19689 1999 Building Blocks toward a WORLD More Effective Public Sector BANK J. Edgardo Campos and Sanjay Pradhan INSTITUTE WBI Working Papers Building Blocks toward a More Effective Public Sector* J. Edgardo Campos Sanjay Pradhan *This paper was prepared for the World Bevelopment Report 1997. We would like to thank Ajay Chhibber, Hamid Davoodi, Gregory Kisunko, Brian Levy, Nick Manning, and Gary Reid for their generous comments and contributions. Economic Development Institute of The World Bank Copyright © 1998 The International Bank for Reconstruction and Development/The World Bank 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. The World Bank enjoys copyright under protocol 2 of the Universal Copyright Convention. This material may nonetheless be copied for research, educational, or scholarly purposes only in the member countries of The World Bank. Material in this series is subject to revision. The findings, interpretations, and conclusions expressed in this document are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or the members of its Board of Executive Directors or the countries they represent. ABSTRACT Given the benefits of better bureaucratic performance (Campos and Root 1996; Keefer and Knack 1995), the critical problem is how to achieve it. Following Campos and Pradhan (1996), we approach the issue through the lens of the new institutional economics. We use theories and concepts developed in this field to help us identify and understand what institutional arrangements may help create incentives that encourage civil servants to perform their duties and responsibilities as best they can. To organize the discussion, we distinguish between two broad tasks of the civil service: the first is policy formulation, given the objectives the political leadership specifies, and the second policy implantation. Under each task, we discuss key arrangements that encourage better performance. Our overall objective is to present a coherent framework within which to think about structuring (or restructuring) the civil service so it can facilitate the growth process. In motivating and/or supporting our arguments, we refer to lessons of experience primarily from East Asia and the OECD countries and when available, to large quantitative cross- country comparisons. Our principal conclusion is that effective public-sector capacity has been built upon the basic foundations of a professional, meritocratic, well-paid bureaucracy with internal coherence and institutionalized links to external stakeholders for feedback and accountability. The basic building blocks for effective state capacity are the same for many low-income developing countries as they were historically for developed and more recently for the East Asian economies: strong central capacity for strategic policy formulation and coordination; merit-based recruitment, long-term career rewards, and adequate pay for a professional civil service; and well-functioning systems of financial accountability. Recent reforms in some OECD countries contain some important elements that can improve performance in developing countries, including a greater emphasis on results, enhanced transparency, and greater contestability in service delivery and policy advice. At the same time, many other components of these reforms will impose unrealistic demands on state capacity and are in any event no substitute for the basic foundations for effectiveness. In developing countries with weak state capacity, the challenge is to restore or create a professional and accountable bureaucracy and substitute institutionalized feedback and accountability for patrimonial patronage. Table of Contents I. The Evolution of Public-sector Management .............................................................2 II. Bureaucratic Capability and Economic Growth ............................................................6 III. The Basic Building Blocks ..............................................................7 IV. Institutions for Policy Making .............................................................8 The Implications of Politics ............................................................. 10 Restraint and Flexibility in Policy Making: Incentive Compatible Institutions .... 10 Weak Capacity and Fragmented Policy Making in Developing Countries ........... 16 The Challenge of Central Policy Coordination in Transition Economnies ............. 19 V. Institutions for Delivery ............................................................ 20 The Legacy of the Past and Pressures for Change ..................................... 22 The Problem of Managing Economies: Balancing Flexibility with Restraints ..... 23 Improving Delivery: The Menu of Institutional Alternatives ................................ 27 Improving Delivery through Markets and the Private/NGO Sector ...................... 27 Improving Delivery through the Broader Public Sector ....................................... 31 Improving Delivery through the Core Public Sector .......................... ................. 35 Strengthening "Voice" Mechanisms: Pressures from Clients ................................ 38 VI. Capable, Motivated Staff ............................................................ 41 Merit-based Recruitment and Promotion ............................................................ 42 Building An Esprit de Corps ............................................................ 50 VII. Conclusion ............................................................ 53 Bibliography ............................................................ 54 i Boxes: Box 1: Laying the Bureaucratic Foundations .............................................................3 Box 2: New Public Management Reforms in Anglophone OECD Countries ..................... 4 Box 3: The "Brains" in East Asia's Miracle Economies ....................................................9 Box 4: Politics, Institutions, and Policy: Coffee Marketing Boards in Kenya, Tanzania, and Uganda .............................................................. 11 Box 5: Australia's Mechanisms for Transparent, Competitive, and Result-oriented Policy-Making ......................................... 14 Box 6: Deliberation Councils as Mechanisms for Policy Input and Restraint .. 17 Box 7: Bangladesh - The Mushrooming of Government ............................................... 23 Box 8: Vouchers and School Choice in the U.S., Netherlands, and Chile ........................ 29 Box 9: The Contract State - Victoria, Australia .......................................................... 30 Box 10: Contracting for NGOs for Better Schooling in Bolivia ...................................... 31 Box 11: Internal Market in the UK Health Service - The Purchaser-Provider Split ....... 33 Box 12: A Performance-based Agency for Tax and Customs Management in Ghana ... 34 Box 13: Household Participation to Improve School Performance .................................. 39 Box 14: Client Surveys to Motivate Service Improvements: India, Uganda, and Nicaragua .............................................................. 41 Box 15: Citizen's Charter in Malaysia and the United Kingdom . . 42 Box 16: India - An Elite Civil Service Gasps in the Political Winds . . 45 Box 17: Recruiting and Grooming the Best and the Brightest: Mandarin and Open Systems ............................................................. 46 Box 18: Public-sector Reform in China ............................................................. 49 Box 19: Building Worker Dedication to the Job - Good Government in Ceara, Brazil .......................................... 52 Figures: Figure 1: Bureaucratic Capability Contributes to Economic Growth .................................6 Figure 2: Shifts in Australian Priorities ............................................................. 14 Figure 3: Policy Costs in Guinea ............................. ................................ 18 Figure 4: Ukrainian Lines of Accountability ......................................... 20 Figure 5: Incentive Environments and Organizational Forms for Delivery ..................... 25 Figure 6: Incentive Environments and Service Delivery ........................................... ...... 26 Figure 7: Meritocracy in Korea's Civil Service .............................................................. 44 Figure 8: Pay and Employment in Twenty African Countries . ......................................... 48 Tables: Table 1: Changing Government Responsibilities in Trinidad and Tobago ................ ...... 22 Table 2: Mechanisms to Improve Delivery .......................................... 28 . . "Sire, a vast majority of civil servants are ill paid. ... The result is that skilled and talented men shun public service. The Government of Your Majesty is then forced to recruit mediocre personnel whose sole aim is to improve their weak pecuniary situation. ... Intelligent, hardworking, competent, and motivated individuals should direct Your Empire's civil service.... A program to strengthen the administration must be formulated. ... Responsibilities will have to be well defined and the spheres of action delineated.... It is Your Majesty's prerogative to introduce the indispensable principle of accountability, without which all progress is retarded and work inevitably destroyed. " Excerpts from The Political Testaments ofAli Pasha, Grand Vizier to Ottoman Sultan Abdulaziz, circa 1871. Source: Andic and Andic (1996) Indeed, the concerns of Ali Pasha are very much the concerns of today. Even rulers who had few credible restraints on their authority struggled long and hard to build institutions for an effective public sector that could act in accordance with their preferences. The same problems that plagued Sultan Abdulaziz's bureaucracy - and for that matter, bureaucracies dating back several centuries - continue to haunt today's public bureaucracies. Structuring an effective public sector continues to be a critical problem in many countries, including some of the more advanced. But why this persistence? Part of the reason lies in politics. Depending upon the nature of political dynamics, political preferences can translate into public sectors with very different degrees of effectiveness. In some instances, politicians and their constituencies may have an incentive to maintain the status quo with a high degree of inefficiency and inequity, and those who lose out from such an arrangement may be unable to provide effective pressures for change. In other instances, the private interests of the political principals may run counter to those of their constituencies. But effective institutional mechanisms may not exist to restrain the arbitrary actions of politicians or hold them accountable for results. Still, the answer is not entirely political. In many instances, politicians have strong incentives and interest to improve public-sector performance. But managing public bureaucracies is a complex business, one that does not lend itself to clear, unambiguous solutions. Building institutions for an effective public sector entails addressing a host of underlying behavioral problems that distort incentives and ultimately lead to poor outcomes. At the core of these problems lies the dilemma between flexibility and restraints. Politicians face information and capacity constraints in formulating and implementing policies that meet the interests of their constituencies. Of necessity, politicians delegate some aspects of policy formulation and all of implementation to bureaucrats, but the costs of monitoring what bureaucrats actually do are high. 1 Politicians can structure and attempt to enforce appropriate contracts that set the parameters for bureaucrats' activities. But the outputs of public agencies are often difficult to measure and not subject to competitive pressures, making it costly or unfeasible to enforce such contracts. These same problems are played out throughout the hierarchy of principals and agents in the public sector. Throughout history, rulers, bureaucrats, and their advisors have struggled to identify institutional mechanisms that can address these complex web of problems. These efforts have met with widely varying degrees of success, often reflecting the particular social, historical, and institutional circumstances. Indeed, while some countries have managed to build relatively effective public sectors, many others are still struggling with the daunting problems that concerned Ali Pasha and others more than century ago. L. The Evolution of Public-sector Management Systems The evolution of public-sector management systems across countries is witness to an enduring set of tensions between flexibility and restraints. Early reforms in developed countries and in the more successful developing countries sought to provide restraints on political patronage and build professional bureaucracies. Recent reforms in some of these countries have sought to provide greater flexibility to respond to fiscal pressures, the demands of citizens, and the competitiveness of the global economy. Many other developing countries, however, are groping for an appropriate balance between flexibility and restraints. In Europe, the development of a professional, merit-based, and capable public sector took many centuries. The boundaries of the first modern states were carved only after centuries of struggle. State-building was almost ancillary to war-making. To support and finance their wars, rulers developed centralized control over what were originally private armies and developed institutions to adjudicate disputes. Like governments in many parts of the developing world today, early modern states in Europe were characterized by patrimonial politics. Not until the nineteenth century did expectations grow in Europe that the civil service should be staffed by disinterested and honest public servants (Brautigam 1996). Only in 1816 did Great Britain's civil servants begin receiving formal salaries, marking the decline of the widespread practice of distributing offices as spoils and prebends (see Box 1). The Northcote-Trevelyan reforms of the British civil service in 1854 heralded the self- conscious construction of a professional meritocracy shaped by the rule of law. Since then, a long historical tradition has helped develop professional bureaucracies in Europe. 