbert, 70
General Prussian Postal Order of 1712, 10 Gomez, Alain, 153-55
Index 507
Gorbachev, Mikhail, 283 Health care providers and insurance networks,

Gore, Albert, 430 377
Gower, 136 Heath, Edward, 104
GPT, 126, 288 Heimburger, Hans, 204
GPT Stromberg-Carlson, 76 Hellaspac, 271
Gradualism, concept of, 436 Hellenic Telecommunications Organization
Grammont, 151 (OTE), 270-73
Grannet plan, 136 Helsinki Telephone Company, 214, 217, 390
Graphic Scanning (Graphnet), 372 Hemenway, D., 358
Gray-market liberalization, 306 Memmeon, J. C., 14
Great Britain. See United Kingdom Henneman, Gerhard, 81
Great Northern Telegraph Company, 226 Herzog, Chaim, 263
Greece, 43, 306, 308, 330, 348, 366 Herzog Commission, 263
cellular telephony in, 273 Heuss, Theodor, 72
equipment industry in, 271-72 Hewlett-Packard, 245, 317
international service in, 341, 344 Heywood, 348
telephony in, 25, 270-71 Hill, George Birbeck, 17
telex in, 271 Hill, Sir Rowland, 17
value-added networks in, 273 Hitler, Adolf, 71
Green, Jeremy, 119, 223 Hoffmann-Riem, Wolfgang, 92, 93
Greenawalt, Kent, 400 Hofmann, Max, 187
Greene, Harold, 112, 323, 391 Holcombe, A. N., 71, 135, 136, 138, 170,
Green Paper (European Commission), 306-7 195, 203, 220, 240
Green parties, 313 Hoist, Gilles, 172
Greewald, B., 34 Honeywell, 324
Gretag, 192 Honeywell-Bull, 160, 161
Grocery networking, 379 Hong Kong, 314, 384
Grosse, Oskar, 13 Hong Kong Telephone Company, 110
Groupe Bull, 147, 149, 159-62, 164, 308, Horton, R., 340
318, 319 Hotel in-room services, 379
Grundig, 173, 318 Hourdakis, 271
GSM standard, 98, 119, 145 Howe, Eric, 397
GTE, 76, 125, 181, 245, 246, 255, 280, 324, Howkins, John, 390
325, 327, 333, 362, 371-72 HT Hirdastechnika Svovetkezet, 276
GTE-Pupin, 280 Hudson, Richard L., 130, 246
GTE Telecomunicazioni, 246 Hughes, 299
G3S, 148 Hull Telephone, 107, 115, 117-18, 120
Guerard, Agnes, 45, 141 Hungarian Telecommunications Corporation
Gulyaev, Yuri V., 287, 288 (HTC), 275, 276
Gustafson, Gunnel, 27 Hungaria Telecom, 277
Gustafsson, Karl Erik, 203 Hungary, 275-77, 287
Guyon, Janet, 248 Hush-A-Phone decision (1956), 59, 89, 90
Huxley, John, 116
Habenicht, Georg, 100, 101
Hafner, Katie, 100, 101
Hai, Shaul, 264, 265 Iberpac, 254
Hambros, 118 IBM Corporation, 47, 49, 64, 65, 76, 107,
Handelsbanken, 208 116, 159-62, 173, 176, 228, 305, 317,
Hapsburg Empire, 3-4, 8-9 355, 361, 372, 373, 377, 385, 391
Hardin, Garrett, 42 IBM France, 147
Hasler, 77, 192, 193 IBM Information Network, 372
Havas, 148 IBN. See Integrated broadband network (IBN)
Hawaiian Telephone Company, 333 Iceland, 230-31, 341, 344
Hayes, Dawn, 119, 246, 377 ICL, 111, 117, 127, 222, 227, 308, 318
Heal, Geoffrey, 27 IDN, 412
508 Index
Imperial Postal Finances Act of 1924 International Institute of Communications, 339

(Germany), 71 International interconnectivity, 430-31
Incandescent lamps, 73 International Network Services Limited (INS),
Income distribution, 27 117
Independent Committee for the Supervision of International Postal Union, 14
Telephone Information Services, 129 International Radio Telegraph Conference, 295
Industrial Development Bank, 272 International Record Carriers (IRCs), 331,
Industrial divide, 49 333-35
Industrial policy, 53, 305, 312-21 International Red Cross, 401
Infonet, 163, 255, 373 International Satellite, Inc., 335
Informational connectivity, 429-30 International services, 111, 114, 128, 171,
Infrared transmission, 52 179, 182, 187, 188, 231, 244, 258-60,
Infrastructure interconnectivity, 430 274, 281, 285, 305-6, 331-49
Inmarsat, 247 International Standards Organization (ISO),
Inquiry Commission on New Information and 303, 360, 365, 413
Communication Technologies (Germany), International Telecom Japan (ITJ), 338
93-94 International Telecommunication Regulations,
Instability in regulation, 422-23 298
Institute for Industrial Reconstruction (IRI), International telecommunications
240, 241, 246-47, 253, 321 organizations, 293-304
Institute de Communicacoes de Portugal, 259 International trade payments, 377
Institute Nacional de Industria (INI), 253, 256 International trade shipment data service, 377
Insurance industry networking, 379 Interorganizational networks (lONs), 366
Integrated broadband network (IBN), 45, 75, Intersputnik, 278, 284, 337
228-29, 309, 361, 367-68 Interstate Commerce Commission (ICC), 420,
Integrated circuits, introduction of, 316 422
Integrated Services Digital Network. See ISDN Interurban Telephone Company of Southern
(Integrated Services Digital Network) Finland, 213
Integrated Telecommunications System (ITS), Intracom, 271
193 Investdata, 376
Intelmatique, 389 IPE, 258
Intelsa, 209, 255 Ireland, 308, 326
Intelsat, 28, 235, 247, 284, 298-300, 331, cable television in, 233-34
335-38 cellular telephony in, 236
Interaction, 417-21 data protection in, 397
Interconnectivity, 39, 40, 111-12, 425-32, equipment industry in, 235
434 international service in, 341, 344
Inter-Institutional Integrated Services postal services in, 233
Information Systems (INSIS), 310 telegraphy in, 232
Internal subsidies, 437 telephony in, 25, 232-34, 348
Internal Telecommunication Union (ITU), 79, value-added networks in, 236
123, 214, 293-97, 348, 360 IRICON, 247
International Bell Telephone Company, 187, Irish Telecommunications Investments (ITI),
203, 220 234
International Broadcasting Union, 297 Irmer, Theodore, 360
International Consultative Committee for Long ISDN (Integrated Services Digital Network),
Distance Telephony, 295 45, 47, 75, 79, 350-68, 412, 414
International Digital Communications (IDC), in Belgium, 183
110, 338 concept of, 350-52
International Electrotechnical Commission in Denmark, 228, 229
(IEC), 303 economies of scale and scope, 354-59, 362
International Frequency Registration Board in Finland, 217
(IFRB), 296 in France, 144, 395
International Institute for Advanced Systems in Germany, 79, 363-64
Analysis (HASA), 402 hegemonic, 363-64
Index 509
in Italy, 250 vertical integration in, 32!

