UL2Eon oCm0MMc DeveEwom-ent T-1-n-at'tu; CXDqs z -CTh Wwll odR 9800 Public Enterprise Reform The Lessons of Experience Mary Shirley John Nellis 91DII IMVV3FCo)PDr@IM §UgMB EDI DEVELOPMENT STUDIES Public Enterprise Reform The Lessons of Experience Mary Shirley John Nellis 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 June 1991 Second printing October 1993 The Economic Development Institute (EDI) was established by the World Bank in 1955 to train officials concerned with development planning, policymaking, investment analysis, and project implementation in member developing countries. At present the substance of the EDI' s work emphasizes macroeconomic and sectoral economic policy analysis. Through a variety of courses, seminars, and workshops, mostof which are given overseas in cooperation with local institutions, the EDI seeks to sharpen analytical skills used in policy analysis and to broaden understanding of the experience of individual countries with economic develop- ment. Although the EDI's publications are designed to support its training activities, many are of interest to a much broader audience. EDI materials, including any findings, interpre- tations, and conclusions, are entirely those of the authors 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. Because of the informality of this series and to make the publication available with the least possible delay, the manuscript has not been edited as fully as would be the case with a more formal document, and the World Bank accepts no responsibility for errors. 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, 27 Congress Street, Salem, Massachusetts 01970, U.S.A. The backlist of publications by the World Bank is shown in the annual Index of Publica- tions, which is available from Distribution Unit, Office of the Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, Banque mondiale, 66, avenue d'Iena, 75116 Paris, France. Mary Shirley is chief of the Public Sector Management and Private Sector Development division of the World Bank's Country Economics Department. John Nellis is principal management specialist in the Public Sector Management and Private Sector Development division of the World Bank's Country Economics Department. Library of Congress Cataloging-in-Publication Data Shirley, Mary M. 1945- Public enterprise reform : the lessons of experience / Mary Shirley, John Nellis. p. cm. -(EDI development studies) Includes bibliographical references and index. ISBN 0-8213-1811-X 1. Government business enterprises-Developing countries. 2. Government ownership-Developing countries. I. Nellis, John R. TI. Title. III. Series HD4420.8.S548 1991 338.6'2'091724-dc2O 91-14974 CIP EDI Catalog No. 345/019 CONTENTS Preface v Introduction 1 1. Reforming the Policy Framework 5 The Shift Toward Market Structures 5 Elements of Macroeconomic Reforn 6 Fostering Efficiency through Competition 7 Market-Oriented Financing 8 Pricing for Efficiency 10 Labor Policies to Foster Efficiency 12 Legal Structures 13 2. Determining the Role, Scope, and Objectives of the State-Owned Sector 16 Noncommercial Objectives 16 Diagnostic Overview 18 Setting Objectives for Individual Enterprises 20 Corporate Plans 21 Perfonnance Contracts 22 3. Decentralized Management: A Delicate Balance 25 The Role of Govemment 26 State Enterprises and the Government Budget 28 Improving the Institutional Framework 29 Core Ministries 31 Central State Enterprise Organizations 32 Holding Companies 35 Boards of Directors 37 Raising Managerial Skills 39 Compensation 40 iii iv Contents Training 42 The Approach to Organizational Change 43 4. Holding Management Accountable 45 Gathering Adequate Information 47 Setting Targets and Indicators of Performance 48 Lack of Competition and Administered Prices 49 Adjusting Profits 50 Other Objectives 52 Performance Evaluation 52 5. Divestiture of State-Owned Enterprises 55 An Overview of Experience 55 The Lessons of Experience 57 6. Sequencing Reforms 63 The Near Term 63 The Medium Term 64 Momentum 65 7. Summary 67 Policy 67 Role, Scope, and Objectives 68 Decentmlized Management 69 Managerial Accountability 69 Divestiture 70 Sequencing 71 8. Public Enterprise Reform in Socialist Europe 72 Appendixes A. Checklist of Topics for Diagnostic Overview 77 B. Minimum Financial Information Assembled by Government Oversight Agencies 80 Index 84 PREFACE In this brief book we try to systematize the lessons learned in the past two decades of public enterprise reform-drawing on our own work as well as that of many theoreticians, practitioners, and governments to correct the perceived deficiencies of state-enterprise sectors in Africa, Asia, and Latin America. Our intended audience includes the economists and management analysts interested in the general question of public enterprise performance, but we have particularly aimed the work at practitioners- policymakers, government regulators, enterprise managers, and members of boards of directors of state-owned firms. Our objective is to give the practitioner a grasp of what reforms have been tried, what has worked or appeared promising, and what areas need further thought and effort. We do not review in detail the many reasons why governments in developing countries have relied on state-owned enterprises for the attainment of their socioeconomic goals. Nor do we present a statistical picture of the economic and financial performance of state-owned firms before or after reform. These topics are covered adequately elsewhere.* We have made our study more of a prescriptive than a descriptive work. The authors would like to thank the many World Bank colleagues who made constructive comments on the various papers that form the basis of this book. In particular, they thank Ahmed Galal and Sunita Kikeri for their ideas and assistance. They acknowledge with gratitude the work of Greg Forte, who edited the study, and they thank Arturo Israel for his guidance and support through all the years of work culminating in this volume. * For a discussion of the reasons for which developing countries have used public enterprise, see Shirley 1983, and Choksi 1979. For studies that attempt to measure the performance of state-owned enterprises, see Short 1983, Jones 1982, and Nellis 1986. v vi Preface References Choksi, Armeane M. 1979. State Intervention in the Industrialization of Developing Countries: Selected Issues. World Bank Staff Working Paper 341. Washington, D.C.: World Bank. Jones, Leroy P. 1982. Public Enterprise in Less Developed Countries. New York: Cambridge University Press. Nellis, John. 1986. Public Enterprises in Sub-Saharan Africa. World Bank Discussion Paper 1. Washington, D.C.: World Bank. Ramamurti, Ravi, and Raymond Vernon, eds. Privatization and Control of State-Owned Enterprises. EDI Development Studies. Economic Development Institute. Washington, D.C.: World Bank. Forthcoming. Shirley, Mary M. 1983. Managing State-Owned Enterprises. World Bank Staff Working Paper 577. Washington, D.C.: World Bank. Short, Peter. 1983. Appraising the Role of Public Enterprises: An International Comparison. International Monetary Fund Occasional Paper. Washington, D.C.: IMF. INTRODUCTION The economic justification for public enterprises, in the developing world and elsewhere, has been based largely on the notion of market failure, that under certain conditions the market produces suboptimal economic outcomes such as low production and extremes of wealth and poverty. For example, so-called "natural monopolies" providing electricity, communications, transport, water supply, and port facilities, were thought to offer classic cases of market failure that justified government intervention. In other cases, non-economic arguments about the weakness, foreign dominance, or "incorrect" ethnic composition of the indigenous private sector led many governments, especially in developing countries, to create large state-owned industrial enterprises. Governments hoped that public enterprises would assist the development of "strategic" sectors, gain access to commercial credit that would be denied to small private businesses, fill "entrepreneurial gaps," empower numerically large but economically weak segments of the population, maintain employment levels, and raise the level of savings and investment. State ownership was not thought to offer any inherent obstacles to the efficient functioning of these enterprises. The private sector, it was assumed, would consume wastefully or remit its earnings abroad. But the hopes for the public enterprise sector dropped because production quantity and quality frequently fell below projections, and the sector saddled governments with increasingly heavy fiscal and managerial burdens. Many came to believe that the problem was a lack of market discipline and the inability of state firms to respond to that discipline where it existed. Reform efforts have centered on government/enterprise management systems to produce more independent and entrepreneurial state-owned enterprises, and on the macroeconomic environment to create a more competitive economy; it was thought that such combined efforts could correct the problems of the state-enterprise sector without I 2 Public Enterprise Reform: The Lessons of Experience dismantling it. Indeed, experience has shown that such a two-pronged attack-institutional and macroeconomic-is critical to the future of the many public enterprises that are likely to remain state-owned for the foreseeable future. The approach has led to improved enterprise performance, and much of this study is devoted to discussing the nature and orchestration of these efforts. Nonetheless, experience has also shown nonmarket approaches to be less effective than had earlier been hoped and market failures to be more tractable than alleged. First, although the theory of nonmarket failure-also called bureaucratic or government failure-is not as well developed as that of market failure, the empirical record of disappointingly weak performance in state-owned enterprises is evident and recognized.t Even with a thoroughgoing reform of state enterprise, public ownership will still impose some disadvantages relative to private ownership, among them a greater vulnerability to political interference, lower compensation for managers, and weaker financial discipline. Also state-owned firms have not proven to be much of a cure against the abuses of monopoly power, governments have made great efforts to prevent their enterprises from gouging the consumer only to find all too often that enterprise losses are gouging the taxpayer. Second, the extent and permanency of market failure now appears to have been exaggerated. Technology in areas such as telecommunications is eliminating or reducing natural monopolies and even some markets dominated by a single producer have proven to be contestable, that is, subject to the possibility that competitors could enter the market. This threat has acted almost as effectively as competition in disciplining monopoly power. Overall, the dominant perception today is that the merits of nonmarket alternatives were oversold, while the potential of the private sector was underestimated. The consequence has been a worldwide shift in approach, away from a preference for the public sector to an emphasis on private options. In many developing countries the largest state-owned enterprises are also the largest enterprises in the economy and are likely to remain in state hands for the 1. Elements of a theory of nonmarket failure are offered by Wolf 1988. See also the works of "public choice" theorists in general and those of James Buchanan in particular, which argue that the justifications for government intervention rest on untenable assumptions conceniing the altruism and competence of governments. Introduction 3 foreseeable future; but careful ann's-length sales of other potentially competitive and viable state firms, and the closing of hopelessly loss- making firms can yield a net benefit to the economy, including stimulating production and distribution efficiency and conserving the government's scarce management reserves. Proper conditions for sale include a market- oriented macroeconomic policy, which can stimulate better perfornance in both the public and private sector, an adequate capital market, and the time to identify the firms that are marketable without the attachment of expensive and counterproductive special concessions to the buyer. Therefore, the view advanced in the following chapters is that reform of the state enterprise sector and privatization can be mutually supporting strategies in the larger objective of creating a more efficient and productive economy. Chapter 1 addresses the macroeconomic adjustments typically required in the effort to fashion a more open and competitive economy- liberalizing trade and ending preferential treatment for state enterprises; shifts toward market-oriented financing in the banking system; moves toward market pricing and reforming the institutional structure for setting tariffs; and establishing more competitive and evenhanded compensation and staffing policies in state enterprises. Chapter 2 argues that with macroeconomic reforms under way, government needs to ground its reform of state enterprises in an assessment of the purposes to be served by its state-owned sector. It must first sort out which state firms should be retained, which liquidated, and which divested; begin the reconstitution of the remaining state enterprises as competitive, market-oriented firms; reform its relationship to them in a manner consistent with the nature of such firms; and find new ways of accomplishing many of the noncommercial objectives previously assigned to state enterprises. As described in Chapter 3, reforming the relationship between government and its enterprises entails striking a delicate balance between autonomy and accountability-maintaining the degree of control necessary to establish goals for the enterprise while allowing management the freedom necessary for the achievement of those goals. Chapter 4 examines the mechanisms used by governments in a variety of developing countries to set goals for state-owned firms and to evaluate their performance. The emphasis should be on strengthening local accounting abilities and devising simple performance targets. 4 Public Enterprise Reform: The Lessons of Experience Chapter 5 addresses the promise and the risks of privatization, whether through outright sales, management contracting, leasing, franchising, contracting-out, or encouraging new entrants into the private sector. The promise, as suggested above, is more productive and efficient use of the assets by the new private owners, the overall strengthening of the market environment, and the unburdening of government. The risk is that if undertaken without careful preparation, privatization can cause some firms to fail needlessly or to be transferred with such heavy protection from the market that the potential for enhancing competition and efficiency will be lost. Chapter 6 offers, not a blueprint, but an overview of the task facing governments with a failing public-enterprise sector. At the outset, the elimination of distortions in pricing, financing, and trade can expose the real situation of state firms and help distinguish those that should be closed from those that might be rehabilitated or sold. Gradually, the government can pursue reform of the institutional framework under which the remaining state firms operate, create a more market-oriented financial sector, and continue the effort to make the overall economy more open and competitive. Above all, the government should move pragmatically and without haste. It should build public commitment to the reform effort through education and by balancing losses and long-term changes with actions that produce quick payoffs. By distinguishing between near-term and longer- range goals, governments can avoid the snares of perfectionism and comprehensiveness and maintain the momentum necessary to the massive task of restructuring. Chapter 7 summarizes the lessons of the study, and Chapter 8 offers a note on the reform of public enterprises in socialist Europe. References Wolf, Charles Jr. 1988. Markets or Governments: Choosing Between Imperfect Alternatives. Cambridge, Mass.: MIT Press. 1 REFORMING THE POLICY FRAMEWORK Although governments in most of the developing world during the postwar period looked to the public sector as the engine of economic progress, practitioners and theorists have increasingly agreed that reliance on the public sector has stretched the managerial abilities of developing countries governments beyond their limits. Governments almost everywhere have become increasingly aware that they must realign their priorities to mobilize the skills and resources of the private sector in the larger task of development, and to concentrate government efforts on essential public services-pruning activities that have become unmanageable, and using all resources more efficiently. The Shift Toward Market Structures To optimally allocate the resources of a country, governments with significant state-owned enterprise sectors must increase market forces, decentralize economic decision making, strengthen managerial capabilities and incentives throughout the economy, and create a new division of labor between the public and private sectors to raise the efficiency of each and raise the contribution of each to development. Inappropriate and weak macroeconomic policies and legal structures create a poor environment for pursuing competitive efficiency, particularly in developing countries. The local market may be small and insulated from import competition, with barriers to entry and a weak capital market. A move to sell a large public firm to the private sector in such circumstances may require granting special privileges to the buyers that protect them from competition; the buyers in tum may grant assurances, such as a promise of no major layoffs or plant closures, that hobble them in competing abroad. Also, governments may be less capable of regulating monopolies in the private sector than controlling them in the public sector. Refashioning the policy environment to build a more open and competitive overall economy lets governments lighten their own 5 6 Public Enterprise Reform: The Lessons of Experience administrative burden and simultaneously improve overall efficiency by applying market tests of viability and performance to its firms; it can allow or force nonperforming state enterprises to close, while carefully transferring many viable ones to the private sector, and grant more autonomy and entrepreneurial flexibility to those that remain public. In practice, state-owned enterprises seldom take advantage of a competitive, profit-maximizing market environment: they are encumbered with noncommercial objectives; they operate in noncompetitive markets; their management is more bureaucratic than entrepreneurial, impeded by government intervention in details of management, lacking incentives to improve performance, and without accountability for results; and state firms have seldom been allowed to go bankrupt. Privatization coupled with policy reform may offer the optimal solution. The criterion for deciding whether an activity or firm is best located in the public or private sector is to judge where, given current government and social resources, it will maximize net benefit to the economy. Elements of Macroeconomic Reform The macroeconomic policies goveming trade, finance, pricing, and labor are the most relevant to state-owned firms. Governments may not be ready to change such policies overnight, but at a minimum they need to recognize the fundamental reforms required and adopt a timetable for change. Efforts to introduce institutional changes without an adequate framework for macroeconomic policy reform run the risk of being self- defeating, undermining the very improvements they are meant to achieve or, worse, actually hardening an undesirable policy stance. For example, a key barrier to the success of performance contracts in some West African countries has been the failure to establish first the basic policies of pricing and finance that would set the framework for the contracts. The fundamentals of economic policy reform have been well documented in other works.1 The following summarizes aspects of these policies that are most pertinent to the reform of the state-owned sector in developing countries. 1. For those published for the World Bank, see, for example, Dornbusch and Helmers 1986; Munasinghe and Warford 1982; Tolley, Thomas, and Wong 1982; Cody, Hughes, and Wall 1980; and various issues of the annual World Development Report. See also Nellis in Thomas and others 1991. Reforming the Policy Framework 7 Fostering Efficiency through Competition Any enterprise, public or private, operates more efficiently when faced with a competitive market. The spur of competition pushes firms to develop their marketing skills, pay close attention to service, keep up technologically, and control costs. But state enterprises must be able to respond appropriately if they are to become more efficient through exposure to competitive pressures. If, instead, the enterprise managers are bureaucrats rather than businessmen and have never developed marketing or financial management skills, if they are not permitted to cut costs through layoffs and closures, and if the capital base of their firm has been eroded by govemment pricing policies and the firm cannot collect on its accounts, then the condition of the firm will deteriorate under competition. The losses will worsen, thereby increasing the budget hemorrhage and potentially creating a backlash against further reforms. In some cases, privatizing the firm may be a rapid way to improve its capacity to compete because it addresses many points of competitive vulnerability simultaneously. In other cases, a government can enhance a state-owned firm's ability to respond to competition by improving its institutional environment and its management (see Chapters 3 and 4). In any case, reform programs should attempt to orchestrate both the stimulus of heightened competition and the response: improvements in the state firms or their divestiture. Competition can be introduced to the domestic market through two channels: international trade (by lowering import barriers and by encouraging exports) and the restructuring of the domestic market. The exposure to imports through trade liberalization and the encouragement to export through such outward-oriented policies as reasonable exchange rates, duty drawback schemes, and easier export licensing, can strongly encourage overall efficiency. For example, a World Bank study of India's public enterprise Hindustan Machine Tools, found that even a small exposure to competition in export markets (8 percent of company sales) markedly increased the company's dynamism and professionalism. State-owned firms in developing countries tend to be large in relation to the local market, and their size and dominant position dampen competition. Nonetheless, the scope for domestic competition can be expanded even in relatively small markets by eliminating state monopoly or monopsony 8 Public Enterprise Reform: The Lessons of Experience power and by generally putting state and private-sector firms on a more equal footing. In response to a loss of market power, the state firm sometimes loses ground; in Somalia, for example, after the Agricultural Development Corporation lost its exclusive right to purchase and market maize and sorghum, its share in purchases of these crops dropped to 1.6 percent. In other cases, state companies respond favorably to competition; when private traders were allowed to compete with the public food importer in Niger, not only did prices fall and foodstuffs become more widely available, but the state firm improved its performance and made a profit for the first time. Surprisingly few governments attack state-enterprise problems by putting public and private competitors on equal terms. Although state- owned enterprises operate under a number of handicaps, notably those of limited financial and operational autonomy, they typically also enjoy a number of advantages over their domestic competitors: explicit or implicit government guarantee of debt, access to