2 Box 1: Laying the Bureaucratic Foundations - Northcote-Trevelyan Reforms in Britain During the nineteenth century, most industrializing states underwent administrative modernization. Early leaders were France, Prussia, and Great Britain. Until the early nineteenth century, the affairs of the State in Britain were administered by public officials who owed their position to political patronage and influence. There was no common system of pay, bribes augmented official salaries, and office holders, who viewed their positions as property that could be sold, often engaged and paid their own staff. Although such a system did not rule out advance by individual ability, it was not a basis for sound administration. The Victorian era in Great Britain was a period of intense reform, driven by social and economic change and the demands of an expanding and educated middle class. Universities, the armed forces, the judiciary, and the bureaucracy at the central and local government level were all reformed. The blueprint for reform of the civil service was the Northcote-Trevelyan Report of 1854. Commissioned by Gladstone, Sir Charles Trevelyan, who was the Permanent Secretary to the Treasury, and Sir Stafford Northcote, it proposed the creation of a modern bureaucracy based on a career civil service. Drawing on similar ideas advanced for the Indian Civil Service by Macaulay, Northcote and Trevelyan proposed the division of work into two classes - intellectual (i.e., policy and administration) and mechanical (i.e., clerical) - and the creation of a career civil service to carry out this work. New staff for intellectual work would be recruited from the newly reformed universities, securing the best talents by tough competitive examinations, supervised by a board of Civil Service Commissioners. Opposition was strong. Although the Commission was established in 1855, many departments continued recruiting in their own way until 1870, when by order-in-council, patronage was finally abolished, and the two grades were made compulsory for all departments. Northcote-Trevelyan was then followed by reforms in the armed forces, the judiciary, and later, municipal government. There were also extensive machinery of govemnment changes. By the end of the nineteenth century, the foundations of modern government had been laid and the classic values of honesty, equity, economy and political neutrality institutionalized in time for the next wave of change - the foundation of the welfare state and the challenge of world war. Lessons to be drawn from the reforms include: the clarity of Northcote-Trevelyan's vision; the importance of political leadership by Gladstone, first as Chancellor and then Prime Minister; and the pressure of social and economic forces for modernization of the State, boosted by the logistical fiasco of the Crimean War. Even so, the blueprint took several decades to be implemented. Source: Stevens (1996) Even though incremental reforms have been carried out, most European countries have maintained the basic constructs of the traditional, rule-based, top-down public sector, which had its foundations laid down centuries ago. Over the last decade, faced with fiscal crises and growing demands from citizens for better value-for-money, a group of Anglophone OECD countries - particularly New Zealand, United Kingdom, and Australia - have embarked upon far-reaching reforms to restructure the manner in which the public service is managed. Among other things, these reforms - usually labeled as New Public Management (NPM) - devolve significant authority over staffing and budgetary allocations to managers within hard budget constraints while holding managers accountable for the outputs or outcomes achieved (see Box 2). The NPM reforms have attracted a great deal of attention and are being emulated to varying degrees in some developing countries in East Asia and other parts of the world. 3 Box 2: New Public Management Reforms in Anglophone OECD Countries In the early 1980s, Australia, Canada, New Zealand, and the United Kingdom faced serious fiscal crises and growing public demands for better value-for-money. In response to these problems, each of these countries embarked on comprehensive reforms that devolved significant operational flexibility to managers within aggregate restraints and accountability for results. Although these reforms differ among themselves in important ways, they share some important features: (i) binding commitment to aggregate fiscal discipline; (ii) devolution of managerial authority; (iii) accountability for results; (iv) transparency in the use of resources and results achieved; and (v) greater contestability in policy advice and in service delivery. The latter involved a much greater reliance on market-type mechanisms for contracting out to the private sector, as well as contracting in the public sector combined with attempts to create internal markets and benchmarks for performance. Source: Governance (1990) The development of capable public sectors in East Asia have followed varied routes, but the essential ingredients for effectiveness are similar to those found in the OECD countries. In Japan, centuries of tradition have shaped today's bureaucracy. Its present day bureaucracy can be traced back to the emperors' need for an elite corps of warriors, i.e. the samurai, to fend off attacks and to conquer new lands. Basic foundations were instituted and nurtured over a long period of time (with considerable experimentation) to build up the quality and capabilities of the bureaucracy. These included meritocratic recruitment through highly competitive examinations, a merit-based internal promotion system, long-term career rewards, adequate pay, and a strong central capacity for strategic policy formulation. Together, these arrangements have created commitment, an esprit de corps, and a sense of corporate identity much like the legendary samurai. In Latin America, Chile instituted deliberate reforms as far back as the nineteenth century to institute professionalism and -financial accountability and weed out patrimonial influence in the aftermath of independence from the Spanish. More recent reforms have reinforced professionalism and accountability, while informal networks among the elite officials (Chicago Boys) and a relatively unified vision have provided internal coherence. By contrast, many developing countries across different regions have not been able to institutionalize the basic underpinnings for a professional bureaucracy. Though formal systems in many developing countries resemble that of the developed countries that once colonized them, informality has often become the prevailing norm governing actual behavior in many developing countries. Merit-based personnel rules are circumvented in favor of procedures that allow employment, promotions, and transfers for reasons of patronage or ties with the clientele. Formal policymaking mechanisms such as budgeting processes often have little meaning. Budgets are unrealistic, and actual decisions are made in an ad hoc, non-transparent manner during budget implementation. Clearly, there are variations among countries along these general characterizations, but patterns remain. For instance, Peru, Mexico, or Venezuela in Latin America have not succeeded in instituting a merit-based bureaucracy except in very small enclaves. In South Asia, the highly capable bureaucracies at the time of independence have eroded over time. 4 A rapid expansion in public employment and erosion in real wages has undermined the motivation of the once-elite civil service. Despite the presence of highly competitive, merit-based recruitment, political influence has grown in the job assignments and frequency of transfers of senior public servants. Many countries in Africa face daunting challenges to build or rebuild their weak public sectors. Compared with those in other developing areas, African states are very new, the vast majority being formed during colonial rule and taken over by the first generation of African leaders scarcely a generation ago. The rapid growth in public-sector employment, wage erosion, lack of discipline, lack of transparency and accountability, and patrimonial politics have pervaded and undermined the administrative capacity of governments. And worse, some aspects of donor assistance have contributed to the problems: the proliferation of fragmented donor-led projects has undermined the coherence of policymaking. Transition economies face altogether different but nonetheless formidable challenges. In addition to managing a fundamental economic transformation of the role of the state, many countries (e.g., the former Soviet Union) must build a nation-state and a central-government capacity to formulate and coordinate policies within the nation-state. Indeed, the task of building or reforming public-sector institutions varies considerably across regions and within regions across countries. While some industrial countries are embarking on new public management reforms, many developing countries need to proceed with caution, given their vastly different institutional capacities and circumstances. Many countries have not successfully laid the basic foundations of a rule- based bureaucracy. Where informal norms deviate significantly from formal rules, introducing new formal rules will not by itself change incentives. But while the new public management reforms themselves are not transferable in these contexts, some elements are. Principal among these are greater attention to results, enhanced transparency, and more competitive pressures in areas of policymaking and service delivery, which the public sector has long monopolized. Over the long term, there is no substitute for the basic foundations of a professional bureaucracy, with sufficient flexibility to design and implement policies but with effective restraints on informality, patronage, and corruption. These building blocks are the same for many low-income countries today as they were historically for industrial and more recently for the East Asian "miracle" economies. For many of today's developing countries, the added kick to economic performance that a well-functioning bureaucracy can contribute may well be worth the effort. 