in Norway, 222 videotex in, 383, 389
open network architecture concept and, Jason, Chris, 112
364-66 Jefferson, Sir George, 105, 114-15
political economy of, 360-68 Jeumont-Schneider, 148
in Sweden, 211 Job searches, 379
technical capability, 367-68 Johnson, L., 358
in United Kingdom, 129 Joint European Submicron Silicon Initiative
in United States, 361-63 (JESSI) program, 310
Iskra, 268 Joint ventures, 318-19
Iskratel, 280 Jordan, Art, 27
Israel Joseph, Sir Keith, 104
cellular telephony in, 265 JOVE, 116, 117
electronics industry in, 262 JTAS, 224, 225, 227
PBXs in, 266 Junta Nacional de Telecommunicaciones, 252
telephony in, 262-64 Justice, U.S. Department of, 112, 182, 322,
telex in, 265 323, 327
value-added networks in, 265
ISRANET, 264 Kapsch, 197-99
Istel, 115, 117, 236, 373 Kami, Shlomo, 27
Istok switch, 287-88 Karvonen, Mali, 218
Italcable, 239, 241, 242, 243, 247, 249 Katz, Michael, 355
Italia Radio, 241 Kav Manche, 265
Italtel, 24, 239, 245-47, 250, 256, 278, 319, KDD, 288, 338
321, 325-26 Keenan, Brennan, 233, 234
Italy, 61, 62, 111, 163, 174, 226, 239, 282, Keketekuty, Geza, 402
288, 308 Kellaway, Lucy, 180
cellular telephony in, 250 Keller, Franfois, 188
electronics industry in, 239, 248-49, 318 Keller, Paul, 171, 197
equipment industry in, 245-49, 316, 320, Kelly, Tim, 99, 280
324-26 KEMA, 177
international service in, 341, 344 Keren, 266
ISDN in, 250 Kienzle, 77
Open Network Provision and, 366 Kingsbury Commitment (1913), 39
postal services in, 240, 242 Kinnevik, 110, 210, 211
telegraphy in, 240, 241 Kiosk system, 386-88
telephony in, 25, 142, 209, 240-44 Kioulafas, Kyriakos, 273
telex in, 243, 244, 247, 249 Kirch, L., 185
value-added services in, 372, 373 Klett, Roland, 80, 86
videotex in, 250, 383, 384, 389, 390 Klingman, David J., 27
Itapac, 249 Knieps, Gunter, 88, 89, 190, 191
Itatel, 174 Knight-Ridder, 391
Ito, Yoichi, 48 Knox, Robert, 27
ITT Corporation, 76, 77, 82, 85, 126-27, Kobelt, Christian, 187
138-40, 146, 151, 152, 154-58, 171, Kohl, Helmut, 157
181, 187, 197, 209, 218, 227, 245, 246, Komatsuzaki, Seisuke, 48
251, 256, 258, 260, 261, 271, 280, 324- Kommertel, 278
26, 328, 383 Koor Industries, 265, 266
ITT-Alcatel, 321 Koornet, 265
ITT Worldcom, 325 K.O.P., 217
Kotsonis, 273
Japan, 288, 319, 320, 329, 330 Kramer, Richard, 388
equipment industry in, 48, 321 Kraushaar, J., 348
international service in, 338 Krzywicki, John, 205
telephony in, 26 KTAS, 224, 225, 228
510 Index
Kvant PBX, 287 Macao, 110

Kvarts switch, 287, 288 McCaw, 131, 279, 330
Kwerel, Evan, 331 McDonnell, Douglas, 131, 372
McDowall, M. E., 340
McGovern, Fergus, 234, 236
Labour party (United Kingdom), 105, 112, McGuire, Martin, 41
117 Madeira, 258
Lafarge, Guy, 45, 141 Magnetic Telegraph Company, 18
Laissez-faire, 62 Magyar Posta, 275
Lakicevic, Ognjen, 279 Maitland Report, 296
Lancaster, K., 358 Malerba, Franco, 75, 162, 173, 315
Landis & Gyr, 192 Malik, Rex, 339
Lange, Bernd-Peter, 95 Malta, 250
Larcher, Gerard, 165 Mammi, Oscar, 243
Large scale integration (LSI), 316 Managed Data Network Services (MDNS)
Lassere, Bruno, 164 project, 305, 373
Latvia, 286, 287, 289 Mannesmann, 98
Latzer, Michael, 196, 198, 199 Manning, D., 107, 221, 222
Lauro, R., 246, 247 Mantell, L., 57
Law on Restructuring the Bundespost of 1989 Manufacturing design, 378
(Germany), 97 Marconi, 74, 118, 128, 295
Leased line resale, 115 Marcus, Michael J., 362
Le Boucher, Eric, 65, 300 Mark, Karl, 49
LEDs, 46 MarkNet, 372
Legal interconnectivity, 426 Marks, Herbert E., 360, 362, 366
Leotard, Francois, 164 Marlborough, Duke of, 21
Liberal party MarTech Strategies, Inc., 255
in Netherlands, 174 Martinez, Miguel, 256
in United Kingdom, 103 Marx, Karl, 312
Library shared cataloguing, 380 Massachusetts Institute of Technology (MIT),
Lieben patent, 74 314-15
Liechtenstein, 194 Mass production theory, 49-50
Lifeline programs, 56, 323, 429 Master television antennas, 52
Lill, A., 373 Materiel Telephonique, Le (LMT), 138, 139,
Link resources, 372 146, 148, 149, 151, 152, 154, 155, 267,
Lithuania, 286, 289 324
Littlechild, Stephen, 103, 107, 108 Matra, 145, 147, 148, 151, 158, 163, 164,
Litton, A. J., 232 326, 337, 389
Local area networks (LANs), 46, 432 Matra-Harris Semiconductors (MHS), 162
Lockheed, 402 Matsushita, 338
Locksley, Gareth, 104, 125 Mauroy, Pierre, 148
Loenqvist, Ivar, 339 Maximilian I (emperor), 3, 8
Logica Consultancy Ltd., 340, 341 Mayak satellites system, 284
London Hydraulic Power Company, 111 MBB, 301
Longuet, Gerard, 163-65 Mbs, 210
Lopez-Escobar, Esteban, 252 MCI, 28, 49, 58, 110, 209, 332-35, 348,
Lorenz, C., 83 373, 410
Louisson, Simon, 264 MCI Data Transport, 372
Louis XI (king of France), 133 MCI Execunet (1976), 39
Lucas, 315 MCI International, 334
Luxembourg, 25, 184-85, 305, 341, 344 Mclnnes, Neil, 159
Luxpac, 185 Medes, 7
Lycom, 227 Medical communications, 379
Lynch, Karen, 119 Meganet, 228
Index 511
Memorialists, 103 Montgomery, Page, 171, 197