government subsidies or low- interest loans, and tolerance of arrears on taxes or other payments. Government firms also enjoy tax exemptions and favored positions in public procurement decisions although private firns can often more easily evade taxes or labor and other laws. Moreover, in some countries a government enterprise cannot be declared bankrupt or subjected to forced liquidation by creditors. As a result, helping state-owned firms become responsive to competition will often entail helping them overcome the effects of years of price controls, inadequate equity, and overindebtedness; for the same reasons, governments may also need to reform laws and regulations that discriminate in favor of state firms. Market-Oriented Financing Reforms in financial policy-increasing competition in the banking sector, moving toward market-determined rates of interest, enhancing the independence of banks in determining credit policy-all create special opportunities and problems for state-owned firms. Typically, government enterprises have been created with little or no equity, and then, with the shortfall in revenues caused by the low prices imposed by governments, they have been forced to borrow to cover operating expenses or to finance all of their investments. State enterprises have encountered few problems in obtaining these loans. They have usually enjoyed privileged access to Reforming the Policy Framework 9 credit and government guarantees, unlimited overdraft facilities, low controlled interest rates on domestic funds, and low-cost foreign funds by virtue of an overvalued exchange rate. Furthermore, public-sector firms have often ignored the debts they owe each other, thereby creating interlocking arrears within the sector and between the firms. This easy access to cheap capital and the government's readiness to take on the debt of a state enterprise or to fund its deficits add up to what Hungarian economist Janos Komai calls the soft budget constraint, which allows managers of state enterprises to feel little pressure to control costs or conserve funds (Komai 1983). The overall consequences are misallocation of resources, selection of excessively capital-intensive projects, growing indebtedness of state-owned enterprises, crowding out of private borrowers in local capital markets, and a general tendency to expand the scope of public operations. These circumstances militate against a sudden switch to market-oriented financial policies: charging state-owned enterprises the correct price for their capital would, at first, make the financial situation worse. Reform operations must first lay the groundwork by finding ways to reduce the extraordinary debt burden of the state sector-by canceling offsetting debts, restructuring the capital base of some enterprises by converting government debt to equity, injecting new capital to reduce the debt overhang, canceling arrears to the banking system and suppliers, and renegotiating the terms of the debt. Debt-equity swaps can be another useful way to address the problem, both for foreign debt and for domestic arrears; in Portugal, for example, municipal power-distributing companies transferred assets to pay their debts to the state-owned power-generating company. The state should assume the debt of a state enterprise or supply it with new capital only after it has first set strict criteria that restrict such financial aid to firms that are economically viable, potentially profitable, and subject to financial discipline. As one report on the topic has noted: Equity in publicly-quoted companies carries its own discipline: the share price will fall if dividend performance is poor. Equity may quite properly be used to maintain a degree of flexibility in the basic capital structure of an enterprise which is expected to be profitable in the long run ... To use it for providing capital to an enterprise which is not expected to be sufficiently profitable to remunerate its equity is 10 Public Enterprise Reform: The Lessons of Experience to perpetrate an illusion. Such equity is essentially the capitalization of a recurrent subsidy, but suffers one major disadvantage compared with such a subsidy in that it is not subject to periodic reappraisal. Once provided, it is there forever, whereas an overt current grant is routinely the subject of annual reappraisal and decision. Thus, in theory, an assessment of the economic benefits and potential profitability of a state-owned firm should be a prerequisite for financial restructuring, which would then be supported by the enforcement of a reasonable dividend policy. In practice, however, few countries have followed this procedure; Chile, which requires a return of 5 percent to 10 percent on assets as a dividend from its enterprises, is one of the exceptions. At the very least, financial restructuring of a poorly managed finn should be delayed until new management has begun an agreed-upon program to increase operating income. To eliminate the insidious effects of the soft budget constraint, governments should stop subsidizing their firms whenever possible, ideally requiring them to borrow principally from private commercial banks and to pay the opportunity cost of capital. If a country's banks are competitive and free to manage their own portfolios and if they subject state enterprises and private firms to the same financing criteria, they can take on much of the burden of monitoring and supervising competitive state enterprises. Developing an arn's-length relationship between the government and the financial institutions-particularly state-owned financial institutions-is thus a critical part of restructuring. This requirement complicates the reform process because in most developing countries the banks lack this autonomy, and in many they are on the verge of bankruptcy. Nevertheless, these reforms are unavoidable because in many cases the banks have become the main prop for inefficient public (and private) firms. Pricing for Efficiency A critical element of any reform in the state-enterprise sector is the move to market pricing where competition is possible and, where it is not, the development of criteria for monopoly tariffs. A wealth of theoretical work has aimed to devise rules for setting monopoly tariffs at the equivalent of efficiency prices (prices equal to the economic cost of the last unit sold plus a markup to clear the market). Such rules are seldom applied Reforming the Policy Framework 11 in practice, however; a study of tariffs in West Africa suggests that improving revenue collection and a simpler approach to pricing is preferable to lengthy and costly long-run marginal cost studies, particularly in the data-scarce countries in that region. Such studies have not been effectively translated into better revenues precisely because of inefficiencies in revenue collection and production. (Julius and Becker 1986; Julius and Alicbusan 1986). Price increases are often needed to end further decapitalization of state- owned firms and to avoid the budget drain, waste, and distortion associated with artificially low prices. Still, governments should avoid automatic cost-plus pricing and instead seek to provide incentives for managers to look for cost reductions as diligently as they look for price increases. In Turkey, for example, the government monopoly enterprises improved their financial situation and reduced the budget burden by raising prices, but observers question whether the firms became any more efficient. As one observer notes, If governments allow public enterprises to operate as profit centers, they had better make sure that the prices managers face in designing and implementing their business strategies come pretty close to reflecting true opportunity costs to society... If public enterprises do not operate in competitive markets or if government-administered prices faced by public enterprises do not approximate real social opportunity costs, then in most cases it would be preferable to treat them as cost centers. (Mallon 1982, p. 26). The government and the enterprise can, for example, set unit cost targets in reference to benchmark indicators ffom other countries and to the firm's past performance. Once a price benchmark such as the border price is set, the firm should be allowed to retain part of any surplus it can generate, at least for a few years, to give managers an incentive to reduce costs. Governments could also well put greater emphasis on reforming the institutional structure for setting tariffs. Ideally, the authority to approve tariffs should be vested in an entity that is free of arbitrary political influences, competent to review tariff decisions and provide an external check on the efficiency of the enterprise, and required to respond quickly and fairly to requests for price increases. Recently several West Africa 12 Public Enterprise Reform: The Lessons of Experience countries have begun to specify pricing formulas and efficiency targets through a performance contract system modeled on the French contract plan. The experience has not been an unqualified success, but it seems to be a move in the right direction (see Chapter 2). Labor Policies to Foster Efficiency Many developing countries enact laws governing the personnel policies of both public and private enterprises but typically enforce these laws more strictly in the state firms, putting them at a competitive disadvantage. In other cases, state-owned enterprises come under the generally rigid and bureaucratic rules of the civil service or similar system. Bureaucratic rules together with the political visibility of public firms and their easy access to government subsidies and credits contribute to the labor problems typical in many state enterprises: redundant workers, strong upward pressures on wages at lower levels, and wage compression between the top and lowest grades. State firms generate wage compression typically by raising pay for the bottom grades and freezing salaries at the top. For example, the ratio of management pay to the income of the lowest-paid unskilled worker in state enterprises in Ghana shrunk from 4.0 in 1982 to only 2.2 in 1985, excluding fringe benefits. A further result of compression is that professionals and executives in state firms usually earn less, often significantly less, than their counterparts in the private sector, whereas unskilled and semiskilled workers in public firms eam about as much as, if not greatly more than, those in the private sector-while enjoying a high degree of protection from layoffs. A few examples will illustrate the extent of such pay differentials: unskilled Nigerian workers in state-owned enterprises in 1982 were paid 95 percent of the wage of their private counterparts, while the parallel ratio for senior management was 49 percent; laborers in Zambian state enterprises in 1983 received 144 percent of the wage paid comparable workers in the private sector, while professionals received only 91 percent; a manual worker with less than a tenth-grade education earned 33 percent more in a state enterprise in Thailand in 1983 than a similar worker in a private firm, while an executive earned 6 percent less. State enterprises are typically more capital-intensive than private enterprises and are not a major source of employment in most countries, Reforming the Policy Framework 13 even in the modem sector. Nonetheless, they tend to dominate the nonagricultural labor market and to exert a strong influence on the the labor policies of private firms.2 State enterprises attract lower-ranked workers, raise the wage expectations of such workers in both public and private firms, threaten external competitiveness, and slow job creation. In addition, the relatively low pay of public-sector professionals and executives makes their ranks hard to fill and reduces their morale, discipline, and productivity. Managers who eam significantly less than their private-sector counterparts will be understandably cool toward efforts to increase their responsibility and accountability even though the price of higher executive salaries in the private sector is usually less job security. Introducing fair but more disciplined labor policies in state firms, matching the pay of all employee groups more closely to the private sector and to the need for competitiveness, can benefit both the public and private sectors by improving efficiency and increasing the opportunities for productive employment. Although naturally loathe to face the massive short-term costs-social, financial, and political-of restructuring pay and firing workers, more and more developing countries have begun giving the managers of their state- owned enterprises greater freedom to set levels of pay and employment. As a result, for example, Ghana's Cocoa Marketing Board laid off 29,000 of its workers during the late 1980s. Such restructuring can nonetheless be expensive if the firm has previously agreed to generous severance packages. Legal Structures Experience shows that as a govemment moves from macroeconomic policy to the details of restructuring its state-owned enterprise sector, it often tends to overlook the need for legal changes or to underestimate the time such changes require. However, a program to move the whole economy toward greater efficiency may both hinge and founder on highly visible changes in the legal environment to support more economic competition and accountability in both the public and private sectors. 2. The share of nonagricultural workers employed by nonfinancial state-owned enterprises is about 6 percent in Latin America, 16 percent in Asia, and 19 percent in Africa (Heller and Tait 1984). 14 Public Enterprise Reform: The Lessons of Experience The important and complex topic of legal reform is beyond the scope of this study. For present purposes, it is enough to emphasize that the laws goveming the firms that will inevitably remain in government hands should support competitiveness and managerial accountability. To achieve this, a country may, like Turkey, concentrate on legislating the reform of its public enterprises or, like Peru, apply the established private-sector commercial code to state-owned enterprises to put them under a more efficient set of rules and to emphasize that they are to behave as if they were private. To the extent that the legal heritage of the country permits putting as many state firms as possible under the same commercial code that govems private firms-and assuring that the commercial code itself supports market structures as well as possible-will generally promote competition and nondiscrimination more effectively than special legal status for state firms. References Cody, John, Helen Hughes, and David Wall, eds. 1980. Policies for Industrial Progress in Developing Countries. New York: Oxford University Press. Dornbusch, Rudiger, and F. Leslie C. H. Helmers. 1986. The Open Economy. New York: Oxford University Press. Heller, Peter, and Alan Tait. 1984. "Government Employment and Pay: Some Intemational Comparisons." Washington, D.C.: IMF. Julius, De Anne, and Adelaida Alicbusan. April 1986. "Public Sector Pricing Policies: A Review of Bank Policy and Practice." Washington, D.C.: World Bank. Julius, De Anne, and Sarah Becker. February 1986. "Public Enterprise Price Formation in West Africa." Washington, D.C.: World Bank. Komai, Janos. 1983. "Comments on Papers prepared in the World Bank about Socialist Countries." Country Policy Department Paper 1985-10. Washington, D.C.: World Bank, 1983:13. Mallon, Richard. February 1982. "Public Enterprise Pricing Policy." Discussion Paper 129. Boston, Massachusetts: Harvard Institute of International Development. Reforming the Policy Framework 15 Munasinghe, Mohan, and Jeremy Warford. 1982. Electricity Pricing. Baltimore: Johns Hopkins University Press. Nellis, John. 1991. "Reform of Public Enterprises." Chapter 6 in Restructuring Economies in Distress: Policy Reform and the World Bank, edited by Vinod Thomas and others. New York: Oxford University Press. Tolley, George, Vinod Thomas, and Chung Ming Wong. 1982. Agricultural Price Policies and the Developing Countries. Baltimore: Johns Hopkins University Press. World Development Report. Various issues. Washington, D.C.: World Bank. 2 DETERMINING THE ROLE, SCOPE, AND OBJECTIVES OF THE STATE-OWNED SECTOR The impetus for advancing from reform of national economic policies to reform of the state-owned sector typically arises when the new policies- fostering a market orientation in trade, financing, pricing, and labor throughout the economy-have begun to change the rules of the game, leaving many state enterprises in serious trouble. Handling these firms in the new environment requires government decisions about the role, scope, and objectives of the state enterprise sector, decisions foreshadowed in the very macroeconomic policy initiatives that have now put the spotlight on these firms. Institutional and political factors tend to inhibit governments from stating clearly what they expected from the state-enterprise sector as a whole; but the performance of state enterprises has often deteriorated to the point where many governments become willing to take this difficult and contentious step. For the government, this clarification entails, in rough order, (a) finding a new approach to accomplishing many of the noncommercial social and political roles historically assigned to the state- owned enterprises; (b) making an initial diagnosis of the firms, given their altered mandate, to sort out which should be assisted, which should be liquidated, and which should be divested; and (c) beginning the reform of its relationship with the remaining state-owned enterprises, finding ways to set objectives with them in a manner more consistent with the operation of competitive, market-oriented firms. Noncommercial Objectives Some of the reasons governments cite for owning enterprises- promoting "strategic" sectors, generating income for the treasury, offsetting economic dominance by foreign or certain national interests, and regulating monopoly power-are compatible in principle with the efficient 16 Determining the Role, Scope, and Objectives of the State-Owned Sector 17 operation of a commercial entity (Choksi 1979; Nellis 1986). But other reasons, called here noncommercial objectives, create problems that need to be taken into account. Noncommercial objectives include the use of public enterprises to promote regional development, job creation, and income redistribution; they often involve taking on or maintaining redundant workers, pricing goods and services below market (sometimes even below costs), locating plants in uneconomic areas, or keeping uneconomic facilities open. The pursuit of a noncommercial objective through state-owned enterprises can have unforeseen perverse effects: * When state enterprises are required to keep prices artificially low, not only do budget deficits and debt grow and shortages worsen, but public and private investment decisions are distorted and the underpriced resources wasted. * When state enterprises must hire excess workers to promote employment, labor productivity falls and unskilled laborers may end up earning more than in the private sector, at taxpayers' expense. * When state enterprises receive subsidies for noncommercial objectives, management tends to become lax, and the resulting inefficiencies create shortages and bottlenecks throughout the economy. The most effective way to promote the efficiency of an enterprise is to require it to maximize profits in a competitive environment. In theory, the goal of profit maximization need not rule out the use of the enterprise to also pursue noncommercial objectives, as when an enterprise is located in a remote area to promote regional development and then told to maximize profits. But even if it operates at peak internal efficiency the enterprise will in all likelihood generate a lower level of profits than if it were located closer to its markets or suppliers. It will be hampered in its efforts to compete. To solve this problem, the government may give the enterprise a transfer to cover its added costs or, in the worst case, to protect the enterprise from competition. But such transfers are notoriously hard to calculate, and ample evidence shows that subventions lead to declining productivity-the firm can attribute its poor performance to the cost of meeting the social goals. Also, because customers of state enterprises are 18 Public Enterprise Reform: The Lessons of Experience often large industrial users, wholesalers, or the middle and upper classes, they-not the poor-may benefit most from the subsidies. The subsidy is ultimately paid by the taxpayer, or, if the deficit is financed through inflationary monetary expansion, by the public at large. Given the regressive nature of taxes in many developing countries and the effect of inflation on the poor, the net result may be to increase income inequalities. If, on the other hand, the government imposes a social objective without an offsetting transfer, enterprises are forced to cover the cost through borrowing or by deferring maintenance and replacement investments. Because the absence of a transfer makes the cost of the social welfare goal invisible to the budget for a time, it can encourage governments to continue programs they can ill afford. Probably the worst distortions caused by forcing noncommercial objectives on state enterprises arise when agricultural commodity boards purchase products at prices below market and even below cost, typically in an effort to keep prices low to the urban consumer and to provide employment at often bloated staff levels. Such behavior further impoverishes the rural poor-the poorest class in most developing countries-and discourages farmers from producing commodities in which the country has a comparative advantage. Most governments embarking on the reformn of their state-owned enterprises have come to recognize, first, that some social objectives are beyond their current means and, second, that in many cases state firms are not the best vehicle for the pursuit of such goals. But dismantling such programs in the short run can be politically difficult, and developing alternative, less costly ways to meet social welfare goals takes time. Interim reform efforts should try to make the social goals explicit, calculate costs carefullly, and finance the added costs through the budget rather than to decapitalize state enterprises. Government should closely monitor agreed indicators of performance to minimize the erosion of efficiency engendered by subsidies. One way to encourage productivity despite price controls may be to evaluate profitability on the basis of trends in constant or shadow prices, as discussed in Chapter 4. Diagnostic Overview Once the government has assessed the fundamental objectives for which the enterprises were created and determined which, if any, of these Determining the Role, Scope, and Objectives of the State-Owned Sector 19 objectives are still valid, it must diagnose the condition of individual enterprises, deciding which need restructuring and which should be privatized or liquidated. The government then sets a timetable for action. Even though the sector's problems may be already well known, putting all the issues and recommendations together in a single document seems to be indispensable. The diagnosis need not be elaborate, especially if the issues have already been studied piecemeal; in that case a review of the earlier studies and an examination of the reasons why the government failed to act or why its reform efforts failed may help avoid the repetition of earlier mistakes. Appendix A provides a checklist of diagnostic topics. In most developing countries, the government owns a core group of state enterprises whose efficient functioning is critical to the economy. This group usually includes electricity companies, railways, ports, telecommunications firms, and very large extraction enterprises. Because of such firms' size and monopoly character, governments are usually reluctant to sell them in the short term, yet improving their performance is of the highest priority. Nonviable firms outside this group clearly should be liquidated; typically these entities include small retail outlets, state farms, and marketing boards, and in some countries their number may be