o DR /3 4I 1 3 6 1z1 World Bank Discussion Papers Strengthening Public Service Accountability A Conceptual Framework Samuel Paul Recent World Bank Discussion Papers No. 77 Higher Education in Latin America: Issues of Effidency and Equity. Donald R. Winkler No. 78 The Greenhouse Effect: Implicationsfor Economic Development. Erik Arrhenius and Thomas W. Waltz No. 79 Analyzing Taxes on Business Income with the Marginal Effective Tax Rate Model. David Dunn and Anthony Pellechio No. 80 Environmental Management in Development: The Evolution of Paradigms. Michael E. Colby No. 81 Latin America's Banking Systems in the 1980s: A Cross Country Comparison. Felipe Morris, Mark Dorfinan, Jose Pedro Ortiz, and others. No. 82 Why Educational Polines Can Fail: An Ovenview of Selected African Experiences. George Psacharopoulos No. 83 Comparative African Experiences in Implementing Educational Polides. John Craig No. 84 Implementing Educational Policies in Ethiopia. Fassil R. Kiros No. 85 Implementing Educational Policies in Kenya. G. S. Eshiwani No. 86 Implementing Educational Policies in Tanzania. C. J. Galabawa No. 87 Implementing Educational Policies in Lesotho. T. Sohl Thelejani No. 88 Implementing Educational Policies in Swaziland. Cisco Magalula No. 89 Implementing Educational Polides in Uganda. Cooper F. Odaet No. 90 Implementing Educational Polides in Zambia. Paul P. W. Achola No. 91 Implementing Educational Policies in Zimbabwe. 0. E. Maravanyika No. 92 Institutional Reforms in Sector Adjustment Operations: The World Bank's Experience. Samuel Paul No. 93 Assessment of the Private Sector: A Case Study and Its Methodological Implications. Samuel Paul No. 94 Reaching the Poor through Rural Public Employment: A Survey of Theory and Evidence. Martin Ravallion No. 95 Education and Development: Evidencefor New Priorities. Wadi D. Haddad and others No. 96 Household Food Security and the Role of Women. J. Price Gittinger and others No. 97 Problems of Developing Countries in the 1990s. Volume I: General Topics. F. Desmond McCarthy, editor No. 98 Problems of Developing Countries in the 1990s. Volume 11: Country Studies. F. Desmond McCarthy, editor No. 99 Public Sector Management Issues in Structural Adjustment Lending. Barbara Nunberg No. 100 The European Communities' Single Market: The Challenge of 1992for Sub-Saharan Africa. Alfred Tovias No. 101 International Migration and Development in Sub-Saharan Africa. Volume I: Overview. Sharon Stanton Russell, Karen Jacobsen, and William Deane Stanley No. 102 International Migration and Development in Sub-Saharan Africa. Volume 11: Country Analyses. Sharon Stanton Russell, Karen Jacobsen, and Williamn Deane Stanley No. 103 Agricultural Extensionfor Women Farners in Africa. Katrine Saito and C. Jean Weidemann No. 104 Enterprise Reform and Privitization in Socialist Economies. Barbara Lee and John Nellis No. 105 Redefining the Role of Government in Agriculturefor the 1990s. Odin Knudsen, John Nash, and others No. 106 Social Spending in Latin America: The Story of the 1980s. Margaret E. Grosh (Continued on the inside back cover.) 13 6 SI World Bank Discussion Papers Strengthening Public Service Accountability A Conceptual Framework Samuel Paul The World Bank Washington, D.C. Copyright C 1991 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing November 1991 Second printing December 1993 Discussion Papers present results of country analysis or research that are circulated to encourage discussion and comment within the development community. To present these results with the least possible delay, the typescript of this paper has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. Some sources cited in this paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequence of their use. Any maps that accompany the text have been prepared solely for the convenience of readers; the designations and presentation of material in them do not imply the expression of any opinion whatsoever on the part of the World Bank, its affiliates, or its Board or member countries concerning the legal status of any country, territory, city, or area or of the authorities thereof or concerning the delimitation of its boundaries or its national affiliation. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center, Inc., Suite 910, 222 Rosewood Drive, Danvers, Massachusetts 01923, U.S.A. The complete backlist of publications from the World Bank is shown in the annual Index of Publications, which contains an alphabetical title list (with full ordering information) and indexes of subjects, authors, and countries and regions. The latest edition is available free of charge from the Distribution Unit, Office of the Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, The World Bank, 66, avenue d'Iena, 75116 Paris, France. ISSN: 0259-210X Samuel Paul, formerly an adviser on macromanagement in the Country Economics Department of the World Bank, is now a freelance consultant. Library of Congress Cataloging-in-Publication Data Paul, Samuel, 1930- Strengthening public service accountability: a conceptual framework / Samuel Paul. p. cm. - (World Bank discussion papers ; 136) Includes bibliographical references (p. ). ISBN 0-8213-1970-1 1. Administrative agencies-Developing countries-Accounting. 1. Title. II. Series. HJ9940.P38 1991 657'.835'0091724-dc2O 91-38260 CIP FOREWORD Public accountability has assumed special significance in the context of the governance of developing countries. As a leading international development agency, the World Bank is concerned about the effective governance and public accountability in its borrower countries as they are pre-requisites for the success of its investment lending and related assistance programs. In recent years, the Bank's research and policy work has begun to focus increasingly on different aspects of these complex issues. This discussion paper, prepared by Samuel Paul of the Public Sector Management and Private Sector Development Division in the Country Economics-Department, is part of a larger effort within the Bank to analyze and understand the problems of governance and accountability in developing countries. The author departs from the traditional ways of anabling these issues an'd offers a conceptual framework that highlights the role of competition and public participation in promoting public accountability. He shows how the nature of the public service and of the publics involved are pertinent to the choice of the accountability mechanisms to be used. Paul's theoretical framework, propositions and policy implications need to be empirically tested, a task which I hope wil be taken up by the author as weli as other researchers in the Bank and elsewhere. Within the Bank, this paper has already been used by the Taskforce on Govemance in its work and in planning an agenda for research on.governance. It complements the research currently under way in the Country Economics Department on public sector management and public economics. As the subject is of interest to the development community at large, the paper is being published in the Bank's Discussion Paper Series in order that it may stimulate further discusion and action on this important subject. /JohannesF. Linn Director Country Economics Department TABLE OF CONTENTS Page No. ACKNOWLEDGEMENTS ..................................... vi EXECUTIVE SUMMARY .......... .......................... vii L INTRODUCTION .................. ........................ 1 IL ACCOUNTABILITY: A HISTORICAL PERSPECTIVE .....5........ 5 IIL A CONCEPTUAL FRAMEWORK: DEIERMNANTS OF ACCOUNTABILT .................... 10 IV. IMPROVING ACCOUNTABILITY: A MENU OF OPTIONS .... ..... 26 V. THDE PAR.ADOX OF CAPTUIRE .40 VL CONCLUSIONS AND POLICY IMPIICATIONS ..... ............. 44 REFRENCES ............................................. 52 -n - The author is grateful to Michael Barzelay, Ahmed Galal, Arturo Israel, Barbara Lee, Brian Levy, John Nellis, Endque Rueda-Sabater, Mary Shirley, AAndrew Stone, and Ashok Subramanian for their comments on earlier drafts of this paper. - vi- EXECUTIVE SUMMARY This paper develops a conceptual framework to analyze the problem of public service accountability in developing countries and to identify alternative ways to strengthen it. An important proposition derived from this framework is that effective public accountability can be sustained only when government's "hierarchical controlr over public service providers is reinforced by the public's willingness and ability to 'exit" (alternative sources of supply) or to exert pressure on the providers to perform (the use of "voice"). The design of effective accountability systems in developing countries should be of interest both to their governments and to international agencies such as the World Bank that finance the infrastructure for public services (e.g., transport, water, electricity, health and education). The theoretical framework presented in the paper shows that in general, the use of exit and voice by the public will depend on their relative costs and on the expected returns to the public from their use in the context of specific public services. The costs and returns associated with exit and voice will in turn be influenced by the degree of market failure underlying the services in question. The specific factors that influence the use of exit and voice for improving public accountability are severaL The relevant service characteristics on the exit side are the presence of economies of scale (monopoly), legal barriers to entry, and spatial barriers to exit. Voice potential is determined by legal, informational and institutional barriers facing the public, the public's level of income and education (and hence political power), the relative importance of the service to the public, and service differentiability. Some of these are natural barriers and -Viii - characteristics whereas others are artificial or policy induced. Economies of scale, spatial barriers, the relative importance of services, and differentiability are natural factors. Legal, informational and institutional barriers, and income and education characteristics (to a large extent), are policy induced factors. Public services can be categorized in terms of the exit and voice potential they afford the public by reference to these barriers and characteristics. An analysis of the features and barriers of public services and of the publics involved can be used to predict the potential for the use of exit and voice in specific service contexts. This analysis along with the knowledge of the extent to which such features/barriers are natural or policy induced provide a basis for the design of new mechanisms for public accountability and improved service performance. A useful approach to the improvement of accountability is to devise ways and means to overcome these barriers or to minimize the adverse effects of these characteristics. Cearly, natural factors are more difficult to overcome than those induced by policy. Exit and voice potential can therefore be more easily realized by first eliminating the policy induced barriers. Minimizing the adverse effects of natural barriers will call for more time, effort and resources. Judged by these criteria, services characterized by weak voice and low exit will be the most difficult to deal with. Natural factors play a dominant role in causing this combination of exit and voice. The implications of this analysis can be seen in the menu of options for improving accountability for services that fall under the different combinations of exit and voice potential discussed in Section IV of the paper. A menu of options that fits the problems of the different exit-voice combinations is a more useful approach to the improvement of public service accountability than one that offers a standard answer for all situations. The set of propositions summarized below can be a guide to the choice of options: - ix - 1. When a public service operates as a local monopoly due to spatial barriers and the public inolved is characterized by low incomes and legal,informational and institutional barriers, improved accountability is achieved through the use of voice. 2. Under the conditions stated above, the use of voice by the public is likely to be stimulated or assisted by the intervention of agents/organizations outside of the local community (e.g., NGOs). 3. When a public service is characterized by large economies of scale andAegal barriers to entry, and its differentiation is difficult while the public involved or a segment of it is not constrained by low incomes and limited information, voice will tend to be used to improve public accountability. 4. Ihe use of voice under these conditions is likely to be initiated by the public and not through the mediation of extemal agents. 5. When a public service can be differentiated, but is not constrained by economies of scale in its production and the public involved faces income, informational and institutional barriem, improved accountability is achieved through the use of exit The phenomenon of capture-the tendency of the stakeholders who manage and control the alocation of public services to appropriate service benefits for themselves and influential elites-can be an important barrier to the improvement of accountability in developing countries Capture, along with goverment monopoly of many public services, the limited capacity x - of the public to demand and monitor good performance, and the problems in measuriig and quantifying the benefits of services, make the improvement of public service accountability an especially complex and difficult undertaking. Exit and voice actions of the public at the level of specific pubUc services (micro) may well compensate for the reduced effectiveness of heavily overloaded macro level accountability systems (e.g., delays in public expenditure audits, inadequate legislative reviews). Even if public auditors and legislators are forced to spread their attention thinly over an expanding and wide ranging set of public services, exit and voice can be used to exert pressure on public accountability from below. On the other hand, those in charge of macro level accountability could use the feedback from below also to induce the HC systems under their charge to be more responsive and accountable to the public. Governments in developing countries can take a number of steps to enhance public service accountability: (1) I the power of exit and voice mechanisms in promoting public accountability is accepted, it follows that systematic attention to their identification and potential use must be considered at the design stage of programs that are meant to deliver services with public good characteristics. (2) In programs whose outputs/services are meant for reasonably well defined publics (kocationally or through targeting), there is merit in eliciting public feedback on government - xi - services periodically. The purpose here is to generate information as a basis for both the government and the public to learn, respond and adapt so that performance can be improved. (3) In public utilities