Mercury Communications, 104, 106, 107, Morgan, Kevin, 103
109-13, 115, 119-21, 132,255, 337, Morley, Arnold, 21-22
338, 341, 348, 389 Morley, Meg, 143
Mestmacker, Ernst-Joachim, 82, 86, 90, 93 Morse, Samuel, 17, 18
Metaconta, 287 Morse telegraph, 71-72
Meteorological satellites, 311 Moschet, Wernhard, 80, 95, 96
Metrocast, 131 Moscow Telephone Network, 219
Metropolitan area, 52 Motorola, 118, 119, 198, 265, 330
Metten Report, 308 Motorola Ireland, 236
Mexandeau, Louis, 306 MT system, 152-54, 156, 158, 287
Mexico, 209, 255 MIX network, 185
Meyer, Hugo Richard, 19, 21 Mueller, Milton, 39
Meyer, John R., 57, 354 Mullard, 173
Microelectronics, 316 Miiller, Jurgen, 88, 89, 181, 320
Micronet 800 service, 384 Multiplexes, 46
Microwave multipoint distribution, 52 Multisat, 171
Microwave transmission, 44, 46, 111, 226 Mulvey, J., 27
Millerand, A., 137, 138 MUPID, 198
Millicam, 210 Mussolini, Benito, 246, 253
Millicom, 118, 119 Must-carry rule, 428
Mine, Alain, 53, 64, 361, 386 Myllo, Heikki, 214
Minicall, 210
Minicom, 144
Minitel, 144, 145, 306, 383, 386-89, 389 Nachrichtentechnik, 288
Minitelnet, 389 Nadiri, Ishaq, 57
Minpromsviaz, 287 Nagai, Susuma, 48
Minsviaz, 283-89 Nambu, Tsuruhiko, 48
Mitchell, B. M., 339, 340 Napoleon I (emperor), 133
Mitchell, James, 233, 234 Napoleon HI (emperor), 135
Mitchell, Pam, 118 Napoleonic Wars, 13, 16, 134
Mitel, 124-25, 131, 321 National Aeronautics and Space Administration
Mitsubishi, 338 (NASA), 311
Mitsui Knowledge Industry Company, 389 National Commission for Communications and
Mitterand, Francois, 146-48, 150, 162, 164, Liberties (CNCL), 164
246, 261, 314, 326 National Communications Union, 105
Mix & Genest, 83 Nationalization, 23, 136-41, 146-50, 153,
Mobile telephony. See Cellular telephony 154
Mobira, 218 National Network, 115
Mobitex, 210 National Pole, 246
Modai, Yitzhak, 263 National Semiconductors, 162
Modems, 52, 86, 90-91, 100, 172, 182, 190. National Socialist party (Germany), 83
205-7, 228, 247, 261, 266, 268, 271, National Telecommunications and Information
279, 286, 305, 308 Administration (NTIA), 328, 351
Modrow, Hans, 100 National Telecommunications Council (Statens
Modularity of networks, 412-16 Telenamnd), 207
Moller, Emil, 227 National Telephone Company (Ireland), 232
Monde, Le, 65, 147 National Telephone Company (United
Monopoly Kingdom), 20-22, 117
defense of, 55-60 Natural monopoly, 31, 57-59, 93, 94, 354,
natural, 31, 57-59, 93, 94, 354, 359, 434 359, 434
Monopoly and Mergers Commission (United NEC, 24, 125, 161, 280
Kingdom), 106, 124, 126, 321 Nederlandchse Bell Telephoon Maatschappij,
Monsen, R. Joseph, 148 169
512 Index
Nederlandsche Standard Electric Maatschappi Nixdorf, Heinz, 77

(NSEM), 172 Nixdorf Computer, 75-77, 308
Neoclassical economic theory, 50 Noam, Eli M., 27, 52, 112, 120, 353, 361,
Neopostel, 170 365, 366, 398, 400, 402, 426, 430
Nerion, 207 Nokia, 145, 217, 218-19, 286
Netas, 268, 269 Nokia Cables, 218
Netcom, 222 Nokia Data Systems, 218
Netherlands, 63, 209, 373 Nokia Electronics, 390
cable television in, 171, 175 Nonprofit institutions, 45
celular telephony in, 171 Nonsustainability theory, 29-30
data protection in, 397 Nora, Simon, 53, 64, 361, 386
elecronics industry in, 172-74 Nordic Mobile Telephone Service (NMT),
entry barriers in, 26 210, 213, 217, 222, 228, 278
equipment industry in, 172, 177, 324 Nordictel, 211, 226, 286
international service in, 341, 344 Nordisk Kabel og Trad (NKT), 226-28, 273
PBXs in, 172, 176, 177 Nordmende, 153, 318
postal service in, 169 Nordtel, 333
radio in, 170 Normalzeit, 75
satellite communications in, 171 North Atlantic Treaty Organization (NATO),
telegraphy in, 170 178
telephony in, 25, 169-71 Northern Business Information, 218
telex in, 171 Northern Telecom, 82, 111, 125, 127, 152,
value-added services in, 175, 177, 372 197, 206, 207, 223, 259, 261, 268, 272,
videotex in, 171, 383, 384 276, 324, 326-28, 377
Network Information Service Company North Pacific cable, 110
(Japan), 131 Norway, 220-23
Network management equipment, 46 cable television in, 222
Networks, 26-42 cellular telephony in, 222
concept of, 26-28 equipment industry in, 223, 324
cost-sharing stage, 31-35, 43 international service in, 344
in future, 407-11 ISDN in, 222
interconnection, 39, 40, 111-12, 425-32, PBXs in, 223
434 telegraphy in, 220
modularity, 412-16 telephony in, 4, 220-23
new coalitions, 432-35 telex in, 222, 223
pluralistic, 43, 49-50, 407, 425-39 value-added networks in, 222
redistributory stage, 35-39, 43 videotex in, 222
social welfare and, 41 Norwegian Telecommunications
subsidies for start-up, 39-41 Administration (NTA), 221-22
technical integrity of, 59 Norwegian Telecommunications Regulatory
theories of emergence of multiple, 28-30 Authority (NTRA), 222
tipping, 30-31, 37-38, 43, 433 Nouvion, Mireille, 137, 138, 151
Network Systems International (AT&T), 246 Nova Industrier, 222
Network termination device (NT1), 362 NOZEMA, 170
Neumann, Karl-Heinz, 75, 99 NREN science network, 430
Newspapers, 4, 81, 384 NsNIIS, 288
New Telephone Company, 21 NTK, 177
New York State Public Service Commission, Nulty, Timothy, 274, 280, 285
323, 350, 366, 401, 426, 428, 430 Nurneris, 144, 395
New York Telephone Company, 130 Nyheim, Jan Henrik, 221, 222
New York Times, 49, 390 NYNEX, 98, 119, 121, 257, 330, 338, 391
Niblett, Bryan, 396
NHR, 288
Nippon Telephone and Telegraph (NTT), 48, OES-E system, 197
321, 373 OFEG, 197
index 513
Office equipment manufacturers, 24 PBX (private branch exchange), 46, 74, 76,
Office of Management and Budget (OMB), 77, 305, 373
398 in Belgium, 180
Oftel (Office of Telecommunications) (United in Denmark, 227, 228, 308
Kingdom), 106-9, 112, 115-19, 121, in Finland, 215
124, 129-32, 321, 428 in France, 140, 152-53
Ohnesorge, W., 71 in Germany, 85, 86, 91
Okavana, Jun, 233 in Iceland, 231
Olinet, 248 in Israe!, 266
Olivetti, 98, 239, 248-50, 308, 317, 318, 325 in Netherlands, 172, 176, 177
Olivetti-Bull, 159 in Norway, 223
Olsen, O. Jess, 225, 226 in Portugal, 261
Omnitel, 188, 250 in Russia, 287
On-line databases, 379, 402-4 in Sweden, 205-7
Ontyme, 131 in Switzerland, 190
Open network architecture (ONA) concept, 39, in Turkey, 268
60, 364-67, 372, 415, 426, 435, 436 in United Kingdom, 124-25
Open Network Provision (ONP), 145, 182, in Yugoslavia, 280
306-8, 365-67, 372, 436 PCI, 372
Open skies policy, 301, 335 Pearce, Alan, 274
Open Systems Interconnection (OSI) model, Pebereau, George, 154, 155, 158, 328
116, 365, 413 Pedersen, Mogens Kuhn, 225-28
Optical switched-star cable systems, 53 Penny Post, 15-17
Orbita-Molniia satellite system, 284 Pentaconta system, 152
Organic Law of 1959 (France), 143 People's Republic of China, 110, 209
Organization for Economic Co-Operation and Peres, Shimon, 263
Development (OECD), 24, 171, 210, Perier, Denis, 388
213, 216, 230, 244, 267, 280, 293, 302, Perry, Charles R., 19
339-49, 363, 393-94, 402 Persians, 7
Orion Telecommunications, 335, 337 Personal networks, 119-20, 408
Orkla Borregaard, 222 Peru, 7, 325
Osterreichische Post und Pervyshin, Erlen, 287
Telegraphenverwaltung (PTV), 196-99 Petersen, Thane, 162
OTC, 288 Pfeiffer, Giinter, 45, 90, 98
Overcapitalization, 49, 58, 108, 328, 354-55 Pharos, 21
Phase E review, 330
Philips, Anton Frederik, 172
Pacific Telesis, 98, 250, 256 Philips, Frederik, 172, 173
Packet-switched networks, 143, 144, 171, Philips, Gerard, 172
182, 188, 211, 217, 222, 227, 228, 236, Philips (company), 24, 75, 119, 148, 152,
242, 249, 268, 271, 280, 286, 351, 371- 154, 156-57, 161, 169, 171-74, 176,
72 177, 183, 196, 256, 271, 279, 308, 317,
Pactel, 119, 121, 330 318, 325, 326, 383
Paging services, 145, 213, 218, 236, 268 Phillips, C., Jr., 427
Paliadent, 266 Phonepoint, 119, 163
PanAmSat, 300, 335, 337 Photovideotex, 384
Panzar, John C., 27, 29, 57, 354 Piano Europa, 244
Papandreou, Andreas, 270-72 Picaud, Michel, 146
Papandreou, Vaso, 271 Pichault, Frangois, 182
Pasok party, 272 Pierrand, Jean, 154
Patents, 40 Piher Semiconductor, 256
Patterson, Paul, 339 Pinhasi, Raphael, 266
Pautrat, Charles, 141, 143 Piore, Michael, 49, 408
Paxnet, 227, 228 Pipe, G. Russel, 393, 394, 402, 403
Payphones, 131, 268, 349 Piquet, Jacques, 192
514 Index
PlanCaiculs, 160-61, 318 Poste, La, 142, 165

