Regulatory Policy and the Social Sciences

Regulatory

Policy and

the Social

Sciences

California Series on

Social Choice and Political Economy

Edited by Brian Barry and Samuel L. Popkin

Regulatory

Policy and

the Social

Sciences

Edited by Roger G. Noll

UNIVERSITY OF CALIFORNIA PRESS

Berkeley Los Angeles London

University of California Press

Berkeley and Los Angeles, California

University of California Press, Ltd.

London, England

© 1985 by

The Regents of the University of California

Library of Congress Cataloging-in-Publication Data

Main entry under title:

Regulatory policy and the social sciences.

Based on papers presented at the Conference on Regulation and the Social Sciences, held at Reston, Va., Jan. 1982, and sponsored by the National Science Foundation’s Regulation and Policy Analysis Program.

Bibliography: p.

1. Trade regulation- United States-Congresses.

2. Social sciences-Research- United Sutes-Congresses.

I. Noll, Roger G. 11. Conference on Regulation and the Social Sciences (1982: Reston, Va.) III. Regulation and Policy Analysis Program (National Science Foundation) HD3616.U47R144 1985 338.973 85-16515

ISBN 0-520-05187-4 (alk. paper)

Printed in the United States of America

123456789

Contents

Contents

Preface

Acknowledgments

PART I An Overview of Social Science Research on Regulation

1 Introduction Roger G, Noll

2 Government Regulatory Behavior A Multidisciplinary Survey and Synthesis Roger G. Noll

PART II

Regulation in the Larger Social Setting

3 The State in Politics The Relation Between Policy and Administration TheodoreJ. Lowi

COMMENT: John Ferejohn

4 On Regulation and Legal Process Lawrence M. Friedman

COMMENT: James V. DeLong

5 A Wide Angle On Regulation An Anthropological Perspective Laura Nader and Claire Nader

COMMENT: Carol MacLennan

PART III The Politics of Regulation and Deregulation

6 Group Concentration and the Delegation of Legislative Authority Morris P. Fiorina

COMMENT: William H. Riker

7 Why the Regulators Chose to Deregulate Martha Derthick and PaulJ. Quirk

COMMENT: Kenneth A. Shepsle

PART IV Applications of Social Scientific Methods: Case Studies

8 Regulation of Risk A Psychological Perspective Paul Slavic, Baruch Fischhoff, and Sarah Lichtenstein

COMMENT: Richard A. Winett

9 Conflicting Regulations Six Small Studies and an Interpretation Theodore Caplow

COMMENT: W. Richard Scott

10 Self-Regulation as Market Maintenance An Organization Perspective Mitchel Y, Abolafia

COMMENT: Roger G. Noll

11 Integrating Themes and Ideas

Neglected Areas of Research on Regulation James Q. Wilson

Focusing Organizational Research on Regulation Philip Selznick

Bibliography

Appendix List of Participants, CalTech/NSF Conference on Social Science and Regulatory Policy, 22-23 January 1982

Preface

A considerable amount of congressional and executive branch government activity consists of creating public policy to redistribute income and to protect or use resources efficiently. Of the many ways available to achieve these objectives, governments frequently chose one despite evidence suggesting that it is not only unlikely to accomplish the objectives but is even liable to create an inefficient use of resources. I refer to regulation—governance by command and control. The study of this phenomenon has become a virtual subdiscipline, especially within economics but also within political science and law.

The ubiquitous use of regulation raises questions that are the province of study for many behavioral and social sciences. Why do organizations, especially governments, choose regulation instead of other options that employ incentives and disincentives? What behavioral reactions does regulation induce? How does government organize to effect regulation, and how successful are its organizational modes? What alternatives to regulation may accomplish the objectives more efficiently?

It is easy to recognize that these questions concern not only economic behavior but political, social, cultural, and psychological behavior as well. Yet contributions to the literature on regulation are mostly from economics. Thus, the purpose of the Conference on Regulation and the Social Sciences, held at Reston, Virginia, in January 1982, was to seek out knowledge from disciplines other than economics and to encourage more direct research on the subject of regulation. This undertaking was founded on the belief that contributions of theory, method, and tools from many disciplines will continue to enliven the study of regulatory phenomena as a legitimate, discrete field of scholarly inquiry. Thus, the National Science Foundation’s Regulation and Policy Analysis Program was delighted to fund the conference and the preparation of this book, which is a synthesis of the findings of that conference.

In this volume leading scholars of anthropology, law, political science, psychology, and sociology describe and analyze the phenomenon of regulation from the perspective of their own discipline and also contrast work in their discipline with that in economics. The volume contains individual studies and proposes research agendas for each of the represented disciplines to pursue.