and other services where the exit option is unavailable, governments need to explore the scope for providing voice mechanisms to enhance accountability. Even where regulatory agencies have been set up to deal with monopoly situations, there is a case for increasing public participation. (4) Voice mechanisms can be built into the device of the memorandum of understanding or contracts to improve the economic performance and accountability of a variety of public enterprises and other agencies. Efficiency measures play a key role in these contracts. But exclusive reliance on efficiency indicators may not cause accountability to improve in all cases. (5) Micro level accountability will be sustained only when the government and the society at large are concerned and wish to improve governance. Comparative information on how different countries perform in respect of key public services can be a powerful tool to generate greater national and international interest in this subject L INTRODUCMION Recent years have witnessed a growing dissatisfaction with the performance of public sector services in many developing countries. Several factors have contributed to this phenomenon. The unprecedented expansion of the public sector to provide services which are regarded as essential to the public (public and quasi-public goods), growing interventions by the state to regulate economic activities for externality and equity reasons, and the overstretching of governmental administrative capacity and resources in the process, are the most important and fiequently cited among these factors. Developed countries have also faced similar problems with public services, though unlike most developing countries, more of them have addressed these problems and in several cases have found interesting and innovative solutions to improve public service accountability (Jabbra and Dwivedi 89; OECD 87; Heald 89; Eaves 90). The economic and financial crisis that gripped the developing world in the 1980s, on the other hand, shifted the focus of governments and donors to the containment of this crisis. While this shift has led to concerted efforts to reduce public sector size, improvement of the efficiency and effectiveness of public services has remained a neglected area. Refocussing attention on this task with special reference to the services that directly affect the public is long overdue.' 1 Government's output can be divided into two categories of goods and services: (1) intermediate goods and services, which though of some interest to the public, are essentially inputs for the production of other goods and services to be delivered to the public; (2) final goods and services which the public or segments of the public receive. Examples of the first are public sector recruitment, investment plans, accounting and budgetary services, etc., that entail inter- agency transactions within governments. Examples of the second are self evident and numerous. They include a wide range of goods and services, economic, social, and regulatory, that public agencies deliver to citizens free or at a price. Ihe primary concern of this paper is with government's final goods and services. -2- Accountability means holding individuals and organizations responsible for performance measured as objectively as possible. Public accountability refers to the spectrum of approaches, mechanisms and practices used by governments to ensure that their activities and outputs meet the intended goals and standards. The concept of accountability applies equally well to all levels of government (central, provincial and local), to public enterprises and to other public agencies that deliver public services to citizens at the micro level, individually or in groups. The accountability mechanisms used at the micro level of public service delivery may differ, however, from those deployed at the macro leveL Elections and legislative reviews, for example, are less appropriate here whereas public hearings, information access, "whistle blowing," and competing sources of supply may be more effective. The objectives of public accountability at both levels are to ensure the congruence between public policy and public actions and services, and to promote the effective and efficient use of public resources. The effectiveness of accountability mechanisms will depend on whether influence of the concerned stakeholders is reflected in the monitoring and incentive systems of service providers.2 When the latter are seen to monitor performance measures consistent with public policy and are motivated to pursue the attainment of public welfare, an effective public accountability system can be expected to exist 2 Public accountability involves three groups of inter-related stakeholders The public and the customers of the service (often a subset of the public at large) are stakeholders interested in service providers being accountable to them for attnbutes that benefit them most. Political leaders and bureaucratic supervisors of service providers are stakeholders who would like the latter to be accountable to them for a mix of public policy and possibly private or parochial goals that interest them. It is the pursuit of private goals by these stakeholders that results in rent seeking activities (Buchanan, et.aL, 1980; Olsen, 1985). Service providers themselves constitute a third category of stakeholders with objectives and interests often different from those of the first two. The public accountability system that finally emerges tends to reflect the relative bargaining power of the different sets of stakeholders involved. Whether public policy goals will be neglected to facilitate rent-seeldng activities, for example, will depend on who dominates the bargaining procs and what their interests are. -3 - A government's performance in respect of public semces can be unsatisfactory for a variety of reasons: (1) Governments often have a monopoly of certain services, leaving littlc incentive for public officials to be efficient Public servants may consider themselves accountable to the public or society at large in an abstract sense, but may not feel accountable to those to whom they provide specific services. Compensation of public servants typically does not depend on their efficiency or responsiveness to the public. (2) In many developing countries, the public/beneficiaries do not have the ability or incentive to demand efficient services or to insist on greater public accountability. It is also possible that some segments of the public (speci interest groups) may capture an undue share of the benefits at the expense of weaker sections (Niskanen 71; Tullock 65; Becker 83). Lack of political power, information gaps and a variety of institutional factors have failitated this tendency. (3) The nature of some public services is such that measurement and quantification of their benefits are not easy.3 Thi further complicates public accountability which has traditionally focused on the internal means of supervision and control (hierarchical oDntrol) in public agencies. Since public servies tend to vary in all these respects, there is no simple or unique answer to the complex issue of public sector accountability. The three sources of poor performance referred to above cearly show that any solution to the accountability problem must take into account the nature of the services involved as well as the charcteristics of the relevant publics. Ihe purpose of this paper is to present a conceptual framework to analyze the problem of public servce accountability and to identify alternative ways to strengthen it. The traditional public accountability mechansm such as expenditure audits and legislative reviews This is not to deny that measurement and quantification annot or should not be improve The term "specificity has been used to decribe this feature and also to link it to the nature of technoko involved. See Arturo srael, lit.iona Johns Hopkins Press, Baldmore 1987. seem unequal to the task of ensuring accountability for public sevices at the micro leveL To be effective, govenmernt-wide accountabiity systems need to be reinforced by new accountability devices for specific services. A promising approach toward this end is explored in the remaining sections of this paper. The paper argues that the public's use of "exit" (competing sources of supply) or "voice" (participation/protest to induce service providers to perform) will enhance public accountability in a given situation when it is consistent with the characteristics of the services and of the publics invohved. An understanding of these characteristics can be used to predict the potential for the use of exit and voice in specific service contexts. It is further argued that public service accountability will be sustained only when the "hierarchical contror (HC) over service providers is reinforced by the public's willingness and ability to exit or to use voice. This is because the only way the behavior of service providers can be made more responsive to the public is through the signals from the HC function (e g., monitoring and incentives) of the agency. When the incentives facing public service providers are wrong, the latter may continue their 'quiet life" despite the exit or voice actions of the public. These propositions challenge the conventional wisdom that competition on the supply side (facilitating exit) or public participation (use of voice) at the micro level alone are adequate to ensure the accountability of public agencies. An important findig is, however, that the phenomenon of "capture," described in a later section, is likely to be a binding constraint on public accountability in many developing countris, and a chalenge to the conventional wisdom about the factors that limit accountability a summarized in the preceding paragraph. As a backdrop to the analysis, the practice of public acuntablity is oxamined from a historical perspecdve in Section IL A conceptual framework -5- for analyzing the determinants of accountability is presented in Section m. Drawing upon this framework, a menu of options for improving accountability is then proposed in Section IV. Section V examines the phenomenon of "capture" as a barrier to change, and its implications for public service accountability. The conclusions and policy implications of the paper are summarized in Section VL H. ACCONTAB : A HISTORICAL PERSPECIVE Governments have, in general, considered public accountability as an essential prerequisite for the efficient production and delivery of public services. Public service performance is, however, a complex phenomenon that depends on a variety of factors. (1) Public policies governing a service tend to influence the financial resources, technology and organizational resources available to the service provider. (2) The characterstics of the public and its environment tend to affect service outcomes. The income and education levels of the public to be served, the size of the public and its heterogeneity, configuration of its values and norms of behavior, and the difficulties in reaching the publc will influence outcomes. (3) Government control over the service provider is a third variable affecting outcomes. The bargaining power of the stakeholders interacting under government auspices and government's internal control systems are key components of this factor. Whfle improving the efficiency of specific public services usually cals for actions on many fronts, the central issue that is common to all such services is the question of how governments and their service providers can be made more accountable to the public for their performance. As implied in the following definition, accountability is the driving force that generates the pressure for the key actors involved to be - 6 - responsible for and to ensure good public service performance. "At its most elementary, public accountability simply requires that public bodies give an account of their activities to other people and provide a justification for what has been done in terms of other people's values, in a way that private bodies do not." (Stanyer and Smith 76) The prevailing concepts and practices of public accountability can be traced to a wide range of developments in economics, political science, public administration and law. Public finance and, in particular, developments in cost-benefit analysis and performance measurement have had a major impact on the information being sought for public accountability at the macro and micro level (Feldstein 85; Premchand 83; Bos 86). Recent developments in public choice theory and the New Institutional Economics (NIE) are also germane to the issue of public accountability. Public choice theory does not consider the state and its agents as neutral bystanders in group interaction. This has given rise to the positive theory of rent seeking which is concerned with the means used by interest groups in getting what they want (Buchanan 80; Bruenan and Buchanan 85; Olsen 65 and 85). Given the bureaucracy's rent seeking tendency, the theory advocates a minimalist state and reduction in the size of the bureaucracy. A branch of public choice theory, the coliective action literature, specifies the conditions under which groups of people tend to get organized to ensure the accountability of public service providers. NIE, on the other hand, focuses on the role of transaction costs, contracts and incentives as the route to greater accountability and performance. The approach here is to enhance accountability by solving the principal-agent problem. Implicit here is the assumption that incentives can be designed to make hierarchical control work. Much of this work is concerned with the behavior of the firm. Whether the approach to the resolution of the agency problem proposed in the NIE is relevant to the public services discussed in this paper is not entirely clear. Those who work in this field do recognize that problems of incentives and governance are enormously difficult in the political or public sector context (Williamson 85). The concept and practice of public accountability have been greatly influenced by political theories and the professions of public administration and law (Jabbra and Dwivedi 89; Schuck 83). They have evolved in the more developed, western countries in parallel with their socio-political evolution. (1) Democratic accountability was the starting point for many countries in the course of the last century. This represents a mix of political and administrative accountability. The government (ministries, the bureaucracy and its constituent parts) in this system is accountable to the political leadership (elected or otherwise) of the country for its actions and performance. Ministers, for example, are accountable to the parliamentJlegislature in democratic countries. Civil servants in turn are accountable to their ministers. Accountability here is seen as a macro level concept as it is difficult for a minister or legislature to supervise or control the individual acts of all civil servants and their departments.4 Furthermore, it