Plessey, 24, 76, 107, 111, 125-27, 209, 260, Post Office Department (U.S. Postal Service),
261, 308, 315, 319 3
Pluralistic networks, 43, 49-50, 407, 425-39 Post Office Engineering Union (United
Pointel, 145 Kingdom), 105, 124
Point-of-sale services, 378 POUNC, 107
Poland, 277-78, 287 Pozzi, Giovanni, 244, 247
Policy directions, 61-65 Prado, Emilio, 253
Polish Post Office, Telegraph and Telephones Presentation Level Protocol (PLP), 382
(PPTT), 277-78 Prestel, 128, 183, 228, 316, 382-85, 389
Polish Transmission Works (PZT), 278 Prevot, Hubert, 165
Political power, 29, 408-9 Price discrimination, 34-36, 355-57, 369
Political telematique, 64-65 Privacy, 392-404, 429-30
Polo, Marco, 7 Privacy Act of 1974 (United States), 398
Pool, Ithiel de Sola, 52 Private networks, 45, 51, 432, 433
Popescu, Virgil, 281 Private optimum size, 33, 34, 36
Porat, Marc U., 44 Prodigy, 391, 409-10
Portugal, 43, 308, 330 Pronay, G., 276
cellular telephony in, 260 Protectionism, 319, 363, 403
equipment industry in, 260-61, 324 Proteo system, 245, 246
international service in, 341, 344 Provider specialization, 50-51
Open Network Provision and, 366 Prussian Post, 10-14
PBXs in, 261 PRX-205 system, 171
postal service in, 258, 259 PSTN, 271
telephony in, 25, 258-59 PTAT cable, 110
telex in, 258 PTOs (public telecommunications
videotex in, 260 organizations), 5, 23
Posner, Richard, 399 PTT Telematics Laboratories (PTL), 170
Postal Administration Law of 1954 (Germany), Public organizations, 45
80 Public Utility Commissions, 420-21
Postal and Telecommunications Services Act Puerto Rico, 325
of 1983 (Ireland), 233 Pupin, Michael, 72, 73
Postal-industrial complex, 23, 47, 52, 64, 434 Pupin coils, 72-73, 139
Postal Savings Bank (Germany), 81 Purton, Peter, 115, 198, 254, 260, 373
Postal service Pye, 173, 383
in Austria, 195
in Finland, 212 Quander, Peter, 300
in France, 133, 142 Quatrepoint, Jean-Michel, 65, 153
in Germany, 8-9, 13-14, 69, 70, 81, 91- Queck, R., 180
92, 94, 97 Quiles, Paul, 164, 165
in Hapsburg Empire, 3-4, 8-9 Qvorturp, Lars, 211, 229
history of, 7-8, 9-17
independence of telecommunications from, 6 Racal, 115, 118-19, 132, 145, 211, 250
in Ireland, 233 Racal Vodac, 118
in Italy, 240, 242 Rad Data, 266
in Netherlands, 169 Radio, 295
in Poland, 277 in France, 139, 146
in Portugal, 258, 259 in Germany, 74
in Prussia, 10-14 in Netherlands, 170
in Spain, 252, 253 in Switzerland, 188-89
in Switzerland, 186 in United Kingdom, 295
in United Kingdom, 14-17 Radio Austria, 196
Postal Workers Union (United Kingdom), 103 Radiocom 2000, 145
Post and Telecommunications Administration Radio France, 139
(Italy), 239, 242 Radio Law of 1927 (Finland), 213
Index 515
Radio Linja, 217 Romania, 270, 281-82