The volume begins with Roger Noll’s introduction and brilliant review and analytic synthesis of the multidisciplinary literature on regulation. These chapters prepare for the rest of the book by describing the work of economists on regulation in reference to that of scholars from other disciplines.

Three chapters in the volume are concerned with the definition and characterization of regulation from the perspectives of different disciplines. Laura Nader and Claire Nader apply the anthropologists’ holistic approach to understanding regulation in complex industrial societies. They view government regulation in the context of the private and governmental, social and cultural organization of American society. Theodore Lowi explores the political science of policy analysis by identifying and characterizing public policies as four types: regulatory, distributive, redistributive, and constituent. He analyzes the organic statutes creating government agencies in the United States and France to classify each agency according to his scheme. Lowi argues that the form of an agency and of its regulations is dependent on the type of legislative rule it enforces. Lawrence Friedman develops the theme that regulation operates through the law and that the legal system is a reflection of the social and moral beliefs of society.

The volume also contains studies of particular aspects of regulation that illustrate the contributions of political science, psychology, and sociology. Morris Fiorina extends a model of the political incentives that lead legislators to delegate regulatory powers to administrative bodies. He finds that a model maximizing pecuniary and nonpecu- niary net benefits to a legislator’s district satisfactorily describes the decision processes resulting in regulatory delegation. Examining three agencies that have deregulated, Martha Derthick and Pau) Quirk search for the reason these agencies made policy changes toward less regulation even before congressional legislation mandated the reform. Their findings give weight to the force of prevalent ideas.

Paul Slovic, Baruch Fischhoff, and Sarah Lichtenstein explore the applications to regulation of research in two areas of cognitive psychology: the limits of human intellectual ability, and the perception of risk. The authors believe that research on intellectual ability suggests that the rationality assumption of economics is not a good predictor of behavior and needs modification. Research on risk suggests the need to consider social issues in setting safety standards.

Theodore Caplow reports on six interview and questionnaire studies of organizations subject to large amounts of regulation. He studies the relation of organizational structures and functions to the management of perceived conflicts between regulations. The study advances a number of interesting explanatory hypotheses and recommends an agenda for further research. Mitchel Abolafia examines the case of self-regulatory behavior and motives in commodity futures exchanges. He advances the view that self-regulation may be pro-competitive in reducing transaction costs to all parties and that cartelizing activities are held in check by the threat of government intervention.

The book concludes with a thought-provoking series of commentaries by three leading scholars of regulation: Bruce Ackerman (law); James Q. Wilson (political science); and Philip Selznick (sociology).

I hope readers find that the book has merit in its own right and that it stimulates more good research on regulation as a social, cultural, political, and legal phenomenon as well as an economic one.

The National Science Foundation thanks all the conference participants for the effort and interest that made this project a success. Special thanks are due to Roger Noll for creating and structuring the conference, carrying it to a successful conclusion, and then contributing to and editing this volume.

Laurence C. Rosenberg

National Science Foundation

Acknowledgments

Several organizations helped support the preparation of this book, and I am grateful to each. The Regulation and Public Policy Program of the National Science Foundation provided financial support for preparation of the papers, the conference on which they were based, and the refereeing process. Final editing was completed while I was in residence at the Center for Advanced Study in the Behavioral Sciences, supported by a fellowship from the John Simon Guggenheim Memorial Foundation and by the Center’s institutional grant from the National Science Foundation. The remaining costs associated with the final preparation of the manuscript were financed by the Energy Policy Studies Program of the Caltech Environmental Quality Laboratory.

During the process of completing the book, I received able assistance from some excellent people. Kathryn Kurzweil performed with dedication and dispatch the numerous secretarial and administrative tasks necessary to coordinate and prepare the final text. Susan Davis, with her amazing administrative talent, organized the conference and assisted me in coordinating the process by which the papers were refereed and revised for publication. Jane-Ellen Long performed the small miracle of editing the manuscript, sometimes quite heavily and always with skill, without offending a single author! Finally, I want to thank Robyn and Kimberlee for their understanding and support as they spent a highly constrained summer vacation in Mammoth Lakes watching me edit this book.

Roger G. Noll

April 1984

PART I

An Overview of

Social Science

Research on

Regulation

1

Introduction

Roger G, Noll

Government regulation of business is a ubiquitous feature of the American economy. Originating in attempts by state government to establish political control of infrastructural industries such as railroads and grain elevators, the concept of creating a government bureau to constrain business activities has mushroomed into a complex, multiagency, and multijurisdictional system that touches every commercial and nonprofit organization in the country.