does not distinguish between government's intermediate and final goods and services. The instruments of macro level accountability include legislative reviews of ministry activities, periodic audit reports on public expenditure, and the practice of questioning ministers in parliament on ongoing government activities. Political leaders/legislatures act as surrogates for the public in the system of democratic accountability. The effectives of thes macro level accountability devices has been eroded a great deal in recent times with the expanding role of the state and the explosion of 4 The term "macro lvr is being used here only to distinguish it from "micro level" accountability which pertains to specific public services. Democratic accountability is applicable not only at the national level, but also at the local leveL The focus is on the totality of a political uniL - 8 - public services. Public expenditure audits and program reviews get delayed or are inadequately done in many developing countries. The overload on the macro accountability system is indeed a worrisome trend. (2) The expansion of public services which required technical expertise in their production and delivery subsequently led to the practice of professional accountability in the public sector. The experts (doctors, engineers and other specialists) involved in these tasks are guided by their professional norms in being accountable for the services they provide. They exert considerable autonomy in defining the public interest, on the basis of professional norms, and in deciding on the nature and content of the services. The criteria of accountability here are heavily influenced by the norms internally (from the supply side) agreed upon by professionals acting on behalf of the public. This type of accountability at the micro-level has much broader application in the context of specific services than is true of the concept of democratic accountability discussed above. It did not supplant democratic accountability, but acted additively and as a response to the increasing complexity of public services. (3) In recent years, judicial systems and specially devised laws in some countries have been used to augment the accountability of individual civil servants and units within government who are responsible for the production and delivery of public services. The growing interactions of the public with the bureaucracy, and the dysfunctional impacts of secrecy and anonymity within government on the efficiency and effectiveness of services have contributed much to the emergence of the legal accountability concept (Smookler 89). By and large, its practice is confined to the more developed countries with an educated public and a democratic political system. Its evolution is associated with the public's right to seek information from -9- government, the right to sue individual civil servants and public agencies in law courts and the power of courts to make the latter financially liable for violations of the public interest. This pattern of the historical evolution of public accountability reveals three interesting features. (1) The original thrust of government accountability to the public rested with the political leadership at the macro leveL Attention to accountability as a means of controlling the behavior of individual civil servants for public services is a relatively recent development. (2) The focus of the key instruments and measures used to effect public accountability has traditionally been on inputs and not on outputs. In most cases, the latter tend to be diverse and too complex to measure and to aggregate though in recent years there has been some progress on this front. On the other hand, public expenditure which is an input and a common denominator can be easily measured and audited. Internal processes can be assessed as there are generally uniform norms about their use within government (e.g., the sequences and procedures to be followed for decisionmaking and implementation of decisions). (3) Except for legal accountability, a recent development, the primary concern of the concept of accountability has been on internal means of controL Political leaders, agencies and bureaucrats act as proxies for the public and hold those reporting to them accountable through control systems within the relevant organizations. Legal accountability introduces, however, the concept of the public intervening directly to ensure accountability. The dominant pattern has thus been one of vertical or "upward" accountability and not of "outward" accountability.' As noted, the dominant upward accountability system is increasingly getting overloaded. Whether outward accountability with iis focus on the public as "customers to be served" can be made to offwet the adverse impact of this 5 "Upward' refers to the hierarchia approach to control and "outward" to the accountability towards the public, a horizontal relationship. An exception is professional accountability where the behavioral norm departs from hierarchical controL See page 7. - 10- overload is a question that merits special attention. IIL A CONCEPTUAL FRAMEWORK DETERMNNANTS OF ACCOUNTABILITY Accountability for public services has traditionally been viewed from the supply side (the standpoint of the suppliers of services) as is clear from the following quote: "Accountable management means holding individuals and units responsible for performance measured as objectively as possible. Its achievement depends upon identifying or establishing accountable units within government departments-units where output can be measured against costs or other cnteria, and where individuals can be held personally responsible for their performance." (Fulton Committee 68) This is a view of public accountability through Whierarchical contror (HC). There are two supply side factors which have received considerable attention in the literature, namely, the design of services and the incentive structure governing suppliers. The supply side argument is that accountability can be augmented by improving the design of the service and by ensuring adequate incentives for the supplier both to design and to deliver the service. If actions are taken on these fronts, it is assumed that public services will improve and that accountability to the public wfll be enhanced. This approach is valid, however, only when the problem of principal-agent relations within public agencies can be readily solved. When multiple principals with conflicting objectves exdst (political actors, the public, and bureaucratic supervisors in this case), and opportunism and market imperfection prevail(information asymmetries, monopoly conditions), - 11 - the scope for collusion increases, and the agency problem remains unresolved (Tirole 86; Levy 87; Fama and Jensen 83). Under these conditions, it cannot be assumed that an accountability system based on HC will serve the interests of the relevant stakeholders. Alternatively, HC might approximate true public accountability for some time (e.g., when good leaders are in charge or soon after a public revolt), but might not be sustained. According to a recent assessment, "there is ample evidence that it is exceptionally difficult to devise and then sustain control frameworks: the sponsorship function from (sponsoring and finance) ministries to enterprise is a very problematical one, not least in terms of objectives, loyalties and timescales. Outcomes can depend heavily upon the personalities involved and on the wider cultural and political context. Moreover, the difference between developed and developing countries should be fully recognized, especially when discussing the viability of public enterprise reform strategies or options for privatization." (Heald 89) In most developing countries, the casualty in the conflicts between the stakeholders in public accountability is the public's viewpoint (IRAS 89;- Dubashi 90). The directional change required in the reform of accountability systems for public services lies in integrating this element into the HC approach. Determinants of Accountability Viewed from the standpoint of the public, there are two basic factors that influence accountability. One is the extent to which the public has access to alternative suppliers of a given public service. The question here is whether there is potential or scope for the public to exit when dissatisfied with a public service. The second is the degree to which they can influence the final outcome of a service through some form of participation or articulation of protest/feedback irrespective of whether the exit option emsts. In other words, can they exert - 12- their voice in order to enhance accountability? Exit and voice are terms that Albert Hirschman (19;70) made popular in his excellent discussion of the ways in which consumers cope with the problem of performance deterioration in the production of goods and services.' He treats exit as an economic response mechanism and voice as a political response mechanism. The former is ubiquitous in the competitive market place whereas the latter is found more often in organizations such as political parties and voluntary agencies.! The analytical framework presented below builds on Hirschman's concept of exit and voice, but with an exclusive focus on public services. Two interrelated questions will be explored here. Under what conditions are the exit and voice options likely to be efficient in enhancing accountability in public services? How and why do public services differ in their amenability to the use of exit and voice as a means to improve their performance? Answers to these questions are essential for designing better approaches to public service accountability. Diagrams I and II use standard cost-price analysis to provide a simple, but generalized answer to the first question.I They demonstrate that the public's decision to use exit, voice or a combination of both will depend on two factors, namely, the expected returns resulting 6 Exit and voice options can be both substitutes and complements to each other depending on certain underlying conditions which Hirschman develops in his book. He further notes that the cost of voice could be higher than that of exit in view of the need for collective action and the risk of the less certain outcomes of voice. While Hirschman recognizes the relevance of these options to both private and public sectors, he focusses almost exclusively on the world of private goods where both options are readily available. Furthermore, he does not examine the issue of the relevance and design of institutional devices or arrangements for facilitating the use of exit and voice in varying contexts and the conditions under which they will be optimal. 'Ths generalzation must be qualified in part. In politics, forms of exit do exit (leaving a party). While in the market, the use of voice is not unknown (a public demonstration). * Readers unfamiliar with cost-price analysis may consult any standard taetbook in microeconomics for an introduction to the basic concepts. - 13 - coensi DIAGRAM I Returrs Costs of Exit and Voice Voice o ~~~~~Ki Services ranked by market failure co.ta/ DIAGRAM 11 Retrns OptiUsa ltExt-and-Voai (for an Individual) B E X F R_r A~~~~~~ o KI Ml Li Services ranked by market failure - 14 - from improved accountability (e.g., better quality, reduction in delays and corruption, responsiveness or other attributes that the public value, etc.), and the costs associated with the use of exit and voice. A person will resort to exit (e.g., moving to another service provider) only if the cost he incurs in the process is less than the gains from the improved service. The same logic applies to the use of voice. The relative costs of exit and voice and their levels may vary depending on the degree of market failure affecting the services.! In Diagram I, costs and returns are measured vertically while public services are ranked horizontally according to the increasing intensity of market failure.'0 Assume that the public is willing to use exit and voice in order to improve public accountability. AB represents the cost of exit associated with each of the services on the horizonal axis and incurred by an individual (a member of the public). CD is the cost of voice estimated on a similar basis. Although the cost of voice is often incurred in a group, each point on the curve could be treated as the individual's cost of voice. Both AB and CD are upward sloping curves because it costs more to a person to resort to exit or voice as he/she moves to the right towards services that are increasingly affected by market failure (e.g., natural monopolies such as electricity, water supply, etc.). As expected, for the public, voice costs more than exit for the range of services facing competition, but as market failure increases, exit becomes more costly than voice. The cost of exit is clearly more sensitive to market failure than the cost of voice. For example, the cost of creating an alternative (exit) in the face of a natural monopoly will be 9 The level of voice costs will be affected also by the prevailing socio-political environment. For emmple, a free press, dissemination of information, legal rights, etc. will reduce costs for the individuaL ' Note that services here differ only in respect of the degree of market failure. Strictly speaking, only one type of market failure should be considered at a time along the Y axis (e.g., natural monopoly). Needless to say, these are highly restrictive assumptions. - 15 - enormous. The cost of organizing the public to seek changes from the monopoly (voice) will also be substantial. But even the cost of voice tends to go up as the information asymmetries associated with market failure increases along the Y axis. Diagram I shows that the public services within the range of OK will find the use of exit more efficient for improving accountability whereas the services to the right of K will find voice more efficient. When city transport services permit both public and private options, the public will tend to use exit, while faced with urban water supply problems the public is more likely to resort to voice. Diagram II shows that the public's decision to use exit or voice will depend not only on their costs, but also on the expected value or returns from performance improvement derived from improved accountability. It is the net returns from the use of exit or voice that matter to the public. EF represents the expected returns to the individual of such performance improvement for a given level of cost of exit or voice. Its declining slope signifies the increasing difficulty in improving accountability as market failure increases. A monopoly, for example, is expected to be less responsive to public pressure. An upward shift in the expected returns curve can occur, however, when supervision or incentives (hierarchical control) are strengthened to make the service provider more responsive."1 Given the exit and voice cost curves, Diagram II shows that the combinations of exit and voice that are efficient for different goods