5 II. Bureaucratic Capability and Economic Growth Does a well-functioning bureaucracy matter for growth? Why should we be concerned about the effectiveness of the bureaucracy? A well-functioning bureaucracy can help improve the efficiency with which resources are allocated and used. For instance, an effective civil service is needed to formulate and implement sound policies that promote growth and improve living standards. Efficient public-sector agencies can deliver critical public goods and services at least cost. Delays in service - such as issuance of licenses, uncertainty with respect to procedures, poor maintenance of infrastructure, etc. - increase costs for firms and thus reduce their ability and their incentives to invest. A civil service of ill repute sends bad signals to society about how their taxes are being used and thus encourages them to evade the taxman (Campos, Levi, and Sherman 1994). These and other reasons suggest why improving the civil service can speed up growth at the margin. A look at some data helps corroborate that bureaucratic capability matters for economic performance and that governments can reap payoffs from an effective public sector. In Figure 1, an index of bureaucratic capability, based upon the perception of investors, is plotted against the part of economic growth that is unexplained by variables such as income, education and policy distortions. As the figure illustrates, there is a high correlation between the index and the residuals. The implication is that more effective bureaucracies can contribute to growth over and above the effect of typical socio- economic variables. Figure 1: Bureaucratic Capability Contributes to Economic Growth 4.. 0 0-1 0.2 .3 0 .4 0.5 Om6 0.7 Index f bureaucratic capability (average for 1964-93) ..~~~~ 4E...,, Note: -- ' Sc o n . . .. 1 6 4~~~ 0 010.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Indes of bureaucratic capability (average for l964-93) Note: controlling for income, education and policy distortions Source: Commander, Davoodi, and Lee (1996) 6 III. The Basic Building Blocks A well-functioning bureaucracy can promote growth and poverty reduction by providing sound policy input for state intervention and by delivering critical public goods and services at least cost. Experience suggests that building institutions for an effective public sector requires putting in place incentives that can generate well-informed and disciplined strategic policies, deliver public services efficiently and effectively, and attract and motivate capable staff to carry out these tasks. The capacity to formulate and coordinate policy is like the brain of the human body. Politicians formulate visions and set goals. But for these to materialize, they need to be translated into strategic priorities through policy formulation and coordination; without this, vision and goals remain unrealized promises. This requires mechanisms that lead to well-informed, disciplined and accountable decisions. A constant challenge for all countries is to set rules that give politicians and their expert counselors the flexibility they need to formulate policies but embed their decision making in processes that allow for inputs and oversight from stakeholders. With information and appropriate "rules of the game" comes the brain. The brain is of no use without the limbs. Policies need to be implemented and the programs and services that implementation embodies need to be delivered. Delivery systems become the limbs that respond to the signals from the brain and make things happen. Here, too, reform involves setting the right balance between flexibility and accountability. For activities that are contestable and easily specified, market mechanisms and contracting out services can often improve delivery dramatically. But for many other services, there is often no substitute for delivery by the core public sector. Giving citizens greater voice here and allowing client feedback can exert pressure for better performance, but ultimately performance will depend on the loyalty of civil servants and their compliance with established rules. But both the brain and limbs need blood to come to life. Ultimately, it is the civil service that breathes life into the public sector. It is the life blood of the executive, without which nothing moves and nothing gets done. Motivated and capable personnel inject energy for the public sector; an anemic staff stifles it. Civil servants can be motivated to perform well through a range of mechanisms, including merit-based recruitment and promotion, adequate pay, and a strong esprit de corps. The rest of this paper looks into the inner workings of the executive and explores the institutional underpinnings that make for a healthy "executive." Each of the key parts -brain, limbs, and life blood - are discussed in turn. 7 IV. Institutions for Policy Making Sustained economic development has escaped many of the world's developing countries. For this reason, there has emerged considerable interest in the experience of successful performers, in particular the so-called miracle economies of East Asia and more recently of Chile in Latin America. While these countries have undoubtedly adopted different strategies, they do share some common features. The most striking commonality lies in the area of policy making. In each of these countries, leadership formulated a long-term vision for the country and created the institutional arrangements needed to translate this vision into a highly focused set of strategic priorities. In other words, leadership created the "brain" required to focus policy-making efforts on long-term priorities that its vision implied. It is sad but true; for many developing countries, a similar capacity at the center of government has been missing, and in a few, even the vision is not there. The result: incoherent policies that reflect ad hoc responses to problems that, in some cases, lead to perpetual crisis management. A long-term vision formulated by political leadership provided the focal point for the "brain." In Japan, for example, it was the LDP's catch up with the West, and in Malaysia, it has been Prime Minister Mahathir's Vision 2020 (Campos and Root 1996). But to magnify this vision and reveal its logical components - the policies needed to translate the vision into reality - powerful and relatively autonomous central agencies with tightly knit personnel (e.g., Ministry of Finance, Ministry of Planning) were created to maintain macroeconomic stability and coordinate policies that were designed to meet long-term priorities (see Box 3). Across the Pacific from East Asia, a potential high performer is emerging: Chile. What is interesting is that Chile shares several common features with East Asia's high performers. Most striking is the affinity of its so-called "Chicago Boys" with Indonesia's "Berkeley Mafia" and in more recent times with Thailand's "gang of four." Central capacity for policy making and coordination has been built up in Chile much like it was in East Asia. Of course, Chile's and East Asia's miracle economies have not been problem-free. However, they do share a common feature of having been successful in varying degrees in formulating effective policies to achieve their principal goals. As the World Bank (1993) and many others have shown, macroeconomic stability and strategic prioritization have worked reasonably well in these countries. One comment often heard about successful performers is that politics in their countries were suited for promoting economic growth. The implication is that their experience is difficult to adapt to other developing countries because of less favorable political conditions. But this is not necessarily the case in the context of policy making. 8 Box 3: The "Brains" in East Asia's Miracle Economies Elite agencies have played a powerful role in formulating and coordinating strategic policies in East Asian countries. The specific roles have varied, but this pattern is common across the spectrum, including Japan, Korea, Taiwan, Thailand, and Indonesia. Japan and South Korea share the tradition of having powerful central agencies that exercise considerable influence over macroeconomic policies and industrial development. MITI in Japan and the EPB in South Korea are considered to be the primary movers behind the formulation of industrial policy and the use of the associated process of administrative guidance. Roughly speaking, administrative guidance involves an application of "carrots and sticks" that induce private firms to engage in preferred areas of manufacturing and industry (at least during the high growth period). While there is still much debate about the efficacy of industrial policy in these countries, there is no disagreement about the extensive involvement of these agencies in the process of promoting industrial transformation. In Japan, the MOF has controlled the budget process and thus fiscal policy. "On an annual basis, the ministry's various budgetary officials fan out through the Kaumigaseki area of Tokyo, which houses most national government offices, to oversee the budget-making process of Japan's individual ministries.... One power is central: the power to set a finite limit on the expansion, from one year to the next, of any agency's budget" (Pempel and Muramatsu 1995: 59). In Korea, the EPB, which is considered a (and is the only) superministry, has control over both budgeting and planning, powers that allow it to coordinate fiscal policy and direct industrial transformation. Taiwan too has had powerful central agencies. But at the pinnacle is the Central Bank, which has a considerable degree of autonomy from government and significant influence over other agencies. Another powerful agency, the Council for Economic Development and Planning (CEPD), was chaired for a long time by the Central Bank Governor. CEPD "is an advisory body to the cabinet and all but one of its twelve councilors are cabinet members. K. H. Yu, governor of the central bank ... has been its chairnan since 1978.... CEPD constitutes by far the largest concentration of economists in government service. Its staff is divided into seven departments ... with responsibilities that (among others) cover the formulation of one, four, and ten-year mnacroeconomic development plans ... and the arbitration of disputes among ministries" (Wade 1990: 197). The Ministry of Economic Affairs, through its Industrial Development Bureau, and the MOF have been loosely responsible for industrial policy. However, their ability to use administrative guidance has been much more circumscribed and does not approach the magnitude and depth to which it has been exercised in Japan or South Korea. Preparation and management of the budget sits with the Directorate-General of Budget, Accounts, and Statistics (DGBAS), the MOF, and a committee consisting of four ministers without portfolio. In Thailand, the Ministry of Finance, the Budget Bureau in the Prime Minister's office, the Central Bank, and the National Economic and Social Development Board (NESDB) - the so-called "gang of four" - consult each other in the preparation of the budget with the express objective of setting a level of deficit that is sustainable over the medium term. They act in unison to put a cap on spending and to constrain inflation. In Indonesia, the Ministry of Finance and Bapennas, the planning agency, have been the guardians of the purse and the brain of the civil service. As historical accounts place it, President Suharto recruited a team of technocrats, often referred to as the Berkeley Mafia, placed them in key ministries of which the MOF and Bapennas were the primary core, and have called on them time and again to address macroeconomic and sector problems. As such, they have become a critical player within the President's coalition of supporters and a core element of the "brain" of government. Whether it is the technocracy in Indonesia or the dedicated bureaucrats in Japan's MII, personnel in each of these powerful agencies have been bound together in varying degrees by an esprit de corps - a common understanding among its members about what is desirable and undesirable behavior, manifested in norms both formal and infonnal and grounded on some set of objectives, and a devotion to upholding the honor of the group based on this common understanding. The esprit has created a high degree of coherence among the group's members, given them a strong sense of purpose and belonging, and effectively created a self-discipline among members that guides them towards achieving the group's objectives. Source: Campos and Root (1996); Leipziger and Kihwan (1992) 9 The Implications of Politics Societal goals and the policies designed to achieve them typically reflect the underlying structure of politics in the country. Policies have distributive consequences for different groups and groups lobby politicians to protect or enhance their benefits. Hence, policies become the instrument through which politicians seek political support. To remain in office, politicians need sufficient support from society, and this invariably means helping promote policies that favor their constituents. This