Who Learns What from Whom: a Review

of the Policy Transfer Literature

DAVIDDOLOWITZAND DAVIDMARSH
Department of Political Science and International Studies,
University of Birmingham
As Rose argues policy transfer or lesson drawing is common:
Every country has problems, and each think that its problems are unique
. . . However, problems that are unique to one country . . . are abnormal . . .
confronted with a common problem.’ policy makers in cities, regional
governments and nations can learn from how their counterparts elsewhere
responded.’
It has always existed but there can be no doubt that the rapid growth in
communications of all types since the Second World War has accelerated the
process. Not surprisingly, the increase in policy transfer has led to the
development of interest in the topic among students of comparative politics and
public policy. This article reviews this literature which is of two types. First,
there are studies which do not use the concept but which often throw consider-
able light on policy transfer. Second, there is a growing body of material
dealing specifically with the process. Both types of study are now common. As
an example, the June 1994 edition of Political Studies included two articles
dealing with the issues discussed here. Walsh’s examination of exchange rate
policy in France and Germany is a good representative of the first type of study.
He does not discuss transfer, but shows how international capital flows gave
rise to tensions in the European Monetary System which, in turn, led to policy
convergence. In contrast, Coleman’s analysis of policy convergence in banking
directly addressed the question of how international economic changes affect
policy goals, policy content, policy instruments and policy style. The aim of
this article is to present a critical review of the literature on policy transfer in
order to both introduce the topic to a wider audience and contribute to its
development.

Our review is both narrower and broader than the major existing review by
Bennett.* Bennett’s approach is broader in that he concentrates upon the
general pattern of convergence between the policies adopted by nations. In
contrast, our focus is more narrowly upon the transfer of specific policies as a
result of strategic decisions taken by actors inside and outside government. At
the same time however, our approach is broader, particularly because we:

consider transfer within nations as well as between nations; acknowledge the

significance of actors outside the government machine in the process; and place
more emphasis upon how governments, international organizations and trans-
national companies can push other governments into policy transfer. The full
implication of these differences will become clear by the end of this review.
We have organized this review around a series of questions which are raised
explicitly or implicitly by the literature. What is policy transfer? Who transfers
policy? Why is there policy transfer? What is transferred? Are there different
degrees of transfer? From where are lessons drawn? What factors constrain
policy transfer? What problems are there with the literature?
What is Policy Transfer?
Policy transfer. emulation and lesson drawing all refer to a process in which
knowledge about policies, administrative arrangements, institutions etc. in one
time and/or place is used in the development of policies, administrative
arrangements and institutions in another time and/or place. We do not think
that the difference in nomenclature is overly significant. However, we do not use
the terms policy transfer and lesson drawing interchangeably, unlike Rose who
refers to the overall process by which policy or institutions are transferred as
lesson drawing3
In our view, the term ‘lesson drawing‘ implies that political actors or decision
makers in one country draw lessons from one or more other countries, which
they then apply to their own political system. This focuses on ‘voluntary’ policy
transfer, which occurs as a result of the free choices of political actors; yet, as we
shall see later, an important category of policy transfer involves one govern-
ment or supra-national institution pushing, or even forcing, another govern-
ment to adopt a particular policy. Policy transfer, a term that can cover
‘voluntary’ and ‘coercive’ transfer, seems to us to be a more appropriate term.
However, we use the term lesson drawing because, in many cases, lessons are
drawn from other places or times which do not result in policy or institutional
change. So, for example, in some cases a negative lesson may be drawn about
how not to proceed.
The study of policy transfer emerged gradually as a sub-set of the compara-
tive politics literature. Before 1940 most comparative studies focused on the
formal institutions of government and were thus ‘state centred’ and overly
descriptive. During the 1940s such approaches became less fashionable and
studies began examining how civil society interacted with the state. By the 1960s
a key focus was upon comparative policy analysis.
As the field of comparative policy analysis emerged, a number of authors,
notably Walker. showed considerable interest in process termed policy
diffusion. These initial studies were not concerned with the specific content of
the policies being transferred, but rather sought explanations of diffusion based
on timing, geographic propinquity and resource similarities; as such, they had a
much narrower focus than later studies of policy t r a n ~ f e r . ~As diffusion studies