will vary depending on the level and slope of the expected return curves. Thus when returns are represented by EF, both exit and voice yield net returns along the range of OKI. Nevertheless, exit is clearly the more efficient option as its U Returns are the monetary equivalent of the gains from improved accountability that the individual receives as a result of the use of exit or voice. The upward shift is treated here as an autonomous step, but need not alwas be so. - 16 - net returns exceed that of voice in this range. Within the KIMI range of services, voice is the superior option though up to M, exit continues to yield net returns. For services beyond OLI, neither exit nor voice seem efficient solutions though an upward shift in the returns curve can reduce their number. Public services can thus be partitioned into three categories by reference to the net returns to performance improvement generated by the use of exit and voice. Within the range OKI, exit yields higher net returns and hence is more efficient than voice. Within KiLl voice is more efficient than exit. Beyond LI neither is efficient as net returns are negative. Four propositions follow from this simple diagrammatic presentation. (1) Exit is more efficient, and hence more likely to be used in services least affected by market failure whereas voice is relatively more efficient as market failure increases. (2) For many services, either option or a combination of the two will still leave some net returns and therefore the choice may depend also on the preferences of the public for different combinations. (3) The level of expected returns has a direct bearing on the extent to which voice will be deployed especially in the range of services most affected by market failure. In other words, the public is likely to use voice only when there is a high probability that the public sector will be responsive and make this investment worthwhile. (4) There are some public services for which the use of neither exit nor voice mechanisms will be efficient. If an upward shift in the returns curve occurs, a decline in their number will follow. These are the services most affected by market failure. We shall now move on to the second question and explore the variables that influence the scope for exit and voice in the context of different types of public services. This will take us beyond the aggregative approach of Diagram II which considered services solely as a function of market failure and the public as an undifferentiated lot. Further disaggregation will - 17 - provide a useful basis for designing improved approaches to accountability. Exit Determinants (1) Economies of scale are important in the production of many public and quasi- public goods. Natural monopolies with increasing returns to scale are well known examples. In a country or a geographical area where there is scope for only one enterprise or agency to produce and deliver a service, the public's potential for exit is limited or non-existent, and hence exit cost is high. (2) Legal barriers to entry may exist in a country which limits the public's scope for exit. Registration of vehicles and trade controls can be viewed as legal barriers which the public cannot escape unless illegal action is resorted to. This is because the state prohibits by law other suppliers from delivering a given public service though there is technically no reason why others cannot supply the same service. In many countries, barriers to entry in sectors such as road and air transport and education illustrate this point. (3) Spatial barriers may limit the potential for exit for some segments of the public. Here the problem lies not in the nature of the good or service but in the characteristics of the public. Thus there may be scope for only one small school or health clinic in an isolated village. The constraint is not the existence of scale economies, but rather that certain features of the public (e.g., location) limit their potential exit. However, the effect on the service (as when scale economics exist) is that it operates like a local monopoly. Migration by people facing poor performance, of course, is a form of exit, but a costly one under these circumstances. - 18- (4) Where the nature of a good or service is such that no member of the public can be excluded from access to it,then exit by definition is ruled out. Pure public goods such as defense or environmental protection are classic examples. Voice Determinants (1) Legal and institutional barriers to voice may exist in a country thus making it difficult for segments of the public to use their voice. In some cases, this could be traced to the nature of the larger political system or ideology. Even where an open or democratic political system exists in a country, its laws and legal and institutional devices may not permit or may constrain the use of voice (e.g., legal barriers to the recognition of user groups, lack of public hearings and denial of the right to sue public service suppliers). It is possible that nothing is wrong with the laws, but the procedures and practices used in their implementation stifle or delay the use of voice (e.g., procedures used by courts, access made difficult through the location of facilities). (2) Informational asymmetries can be a severe constraint on the public's use of voice. Service providers often possess information that is not available to the public. Governments may restrict the public's access to information or limit the scope for the media to challenge or publicize the poor quality and other attnbutes of services. Dissemination of information may also be limited by the poor technologies available in the country. Inadequate telecommunications and TV facilities illustrates this problem. Those who have privileged access to the relevant information on services such as elite groups may take advantage of it at the expense of other segments of the public. The net result is that for ordinary people, the transaction costs - 19- of voice will be very high. (3) Non-differentiation of public services can aid the use of voice under certain conditions.' A quasi public good such as drinking water and public parks are non-differentiable products. Since all segments of the public have an interest in their supply and quality, those with a weak voice also gain from the voice of the stronger segments of the public (Wade 88). Non- differentiation of services thus creates an "externality" effect on voice. In the case of education, for example, it is possible to differentiate services (schools with varying quality), and hence the eaternality effect does not obtain. (4) Income, education and related attnbutes of the public increase their ability to use voice. Even when information is available, lack of education may limit its proper analysis and use by the people. The cost of voice can be too burdensome for low income people. Lack of knowledge and skills constrains them in their assessment of options and in demanding better service or access. The poor and illiterate, therefore, are usually the weakest in respect of voice, though their numbers may be large. (5) The relative importance of a service to the public also influences voice. Thus if a person spends a sigoificant proportion of his income or time on a service, or develops a continuing relationship with the service provider ("product involvement"), his incentive to use voice is greater than when the service is of little consequence or is not durable in terms of future This term refers to the concept of "product differentiation" frequently used in studies of indusrial organization and marketing. - 20 - relationships or benefits.'3 This is true even when the individual has an exit option. Examples are housing and health services (doctor-client relations). In health care, there is a clear information asymmetry between doctors and patients. The latter depend on the advice and expertise of the doctors they trust. Exit will, therefore, be a last resort and is likely to be preceded by the use of voice. The foregoing discussion of the factors underlying exit and voice shows that both the nature of the good or service and the characteristics of the public exert an important influence on these options. Natural or artificial (policy induced) monopoly turns out to be an important attnbute of many public services (e.g., electricity, regulatory services) that tends to limit the public's scope for exit. The ability and willingness of the public to respond to this condition through voice can in part be policy induced, but are also a function of certain attnbutes of the public (e.g., income, education, location, etc.). In the final analysis, the factors identified above influence the costs facing the public and thus their choices between the two options. For example, if spatial barriers are high for a person, the cost of talidng the exit route is likely to be substantial for him. Hence he is likely to explore the voice option first before resorting to exit. On the other hand, where spatial barriers are low or when a service is relatively unimportant to an individual, he may find the cost of voice to be higher relative to that of exit. Thus the public may face high or low natural or policy induced barriers with respect to exit or voice or both. Public services can be categorized by the severity of these barriers and characteristics. I Product involvement is a term uswed in the field of marketing, and is a proxy for the relative importance of goods to consumers. - 21 - Combinations of Exit and Voice Public services are categorized below into different groups according to the criteria of exit and voice. First, services can be classified according to whether there is scope for the public to exit if dissatisfied with what they get. This is a proxy measure for the feasibility of competition in the production and delivery of services. Second, public services differ in the degree to which their beneficiaries can make their voices heard if dissatisfied with the outcomes. For analytical purposes, Diagram III depicts four different combinations of exit and voice that provide a useful basis for the classification of public services. These should not, however, be treated as watertight divisions. As the vertical and horizontal arrows imply, they reflect degrees of actual or potential possibilities for the use of exit and voice. Examples of public services that fit the different combinations of exit and voice are given in Diagram IV. Cell 1 above represents public services that fit the low exit-weak voice combination. The nature of the services in this category is such that exit will be extremely difficult or costly for the public. At the same time, the beneficiaries involved are disadvantaged in the sense of being either poor or illiterate or both. Hence their weak voice. They have limited capacity to use voice if dissatisfied with the services. This can be exacerbated by legal, informational and institutional barriers. The public services in this cell can be produced efficiently on a small scale (e.g., to meet the needs of a small town or village). While economies of scale are thus not a barrier, spatial barriers to exit tend to operate. To illustrate, a village needs only a single primary school which can be operated efficiently. If, however, some villagers are dissatisfied with the.school's services, the exit option may not be open to them as the next school may be located in a far away place. Distance and not scale economies is the source of the - 22 - monopoly condition present in this case. Furthermore, the public in this cell do not offer any scope for product differentiation. Health services, agricultural extension, rural water supply and regulatory services affecting the poor (e.g., licensing of vendors) are other examples of services which belong to Cell 1. DIAGRAM III Voice Weak ----…---------------…> Strong Exit Exit - high spatial barriers - large economies of scale Low - local monopoly - high legal barriers to entry Voice Voice - low differentiability - low differentiability of services of services - high income barriers - low income barriers - high legal/institutional - low or moderate barriers information barriers - high information barriers 1 2 Exit 3 4 Exit Exit - legal barriers to entry - low to moderate economies - low economies of scale of scale Voice Voice VI / - high differentiability - high differentiability of services of services High - high income barriers - low income barriers - high legal/institutional - high product involvement barriers - high information barriers - 23 - DIAGRAK IV Voice Low > High - rural primary education - urban water supply Low - rural health - Electricity - law and order (for the - regulatory services poor) - irrigation(large scale) 1 2 Exit 3 4 - urban low income housing - urban transport High - urban primary education - university education (low income) - specialist health care - welfare/nutrition services A combination of low exit and strong voice characterize Cell 2. Here again, exit for those dissatisfied with a service is impossible or expensive, but the scope for voice is considerable. Scale economies matter a great deal in the production and delivery of services in this cell. Natural monopolies owned or regulated by government illustrate the problem. Voice can be strong here for two reasons. Given the nature of the services, product differentiation is not possible or is extremely difficult to design with the result that the entire public gets more or less the same quality and type of service. Even if only a segment of the concerned public is- capable of exerting voice, everyone stands to benefit from such action due to the externality effect explained above. Utilities such as electricity, telecommunications, urban water supply, and irrigation are examples of services in this cell. Regulatory services affecting publics with a strong voice also belong to Cell 2 (e.g., industrial licensing, foreign trade regulation, etc.). The public here is capable of using voice on its own initiative. Organized groups of producers and consumers with strong interest in service outcomes are found in these sectors at least in some countries. - 24 - If service characteristics such as product differentiation are in fact present in a service in Cell 2, the expected positive externality effect on the use of voice is unlikely to emerge. Thus when there is scope for product differentiation in a service, externalities enjoyed by the weaker segment of the public (low voice) will tend to disappear and the use of voice by the stronger segment will not improve accountability for all. Under conditions of extreme inequality, this may be the case in irrigation, for example, where large farmers could use voice to get a disproportionate share of water at the cost of smaller farmers. Similarly, in the case of electricity, the basic product cannot be easily differentiated, but the related elements of repair and maintenance can be differentiated between different segments of the public. Improved accountability here would still require the use of voice, but possibly