Radio Suisse, 188, 189 Romans, 7
Raduga satellite system, 284 Roos, Ingevar, 339
Rafferty, C. D., 235 Rosenbaum, Andrew, 243
Raggett, R. J., 119, 233, 269 Rothenberg, Jerome, 41
RAI, 243, 248 Roulet, Marcel, 163, 164
Railroads, 10, 55-56 Roussel, Anne-Marie, 144, 164, 246, 330,
Railway and Canal Commission, 22 373, 388
Ramsey pricing, 355 Roux, Ambroise, 152
Ram Telex, 265 Roux, Dominique, 147
Rank, 383 Royal Post, 14-16
Rathenau, Emil, 70, 72 R6 switching system, 139, 151
Rathenau, Walter, 74 RTVE, 252, 253, 255
Rausch, Jean-Marie, 165 Rubinstein, Amnon, 263
RCA Americom, 335 Rudolf II (emperor), 9, 91
RCA Corporation, 75, 153, 154, 159, 316, Rural Electrification Administration, 56, 323
333, 391 Russia, 256, 275, 281, 282-89
RC International, 227 Rutkowski, Anthony M., 99, 294, 296, 297,
Reagan, Ronald, 125, 246, 326 350, 360
Redistributory networks, 35-39, 43 Ryf, J., 187
Regie des Telegraphies et des Telephones
(RTT), 178-83, 305 Saba, 153, 318
Regional collaboration, 53 Sabel, Charles, 49, 408
Regnecentralen Computer (RC), 227 Sagem, 140
Regulation, 418-23, 425-28, 431, 435-36 Saint Gobain, 151, 162
Reichspost, 70, 71, 73, 74, 82, 83, 85 Salo, T., 217
Reis, Philipp, 69 Saloner, Garth, 358, 359
Rekola Commission, 215-16 Salop, S. C., 358
Renault, 148 Samsung, 278
Research and Development in Advanced Sanoma Corporation, 390
Communications Technologies for Europe SARIN, 390
(RACE) program, 305, 309, 318 Sarwar, I., 170
Reseau Numerique a ['Integration de Services Satellite communications, 44, 46, 52, 98,
(RNIS), 144 334-38, 432
Rete Fonia-Dati (RFD), 249 in Denmark, 229
Retevision, 252 Eutelsat, 163, 248, 278, 280, 282, 301,
Reuters, 376, 377 311, 336
Revel, Jean-Francois, 150 in France, 163, 164
Reynolds, Albert, 233 in Germany, 163
Riboud, Jean, 162 Intelsat, 28, 235, 247, 284, 298-300, 331,
Richardson, Jeremy John, 27 335-38
Rifa, 208 in Netherlands, 171
Ring, G. A., Company, 223 in Poland, 278
Robber barons, 62 in Russia, 284
Robinson, Kenneth, 328, 329 in Turkey, 268
Robinson, Peter, 394 in United Kingdom, 128
Robotron, 288 in United States, 299
Roche, Edward M., 394 Sauvant, Karl P., 394
Robotron, 288 Scandinavian Airlines (SAS), 211
Roche, Edward M., 394 Scanvest Ring Communications, 223
Rocks, David, 276, 277 Schankerman, Mark, 57
Rockwell International, 125, 159 Schares, Gail, 91
ROGALAND radio, 222 Schelling, Thomas, 41
Rohlfs, J., 368 Schenker, Jennifer, 79, 162, 185, 244, 271,
Rolm, 76 308, 373
516 Index
Scherer, Joachim, 83, 90, 298, 300 308, 315, 316, 318, 319, 325, 326, 328,
Schiller, Dan, 45, 350 329, 340, 374
Schlumberger, 162 Siemens-Nixdorf Information Systems, 77
Schneider, Friedrich, 192, 193 SIFRANET, 265
Schneider industrial group, 160 Sigismund Augustus (king of Poland), 277
Schon, Helmut, 363 Signetics, 173, 317
Schrack, 197-99 Silicon Valley, 47, 314
Schulte, Josef, 45, 78, 89 Simitis, Spiro, 404
Schulte-Braucks, Reinhard, 306 Simonsen Electro, 223
Schwarz-Schilling, Christian, 79, 93, 96, 98 Sims, Calvin, 119
Scientific-Atlanta, 125 Sindelka, Josef, 198
Scott, John Dick, 72 Singapore, 110, 163, 314, 320
SEA, 160 SINTEL, 253
Sears, 391 SIP. See Societa Italiana per 1'Esercizio
SEAT, 248, 390 Telefonico (SIP)
Securicor, 118 Sirm, 242
Seita, 148 Sitter, R., 266
Selenia-Elsag group, 248 Skendi, Stavro, 282
Selwyn, Lee, 171, 197 SKF, 395
Servan-Schreiber, Jean-Jacques, 314 Skyggnir earth station, 231
Service dispatch, 378 Slovenia, 280
Service economy, 44, 48 Smart Card, 316
Service quality interconnectivity, 430 Smau, 244
SESA, 144, 153 Smith, Adam, 12, 31, 49
SET, 240 Smith, Diana, 261
Seven Years War, 12, 13 Smyth, Eleanor C., 17
Sevres porcelain works, 148 Snow, Marcellus S., 89
SFT, 146, 152 Soares, Mario, 261
SFT-Ericsson, 151, 152 Social Democratic party
SOB, 318-19 in Belgium, 180
SGS, 154, 162-63, 248, 319 in Denmark, 226
Shapiro, Carl, 355 in Germany, 78, 92, 95, 396
Shared-tenant services, 46, 366, 410, 432 in Portugal, 261
Shas party, 266 Social interconnectivity, 429
Shaye, 119 Socialist party
Shearson-Lehman-Hutton, 98 in France, 146, 147, 149-50, 152-53
Shell, 119, 169 in Italy, 243
Shenda telephone company, 110 in Portugal, 261
Shepherd, William, 29, 57 in Spain, 255
Sichel, Bertha, 386 Social optimum size, 33-34
Siemens, Carl Friedrich von, 74 Social Partners (Austria), 196, 197
Siemens, Georg, 70-74 Social subsidies, need for, 5556
Siemens, Heinrich, 72 Social welfare, 41
Siemens, Sir William Carl, 72 Societa Finanziaria Telefonica (STET), 154,
Siemens, Werner, 69, 70, 72 163, 239-43, 288, 321, 390, 437
Siemens-Albis, 188, 192, 193 Societa Generale di Telefonia, 240
Siemens & Halske, 71-77, 82, 83, 85-88, 90, Societa Generale Italiana de Telefonia e
91, 95, 98, 100, 102 Applicazione Elettriche, 240
Siemens & Schuckert, 72, 74, 75 Societa Italiana per 1'Esercizio Telefonico
Siemens Brothers, 72 (SIP), 61, 239, 241-47, 249, 250, 326,
Siemens (company), 24, 27, 69, 70-72, 124- 390
26, 138, 139, 151, 152, 156-58, 161, Societa Telefonica della'Alta Italiana, 240
162, 173, 181, 183, 197, 198, 199, 217, Societe Anonyme de Telecommunications
218, 227, 228, 231, 245, 246, 260, 261, Radioelectrique et Telephonique (SAT),
268, 271-73, 276, 278, 279, 282, 288, 148, 152, 155
Index 517
Societe Europeenne des Satellites (SES), 184, See also ISDN (Integrated Services
185 Digital Network)
Societe Franchise du Radiotelephone (SFR), Standard Radio and Telefone AB, 324
145 Standards Promotion and Applications Group
Societe Generate de Belgique, 157, 181, 249 (SPAG), 303
Societe Generate de Telephone (SGT), 136, Standard Telefone og Kabelfabrik (STK), 223,
137 324
Societe Grammont, 139 Standard Telephone and Radio (STR), 187,
Societe Industrielle du Telephone (SIT), 138- 192, 193, 324
40, 151, 152 Standard Telephone & Cable, Lid. (STC),
Society of Telecom Executives (United 111, 119, 124, 126-27, 324
Kingdom), 105 Standard Telephone Company (STC), 324
Socotel, 152, 155 Stanhope, John Lord, 14, 15
Sofia Telecommunications Plant, 280-81, 287 Stanley, Kenneth, 331, 339
Solana, Luis, 254, 255 Staple, Gregory, 128, 244, 331, 339
Somitomo, 338 Stapley, Barry, 124
Sony, 119, 173 Star network architecture, 120
Sorman, Guy, 150 State, U.S. Department of, 323
Sotolec, 155 State Agency for Telephone Services (ASST)
South Korea, 314, 320, 326 (Italy), 239, 240, 242, 243, 249
Southwestern Bell, 349, 389 State regulatory commissions (United States),
Sovam Teleport, 285 323
Sovereignty, 402 Stationary cellular radio, 52
Space development, 311 Steckel, Marie-Monique, 45, 143, 144, 148,
Spain, 139, 163, 209, 251-57, 330 163
electronics industry in, 252 Steenbergen, T., 175
equipment industry in, 255-56, 308, 321, Steenbergen Commission, 175-77
324, 326 Stenbeck, Jan, 210
international service in, 341, 344 Stentor, 223
postal service in, 252, 253 Stephan, Heinrich, 4, 10-14, 17, 69-70, 72,
telegraphy in, 252-54 133, 294
telephony in, 25, 251-54 STET. See Societa Finanziaria Telefonica
telex in, 252-54, 308 (STET)
value-added services in, 253, 373 Stette Commission, 221
vertical integration in, 321 STIPEL, 240
videotex in, 254, 389 Storno, 226, 227
Spare-parts service, 378 Stourdze, Yves, 137
SPC switches, 171, 172, 206, 216 Strassman, Paul, 44
Special drawing rights, 333-34 Strategic private networks, 51
Specialization, 49-51 Stromberg-Carlson, 125
Special Telecommunications Action for Strowger system, 73, 103, 124, 139, 151,
Regional Development (STAR), 234, 188, 256, 260
310 Stuebing, Heinz Volkmar, 80
Spence, A. M., 358 Suard, Pierre, 158, 246
Sp-1 system, 152 Submarine cables, 19, 110, 127-28, 135, 153,
Sprint Networks U.S.S.R., 285 179, 187, 241, 247, 258, 265, 289, 294,
Staatsbedrijf der Posterijen, Telegrafie en 334, 336-38
Telefonie, 170 Submarine Lightwave Cable Company
Stage coach postal service, 16, 17 (SLCC), 337
Standard Electrica, 258, 260, 321, 324 Subnetworks, 39, 43, 377, 407, 425, 435
Standard Electric Kirk A/S, 324 Subsidies for network start-up, 3941
Standard Electric Puhelinteollisuus, 218 Supply-oriented perspective, 28
Standard Elektrik Lorenz (SEL), 75-77, 82, Supraregulation, 420-22
83, 85-87, 90, 102, 324, 326 Swaffield, J. J. E., 170
Standardization, 27, 358-59, 425-26, 433. Swarttouw, Frans, 174, 175
518 index
Swarttouw Commission, 174-76 Tassis family, 240