large. Usually, many enterprises fall between these two extremes and could be privatized or remain in the state sector under the new economic rules, perhaps with some assistance. Governments tend to classify enterprises as "strategic" or "essential" and "nonstrategic" or "unessential." Economic analysis, on the other hand, usually attempts to divide enterprises into "viable," "potentially viable," or "nonviable." The government of Somalia adopted the following classification and action plan, using both sets of terms: Category Action Essential/strategic, viable Retain Essential/strategic, nonviable Retain, take specific steps to improve Nonessential/nonstrategic, Divest in whole or in part to viable the private sector Nonessential/nonstrategic, Liquidate nonviable 20 Public Enterprise Reform: The Lessons of Experience The Ministry of Agricultural Production and Agricultural Reform in Madagascar applied a similar classification to its state agricultural operations: "strategic" meant that the enterprise carried out essential activities that "for the moment and under present policy conditions, cannot be undertaken by any other mechanism." "Viable" meant that the enterprise either did now or could conceivably (under present policy conditions and without fundamental macroeconomic changes) run along commercial lines at a profit. Obviously, these definitions leave many areas to interpretation. "Strategic" may mean that the enterprise is seen as fulfilling a crucial military or economic role, or it may imply merely that the political costs of disinvestment outweigh the potential benefits. "Viability" is also a hazy concept. A private firm will usually be liquidated if it is insolvent. Insolvency can be straightforward, as when a firm is unable to meet its maturing obligations or it can be judgmental, as when total liabilities exceed the fair market value of assets. But a state-owned enterprise may be insolvent because the government failed to capitalize it adequately or it may be solvent because it has access to subsidized funds or enjoys monopoly privileges. Under the definition in Madagascar, for example, an enterprise that earned financial profits at great economic cost because of distortions created by "present policies" could be considered viable. Potential viability, therefore, should be based on an assessment of future economic and financial viability under reformed policies, weighing the benefits of continued operation against the costs of needed financial or physical restructuring. Classifying all firms at the outset is not necessary; quick action to liquidate, sell, or restructure some of the extreme cases is sometimes more important and lowers the risk that the reform effort will become an academic exercise. Setting Objectives for Individual Enterprises With the role and scope of the state-owned sector initially decided, the government should move to reform its relationship with its state enterprises through the following four actions: (To be effective, these actions need to be accompanied by internal improvements in the management of finance, production, investment, and personnel, which do not differ markedly whether the firm is public or private.) Determining the Role, Scope, and Objectives of the State-Owned Sector 21 * Set clear and attainable objectives compatible with the commercial operation of the firm. * Give management greater autonomy over the operation of the firm and select managers capable of operating independently. * Establish clear rules, procedures, and limits for government involvement in decision-making. * Hold managers accountable by negotiating targets, monitoring and evaluating results, and rewarding managers and staff on the basis of performance. Two partly complementary means of addressing these four action areas are corporate plans and perfonnance contracts. Corporate Plans Some governments, including those of India, Pakistan, and Thailand, make their enterprises set objectives and medium-term strategy through the preparation of corporate plans. Ideally, a corporate plan should analyze the business environment of the enterprise, assess how it is likely to change, and enunciate goals and strategies for the future. The plan should include targets, benchmarks for monitoring their achievement, and an investment program. Although the plan can be a means of negotiating objectives with government, its primary purpose is to enunciate management's vision of the future of the firm. To do so, the plan must become a part of the culture of the enterprise and be reflected in its annual budgets and investment projects. In the public sector, corporate planning has often proved fruitless, requiring as it does management with independence, continuity, and commitment. Managers of state-owned enterprises face a number of disincentives and obstacles to adapting to medium-term changes in their environment through corporate plans or through any other means: Their market is protected and their prices controlled, subsidies or loans are readily available to pay for past mistakes, and the penalty for failure is remote; frequent rotation of managers, directors, and supervisors reduces the commitment to long-term needs; and the government's social welfare aims may take precedence over the long-term health of the enterprises. 22 Public Enterprise Reform: The Lessons of Experience These obstacles have led many observers to recommend that the state particularly avoid, or divest ownership in, "entrepreneurial" industries in which dynamic adaptation to changing consumer tastes is paramount, such as electronics. Performance Contracts Performance contracts have an advantage over corporate plans in generating a dialogue between enterprise and govermment and in the two- way nature of the obligations they create. Under performance contracts, governments pledge to meet their financial and other obligations to the enterprise and to renounce ad hoc interference; in exchange, the enterprise accepts negotiated perfonnance targets. Contracts between the enterprise and the government have been used in France since the late 1960s. The nature of the French contracts has changed over time, but they continue to generate a dialogue between the enterprises and the government and to clarify the objectives and operating framework of the firms. In 1980 their use spread from France to Senegal and thereafter to most of the rest of francophone Africa. In recent years, variations on the concept have been introduced or proposed in some African anglophone countries as well as in Argentina, Brazil, Mexico, India, and Bangladesh. The experience in France and Senegal shows that quantified performance targets are less important than the process of preparing and negotiating them. The agreements can help both parties translate vague intentions into specific goals, make the costs of achieving objectives more apparent, and thus allow a more rational consideration of costs and benefits. Also, the two-way nature of the negotiations improves each side's understanding of the other's operations and constraints. In Senegal, a majority of managers of enterprises with contracts believe that the contract helps give them more autonomy to operate commercially. Also, some slim evidence shows that Senegalese enterprises with contracts had lower operating costs than those without (Nellis 1989). Performance contracts have several disadvantages. Governments can and often do violate them without sanction, a situation that makes "agreement," a term used to describe the document in some anglophone and Latin American countries, more appropriate than "contract." They are complex, can take a long time to negotiate (as much as two to three years in Determining the Role, Scope, and Objectives of the State-Owned Sector 23 France and Senegal), and put a heavy burden on the limited supply of skills and information in developing countries. And the complexity of the process limits the negotiation of the contract to only a small proportion of the state-owned sector at one time. The experience so far with performance contracts provides several lessons. First, the contract system should be one part, rather than the foundation, of any reform effort. Before any contracts are negotiated, the government and the enterprise should agree on general principles of operation, and the enterprise should develop a vision of its future and a corporate strategy that fits the government's macroeconomic strategy. When introduced, the process should be recognized as experimental; and the use of external consultants should be minimized because the value of the device is its promotion of a dialogue between government as owner and management. Second, at the outset, governments should choose an enterprise that is not operating in a competitive environment and that is not in desperate financial condition. Firms operating in a competitive market without a heavy social-service burden usually do not need contracts; they should be privatized or, if that is not acceptable, regulated by the market with minimal government supervision of investments and debt. And although governments typically wish to apply contracts to their poorest performers, the French experience, confirmed in Africa, is that the weaker the firm, the harder will be the negotiation of a contract. Third, the contract's stipulation of payments from the government to the firm are often unrealistic given the financial constraints on the government, and governments often do not bargain in good faith on the matter. The government's subsequent failure to honor its financial obligation calls the entire contract into question. The lack of realism and the lack of good faith are hard to overcome, but contracts (like some in Senegal) may try to do so by identifying the services that will be suspended in case of nonpayment. Some contracts also oblige the government to budget a specific line item to cover anticipated payments to the enterprise concemed. Another requirement for success is a supervisory body that on the one hand is strong and competent enough to push the financial authorities to honor their obligations and on the other does not become an additional layer of control over the enterprises. 24 Public Enterprise Reform: The Lessons of Experience Fourth, the negotiation of the contract should be part of the government's budget process; when it is not, the contract is in jeopardy of being disregarded in budget decisions, and negotiating outside of the budget process places a heavy burden on management. Also, to ensure that the agreements reached in the contract are honored elsewhere in government, other decisionmakers, especially the finance minister, must be party to the negotiations or be otherwise bound by the contract. Finally, contracts, especially initial ones, should be limited in scope and provisional. The French have simplified their contracts over time-moving toward shorter, more general statements of intent and broad policy and away from detailed quantitative specifications-to put more emphasis on the process than on the product. In Africa, where the financial and economic situation is far more weak and uncertain than in France, the contracts have of necessity remained more detailed and complex. Senegal has tried to address the uncertainty in the contract process in part by shortening the intervals between periods of review and modification. Although the contract process can, as in Africa, become ensnared in the problems of the economic environment, its underlying idea is quite simple and potentially beneficial: agree on what the enterprise should do and on what it needs in order to do it. The fulfillment of the contract is contingent on the commitment of the two sides, but particularly on that of the government, to honor their obligations. A more complex contract will not fare any better than a simple one if that commitment is lacking. References Choksi, Armeane M. 1979. State Intervention in the Industrialization of Developing Countries: Selected Issue. World Bank Staff Working Paper 341. Washington, D.C.: World Bank. Nellis, John. 1986. Public Enterprises in Sub-Saharan Africa. World Bank Discussion Paper 1. Washington, D.C.: World Bank. Nellis, John. 1989. Contract Plans and Public Enterprise Performance. World Bank Discussion Paper 48. Washington, D.C.: World Bank. 3 DECENTRALIZED MANAGEMENT: A DELICATE BALANCE Whether or not a government deals with its firms through the mechanism of a performance contract, the shift toward market structures requires changes in governance that will free state-enterprise management to respond as effectively as possible to competition. The developing countries that have been most successful in promoting efficiency in their state-owned enterprises (Chile and Korea, for example) have strengthened budgetary discipline; pushed competition wherever possible; followed appropriate pricing policies; and relied on arm's-length, performance- oriented controls that focus on operating efficiency. These countries have also benefited from an ample supply of the skilled administrators and managers necessary to implement these policies. Circumstances in most developing countries make it harder to implement the same solutions. Many governments in sub-Saharan Africa, for example, own large numbers of enterprises, which usually dominate the small local market and are protected from import competition. Governments have attempted to control this monopoly power and achieve social goals by regulating the enterprises through an elaborate institutional structure. Skilled manpower to staff the regulatory agencies and manage the enterprises are in short supply, and typically the controls stifle managerial initiative and accountability and fail to achieve their purpose. Even in countries with no shortage of administrative skills, regulation of state enterprises has been pointed at detailed control rather than high performance. Oversight agencies typically spend inordinate amounts of time approving trivial decisions best left to the managers: line item budgetary expenditures, routine maintenance, relatively small purchases, and promotions or dismissals of employees. For example, the Department of Statutory Bodies in Malawi approves every appointment or promotion and has issued guidelines on the amounts that its enterprises can spend on uniforms; the 25 26 Public Enterprise Reform: The Lessons of Experience Bureau of Public Enterprises in India at one point issued specifications for fencing factory perimeters. The tight focus on small decisions prevents adequate control of major ones, and at the level of the state-owned sector, the preoccupation with detail often leaves large and powerful enterprises free to spend millions while shackling smaller firms. Managers of powerful enterprises often bypass the control agencies and appeal directly to top authorities in government (who typically appoint them). In some developing countries, excessive autonomy has become more important than misdirected control as the cause of poor performance. In Brazil, Argentina, and the Philippines, for example, many state enterprises have enjoyed monopoly powers, privileged access to credit, and tax revenues earmarked for their benefit; until recently, they were able to evade government oversight when contracting debt, creating subsidiaries, making investments, and so on. For example, until the creation of a special state secretariat in 1980, Brazilian authorities did not have adequate information on the number of state-owned enterprises in the country, much less on their plans and performance. The challenge in all cases is to establish a delicate balance between autonomy and accountability. Accountability is meaningless without autonomy, for how can owners of an enterprise hold managers responsible for performance unless the managers have the freedom to decide? But autonomy without accountability is license. Unfortunately, governments tend to embrace control and accountability and ignore their pledge to increase managerial autonomy. Moreover, the effort to increase autonomy may be hampered by the fact that reforms are usually enacted in times of austerity, when decisions tend to become more centralized. And existing legal frameworks often allow government to intervene at will and may even mandate needless interventions. Building legal and procedural safeguards against undue intervention is therefore important, as is a program to grant state-owned enterprises greater independence as competition and skills increase. The Role of Government The first step in achieving a balance between autonomy and accountability is to identify the issues that are the proper concern of the government. The experience of private-sector conglomerates with their Decentralized Management: A Delicate Balance 27 subsidiaries has led Leroy Jones of Boston University to suggest the following roles as legitimate for a government seeking more productive control of its state-owned sector (Jones 1987): * Set the basic objectives (they must be fundamental, such as make a profit, or diversify into new export areas, or privatize part of the activities). * Appoint the managing director and the board members. * Evaluate performance against the basic objectives, and reward or penalize the managing director accordingly. * Review financing decisions that affect public funds (for example, requests for government equity, for debt with government guarantee, or for reinvesting profits instead of paying dividends). * Monitor prices if the enterprise is a monopoly. * Plan for the long term with regard for the interdependencies of enterprises (for example, to determine industrial, energy, or agricultural policies; to develop sector plans; or to phase out activities that might best be left to the private sector). * Do nothing else. Some items in this list are open to dispute-perhaps the board should appoint the managing director, perhaps the government should review even self-financed investment programs if they have important macroeconomic implications. But the state-owned sectors in most developing countries are so far from the independence implied by these actions that such finer issues have little immediate relevance. The more common and basic problem is to identify the layers of excessive control that need to be peeled away and establish a timetable for action. The autonomy that can be afforded to a competitive, financially independent firm is greater than that appropriate for an insolvent enterprise or a monopoly; hence, an increase in the scope of competition eases the government's oversight burden while creating pressure for efficiency. Thus, as market solutions take effect, the government will need to reduce direct controls and allow management to respond to competition. Portugal, Thailand, and Zambia have coordinated the introduction of market-oriented policies with reductions in controls by classifying each state-owned enterprise according to its degree of independence from government support and the degree of competitiveness in its market. The classification 28 Public Enterprise Reform: The Lessons of Experience is a dynamic one-as each firm becomes less dependent on government for current or capital operations and can borrow without government guarantee, and as its market becomes more competitive, it is granted more autonomy. In the ideal case, the government can give a competitive, financially independent state enterprise a profitability target at the beginning of the year together with supplementary indicators for planning, maintenance, and so on. The firm then would set prices and determine expenditures at will, with the government reviewing only those investments whose costs pass a predetermined threshold, as well as quarterly reports, performance evaluations, and annual accounts. In the case of monopolies, the government at the beginning of the year would set a formula or ceiling for prices. Financially dependent firms would also be required to submit a budget and an investment plan for approval. To charge prices higher than the formula or ceiling permits, the monopoly would have to present supporting evidence to the pricing authority, which would have a fixed period, perhaps a month, to object before the increase went into effect. Typically a number of monopolistic state enterprises will remain as such, some because they are natural monopolies, others because they dominate a small market with high import barriers that the government is unwilling to reduce, or because they are the only local producers of nontradables where barriers to entry are high, and so on. In small economies, such as those in sub-Saharan Africa, state-owned monopolies are likely to dominate the state-owned sector in the near term even with trade reforms; this fact does not preclude the possibility of introducing a more arm's-length, performance-oriented system that will increase their autonomy while also protecting the consumer. Indeed, more selective controls are especially important in countries where government capacity is stretched thin. State Enterprises and the Government Budget In general, developing countries need to change the way they handle the budgets of state enterprises. First, governments must refrain from reviewing the detailed, line-item expenditures in the budget. If the enterprise is competitive and does not require financial support, its budget should merely provide supporting information on how the enterprise Decentralized Management: A Delicate Balance 29 intends to achieve its targets, and the budget should not be subject to approval beyond the board of directors. If the enterprise requires government support or is a monopoly, the Minister of Finance will approve the budget, but again the government should not get involved in the details of current expenditures. Second, the budget for the enterprise should not be consolidated into the government budget; such a move invites intervention in the details of enterprise planning, forces the manager to get ministerial permission for even minor changes in line item expenditures, and leads to rigidities and costly delays in decision-making by the enterprise. Nonetheless, the government must correctly budget for subsidies, transfers, payments for goods and services, and other outlays to state enterprises. Indeed, one barrier to reform has been the failure of governments to allocate sufficient resources to cover such costs of adjustment as the payment of arrears, severance pay and redeployment costs, and explicit subsidies for social welfare objectives. Finally, performance evaluation should not focus on a comparison of actual and budgeted expenditures (Why did you buy eight trucks instead of seven?) but rather on a comparison of actual and targeted achievements in efficiency and profitability (Why did your return on equity fall? Why did you have 20 percent more down-time than your competitors?). The principle underlying these guidelines is that self-financed expenditures should be under relatively less control, while government- financed expenditures-including those financed with government- guaranteed debt or through taxing powers ceded by the government-and investments should be more tightly controlled. In making investment decisions, governments should consider not only the feasibility and desirability of the project but also the viability of the enterprise. Pakistan, for example, sharply curtailed investments in its state-owned industrial enterprises because of the failure of those firms to operate their existing capital stock efficiently. Improving the Institutional Framework The oversight structure for state-owned enterprises varies greatly from country to country. Figure 1 gives a quick° overview of some of the responsibilities assigned to the five levels of organization most commonly involved in enterprise management: the ministries, the enterprise Figure 1. The Institutional Framework: Responsibilities of Key Oversight Bodies and Boards of Directors Finance ministry Sector mnistry Central SOE Holding company Board organization Review and approve Set sector policy Evaluate and rnoritor Appoint board and Approve budgets and Subventions Review and approve performance managing director corporate plans Investments which corporate plans and Suggest follow-up to of subsidiaries Approve annual affect government investment evaluation Approve budget, accounts finance