political logic has implications for the type of institutional arrangements that countries adopt to facilitate the process of policy making. Just as policies reflect the political realities in a country, so do such arrangements. The experience of Kenya, Uganda, and Tanzania with marketing boards illustrates this very clearly (see Box 4). In sum, politics and institutional arrangements are invariably linked. Understanding the latter requires understanding the underlying incentives of politicians and their constituents. Despite the primacy of politics in the structuring of institutional arrangements, there remains considerable room for adapting generic arrangements in the realm of policymaking. Because politicians face collective action problems among themselves - which tend to significantly reduce their net gains (from being politicians) - they have strong incentives to support the creation of institutional arrangements that help resolve these problems. This section attempts to identify and explore which aspects of policy making embody this compatibility and present good opportunities for building institutions within the executive that promote development. Restraint and Flexibility in Policy Making: Incentive Compatible Institutions MACROECONOMIC MANAGEMENT. Politicians are not a monolithic entity; they are a collection of individuals with different preferences. Each desires and pursues a different package of policies. But while doing so may be individually rational, the collective outcome of such endeavors could make everyone worse off and ultimately affect everyone's tenure. This dilemma is vividly manifested in the arena of macroeconomic policymaking. Politicians desire to have policies that cater to their respective constituencies. Hence, come budget time, they push to have funds allocated to programs that promote these policies. Given common pool problems that characterize public-sector budgeting, this generally leads to unsustainable budget deficits and ultimately to macroeconomic instability manifested primarily in constant bouts with high inflation (Campos and Pradhan 1996). As the '70s and the '80s have shown, this creates considerable political fallout since the population and particularly the business sector react and respond negatively - e.g. capital flight, withdrawal into the informal sector - leading to a dynamic that erodes the economy's performance further and leads to stagnation, if not decay. The usual consequence is that the politicians are booted out of power. Through experience, politicians recognize this problem and strive to create institutional arrangements that help prevent such perverse collective outcomes. 10 Box 4: Politics, Institutions, and Policy - Coffee Marketing Boards in Kenya, Tanzania and Uganda Coffee is the major export of Kenya, Tanzania, and Uganda. Tanzania's arabicas are marketed along with those of Kenya among the so-called Colombian mild coffees, the highest quality and highest priced coffees in the world. Uganda and Tanzania export robusta coffee as well. In addition to specializing in the export of a similar crop, the three nations share a common colonial heritage: prior to independence, all were governed by Britain. The legislation and regulations that they employ reflect this history: as the coffee industries of each expanded, their governments adopted highly similar means of regulating them. The legislation called for the creation of Coffee Boards that would provide public services, such as research into the control of pests and diseases; regulate production to defend reputations for quality in international markets; and provide infrastructure for the industry by promoting cooperatives through which to distribute seeds, chemicals, implements, and farm credit. The agencies thus were designed to act as economic institutions to maximize the value of the production of the industry and the value of exports. In the early 1960s, the three nations joined the International Coffee Agreement in an effort to raise the value of exports. The intention was to cooperate with other producers in reducing supplies; given the inelastic demand for coffee among consumers, the result would be an increase in prices in world markets and an increase in revenues. Each government therefore strengthened the powers of their Coffee Boards so as to secure adherence to export controls. Each government made its coffee board a monoposonistic purchasers of the coffee crop and its sole legal exporter. Adopting similar institutions and sharing comparable policy objectives, the three governments nonetheless secured very different policy outcomes. With the passage of time, the regulatory environment changed in each nation. In Kenya, the govemment initially employed the code board to restrict production by banning the further planting of coffee. By rigorously restricting quantities, it was able to pass on the higher world price to domestic producers. In Uganda and Tanzania, the governments employed pricing policies rather than quantity constraints to secure adherence to the quotas. They imposed a tax wedge between the higher world price, secured by the international agreement, and domestic price, paid to farmers. The forms of intervention varied not only by nation but also over time. The performance of the economic institution depended upon its political environment. A major reason for these variations over place and time was the differential effect of political forces, including the power of regional interests and incentives created by the competition for political office and the revenue interests of the state. Source: Bates (1996) The typical arrangement involves politicians tying their hands by granting central ministries the authority to keep individual pursuits from leading to bad collective outcomes and appointing well-trained, reputable individuals to key positions in these ministries. Such arrangements delegate the authority to resolve coordination problems among politicians to bureaucrats and insulate (for the lack of a better term) these bureaucrats from interference from individual politicians. They also help signal the business community that the government intends to restrain unbridled behavior that destabilize the macroeconomy. More specifically, the exercise of collective restraint has involved the delegation of aggregate monetary and fiscal policy to relatively autonomous and powerful central ministries, such as the Ministry of Finance and the Central Bank, and the recruitment of reputable technocrats to manage these agencies to signal the business community that a stable macroeconomy will be maintained. As Box 3 shows, many of East Asia's miracle economies in part owe their remarkable performance to central agencies and their senior I1 officials, whose power and influence were derived from the support of leading politicians (World Bank 1993). But East Asia is not the only region where politicians have gone this route. In recent years, much of Latin America has done so as well in response to the debilitating experience with hyperinflation in the late 1970s and most of the 1980s (Geddes 1994). Most recently, even India, which has not had serious problems with inflation, appears to be going this route. Despite the fall of the centrist reform government and the rise of a more left-of-center government, a fiscal conservative has been appointed as the Finance Minister with the express mandate to maintain the government on a path of economic reform and to signal its credibility to investors. Interestingly, this political logic is economically rational as well. Politicians and leadership in particular recognize they need to delegate authority to the bureaucracy for the simple reason that they cannot do everything. However, by delegating, they create room for bureaucrats to abuse delegated authority, i.e. the principal agent problem. To prevent this, politicians must be able to monitor the performance of bureaucrats and hold them accountable for their actions, but the performance of bureaucrats must be easily measurable or verifiable; otherwise, transaction costs will make monitoring difficult, if not impossible. Because indicators of macroeconomic performance are among the most objective and are relatively simple to quantify, it is easy for politicians to monitor the performance of economic technocrats and the agencies they manage. The usefulness of this arrangement can be appreciated from another perspective. For many developing countries, scarcity of skilled personnel at senior levels is a particularly serious problem. Thus, the assignment of such personnel to the more important agencies becomes imperative. As the East Asian Miracle (World Bank 1993) tells us, stabilizing the macroeconomy is a necessary condition for sustained economic growth: without it, nothing much follows. Hence, if good senior people are in short supply, the logical places to assign them are to the agencies responsible for macroeconomic management and policy coordination if economic growth is a primary objective. This coincidence of the political and economic logic makes it possible to construct a flexible commitment mechanism, one that establishes restraint on politicians but provides flexibility for bureaucrats to respond to changing circumstances. Through delegation of macroeconomic policy making to central agencies managed by reputable technocrats, politicians create a credible mechanism to restrain their individual pursuits from destabilizing the macroeconomy. This, of course, opens up possibilities for misbehavior on another front: the bureaucrats in these agencies. But because performance of macroeconomic agencies is easily monitored, deviant behavior can be controlled. Thus, these agencies can be made reasonably autonomous to give them the flexibility to formulate policies to respond to constantly changing conditions that affect the macro economy. 12 STRATEGIC PRIORITIZATION. The political battle over policies manifests itself most strongly in the area of prioritization of govemment expenditures. This prioritization involves give-and-take among politicians while they are adopting policies. A policy embodies different (implicit) gains and costs to individual politicians. For some, it may represent a net gain, and for others, a net loss. Politicians must necessarily differ in their preferences over policies. While this may seem to create chaotic conditions, it presents opportunities for exploiting "gains from trade" among politicians. Given the aggregate constraints implied by the institutional arrangements for macroeconomic management, politicians need to make trade-offs among policies and the programs that promote them. Hence, they will be inclined to trade their support for policies from which they gain less for policies from which they can gain more. In order for such trades to occur, however, there is a need to establish arrangements that facilitate coordination among policies. Trades among politicians generally involve timing problems. For example, in democratic settings, votes on different sets of policies occur at different points in time. A politician may offer to support another politician's preferred set of policies in return for the latter's promise to support his or hers, but in between, the situation may change so that the latter may decide to renege. Hence, the former has no assurance that the latter will indeed deliver his or her vote. Politicians therefore have an interest in creating and supporting coordination mechanisms - mechanisms that resolve timing problems and allow trades to take place. In Cabinet-based systems among the OECD countries, the institutional arrangements that facilitate coordination of policies among politicians and thus facilitate trades is grounded on the so-called portfolio allocation system among ministries. In its simplest form, "Policy issues are assumed to be bundled into mutually exclusive packages; each of these packages falls into the jurisdiction of a particular government ministry with the authority and indeed the responsibility to act in these matters. The allocation of particular issues to particular ministries creates a division of labor within the cabinet.... Cabinet ministers are presumed to do their best to implement (their) party's policy within their ministry's jurisdiction" (Laver and Shesple 1994: 288-289). The system thus allows parties to trade votes among different policy packages. In order for trades to occur, there must be sufficient information about what is being traded. This means the benefits and costs of policies must be threshed out and information thereon made available to the whole legislature. The portfolio allocation system does not in and of itself offer a means to address this. In Australia, this issue came to the forefront of the public-sector reforms that began in the early 1 980s (see Box 5). 