Swedcom, 207 Taxis, Alexander von, 13
Sweden, 110 Taxis, Franz von, 8
cable television in, 210 Taxis, Leonhard von, 9
cellular telephony in, 209, 210 Taxis family, 3, 8-10, 13
data protection in, 393, 397, 402 TDF-1, 145, 163
equipment industry in, 206-9, 321, 324 TDF-2, 163
international service in, 339-41, 344 TDM system, 84, 152, 153
ISDN in, 211 TDX Systems, Inc., 110, 330, 338
PBXs in, 205-7 Techint Group, 254
telegraphy in, 203 Technical incompatibility, 59
telephony in, 203-6 Technical interconnectivity, 425-26
telex in, 207 Technical Recommendations Applications
value-added services in, 373 Committee (TRAC), 307
vertical integration in, 321 Technical Telecommunications Committee
videotex in, 211 (TTC), 298
Swedish Telecom. See Televerket TECHNOMEN, 218
Swedtel, 207 TEFISA, 253
SWIFT (Society for Worldwide Interbank Tekade, 75
Financial Telecommunications), 305-6, Tekede, 85
376-78, 409, 433 Telcosur, 254
Swiss Broadcasting Corporation (SBC), 188, Telebild, 211
189 Telecom Denmark, 224, 225
Swisscom, 192 Telecom Eireann, 232-36
Swiss Industry Association, 192 Telecom Eireann Information Systems (THIS),
Swiss FIT, 186-94, 373 235
Switzerland, 70, 171, 295 Telecom Factbook, 48
cable television in, 189 Telecommunications Association (Ireland), 235
cellular telephony in, 188 Telecommunications Domestic Contract
data protection in, 397 (Germany), 83
equipment industry in, 192-94, 324 Telecommunications Installation Act of 1990
international service in, 344 (Germany), 91
PBXs in, 190 Telecom Nederland, 169-72, 174-77
postal service in, 186 Telecom-1, 163
radio in, 188-89 Telecom Portugal, 258, 259
telegraphy in, 186-87 Telecom Systemes Mobiles Alphapage, 145
telephony in, 5, 186-91 Telecom Telfa, 278
telex in, 188 Teleconferencing, 250
Symphonic, 163 Teleconseil, 192
Syncordia, 131, 373 Telecottages, 211
SysScan, 223 Telectron, 235
System 12, 77, 85, 127, 157, 158, 181, 183, Teledata, 228, 266
193, 223, 260, 268, 278, 280, 281, 288 TeleDenmark, 5, 226, 228
System 55, 83 Telediffusion de France, 309
System X, 107, 117, 125-27, 129, 131, 209 Telefax, 86-88, 172, 188, 265, 268, 271
System Y, 107, 127 Telefinans, 207
Telefonbau, 75
Tadiran, 265, 266 Telfones de Lisboa e Porto (TLP), 258-61
Tadiran Elisra, 266 Telefonica Argentina, 254
Tagliabue, Paul, 172, 173 Telefonica de Espana, 119, 159, 174, 246,
Taiwan, 314, 320, 326 251-56, 321, 326, 373, 437
Tanberg Telecom, 223 Telegrafverket, 204
Tandy Computer, 317 Telegraph Act of 1869 (United Kingdom), 19
Tarjanne, Pekka, 214, 296 Telegraphs and Telephones Act of 1904
Tassis, Janetto, 8 (Netherlands), 170
Index 519
Telegraphy, 351, 392 in Denmark, 25, 224-26, 348

in Austria, 195 in Finland, 212-18, 348
in Finland, 212-13 in France, 25, 45, 56, 135-44, 151-55,
in France, 133-35, 137 163-66, 348
in Germany, 14, 69-70, 81, 98, 100 in Germany, 14, 25, 45, 56, 69-74, 77-79,
International Telecommunication Union and, 81-101, 140, 141
293-96 in Gibraltar, 256-57
in Ireland, 232 in Greece, 25, 270-71
in Italy, 240, 241 in Hungary, 275-76
in Luxembourg, 184 in Iceland, 230-31
in Netherlands, 170 in Ireland, 25, 232-34, 348
in Norway, 220 in Israel, 262-64
in Russia, 286 in Italy, 25, 142, 209, 240-44
in Spain, 252-54 in Japan, 26
in Sweden, 203 in Luxembourg, 25, 184-85
in Switzerland, 186-87 in Malta, 250
in United Kingdom, 17-19, 127-29, 294 in Netherlands, 25, 169-71
Telegraphy Law of 1919 (Finland), 213 networks (see Networks)
Teleindustrier, 207 in Norway, 4, 220-23
Teleinvest, 207 in Poland, 274, 277-78
Telekomunikacj Polska SA, 277 in Portugal, 25, 258-59
Telekurs AG, 376, 378 provider specialization, 50-51
TeleLarm, 207 in Romania, 281-82
Telelogicab, 207 in Russia, 283-86
Telemap, 384 in Spain, 25, 251-54
Telemar, 242 in Sweden, 203-6
Telemarketing, 380 in Switzerland, 5, 186-91
Telemetry, 375 in Turkey, 267-68
Telenet, 371-72, 372, 382 in United Kingdom, 19-22, 25, 103-13,
Telenokia, 218 117-18, 128-32, 140-42, 209
Telenorma, 75, 77, 82, 83 in United States, 3-5, 46, 47, 49, 56-59,
Telepac, 188 61-62, 71, 112-113, 140, 349
Telepak, 211 universal penetration, 36, 45, 55, 56, 64,
Telephone, invention of, 69 111, 115, 323,429, 437
Telephone (audio) conferencing, 374-75 user demand, 44-45
Telephone Cable Manufacturers Association, waiting time, 347, 348
124 in Yugoslavia, 279-80
Telephone Company of Ireland, 232 Telepoint (CT-2), 119-20
Telephone Development Association, 103 Teleryhama, 217
Telephone Equipment Manufacturers Teleset, 390
Association, 124 Teleshopping, 379
Telephonometric Laboratory, 297 Tele Sonderjylland, 224
Telephony. See also Value-added networks Telespazio, 239, 241-43, 247-48
(VANs) Telesuisse, 192
activist role by equipment industry, 45-46 Telesystemes, 142-43
in Albania, 282 Teletas, 268, 269
alternative transmission paths, 51-52 Teletel, 145, 386-90
in Austria, 195-97 Teletex, 227
in Belgium, 25, 178-80, 182 Teletext. See Electronic mail
in Bulgaria, 280 Teletra, 278
cellular. See Cellular telephony Telettra, 77, 158-59, 245, 246, 255, 278,
comparison of charges, 339-46 279
in Cyprus, 273 Tele 2 AB, 110
in Czecho-Slovakia, 279 Televerket, 203-11, 250, 278, 286, 321, 339,
defense of monopoly in, 56-59 340, 437
520 Index
Televerkets Konkurranseorganisasjan (TBK), Three-Penny Post, 16