advanced the weaknesses of the approach emerged; so that by the 1980s a major
critical review argued: ‘(the) major problem of this research tradition is that it
reveals nothing about the content of new policies. Its fascination is with process
not s~bstance’.~It was as a result of this perceived need to answer questions
ignored by diffusion studies that comparative policy analysts began discussing
lesson drawing and policy transfer.
Initially transfer studies primarily focused upon voluntary transfer, seeing the
process as one in which policies implemented elsewhere were examined by
rational political actors for their potential utilization within another political
system. More recently the literature has begun to address questions concerning
coercive transfer.
Who Transfers Policy?
This literature identifies six main categories of actors involved in policy transfer,
although in any specific case of transfer more than one category of actor is likely
to be involved: elected official$ political par tie^;^ bureaucrats/civil servants;*
pressure groups;’ policy entrepreneurslexperts; and supra-national institutions.
The role of the first four is well rehearsed, but the last two categories merit brief
discussion.
Policy EntrepreneurslExperts
As Rose points out, policy entrepreneurs are not only important to lesson
drawing because of their advocacy of lessons, but also because: ‘their concern
with a special subject . . . leads them to build up a nationwide or international
network of contacts that are a source of ideas for new programs’.’’
The role of such policy entrepreneurs in the international spread of an idea is
clearly illustrated in Rowat’s study of Ombudsmen. He argues:
There is no doubt that some of the ombudsmen themselves have been very
influential in bringing about the further spread of the institution,
particularly Professor Hurwitz of Denmark . . . and Sir Guy Powles of
New Zealand.”

In this way experts from inside and outside government interact to spread

ideas in what Haas terms epistemic communities.’*

Supra-national Organizations

As Rose argues:

Intergovernmental and international organizations encourage exchanges of

ideas between countries . . . The European Community and OECD

encourage exchanges among advanced industrial nations . . . and the

World Bank and the United Nations agencies focus on programs of

concern to developing countries.’3

Rose goes on to emphasize that the European Community: ‘promotes com-

parison . . . so that member states can become aware of what their competitors

are doing and decide which elements of foreign programs they may wish to copy

or adapt.14 Similarly, Bennett found that international organizations were

important actors transferring data protection p01icies.I~ One problem with

these studies is they pay too little attention to the fact that international

organizations are also involved in coercive transfer, an issue we shall return to

later.

Why Engage in Policy Transfer?

Lessons can be used in a number of ways and for a variety of reasons. Both

supporters and opponents of various policies use lessons selectively to gain

advantage in the struggle to get their ideas accepted. As Robertson says: ‘policy

lessons from abroad often are put forward as politically neutral truths. Beneath

this . . . adversaries are just as often using such lessons as political weapons’.I6

In attempting to explain why transfer occurs we must return to the distinction

between voluntary and coercive transfer, although it should be acknowledged

that at the margin the two types merge. In addition, a further distinction needs

to be made between cases of direct coercive transfer and those cases in which the

push factors leading to policy transfer are more indirect.

(a) Voluntary Transfer

Most authors suggest that the primary catalyst of voluntary transfer is some

form of dissatisfaction or problem with the status quo. Supporters of the

dissatisfaction model presume that when governmental policies are functioning

properly there is no need to search for lessons; everything can operate through

established routines. Only when routines stop providing ‘solutions’ is it

necessary to search for lessons.”

A number of causes of dissatisfaction are identified in the literature, but most

authors argue that dissatisfaction usually results from a perception, either by
the Government or the public, of policy failure. However, there is a clear
problem with this analysis because it assumes that policy failure is a non-
contentious and easily measurable concept. In fact, as Anderson points out,
political actors definitions of policy problems are subjective and ‘political’.I8
For example, although the dramatic increase in world wide unemployment led
President Clinton to call for a ‘Job Summit’ with G-7 nations in 1993, the
Thatcher Administration, in contrast, viewed inflation, not unemployment, as
the key economic problem.
As Polsby argues, elections may also force candidates to search for lessons
because:
A central part of the regular quadrennial public competition for the
presidential office is that candidates . . . must find programs to which they
can attach their names. (As such) Presidents who wish to differentiate
themselves from their predecessors and . . . make an individual mark on
history . . . provide a steady market of policy innovation.”
Certainly, Heclo amply illustrates the role party competition played in the
development of social policies in Britain and Sweden.20
As Bennett argues: ‘evidence is used in the policy process in highly selective
ways to legitimate decisions already taken. Thus, information about the effects
of a program elsewhere enters debate to justify prior positions’.*’ This process is
clear in a study by Henig, Hamnett, and Feigenbaum of the international
spread of privatization. In this field, government’s, committed to ‘New Right’
free market economic policies, used lessons drawn from countries which had
already privatized to legitimate their decisions to implement particular forms of
privatization.22
Uncertainty about the cause of problems, the effects of previous decisions or
the future causes actors to search for policies they can borrow. As Haas argues:
‘the hallmark of complex interdependence is uncertainty . . . International
collaboration . . . is an attempt to reduce uncertainty’ 23 Indeed, he demon-
strates how uncertainty led policy makers to turn to international experts in
their attempts to regulate pollution in the Mediterranean.
(b) Direct Coercive Transfer
The most direct method of coercive policy transfer is when one government
forces another to adopt a policy. Majone provides an excellent example in his
analysis of the reasons countries pursue regulatory policies. In particular, he