Analyze sectoral investmnents, Momitor performance Dividend payments Other functions trends and mocro- accounts, and on quarterly basis economic impact borrowing of and advise manage- Other functions Serve on board Improve coordination, subsidiaries ment Nomninate or appoint government Approve creation or Approve investments Serve on board management team oversight dissolution of Approve major Approve monopoly and much of Board Safeguard managerial subsidiaries procurement prices Approve budgets autonomy Do company Norninate or appoint b Approve budgets Approve borrowing Standardize reports corporate plan and management Approve borrowing Approve major and maintain central approve subsidiary Approve major Apptove seks, liqki- procusement data bank plans changes in corpo- dation or creation of Recommend creation, Develop files on rate poliey vis-a-vis SOEs sale, or dissolution candidates for Otherjfunctions staffing, marketing, Evaluate and monitor managerial slots intemnal controls performance Shift funds, invento- Approve sales of Other functions ries, other assets assets from subsidiary to Approve budget subsidiary Approve investments Borrow and distribute Approve personnel funds actions Provide centralized Set sectoral standards services (training, for labor policy and computerized MIS) corporate practices Appoint or second key officers (finance directors) to subsidiaries Decentralized Management: A Delicate Balance 31 boards, and, in the middle, the central oversight organization. The "core functions" are critical under most circumstances and therefore are common to most countries. The "ancillary functions" in the figure are also often found, but some of them may be counter-productive. Many other entities besides the five in the figure may manage aspects of state enterprise operations: the labor ministry or civil service commission may oversee personnel policies; the central bank may need to approve borrowing and foreign exchange allocations; the commerce or trade ministry may issue export and import licenses; the public auditors will either supervise or carry out audits of the enterprises; and parliamentary committees or special courts may evaluate performance and investigate problems. Virtually all developing countries have attached their state enterprises to a sector ministry, which signs off on the major decisions, sometimes consulting the finance ministry; in French- and Spanish-speaking countries, the sector ministry normally cedes oversight of all financial decisions to the finance ministry. In some cases a minister of portfolio further complicates the reporting relationships. Two major goals in the reform of the institutional framework are (1) to improve the quality of oversight by shifting control in the direction of ex post performance evaluation and away from prior intervention in decisions more efficiently handled by management or the board and (2) to create groups in the sector and finance ministries trained in the review of state enterprises. Making the central state-enterprise organization a more effective buffer between the ministries and the enterprises can often help realize these goals. Holding companies can also enhance the autonomy of state enterprises if their tendency to become overly controlling can be avoided. Lastly, the enterprise boards of directors should be structured to maximize knowledgeable oversight and continuity. Core Ministries The finance ministry plays a key role in determining the course of state enterprises, especially in times of austerity. Where such a ministry is weak, a central state enterprise organization capable of undertaking the analyses required for informed decisionmaking can compensate in the short run; indeed, where skills are in short supply, putting such an organization in the finance ministry may make sense. But over time, the importance of 32 Public Enterprise Reform: The Lessons of Experience the finance ministry will require its analytical capabilities to be strong if the prospects of state enterprises are to improve. Improving oversight provided by the sector ministries is more problematic-they are usually the main culprits in countries where excessive intervention is rampant; and they have, appropriately, often been the chief victims of reforms-that is, they have lost their powers. They also tend to advocate expansive investments for state enterprises, provide ineffective performance evaluations, and effectively resist divestiture. Reorienting and improving these ministries are difficult tasks, especially where skills are scarce. Some countries, such as Brazil and Zambia, have virtually bypassed the sector ministries with varying degrees of success. In most countries, the ancillary activities of these ministries (in Figure 1) need to be curtailed and the areas of permissible intervention defined. As in the case of a weak finance ministry, central state enterprise organizations can step in quickly to improve oversight, giving reform efforts time to focus on those sectors, such as energy, in which the sector ministry provides an especially strong link between government policy and the relevant state enterprises. Ministries of portfolio, also called public enterprise ministries, have- with one or two exceptions-proved to be ineffective. They suffer from the same weaknesses as the sector ministries, and they lack detailed understanding of the sector. In the worst cases they offer nothing but an added layer of control because the state enterprise must still also report to the sector ministry. Like multisectoral holding companies, ministries of portfolio tend toward a bureaucratic focus on administrative controls. Central State Enterprise Organizations A central oversight or coordinating organization has been created or improved in many countries (including Ghana, Mali, Senegal, and Pakistan) to block ministerial interventions in state enterprises by breaking the one-on-one relationship between the sector minister and the managing director. These central organizations often answer to the office of the president or prime minister, to the cabinet, or to a special interministerial group; they coordinate the multiple actors involved in state enterprise decisions and can take an interventionist ministry to task for not playing by the rules and for ignoring the wider interests of the economy. Decentralized Management: A Delicate Balance 33 The track record of central oversight organizations is mixed. The oldest-the India Bureau of Public Enterprises-has been described at different times as being supportive and a spur to efficiency, as highly interventionist and bureaucratic, or as ineffectual and generally ignored, depending on who was in charge of the bureau and of the ministry of finance to which the bureau is attached. The mere existence of such organizations is not enough to assure the right balance between enterprise autonomy and accountability, but under the following circumstances they can help reduce and coordinate intervention. First, the central state enterprise organization should be small and have a well-qualified staff. The ratio of staff to enterprises strongly influences the degree to which the organization may be tempted to intervene in decisions that should be management's prerogative. Moreover, a small size is particularly important where skills are scarce. As to the importance of skills, the organization is pointless unless the quality of its work is assured. The Bureau of Public Enterprises (BEP) in Mali has some 7 professional staff to oversee 58 state enterprises. The State Enterprise Secretariat in Brazil's Planning Ministry, which has had some success in curbing the excesses of an overly autonomous sector, has about 60 professionals for 250 enterprises. The National Investment Board (NIB) in The Gambia has 5 professionals to monitor 17 enterprises. The appropriate number of staff depends on the decision about the organization's functions; the aim should be to create a small, nonbureaucratic elite. Second, the central organization should have access to the highest level of power; it will need top backing to deal credibly with ministers of industry and finance and directors of petroleum and mining monopolies. In The Gambia, the NIB has access to the president through the secretary to the cabinet. In Mali, the BEP reports to an interministerial committee. In Ghana, the head of the State Enterprise Commission is a member of the cabinet. Either the head of the central organization or the person to whom the organization reports must assure that its opinions are considered before the government makes any major decision on state enterprises-a condition requiring access to the economic cabinet and, in smaller and more hierarchical countries, ultimately to the head of state. The choice of bureaucratic location is driven by this need for access and the desired scope of power. In Brazil and Mauritania the organization is located in the 34 Public Enterprise Reform: The Lessons of Experience planning ministry; in The Congo, in the ministry of finance; in Kenya and Burundi, in the office of the president. Finally, its mandate should be clear, with its scope of decisionmaking authority and its relationship with the relevant ministries spelled out. Its functions should not overlap with those of other agencies. The tasks assigned to such an organization, beyond the core responsibilities shown in Figure 1, will vary from case to case. In The Gambia, for example, the central organization is purely advisory. In Brazil, the Secretaria de Contr6le de Empresas Estatais (SEST) approves and monitors enterprise budgets, all foreign and domestic credit operations, and proposals to create, expand, or liquidate state enterprises. SEST establishes firm ceilings for major spending categories, which are summarized in an annual SEST budget authorized by the president. A proper central organization should, it is argued, be given decision- making powers because it would be staffed with some of the best available people to make decisions; it would be freer of political pressures than the traditional agencies; and it would be small, focused on performance, and therefore less likely to intervene than the ministries. Moreover, for smaller countries and countries with few enterprises, the scarcity of skilled people makes the centralization of enterprise decisions outside the ministries most logical. The arguments against an effective, and therefore powerful, central organization are that the ministries will never relinquish their powers, so the central organization will simply add another layer; the system of checks and balances is better served by an advisory body than by a decision- making one; and the abuse of power will continue-the central organization will simply replace the ministry as the interventionist body. Also, in countries with large and powerful state enterprises, a central organization may be hard-pressed even to monitor the sector adequately; decisions should usually rest with sector ministries, which deal with a manageable portion of the state enterprises. Thus, the decision of how much power to give a central state enterprise organization depends on the size and scope of the sector; the quality of oversight in the sector ministries; the support the central organization will receive from above over the long run-that is, if it advises a powerful official or body, the central organization should not need its own power base; the scarcity of skills in the country; and whether power can be realistically transferred. Decentralized Management: A Delicate Balance 35 For example, Thailand's relatively small and shrinking public enterprise sector is monitored by a large number of government bodies and could benefit from consolidating oversight into one or two committees comprising the major actors. The number of state enterprises in Thailand dropped 13 percent between 1984 and 1988, from 69 to 60. Sixteen of the remaining firms provide virtually all of the country's electric and petroleum power, air, rail, port, and mass transit service, telephone communications, and water; these 16 firms also dominate the nonfinancial state enterprise sector, with 85 percent or more of its assets, employment, and investment, and 60 percent of its government subsidies and transfers. Supervision of these and other Thai state enterprises is scattered among the national development bureau, the budget office, the comptroller general, the auditor general, and the fiscal policy office; also involved are a supervisory ministry, sector committees, and other sector agencies. A case in point is the state-owned Bangkok Mass Transit Authority, one of the two largest state-owned money losers; it suffers from the inadequate coordination and duplication of effort that arises from the fact that not only all the central government agencies just mentioned but more than 40 other statutory and ad hoc agencies and committees deal with some aspect of transport in the capital. With centralized oversight, all these groups might work through the Ministry of Transport, which would represent their input to one entity (such as the Comptroller General, which already has monitoring responsibility, or a committee of staff members from the major central govemment offices). At the top of this system could be a committee of senior representatives of the prime minister and the other major offices to issue final approvals. Holding Companies Holding companies are often suggested as a vehicle for advancing the goals of state enterprise reform. By creating conglomerates and thus increasing the size and power of the state enterprise sector compared to the ministries, holding companies can work in favor of enterprise autonomy. Moreover, subsidiaries of state holding companies are often explicitly exempted from some of the controls exerted over directly owned enterprises. Hence, most state enterprises in Mexico which are not part of conglomerates are under more direct state control than those in Brazil, where holding companies are common. Likewise, managing directors in 36 Public Enterprise Reform: The Lessons of Experience the subsidiaries of the Malawi Development Corporation are not held to the standardized compensation system for state enterprise managers and hence earn more than the management at headquarters. Among the other factors favoring holding companies are their ability to exploit economies of scale (thoiugh bulk procurement and centralized training, for example) and their ability to function more effectively than smaller firms in international export and capital markets. Also, the liquidation of a nonviable subsidiary can be easier than that of a freestanding state enterprise, and that is so because, in general, holding companies can be an effective buffer against political interference. But public-sector holding companies, especially large conglomerates that combine unrelated subsidiaries, tend themselves to become political, bureaucratic, and control-oriented. They may promote monopolistic or oligopolistic behavior. And they often "cross-subsidize," that is, shift funds, inventories, and skilled staff from profitable companies to keep alive nonviable firms, thus dragging down the performance of all subsidiaries. Accountability can become a serious problem; in some countries, funds are diverted within the holding company for cross- subsidization among enterprises in ways that are hard to trace. Finally, holding companies add the complication of an additional managerial layer and require managerial skills where they may be in short supply. One way to gain the benefits of conglomerates while minimizing the problems is to avoid the large multisectoral holding company, such as Italy's Istituto per la Ricostruzione Industriale with its 400,000 employees; smaller, sectoral holding companies might be more beneficial if they are subjected to proper safeguards against monopoly power. Another protection against the negative tendencies of holding companies is to limit the size and powers of the front office. For example, the Zambia Industrial and Mining Corporation, which operates more as an oversight organization than as a holding company, has only 20 people in its front office to oversee 72 enterprises, and it is prohibited from shifting resources from one subsidiary to another. Holding companies clearly are not a panacea-countries (Pakistan and Egypt, for example) have introduced them in one reform, removed them in another, reintroduced them, and so forth. The following factors should be considered in deciding whether to establish a holding company for state- owned enterprises: Decentralized Management: A Delicate Balance 37 1. The effect on competition. In most cases, a reduction in competition should be avoided, although a holding company could conceivably reduce domestic competition but allow the conglomerate to compete in export markets. 2. The size and maturity of the state enterprises. A holding company of powerful firms could itself easily become too powerful; small, new, and vulnerable firms are more likely to benefit from the conglomerate. 3. The size and diversity of the holding company. Large, unwieldy groups tend to become inefficient and bureaucratic in both the private and public sector, a problem that partly explains the trend toward smaller groups in the United States. 4. The nature of administration in the country. If the government is interventionist, poorly staffed, and paralyzed by red tape, how can these traits be guarded against in the holding company? Boards of Directors A small number of directors who have been appointed, for the most part, on the basis of their technical expertise and experience rather than as representatives of government ministries, who are properly compensated, and who take the job seriously can be an important asset to a company. In principle, government can then be encouraged to decentralize power to the board as a way of depoliticizing decisions and advancing the company's interest. Few governments have completely followed New Zealand's example and relinquished ministerial representation on the board, permitted the board to hire and fire the chief executive officer, and allowed it to exercise all the responsibilities shown in Figure 1 without further approval; the exceptions are subsidiaries of holding companies. Pakistan illustrates the usual approach. There, the boards of industrial state enterprises are legally the ultimate authority for many major decisions, but in practice the boards usually first clear major decisions informally with the appropriate ministry. Furthermore, the amount of power a government can realistically be expected to delegate to a board of directors is limited. Boards of private- sector companies in developed countries have sometimes acted against the best interests of the shareholders even when they would be held liable for 38 Public Enterprise Reform: The Lessons of Experience doing so. Clearly, the board cannot be the only way to hold management accountable, particularly when public monies are at risk. Another problem in smaller countries is the difficulty of finding sufficient qualifed directors from the private sector whose interests do not conflict with the welfare of the enterprise. Nevertheless, boards of state-owned enterprises can play an important and constructive role, particularly if the following guidelines are observed: * The board should be small (ideally, seven to nine directors) to conserve the short supply of skilled individuals, assure that meetings are fully attended, and minimize the financial burden the board places on the enterprise. * Appropriate qualifications for directors should be spelled out and enforced; where necessary, directors should receive training. Directors should be encouraged to nominate new members; and directors should be dismissed for incompetence. . Directors should have fixed terms of three to five years, with a staggered turnover to assure continuity. * Boards should hold regular meetings quarterly or, at most, six times per year and should always produce a report. * Directors should be compensated adequately but not lavishly. Their salaries should be treated as compensation for work performed. * Management should be required to submit a report and other appropriate information to directors well before regular meetings. * The oversight organization or the line ministry should regularly evaluate the board, and if it finds the board incompetent, dismiss it. * Government representation on the board should be limited to one or two members and alternates, who should be held to the same standards of preparation and performance as the other members. Directorships should not be treated as a sinecure for senior officials or politicians. In no case should the government be in the majority, and a minister should never chair the board because that destroys its collegial nature and makes it an arm of the ministry (Ayub and Hegsted 1986). Elements of a well-functioning board system can be seen in Korea, which in 1983 dramatically changed the way it manages the largest and Decentralized Management: A Delicate Balance 39 most important group of its public firms, the 26 Government Invested Enterprises (GIEs). Among the reforms was the restructuring of the GIE boards of directors. The government eliminated their role as standing executive bodies, thereby placing responsibility for implementation completely in the hands of management and emphasizing the policymaking and review functions of the boards. The boards now meet less frequently, once each quarter. The position of chairman and director for each board, previously permanent and paid, and sometimes held by persons at least as interested in the remuneration as in the welfare of the enterprise, is now limited to a three-year term and paid only for expenses. Government representation on the boards was reduced to two positions; the other directors now are businessmen, accountants, professors, and others who can assist management in the conduct of the business. Raising Managerial Skills Organizational reforms will be effective only if management is competent to operate a commercial state enterprise. Some reform programs have taken away from political authorities the power to select managers and vested it in a technical oversight agency or board of directors. But few governments have been willing to relinquish this political power, except in the case of state-enterprise holding companies, which have often won the right to nominate or even appoint the managers of their subsidiaries, while government appoints each holding company's chief executive officer (CEO). Even if the decision maker is political, however, the appointment decision need not be. Oversight organizations should be empowered to specify the qualifications required of state-enterprise managers, to conduct a management search, and to comment on the qualifications of candidates or nominate a short list of candidates. Where feasible, a more public appointment process, in which the qualifications of the candidates and the skills required for the job are widely known, can also help prevent the appointment of inadequate managers. A good performance evaluation system can also improve the quality of management if poor performers are fired, denied bonuses, or at least not promoted. The secondment of civil servants to state enterprises should be avoided. The skills of a bureaucrat are not necessarily those of a business person, and every effort should be made to distinguish the two careers. Enterprise 40 Public Enterprise Reform: The Lessons of Experience managers should face greater responsibility and risk than civil servants and be paid correspondingly; seconded civil servants who have the option of retuming to their ministries defy this premise. The CEO of the enterprise should be allowed to appoint, or at least nominate, his or her top staff; commanding a team selected and appointed by a ministry is difficult if not impossible. Changing managers for reasons other than performance should also be avoided. Enterprises, especially those that are immature or operate under considerable uncertainty, need continuity of top staff. Yet some countries allow every new administration, or even every new sector minister, to appoint new CEOs. When ministers (and in some countries, governments) have an average tenure of less than two years, the