222 Thurn-and-Taxis Reichpost, 9, 13, 91
Telex, 74, 274, 306, 351 Thurow, Lester, 314
in Austria, 198-99 TIDSA, 253
in Belgium, 182, 308 Tiebout, Charles, 41
in Denmark, 227, 228 Tietgen, F. C., 226
in Finland, 218 Time, Inc., 390
in France, 140 Times Mirror, 391
in Germany, 79 Time-Warner, 47
in Greece, 271 TIMO, 240
in Iceland, 231 T&N, 83-86
in Israel, 265 Tomasson, Olafur, 230
in Italy, 243, 244, 247, 249 Tombras, Theofanis, 272, 273
in Luxembourg, 185 Tonnemacher, Jan, 385
in Netherlands, 171 Toussing, 185
in Norway, 222, 223 Toyota, 338
in Poland, 278 Trachsel, Rudolf, 193
in Portugal, 258 Trade unions, 4, 23
in Russia, 286 "Tragedy of the Commons, The" (Hardin), 42
in Spain, 252-54, 308 Transatlantic trade friction, 322-30
in Sweden, 207 Transborder data flows, 392-404, 419
in Switzerland, 188 Transcom, 144
in Turkey, 268 Transdata, 259
in United Kingdom, 121-23, 129, 298, Transdyn, 144
305, 332, 337, 422 Transistor, invention of, 315
TELI, 206-8, 321 Transmic, 144
Telic/Alcatel, 231 Transnationalism, 438-39
Telidon, 382 Transpac, 143, 144, 244, 388, 389
Telit, 245 Treaty of Rome, 90, 122, 298, 301, 306, 307,
Tel-Optik, 337, 338 319
Telpac, 259 Trilogy, 161
Telrad, 266 Triumph-Adler, 77
Telset, 390 Trofimov, I. F., 288
Telstar, 128 TRT, 152, 155, 156, 333, 335, 389
TELVE, 240 TRW, 299
Temex, 396 TsNIIS, 288
IOC, 287 Tube, Sir Brian, 14
Territoriality, 407, 432 Tullock, Gordon, 41
Tesla Electronics Group, 279, 280 Tully, Shawn, 158
Tesla Karlin, 279 Turkey, 43, 348
Tesla-Liptovsky, 279 cellular telephony in, 268
TETI, 240 equipment industry in, 268-69
Texas Instruments, 316 international service in, 341, 344
Thatcher, Margaret, 63, 104, 105, 125 PBXs in, 268
Thiemeyer, Theo, 91 satellite communications in, 268
Thirty Years War, 9 telephony in, 267-68
Thomas, Hilary, 389 telex in, 268
Thomson, 162, 164, 248, 308, 318, 319 Turkpak, 268
Thomson-Brandt, 161 Turn, Rein, 393, 403
Thomson-CSF, 125, 146, 148, 152-56, 163, Tutt, Nigel, 171, 182
287 TWA, 47
Thomson Group, 24, 147, 148, 149 Two-Penny Post, 16
Thomson-Houston, 139, 151 Tydeman, John, 383
Thorn-EMI, 119, 127, 209, 318, 383 Tymnet, 131, 372, 373, 382
Thorngren, Bertil, 207, 210 Tymshare, 372
Index 521
Unequal user size, 38-39 telephony in, 3-5, 46, 47, 49, 56-59, 61-
Ungerer, Herbert, 366 62, 71, 112-13, 140, 349
Unidata, 15 value-added services in, 115, 369-70, 380
Unified Automated System of videotex in, 374, 383, 390-91, 409-10
Telecommunications (EASS), 284 United States International Trade Commission,
Uniformity in regulation, 421-22 329
Uniform Settlements Policy, 333 United Telecommunications, 372
Unilever, 169 United Telekom (UT), 278
Uninet, 372 United Telephone Company (Ireland), 232
United Kingdom, 52, 103-32 United Telephone Company (United
cable television in, 53, 120-21 Kingdom), 19, 20
cellular telephony in, 118-20 Unite!, 119
data protection in, 396-97 Universal Communications, 223
duopoly review, 113, 132, 361 Universal Postal Union (UPU), 297
electronics industry in, 47-48, 104, 318 Universal service, 35, 36, 45, 55, 56, 64,
entry barriers in, 26 111, 115, 323, 429, 437
equipment industry in, 104, 115, 123-27, UNIX-based systems, 227
317, 320-21, 324 Unlisted telephone numbers, 400
international service in, 339, 344 URA 2G system, 153
ISDN in, 129 US Sprint, 209, 285, 332, 333, 348, 410
PBXs in, 124-25 USTT, 325
postal system in, 14-17 Ustwer, 192
radio in, 295 US West, 119, 121, 276, 279, 288, 330, 349,
satellite communications in, 128 389
telegraphy in, 17-19, 127-29, 294
telephony in, 19-22, 25, 103-13, 117-18, Vail, Theodore, 4
128-32, 140-42, 209 Value-added networks (VANs), 5, 6, 28, 51,
telex in, 121-23, 129, 298, 305, 332, 337, 56, 107, 111, 113-17, 303, 308, 369-
422 81, 433, 434
Thatcher government in, 48, 63, 104, 105 basic packet transmission, 371-72
value-added services in, 113-17, 372 in Belgium, 182
vertical integration in, 320-21 conceptual framework for, 370-71
videotex in, 128, 316, 320-21 in Denmark, 228
United Kingdom Telegraph Company (UKTC), in France, 144-45, 164, 372
18 generic services, 370, 373-76
United Nations Center on Transnational in Germany, 372, 373
Corporations, 394 in Greece, 273
United Nations Intergovernmental Bureau for in Ireland, 236
Informatics, 394 in Israel, 265
United States in Italy, 372, 373
cellular telephony in, 329 in Netherlands, 175, 177, 372
containment of equipment industry in in Norway, 222
Europe, 323-25 in Spain, 253, 373
data protection in, 398, 402-4, 429 specific applications, 376-80
deregulation in, 23, 62, 65 in Sweden, 373
electronics industry in, 47, 315-17 in United Kingdom, 113-17, 372
European exports to, 327-30 in United States, 115, 369-70, 380
European interpretation of Value-added tax, 74, 429
telecommunications policy of, 322-23 Valvo, 173
international services, 331-36, 338-40, 344 VAN Infonet, 144
International Telecommunication Union and, van Rosenthal, C. Jansen, 310
295-96 Vasseur, Frederic, 151
ISDN in, 361-63 VCRs, 173
open network architecture in, 364-66, 436 Veblen, Thorstein, 312
satellite communications and, 299 Vedel, Thierry, 56, 386, 387
522 index
VEF, 287, 288 Warren, Samuel D., 399