Not surprisingly, the growth of regulation caused its size and scope to become a major political issue in the 1970s. Two successive successful presidential campaigns—Carter’s in 1976 and Reagan’s in 1980—included regulatory reform and selective deregulation as principal campaign issues. Certainly this political prominence is a far cry from the political invisibility that scholars and regulators in the 1950s claimed made reform of regulatory policy difficult. The new- found political salience of regulatory policy has led to widespread changes in regulation since the mid-1970s: deregulation in several spheres of the economy, various forms of mandatory impact analysis in social regulatory agencies that are part of cabinet departments, and new approaches to monitoring regulatory decisions by Congress and the courts.

Research by social scientists has played an important role in the reform of regulatory policy. Pferhaps the best evidence that scholarly research has affected regulatory policy is that several scholars whose reputations were built, at least in part, upon their research on regulatory policies have been appointed to important government positions in regulatory affairs. Notable examples are Alfred Kahn, Elizabeth Bailey, and Michael Levine at the Civil Aeronautics Board; Paul MacAvoy, Murray Weidenbaum, and George Eads at the Council of Economic Advisers; Stephen Breyer as the staff person who organized the investigations by Senator Edward Kennedy that led to airline deregulation; Darius Gaskins at the Interstate Commerce Commission; and James Miller and George Douglas at the Federal Trade Commission.

All the social sciences contain some research on regulatory policy; however, only in two—economics and political science—has the work been sufficiently extensive to constitute a subheld of the discipline. Moreover, a fair assessment of the impact of this work is that research by economists has been by far the most influential in the policy process. Again, evidence in support of this proposition is the dominance of economists in the roster of scholars who have occupied important positions in regulatory affairs since the wave of reform began in the mid-1970s.

The purpose of this book is to explore the scope of research on regulation in social science disciplines other than economics. The premise of the book is that research on regulatory policy in disciplines other than economics can be of both intellectual and practical value and ought to be encouraged. In political science, of course, there is already a body of interesting scholarship; however, with few exceptions it has lacked practical focus and has had little influence on policymaking.

Although research on the economics of regulation is both powerful and useful, its focus does not include the full range of social scientific questions that could have important practical significance. Economics is especially useful for analyzing the effect of institutional arrangements on economic efficiency. With respect to regulation, the principal contribution of economics has been to show theoretically and empirically how regulatory policies alter the incentives of businesses and how changes in incentives in turn affect the efficiency of the markets in which the regulated businesses operate. Economists have also analyzed the effects of regulation on income distribution. This work has been especially useful in three respects. First, it has identified instances in which regulation achieves little more than protecting a narrow special interest in a manner that creates substantial inefficiencies, and thereby it has provided the informational foundation for extensive deregulation in energy, transportation, communications, and financial institutions. Second, it has identified alternative approaches to command and control regulation when the purposes of intervention can be accurately characterized as economic in character. The impact of this work has been less sweeping than the effects of the work supporting deregulation but has sometimes led to important policy reform—notably, in the movement to ward greater use of incentives in environmental regulation. And third, this work has been sufficiently influential to cause its approach to be institutionalized in most regulatory agencies by the development of mandatory benefit-cost analysis and by the creation of economic analysis groups that play an important role in developing policies and priorities in regulatory agencies.

As important as this work has been, it does not provide a solid intellectual foundation for rationalizing all aspects of regulatory policy. To the extent that regulatory policy serves objectives that are not narrowly economic in character, analysis that focuses exclusively on technical efficiency and income distribution will not constitute a complete, comprehensive basis for policy actions. In its broad sense, economic efficiency can incorporate all the values normally assumed to be missing from economic theory: fairness, justice, equity, altruism, legitimacy, etc. Efficiency arguments in economics turn on the ability of the economic system to do the best possible job in serving individuals as final consumers. If individuals are willing to sacrifice more narrowly economic objectives in order to gain these other values, economic theory can incorporate this phenomenon in a rather straightforward way. But practically speaking, theory and measurement are vastly better developed for analyzing prosaically economic activities and objectives of members of a society. Presumably, scholars whose research focus is on these noneconomic aspects of life know best how to incorporate these aspects into a broader view of the regulatory process.

Research on the economics of regulation has also paid relatively little attention to the problem of designing maximally effective organizations for promulgating regulatory policy. Again, the problem is not that in principle economics has nothing to say about designing organizations: economists have played an important though not dominant role in the development of organization theory, especially concerning the internal incentive structure of organizations. But little of this work has found expression in research on regulation, just as relatively little work in organization theory from other disciplines has focused on problems of designing regulatory agencies. Consequently, the scholarly community has been able to provide little useful advice on how reorganizing agencies or restructuring the relationships among agencies, courts, and the legislature might affect regulatory decisions. If the primary issue is whether to have regulation at all (if, for example, the most important question is whether to deregulate or whether to switch to another instrument of public policy such as taxation, subsidization, or a new form of tradeable and enforceable property right), the application of organization theory to the design of regulatory agencies is not very important; however, wherever regulation is the preferred policy instrument, organization theory becomes the primary lever for improving performance and hence assumes great importance.