effect on performance is severe. Giving managers a contract renewable after no less than three years may be one way to reduce turnover, but such a contract has not been widely used in the public sector. An important feature of state enterprise reforms in Korea has been the change in the source of appointments to managerial vacancies; previously, almost half of all vacancies in the senior staff were filled from outside the enterprise, often with unqualified people, but preference is now given to intemal appointments, which are often made with some merit assessment. Senior staff in Korean state enterprises work under a three-year contract and can be dismissed only for incompetence. In many developing countries, particularly in sub-Saharan Africa, the pool of managerial talent is small because the number of entrepreneurs is small. This situation can be eased somewhat by reducing the number of state enterprises through mergers, liquidations, and sales and by tapping private managerial skills through leases and management contracts. Compensation Experience shows that the compensation of state-enterprise managers does not have to be on a par with that of domestic private-sector managers in order to attract and retain good professionals. Managers and other workers in countries as intemally diverse as India, Pakistan, and Mexico, for example, stay in state enterprises because they offer greater opportunities and stature than do family-dominated private businesses. (Those managers with skills that are internationally traded, such as petroleum engineers, are an exception.) In Pakistan and Mexico, to continue the example, the compensation package provided by state Decentralized Management: A Delicate Balance 41 enterprises exceeds the threshold standard of living that enterprise staff members see as their due. When the pay falls below that threshold, as happened some years ago in Pakistan, tumover increases markedly. Unfortunately, most governments don't provide this threshold standard of living or even attempt to understand the market for professionals. Instead, many developing countries place managers and nonunionized staff of state enterprises under the civil service pay scale or under a parallel system. Here again, exceptions are often made for holding companies or very large state enterprises. This approach to compensation has several disadvantages, as noted in Chapter 1: it rewards managers according to the size or economic importance of the state enterprise rather than on the basis of perfonnance; the civil service compensation system is typically rigid and therefore cannot adapt to the vast differences that exist among state enterprises or within the skills market; the pay differential between top management and workers is often inadequate to motivate the manager, and the link with the civil service results in compensation too low to attract, retain, and motivate skilled managers and other professionals. Furthermore, a bureacratic salary system, combined with pay increases that make little distinction between laborers and management, invites abuse through excessive overtime, bonuses, and fringe benefits and, worse, through corruption, for which managers of state enterprises tend to have ample opportunity. In Thailand, one profitable state enterprise paid 100 percent of its profits before taxes and dividends as bonuses in 1987, and by counting government subsidies, even some state enterprises with losses have paid bonuses; the government reports that overtime in some state enterprises constitutes nearly 80 percent of the wage bill, and some firms report fringe benefit packages that equal up to 25 percent of total remuneration. Governments typically respond to such abuse by trying to regulate fringe benefits and by putting controls on financial decisions, which in turn increase the system's rigidity and impair the performance of the enterprise. The problem of abuse is best met through the introduction of a performance evaluation system. If managers can be held strictly accountable for results, their tendency to squander money will be curbed far more effectively than through controls. The most important step toward an effective compensation system is to move away from a rigid link to the civil service. Most govemments will 42 Public Enterprise Reform: The Lessons of Experience insist on a pay ceiling that is a multiple of the top civil service rank but may accept greater flexibility within that ceiling. The best system is one in which the compensation package and increases are decided for each enterprise on the basis of its situation and performance. Giving managers an incentive to control labor costs and at the same time push for competitive wages for scarce skills can counteract the the pressures for upward creep in labor costs; where possible, state enterprises should be encouraged to monetize benefits if not eliminate them in exchange for higher salaries. Training Private enterprises in developing countries invest a great deal of money and effort in training their managers; state enterprises seldom do, attending more to technical expertise than to managerial skill. But under a reform program, training becomes especially important in helping state enterprise managers adapt to the new rules of the game. Managers who have operated for years in protected markets with restricted autonomy cannot be expected to respond automatically to competition and freedom by cutting costs, seeking new markets, trimming staff, closing production lines, and so on. While operational assistance can contribute in the first instance, training is necessary to help create the dynamic, adaptive attitude required for a sustained improvement. Such training, however, should be an explicit and on-going phase of staff development, not an add-on by consultants whose main assignment is operational (Operations Evaluation Department, The World Bank 1985; Lethem and Cooper 1983). Besides formal courses and seminars, training from suppliers, contractors, and purchasers can be useful. Twinning a state enterprise with a similar, more developed enterprise elsewhere has proven beneficial to managers and other staff of state firms in Tanzania and Ecuador for example. Joint ventures can also make an important contribution to training local staff if the initial agreement includes the transfer of knowledge. Persuading government to allow greater access to foreign suppliers may be a necessary first step toward increasing technology transfer. Among the problems common to training programs are overly ambitious agendas; irrelevance to the practical needs of those being trained, especially in the case of overseas training; poorly motivated trainees and a consequent high rate of absenteeism; lack of a follow-up course or Decentralized Management: A Delicate Balance 43 permanent training program; diversion of trainers to other tasks at the expense of the trainees; and inadequate preparation of the trainees-for example, experts assigned to train members of the Congo's Commissariat National aux Comptes in the oversight of state enterprise accounts found that the trainees did not know basic accounting. These problems are not new. A basic strategy to address them would include a separate training program based on realistic expectations (particularly, with experienced trainers not burdened with other tasks); a target group likely to benefit from the training-such as local staff with appropriate positions of responsibility-and motivated by specific incentives to take the training seriously; and phase-out or continuation mechanisms. The Approach to Organizational Change Some reform projects have devoted considerable effort to the organizational changes just described while neglecting aspects of decisionmaking, such as the budget process, that are critical to enterprise autonomy. New organizational arrangements can contribute to autonomy and accountability, but by themselves they are never a sufficient condition for improvements in efficiency-many countries have a long history of introducing, eliminating, and changing organizations with little positive effect on performance to show for the effort. As we argue throughout this book, corresponding changes must be made in the policy environment and in the content and rigor of supervision if the reform of state enterprise is to be effective. Although generalizing about the complex issue of institutional change would be foolhardy, the established rules of institutional development offer useful guidelines for most circumstances (Israel 1983-84). First, work with existing institutions and processes as much as possible to avoid bureaucratic proliferation. For example, temporary implementing bodies that are expected to disappear after the reform tend to become permanent. Second, prepare for the second (or third) best solution. Can the new organization still make a contribution even if its staffing, budgeting, or functions are not ideal, or will it become a bottleneck in its own right? Third, take seriously the institutional history. For example, have holding companies in the sector already been created and eliminated? If so, can this effort succeed where others have failed? 44 Public Enterprise Reform: The Lessons of Experience And last, prevent the institutional changes from merely reshuffling the boxes on the organizational chart-make them part of a broader adjustnent process that will change policies, decisionmaking procedures, and the location of responsibilities. References Ayub, Mahmood, and Sven Hegsted. 1986. Public Industrial Enterprises: Determinants of Performance. Industry and Finance Paper 17. Washington, D.C.: World Bank. Israel, Arturo. 1983-84. World Bank Technical Notes, Nos. 1-6. Washington, D.C.: World Bank. Jones, Leroy. 1981. Towards a Perfornance Evaluation Methodology for Public Enterprise with Special Reference to Pakistan. Paper presented at UNDTCD conference, November 1981, in Pakistan. _________. 1987. Speech prepared for World Bank seminar on the reform of public enterprise, February 25-27, 1987, in Annapolis, Maryland. Lethem, Francis, and Lauren Cooper. 1983. Managing Project-Related Technical Assistance. World Bank Staff Working Paper 586. Washington, D.C.: World Bank. Operations Evaluation Department. 1985. Handbook on Consulting Services. Washington, D.C.: World Bank. 4 HOLDING MANAGEMENT ACCOUNTABLE Evaluating the performance of a state enterprise manager, who typically faces administered prices, social welfare objectives, and governmental interventions and cannot lay off workers or close down lines, is very tricky. Ingenious and sophisticated schemes have been applied to the problem, with only mixed results. Nonetheless, performance evaluation is an essential part of the reform of the state-owned sector. If good performance and bad performance alike are ignored, then stagnating or falling productivity should surprise no one. Moreover, without a good monitoring and evaluation system, government is in the dark about the effects of its policies on its enterprises.1 A complete performance evaluation system includes a reliable and timely flow of appropriate information in standardized form; objectives, targets, and specific criteria for evaluation; an objective oversight body to monitor performance and evaluate results; a decision making body to act on the findings; and a managerial incentive program. Reform programs have varied from region to region and even from country to country. Efforts to upgrade performance evaluation in sub- Saharan Africa, for example, have focused on improving information and accounting at the enterprise level and generating a timely flow of a few key indicators to government authorities. Elsewhere, some countries are working to improve information on the macroeconomic influence and performance of the state-enterprise sector as a whole. Two countries with the necessary administrative skills, Pakistan and Korea, have implemented more elaborate systems to target, monitor, and improve performance. The early experience with performance evaluation in Pakistan and Korea shows that even technically flawed indicators seem to improve efficiency, and they clearly raise the attention paid to performance by both 1. The conceptual framework for the performance evaluation of public enterprises in developing countries has been established largely by Leroy Jones of Boston University. For more on the concept, see Jones 1981. 45 46 Public Enterprise Reform: The Lessonis of Experience management and government.2 After three years of the perfonnance evaluation in Pakistan, the 33 industrial state enterprises under the program nearly doubled their aggregate nominal private profits, and 7 out of a sample of 12 examined in detail increased their constant-priced public profits in the first two years. Managers of companies that had received bonuses for reaching their targets say that their staffs are highly aware of the targets and know what they have to do to achieve them. Under the influence of the performance target system, Pakistan's Petro Carbon, a small and inefficient producer of carbon black, converted a R9 million loss in 1984-85 to a R7.5 million profit in 1985-86 and brought down average production costs more than 40 percent per ton. Korea's 26 Government Invested Enterprises (GIEs)-the core of its state-owned enterprise sector-have shown significant improvement since a performance evaluation scheme was introduced in 1983 along with reforms to enhance the autonomy of management and streamline oversight. Public profitability in constant prices, which is the most meaningful overall financial yardstick for state enterprises and which is used as a target for the five largest GIEs, rose from about 13 percent in 1982 to more than 18 percent in 1987, largely on the sustained strength in the largest two of the five-the Korea Power Corporation and the Korea Telecommunication Authority. All GIEs are given some target for cost reduction and revenue growth; a study of the three years following the introduction of reforms found that all GIE sectors, financial, promotional, and business/industrial, had improved on earlier trends in the ratio of costs to revenues-for all GIEs the ratio was about 68 percent in 1986, compared to an expectation of about 73 percent based on past performance (Song 1988). More generally, the Korean reform program has sharply curtailed government intervention and raised the attention to performance in the operation of the GIEs. Most of the firms have created an office to develop objectives, negotiate targets, monitor achievement, and write evaluations of the results; the significance of these offices is suggested by the fact that 2. Besides formal performance evaluation, financial intermediaries can play an important role as a check on the performance of state enterprises. This study does not deal with this issue. In outline, an independent banking system that can freely assess management and creditworthiness will give clear signals about enterprise performance when deciding on loans. Governments that intervene in credit decisions thus rob state enterprises of an incentive to improve managerial performance as well as distort the allocation of resources. Holding Management Accountable 47 several executive directors have recently been elevated from their staffs. Many GIEs have developed detailed targets for departments, and promotions take the results into account. These results show that the performance of public enterprises can be improved. Gathering Adequate Information Developing a streamlined flow of timely and accurate data in a form usable to government oversight agencies requires the introduction of cost accounting and the training of accountants and auditors. Yet, despite considerable improvements, the information on state enterprises in some countries, particularly in Africa, is not timely or accurate enough to manage one enterprise properly, much less a group of enterprises. Commercial state enterprises should be required to comply with at least the basic accounting standards normally applied to private enterprises, and government may need additional information for oversight. Many governments have designated a public accountancy office or accountant general as the state-enterprise auditor, a move that should be resisted in most cases, for several reasons. First, the decision about auditing firms should normally be a management decision. The proper role for the accountant general would be to certify the firms that state enterprises could use as auditors and to review enterprise audits. If the enterprise's board of directors, or any of its oversight agencies, or the accountant general are dissatisfied with the quality of the audit, they can appoint their own auditors or the accountant general's office can undertake its own investigation. Second, government should not impose a monopoly when competition could help ensure quality. Government could help promote competition in auditing by encouraging enterprises to use local firms through tax advantages or the like, by encouraging local offices of foreign firms to train nationals in accounting, and by helping to find funds to develop accountancy training. Third, the services provided by the accountant general's office are often inadequate because it lacks the necessary skilled staff and has experience only with government audits. Even in countries where the accountant general's office can provide adequate service, it still should compete with private firms for the state enterprise market rather than be given a state monopoly. 48 Public Enterprise Reform: The Lessons of Experience Efforts to improve the quality and quantity of information have been the most common components of reforms supported by the World Bank, and they have produced the most concrete results in the short run. For example, a program to improve auditing of state-owned enterprises in Senegal contributed not only to the flow of reliable data-state enterprises began to produce quarterly and monthly financial statements-but also to a strengthening of local accountancy firms. By creating demand, the program led a number of Senegalese accountants to return from abroad; it also set new accountancy standards and promoted joint ventures between local and intemational firns. A sample of the most basic information a state enterprise should supply for govemment oversight is shown in Appendix B. If the enterprise cannot supply these data, then it cannot be managed efficiently, and top priority should be assigned to improving information and control. Another problem for public and private firms alike is the failure to fully account for inflation. Some developing countries, especially in Latin America, partially adjust their accounts for inflation, usually to revalue fixed assets. But inflation can have a major effect on other items in the accounts, including the profit and loss statement, and they should be adjusted accordingly (Veneroso and de Melo 1986; Tybout 1984). Setting Targets and Indicators of Performance If performance indicators are to give managers a clear signal, they should be few, nonduplicative, easily measured, and weighted by priority. The long history of performance evaluation in the Soviet Union and Eastern Europe shows the problems caused by partial indicators: production targets fail to signal managers to conserve costs or control quality; cost targets lead to neglect of items such as maintenance or training; combinations of production, quality, and cost targets make it virtually impossible to calculate a composite picture of performance. One message from this history is that, as a weighted indicator of revenues minus costs, profits are a powerful indicator of performance. Thus, for potentially competitive state-owned enterprises, government should increase competition and decontrol prices so that financial profits can be used to judge their performance. But profits are a poor indicator when the enterprise faces imperfect or no competition; faces government- administered input and output prices that are far removed from efficiency Holding Management Accountable 49 prices; does not adjust for differences between profitability measures in the private sector and in the public sector; and has objectives that differ from, and may conflict with, profit maximization. Evaluating the performance of such enterprises requires special techniques and is critically important because some of the largest, most economically influential, and heavily subsidized state enterprises-such as the utilities and railroads-are in this category. Lack of Competition and Administered Prices The schemes that allow governments to evaluate noncompetitive enterprises despite price distortions-such as total factor productivity or the measurement of profits in constant or shadow prices-require a great deal of data and a sophisticated staff to interpret them. They can also be hard to explain. Pakistan has been calculating profitability in constant prices for several years, but government officials and enterprise managers found it hard to accept it as the criterion for awards, so targets are set and bonuses are awarded principally on the basis of profits in current prices- even though many of the enterprises being evaluated are monopolies with cost plus pricing. Monopolies can be evaluated with techniques simpler than shadow or constant prices. The simplest measure, financial profit, could be used if prices could be set to approximate costs to society, but this is seldom achieved. An intermediate solution is an indicator of cost or quantity. Railway action plans, for example, typically include physical indicators (locomotive availability, staffing ratios) and targets for cost minimization. International benchmarks are often used for cost comparisons. The problem is that such approaches use multiple indicators, which give confused signals and are hard to evaluate, or they use partial indicators, which distort performance. Cost minimization targets, for example, can cause managers to neglect expenditures with high rates of return.3 Nevertheless, even a flawed evaluation is better than none. Less distortionary, more sophisticated indicators can be introduced gradually, 3. A related problem is asymmetric counting. For instance, giving an enterprise the three simultaneous targets of minimizing wages, reducing employees per ton, and reducing financing charges would cause the enterprise to count labor costs twice and capital costs once. The result would be a bias toward capital-intensive production. 