demonstrates how American regulators forced European states to adopt

American anti-cartel laws:
It is well known that the anti-cartel clauses of the ECSC treaty . . . were
significantly influenced by the American model . . . Washington, repre-
sented by John McCloy . , . insisted more than once on a particular wording
of individual article^.'^
The direct imposition of policy transfer on one country by another is rare.
However, supra-national institutions often play a key role in coercive policy
transfer. For example, supra-national institutions have played a crucial role in
the spread of Western monetary policies to Third World countries. Inter-
national Monetary Fund or World Bank loans are much cheaper for these
countries, but in return the IMF will stipulate certain economic policies that
have to be implemented if the loan is to be granted.25 The European Union
(EU) and the European Court of Justice are other supra-national institutions
forcing policy transfer upon member nations. Indeed, Shapiro shows how the
EU has functioned as a policy-pusher, using its power to issue directives and
regulations, while the European Court of Justice has forced Governments to
adopt policy directives the EU has issued.'6 Trans-national corporations
(TNCs) can also force governments into policy transfer because they hold a
trump card. They can threaten to take their businesses elsewhere. This
potential mobility can force governments to adopt policies capable of
attracting industries. Certainly, Heidenheimer et al. suggest that such threats
have constrained some Governments in relation to environmental regula-
tion.?'
(c) Indirect Coercive Transfer
Hoberg emphasizes the potential role of externalities, or functional inter-
dependence, in indirect coercive transfer. His analysis shows that the Canadians
looked to America for lessons which could be used in drafting their environ-
mental regulation, largely as a result of the indirect effect US pollutants, and
regulations, had on the Canadian side of the border.28
There is little doubt that externalities. which result from interdependence,
push government to work together to solve common problems. In this way,
Haas demonstrates how such interdependence was responsible for policy
transfer and the development of common environmental policies in the
M e d i t e r r a n e a ~ ~ ~Similarly, Majone discovered that externalities stimulated the
development of European chemical regulations; in response to American

legislation controlling imports of toxic substances, European countries turned

to the EC to develop common r e g ~ l a t i o n s . ~ ~
Technology can also push governments into policy transfer because of the
speed with which it forces change. Governments, not knowing how to deal with
the issues technological advances create, turn to each other for precedents and
ideas.31
The world economy constrains individual governments and economic
pressures can lead to transfer. In this vein, M ~ r a n ~ ~and Coleman33 show
how global economic integration led to convergence in the institutional
structures and trading practices in the securities market and the banking sector
respectively. When one country depends upon another as its primary market it
may also be pushed into adopting similar policies. Hoberg argues: ‘historically,
much of Canadian economic policy has been focused on resisting the natural
north-south pull of economic activity in N ~ r t h - A m e r i c a ’ . ~ ~
A country can also be indirectly pushed towards policy transfer if political
actors perceive their country as falling behind its neighbours or competitors. As
Bennett puts it: ‘fears of being left behind on an important public issue can
trigger attention. The cumulative effect of action elsewhere may translate into a
feeling of insecurity about being the ~ d d - m a n - o u t ’ . ~ ~Certainly, Hoberg argues
that national embarrassment, caused by the perception that Canada was falling
behind the United States in environmental regulation, pushed policy makers to
emulate American auto-emission standards during the 1 9 8 0 ~ ~ ~
The emergence of an international consensus may also act as a push factor.
When the international community defines a problem in a particular way, and
even more when a common solution to that problem has been introduced in a
number of nations, then nations not adopting this definition or solution will
face increasing pressure to join the international ‘community’ by implementing
similar programmes or policies. In this way Bennett explains Canada’s intro-
duction of data protection legislation as a response to the pressures associated
with the international consensus emerging around the importance of data
protection during the 1 9 8 0 ~ . ~ ~
What is Transferred?
As is clear from this review to date, it is wrong to think of policy transfer merely
in terms of the successful transfer of policy. Bennett’s review focuses largely
upon the transfer of policy goals, content and instrument^.^^ In our view this
focus is still too narrow. In fact, we identify seven objects of transfer: policy