with some external assistance for the weaker segment. Cell 3 depicts services with a high exit-weak voice combination. As in Cell 1, the services here are not sensitive to scale economies. This means that in a given geographical area like a city, several units can efficiently produce and deliver the services. When performance deteriorates, this will enable the public to exit provided multiple service providers exist. At the upper end of the spectrum in Cell 3, the scope for exit may be reduced through policy induced legal barriers to entry."4 Thus a government may permit education only under public auspices although voluntary agencies or private entities could provide this service. Product differentiation, however, is feasible in these services, thus making it possible for those with strong voice to opt for the quality they prefer. This tendency separates the segment of the public with weak voice from the former. Even if product differentiation is not present, policy induced legal, informational and institutional barriers to voice may also weaken the public's voice. Urban services which lend '1 Note that scope for exit increases as one moves down vertically from Cell 1 to Cell 3. -25 - themselves to differentiation will fail into this cell, e.g., low income housing, and health clinics in poor areas. Between exit and voice, for the weak public in this cell, exit is the more efficient option though voice may play a supplementary role when policy-induced, informational and institutional barriers are eliminated. Finally, Cell 4 refers to some of the quasi public and private services that governments might provide under public auspices for a variety of reasons. Here, services are characterized by low to moderate economies of scale so that potentially several units could produce them in a given geographical area. As a result, the potential for the public to exit is high when faced with performance deterioration. The relative importance of these services to the public is high thus causing their 'product involvement" to be high. This, coupled with the fact that services can be differentiated for the benefit of the public with high income, education, etc., may facilitate the use of voice by the latter. These services may often be the same as those in Cell 3, but differentiated for the higher income groups who normally can exert a strong voice. High quality schools, hospitals, and air transport are examples. Some of the barriers and characteristics discussed above are natural whereas others are policy induced or political in nature. Economies of scale, spatial barriers, the relative importance of services, and infeasibility of product differentiation are natural factors. Some of the policy-induced factors may originate in political discrimination or denial of rights, e.g., low income or education and some legal barriers. In general, political barriers are reflected in policy- induced factors. Legal, information and institutional barriers, and income and education characteristics (to a large extent), are thus policy-induced factors. Public services can be categorized in terms of the exit and voice potential they afford the public by reference to these - 26 - barriers and characteristics. The mix of barriers and characteristics in the four cells of Diagram mI can be used to predict the degree of exit and voice potential that different public services can have. The basis of the prediction is an analysis of the services in terms of these characteristics/barriers and the extent to which they are natural or policy induced. There are, however, some services which are similar in nature between certain cells. This is because when provided to different types of population the same service can be characterized by differing degrees of exit or voice potential. For example, primary education is found in both Cells 1 and 3 of Diagram IV. Nevertheless, from the standpoint of accountability, this seeming overlap does not imply a duplication. Contextual features and attributes of the public need not be the same even for identical services. An urban primary school, unlike a rural school, may leave an exit option for the public. If the relative importance of this service to a segment of the public is high (product involvement), voice may be resorted to before exit. This difference is relevant to the determination of the approach to accountability to be adopted. What distinguishes each cell is the positioning of a set of public services in the context of a unique exit-voice combination. IV. IMPROVING ACCOUNTABILITY: A MENU OF OPTIONS. What options are available to improve public accountability for the services in the four cells discussed above? This question can best be answered by viewing the features of each service against the constraints and opportunities for exit and voice being faced by the relevant publics Options here refer to alternative remedies or solutions to the problem of accountability. - 27 - Given the diversity in the characteristics of services and their publics, a menu approach is clearly superior to an approach that promotes a single or standardized solution. Some implications for improving public service accountability are self evident in Diagram m. First, the accountability problem appears to be most severe in Cell 1 where the potential for both exit and voice are low. Second, the search for improved accountability as one moves down to Cell 3 should focus more on exit (for example, private or competitive delivery) than on voice. To improve accountability in Cell 2, on the other hand, the thrust should be on voice rather than on exit. The greatest need for improving accountability is in these three cells with Cell 1 and 3 deserving priority attention. Third, the search process should first explore the scope for eliminating policy induced barriers and characteristics as a means to improve accountability. Natural barriers or features are more difficult to deal with. Hence the payoff from the latter will be smaller and the time taken to achieve improvements will be longer. A suggestive menu of options for improving accountability through the use of exit and voice mechanisms is offered in Diagram V. The specific choices to be made will depend on the exit-voice potential combination that exists in a given context. To illustrate, the public in Cell 1 whose exit and voice potential is low, faces natural exit barriers whereas most of their voice barriers are policy-induced. Accountability can be achieved here not through exit, but through the use of voice aided by external agents (e.g., NGOs or public evaluation). For the public in Cell 2 who have strong voice, mechanisms such as public hearings, and participation in decision bodies can help improve accountability. This group may well deploy voice mechanisms on their own rather than depend on external support or initiative. Those in Cell 3 who are weaker in voice can explore exit mechanfism such as vouchers and contracting out. - 28 - Diagram V shows that exit/voice mechanisms do not influence accountability in isolation. The proposition is that exit, voice, and HC determine accountability in an interactive mode. This is important because the test of the impact of these mechanisms lies in the behavior of the public agencies involved, though the case for the use of exit and voice to improve accountability is theoretically persuasive. Some have argued that while exit might cause private providers to be more efficient, a competing public agency might revert to the "quiet life" as it faces less demand pressure from the public. The crux of the matter is in the medium through which the signals from exit or voice are transmitted to service providers. The only medium that can directly pass on these signals to them is HC as the incentives that influence the behavior of service providers operate through HC. The tendency of public service providers to resort to the quiet life in the face of exit can therefore be explained in terms of a missing or inadequate HC. DIAGRAMV Mechanisms for Accountability Exi=t Voice vouchers and grants c-> participation in decision making contracting out boards, referenda (incl: mgt contracts, leasing) local governments public competition public hearings/panels self help community organizations deregulation media forums/interventions privatization public surveys/evaluation migration ombudsmen technological options external organizations (NGOs) (long term breakthroughs for exit) legal challenge through courts Aids to Exit and Voice information organizational incentives (HC) monitoring and other supply side systems (HC) - 29 Will the use of exit and voice mechanisms result in the improvement of public accountability? A positive answer to this question will depend on two conditions. First, the use of exit and voice should be consistent with the characteristics of the services and the publics involved. An analysis of the features and barriers of public services and of the publics involved can be used to predict the potential for the use of exit and voice in specific service contexts. This analysis along with the knowledge of the extent to which the featureslbarriers are natural or policy induced provide a basis for the design of new mechanisms for public accountability and improved service performance. The following propositions specify the conditions under which accountability will be improved through the use of exit and voice. * When a public service operates as a local monopoly in small and isolated communities, and the public involved is characterized by low incomes and legal,informational and institutional barriers, improved accountability can be achieved through the use of voice. * When a public service is characterized by large economies of scale and/legal barriers to entry, and its differentiation is difficult while the public involved or a segment of it is not constrained by low incomes and limited information, voice can be used to improve public accountability. * When a public service can be differentiated, but is not constrained by economies of scale in its production and the public involved faces income, informational and institutional barriers, improved accountability can be achieved through the use of exit. - 30 - When a service is characterized by "product involvement" and offers potential for both exit and voice, accountability can be strengthened through a combination of exit and voice mechanisms. Second, exit and voice should cause the behavior of the public service provider to change. Since HC is the instrument for achieving behavioral changes in the service provider, the condition to be met is that its monitoring and incentive systems are adapted in response to exit and voice (see Diagram V). This is important because if HC has not been adapted, the public service provider's behavior is unlikely to be affected and exit and voice would not have had their desired impact on accountability. With the use of exit, for example, some segments of the public may benefit by moving to private providers. The public service provider may still remain as inefficient and ineffective as before if he does not get the right signals through an adapted HC. A measure of the adaptation is the extent to which monitoring by the provider spans efficiency, access and quality instead of being limited to inputs and compliance with internal rules. When such monitoring is reinforced by appropriate incentives to the provider, HC can be said to have adapted to exit and voice. If an adaptive response does not follow, exit and voice mechanisms may be further modified so as to give stronger signals to the provider. When mutual consistency between HC, exit and voice is achieved, the second condition for improving public accountability is satisfied. We shall now examine the relevance of the exit and voice mechanisms of Diagram V for improving the accountability of services that fall under the different cells of Diagram III. Though they are important, less attention is paid here to the role of HC mechanisms (monitoring, inoentives, etc.) mainly because there is an extensive literature on the subject (Heald 89; - 31 - Herzlinger 79; Ramanathan 82). Low Exit - Weak Voice Since both ^it and voice are difficult or too costly for the public in Cell 1, these services pose the most severe challenge to those who design accountability systems . When there are severe constraints on exit and voice, the traditional approach of improving accountability through better HC will have serious limitations, chiefly because the principal-agent problem referred to in an earlier section cannot be satisfactorily resolved in a non-competitive setting with multiple principals who have conflicting goals or preferences. The answer therefore lies in searching for ways to expand the exit and voice options, and integrating them with appropriate HC mechanisms. Given the constraints on exit, however, the focus of accountability in Cell 1 has to be on voice. The exercise of voice is not only costly for the poor in terms of time and effort, but also more risky than exit as the outcomes of voice are less certain given the severity of market failure (see Diagram II). The poor are unlikely, therefore, to readily invest in the use of voice. Given the high costs of voice facing them, the endeavor should be to search for voice surrogates. These are external agents who mobilize or organize the local public in order to demand and monitor better service performance. Non-governmental organizations (NGOs) are well known for playing this role. In some cases, they may also act as service providers. For example, the public tends to trust NGOs in difficult services such as health care for the poor and the aged because the quality of care is not easily monitored (Weisbrod 88). Their voice augmenting role tends to reduce the cost of voice to the local public who are generally poor and illiterate. Under - 32 - monopoly conditions involving the poor, this form of cost internalization by an external agency can tilt the balance in favor of the use of voice by the public. In health services, population control, and education, NGOs have played this role admirably in many developing countries (Paul 87; Esman and Uphoff 84). The alternative of a local community organizing itself to exert voice is also feasibleU5. The cost of voice in this case has to be borne fully by the community, a burden that many poor groups are unable to bear. In both cases, the incentive within the organizations to ensure accountability is aided by a sense of professionalism and commitment rather than by financial rewards. Dissemination of information about services can aid the public in demanding greater accountability from service providers. More systematic pressure can be exerted through a periodic public monitoring and evaluation of the relevant public services. This can be done under public or private auspices. An expert group, for example, can be commissioned to poll the beneficiaries or gather data from them as well as the service providers for analysis and comparison across locations and over time. Compared to the direct use of