13 Box 5: Australia's Process for Transparent, Competitive, and Result-oriented Policymaking One of the main objectives of Australia's public-sector reforms has been to institute a process to discipline, coordinate, and contest policies. Some of the challenges faced by the Labour government when it came to power in the early 1980s were similar to those confronting many developing countries today: fiscal crises and the unsustainable fiscal commitments of previous policies. To discipline policy formulation and gain political support for strategic prioritization, the new administration decided to publish estimates of existing policies. In the early years, these estimates projected sizable and unsustainable real growth in future spending requirements, undersconng the need to scale back the size and scope of government activities. Once the government had published the estimates, however, it was obliged to continue to do so to show continuing declines in future conunitments -the so-called "falling man" (see Figure 2). Open financial markets served as an additional disciplining device on the publication of the data. The reforms also required the government to publish a reconciliation table showing the deviation between the forward estimates of existing policies and those of new policies. These measures helped make transparent changes in the government's strategic priorities, as well as the medium-term costs of new commitments. In addition, forward estimates provided predictability in resource flows to line ministries, since forward estimates were automatically rolled over into the budget if there were no changes in policy. This approach helped improve decision making and the operational efficiency of line agencies. The reforms also required that any new policy or changes in policy proposed by line ministries be accompanied by offsetting savings to stay within the resource envelope agreed in the cabinet. The cabinet focused on changes in strategic priorities to meet targets for net spending or savings. Policy proposals were contested in the cabinet as well as by the requirement that all ministries and agencies affected by the policy submit written cornments on the soundness of the proposals. This approach helped legitimize and build consensus on policy priorities. Finally, the reforms focused attention on policy results through mandated periodic evaluation of new and existing policies and reporting on performance and outcomes. The results? The deficit, 4 percent of GDP in 1983, became a surplus by the decade's end. At the same time, there were significant changes in the composition of public expenditures, reflecting both the strategic shifts in intersectoral priorities identified by the Cabinet and changes in intraministerial priorities, often identified by line agencies themselves (see Figure 2). Figure 2: Shifts in Australian Priorities Projected growth Changes by category 1983/84 to 1992/93 Billions of Percent (cumulative) Australian dollars March 1984 5 . 'W/ ~ ~ 4 .New spending Projection dates Netchange May 1985 2 November 1985 -1 December 1986 -2 Aug st Savings from 19 e existing policy -4 1982/83 1983/84 1984/85 1986/87 1988/89 Defen-s Ed.c.- He-tUb Sociat Tr,.s- Other tion security portation Fiscal years Source: Campos and Pradhan (1996); Dixon (1993) 14 The cornerstone of the reforrns has been a system of forward estimates, which are three-year forecasts of the minimum cost of existing policies and programs that are published and are made available to the general public. The estimates have several virtues, one of which is to provide a baseline upon which various ministries through their respective cabinet ministers can compare competing policies, policy proposals generated by ministries/departments are contested subjecting them to written reviews from other ministries/departments. Another virtue is that it reduces the cost of obtaining information. The estimates are automatically rolled into budgetary allocations if there are no changes in policy; moreover, a ministry or department that requests a change must defend this before the whole legislature. This removes from ministerial consideration (i.e. the discussion of trades) outlays in any budget that do not involve any changes in policies, which, given the transactions costs of proposing changes, form the bulk of the budget. In effect, the system of forward estimates reduces the costs of information and enables trades among ministries to be better informed. In the United States, the congressional committee system evolved primarily as a means to make intertemporal vote trades feasible. "In a large, diverse republic, political interests will be heterogeneous - things dear to one set of representatives will be of little value to others, and vice-versa. Thus, logrolling can produce gains from trade.... If opportunities to make the same sorts of trades are expected to occur repeatedly, transactions costs can be cut by institutionalizing the trades. ... Thus legislatures establish committees with specialized jurisdictions, permit members to self-select positions on committees with greatest interest to them ... and adopt 'norms' that lubricate the gears of the system" (Fiorinan 1987, p. 338). The result is an "institutional arrangement for exchanging support that is superior to a spot market" (Shepsle and Weingast 1994: 156). The system makes bargains among politicians credible and thus enables policies to be coordinated in ways that are politically feasible.' U.S. executive agencies, for their part, are governed by the Administrative Procedures Act of 1946, which provides flexibility on the substance of policies while imposing procedural requirements - such as public announcement of new policy - that are enforceable in the courts. This procedures-oriented approach to policy formulation allows legislators to shift substantive policymaking to specialist agencies and other interested parties closer to the problem. This kind of decentralized mechanism uses citizen voice and the judiciary for ensuring accountability, albeit with the inevitable side-effect of slowing the decision-making process. Likewise, many continental European countries rely on administrative law and specialized courts for judicial review of administrative actions. Citizens can challenge administrative decisions on legal grounds or for factual errors. The European Union has adopted this system to allow the European Court of Justice to oversee decisions made by European institutions. 1 Unlike in many other countries, control of unbridled behavior among politicians and the setting and pursuit of long-term priorities is effectively imposed through party leadership and discipline (Cox and McCubbins 1993). The role of the Ministry of Finance is much more subdued. However, the Federal Reserve Bank, an independent body, does provide some restraint on macroeconomic policy making. 15 In Japan and South Korea, the institutional arrangements underpinning industrial policy formation are well known (Chung 1992; Aoki, Kim, and Okuno-Fujiwara 1997; Campos and Root 1996). MITI and the MOF in Japan and the EPB in Korea were given considerable leeway to establish sectoral priorities and to influence the investment patterns of private firms through the instruments of administrative guidance. Information about costs of different policies was extracted through the system of deliberation ccuncils: tightly organized consultative committees, private-sector firms, and central ministry bureaucrats constantly exchanging information about the potential and actual impact of policies. This restrained pursuit of costly programs. Policymaking mechanisms in East Asia, therefore, provided flexibility to strong central agencies while restraining and embedding them in institutionalized mechanisms of input and oversight through deliberation councils (see Box 6). The examples illustrate that politicians are willing to impose restraints on themselves in order to make more informed decisions about policies and to enable politically beneficial trades to materialize; rules that govern the "gains from trade" system put restrictions on politicians. But the rules generally pertain to procedures rather than to specific do's and don'ts, so in effect, flexibility is built in. While it restrains politicians from pushing preferred though costly policies, the Australian system of forward estimates combined with the portfolio allocation system actually encourages a focus of discussions and negotiations on a small number of policies, those for which changes are most needed. A good "brain" embodies a reasonable balance between restraint and flexibility. Weak Capacity and Fragmented Policy Making in Developing Countries In many developing areas, such as Africa, Central America, and the Caribbean, there is only weak central capacity for forming and implementing macroeconomic and strategic policies. Available capacity is thinly stretched among a few senior officials who must attend to many tasks. Consequently, there is an absence of a critical mass with capacity and internal coherence to formulate and coordinate strategic policies. The capacity that exists is not embedded in a larger network of external think-tanks that provide input and contest policies. Beyond this, a central problem in many countries is a lack of transparency in decision-making. For instance, the original published budget is often a mirage, and actual policies are remade during budget implementation. Consequently, there are large variations between budgets and actuals. For instance, in recent years, the average variation between budgeted and actual recurrent expenditures across ministries has been over 50 percent in Tanzania and over 30 percent in Uganda. Further, considerable lags in the production of financial accounts and audits often impedes the ability of policy makers to make informed decisions. To address these problems, countries are undertaking reforms to strengthen financial information systems (e.g. Integrated Financial Management Systems in Latin America). The capacity of central agencies to more accurately program revenues and macroeconomic estimates is also being strengthened as in the Philippines and Ghana. 16 Box 6: Deliberation Councils as Mechanisms for Policy Input and Restraint No matter how qualified the policy formulators in the government are, they need input from those who will be affected by the policy. Institutionalized mechanisms for input and oversight from key stakeholders provides important information and restraints to policy makers. A mechanism that has proven very useful in some cases is the deliberation council. A council is a forum through which stakeholders and the executive/bureaucracy can regularly exchange information and discuss policies that would be most appropriate to resolving specific problems. Its value comes from its reduction of uncertainty in the policy environment; stakeholder groups - or at least those who are best organized- know that any changes to the existing policy regime must first be discussed within the council - it is a commitment device. The Joint Public Sector-Private Sector Consultative Committee in Thailand during the military rule in the mid-80s, the Monthly Export Promotion Meetings in Korea during General Park's time, and the informal elite based consultative committees in Chile during the latter half of General Pinochet's rule are concrete examples of successful deliberation councils despite the absence of an information filtering mechanism through the legislature. Even in societies with active legislatures, deliberation councils have served as useful consultation and commitment mechanisms. The numerous industry-based councils in Japan, the National Wages Council (NWC) in Singapore, the Budget Dialogue Group in Malaysia, the consensus councils in North Dakota, and the business-labor sector councils in Canada are concrete examples. Hence, it is not that sensitive to the type of political regime. The key to success lies in the council's capacity to act as a credible commitment mechanism. There are several factors that help build this capacity. One is low turnover of participants, especially those from the government. The repeated game feature of a council creates incentives for participants to work cooperatively. If people on one side of the table change frequently, then those in the other side will have less