Another set of research questions arises from the tendency of nearly all research on regulation by economists to assume a strong form of economic rationality. On questions relating to decisionmaking in environments characterized by substantial uncertainty, the theory is more normative than positive, in that its emphasis is on characterizing the nature of a rational approach to decisionmaking rather than describing how people actually behave. This work provides useful techniques for decisionmakers who are faced with uncertainty, and it provides a means of evaluating the quality of decisions in any given decision making environment. But other interesting issues remain. One is how to educate decisionmakers to behave more rationally when facing complex decision problems characterized by incomplete and probabilistic information. Another is how to construct decisionmaking environments so that people are likely to make rational decisions—or at least how to avoid situations in which people systematically make bad mistakes in dealing with uncertainty.

Finally, economic research alone cannot be relied upon to illuminate the process by which regulatory policy is adopted and promulgated. Again, the problem is not that economists have nothing, or only incorrect things, to say about the connections between economic policy and political processes: economists have played an important role in the development of explanations of how narrow economic interests influence the political process. The difficulty is that there is ample evidence that there is more to regulatory politics than this. The deregulation movement of the 1970s and the concomitant influence of economists in reforming regulatory policies were certainly not predictable from purely economic models of the political process. There is some irony in the fact that in the act of demonstrating the emptiness of a public-interest theory in many areas of regulation, several influential economists played a major role in causing a pronounced shift to serving what most would regard as the public interest in those very areas. An important question is how and why many poorly performing regulatory policies were turned around in the late 1970s, and what if anything can be done to the structure of government to insure that this is not a temporary phenomenon.

The purpose of this book is not to provide the last word on any of these questions. Instead, it is to assemble papers that suggest in some detail promising avenues of policy-relevant research on regulation in disciplines other than economics. The book is addressed to two audiences: research scholars who might find useful ideas that will influence their future research, and government officials who might gain a greater comprehension of the potential contribution of social science to understanding and improving the regulatory process.

The genesis of the book was an interest on the part of the National Science Foundation in expanding the scope of its Regulation Program. To this end, the Regulation Program provided financial support for a conference to explore what useful research on regulation might be done by scholars in such disciplines as political science, psychology, and sociology, as well as possible social-scientific perspectives in legal scholarship. Scholars in each of these disciplines were told the purpose of the conference and asked to write papers that would explicitly identify valuable research questions. Two forms of papers were solicited: articles reporting the results of a research project as an example of a general category of useful research for scholars in social science, and essays that explained how a research literature in social science not directly focused on regulatory policy might usefully be applied to studying regulation.

To provide background and coherence to the collection of papers, before they began work on their papers all the authors were supplied with two monographs. One was a survey by Paul Joskow and me (1981) of the economics literature on regulation. The other was an earlier version of Chapter 2 of this book. Chapter 2 addresses the problem of building a theory of the regulatory agency based not only on what scholars in the social sciences have had to say specifically about the regulatory process, but also on the more general literatures on political processes, bureaucracy, organization theory, and the law.

First drafts of all the commissioned papers were discussed at a conference in January 1982. The formal discussant of each paper was a scholar from the same discipline as the author. In addition, conference participants, including several economists, were invited from government agencies and universities. (A complete list of participants is given in the Appendix.) The purpose was to address regulatory research in a multidisciplinary fashion, giving authors maximal opportunity to profit from the knowledge of practitioners and scholars in other fields. Each paper was then revised—most, substantially so—on the basis of the discussion at the conference. Each paper except Chapters 2 and 11 was then submitted for review by an anonymous referee who is a distinguished scholar in the same discipline as the author. The referees were asked to evaluate the papers according to the standards of leading journals in their fields. Finally, two additional referees were asked to review the entire manuscript. In a few cases, the refereeing process led to another revision.

Part I provides the framework for the book by explaining its origins and providing a unifying structure in which the remaining chapters can be placed. Part II contains essays that examine how regulation fits into a broader pattern of social institutions and values. Part III focuses on the politics of regulatory change: political actors’ motives in writing regulatory legislation and how politics affect the decisions of the agencies. Part IV contains detailed studies of specific issues in regulatory policy. The examples are the use of research in cognitive psychology in developing regulatory policies to deal with risk; the use of organization theory to shed light on the problem of conflicting regulations—how it arises and how it can be ameliorated; and the use of organization theory to understand self-regulation— why it develops and how it can be used for valid social purposes without leading to cartelization. Part V contains three general statements about promising lines for future research on central questions about regulatory processes.