goals, structure and content; policy instruments or administrative techniques;

institutions; ideology; ideas, attitudes and concepts; and negative lessons.
In several studies Wolman traces the transfer of urban development policies
between the United States and Britain and argues that: ‘what potential
borrowers appear to be most concerned with is program structure rather than
effect‘.39Wolman demonstrates how the transfer of the US UDAG programme
to Britain involved the borrowing of the entire programme structure. However,
he also points out that both general policy ideas and specific policy instruments
can be transferred, but that the borrower may pick and choose what to borrow:
‘in many cases a specific policy idea . . . may be borrowed, but the specific design
or structure through which this occurs in the original country may not be’.40
Robertson and Waltman show how various leaders and administrators used
policy instruments or administrative techniques, for example ‘experience rating’
and productivity measures, borrowed from other countries. Indeed, they argue:
’most public policy borrowing appears to involve administrative techniques . . .
rather than a change of policy dire~tion’.~’However, this is contentious, for
Majone concludes that, while there is considerable policy transfer between
nations, few cases involve policy in~truments.~?
The institutions used to implement policy can also be transferred. Haas
demonstrates that the establishment of many environmental Ministries to
protect the Mediterranean involved substantial borrowing of ideas and
Similarly. Training and Enterprise Councils (TECs) in England
and Wales and Local Enterprise Companies (LECs) in Scotland were based
upon American Private Industry Councils ( P I C S ) . ~ ~
Robertson and W a l t m a r ~ ~ ~demonstrate that President Reagan and Prime
Minister Thatcher adopted each other’s ideological rhetoric to justify and
spread political programmes based upon ‘New Right’ principles. More specifi-
cally, the analysis by Henig ef al. of the development of privatization politics
and policies in Britain, France and the United States indicates that other
governments sent delegations to Britain to learn about both the details of the
privatization programme and the ideology underpinning it.46
Political actors can also transfer ideas, attitudes and concepts; although this
category overlaps the last one. During the 1980s the rhetoric adopted to justify
changes in the British welfare system was drawn directly from American
writings which emphasized dependency, reciprocal obligations and using
welfare to help recipients help them~elves.~’
Similarly, in his analysis of the development of data protection legislation,
Bennett shows that transfer led to policy convergence around the principle of
‘fair i n f ~ r m a t i o n ’ . ~ ~Interestingly, while the countries adopted a shared
comprehensive set of data protection principles and legislated to protect
citizens against abuses, none designed the same policy instruments to enforce
these principles.
Lessons can be negative as well as positive. As an example, after examining
1970s American auto-emission standards, Canadian legislators made an explicit
decision not to emulate American standards, believin them to be unnecessarily
the development of American data protection legislation, legislators and
bureaucrats specifically rejected the Hessian statute as inappropriate as it
protected the interests of the state over those of the people.50
restrictive, given Canada’s environmental s i t ~ a t i o n . ~!?In the same way, during
Are there Different Degrees of Transfer?
When engaged in policy transfer actors have a range of options on how to
incorporate lessons into their political system, Rose identifies five: copying;
emulation; hybridization; synthesis; and in~piration.~’We prefer to combine
the two related categories of hybridization and synthesis.
Copying occurs when a country adopts a programme in use elsewhere
without any changes. The easiest way to prove that copying has occurred is to
examine the wording of the legislative bill authorizing a p r ~ g r a m m e . ~ ~
Emulation happens when a country: ‘reject(s) copying in every detail, [but]
accepts that a particular program elsewhere provides the best standard for
designing legislation at home’.53Hybridization and synthesis involve combin-
ing elements of programmes found in two or more countries to develop a
policy best-suited to the emulator. Finally, studying familiar problems in an
unfamiliar setting can expand ideas and inspire fresh thinking about what is
possible at home.s4
From where are Lessons Drawn?
When policy makers begin searching for lessons, their own country’s past is the
logical place to begin.55By searching the past, agents learn not only what has
worked but can learn what not to repeat. So, Rose claims: ‘the defence policies
of Germany and Japan today represent lessons learned about what not to d ~ ’ . ~ ~
Similarly, in examining the rise of Keynesianism during the Depression, Weir