incentive to cooperate and are likely to cease participating. Many of the above mentioned councils have had very stable membership, at least during the period when they were effective. In Korea, for example, when parliamentary democracy once again emerged, the Monthly Export Promotion Council and other similar councils went by the wayside. One reason was a change in the guard within the executive and the key players in the bureaucracy who championed the creation and/or use of these councils. Another factor is focus on a clear, relatively narrow set of issues. The broader the set of issues, the larger the number of stakeholders and the more difficult it becomes to build a consensus or agree on a compromise. The most successful and among the longest lived of the above mentioned councils, the NWC in Singapore, focuses mostly on wage policies such as guidelines for annual wage increases; every year, the NWC issues guidelines for wage increases, and every year, both business and government have conformed to those guidelines, including a de facto wage cut in the mid-80s (see Box 18). The sector councils in Canada have been reasonably successful in resolving issues pertaining to worker training and retraining (addressing interfirm externalities) but have had difficulty moving beyond these issues. In Japan, where the council system is most well developed, councils are subsector specific. Source: Campos and Root (1996) In aid-dependent countries, donors both alleviate and exacerbate the problem of weak central capacity. To the extent that donors provide policy advice that supplements weak government capacity at the center, the short-term problem is mitigated. However, this works against building long-term capacity, if the need for local counterparts is not acutely felt and demanded by politicians. Donors also impede central capacity for policy formulation through the fragmentation in decision-making introduced by individual donor projects. Individual 17 ministries and donors enter into bilateral deals on individual projects, though the accumulation of these individual projects may not be mutually consistent. In Zambia, for instance, there were over 180 projects in the agricultural sector alone funded by a large number of donors. Each of these projects had different objectives, often different from Zambia's overall sectoral strategy. Furthermore, the cumulative recurrent costs of these may not be sustainable. The Public Investment Program often becomes a passive repository of donor- driven lists of projects. In several instances, the lack of coordination between the Ministry of Planning and the Ministry of Finance impedes the integration of capital and recurrent expenditures. Together, these factors result in weak central capacity and institutional mechanisms to coordinate and discipline decision-making. The problems that typically result are illustrated in the case of Guinea. Even though the government's stated priorities were primary education, public health, and road maintenance, actual expenditures were allocated to other programs. There was also no system of costing or contesting competing policies. An exercise to do so revealed that the share of these priority programs would need to triple over the next four years, implying drastic cuts in other expenditures. The problem of unsustainable recurrent costs of donor- led investment projects is illustrated in Figure 3. There is a growing gap between actual recurrent expenditures and requirements. Figure 3: Policy Costs in Guinea Guinea: Costs of Sectoral Policies 1994 Guinea: Recurrent Costs of Investment Projects 85% Millions of constant GNF 400 250 2000 200 150 56% 100 50 44% 1996 1997 1998 1999 2000 _ capital expenditures Pnmary Education Road maintenance -rU-ecurrent e:penditure requirerDeats | Public Health Others --recurrenlt actual Source: World Bank (1996b) 18 A number of important initiatives have been launched in some countries to address these problems. Governments and donors have together launched Sectoral Investment Programs, which coordinate donor assistance to support coherent policies such as the Agricultural Sector Investment Program in Zambia. Countries like Malawi and Uganda are moving beyond coordinating donor policies for a given sector to undertaking a systematic process for strategic prioritization across sectors within aggregate expenditure constraints. This has entailed top-down priorities as well as bottom-up costing of policies in line ministries. In Malawi, for instance, four pilot ministries - education, health, works and agriculture - are costing out policies within a hard budget constraint and explicitly focusing on what competing policies are designed to achieve and how much they cost for the first time. A few countries, such as Colombia, are undertaking comprehensive efforts to institute an ex-post evaluation system to better assess whether policies and programs are achieving their intended outcomes. Changes in policy formulation are also evident in procedures for preparing submissions to Cabinet, in the structure of the secretariat, in the mechanisms employed for ensuring coordination of complex and diverse proposals, and in the implementation of monitoring arrangements. Proposals for changes in Cabinet procedures under consideration by the Government of Lesotho will focus Ministry attention on the need for policy advice to be coherent and consistent. The Cabinet Office in Zambia has recently established the Policy Analysis and Coordination (PAC) Division. The goal is to strengthen coordination capacity and to change the rules of the game for policy formulation and contestability. The Challenge of Central Policy Coordination in Transition Economies The experience of the former Soviet republics exemplifies problems that arise in the absence of a coherent policy-formation mechanism. Under the Communist regime, policy making and coordination were the function of the Central Committee and the Politburo in Moscow. The function of the state administrative was simply to implement Party policy. The abolition of the former Communist Party removed the central decision-making apparatus (the brain) for coordinating the various ministries and departments (the limbs). These countries are therefore confronted with haphazard, overlapping responsibilities and multiple rather than collective accountability, a definite formula for disastrous policy formulation. Ukraine exemplifies such problems in the extreme. Following independence in 1991, a central machinery of government was established that reflects many of the features of a former Soviet system. Decision making remains highly centralized. The Apparat of the Cabinet of Ministers has retained responsibility for policy formulation and coordination and directs the activities of central government departments. The number of central government bodies remains large (110), their responsibilities often overlap, and lines of accountability are unclear (see Figure 4). The cumbersome structure makes coordination difficult, delays decision making, and reduces transparency. However, efforts to reform this system are under way following the adoption of the new constitution in July 1996. 19 Figure 4: Ukrainian Lines of Accountability Vrc Prime ister A First Vic Serrime, irector Is ic Ve m ity Line rinistr Lin e min e stinister of Minister ~Minister /Ca4ne 0f Minister State S E l ~APPARAT OF CABINET OF NMlNISTERS President's - : : : : : : SAdminestration Eae DenarErpents, Services, wDirectorh4s -------- -=\-a/~ Line ministry Lirne miniistry Line ministry Agency head Source: The WNorld Bank (1997a) Some Central and Eastern European countries with simiular, though perhaps less severe, problems have initiated promising reforms of their central decision-making mechanisms. Poland and Hungary have both introduced reforms to streamline multiple and conflicting responsibilities and speed up the decision-making process. In Georgia, improved streamlining has removed overlapping and conflicting positions, and draft laws are decided in the presence of all members of the President's Economic Council before submission to Parliament. Such reforms have facilitated consultation and coordination in central government decision making, but most countries have a long way to go to build the institutional capability to respond effectively to the many demands of transition. V. Institutions for Delivery Even well-designed policies have experienced major implementation problems in many countries. Services are simply not delivered, or they are delivered badly. Poor quality, high cost, waste, fraud, and corruption have marred delivery in many developing countries. The central problem has been poor incentives, stemming from a lack of both flexibility and restraints. Politicians have often intervened in the day-to-day operations of public agencies, and public-sector managers have had limited managerial flexibility in delivering services. At the same time, there have been few credible restraints in the form of probity or accountability for results. More fundamentally, the public sector has assumed or tried to assume a monopolistic role in delivery, and there have been few external pressures for better performance. 20 In recent years, mounting citizen dissatisfaction, continuing demands for fiscal stringency, the pressures for better value-for-money, and the imperatives of global competitiveness have added new pressures on States and politicians to improve delivery. It is not surprising, then, that revamping delivery systems has become a central aspect of State reform in many countries. Many countries are actively exploring alternative institutional mechanisms for delivery. A range of institutional alternatives is being explored, but the shared principle is to introduce greater contestability through actual competition or through competitive surrogates. These are providing users more "exit" options for better quality at lower cost. Contracting out to the private sector and NGOs is a growing trend based on the realization that even if the State has a role in ensuring the provision of a service, it does not have to deliver the service itself. In addition, user participation, client surveys and published benchmarks are providing more options for using "voice" in service provision. Finally, several developed countries in particular are seeking to improve service provision through internal incentives: entering into formal contracts with public-sector agencies and providing them greater managerial flexibility while holding them accountable for results. Other countries are seeking to rely on more traditional bureaucratic forms through meritocratic recruitment and promotion and the creation of an esprit de corps to build loyalty for better performance. Several of these initiatives hold great promise for improving delivery in developing countries. At the same time, it is crucial to recognize that many of these initiatives do not obviate the role of the State in delivery. Rather, the role of the State shifts from being a direct provider of several services to a regulator or financier of services delivered through private or public-sector organizations. Specifying and enforcing these implicit or explicit contracts requires effective State capacity. The capacity requirements vary depending upon the characteristics of the service, including how contestable it is and how difficult or easy it is to specify outputs. Countries differ widely in their strength of underlying institutions that influence their capacity to enforce internal and external contracts for services of varying degrees of complexity and contestability. For some countries with strong institutions, there are considerably greater opportunities for exploring alternative delivery mechanisms for a range of services. Even States with weak capacity may be able to contract out easily specifiable and contestable outputs, depending upon the strength of the private market for those outputs. But there are important cautionary tales from past experience. The State in many countries has not been effective in enforcing contracts even for relatively easily specification and contestable outputs. The dire performance of State-owned Enterprises (SOEs) in many countries is a sad testimony to the problems governments have faced in enforcing contracts within the public sector. The challenge for state reform in delivery, therefore, is to match institutional arrangements that improve performance with the capacity of the State to enforce internal or external contracts. This entails unbundling public services and exploring the fit between flexibility and restraints suited to the country's institutional endowment. 