50 Public Enterprise Reform: The Lessons of Experience following an effort to persuade management and govemment of their validity. Adjusting Profits Even under perfect competition, the private-sector measure of profits must be adjusted to be a useful measure of public-sector efficiency. The Korean system and the original design of the Pakistan system uses the concept of public profits, defined as Private profits after tax + taxes + depreciation + interest - nonoperating income - the opportunity cost of working capital. Taxes are added back because they are a return from the govemment's point of view, and their inclusion avoids giving managers an incentive to reduce them. Adding back depreciation avoids favoring older plants over newer ones, prevents profits and profitability from increasing (assuming no new investment) unless efficiency rises, and removes the incentive for managers to underdepreciate or to change their accounting practices. Interest is added back because changes in interest payments do not reflect changes in efficiency but merely transfers from one part of society to another. Adding interest also reflects the fact that central government, and not the enterprise manager, controls the capital structure of the business. Decisions on investment and debt are best handled through separate control systems and not through performance evaluation to assure the most efficient allocation of capital. Nonoperating income is excluded because it does not reflect operating efficiency. A charge is added for the opportunity cost of working capital (which in Pakistan was figured as about 10.5 percent times the sum of inventories, cash, demand deposits, accounts receivables, and the like), under the assumption that managers can control their working capital. The profit rate includes a charge for fixed capital through the inclusion of fixed operating assets in the denominator. Holding Management Accountable 51 An important argument against public profitability is that by ignoring taxes, interest, depreciation, and nonoperating income, the measure discourages managers from minimizing costs and maximizing income. But in fact most public managers cannot control some of these items, such as debt and interest charges, which are often determined by government; and they can easily manipulate others, such as depreciation or nonoperating income, to hide inefficiencies. The Korean system uses public profitability in constant prices to evaluate managerial performance and uses liquidity measures-the ratio of debt to equity-and other indicators to assure that financial solvency is improving. The government of Pakistan decided not to use public profitability as a target for its enterprises, choosing instead to rely principally on unadjusted private profits after taxes. As a result, several enterprises have been able to achieve a top grade because of nonoperating income such as interest on deferred foreign credits or because of a drop in interest following government debt relief, developments that have little or nothing to do with efficiency. On the other hand, for most of the 12 enterprises studied in detail, public profits were higher than private profits, mainly because of interest and depreciation charges-items largely beyond their control. To the extent that high interest charges resulting from the government's initial decisions on investment and capital structure make it impossible for managers to achieve their private-profitability targets, managers will have no incentive to improve the factors that they can control. Moreover, some of the factors under the control of management, notably working capital, are not measured by private profitability. The administrators of Pakistan's evaluation system have worked to correct these anomalies but, of necessity, only after the fact and with only limited success.4 In deciding whether to apply the concept of public profitability, the govermment of Pakistan worried, understandably, that its use might allow unprofitable companies to meet their targets and force the treasury to subsidize the cost of the bonuses. Indeed, the same worry causes the government to base all its state-enterprise targets on break-even or better. However, a strong case can be made in favor of rewarding managers who 4. During its evaluations, for example, the evaluators may discount performance because of an increase in administered prices or a windfall gain in nonoperating income. Managers have labeled these realistic adjustments "arbitrary" and have usually forced the evaluators to drop them. 52 Public Enterprise Reform: The Lessons of Experience improve the efficiency of unprofitable firms; a drop in losses is at least as beneficial to the treasury as a comparable increase in profits and, from a managerial point of view, is typically harder to achieve than an increase in profits; furthermore, attracting good managers to money losers is difficult enough without forcing them to forgo bonuses until they achieve break- even. Other Objectives State-owned firms have many objectives different from, and perhaps in conflict with, profit maximization. Some noncommercial objectives- national or state control of a sector or development of an industry neglected by private business-need not affect profit; and the adjustment to profits described above can avoid undesirable rent seeking such as tax evasion or interest arbitrage. But if government assigns to its enterprises objectives that reduce profits, government should subsidize the cost with a direct payment or reduce its targets accordingly (see Chapter 2). In some cases supplementary indicators can take account of the quality of service or give a longer time horizon. Under the Korean performance evaluation system, for example, 30 percent of the target is composed of qualitative indicators in the following areas: corporate planning, research and development, management information systems, internal coritrol systems, and the quality of service. Performance Evaluation Judging from the experiences of Pakistan and Korea, the agency empowered to implement the performance evaluation system must be expert, independent, objective, and credible, and must have the ear of powerful officials. The Pakistani performance evaluation system, created in 1983, has been applied only to industrial enterprises under the Ministry of Production. As noted earlier, the evaluations are performed by the Experts Advisory Cell (EAC), a semiautonomous agency attached to the ministry. The EAC has a reputation for objectivity and for having the respect of the minister, hence, its evaluations are taken seriously by the enterprises managers. The EAC has been able to attract skilled professionals because its salaries can exceed the limits in the civil service. The fact that the EAC is funded by a levy on the enterprises adds to its independence and the seriousness with which it is treated. Holding Management Accountable 53 In Korea, the Public Enterprise Evaluation Bureau works closely with a research institute to develop the performance indicators. At the end of each annual review period, the state enterprises that are part of the system submit their own evaluations of their performance. Separate task forces- created for each enterprise with representatives from private business, the state enterprises, and the research institutes-review the documents and consult with their respective enterprises before doing their own evaluations, which they submit to a council of relevant ministers and commissioners from the private sector. The high-level political support for the Korean system, the involvement of an outside research institute, and the opportunity for management to do its own evaluation, have all contributed to the widespread acceptance of this system among the Korean enterprises. Also noteworthy is the attention in Korea and Pakistan to the training of government officials and enterprise managers in the rationale and workings of the system (Park 1986; Jones 1989; Shirley 1989). The final evaluation must take into account the difference between the performance of the enterprise and that of management. If, for example, the government did not allow management to lay off superfluous workers or to raise prices, the manager should not be penalized for not meeting a profitability target. In Pakistan these factors are brought out during a review meeting between government and management to discuss the draft evaluation; in Korea they are highlighted because the enterprise does its own evaluation. More broadly, the evaluating agency could produce an overview report on the sector as a whole, pointing out performance shortfalls caused by common factors and suggesting remedies. The Pakistani and Korean systems have not been used much for this purpose, but that could change as the systems become established. Much simpler systems could also serve this purpose-for example, the Tanzania Audit Corporation's annual report on enterprises, which has consistently been a hard-hitting, critical, but constructive document. Follow-up is critical if the evaluation is to be more than an academic exercise. In Pakistan, the performance ranking of an enterprise determines the bonuses awarded the managers of that enterprise; in Korea, the rankings determine the bonuses of the work force as well as those of the managers. Each performance indicator is weighted to yield a composite indicator of achievement for a firm. All the firms are then ranked in categories according to their composite indicator, and a bonus of a certain S4 Public Enterprise Reform: The Lessons of Experience number of months' pay is awarded in each category. As a motivating factor, the bonuses have been important in Pakistan; they have been less so in Korea, where the recognition and honor awarded to good performance are at least as important as the bonus. References Jones, Leroy. 1989. "Towards a Performance Evaluation Methodology for Public Enterprise with Special Reference to Pakistan." In PPR Working Paper Series 160. Washington, D.C.: World Bank, Country Economics Department, Public Sector Management and Private Sector Development Division. Park, Young. 1986. A System for Evaluating the Performance of Government-Invested Enterprises in the Republic of Korea. World Bank Discussion Paper 3. Washington, D.C.: World Bank. Shirley, Mary. March 1989. "Pakistan: Assessment of the Industrial Public Enterprise Performance Evaluation System." In PPR Working Paper Series 160. Washington, D.C.: World Bank, Country Economics Department, Public Sector Management and Private Sector Development Division. Song, Dae Hee. November 1988. "Korea Public Enterprise Performance Evaluation System." Seoul: Korea Development Institute Tybout, James. July 1984. "The Algebra of Inflation Accounting." Discussion Paper No. 84. Washington, D.C.: World Bank. Veneroso, Frank, and Martha de Melo. June 1986. "Inflation Accounting and its Policy Implications for Portugal's Public Enterprises." Washington, D.C.: World Bank. 5 DIVESTITURE OF STATE-OWNED ENTERPRISES As the sense grew in many developing countries that the state sector had become too large and costly and that peripheral endeavors were diverting public money and managers away from priority activities of government, many states increasingly looked to divestiture to reduce their fiscal and credit pressures and their management burdens. Governments also see divestiture as a way to increase competition in the economy and thereby develop a private business sector that is more flexible, more responsive to consumers, and more efficient than the public alternatives. The hope in the latter regard is that divestiture will rid the enterprises of the managerial deficiencies common in state-owned enterprises and produce a more entrepreneurial management, one that will seize opportunities rather than simply obey regulations. Moreover, financial transactions between government and private firms are expected to be more explicit, more public or "transparent," and hence, more carefully considered. Divestiture will not solve all problems, but enough experience has accumulated to suggest that it is often worthwhile, and that ways exist to make it work best. An Overview of Experience Despite the widespread interest in divestiture, only a few developing countries have made substantial sales of state enterprise assets and equity. Among them are Chile, which has sold or transferred about 470 enterprises since 1973; Mexico, which has sold, liquidated, merged, or transferred some 700 enterprises, equivalent to about 20 percent of state assets; Jamaica, which has sold or leased 40 firms; Togo and Guinea, which each sold 15 firms and closed several more; Tunisia, which sold 40; and the Philippines, which sold 29. A number of other countries are currently joining this list of leaders. However, except for the Chilean enterprises, 55 56 Public Enterprise Reform: The Lessons of Experience most of the companies sold to date have been small and of low value, and most of the liquidated firms were not operating. Privatized enterprises have most often been manufacturing or service firms that originated in the private sector and were later nationalized. Formal liquidations have been few; more common is mothballing, usually informal-instead of shutting down the enterprise; selling off its assets, and legally ending its existence, governments tend to close the firm, cut off most or all subsidies and new credit, and gradually reduce the staff through attrition. Sometimes the company is allowed to meet its payroll through the sale of inventories or through peripheral activities; for example, workers at a closed sugar mill in Panama used the company's trucks to sell transportation services (Berg and Shirley 1987; Nellis and Kikeri 1989; Galal 1989). The overarching criterion for judging the ultimate success of privatization efforts, as we argued in Chapters 1 and 2, is not the sale or transaction itself, nor the price paid to the government, nor even the survival or expansion of the enterprise. Rather the test is whether the transaction yields a net benefit to the economy as a whole. These benefits are by no means assured-privatization is a relatively new and experimental activity for most developing countries and they may not have all the necessary conditions for success. Because privatization is a relatively recent phenomenon, a track record of performance is only now beginning to accumulate and to date has not been studied with any rigor (Galal 1989). Many anecdotes are encouraging. In Jamaica, for example, the divestiture of the telephone holding company led to a 55 percent increase in outgoing international calls; in Mexico, the divestiture of the state-owned auto parts industry led to greater use of its technological capacity, produced overall cost reductions, and better management. Other anecdotes are discouraging. Enterprises have been closed without legal authorization and their assets sold without competitive bidding. Purchasers have lacked the necessary managerial capacity, experience, or technical expertise, and worse, some have acquired special privileges-in one country, for example, a cigarette manufacturer received heavy protection, including confiscatory taxes on competing production and a monopoly on both production and imports. Other disappointments have been tax exemptions, subsidized loans, and lax enforcement of terms; in Divestiture of State-Owned Enterprises 57 such cases the efficiency gains from the transaction are likely to be far less than when the privatized enterprise is forced to compete and pay the opportunity cost for capital. As a procedural matter, divestiture of specific enterprises can be handled through several channels other than sale of majority ownership: liquidation; formal or informal mothballing, in which the enterprise suspends operations but retains a legal and economic life; partial sale; and privatization of management through leases and management contracts. Beyond the divestiture of individual firms, privatization in a broader sense can include the general reassignment of property rights from the state to the individual (as in Chinese agriculture); contracting out the delivery of public services to the private sector (garbage collection and cleaning of public buildings are common examples); and cutbacks in state activities to allow room for private initiatives. The Lessons of Experience First, divestiture should be viewed, not as an end in itself, but as part of a broader program of reforms designed to promote a better allocation of resources, encourage competition, foster a supportive environment for entrepreneurial development, and develop the capital market. Divestitures that follow or parallel macroeconomic and institutional reforms have shown very much better outcomes than operations undertaken in isolation from such reforms. Although this view may seem obvious, the pace of privatization has not always been well orchestrated with the removal of distortions and the development of a supportive institutional, managerial, and financial enviromnent. Togo, for example, recently in the forefront of privatizers, was sufficiently worried about the pace and terms of privatization that it announced a pause in divestitures to study the problem. Similarly, an emphasis on speedy privatization in Chile from 1974 to 1982 led the government to sell 232 state enterprises, in addition to 240 that were returned to their previous owners, for little equity to groups that also controlled much of the financial system. These divestitures increased the concentration of economic power by strengthening the banks' control of industrial enterprises, which in turn severely distorted the banks' lending decisions. These effects caused the government to take back many of the enterprises and resell them later when the economy, especially the capital market, was better able to absorb the transactions. 58 Public Enterprise Reform: The Lessons of Experience Second, unless the government considers the trade-offs between objectives before taking action, it can make costly errors. Many sales programs have focused on the short-run revenue effects, but long-run net fiscal benefits are by no means assured; if, say, the sales price of a profitable firm plus the present value of future tax revenues are less than the present value of the net future income stream under state ownership, the sale worsens the government's financial position. Whether the economy is worse off depends on how the government spends its money. But while it eliminates a fiscal drain on the government, the sale or lease of an unprofitable firm can leave the economy worse off if extensive concessions have to be extended to the buyer. For example, short-run revenue gains may be maximized by selling monopolies, but the failure to exploit opportunities for competition-such as trade reform, breaking up monopolies, and ending special privileges-can impose considerable long- run costs on the economy; not only must downstream firms and final consumers pay higher prices if a monopoly is kept in place, but the absence of competitive pressure allows new technologies and markets to be neglected. These circumstances suggest that the selection of a particular method of divestiture-liquidation, contracting for activities previously handled within government, leases and management contracts, sales through private placement, competitive bidding, and share issues-should be conditioned on the government's objectives, the condition of the enterprises, and the circumstances of the country (Vuylsteke 1988). Third, weak capital markets make it imperative that the financial prerequisites for divestiture be assessed carefully and be included in the design of sales; otherwise, anomalies can result. The Philippine government, for example, at one time competed with its own sales by offering Treasury bills with a 25 percent real return. Many governments have reduced the market for their assets by limiting or excluding participation by foreigners or even local minorities. Selling shares in state enterprises may be a way to promote the stock market and bring in new savers, but unless the sophistication of market information, analysis, and regulation increases apace, the risk to investors, and in turn to the viability of the market, will be great. Furthermore, spreading share ownership among a large number of small investors will not assure the strong guidance needed to turn a firm around. The purchasers of stock in a state enterprise are likely to assume that the Divestiture of State-Owned Enterprises 59 government is providing an implicit guarantee of their share value, and indeed the govenmment may feel obliged to intervene if the value falls. The government of British Columbia transferred a group of enterprises to the public in the province by fixing a value and giving all citizens an equal number of shares; they also sold shares to raise additional capital. Although the share transfers were quite popular, the capital raised through the sales was misinvested by the enterprise, which soured many on the experience. One possible, if risky, way to use shares to improve the incentive for a turnaround is to reserve a substantial block of shares that might initially be the only voting shares, an approach similar to that taken by the French when they issued "core shares" of divested enterprises to groups selected for their expertise in the relevant industries. Private placements, on the other hand, can also create a variety of problems: they can be more politically visible, harder to finance, and greatly increase economic concentration, as did the first round of sales in Chile. Some governments have entirely financed these placements by taking back notes for the purchase money; the buyers, having no equity in the enterprise, have little incentive to work hard at making the firm profitable, and indeed such deals have left the government with massive defaults by the buyers. Fourth, designing a strategy for divestiture and classifying state enterprises according to the type of action they should receive- liquidation, sale, lease, and so on-have been useful steps in clarifying the government's objectives and approach. These preliminaries come with a trade-off, however; they require a good deal of expensive advice and they slow down the process by making agreement among the many actors harder to achieve. Governments must be aware that their attempts to rationalize and regulate privatization can and often do bring the process to a halt. Experience shows that outside experts are best used for specific technical tasks in law, finance, and accounting. Advisors with broader mandates and part of whose fees are paid upon sales have tended to ignore the public policy issues and press for speedy sales, a problem encountered in Canada, for example, where outside experts eager for sales made overly optimistic price projections. Such advisors may be essential for the sale of large, complex companies, but they have been used excessively for smaller firms in some developing countries. 60 Public Enterprise Reform: The Lessons of Experience In some countries, the potentially marketable state enterprises can be identified relatively quickly without lengthy formal classifications. In other cases, studies and schemes merely offer a convenient excuse to postpone action. Another risk posed by formal schemes is that they may encourage authorities to take nonviable enterprises out of mothballs in a vain attempt to sell them. Fifth, privatization requires special administrators. Managing a privatization program is a complex matter, and government officials seldom have the needed skills. Moreover, the government may be in a weak bargaining position: publicly committed to privatization, burdened by unattractive enterprises, and short on the information and experience required to make good deals. A common outcome under such circumstances is that assets are undervalued and offered with unduly favorable financing. Alternatives to sales-leases and management contracts-are more acceptable politically but they may be hard to manage: costs are not easily tracked or anticipated, and once again, the absence of an equity stake reduces the commitment of the new private management. To overcome these weaknesses, some sort of central administrative unit must oversee the privatization process and keep decisionmakers informed. Technical ministries are typically reluctant to see "their" subordinate enterprises stripped away. With a central unit, responsibility for privatization can be vested in a group with an interest in its success. Such a body can do the analytical work necessary to decide between liquidation and divestiture, make economically rational choices about the terms of the sale, negotiate the subsequent privileges for buyers, and so on. Examples of such special offices are the Bangladesh Divestment Board, created to decide on sales, and its associated working group, which values assets and recommends prices to the board; and the Joint National Investment Commission of Jamaica, assisted by a Divestment Secretariat. Such a group need not, indeed should not, be large. In Mexico, only a handful of people have been responsible for the sale of 250 enterprises; Tunisia's Technical Commission for Restructuring Public Enterprises is composed of 10 professional staff members. Sixth, the immediately visible social costs of privatization can be severe in the short run, whereas the growth of benefits and increases in employment and investment may not appear until later. To demonstrate awareness of these costs, which cannot often be met with the proceeds of Divestiture of State-Owned Enterprises 61 sale, and to offer hope of an orderly transition, the divestiture programs should include a realistic assessment of interim dislocations and a feasible financing plan for layoffs, the settling of arrears, the rehabilitation of firms before sale, and the like. The Tunisian and Philippine divestiture programs, for example, have been able to defuse potential worker opposition through generous severance arrangements and careful communication with labor. Contingent liabilities can prove costly; the severance packages in Ghana have been generous enough to give the government considerable difficulty in financing them. Privatization tends to be associated with layoffs. Unions in Thailand, which are much stronger in the state enterprises than in the private sector, have made opposition to sales-and even to management contracts and leases, where no layoffs were involved-a rallying cry. Severance packages for employees have been important for two reasons: few countries have a social safety net for laid-off workers (the Minimum Employment Program in Chile was an exception); and retraining and redeployment schemes have a poor record of placing workers, partly because divestiture has tended to occur during economic contractions. But layoffs are likely to be part of any effort to upgrade public enterprises, whether or not they are privatized. The actual experience with layoffs can be instructive: Britain shed redundant workers before selling firms, but average employment in the firms since privatization has risen; the blue- collar work force in one set of privatized Mexican firms declined 5 percent, whereas the employment of managers fell 24 percent. Argentina is planning to leave layoffs to the new owners of privatized firms in the hope they will be able to employ them elsewhere. A final and related issue is openmess and public confidence. Experience shows that the merits of divestiture will be debated regardless of whether information is hoarded or released. The challenge is to make the debate an informed one. The public should know more, for example, about how much implicit and explicit subsidizing goes on with state ownership and about the costs, including forgone activities, of such ownership. In New Zealand, laying out to the public the costs of weak performance by state enterprises persuaded the electorate that privatization was necessary. The fact that transition costs will have to be paid regardless of ownership should be clarified; and if excess labor has to be laid off, that, too, should be declared. But, in general, the divestiture debate in most countries is too 62 Public Enterprise Reform: The Lessons of Experience much about costs-unemployment, increased imports, liquidating some seemingly modem enterprises-and too little about benefits, which have to be added to the dialogue if acceptance is to spread. References Berg, Elliot, and Mary Shirley. 