Velazquez-Gaztelu, Candido, 254 Waverman, Leonard, 57
Venezuela, 209, 255 Way-leaves (rights-of-way), 20, 21
Venture capital, 317-18 Wealth of Nations, The (Smith), 12
Vertical integration, 50, 320-21 Weinstock, Lord, 125
Very large-scale integration (VSLI), 316, 318 Western Electric Company, 76, 138, 140, 151,
VET, 287 327, 328
Victoria (queen of England), 19 Western Union, 3, 209, 332
Video broadband, 426, 427 Western Union International, 332, 334
Videoconferencing, 128, 188, 375 West Germany. See Germany
Videotel, 250, 390 Westinghouse Company, 72
Videotex, 45, 51, 79, 98, 382-91 Wheatstone, Charles, 17, 18
in Austria, 196, 198 Wheatstone pointer telegraph, 71-72
in Belgium, 183, 384, 389 White, David, 253
in Denmark, 228 Whitehead, Alfred North, 42
in Finland, 215, 217, 390 Whitehouse, Bob, 207, 214, 372
in France, 145, 382, 383, 386-90 Wide area networks, 432
in Germany, 79, 91, 382, 383, 385-86, 389 Wiegner, Kathleen, 317
in Italy, 250, 383, 384, 389, 390 Wieland, Bernhard, 45, 90, 98, 169, 174, 176
in Japan, 383, 389 Wigand, Rolf T., 397
in Luxembourg, 185, 389 Wild, T. p., 187
in Netherlands, 171, 383, 384 Wilde, Candee, 174
in Norway, 222 Wilderness and Island Telephone Network,
in Portugal, 260 218-19
in Spain, 254, 389 Wiles, Peter, 312
in Sweden, 211 Wilhelm II (emperor), 74
in Switzerland, 188, 384, 389 Williamson, John, 214, 217, 256
in United Kingdom, 128, 316, 382-85 Willig, Robert D., 27, 29, 57, 354
in United States, 374, 383, 390-91, 409-10 Wilson, C., 368
Videoton, 276 Wilson, Robert W., 57
Viditel, 171 Windebank, Lord, 15
Vienna Congress of 1815, 13 Wireless Telegraphy Act of 1904 (United
Vienna Privat-Telegraphen-Gesellschaft, 195 Kingdom), 127
Viewtel, 383 Withering, Thomas, 15
Virgin Islands Telephone Company, 325 Witte, Eberhard, 80, 86, 92, 95, 96
Virtual private networks, 51 Wittiber, 85
VISTACOM, 218 Wolf, Roswitha, 89
Vodaphone, 107, 118-19 Wolf, Thomas, 5
Voge, Jean, 143 Woollacott, Emma, 128, 373
Voice mail, 51, 373-74 World Bank, 276
Voice retrieval (audiotex), 51, 374, 410 World Center for Computer Science and
Volkswagen, 77, 248 Human Resources (Centre Mondial de
Volvo, 211 ['Information et Resource Humaine), 314
von Beust, Christopher, 8, 9 Woroch, Glenn A., 29
von Weizsacker, Carl Christian, 88, 89, 190, Wiihrmann, Karl, 189, 194
354, 376
VSAT (very small aperture terminals) service,
337 Xenophon, 7
Xerox Corporation, 355
Waden, Daniel Johannes, 212
Wagner, Adolph, 91
Waiting time, 347, 348 Yaacobi, Gad, 263
Wallenberg Group, 208 Yellow Page service, 379
Wallonia, 180, 181 Young, Peter, 126
Walters, A. A., 108 Yugoslavia, 279-80, 282, 287
Index 523
Zacharisen, W. M., 21 Zippori, Mordechai, 263

Zegveld Commission, 177 Zurich Telephone Company, 187,
Zellweger, 192, 193 190
Zenith, 162 ZVE1, 85, 86
Zero sum society, 314 ZVTf Banska Bystrica, 279
Zion Cables, 266 ZWUT, 278

Telecommunications

Uploaded by

Copyright:

Available Formats

Telecommunications

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Telecommunications

Uploaded by

Copyright:

Available Formats

Telecommunications in Europe

COMMUNICATION AND SOCIETY

New York Oxford

Copyright 1992 by Eli Noam

Library of Congress Cataloging-in-Publication Data

New York E.N.

I TRADITION AND CHANGE

1. Public Telecommunications: A Concept in Transition 3

II THE TELECOMMUNICATIONS SYSTEMS OF EUROPEAN COUNTRIES

Telecommunications in the Benelux and Alpine Countries

10. The Netherlands 169

Telecommunications in Scandinavia and the North Atlantic

15. Sweden 203

19. Iceland 230

Telecommunications in the Mediterranean Countries and Eastern Europe

21. Italy 239

III BATTLEFRONTS IN TELECOMMUNICATIONS POLICY

28. The International Organizations of Telecommunications 293

IV THE FUTURE OF TELECOMMUNICATIONS

38. Networks in the Future 407

The Monopoly System

A number of far-reaching changes in telecommunications policy originated in

Early Postal Systems

The key institutions of European telecommunications are the PTTsthe gov-

Private Postal Service: The Taxis System

able. In today's language, they engaged in "cream skimming," though it was

The Development of the Classic PIT: The Prussian Post

The Establishment of the PTT Monopoly in Britain

The Struggle for a Telegraph Monopoly in Britain

an analog system in which a needle pointed to letters of the alphabet. The

The Struggle for Telephone Monopoly in Britain

municipal responsibility, as were gas and water. He went on to predict: "gas

The PTT System at Its Maturity

Developments similar to those described for German and British telecommuni-

private companies (Siemens in Germany, CGE and Thomson in France, GEC

"unborn technological generations." Organizations of large users within Eu-

According to John Kenneth Galbraith, the "great advantage of being in the

The Concept of Network

Networks are an important concept in society and economy. They abound as

beyond their explanatory power. Used single-mindedly, they have degenerated

Theories for the Emergence of Multiple Networks

1. "More powerful technology leads to new transmission options, thereby

Market Structure Explanations

A New Approach: The Tipping of Network Coalitions

Figure 3.1. Stages in Network Expansion.

The Basic Model5

is declining, at least at first.

which holds where the marginal utility of an added subscriber is

Entitlement Point and Universal Service Obligation

network at least to where C'(ri) = P = u(n), leaving fixed costs to be distrib-

Exit from the Network

Political Price Setting and Redistribution

This, then, is the redistributory outcome, assuming no discrimination within

Monopoly and Expansion

Unequal User Size

Subsidies for Reaching Critical Mass

might subsidies to the start-up of a network as a whole have to be provided

Social Welfare and Multiple Networks

If network associations can control their memberships, stratification is inevita-

together only as long as arbitrage is prevented, as long as the minority cannot