21 Table 1: Changing Government Responsibilities in Trinidad and Tobago 1955 1996 Education Education Customs Foreign affairs Public health Health Police National security Land survey Planning and development Roads Works and transport Judiciary Legal affairs Agriculture, land, and marine resources Community development and women's affairs Energy Housing Labour and co-operatives Public utilities Social development Sport and youth affidrs Tourism Trade, industry, and consumer affairs Source: Manning (1996) The Legacy of the Past and Pressures for Change The present-day challenges confronting public delivery systems has been a product of the expanding role of the State in a service provision over the last few decades. In developing countnes, governments embarked on development strategies that accorded a dominant role to the State in economic transformations. They set up state-owned enterprises to produce a variety of goods, including commercial products, mining, and heavy industry. Governments also viewed themselves as the dominant - if not the only- provider of many services to society. They undertook significant infrastructure expansion through monopoly provision and set ambitious targets for providing social services. In Trinidad and Tobago, for instance, the changing government responsibilities increased the number and responsibilities of service providing ministries (see Table 1). While there were continued demands to provide services to meet the diverse needs of a growing population, the expansion in the State's role gave rise to opportunities for politicians to use the growth in government for political patronage. In Bangladesh, for instance, this led to a mushrooming of government, with a proliferation of agencies to offer ministerial positions and more jobs for political considerations (see Box 7). The result of similar patterns in many countries has been that governments have become overextended, trying to do too much and doing so inefficiently. Services have not been delivered or have been delivered badly. A survey of domestic businessmen in 31 developing and transition economies revealed that over 70 percent rated the delivery of government services as being mostly to very inefficient. The same pattern holds across many services. 22 Box 7: Bangladesh - The Mushrooming of Government Since independence in 1971, the Government of Bangladesh has virtually doubled the numbers of ministries, departments and officials. The number of ministries increased from 21 to 35 over the last two decades, while the number of departments and directorates more than doubled over the last decade, from 109 in 1990 to 221 in 1994. Employment in the public sector increased from about 450,000 in 1971 to almost one million in 1992, i.e., at a compounded rate of 3.6 percent per annum, compared to the population growth rate of 2.5 percent during the same period. Civil service pay has declined considerably, especially for those at the top. The basic pay of a Permanent Secretary has declined by 87 percent in real terms over the last three decades. New ministries, divisions and departments were created in part to meet emerging needs such as environmental concerns and women's issues. But the State has also spread its wings more and more into commercial economic activities. Indeed this growth has often been stimulated by political considerations. The increase in ministries accommodated more intra-party groups and offered more ministerial positions, and it created more jobs to be dispensed by political leaders. Aside from its budgetary impact, expansion has stretched implementation capacity, compounded coordination problems, and exacerbated regulatory intrusiveness. It has also created vested interests that have blocked efforts at rationalization and reforn. Source: The World Bank (1 996a) Power system losses in low-income developing countries are over two times those in other countries. In China, almost I million hectares have been taken out of production since 1980 because of insufficient maintenance. Residents of Jamshedpur, India, undertake costly investments in tubewells and digging wells to avoid dependence on the unreliable public water system. In Uganda, a recent tracking exercise revealed that a significant percentage of funds allocated for basic social services never reached schools' health clinics, and there was no discernible improvement in service delivery. Public delivery systems today face growing external pressures for better performance - from citizens, markets and global competition. Continued fiscal pressures have created demands for better value-for-money, for greater efficiency, and for effectiveness in service delivery. These pressures have provided incentives for politicians to improve delivery. The Problem of Managing Agencies: Balancing Flexibility with Restraints The problem to be solved is that of better managing the principal-agent problems inherent in delivery. Of necessity, politicians delegate policy implementation and delivery to bureaucrats. However, there is nothing to guarantee that bureaucrats (agents) will act in accordance with politician's (principal's) goals. They can subvert (substitute private goals for collective interest) or shirk (exert inadequate effort) - a danger, in short, that the agent will behave opportunistically. To minimize opportunism, the principal needs to devise and enforce contracts that produce a high degree of compatibility between the incentives to which agents actually respond and those that can elicit behavior in line with 23 principal's objectives. Because the principal cannot possibly monitor all the actions needed for implementation, it has to create mechanisms that help ease the task and reduce the costs of monitoring and thus enable it to address what otherwise would be an unmanageable problem. At the same time, the problem lies with the principal as well. Precisely because it is difficult to monitor what happens at the micro/project level, politicians have incentives to intervene on their own behalf: if it is difficult for politicians to monitor bureaucrats, then they will also have difficulty monitoring what their fellow politicians do individually. And they do. This creates credibility problems with bureaucrats, the agents, and takes away legitimacy of any contract with the principals. Bureaucrats thus feel less compelled to achieve the objectives set forth by politicians. Services have been delivered poorly in developing countries because neither the principals nor the agents have lived up to their side of the implicit bargain. The difficulty has arisen in essence because of poor enforceability of implicit or explicit contracts within the public sector. The central problem has been poor incentives, stemming from a lack of both flexibility and restraints. Services have been delivered by public-sector organizations, operating under tight command and control systems of the civil service. Public-sector managers have had insufficient managerial flexibility and there has been considerable political interference in day-to-day operations. At the same time, there has been few restraints in the form of probity or accountability for results. Indeed, public agencies have often operated with monopoly power in the market, with few external pressures for contestability or better performance. The reason service delivery has been poor can be seen by setting out the whole picture for solving the agency problem (Manning 1996). As shown in Figure 5, services can be delivered through a range of organizational forms operating in different incentive environments. The public sector itself can be seen as a series of different incentive environments within which the tasks of government can be performed. In most settings, the civil or public service lies at the center of the public sector. This may have constitutional control bodies at the very center, with line Ministries surrounding it. The civil service is, by tradition, a disciplined body of people who have agreed to accept more controls on their behavior in exchange for additional job security. The broader public sector is distinguished from the civil service by the relative flexibility of the financial management regime, and the greater freedom available to managers in recruitment and promotion. That broader public sector may contain a large number of special purpose agencies and, at its outer limits, the state-owned enterprises. Beyond the public sector, lies the domain of markets and civil society. Services can be delivered by for-profit, non-profit or community organizations. Service delivery can be located in these outer boundaries depending upon the nature of the outputs and the mechanisms by which their delivery can be managed by the public service. 24 Figure 5: Incentive Environments and Organizational Forms for Delivery \ \ * agencies /~~agccie \~~~~~~~raepbi secto for-profit / m7arkets and the pn'vate sector Source: Manning (1996) Put broadly, these organizational contexts are better or worse at providing an environment in which the civil service or public service, acting on behalf of the government, can be sure of obtaining the specified results from the service providing agency. It follows that improving service delivery entails locating service delivery in the right context and regulating that context to ensure that there are appropriate incentives for delivering the services efficiently and effectively. As shown in Figure 4, providers responsible for services that are hard to specify, budget-financed and with a constitutional watchdog role (e.g., Ombudsman, Court of Accounts) generally require the focus on inputs and processes found in the public service. By contrast, easily specified services in contestable markets can be located successfully in the private sector. Reforms - particularly in the developed countries - have sought to move delivery away from the center of circle, by commercialization through user fees, corporatization, privatization, greater contestability, and arm's-length contracts with private and public-sector organizations. New Zealand provides the most dramatic example of a systematic set of reforms designed to push delivery to the outer limits of the circles, with formal contracts to ensure technical efficiency. Starting in the early 1980s from an expansive public sector not dissimilar to that of many developing countries today, the Government broke up conglomerate ministries and created agencies with focused tasks. Agencies producing commercial activities were hived off, corporatized, and mostly privatized. 25 The Government introduced competition in markets and relied extensively on contracting out to the private sector using competitive mechanisms. For the activities that remained in the core public sector, the government entered into formal contracts with chief executive officers on fixed-term, performance contracts, while providing them autonomy in the management of financial and human resources. These reforms helped transform the budget deficit of 9 percent of GDP in 1983 into a surplus in recent years, and unit costs of services delivered are reported to have been reduced by over 20 percent in a sample of agencies (Campos and Pradhan 1996). But what is feasible in New Zealand is beyond the realm of possibilities in many developing countries. It requires considerable capacity to write and enforce contracts, especially for outputs which are difficult to specify precisely. It is an open question whether the way to achieve better performance is through formal contracts. Be that as it may, how far countries can push activities to incentive environments in the outer circles will depend upon the nature of the outputs and the capacity of the country to enforce implicit or explicit contracts. Figure 6: Incentive Environments and Service Delivery More Strengthen possibiities ~ ~~ voice possibilities mechanisms \ \ ~~~~~broader // \ publi~~~u c sector/ Improve compliance and loyalty Source: The World Bank (1997b) 26 Improving Delivery: The Menu of Institutional Alternatives Faced with growing pressures and demands to improve delivery, governments in developing countries are adopting and exploring a range of institutional alternatives. In general, these can be seen as moving to the outer circles, subjecting service delivery to different sets of incentive environments depending upon the nature of the activities. The menu of institutional alternatives can be classified into three broad categories - strengthening exit, voice, and compliance-loyalty - that apply in