voice, this is less demanding and less costly for the public. Again, this is a surrogate for voice through external intervention and is designed to promote greater accountability. Its limitation is that such evaluation invariably calls for public resources. Public surveys have been used extensively in Australia, France, Canada and the Netherlands as a usable means of feedback on tax administration (OECD 87). Canada, for example, gathers feedback on taxation from clients on a continuing basis. In this context, transparency of the methods and data used and care in the dissemination and use of findings assume special importance. u. Since the 1960s, many experiments in the use of voice have been attempted in the developing world with a special focus on the poor. - 33 - Complaints procedures and the institutional mechanism of the "ombudsman are ways of providing voice to the people when there is no need or incentive for collective action. Here the attempt is to let individuals make known their problems to designated authorities for redressal. "Hotlines" can be established to help the public respond fast to unsatisfactory services or decisions by public agencies. A review of international evidence in this area shows that these mechanisms are in place in many countries, but that their utilization and impact vary widely. An important finding is that in poor societies, it is the elites who tend to use these devices more effectively. In an African country where the ombudsman did not evoke any response from villagers, a special effort was made to reach out by adopting a mobile approach, but with limited results (Eaves 90). Though spatial barriers are the binding constraint on exit, it is possible that new technologies for service delivery can be used to augment the scope for exit. For example, under certain conditions, mobile courts, schools and hospitals can cross spatial barriers and offer more options to the public in the fields of criminal justice, education,and health.' As Hirschman has noted, there is a risk that easier exit may not necessarily eliminate the performance deterioration in the local public school or clinic The fact remains, however, that technology can be a source of ent expansion under certain conditions and may cause overall accountability to improve. Low Exit - Strong Voice In view of the dominance of scale economies, the scope for improving X A mobile service assumes that road transport is available. For example, even if a 'village has only one clinic or one doctor, another doctor may also visit and offer services to the people through the use of a van equipped with supplies - 34 - accountability through the exit option is limited in this case. Supplementary generation of electricity by agents other than an existing monopoly can be thought of, for example, as a means of increasing competition in the utility field. But the impact of such measures remains relatively insignificant. The improvement of accountability must therefore be sought in devising creative ways to use voice. Since strong voice characterizes this cell, an important option to explore is the participation of the public's representatives in the decision making bodies of service providers. For example, users' representatives can be made members of the boards of directors or committees of these organizations. Where appropriate, users may hold stocks of utilities so that they have a stake in their management and performance. In most developing countries, such use of voice or 'public participation" is conspicuous by its absence. Irrigation projects, for example, are typically owned by governments. Yet, beneficiary farmers could be made joint investors, or responsible for management and maintenance, thereby creating a strong incentive for them to demand and facilitate good performance. In developed countries, public participation in the regulatory process is an established practice. In the U.S., many regulatory agencies have a statutory duty to provide a forum for disputes arising in the course of the agency's rule making, or between those regulated and those affected by the actions of the industry concerned. Since public participation in this process was limited relative to that of the regulated industries, intervenor funding mechanisms have been created in some cases to deal with this imbalance. Thus the California Public Utilities Commission has established a mechanism for after-the-fact funding to groups and individuals who have a financial hardship and who have made unique contributions to a decision adopted by the -35 - Commission. Another example of public involvement is in 'negotiated rule making" that brings together representatives of an agency and the relevant interest groups to negotiate the text of a proposed rule. Regulatory negotiation has been used by the Environmental Protection Agency and other departments of the U.S. Government. Where direct participation of users in decision making is difficult or inappropriate, it may be possible to consult them on important issues or to afford them opportunities to express their views on key decisions directly affecting them. Public hearings on the revision of rates in utilities is a good example of this approach. Advisory panels of a standing nature are another institutional mechanism. These devices can be set up by service providers or through regulatory agencies which are in any case necessary given the monopolistic nature of the activities involved. Since large scale organizations are typical of the servimces in Cell 2, any direct or indirect voice mechanism will act as a useful countervailing force against the potential collusion implied in the multiple principal-agent relations alluded to earlier. Fnally, public evaluation of the type diussed under Cell 1 and provision for legal redress of complaints can (eg., hotlines, ombudsman, etc.) play a positive role here. The dissemination of comparative performance data, including the pubLication of performance indicators on costs, access and quality, etc., can be cpected to semitize users to demand better service and thus to create further pressure for accountabiity. - 36 - High Exit - Weak Voice Since exit is relatively less costly than voice for the set of services in Cell 3, the basic thrust of accountability improvement here should be on expanding the scope for exit. Deregulation of services is an obvious option to consider since there is usually space for both public and private service providers to coexist and compete. For example, deregulatory measures in the education sector may induce the establishment of private schools along side public schools. Monopolies granted to trade agencies, when abolished, will encourage private traders also to enter the field. Since the public involved is poor(weak voice), the public function of regulation and possibly subsidization is likely continue. One option is to make grants to the private providers so as to minimize any adverse economic impact of deregulation on the poor. Grants -in-aid to NGOs have been used in both health and education sectors to achieve this objective. Food stamps or ration cards can be used for the same purpose in the area of essential goods distribution. If for some reason, certain services require close government supervision, contracting them out to multiple private providers can be another option. This approach would permit careful monitoring and quality control while affording increased scope for exit to the public. Municipal services of various kinds (garbage disposal, road maintenance, tax or fee collection, etc.) have been contracted out in many developed countries. An important way in which this mechanism impacts accountability and performance is through its "contestability effect." A more direct impact on accountability can be made through the use of vouchers. - 37 - People who deserve special support for income or other reasons can now receive subsidies for certain services while choosing the service providers that meet their needs most efficiently. The administration of vouchers could be more cumbersome than that of grants to the service providers. On the other hand, vouchers are a superior mode for the exercise of exit in comparison to the grant system. In several European countries, housing vouchers have helped improve living conditions and reduced excessive rent burdens by providing low-income renters an option to exit substandard public housing. Chile successfully replaced housing subsidies through low interest rates with the more direct subsidized method of a voucher program. Under this system, beneficiaries pay for the value of a specified house with their own savings, the voucher and a credit obtained from commercial banks at market interest rates. Initially vouchers did not work well because there was no private supply of very low cost housing and because the commercial banks were not interested in financing low income earners. Eventually, the construction industry responded as increased financing was provided by the State Bank and the Ministry of Housing (Howenstine 86). The private sector has now become so active and specialized in low cost housing that per unit costs have dropped. The voucher system works best under certain conditions: * Individuals' preferences for a service differ significantly, and the differences are legitimate. * People are motivated to shop aggressively for the service. * Individuals are well informed about market conditions, including the cost and quantity of the service and where it can be obtained. - 38 - * Many suppliers of the service are already in competition, or find it a relatively easy field to enter. * The quality of the service is easily determined by the user. * The service is relatively inexpensive and is purchased frequently, so the user learns by experience (Allen 89). Hrgh Exit - Strong Voice -his is the simplest case of all since services in this cell rank high on both exit and voice. To the extent that some of these are private services, the options to follow are fairly self evident. Privatization is clearly the first option to explore. When the potential for exit and voice are both high, market competition can be expected to ensure accountability. Where quasi public goods are invohved, governments may continue to perform a regulatory role (e.g., standard setting, quality control etc.). For example, city transport may be left to competing private transport operators, but under the watchful eye of a regulatory body, advisory councils involving the public, etc. Where the private sector is not adequately developed, there may be a case for the pubbc sector to continue to play a servce provider rol. Instead of outright privatization, public- private competition may then be encouraged. For example, private banks may be permitted to - 39 - compete with public sector banks. Both public and privately owned airlines may coexist. Both privatization and public-private competition can be expected to ensure accountability through the exit option. As Hirschman points out, however, services which entail continuing relations between the public and the providers face a unique problem. The public are unlikely to exit without first attempting to use their voice to improve performance. This need not pose any difficulty in Cell 4 as the public involved are characterized by strong voice. Hierarchical control within public agencies and the exit and voice options discussed above should be mutually reinforcing. For example, the internal incentives and control systems and practices of service providers(HC) should be congruent with what is being attempted on the exit and voice fronts. If not, they will work at cross purposes and weaken public accountability. If service providers do not have the incentive to improve quality as indicated by the feedback from a public evaluation or a panel, accountability and performance will not improve. The commercial nature of some services (utilities, public enterprises with priced services, etc.) will make it easier to use financial and other related incentives to motivate the employees of the service providers. The primary task here is to ensure that such incentive and control systems are consistent with the requirements of accountability to the public. - 40 - V. THE PARADOX OF CAPTURE A common feature of the two right hand cells in Diagram IIl is that the publics involved command 'strong voice." This means that all or some of the beneficiaries of the services belong to the better off and more articulate sections of the population. The options to improve accountability analyzed in Section IV highlight a variety of ways in which the voice mechanism could be used for this purpose. Furthermore, services in the lower right hand cell could be made to perform better through the exit option. One would expect to see in developing countries many examples of the use of voice and exit in improving the services depicted in the different cells. Yet the reality is that such examples are few and far between. If voice is not being used to improve public service accountability in Cells 2 and 4, and exit is not being resorted to in Cells 3 and 4, it is reasonable to conclude that service performance will remain less than optimal. What accounts for this paradox? A closer look at this paradox shows, however, that it can indeed be explained to a large extent by reference to a set of factors that exist in the public sectors of many developing countries. It is reasonable to hypothesize that the interaction among these factors, as detailed below, is the primary explanation of the paradox. (1) The population with a strong voice who have an interest in the services in the right hand cells of Diagram IH includes politicians, bureaucrats and other influential persons who happen to play an active role in the management and control of the same services. They need not be large in number as long as their influence is considerable and their networks strong. They are members of both groups, beneficiaries and service providers, and are mutually - 41 - reinforcing. (2) As a group, they have a strong disincentive to use exit and voice, and are inclined to prevent others from using these options. Exit which implies the existence of competition can disrupt the "quiet life" of service providers who are part of or linked to this interest group. Voice entails a risk for the users in so far as its outcomes are uncertain. In fact, voice may well force the elite group to share the benefits of a service with the rest of the public. It is in the interest of the group, therefore, to seek alternative ways to maximize its share of the benefits. (3) The group referred to above is relatively small and well organized in contrast to the larger public of which they are a part. It is easier for them to engage in effective collective action than for the public at large (Olsen 65). Their strong voice, their joint status as beneficiaries and controllers of services, and their membership of a compact and organized sub- group seem to give them the capability, opportunity and a strong incentive to capture a disproportionate