2

Government

Regulatory Behavior

A Multidisciplinary Survey

and Synthesis

Roger G. Noll

Government regulation is a pervasive feature of the American economic system. It is a uniquely American approach to the political control of market processes. The purpose of this chapter is to review the range of theories to explain the development and direction of regulatory policy and to point out lines for further research.

First, a definition of regulation is in order. All levels of government attempt to control some private-sector economic decisions to which the government is not a party. One such method of control is to assign to a government agency the responsibility of writing rules constraining certain kinds of private economic decisions, using a quasijudicial administrative process to develop these rules. This process has two key features. First, the job of the agency is to channel and to alter the direction of an economic activity that is generally regarded as desirable to society. Second, because private ownership of property and its exchange through markets are protected constitutional rights, the agency must satisfy elaborate procedural and evidentiary

Part of the financial support for preparing an earlier version of this essay was provided by the National R&D Assessment Program of the Natural Science Foundation, grant DA 39495. Research assistance was supplied by John Allen, Marcia Bencala, Maryly Crutcher, and Barry Weingast. Especially helpful comments on an earlier draft were provided by Ross Eckert, Paul Joskow, Kai Lee, Michael Levine, and Charles Perrow.

rules if it seeks to constrain market activities. These rules give rise to the quasi-judicial process by which the agency develops its policies and means of controlling economic activities within its sphere of responsibility. The bureaus organized to undertake such tasks are herein referred to as regulatory agencies, a definition that is developed and defended in Noll (1980). Included in this classification are agencies that control aspects of transactions such as the price or the quality of the good transacted, that mandate certain features of the production process such as emissions-control methods and worker-safety requirements, or that control entry, as by licensing. Excluded are agencies that are public enterprises in that their primary function is either to procure private goods or to produce government goods, that try to prevent certain types of behavior or transactions rather than manage them, that manage government financial affairs such as tax collection or control of the money supply, or that try to alter market behavior by subsidy or by placing conditions on government procurement when the government is an important but not the sole entity on the demand side of the market.

The reason for this focus on regulation is more practical than theoretical. It may well be the case that the theory of regulation is very close to the general theory of government policy, but regulation is a distinct kind of policy that has spawned a distinct theoretical and empirical literature; indeed, in economics and political science its study has been elevated to the status of a subdiscipline. Moreover, because the state of knowledge about the development of the character of public policies is still rather primitive, the focus at this stage should be on specific kinds of policies.

Regulatory agencies come in many sizes and forms. Some are headed by commissions—a group of coequal heads who make decisions by voting on formal proposals, much like a legislature—while others have a single administrative head. Some are independent agencies technically outside the President’s administrative control, while other are lodged in executive branch departments. Some are what amounts to the first court in the judicial system, with the power to fine regulated firms or even to ban them from markets, while others must achieve their ends by fighting regulated firms in the federal courts. Some have very narrow responsibilities, such as the appropriately named Packers and Stockyards Administration or the Commodities Futures Trading Commission. Others, like the Occupational Health and Safety Administration, regulate every business in the nation.

Despite their variety, all regulatory agencies make many decisions that, in principle at least, affect economic efficiency. The economics literature as well as regulatory law emphasizes the effects of regulation on static efficiency—that is, the effect of regulatory decisions on costs, prices, and product quality, given unchanging technology and consumer tastes. By controlling prices, profits, entry, and the attributes of products or processes, regulators directly alter the net economic benefits derived from the regulated industry. To the extent that regulatory rules counteract market imperfections, they contribute to economic efficiency; to the extent that they reduce production efficiency or confer monopoly market positions on regulated firms, they reduce efficiency.

In addition, the policies of regulatory agencies also effect the dynamic efficiency of regulated markets, that is, whether the rate and pattern of technological change in regulated industries are economically efficient. In some cases, agencies have an explicit mandate to influence technology, either by subsidizing research and development (as was the case with the Atomic Energy Commission prior to 1974) or by imposing technical requirements intended to improve the performance of a particular industry (as was the case in the promotion of UHF television by the FCC [Webbink, 1969]). More commonly, agencies play a relatively passive role, approving or disapproving adoption of new technology on the basis of other policy mandates (Ackerman and Hassler, 1981; Capron, 1971; Peltzman, 1974; Warford, 1971). Finally, agencies make decisions on matters not directly related to the choice of technology but that, perhaps unexpectedly to the agency, indirectly influence technological change. Thus, regulatory lag—the time required for regulators to change regulations in response to changed conditions—may change the incentive of regulated firms to innovate, and rate-of-return regulation may bias innovation of regulated firms in favor of more capital-intensive technologies (Westfield, 1971).