and Skocpol found social democratic policy ideas were re-articulated from past
attempts to implement Keynesian type programme^.^^
Searching for lessons within a political system‘s past has the advantage of
saving time and resources. However, searching the past involves subjective
evaluation; while history is constant it is open to many interpretations and
current situation’s may not be truly analogous to past ones. Indeed, when
drawing lessons actors might not truly understand the past or its relation to the
present. At the same time, given that most policy develops over time, deciding
from which period to draw lessons can critically affect the lesson drawn.
Finally, as Rose points out, time is not a constant when actors are engaging in
policy transfer: ‘obstacles to lesson-drawing are not permanent; in the course of
time many obstacles become variable^'.^^
When searching within a nation, actors may start with their own organization
because different branches and divisions provide opportunities for transferring
administrative techniques and procedures. As Rose emphasizes: ‘the first lace
Actors can also draw lessons from other political systems within their own
country if its constitutional structures create a series of similar sub-national
units of government within a relatively harmonious political culture. As an
example, the United States 1988 Family Support Act, was developed based
upon the experiences of state welfare programmes during the 1980~.~’
Although constraints exist, it is common for governments and agents to
transfer policies from one nation to another. So, for example, in the welfare
area international policy transfer has played a crucial role in the spread of poor
laws and unemployment legislation since the nineteenth century. Not only were
British poor houses adopted in the United States during the nineteenth century,
but, as Heclo shows: ‘experience in Denmark and New Zealand were important
in Britain . . . however, it was experience in Germany that stirred greatest
international interest’.6’
Some countries and regional governments commonly export policies to other
nations and regions. While examining policy transfer between the United States
and Canada, Robertson and Waltman found that particular American States
and Canadian Provinces acted as policy leaders: ‘(their) innovations are
disproportionately copied by neighbouring jurisdictions’.62 Heclo shows that
Germany played a similar role in the transfer of unemployment insurance to
Sweden and Britain during the early twentieth century.63 Similarly, Bennett
found that the 1973 Swedish Data Act acted as an exemplar to other nations
instituting data protection laws.

The basic assumption involved in drawing lessons from other nations is that:
‘similarities are greater within a given program across national boundaries than
among different programs within a ~ountry’.~’When drawing lessons across
nations, geographic propinquity does not equate with policy transfer because
ideological and resource similarities are necessary preconditions to adapt
lessons from one country to another and neighbouring countries do not always
meet these preconditions. We return to this question below.
What Factors Constrain Policy Transfer?
The complexity of a programme affects its transferability; the more complex a
policy or programme is the harder it will be to transfer. Although the effect of
complexity on transferability is an understudied area, Rose suggests six
hypotheses:
1 programmes with single goals are more transferable than programmes
with multiple goals;
2 the simpler the problem the more likely transfer will occur;
3 the more direct the relationship between the problem and the ‘solution’ is
perceived to be the more likely it is to be transferred:
4 the fewer the perceived side-effects of a policy the greater the possibility of
transfer;
5 the more information agents have about how a programme operates in
another location the easier it is to transfer;
6 the more easily outcomes can be predicted the simpler a programme is to
transfer.66
These hypotheses offer a basis for future research but attention also needs to
be paid to the way complexity interacts with the other factors discussed here to
help shape what is transferred and in what form.
Rose makes a key point which students of policy neglect at their peril:
Policy makers are inheritors before they are choosers; as a condition of
taking office they swear to uphold the laws and programs that predecessors
have set . . . new programs cannot be constructed on green field sites . . . they
must be introduced into a policy environment dense with past commitment ~ . ~ ’
Past policies constrain agents as to both what can be transferred and what
agents look for when engaging in policy transfer. In this way, Heclo demon-
strates how past policy constraints influenced the approaches Britain and
Sweden adopted when developing social policies during the twentieth century.68
Similarly, Robertson and Waltman show how existing policy commitments
constrained transfer between the Reagan and Thatcher Administrations,
despite the fact they shared ideological goals.