share of the services (for themselves) and possibly also to benefit others closely linked to them. It is almost as if they are well positioned to appropriate these services with much less effort or cost than will be the case if they had to resort to exit or voice. The paradox that is evident in many countries is therefore more apparent than reaL Examples of the paradox of capture abound in developing countries. Governments often give preferential allocations and quotas of scarce services (including goods) to politicians, bueaucrats and business Telephones, housing, permits of various types, etc., illustrate the point. -42 - * In one country, housing development boards(public agencies) have a priorty quota for civil servants that limits the supply available to the public at large. * In another case, all higher level civil servants can get one telephone connection( a highly scarce service) on a priorty basis upon retirement. * An electricity board( state owned enterprise) provides electricity free to all its employees while the public faces power cuts and rationing. Sometimes, informal arrangements and contacts are used to get allocations of scarce items. Thus queue jumping tends to be widespread, but without rousing any public outcry largely because of the lack of transparency of procedures (e.g., admission to quality schools, hospitals or in the allocation of urban land). I£, through such devices, influential segments of the public are able to capture the public services they require, there is no incentive for them to resort to the more costly exit and voice options. Those responsible for the actual service delivery, as shown above, may also collude in this process as exit would be a threat to their quiet life. This then is the explanation for the paradox of scarce and often poorly performing public services coexisting with the reluctance of the influential publics involved to make use of the available exit and voice options to improve performance for alL In many developing countries, the phenomenon of capture is a dominant barrier to the increased use of exit and voice and hence to the improvement of accountability. Capture can now be added to the three other barriers, listed in Section I, namely, governments' monopoly of public services, the public's limited capacity and willinge to demand service, and the difficulties - 43 - in measuring and quantifying the outputs of services. There are three important implications of capture for public service accountability. First, capture can be considered a case of pseudo-exit. Since the elites have a way out of the consequences of the poor performance of public services, there are no internal incentives for the agencies involved to improve their accountability. Genuine principal-agent relations do not exist in this setting. Collusion between the two will be pervasive under these conditions. Furthermore, voice of the public at large is usually unequal to the task of augmenting public service accountability. Second, the longer the paradox of capture persists, the greater the difficulties in dismantling it. Over the long haul, the capture phenomenon may cause performance for the rest of the public to decline so much that it sets in motion strong voice mechanisms (mass protests, political upheavals). This may cause major reforms of the prevailing regime to be undertaken leading to a distinct improvement of accountability. Ihe implication here is that no perceptible improvement of performance needs to be expected unless and until a severe crisis erupts (Olsen 85). Increasing the supply of services is the only other means to weaken the incentives for capture. lhird, continued public ssemnation of information on public service performance and the creation of greater pubLic awarenes of these issues on a continuing basis can be used interim to influence accountability if the crisi scenario does not materialize. In the final analysis, capture can be countered only by the voice of the rest of the public. There is some evdence that this approach can indeed lead to improved public sector peromance and - 44 - accountability. For example, the publication of comparative data on infant mortality is said to have put pressure on some governments (obviously poor performers) to expand and improve their public health programs. One of the states in India was ranked second from the bottom (compared to other states) in terms of nutrition indicators some years ago. The shock of this revelation is said to have triggered a new and successful nutrition program in this state which today has improved its position to the medium range. Comparative studies of public services and the wide dissemination of their findings can, thus, help mobilize public opinion and initiate public action. This approach promises to be a useful surrogate for competition in the public sector. VL CONCLUSIONS and POLICY IMPLICATIONS The basic argument of this paper is that public accountability is an important determinant of public service performance and that its impact on performance can be augmented by moving away from an exclusive reliance on HC mechanisms such as monitoring and use of organizational incentives to a system that uses exit or voice mechanisms in conjunction with HC. Whether the public will resort to exit or voice will depend on the relative costs associated with these options and the expected value to them of the performance improvement resulting from their use in a specific context. The costs and returns associated with exit and voice will in tum be influenced by the degree of market failure underlying the services in question. Exit and voice at the micro level can be expected to impact accountability both directly and indirectly. The contestability effect of some of the exit and voice mechanisms can have a direct effect on the behavior of service providers. Through their impact on HC - 45 - mechanisms, they can also have an indirect and positive effect on the same set of agents by keeping them "in line". The nature and extent of this impact will depend on the relative bargaining power of the multiple stakeholders vis-a-vis the agents involved. The public need not therefore rely solely on macro level accountability mechanisms such as public expenditure audits and legislative reviews of ministries to ensure the accountability of agents. In fact, accountability mechanisms at the macro and micro levels can be mutually reinforcing. Exit and voice operating at the level of specific public services (micro) may well compensate for the reduced effectiveness of macro- level accountability caused by the overload problem alluded to in Section II. Even if public auditors and legislators are forced to spread their attention thinly over an expanding and wide ranging set of public services, exit and voice can be used to exert pressure on public accountability from below. Similarly, working with eaxt and voice mechanisms to improve services, the public is likely to become more sensitive to accountability issues in general and to support and strengthen the institutions and practices that are essential to macro level accountability. Macro level accountability, on the other hand, can reinforce the impact of exit and voice mechanisms at the micro level in three ways. First, systematic and repetitive practices of public audits and legislative reviews of national scope tend to create an enabling environment for micro-level mechanisms such as exit and voice to function. In their absence, the latter may be initiated or may survive for some time, but may not be sustained. Second, even if conventional accountability mechanisms like expenditure audits and reviews do not address issues of performance and outcomes adequately, feedback from below through devices such as exit and voice wfll provide a basis to raise such issues at the national level more often than will be the case - 46 - otherwise. Third, those in charge of macro level accountability could use the feedback from below also to induce the HC systems under their charge to be more responsive and accountable to the public. The factors that influence the potential for exit and voice in the context of different public services are severaL The key characteristics on the exit side are the presence of economies of scale (monopoly), legal barriers to entry, and spatial barriers to exit. Voice potential is determined by legal, informational and institutional barriers, the public's level of income and education (and hence political power), the relative importance of the service to the public, and service differentiability. Some of these are natural barriers and characteristics whereas others are artificial or policy induced. Economies of scale, spatial barriers, the relative importance of services, and differentiability are natural factors. Legal, informational and institutional barriers, and income and education characteristics (to a large extent), are policy induced factors. Public services can be categorized in terms of the exit and voice potential they afford the public by reference to these barriers and characteristics. An analysis of the features and barriers of public services and of the publics involved can be used to predict the potential for the use of exit and voice in specific service contexts. This analysis along with the knowledge of the extent to which such features/barriers are natural or policy induced provide a basis for the design of new mechanisms for public accountability and improved service performance. A useful approach to the improvement of accountability is to devise ways and means to overcome these barriers or to minimize the adverse effects of these characteristics. Cearly, natural factors are more difficult to overcome than those induced by policy. Eit and voce potential can therefore be more easily realized by first eliminating the policy induced - 47 - barriers. Minimizing the adverse effects of natural barriers will call for more time, effort and resources. Judged by these criteria, services characterized by weak voice and low exit will be the most difficult to deal with. Natural factors play a dominant role in causing this combination of exit and voice. The implications of this analysis can be seen in the menu of options for improving accountability for services that fall under the different combinations of exit and voice potential discussed in Section IV. A menu of options that fits the problems of the different exit-voice combinations is a more useful approach to the improvement of public service accountability than one that offers a standard answer for all situations. The set of propositions summarized below can be a guide to the choice of options: 1. When a public service operates as a local monopoly due to spatial barriers and the public involved is characterized by low incomes and legal,informational and institutional barriers, improved accountability can be achieved through the use of voice. 2. Under the conditions stated above, the use of voice by the public is likely to be stimulated or assisted by the intervention of agents/organizations outside of the local community (e.g., NGOs). 3. When a public service is characterized by large economies of scale and/legal barriers to entry, and its differentiation is difficult while the public involved or a segment of it is not constrained by low incomes and limited information, voice is likely to improve public accountability. -48 - 4. The use of voice under these conditions is likely to be initiated by the public and not through the mediation of external agents. 5. When a public service can be differentiated, but is not constrained by economies of scale in its production and the public involved faces income, informational and institutional barriers, improved accountability can be achieved through the use of exit. 6. When a service is characterized by "product involvement" and offers potential for both exit and voice, accountability can be strengthened through a combination of exit and voice mechanisms. The phenomenon of capture is an important barrier to the improvement of accountability in developing countries. Capture, along with government monopoly of many public services, the limited capacity of the public to demand and monitor good performance, and the problems in measuring and quantifying the benefits of services, make the improvement of public service accountability an especially complex and difficult undertaking. Nevertheless, governments can take number of steps to enhance public service accountability in developing countries in the medium term. (1) If the power of exit and voice mechanisms in promoting public accountability is accepted, it follows that systematic attention to their identification and potential use must be consiered at the design stage of programs that are meant to deliver services with public good - 49 - characteristics. During program preparation, the barriers to accountability discussed above need to be identified and assessed, taking into account the nature of the service and of the public involved. While it is appropriate to strengthen the internal control system in a public agency, one should not assume that accountability for performance will follow simply because hierarchical controls are in place. Reinforcing accountability through a search for exit/voice mechanisms is essential especially in poverty oriented projects where diversion of benefits may not be difficult. Often, the mandates and behavior of service providers are dominated by their own preferences or the priorities of their supervisors and influential elite groups. The weakest stakeholder is the public or its segments who do not have an adequate 'voice" for the reasons discussed in this paper. It is important to examine the kinds of exit and voice mechanisms that can be mustered to correct this imbalance. Since the costs and benefits of the different options tend to vary a great deal, there is merit in assessing a variety of experiences in this regard. Innovations in the design of improved accountability systems must be informed by such evidence on the ground. (2) In programs whose outputs/services are meant for reasonably well defined publics (locationally or through targeting), there is merit in eliciting public feedback on government services periodically. The purpose here is to generate information as a basis for both the government and the public to learn, respond and adapt so that performance can be improved. A suitably designed public evaluation should not be a one time affair, but one that is undertaken periodically as a mechanism to enhance transparency and accountability. Task forces or committees of experts should undertake this assessment the findings of which should be publicly disseminated. (3) In public utilities and other services where the exit option is unavailable, -50 - governments need to explore the scope for providing voice mechanisms to enhance accountability. Even where regulatory agencies have been set up to deal with monopoly situations, there is a case for increasing public participation. Public hearings, induction of non-officials/experts on