1987. Divestiture in Developing Countries. World Bank Discussion Paper 11. Washington, D.C.: World Bank. Galal, Ahmed. May 1989. Ex-post Performance of Divested State-Owned Enterprises. World Bank research proposal. Nellis, John, and Sunita Kikeri. 1989. "Public Enterprise Reform: Privatization and the World Bank." World Development 17(5):659-72. Vuylsteke, Charles. 1988. Techniques of Privatization of State-Owned Enterprise. World Bank Technical Paper 88, Vol. 1. Washington, D.C.: World Bank. 6 SEQUENCING REFORMS The reform of the state-enterprise sector must be handled in stages if it is not to self-destruct. Experiences over the last decade offer some lessons on such staging, or sequencing. A main lesson is that a high priority must be accorded the legislative process; governments often agree to economic reforms in principle while hesitating to take the more visible step of advocating and passing the required enabling laws. Such reluctance, especially in the early stages, could throw an otherwise carefully structured reform program into prolonged limbo. The degree to which legal changes are a potential stumbling block varies from country to country. Francophone African countries, in particular, tend to have strong legal traditions, which they are reluctant to change. With this caution about the fundamental issue of law, the sequencing of the sectoral reforms can proceed as follows. The Near Term The state-owned sector in the average developing country tends to be large and diverse and is increasingly illiquid or on the verge of financial collapse. Although approaches must be tailored to fit individual circumstances, experience has shown that the more successful efforts have begun by eliminating major distortions in prices, interest rates, and exchange rates; halting most new investment; and cutting off most subsidies. These actions clarify the true situation of the enterprises and thereby facilitate decisionmaking. For example, studies of state-owned industrial enterprises in Tanzania and Egypt found a surprising number of firms showing negative value added under intemational prices. Trying to make sensible decisions about rehabilitation or new investment under such circumstances is nearly impossible. The questionable viability of a number of Tanzanian enterprises, for example, became much clearer after moves were made toward market pricing and trade liberalization. 63 64 Public Enterprise Reform: The Lessons of Experience An effective second step has been to identify key enterprises that need rehabilitation and that will likely remain in the state sector for some time. Most countries have no more than 10 or 15 enterprises in this group; typically they include the companies providing electricity, water, and post and telecommunications services, the railroad companies, state financial institutions, and some mining or petroleum firms that cannot be privatized easily. The rehabilitation effort should focus on meeting the emergency needs of critical firms such as these, which provide services essential to the rest of the economy. The remaining enterprises should then begin to fend for themselves as the government removes special privileges, ends state monopolies, encourages competition with domestic and foreign firms, permits market pricing, and gives managers the flexibility to react to market signals. Enterprises that cannot survive this regime should be liquidated or at least be permitted to wither away. Some of the survivors can begin to be privatized, particularly smaller industrial enterprises. Privatization at this stage can help add to the competitive pressures, increase the space for private initiative, eliminate a fiscal and managerial burden on the state, and add to the credibility of the reforms. The Medium Term The process outlined here is not a neat one, and discontinuities will arise along the way, especially as the process enters the subsequent phases. After the government initially sorts out the enterprises, it will have to deal with the more pressing medium-term reforms. The institutional framework will have to change along the lines described in Chapter 4, an extended process that should start with the creation or strengthening of some central oversight group to monitor the process and advise on the rules of the game. Such oversight should focus first on assuring that resources flow only to priority enterprises and that management of these enterprises begins to improve. Reforms in the financial system will also become pressing: putting state and private enterprises on an equal and competitive footing, assuring an arm's length relationship between banks and the state, and creating an effective mechanism for the efficient allocation of resources. As market pressures build, some of the previously implicit decisions must become explicit and transparent. That is, the government will need to Sequencing Reforms 65 decide whether more of the enterprises outside the core will be sold, liquidated, or rehabilitated. Now that price distortions are fewer, the costs and benefits of rehabilitation and continued public operation can be determined more easily. Some viable enterprises that cannot compete because public operation has left them decapitalized or without proper management may need financial and technical assistance, which could come from the public or the private sectors. Again, this sequence is by no means a blueprint. In some countries the process begins with divestiture because the government is ready and sees a window of opportunity. In contrast, institutional reforms may have first priority where the policy framework is sound and the state enterprise sector is small. In any case, simplistic blueprints have proven counterproductive, and efforts to resolve the government's administrative weakness with large amounts of technical assistance have seldom succeeded. Momentum A combination of circumstances has created a favorable environment for the reform of the state enterprise sector: the lesson of such reforms in developed countries and the successes of market-oriented strategies among developing countries; the acute financial crisis, which has raised the economic and political costs of doing nothing; and the installation of a number of new, reform-minded governments. Yet the resulting movement toward reform is inherently temporary; the momentum must be sustained through actions such as the following (Heaver and Israel 1986): - Design a reform package that is perceived as equitable. For example, layoffs that hit all sectors typically raise fewer objections than those more narrowly targeted. Enlist those affected by the early reforms in assisting with the push for continuation. * Try to balance losses with gains. For example, farmers will be less opposed to paying cost-recovery rates for rail freight if trade reform has lowered the cost of imported inputs. * Build commitment to reform through workshops and seminars; keep affected groups informed and consult them frequently, but give no group special treatment. Show the costs of standing still; for leaders, make the status quo difficult to defend and advantageous to repudiate. 66 Public Enterprise Reform: The Lessons of Experience * Balance the long-term nature of institutional reform with the politician's short-term agenda by building in quick payoffs and concentrating on actions where prospects for success are high. For example, a program could begin with the emergency rehabilitation of a few viable state-owned enterprises, with longer-term restructuring to come later. Thus, in the Congo, where reform has focused on ending state enterprise monopolies, the quick response of private producers, in pharmaceuticals, for example, has helped build support for the program. * Avoid perfectionism and excessive comprehensiveness; it leads to a lengthy start up and a slow, contentious course of implementation. References Heaver, Richard and Arturo Israel. 1986. Country Commitment to Development Projects. World Bank Discussion Paper 4. Washington, D.C.: World Bank. 7 SUMMARY The world is witnessing a historic shift away from rigid, formulaic approaches to economic organization. In developing countries, the era of reliance on the public enterprise sector to drive the economy to prosperity has largely come to an end, with state firms having often proven to be financial burdens rather than creators of wealth. We have examined the weaknesses of the public sector in developing countries and described strategies toward its reform. The approach includes liquidation and privatization of state firms; but while the old preference for public over private may have ended in much of the developing world, public enterprises will still have a central role there even as the sector itself is transformed. Where they are already state-owned, core enterprises such as ports, utilities, and resource extraction companies will likely remain public for the foreseeable future. Their fate and the fate of the wider national economy in which they operate are linked, and the principles of competition and efficiency that underlie the prescriptions in this book are central to their prosperity. To summarize the main points: Policy * No amount of reform of managerial procedures or institutional arrangements can substitute for liberalization in trade policy and the lifting of controls on exchange rates, interest rates, and other basic prices. It is essential to "get the prices right." * Ownership matters. In theory, ownership may not affect efficiency, but in practice it almost always does-because governments tend to tilt the economic/financial playing field toward "their" firms in ways that are difficult to perceive and harder to correct. Thus, the telling argument for privatization is that it enhances competition. Privatization also preserves the gains laboriously achieved under, 67 68 Public Enterprise Reform: The Lessons of Experience but continuously threatened by, public ownership. The lesson of experience is that governments cannot resist interfering in public enterprises, regardless of the barriers they may erect to prevent such tampering. * The financial sector has assumed a central role in the affairs of public enterprises, which find they can no longer rely on government to subsidize their operations. An independent and competitive banking system can hold public enterprises to strict standards of performance and help impose financial discipline on their managers. * Cost-plus pricing for public enterprise goods and services is a subsidy that decreases efficiency. Even when marginal-cost pricing is not feasible, efficiency targets can be constructed that distinguish reasonable price increases from those that pass the costs of inefficiency on to the consumer. * Many public enterprises pay managers too little and their subordinates too much. Remedies must include incentives for managers to control costs, including wages. Retrenchment is a sensitive, indeed potentially explosive issue; all governments that have succeeded in downsizing labor forces have paved the way by devoting special attention and resources to the construction of a social safety net woven of severance packages, training, credit schemes, and resettlement programs. Role, Scope, and Objectives * Efforts to clarify and simplify the objectives of the state enterprise are well worth the investment (which can be considerable). Such efforts are not academic; few developing countries can afford to prop up indefinitely an enterprise that persistently fails to achieve its objectives or that is no longer useful. * The assignment of noncommercial objectives to state enterprises has such an insidious effect on performance that the burden of justifying their imposition should be placed on those who advocate them. * In developing countries, the use of performance contracts-to negotiate objectives and specify the rules of the game between government and enterprise-has not yet dramatically improved Summary 69 performance. But the underlying idea is simple, appealing, and essential: agree on what the enterprise is to do and on what it needs to do it. Decentralized Management * Countries that have raised the efficiency of their state enterprise sectors-Chile and Korea among them-have done so by increasing the responsibilities of enterprise managers while shifting government supervision from control of financial transactions to the evaluation and stimulation of results. * Performance evaluation, a critical component of success, has to be coupled with management autonomy, which in turn requires a supply of good management skills. * The importance of the institutional setting in which state enterprises operate-including holding companies, boards of directors, and oversight agencies to coordinate government supervision of enterprises-is not trivial but has been oversold. Without fundamental shifts in the location of decisionmaking authority, organizational changes will do little to increase enterprise autonomy and may add to the bureaucratic hurdles management must overcome. * A critical element of successful reform is higher incentives for managers. Bonuses and more decisionmaking authority, clearly linked to performance, have proven useful. The corollary, of course, is that governments must have the means to assess enterprise performance and be ready to dismiss managers who fail. Managerial Accountability * A flawed or partial system of setting enterprise targets and evaluating results is better than no system at all. Financial disasters occur when governments ignore their enterprises. * A proven way to hold public enterprise managers accountable is to isolate the factors that are within their control and then let them know that they will be judged according to changes in these variables. 70 Public Enterprise Reform: The Lessons of Experience Performance evaluation systems work because they provide management with a clear statement of the owner's objectives and government with a clear picture of performance. Without such clarity, managers cannot determine whether any proposed investment represents the best use of resources. * Performance evaluation is not a substitute for an assessment of the economic and financial viability of the enterprise as a whole or for cost-benefit assessments of new investments in the enterprise. Divestiture * When correctly conceived and executed, divestiture increases enterprise efficiency; rids the state of administrative, managerial, and financial burdens; increases competition in the product market; and widens and deepens financial markets. X Divestiture is more likely to live up to its potential when it is part of a broader program of macroeconomic reforms designed to remove price distortions, increase competition, and develop the private sector. e Divestiture can go wrong; particularly in developing countries, the factors that cause public sector inefficiency also impinge-though less severely-on private owners. * Divestiture is not an end in itself. Rather, it is one arrow available to the policymaker aiming at increased efficiency and productivity. - Because the numerous goals usually advanced for divestiture conflict with one another, the trade-offs must be recognized and objectives consistently ranked. * Divestiture requires special skills that govemments lack. Experience shows that outside expertise is indispensable-and painfully expensive. * External assistance can provide support for the hard decisions, but the divestiture process is inescapably political; the political authorities, not consultants or technicians, must ultimately decide on the purchasers, prices, and structure for such transactions. * Special organizations-technical secretariats on divestiture-have proven useful in many countries. These agencies devise strategies, Summary 71 identify candidates, facilitate the myriad parts of the inevitably tangled process, and evaluate results. They allow the political authorities to concentrate their energies on the big questions and build public confidence by helping to assure an open, transparent process. * The visible social costs of privatization must be directly addressed through severance pay and the elaboration of social safety nets. Sequencing * The reform of the state enterprise sector is daunting because it touches simultaneously on so many facets of the economy and society. Yet not all state enterprises need to be dealt with at once, and some not at all. Some can be left to fend for themselves without subsidy or support. Those that cannot survive should be allowed to go bankrupt, a painful outcome that is increasingly tolerated. * Government should concentrate its efforts on the larger, more financially significant enterprises, where the benefits from performance improvement measures-including privatization- justify the costs. 8 PUBLIC ENTERPRISE REFORM IN SOCIALIST EUROPE This study on enterprise reform in developing countries was written as the countries of Eastern Europe began their historic shift away from socialism. Hence the question arises more forcefully now than it might have earlier: How much of the foregoing analysis applies to socialist countries? On the face of it, the general problem is the same; that is, among developing countries and socialist countries alike, property belonging to the state, to society at large, is too often treated as though it belongs to no one and has thus suffered from neglect, misuse, and waste. The problem has been as severe in Peru as it has been in Poland. The significant difference is that, in developing countries, the state- owned sector covers only a part of the economy, albeit a large part-it accounts on average for between 15 and 20 percent of GDP. The economies of all but a few developing countries contain a nonstate sector, which normally accounts for much more than half of domestic output and operates with much less supervision and control than does the state sector. By contrast, the state sectors of the socialist economies of Europe are in effect the totality of their economies, making up the bulk-between 65 and 90 percent-of aggregate economic activity and the overwhelming mass of nonagricultural employment. And because of their primary position in the distribution of housing, health, and leisure services, they play central noneconomic roles as well. In these circumstances, the reform of state enterprises is not a matter of incremental adjustments to elicit higher levels of efficiency; it represents the fundamental transformation of the entire society. This major difference produces at least two results. First, and obviously, it makes the enterprise reform process exponentially larger, more complex, and important in socialist economies than in other countries. Second, it negates or calls into question at least one of the 72 Public Enterprise Reform in Socialist Europe 73 lessons of experience derived from capitalist and mixed economies: that privatization is but a means to an end. To illustrate the former effect, one can contrast the average of 75 public enterprises in the countries of sub- Saharan Africa (a region of heavy state intervention), to the 8,700 public enterprises with which Poland begins its transition process. Or one can note that while many developing countries administer some key prices, most socialist countries administer all prices. As for the second result, the fact is that the leaders and advisors of many socialist countries of Europe do not regard privatization and the creation of private property rights as a means toward the end of greater efficiency; they see it as tantamount to the destruction of the command system and hence as an end in itself. Many in these countries believe that if they were to wait until the conditions for obtaining greater efficiency from each privatization were clearly propitious, years would pass before any substantial portion of assets was in private hands. The economic argument is that one must trade the possibly high costs of suboptimal privatization for the definitely high costs of the present operations of state enterprises. Moreover, on the political front, most reformers feel that they-and their suffering and demanding electorates-cannot tolerate a long wait. If the state sector remains dominant, it constitutes a base of operations for those who may want to restore the system of administered security. Thus, assert the officials in charge of the Czechoslovak privatization program, It is axiomatic that there can be no market without private ownership; privatization is thus conceived of as a worthy goal in and of itself. In the highly improbable case that privatization engenders- in the short run-declines in growth and efficiency it would remain the government's policy to privatize (Triska and Jelinek-Francis 1990; Lipton and Sachs 1990). David Lipton and Jeffrey Sachs, at the 1990 Ljubljana,Yugoslavia, Conference on Privatization in East Europe, argue that "Even though we favor rapid privatization we doubt that privatization will produce immediate large increases in managerial efficiency or enterprise productivity... Nonetheless, we believe that.. .it is necessary to proceed rapidly and comprehensively on creating a privately owned, corporate-based economy in Eastern Europe... .the real risk in Eastern Europe is not that privatization will be less than optimal, but that it will be paralyzed entirely. We believe 74 Public Enterprise Reform: The Lessons of Experience that unless hundreds of large firms are quickly brought into the privatization process in each country, the political battle over privatization will soon lead to stalemate in the entire process, with the devastating long- term result that little privatization takes place at all." However, like the developing countries, the European socialist countries are finding that privatization takes far longer than initially anticipated. Judging from the intemational experience and the particular obstacles to rapid privatization within this set of countries, we conclude that socialist countries will have great difficulty