As Wolman shows, the institutional and structural constraints faced by

agents transferring policies are crucial. So, Wolman demonstrates how the
United States’ federal structure acted as a constraint on the transfer of policies
from the unitary British system, while the British system facilitated the Govern-
ment’s ability to transfer policies from the United States.70
Policy transfer is also dependent upon the transferring political system
possessing the political, bureaucratic and economic resources to implement the
policy. Robertson suggests that both transfer, and the success of transfer, is
more likely if the policy is consistent with the dominant political ideology in the
‘host’ c o ~ n t r y . ~ ’Certainly, ideological similarities between countries can be a
key factor when actors look for lessons.’* Examining the transfer of environ-
mental policies between the United States and Canada, Hoberg found that one
reason for the Canadian adoption of American environmental standards was
the ‘value consensus’ which existed between the two nations.73 Similarly,
Kelman explained the convergence of safety and health regulations in Sweden
and the United States as a product of the ideological consensus between
government officials.74
Rose emphasizes that the bureaucratic size and efficiency may influence
transfer.75 Similarly, even desirable programmes will not be transferred if
implementation is beyond a nation’s technological abilities. For example,
developed nation’s emission standards generally require the installation of high
technology monitoring equipment and filters which are beyond the techno-
logical and monetary resources of most Second and Third World countries.
Even developed countries might decide not to transfer policies because of the
technological complexities involved. On several occasions Canada explicitly
rejected particular American environmental protection policies because the
technology used to implement them was too expensive and r e ~ t r i c t i v e . ~ ~
Implementation costs money so economic resources are another critical
constraint for agents engaged in policy transfer. In this way, Bennett demon-
strated the importance economic constraints played in the transfer of data
protection principles during the 1 9 7 0 ~ ’ ~
What are the Problems with the Literature?
(i) Is it a Dependent or an Independent Variable, or Both?
Some authors, like Bennett, use policy transfer as an independent variable; the
process of policy transfer is used to explain why a particular policy was
adopted. In contrast, other authors, for example Rose, treat it as a dependent
variable; attempting to explain why transfer occurs. Obviously, both concerns
are legitimate, but the second seems to us more interesting

There is plenty of evidence that policy transfer occurs and is an important

cause of policy innovation. However, we need much more work on how and
why it happens; perhaps organised around the series of questions identified in
this review.
(ii) Beyond Pluralism
Policy transfer has been strongly associated with a pluralistic perspective on
power and the policy process. Most authors discussing policy transfer believe
that the process brings new actors and ideas into decision making, expanding
the number of actors involved in the policy process. However, at best, this is one
of two competing hypotheses. The pluralist origins of much of the literature are
clear from: the overemphasis upon politics and, particularly, the role of
politicians/bureaucrats and upon intentional, as distinct from structural,
explanation; and the assumption that political actors are rational, calculating
subjects.
If policy makers are looking to draw lessons from polities which are similar in
institutional, economic and cultural makeup, it might be argued that, instead of
expanding the number of ideas and actors involved in the decision-making
process, policy transfer enhances the power of a relatively small circle of actors
who consistently draw lessons from each other. So, if policy transfer occurs
within relatively closed international policy communities,instead of introducing
new ideas, lesson drawing simply reinforces the existing system, for the same
ideas will be circulated amongst like-minded nations, thus maintaining the
status quo. Of course, to a large extent this is an empirical question; we can
analyse the extent to which policy transfer limits participation and underpins the
existing order. However, policy transfer is underpinned by pluralist assumptions
and the key question of whose interest is served by this process goes unasked.
Obviously, policy transfer involves strategic political judgements and, as we
saw, the key actors involved in the process are politicians and bureaucrats. As
such, the concept fits in well with the ‘statist’ developments in both pluralism
and other state the~ries.’~However, it is not possible to explain policy transfer
solely in terms of the decisions of political actors. The strategic decisions of
actors are taken in a context which is mostly not of their making and which is
characterized by structured inequality.
Some of these constraints are national and political, resulting from a
country’s own institutional structures and processes. Rose, like most analysts of
policy transfer, plays down such constraints:
Even though institutions are necessary, it does not follow that they are
important . . . A new program can usually be administered in more than one
way; insofar as this is the case, one institution can be substituted for
another.79
This conclusion is misguided. Indeed, the general trend in the wider
comparative politics literature, especially in the New Institutionalism and
state-centred approaches, is towards emphasizing the importance political