boards/advisory panels, public access to important deliberations/meetings, and surveys of public satisfaction are some of the mechanisms that deserve attention. There is a strong tradition in governments of taking key decisions pertaining to such public utDities/agencies within ministries and in secret. Public dissemination of key performance indicators and the use of local research institutions or media to undertake the comparative analysis and publication of such data and their findings will further reinforce the pressure for accountability. (4) Voice mechanisms can be built into the device of the memorandum of understanding or contracts to improve the economic performance and accountability of a variety of public enterprises and other agencies. Efficiency measures play a key role in these contracts. But exclusive reliance on efficiency indicators may not cause accountability to improve in all cases. There are issues of the effectiveness of the goods and services covered by performance contracts that cannot be easily judged solely by these indicators. Thus if the quality or some other attributes of a service can be assessed only through user feedback, voice mechanisms may be the only valid basis to follow. From the accountability standpoint, a major gap in public agencies is the lack of monitoring systems to track the relevant performance dimensions. Often inputs and technical and financial aspects will be monitored, but dimensions that reflect public responsiveness and satisfaction (access, quality, reliability, etc.) may be neglected. Public surveys of users may then have to be built into the contracts. This may be necessary also to counter the tendency of interest groups in large agencies to " manage" technical data in ways that suit them and as a cross-check with other sources of data on the more difficult to measure attributes. -51. (5) Accountability at the mico lvel will be suswtained only when the gvernment and the society at large are concerned and wish to improve goverance. Comparative information on how different countries perform in respect of key public senices can be a powerful tool to generate greater nadonal and international interest in this subject. Selcted indicators of performance based on good professional judgment can be assmbled on these services from as many countries as possible for wide dissemination. It wfll provide the data for governments, scholars, media and the public at largc to see where thek countries stand in respect of the level and quality of their services. The public's awarene of these matters can often lead to an increased demand for public accountability. - 52 - REFERENCES Allen, J. W., et. al., The Private Sector in State Service Delivery: ExampIes of Innovative Practices The Urban Institute Press, Washington, D. C, 1989. Barzelay, M. and Armajani, Breaking Through Bureaucracy Urban Institute Press, Washington, D. C., forthcoming. Becker, G., "Competition Among Pressure Groups for Political Influence," Quarterly Joumal of Economics. August 1983. Bos, D., Public Enterprise Economics: Theory and Applications North-Holland, Amsterdam, 1986. Bruenan, G. and Buchanan, J., The Reasons for Rules, University of Chicago Press, Chicago, Illinois, 1985. Buchanan, J., Tollison, R. D., and Tullock (eds), Towards a Theory of Rent Seeking Society A & M University Press, Texas, 1980. Colander, D. C. (ed), Neoclassical Political Economy: An Analysis of Rent Seeking and DUP Activity Ballinger, Cambridge, MA, 1984. Eaves, Paige, Accountability in Public Services: A Review of the Literature, mimeo, CECPS, 1990. Esman, M. and Uphoff, N., Local Organizations: Intermediaries for Rural Development. Cornell University Press, Ithaca, N. Y., 1984. Fama, E. F. and Jensen, M. C., "Separation of Ownership and Control," Journal of Law and Economics 1983. Feldstein, M. (ed), Handbook of Public Economics North-Holland, Amsterdam, 1985. Finsterbusch, K and Van Wicklin Im, Waven, "Beneficiary Participation in Development Projects: Empirical Test of Popular Theories," Economic Development and Cultural Change 1989. Galal, A., Institutional Framework for Efficient and Sustainable Restructuring of State-Owned ,niiwd~ .1989. Hardin, R., Collection Action. Resources for the Future, Washington, D. C., 1982. Hartmark, ief, Accountability, Efficiency, and Effectiveness in the State University of New York," SUNY-Albany: ComRarative Dewlopment Studies Cente. 1975. - 53 - Hatry, Harry et_al., How Effective Are Your Community Services?, The Urban Institute, Washington, D.C., 1977. Heald, D., 'Performance Measurements in Public Enterprises," in M. Neuman and K Roskamp (eds.), Public Finance and Performance of Enterprises. Detroit, 1989. Hirschman, A., Exit. Voice and Loyaltv. Harvard University Press, Cambridge, MA, 1970. Howenstine, E., Housing Vouchers: A Comparative International Analvsis, Center for Urban Policy Research, New Jersey, 1986. Jabbra, J. G. and Dwivedi, 0. P. (eds.), Public Service Accountability: A Comparative Perspective Kumarian Press, Hartford, CT, 1989. Kanter, R. M. and Summers, "Doing Well While Doing Good: Dilemmas of Performance in Non- Profit Organizations and the Need for a Multi-Constituency Approach," in Powell, The Non-Profit Sector, Yale University Press, New Haven, 1986. Levy, B., 'A Theory of Public Enterprise Behavior," Journal of Economic Behavior and Organization, 1987. Nellis, John, Contract Plans and Public Enterprise Performance PRE Working Paper, World Bank, 1988. Niskanen, W., Bureaucracy and Representative Government Aldine, 1971. OECD, A Survey of Initiatives for Improving Relationships Between the Citizen and the Administration Public Management Studies Series # 1, Paris, 1987. OECD, Administration as Service. The Public as Client. Paris, 1987. Olson, M., The Logic of Collective Action Harvard University Press, Cambridge, MA, 1965. Olson, M., The Growth and Decline of Nations Yale University Press, Hartford, CT, 1985. Paul, S., Community Participation in Development Projects: The World Bank Experience. Discussion Paper No. 6, World Bank, Washington, D. C., 1987. Premchand, A., Government Budgeting and Expenditure Controls. IMF, 1983. Ramanathan, K. V., Management Control in Non-Profit Organizations, Wiley, New York, 1982. Rose-Ackerman, Susan, "Reforming Public Bureaucracy through Economic Incentives," Journal of Law. Economics, and Organization. 1986. Smookler, H. V., "Accountability of Public Officials in the U. S.," in Jabbra and Dwivedi (eds.), Public Service Accountability: A Comparative Perspective. Kumarian Press, Hartford, CT., 1989. -54 - Stanyer, J. and Smith, B. C, Administering Bdtain Fontana, London, 1976. llole, L., 'Hierarchies and Bureaucracies: On the Role of Collusion in Organizations," Journal of Law, Economics and Organiao 1986 Tuflock, Ihe Politics of Bureaucracy. Public Affairs Press, Washington, D.C., 1965. Wade, R., Village Republics Cambridge University Press, Cambridge, 198& Weisbrod, B. A., Mme Nonprofit Economy Harvard Univesity Press, Cambridge, MA, 1988. Distributors of World Bank Publications ADGliNTINA FINIAN MEXICO SOUTH AFRICA. BOTSWANA Cado Hawh SRL A Khiakupp WFOTEC Fiovlai. illw Galeri Gue_a Y.O. DMx 1i Ap,wdo Fusial 22-860 Aofrd Univaity Paem FRida 165. 4ala Flwr-Oc 453/465 SF0M0I1 14I061OP. MadD.F. SnoutdeAfrica 13331 enowAir B alOdOS 10 P.O. B- 1141 MOROCCO C§peTcwn EKU AUSTRALIA4 PAPUA NEW GUINEA. FRANCE Sudh d'Etudc Mskf Mare n FlUL SOLAOMON ISLANDS, Wndd Bk Publicatons 12 - Mcs.t Bd. d'Aofa For,icr.c0 VANUATU, AND WESTERN SAMOA 646 aenued'dha Coablan Intertiownal Suboalln Swite DA.BDd%e&Jwol1a 7S116Pbl. P.O. Bx 41095 64 WIld" Road NETHERLANDS Cn4ighall Mii 3132 GERMANY InOr-Publikatb.v. Jdannbu%g 2024 Vidcr UNO-Va-la P.O. B-c 14 POppq^du r Alhe SS 724D DA Lodne SPAIN AUSTRIA D-53WDDEnn1 Mundi-P i_an Ub1 SA. Gewcd cnd Co. NEW ZEAILAND Cilo37 Graba,31 GREECE He iUbnay cd Iriformation Smite 21 I Madid A-lOll Wie KME PdvaMeB46 24. IppodamoumrPLAIa Plas NSw Market Ultrfa Inie-al ABDOS EAHRAIN Ath-11635 Audand Coel do Cen 391 Eagnin tRearch od Conodtlby 0M9 Barcloc Aisodal Lad. GUATEMALA NIGERIA P.O. Bc 22103 UlAria Piedrai l Sale lventy Prom Lmited SRI LANKA AND THE MALD WEE M Town317 IS. Cal. 7-SS ThiCrawnsSuDidingJldrcho Lako Hio.oleocokhop Zva I PrivateMa[IDalSC°5 P. O Bc 244 BANGLADIO GuarIatBCity ibadan 100,SirC1utnp.IlnaA. MiasoinduF"Devd_numi C4iGdner Mawatlh Aoihtame Sodey OhDDAS) HONG KONG, MACAO NORWAY Colombo 2 HoaS, Road 16 Ada 2 tad. Narveoenub hnfoaonC ter tlo odi R/Ars 4SWdhamSiaret Bock Dwtnent SWEDEN iha7 1209 Whorms Cnlre P.O. BE 6125 EtaLad Fows.ik Wk. 7nd Plow N06D2 01o 6 Pdite P 6obraloret BSwdoep Control Halg o8Regw w 12. Dc 16356 Matn Road OMAN 5-l03 27 Stddol M4daCCourt iNDIA MEMRB tonoabion Sewvm Nobakla -358D Alied PublbeaPNvaae Lad. P.O. BD 1613, SetbAlopst Farsiwwino 751 Mount load Muscat Wernpnen-WilflamsAB 76, KDA Avenue M O -6ODOO2 Bcx 3D004 ICrola PAKI STAN S-104 25 Sioonchlm B,coaA~6tx HoMin Bock Agey ELGIUM ISN. Herdi Mag 6. Mahra.h idde,Azxm SWITZERLAND jean De Larmny Blad Edae P. O B- No. 72,° F-oir 65l1 Av. du Rad 22 BEcnby - 40D iU4 L1 3 Ubrairie Payot 10D Brunnia 6, orooGato.r 13/14 AafAl oad PERU Caseopotle 381 CANADA New did -110102 EdiSriDal 1 oklo SA CHM1211Genea11 LDdffumar Apitado 3924 CP. 85,515orueAmphre 17 CtdiwanlnAvenue Uma Fmmumawo cerw Doodamilc QaOule Calcuna - 7i0 a2 Ubrrie Pyt J4B SE6 PHILIPPiNES Service dt Akoanaenbt Jlyadoa HowteldBuiding hab_ctioc Dock CuI Cas pohtale 3312 CNINA Skh Man oRd GwAdhingr Flfth lonrRipfet WoDBuiling CH 1102 Lauawno Chtna Plnodal & Ecmxnic BDogeo-560 c09 Ayda Avenuc, Makat PudAbYng Houae MetroMatll TANZANIA 86D. PoSDO,gJ& 3.51129 Xaddguds CocosRload Odwd UnlvdAty hPe BdE4 Hydeabad - SOD0i7 PORAND P.O. B-c _9S W"AN Dur a S.aamn COLiOMNtiA Prubthanaw, 2nd Pow Pola ' ulyiNauld lIfoalsos Lad. Ner ThakwareEau Na qwa Osur01 Wararaws THAILAND Aprtadc oA 34260 Ahnedabsd-3S00DS Centr Dop.natS1uo Dco Dz PORTUGAL 3D6Sicrt Road Poaala l,,,o Lanrad Pwtugal D.gkk COTEDtVOIIIE 16.AAelkMarg RUv Do Carmo 70-74 Cealz d'Edims et de Dfibnao Lu ao-226C1 12I7 Lio7sD TRINIDAD & TOBAGO, ANTIGUA Ahtrjm (CSDAJ IARUD4A EARBADOS, 0DB.P. 541 INDONESIA SAUDI ARABIA, QATAR DOMINICA, GRENADA. GUYANA, AbidjsnolPIaLou Pt Wdire laied jwdrBockSice JAMAICA, MONTSBRRAT, ST. It Sam loulact 37 P.O. Bo 3196 KIIT S INEVIS, ST. LUCIA CYPRUS P.O. Box 181 Rtydhla11471 ST.VINCENTi&GRENADINES MEMRB hoimcaa on Seryio Jak Puast Syana. StudiaLs Unit PD. BDS 26 MWARS h1laadck Savio E9Wa o Niace ITALY Bnmk a Curepe UCeA C c_mimwaa Sanaod SPA Al Aba Stret Trinidad. Wod Indio DENMARIK Via Buaedntlc FaUl 12D/10 Al Dalu Curoe SamfondeLith-tur Caala Pcbo3552 Ph*a Pior UNITE ARAB EMIRATES RmWAOAUAId 50125 RSuace P.O.En71U8 MEMRBGuUICa DKI-I 9PrSduadb,C Rtyadh P.O. Bcx 67 JAPAN Sih DOMINICAN REPUUC BEen Bock Swain. H#AbduRlbhAib BuIdk g EdIoTa1I r,CpwA. 37S4 Hoago 34aes. Bunkyo.ktal13 K4K wtedSrut UNiTED KINGDOM REu raddse sham Is CstdEe 30 Tokyo P.O. Bw 3969 Mioco Ltd Apaubdc No PS 2190 DDmn POD B8Y 3 S_toDanldago KENYA Alt, HampobtreGUt342PG AikeDBock Swvicc(ESA.lLd. 33,Mo_ammedlianAwadStre England EL SALVADOR P.O. acx 45245 P. O B8Y PS Pumdf NaWst Jeddeh VWNEZUELA AdwAds Mammd Earie Ar&M #353D IJveIa del Bale EBIE6oSUA.r Ploi KOREA, REPUBUC OF SINGAPORETAIWAN, Aptdo. 0337 Sm nSlvsdwr Pas ICwos Dok Cm earaton MYANMA,SEiUNEI Caa 1065-A P.O. Bx 101, Kwagwbmu Wdaoaan Publcaio, EGYPT, ARA BRPULIC OF Sl PFewt Lad. YUGOSLAVLA Al Ahrm 0-061 1tE. Pta induahial j)adoudoKsjge Al Gaa Sia KUWAIT DId6F P.O. BD 36 Cmho MEMESB Iwno Suvio 24 Now Induratol Road Trg RepubIe P.O. BcOX 465 Slogaper 963 YUL1001eiVade ThIMiddleBo_timrv 41, ik6 SYt MALAYSIA C-lu Univerity dOfMalaya Cnpttve P.Be St11jauln Pdana b Lumpr I Recent World Bank Discussion Papers (continued) No. 107 Kenya at the Demographic Tuming Point? Hypotheses and a Proposed Research Agenda. Alen C. Kelley and Charles E. Nobbe No. 108 Debt Management Systems. Debt and Intemational Finance Division No. 109 Indian Women: Their Health and Economic Productivity. Meera Chatterjee No. 110 Social Security in Latin America: Issues and Optionsfor the World Bank. William McGreevey No. 111 Household Consequences of High Fertility in Pakistan. Susan Cochrane, Valerie Kozel, and Harold Aldermnan No. 112 Strengthening Protection of Intellectual Property in Developing Countries: A Swrvey of the Literature. Wolfgang Siebeck, editor, with Robert E. Evenson, Wiiam Lesser, and Carlos A. Primo Braga No. 113 World Bank Lendingfor Small and Medium Enterprises. Leila Webster No. 114 Using Knowledgefrom Sodal Science in Development Projects. Michael M. Cemea No. 115 Designing Major Policy Reform: Lessonsfrom the Transport Sector. Ian G. Heggie No. 116 Women's Work, Education, and Family Welfare in Peru. Barbara K. Herz and Shahidur R. Khandker, editors No. 117 Developing Financial Institutionsfor the Poor and Reducing Barriers to Accessfor Women. Sharon L. Holt and Helena Ribe No. 118 Improving the Peformance of Soviet Enterprises. John NelDis No. 119 Public Enterprise Reform: Lessonsfrom the Past and Issuesfor the Future. Ahmed Galal No. 120 The Information Technology Revolution and Economic Development. Nagy K. Hanna No. 121 Promoting Rural Cooperatives in Developing Countries: The Case of Sub-Saharan Africa. Avishay Braverman, J. Luis Guasch, Monika Huppi, and Lorenz Pohlmeier No. 122 Performance Evaluationfor Public Enterprises. Leroy P. Jones No. 123 Urban Housing Reform in China: An Economic Analysis. George S. Tolley No. 124 The New Fiscal Federalism in Brazil. Anwar Shah No. 125 Housing Reform in Socialist Economies. Bertrand Renaud No. 126 Agricultural Technology in Sub-Saharan Africa: A Workshop on Research Issues. Suzanne Gnaegy andJock R. Anderson, editors No. 127 Using Indigenous Knowledge in Agricultural Development. D. Michael Warren No. 128 Research on Irrigation and Drainage Technologies: Fifieen Years of World Bank Experience. Raed Safadi and Herve PlusqueDec No. 129 Rent Control in Developing Countries. Stephen Malpezzi and Gwendolyn BaDl No. 130 Patterns of Direct Foreign Investment in China. Zafar Shah Khan No. 131 A New View of Economic Growth: Four Lectures. Maurice FG. Scott No. 132 Adjusting Educational Policies: Conserving Resources While Raising School Quality. Bruce Fuler and Aklilu Habte, editors No. 133 Letting Girls Leam: Promising Approaches in Primary and Secondary Education. Barbara Herz, K. Subbarao, Masooma Habib, and Laura Raney No. 134 Forest Economics and Policy Analysis: An Overview. WiDiam F. Hyde and David H. Newman, with a contribution by Roger A. Sedjo No. 135 A Strategyfor Fisheries Development. Eduardo Loayza The World Bank Headquarters European Office Tokyo Office ir l -u 1818 H Street, N.W. 66, avenue d'16na Kokusai Building _ Z Washington, D.C. 20433, U.S.A. 75116 Paris, France 1-1 Marmnouchi 3-chome Chiyoda-ku, Tokyo 100, Japan Telephone: (202) 477-1234 Telephone: (1) 40.69.30.00 Facsimile: (202) 477-6391 Facsimile: (1) 40.69.30.66 Telephone: (3) 3214-5001 Telex: wui 64145 WORLDBANK Telex: 640651 Facsimile: (3) 3214-3657 RCA 248423 WORLDBx Telex: 26838 Cable Address: INTBAFRAD WASHINGTONDC 0~ ISBN 0-8213-1970-1