quickly seling off the bulk of their state-owned enterprises; indeed, it will prove difficult even to give them away. Preliminary experience bears this out: few large enterprises have been privatized to date, even in the most actively reforming socialist countries of Yugoslavia, Poland, Czechoslovakia, and Hungary. Only Hungary has had more than a few "spontaneous privatizations" (that is, management buyouts) of large firms. Between 65 and 70 closely supervised buyouts have recently been concluded in Hungary, and the authorities hope to increase the scope and speed of the supervised buyout method; 20 of their large firms are about to be placed on the open market, and another group of large firms is being selected for trade sales. All this activity is encouraging, but even in Hungary the number of firms in the sector-2,200-has not yet been seriously reduced. Elsewhere, the difficulties seem more severe. Polish authorities are struggling to develop a privatization approach acceptable to the contending interests in the country. In Czechoslovakia, the decision to restore to previous owners or their direct descendants the property taken from them through Communist nationalizations is complicating matters. And in Yugoslavia, the problems involved in reconciling the imputed property rights of workers with the rights of the provider of capital, the state, are proving intractable. Throughout the region, one finds intense debates over the desirability of foreign ownership. This ferment delays the conception and implementation of privatization plans. Moreover, even if a political consensus emerges on a strategy for divestiture, the number of interested buyers and managers may well be inadequate to absorb the thousands of filns that will shortly be coming to the market Privatizations of small firms-retail shops, restaurants, and so on- have moved more rapidly. Poland has reportedly already divested several Public Enterprise Reform in Socialist Europe 75 thousand small entities, and Czechoslovakia and Hungary are embarking on the process. The former East Germany is a special case; there the question is less one of transition to the market than of absorption into a flourishing market economy. As of mid-November 1990, reports put the number of negotiated privatization transactions in East Germany at 200, of which 60 had already become "legally effective." Still, even with the extraordinary capital and expertise available, the divestiture process is proving immensely difficult. These problems heighten the importance of the question with which we opened this discussion of Europe: Are the strategies for enterprise reform described in the preceding chapters-short of ownership change- applicable to socialist countries? We believe the answer is yes. The strategies, which aim at the "corporatization" of state enterprises, address the problems of unclear and ambiguous objectives, interference by politicians in the day-to-day management of the firm, and soft budget constraints-problems that beset state enterprises in the socialist countries of Europe as well as in developing countries. Indeed, our assessment is that, not only is this package of reforms appropriate, but that eventually it will be applied in the European socialist countries. Why eventually? Why not at once? Two obstacles bar the way to adoption of the corporatization package. The first is that, paradoxically, the package requires a reassertion of state authority. That is, for the state either to sell or to demand performance improvements in an enterprise, the state must be the unquestioned owner, and for a variety of tangled reasons, that is not now the case. The simplest way to clarify ownership is, in effect, to renationalize the enterprises and declare them to be joint stock companies with their shares held by the government. But this approach is rejected by a broad spectrum of reformers, who remember well the many failed attempts at partial reform and fear that another effort would result in the re- emergence of state control. One must not underestimate the pervasive fear and mistrust of the state that exists in Eastern Europe. This aversion is a prime difference between the reforming socialist countries and the developing countries. The European reformers therefore seek ways to proceed directly to private ownership, without passing through an intermediate stage of reimposed state control. This in turn leads them to downplay and in some cases reject partial reform packages. 76 Public Enterprise Reform: The Lessons of Experience The second and closely related difficulty is that the package cannot be applied until the prior, constraining legal system is swept away and replaced by one providing individual property rights. This difficulty is more than a question of legislative drafting; it involves the resolution of elemental political issues and hence will require considerable time and effort. These are serious obstacles; indeed, their daunting nature is one of the reasons that reformers have opted for instant privatization. Our conclusion is that the unusual circumstances of the post-Communist societies justify the efforts to privatize for the sake of privatization, and these activities should be vigorously supported. However, because the obstacles to the success of this policy are many and formidable, these countries should- and, eventually, will be forced to-embark on the corporatization reforms we have outlined. References Lipton, David, and Jeffrey Sachs. 1990. Privatization in Eastem Europe: The Case of Poland. Paper presented to the Conference on Privatization in East Europe, November 1990, in Ljubljana, Yugoslavia. Pohl, Gerhard. 1990. Privatization and Restructuring in East Germany: A First Progress Report. Geneva: World Bank. Triska, Dusan, and Charles Jelinkey-Francis. 1990. A Study of Privatization in the Czech and Slovak Federal Republic. Paper presented to the Conference on Privatization in East Europe, November 1990, in Ljubljana, Yugoslavia. Appendix A CHECKLIST OF TOPICS FOR DIAGNOSTIC OVERVIEW I. Performance A. Financial. Profitability, debt and liquidity, arrears B. Macroeconomic impact. Production, value added, exports, imports, employment, wages C. Fiscal impact. Contribution to tax revenues; subsidies and other curent transfers; equity, grants, loans and other capital transfers; dividend payments; share of public intemal and extemal debt 1I. Causes of poor performance A. Mismanagement (lax cost control, poor plant management, inadequate maintenance) B. Inadequate capital structure C. Poor investment decisions (over- or under-sized plant, mismatched equipment, insufficient market, heavily import dependent) D. Government interference in internal operating decisions; bureaucratic delays and red tape E. Selection of inappropriate managers and directors, frequent turnover of managers; lack of managerial incentives and accoutability; insufficient compensation F. Changes in market; adverse trends in relative prices; exhaustion of raw materials; civil war, cutoffs in essential services (power, water, transport, fuel,) or supply of inputs III. Past reform and reasons for success or failure IV. Strategic issues of state ownership A. Size and role of state-owned sector; objectives of government involvement; how objectives and role of sector may have changed 77 78 Public Enterprise Reform: The Lessons of Experience over time: efficacy of using state-owned-enterprises to achieve various objectives (low SOE prices versus target income subsidy, or nationalization of a sector versus regulation) B. Classification of enterprises by nature (commercial or noncommercial); market (monopoly versus competitive); dependence on Treasury support C. Principles of classification of SOEs to be closed, rehabilitated, or privatized in whole or in part D. Division of labor with private sector: removal of discrimination and barriers to competition between public and private enterprises V. Macro policy issues affecting SOEs A. Pricing B. Labor C. Finance and Investment D. Trade VI. Institutional framework A. Present set-up 1. Main actors and responsibilities, formal and informal (role of ministries, boards, holding companies, management, parliament, others) 2. Procedures for making major decisions, including pricing, investment personnel 3. Reporting relationships (flow of information) B. Improving the institutional framework 1. Setting clearer objectives and targets 2. Better division of responsibilities and coordination 3. Decentralization, greater managerial autonomy where appropriate 4. Improving accountability VII. Managerial capacity A. Selection, rotation, firing B. Compensation C. Incentives D. Training Appendix A 79 VIII. Medium-term reform program A. Policy reform B. Institutional changes C. Divestiture D. Financial restructuring and rehabilitation E. Upgrading managerial capacity F. Intemal management systems IX. Sequencing reforms and selecting priority enterprises Appendix B MINIMUM FINANCIAL INFORMATION ASSEMBLED BY GOVERNMENT OVERSIGHT AGENCIES Flow of Funds Analysis Net Profit + Depreciation = Funds from operations + New borrowings-government + New borrowings-other + New paid-in capital - Total funds available Capital expenditures (Repayment of debt to government) (Repayment of other debt) + Total debt repayment Less dividends paid + Other expenditures T Total funds used Net change in working capital Ending networking capital Government Impact Analysis Subventions + Advances 80 Appendix B 81 + New loans + New paid-in capital = Total cash received frm government Debt repayment + Interest paid + Taxes + Dividends + Other = Total cash paid to government Net cash impact on government Income Statement Commnercial revenue + Subventions = Total revenue Direct costs + Other administrative expenses + Depreciation = Total operating expenses Operating profit - Interest - Taxes - Exceptional items = Net profit Balance Sheet Other current assets + Inventory = Total current assets + Net fixed assets 82 Public Enterprise Reform: The Lessons of Experience + Other assets = Total assets Overdrafts + Other current liabilities = Total current liabilities Advances and other + Long-term loans-government + Long-termn loans-others = Total long-term liabilities Net worth Total liabilities and net worth Ratio Analysis Liquidity Net working capital Current ratio Working capitalAtotal assets Quick assets Quick ratio Inventory turnover Debt service coverage Long-tenn debt/equity Government investment/total assets Income Statement Total Assets Commercial revenue Subventions Total revenue Costs of goods sold Other administrative expenses Appendix B 83 Depreciation Total operating expenses Operating profit Interest Taxes Net profit Net profit/net worth Government cash flow/total assets Percentage of Revenue Analysis Nonsubvented revenue Administrative expenses Total operating expenses Interest Net profit Balance Sheet/Total Assets Inventory Other current assets Total current assets Net fixed assets Other assets Total assets = 100% Overdrafts Other current liability Total current liability Advances Long-tern loans-government Long-term loans-other Total long-term liabilities Net worth Source: Adapted from "Malawi: Report on Parastatal Restructuring," Internal World Bank Document. INDEX Accountability: balancing with Boards of directors, 69; productive autonomy and, 26, 33; control of enterprise and, 27; role government and, 26-28; holding of, 37-39; state enterprise budgets companies and, 36; managers and, and, 29 21, 69-70 Bonuses, 41, 46, 49, 51 Accounting standards, 47 Brazil, 22, 26, 32, 33, 35 Africa, 73; mangerial talent and, 40; Budgets of state enterprises, 28-29 monopolies and, 28; performance Bureau of Public Enterprises: in contracts and, 6, 12, 22, 23, 24; India, 26, 33; in Mali, 33 performance evaluation and, 45; Burundi, 34 sequencing of reforms and, 63; tariff study in, 11. See also names of specific African countries Canada, 59 Agricultural commodity boards, 18 Capital, state enterprises and, 9-10 Argentina, 22, 26 Chile, 10, 55, 59, 61, 69 Autonomy, 8, 21; balancing with China, 57 accountability and, 26, 33; Classification of state enterprises, government and, 26-28; managers 19-20, 27-28, 59-60 and,42 Cocoa Marketing Board (Ghana), 13 Commissariat National aux Balance sheet (oversight Comptes (Congo), 43 information), 81-82, 83 Competition, 25; fostering Bangkok Mass Transit Authority, efficiency through, 7-8; holding 35 companies and, 37; noncommercial objectives and, Bangladesh, 22,60 17; oversight and, 27-28; Banking sector, 8; canceling arrears performance evaluation and, 48, to, 9 49; privatization and, 7, 23, 67; Bankruptcy, 8, 71 sale of public firms and, 5 84 Index 85 Conference on Privatization in East 67, 68. See also names of Europe, 73 specific countries Congo, 34, 43, 66 Divestiture: administration Contracts. See Performance considerations and, 60; advantages contracts and disadvantages of, 70-71; Corporate plans, 21-22 classifying enterprises and, 59-60; competition and, 5; experience Corporatization, 75, 76 with (overview), 55-57; financial Cost minimization targets, 49, 51 prerequisites and, 58-59; Credit, 26; state enterprises and, 8-9 Madagascar's classification Czechoslovakia, 73, 74 scheme and, 20; as part of broader program of reform, 57; proper conditions for, 3; public debate Data: oversight financial and, 61-62; revenue effects and, information and, 80-83; 58; social costs of, 60-61, 71; performance evaluation and, 47- Somalia's action plan and, 19. 48 See also Privatization Debts: price control by public enterprises and, 17; public sector, 9 East Germany, 74-75 Deficits: financing of, 18; price Economic reform. See Reform control and, 17; state enterprise, Ecuador, 42 9 Efficiency, 25, 73; competition Department of Statutory Bodies and, 7-8; focus on, 25-26; labor (Malawi), 25 policies and, 12-13; performance Depreciation, 50 monitoring and, 18; pricing for, 10-12, 68; profits and, 17 Developing countries: budgets of state enterprises and, 28-29; core Egypt, 63 group of state enterprises and, 19; Employment agricultural inflation adjustment and, 48; commodity boards and, 18; hiring labor policy and, 12; performance of excess workers and, 17; contracts and, 68; sequencing of privatization and, 61; socialist reforms and, 65; shift toward economies and, 72; state market structure and, 5-6; state- enterprises and, 12-13 owned enterprises and, 2-3, 25, Exchange rate, 9, 67 86 Index Experts Advisory Cell (EAC, Imports: barriers to, 28; Pakistan), 52 competition and, 8 Export licenses, 31 Income statement (oversight Exports, 27; Hindustan Machine information), 81 Tool Study and, 7 Income statement total assets Finance: divestiture and, 58-59; (oversight information), 82-83 govemment control and, 27; India: Bureau of Public Enterprises information and oversight and, in, 26, 33; corporate plans and, 80-83; market-oriented, 8-9; 21; Hindustan Machine Tool reform and, 68; sequencing and, Study and, 7; managers in, 40; 64 performance contracts and, 22 Finance Ministry: budgets and, 29; Inflation, 48 consulting with, 31; importance Institutional changes: approach to, of, 31-32. See also Ministries 42-43; competition and, 7; Flow of funds analysis, 80 diagnostic overview and, 78; France, 59; performance contracts macroeconomic reform and, 6; and, 12, 22, 23, 24 oversight structure and, 29-31; sequencing and, 64, 66; tariff Gambia, 33, 34 setting and, 11 Germany (East), 74-75 Interest rates, 9, 67; adjusting Ghana, 12, 13, 32, 33 profits and, 50 Investment, 18; artificially low Government impact analysis,0 prices and, 17; govenment and Jovernment Invested Enterprises board and, 27; Pakistan and, 29; (GIE, Korea), 39,4647 profitability target and, 28 Great Britain, 61 Italy, 36 Guinea, 55 Hindustan Machine Tool study, 7 Jamaica, 55, 56, 60 Holding companies, 35-37, 69 Joint ventures, 42 Hungary, 74 Jones, Leroy, 27 Import licenses, 31 Kenya, 34 Index 87 Korea, 38-39,40,69; performance and, 21; improving skills of, 39- evaluation in, 45, 46-47, 50, 51, 40; in Malawi, 36; policy 52-53 implementation and, 25; pricing Kornai, Janos, 9 and cost reduction and, I 1; training of, 42-43; wages and, 12, Labor: efficiency and policies for, 13, 68 12-13; hiring of excess workers Market structure: developing and productivity of, 17 countries shift toward, 5-6; Latin America, 25, 48; perfonnance financing and, 8-10 contracts and, 22-23 Mauritania, 33 Legal structure, 13-14 Mexico, 22, 35, 40, 55, 56, 60, 61 Liberalization (trade), 7, 63, 67 Ministries: board of directors and, Lipton, David, 73 37; budgets and finance, 29; Liquidation, 8,67 central organizations and, 32-34; Liquidity (oversight information), consulting finance, 31; oversight 82 and sector, 31; privatization and, 60; role of, 31-32. See also Long-run marginal cost studies, 11 ' o names of specific ministries Ministry of Agricultural Production Madagascar, 20 and Agricultural Reform Malawi, 25, 36 (Madagascar), 20 Mali, 32, 33 Ministry of Finance, 29, 31-32 Management board of directors and, Ministry of Production (Pakistan), 38; decentalized, 69; diagnostic 52 overview and, 78; government Monopoly, 26, 47; budgets and, 29; and decentralized, 26-28; competition and, 7-8; Congo and, privatization and, 60; reform and, 66; evaluation and, 49; natural, 1, 67; reform objectives and 2; price monitoring and, 27; autonomy and, 21. See also pricing and, 10-11, 28; public Performance evaluation enterprise abuses and, 2; Managers: accountability and, 21, technology and, 2 69-70; budgets and, 29; bureaucrats as, 7; compensation analysis and, 40-42; control National Investment Board (NIB, agencies and, 26; corporate plans Gambia), 33 88 Index New Zealand, 37,61 Performance evaluation, 27, 34; Niger, 8, 12 abuse problem and, 41; budget Nigmarker 8, ilur12 2 expenditures and, 29; competition Nonmarket failure, 2 and, 49; data and, 47-48; decentralized management and, 69; diagnositic overview and, 77; 38v budget review and,e2s-29; difficulty of, 45; Korea and, 45, central or coordinating 46-47, 50, S1, 52-53; manager organization and, 32-35; accountability and, 70; oversight decentralized management and, and, 48; Pakistan and, 45-46, 49, 69; financial infornation and, 80- 50, 51, 52-54; prices and, 49; 83; government and autonomy profits and, 50-52; use of targets and accountability and, 26-28; and, 48-49 institutional framework for, 29- Peru, 14, 72 31; managers and, 39-43; Philippines, 26, 55, 61 ministries and, 31-32; Poland, 72, 74 performance evaluation and, 48; Policy: competition and, 7-8; Philippine state enterprises and, finance, 8-10; labor, 12-13; legal 26; trivial decisions and, 25 structure and, 13-14; Ownership: disadvantages of public, macroeconomic reform and, 6, 2; reform and, 67-68; socialist 78; pricing, 10-12; reform and, economies and, 73-76; strategic 67-68; shift to market structure issues of state, 77-78 and, 5-6 Portugal, 9, 27 Prices: agricultural commodity Pakistan, 32; board of directors and, boards and, 18; cost-plus, 68; 37; corporate plans and, 21; monopoly and monitoring, 27; as investment decisions in, 29; noncommercial objective of state managers in, 40-41; performance enterprise, 17; performance evaluation in, 45-46, 49, 50, 51, evaluation and, 49; profitability 52-54 targets and, 28; reform and Performance contracts: advantages getting it right, 67; sequencing of and disadvantages of, 22-24; reforms and, 65; socialist barriers to (West African economies and, 73; Tanzania and, example), 6; French plan for, 12, 63 22, 23, 24; reform and, 68-69 Pricing (for efficiency), 10-12 Index 89 Private sector: mobilizing, 5; wage developing countries and, 2-3, 25, comparison (public sector) and, 67, 68; diagnostic overview 12; checklist and, 77-79; divestiture Privatization, 6, 19, 27; and, 3, 5, 19, 20, 55-62, 70-71; competition and, 7, 23; reform economic justification for, 1; and, 67-68, 71; sequencing of finance and, 8-10; holding reforms and, 64, 65; socialist companies and, 35-37; economies and, 73-76. See also institutional framework and, 29- Divestiture 31; labor policies and, 12-13; Productivity: evaluating profit- legal structure and, 13-14; Prodeiity:and, subveatinpofit- andmanagement skill improvement ability and, 18; subventions and, and, 39-43; monopoly and abuses 17 of, 2; mothballing of, 56; Profitability, 10; argument against noncommercial objectives of, 16- public, 51; calculating, 49; 18; nonmarket failure and, 2; productivity and, 18; target for, performance contracts and, 6, 12, 28 22-24; pricing and, 10-12; Profits: efficiency and, 17; rehabilitation and, 64-65; reform govemment control and, 27; and, 1-2; reform and focus of performance evaluation and, 46, operating efficiency and, 25-26; 48, 49, 50-52; used as bonuses, reform objectives for individual 41 firms and, 20-21; sectoral Public Enterprise Evaluation Bureau development and, 1; socialist (Korea), 53 economies and reform in, 72-76; Public enterprises: balancing twinning and, 42 autonomy and accountability and govemment and, 26-28; basic Reform: boards of directors and, 37- objectives of, 27; boards of 39; budget and, 28-29; central directors and, 37-39; budget issues in, 1-2, 79; central handling and, 28-29; central organization and, 32-35; organization and, 32-35; competition and, 7-8; corporate classification of, 19-20, 27-28, plans and, 21-22; diagnostic 59-60; competition and, 7-8; overview and, 18-20, 77-79; coordinating organizations and, divestiture and, 3, 5, 19, 20, 55- 32-35; core ministries and, 31-32; 62, 70-71; finance and, 8-10; corporate plans and, 21-22; focus on operating efficiency and, corporatization and, 75-76; 25-26; government intervention 90 Index (Korea) and, 46; government's Sector ministries: bypassing of, role and, 26-28; holding 32; oversight and, 31. See also companies and, 35-37; Ministries institutional framework and, 29- Senegal, 22, 23, 24, 32, 48 31; labor and, 12-13; legal Sequencing of reforms, 71, 79; structure and, 13-14; medium term, 64-65; momentum macroeconomic, 6; managerial of, 65-66- near term, 63-64 skills improvement and, 39-43; Sof, 6-6noea , 3-764 ministries (core) and, 31-32; S noncommercial objectives of state Somalia, 8; classification and action enterprises and, 16-18; objectives plan of, 19-20 of, 68; objectives for individual Staff, 25; bloated levels of, 18; firms and, 20-21; organizational central organizations and, 33; changes and, 43-44; performance hiring of excess workers and, 17; contracts and, 6, 12, 22-24; intervention and size of, 33; policy and, 67-68; pricing and, manager's control of, 7; managers 10-12; sequencing and, 63-66, 71; and retrenchment and, 68; socialist economies and, 72-76; privatization and, 61; wage training and, 42 compression and, 12; wage Regulatory agencies, 25. See also expectations and, 13 Oversight State Enterprise Secretariat (Brazil), Resource allocation, 5 33 Resource misallocation, 9 State-enterprise sector. See Public Revenue, 11, 26; divestiture land, enterprises 58. See also Taxes Subsidy, 21, 29, 41; holding Revenue analysis (oversight companies and, 36; information), 83 noncommercial objectives of public enterprises and, 17; parties who benefit from, 18; reform and, 68; state enterprise, 10 Sachs, Jeffrey, 73 Sale of public enterprises. See Tanzania, 42, 53, 63 Divestiture; Privatization Tariffs, price reform and, 10-11 Secretaria de C6ntrole de Empresas Taxes, 8, 29; calculating profits Estatais (SEST, Brazil), 34 and, 50. See slso Revenues Index 91 Technology, natural monopolies Unions, 61 and, 2 Thailand, 12, 27, 41, 61; corporate Wages: labor policy and, 12; plans and, 21; oversight and, 35 managers and, 4042, 68; staff Togo, 55 and, 12,13 Trade, 63, 67; competitions and, 7- World Bank, 7,48 8. See also Exports; Imports Training (managers), 4243 Yugoslavia, 73, 74 Tunisia, 60, 61 Turkey, 11, 14 Zambia, 12, 27, 32, 36 The World Bank 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. Telephone: (202) 477-1234 Facsimile: (202) 477-6391 Telex: WUI 64145 WO}LDBANK RCA 248423 WORLDBK Cable Address: INTBAFRAD WASHINGTONDC 66, avenue d'I6na 75116 Paris, France Telephone: (1) 40.69.30.00 Facsimile: (1) 47.20.19.66 Telex: 842-640651 Kokusai Building 1-1, Marunouchi 3-chome Chiyoda-ku, Tokyo 100, Japan Telephone: (3) 214-5001 Facsimile: (3) 214-3657 I < Telex: 781-26838 11811 /AJP 0-8213-1811-X PUBLIC ENTERPRISE REFORM $9.95 696 .~~~~~~~~~~~~~~~~~~6,3 0M-o!2-