356 Review Section

institutions play in shaping policy. Institutions structure the actions and values
of actors working within them: ‘political institutions determine, order [and]
modify individual motives . . . in terms of institutional interests’.*OAdditionally,
as March and Olsen point out: ‘expectations, preferences, experiences and
interpretations of other actions are constructed within political institutions’.’’
More specifically, they also influence which actors search for lessons and where:
‘political behaviour [is] embodied in institutional structures of rules, norms,
expectations, and traditions that [limit] free play of individual will and
There are other international constraints involving either international
agencies or TNCs. Indeed, in some cases of coercive transfer actors have little or
any autonomy of action. Certainly, a political leader in a Third World country
has little alternative but to accept the policies imposed by the World Bank or
the IMF, given that the consequences of refusal are deepening debt and
economic and, probably, political crisis. Similarly, even in a developed country,
a TNC may have sufficient economic power to influence Government policy in a
way which is analogous to an international beauty contest; a TNC threatening
not to invest in a given country unless policy concessions are made.’3 More
generally, changes in the international economy, and in the location of a
particular nation within it, constrain the autonomy of politicians and may push
them towards the emulation of solutions which are favoured by international
companies, financial institutions and, perhaps specifically, holders of the
country’s currency or debt.
This is not to deny the role of individual actors in policy transfer; but neither
should one deny the crucial role of political structures. However, pluralism
tends to downplay structural and economic explanation and over-emphasize
intentional and political explanation. This is a plea for a broader approach. We
need to recognize that political actors operate within structural constraints.
Some of the constraints derive from the political institutions within which
actors operate, other are economic in origin based upon the structured
inequality which characterizes societies.
Much of the literature on policy transfer also assumes that actors involved in
transfer are rational and calculating. In this view, the process is driven by
dissatisfaction. Policy makers act to to reduce the possibility of incurring
electoral defeat. It is also assumed that decision makers evaluate the causes and
the effects of problems. The key to policy transfer is thus prospective evalu-
ation: ‘prospective evaluation starts by observing how a program operates in
another country and develops a model of what is required to produce its effects
However, dissatisfaction is not the sole cause even of voluntary transfer.
At the same time, policy making is not inevitably, or perhaps even usually, a
rational process. Rather, it is often a messy process in which different policy,
solution, and problem streams need to combine at the appropriate moment for
a policy to develop.85 A few authors have begun to examine the process of
policy transfer in the light of more sophisticated models of policy making.86
More work of this sort is clearly needed.
There are three macro-theoretical positions which, potentially, offer alterna-
tives to pluralism, and within which it may be possible to interpret and explain
policy transfer: new institutionalism; rational choice theory; and neo-Marxism.
Each is dealt with in a longer version of this paper available from the authors.
The key point here is that policy transfer is a meso-level concept which does not
necessarily have to be integrated into a pluralist perspective.
(iii) Too Much Positivism
Most policy transfer studies have an inadequate conceptualization of the role
subjective perception and judgements play in the definition of problems and
solution. Some authors pay lip service to the problem: ‘the definition of a
satisfactory, or at least a “not unsatisfactory”, program is problematic. The
aspirations against which achievements are judged are not given, but a social
c o n s t r ~ c t i o n ’ . ~ ~However, few scholars look at how the definitions of problems
or solutions are socially constructed. More importantly, they ignore the way
that subjective definitions affects how and where searches are conducted and,
also, what solutions are considered. Both Anderson88 and W 0 1 m a n ~ ~attempt to
demonstrate the influence subjective perceptions have upon definitions of
public problems but, again, much more work is needed.
Conclusion
Policy transfer refers to the process by which actors borrow policies developed
in one setting to develop programmes and policies within another. It is an
important area to study because transfer is a common phenomena. We have
suggested a series of questions which can be used both to organize our current
knowledge of the process and to guide future work. However, it is crucial that
the concept is developed beyond the pluralist perspective which has almost
exclusively underpinned research in this area. We have suggested a number of
alternative frameworks within which the process of policy transfer can be
understood and explained. However, further consideration of these broader
questions is essential if the literature is to progress beyond description.

’ R. Rose, ‘What is lesson drawing?’, Journal of Public Policy, 11 (1991), 3-30, p. 3.

’ C. Bennett, ‘What is policy convergence and what cau