The research literature on the effects of regulatory agencies on economic efficiency reaches generally harsh judgments. Government agencies are pictured as ineffective in dealing with market-failure problems such as environmental externalities or monopolistic control of markets, while they generate serious liabilities by protecting business against competition, thwarting warranted technological and economic changes, and imposing significant costs on consumers without producing many benefits.

This chapter does not evaluate the empirical literature on the undesirable economic effects of regulatory agencies (for a review of this work see Joskow and Noll, 1981). Instead, it presumes that agencies do have such effects. The purpose here is to examine the literature on government processes and bureaucratic organizations in search of plausible explanations of the selection of regulatory policies. In principle, theories of government policymaking behavior could provide some basis for evaluating proposals to reorganize a regulatory agency, to redesign its mandate, or to replace regulation with some other form of government policy, and for predicting the likely longterm effect of a regulatory agency.

Let us examine an illustration. During the first term of the Nixon Administration, the President’s Advisory Commission on Executive Organization (1971)—the Ash Council—proposed a sweeping reorganization of the executive structure of the federal government. The council’s report on regulatory agencies contained numerous proposals motivated in part by the perception that independent regulatory authorities create economic inefficiency. The proposals included transferring the agencies to the executive branch, replacing commissions with single administrative heads, streamlining the decisionmaking process, and combining regulatory agencies with related responsibilities into single agencies. A relevant question is whether existing knowledge concerning the operation of government bureaucracies provides support for the notion that these organizational changes would significantly improve the policies of regulatory institutions and the performance of regulated firms.

Dissatisfaction with the performance of a regulated industry can lead to reform proposals of four types.

1. Reorganization. The theory underlying reorganization proposals is that the location of regulatory responsibility within the governmental hierarchy and the organizational structure of the agency significantly affect policy outcomes. Since the early 1970s, this has been a particularly popular category of reform. Some examples are:

Moving an independent commission into an executive department. For example, the Federal Power Commission, an agency responsible for regulating natural-gas prices both at the wellhead and in pipelines, interstate wholesale sales of electricity, and hydroelectric projects, was renamed the Federal Energy Regulatory Commission (FERC) and placed in the Department of Energy when the latter was created in 1978.

Combining several offices with related responsibilities into a new agency (perhaps with additional powers and responsibilities). Two examples are the creation of the Consumer Product Safety Commission out of a product-safety office in the Food and Drug Administration and a hazardous-product labeling responsibility at the Federal Trade Commission, and the creation of the Environmental Protection Agency largely by taking some responsibilities from the Department of the Interior.

Changing the internal structure of an agency. One way is by creating new administrative subunits. In the early 1970s, the Fed* eral Communications Commission (FCC) created the Cable Tele* vision Bureau and the Office of Plans and Policy, the latter a group that undertakes economic analyses of the effects of pro* posed changes in regulation. Another structural reform is to change the number of administrative leaders of an agency. In the early 1980s, the number of commissioners at the Interstate Com* merce Commission (ICC), which regulates surface transporta* tion, and the FCC were reduced from eleven and seven, respectively, to five apiece.

Reshuffling policy responsibilities among agencies. One example is the transfer of the program for subsidizing the maritime industry from the independent Federal Maritime Commission (which also regulates the industry) to the Department of Transportation. Another is the move of oil-pipeline regulation from the ICC to the FERC.

2. Procedural Reform. Administrative law, legal precedent, and the operating rules adopted by agencies determine the flow of information into administrative proceedings and, in principle, constrain the decisionmaking power of agencies. A common complaint is that these procedures slow agencies’ decisionmaking and limit their flexibility, particularly in adjusting policies to changing external circumstances. Another complaint is that agencies systematically give insufficient weight to some types of information. Reform proposals include changing the rules of evidence, relaxing the requirement that decisions take account of all evidence submitted but of no other factual material, imposing deadlines on agency decisions, and requiring benefit-cost analysis.

Since the early 1970s, a series of laws and executive orders have imposed more comprehensive reviews of regulatory decisions in and out of agencies. Beginning with the National Environmental Policy Act in 1970, regulatory agencies have faced an ever-expanding array of requirements to perform benefit-cost analysis. Presidents Ford, Carter, and Reagan, in somewhat different ways, all assigned to a special staff in the Executive Office of the President the responsibility of reviewing the economic impact of major new regulations.

3. Changing the Mandate. Another locus of criticisms of the regulatory process centers on the objectives, methods, and powers given to agencies by Congress through legislation and, to a lesser degree, by the President through executive order. If agencies adopt bad policies or overlook key issues, one possible solution lies in clarifying or correcting the mandate (Friendly, 1962). Another solution is to use a method other than regulation to achieve the same objectives, such as taxes (Mills and White, 1978) or tradeable permits (Hahn and Noll, 1982) to control pollution.

Like the agencies themselves, regulatory legislation varies widely in the precision with which its policy directives are stated and the powers granted to the agency. Many agencies have been told that their purpose is to serve the public interest, convenience, and necessity—whatever that may mean—or have been given a wish list of conflicting objectives with no guidance about how to make trade-offs among them. Other regulatory laws are far more explicit. For example, when Congress decided to regulate the fuel economy of automobiles, it enacted legislation that contained explicit standards for the first few years and that gave the regulators the responsibility to devise a way to measure fuel efficiency accurately and to enforce the regulations, as well as to develop further standards for future years not covered by the standards in the law.

4. Altering the External Environment. If the interactions among an agency, a particular industry, other groups with a stake in the industry’s performance, the courts, and Congress produce unsatisfactory results, one approach to reform is to restructure these external institutions so that the agency is better able to produce the desired performance. For example, the single-entity proposal for international telecommunications (Peck, 1970)—that the companies providing international telephone service be merged— was based on the presumption that the FCC could regulate the introduction of satellite technology more effectively if the industry were monopolized, rather than if a single satellite company, Comsat, were competing with several companies that owned only transoceanic cables. At the other extreme, the antitrust case that successfully sought the dissolution of the American Telegraph and Telephone Company (AT&T) was based in part on the belief that the size, wealth, and political power of AT&T made it essentially unregulatable. Public ownership is another alternative arrangement for the managed industry. Other reforms of this type include proposals to create a special type of court to review regulatory decisions, to create a special agency to participate in regulatory decisions as the representative of consumer interests, and to reorganize the congressional oversight process.

All four types of proposals rest on theories concerning the behavior of government agencies, private organizations, and elected political officials. Yet the proponents of reform rarely justify their proposals by reference to explicit theories and empirical observations of the interactions among these groups. Nor is much thought normally given to the organizational problems of structuring an effective agency at the time that the legislation is adopted establishing a new regulatory role for the government.

A useful theory of regulatory policy would provide empirically verifiable propositions on the relationships among structure, responsibility, powers, and performance of a regulatory agency. It would provide subsidiary propositions about how private organizations influence policy design and cope with the particular government regime that is established. It would predict the capacity of an agency to control various aspects of private market behavior and would identify optimal choices of agency organization, procedures, and policy instruments for a given policy objective and a given structure of the private market to be managed. A useful theory would explain why an agency responds favorably to some technological innovations and unfavorably to others, and the extent to which an agency’s response is affected by controllable characteristics such as structure, procedures, powers, methods, and responsibilities.

The remainder of this chapter examines the existing literature relating to the behavior of regulatory agencies in search of generalizations that are useful in the sense described above. While substantial work has been done in recent years on the theory of regulatory agencies, the results are still fragmentary. Consequently, most of the literature surveyed here is not addressed specifically to the question of regulatory policy. Some is very general, addressed to the properties of all government policymaking processes or even all decisionmaking organizations. Some is more specific but is addressed to organizations other than regulatory agencies. Indeed, most research on the public policy process and the behavior of government officials has not dealt explicitly with regulation. Consequently, any attempt to suggest in a comprehensive fashion how work in most of social science might enhance our understanding of the regulatory policymaking process must extract inferences from research that was undertaken for other purposes.

The analysis of the behavior of regulatory agencies can be usefully divided into two parts. The first deals with organic theories of regulatory behavior: theories that characterize a regulatory agency as a coherent whole, having objectives that it rationally pursues. The second focuses on structural theories: explanations of regulatory performance that are based on analyzing agencies as collections of individuals with conflicting objectives whose behavior is coordinated by the selection of operating rules, hierarchies and methods of communication.

Organic Theories

One approach to studying the behavior of organizations is to adopt the metaphor of an organization as a rational actor having ex* plicit objectives and choosing among alternative actions on the basis of their expected contributions to organizational goals. This approach abstracts from the influence the structure of the organization and the experiences of its members may have on organizational outcomes. It also abstracts from the problems of managing the behavior of members of an organization so that their performance will be consistent with organizational objectives.

The appeal of so-called rational-actor models lies in the powerful, empirically testable hypotheses that can be derived from their simplistic motivational assumption. The power of organic theorizing is best demonstrated by the microeconomic theory of the firm, which assumes that private businesses are single-mindedly devoted to the maximization of some index of their financial success, such as profits or sales. From this simple assumption, plus characterizations of technological possibilities, the demand for a firm’s product, and the supply of resources used by the firm, microeconomic theory has produced a variety of testable hypotheses about how firms respond to various changes in market conditions, when they decide to invest in new facilities, and whether the firm, if unregulated, will be efficient.

Goal-directed theories are not universally accepted, especially outside of economics. Before proceeding with a detailed discussion of them, a few remarks about the critics of these theories are in order.

Most criticisms of rational-actor theories are essentially commentaries on the failure to consider: (1) how goals are formulated; and (2) how structural phenomena affect behavior. In principle, both points could-be-important. If one observed considerable differences among organizations engaged in the same general type of activities, more powerful theoretical predictions might be obtained by incorporating goal-formulation and structural features into a model of organizational behavior. In the following discussion, the process of goal formulation and emendation is accorded attention in several of the expositions of various organic theories, while the issue of the relationship between structure and performance is examined at length in the next section. In the end, whether one should abandon the simplicity of the model by incorporating these features rests on the empirical success of the simple model in providing good predictions about organizational behavior.

Another criticism of organic theories is that the concept of organizational goal lacks meaning and therefore should not be an admissible assumption in organization theory. Furthermore, such theories are said to be normatively inadmissible on the grounds that because they incorrectly assume the presence of an organizational goal, researchers necessarily find all organizations inefficient since they inevitably uncover some output of the organization that is inconsistent with the goal presumed by the researchers (Etzioni, 1960).

The alternative approach proposed by such critics is that the objectives of the organization should be regarded as whatever the organization is currently doing and, according to its members, would like to do now and in the future. The normative standards forjudging organizational effectiveness should be: (1) whether it can do whatever it now does or wants to do more effectively or efficiently by reallocating its resources (Etzioni, 1960); and (2) whether it is succeeding in obtaining sufficient resources to do most effectively what it is now doing (Yuchtman and Seashore, 1967).

The alternative formulation offered by Etzioni and his followers has very little merit. It is useful for predicting and explaining organizational actions only if it is converted to a rational-actor theory (Mohr, 1973). Once one has observed organizational outputs and learned from its members what else they would like to do, one has a multidimensional goal for the organization. Moreover, this information is useful only if one presumes regularity in goals through changes in time and conditions. If the Etzioni approach makes an assumption of regularity, then it boils down to an argument that better predictions will be made if the goal of the organization is stated more carefully. If these regularities do not occur, there is little point from an explanatory perspective in defining the original goal, because each observation on organizational performance will be associated with a different, randomly selected objective.

From a normative point of view the two criteria offered by Etzioni and his followers raise legitimate questions, but there is no reason to exclude others. In particular, citizens and government officials outside a specific bureau can legitimately choose to judge an agency’s performance by standards other than those the bureau might use for self-evaluation. From the perspective of an analyst, it is a scientific, not a normative, issue to inquire whether the behavior of an agency is consistent with a particular objective. The Etzioni model explicitly disallows questions about whether the organization is doing the wrong things. From the standpoint of policy formulation and evaluation, then, such a model is essentially useless.

The remainder of this section discusses several different organic, rational-actor theories of regulatory agencies. The key element of a rational-actor theory is specification of the organizational objective.

The number of organizational goals that have been hypothesized is nearly as large as the number of organization theorists (for an interesting survey of the problem of defining organizational goals, see Mohr [1973]). But, with respect to regulation, organic theories are usually based on some form of one of three assumptions about the motives of regulators: they seek to serve the public interest, implicitly maximizing some—usually undefined—measure of national welfare; they try to serve the interests of particular client groups, often by creating a legally enforceable cartel arrangement in an industry that would otherwise be competitive; or they attempt to maximize their own economic rewards or some measure of the success of their bureau.

Serving the Public Interest

A useful point of departure in a catalogue of theories of bureaucratic behavior is to consider models based on the straightforward assumption that government agencies seek to maximize social welfare or the public interest. Numerous conceptual models of government bureaucratic behavior are based on this assumption. These models differ according to their descriptions of how government officials develop perceptions of the public interest they seek to serve. Of course, two related paradoxes in economic theory cast grave doubts on the existence of a well-defined, consistent public interest.

One is the controversy surrounding the compensation principle (Hicks, 1939; Kaldor, 1939; Scitovsky, 1942). The original proposition was that a public policy ought to be adopted if the beneficiaries of the policy could fully compensate the losers and still be better off if the policy were adopted. This principle