varying degrees in different incentive environments through which services can be delivered (see Figure 6). Markets and contracts with the private sector offer primarily exit options for better performance. The broader public sector - including corporatized public enterprises and performance-based agencies - offer fewer exit options, but voice begins to become influential. For activities that are difficult to specify and not contestable, the core civil service provides no realistic exit options and has considerably less flexibility in financial and personnel management. As exit completely disappears as an option, incentives switch to voice and improved rule-based compliance and loyalty of the civil service. DIFFERENT INCENTIVE ENVIRONMENTS OFFER DIFFERENT OPTIONS FOR IMPROVINAG DELIVERY. The appropriate mechanism for improving performance depends on both the characteristics of the service and the capability of the State to enforce internal and external contracts (see Table 2). For instance, for services that are contestable - such as commercial products and, more recently, telecommunications and power generation - market mechanisms can generate powerful competitive pressures for improved delivery. For services whose outputs the State can specify and enforce at low transactions costs, contracting out to private firms and NGOs is an attractive option. Countries with strong capacity and commitment are setting up performance-based agencies and formal contracts even for complex activities within the core public sector (defense, education, health care). But countries with scarce capacity to enforce complex contracts and weak bureaucratic controls to restrain arbitrariness under more flexible management regimes need to proceed with caution. Improving Delivery through Markets and the Private/NGO Sector STRENGTHENING AND CREATING MARKETS FOR SERVICE PROVISION. Changes in technology have created new scope for competition in markets where none existed before. This has created significant opportunities for improving the efficiency of services provided. Deregulation in telecommunications and power generation has led to significant reduction in unit costs and a rapid expansion in services provided. At the same time, regulation becomes important to ensure that the erstwhile monopolist does not abuse its market power and that new providers are able to interconnect with the incumbent provider on reasonable terms. 27 Table 2: Mechanisms to improve delivery Service characteristics and State capability Contestable Easy to specify and enforce Difficult to specify and enforce Markets and * Strengthen markets: * Contract out to for-profit or credible regulation nonprofit agencies B.. the private * Create markets: sector vouchers + Enhance internal * Set up performance-based competition agencies Broader + Create hard budgets + Corporatize and establish public and divest State firms enforceable performance B contracts sector * Strengthen voice mechanisms Core 3 f 83-+Ensure clarity of purpose and task public * Improve rule compliance sector * Increase loyalty 3 Strengthen voice mechanisms Source: The World Bank (1997b) Recent innovations such as vouchers and capitation grants have greatly increased the scope for competition in the provision of social services. Experience with these mechanisms is confined to education (see Box 8). Chile has a capitation grant system through which students can enroll in any school, which then receives government financing according to the number of students enrolled. The short experience has shown an increase in private enrollment, but there is no clear indication yet of what happened to schools' performance. Colombia has also introduced a voucher system for secondary schools. Kenya has established a workers training fund on the basis of vouchers and a contract scheme that introduces competition among public and private providers. A similar scheme operates in the Philippines. 28 Box 8: Vouchers and School Choice in the U.S., Netherlands, and Chile One solution to the problem of poor-quality schools has been widely discussed in the United States is the school choice or voucher system. Under this system, individual students would be given vouchers funded by public tax dollars but redeemable in either private or public schools. Schools would then have to compete for student patronage. Opponents of the voucher system predicted that it would lead to a mass exodus of public-school students - especially the better ones - that would gut the public system. Yet in a 1993 pilot school- choice program in Puerto Rico, the 18 percent of participating children who did transfer to private schools was largely offset by another 15 percent who actually transferred from the private to public schools - hardly the mass exodus predicted. Puerto Rican experiment has been so successful, moreover, that by its second year the number of applicants had jumped from 1,600 to 15,500. Public funding for private schools is nothing new. In the Netherlands, two-thirds of all students attend publicly funded private schools. While Chile reformed its education system in fiscal 1980, moreover, the Ministry of Education began providing per-student payments in both public and private schools. Because the payment was based on the average cost of education in the public sector and expenditures per student were 70 percent less in the private sector and private schools vied eagerly for students. By 1986, although primary enrollment in private schools had more than doubled, going from 14 to 29 percent, enrollment in private secondary schools increased almost four-fold. Source: van der Gaag (1995) Several potential risks have been mentioned with vouchers, and the jury is still out on their effectiveness. The most common is an increased stratification among services and polarization among users. For instance, some analysts have criticized the unfettered rush towards school choice in the former Soviet republics, arguing that it will exacerbate social tensions in polarized societies (Heyneman forthcoming). The underlying concern is that in the absence of national controls, school curricula will be divisive and parochial, and an essential role of the State - social cohesion - will be undermined. In each of the above cases, however, achieving positive results requires an effective regulatory capacity. This is not always easy to achieve. For the social sectors, the difficulties of regulation are even more daunting than in infrastructure subsectors. The private health-care delivery system in many developing countries, for instance, is not regulated, either because of capacity constraints inherent in regulating a large number of small-scale providers (individual doctors, including "quacks") or because of denial that the private sector is particularly large. Consequently, few developing countries - Brazil is a notable exception - have established regulatory mechanisms for private providers, but the situation is changing. CONTRACTING OUT TO THE PRIVATE SECTOR AND NGOS. In developed countries, contracting out of services to the private sector has become a prominent phenomenon. Generally, initial resistance to contracting out from public-sector staff has been high, and some countries have chosen to follow a policy of "compulsory competitive tendering" to expose service providers to external competition. The case of Victoria State in Australia provides a particularly dramatic example, where each local council is required to subject at least 50 percent of its annual budget to competitive tender (see Box 9). 29 Box 9: The Contract State - Victoria, Australia In Victoria State, Australia, each local council is required to identify services totalling at least 50 percent of the annual budget and subject them to competitive tender by 1997. Within this overall requirement, the state government has adopted a guidelines driven approach with the details of key decisions concerning the criteria to be used in awarding contracts and the rate of return required of successful in-house bids left to the council. However, the legislation does prescribe how the tendering process is to take place and limits the duration of contracts to 5 years. There is comparatively little extemal monitoring that councils are adhering to the competitive tendering requirements. Self-reporting is emphasised, with councils required to submit an annual competitive tendering statement to the state government each year. However, audit requirements have been changed, and external auditors must confirn that the requirements of the legislation have been satisfied. There is considerable scope within the legislation for the state government to adopt a more directive approach if it considers this necessary. This approach of setting a target for the proportion of services that are to be subjected to competitive tendering while leaving the detailed implementation in the hands of the council differs strikingly from the approach adopted in the UK. Compulsory competitive tendering legislation introduced in the UK defined lists of local authority services that were to be subjected to competitive tendering within a centrally prescribed timetable and within tightly defined tendering procedures. However, the UK compulsory competitive tendering is still largely, although not entirely, targeted at the physical services of Local government (cleaning, grass-cutting, catering). By contrast, it is probable that the approach adopted in Victoria will impact on complex community care services at an early stage. Source: Manning (1996) This will inevitably lead to competitive tendering for complex community care services that constitute a principal domain of council activity. The general success of contracting out for easily specifiable services in developed countries has led to increasing confidence in subjecting services to competitive tendering. For instance, as a result of the Competing for Quality Initiative in the UK, between 1992 and 1996, services to the cost of $6 billion had been subjected to market-testing with consequent annual savings of $1 billion and a reduction in civil service numbers of 20,000 (Manning 1996). In developing countries where both markets and state capacities are weak, the options for contracting are not as pervasive as in the developed countries. Nevertheless, where outputs are relative easy to specify and direct competition is not possible, efficiency can be increased by competition managed through contractual arrangements, ranging from service contracts to management contracts to leases to long-term concessions. Thus there is competitionfor the market even though there is no direct competition in the market. Contracting easily specifiable activities such as road maintenance, building construction, or food catering can lead to potential savings. In Brazil, contracting road maintenance to private contractors led to 25 percent lower costs than using government employees or force accounts. Leases have increased technical efficiency of water supply in Guinea and the operation of Malaysia's Port Kelang. In Cote d'Ivoire, a concession contract for SODECI, a private-sector company, has enabled rapid expansion in access, including for the poor and recovery of full costs (World Bank 1994). 30 Box 10: Contracting NGOs for Better Schooling in Bolivia In an experimental program in Bolivia, the government contracted the church-based Fe y Algeria to manage a certain number of public schools, mostly at the secondary level. Before agreeing to do so, Fe y Algeria demanded and received the right to appoint principals and teachers and to allow teachers to work both the morning and afternoon shifts - rather than the 3.5 hours allotted for instruction in public schools. In all other ways, Fe y Algeria schools are identical to other Bolivian public schools. Although hand- picked, their teachers are ordinary normal-school graduates, who receive little special training and are paid the same salary as other public-school teachers. Fe y Algeria schools receive no additional money for books or supplies, and their curriculums and teaching methods are the same as those used in the public schools. The only comparative advantage Fe y Algeria schools can point to is an exceptional esprit de corps among students, parents, and staff. For both teachers and students flock to Fe y Algeria schools, despite their similarity to public schools, with many families paying extra school fees to attend them. On the rare occasions when innovative teaching methods have been tried, such as a math course transmitted over a public radio station, they have also proved very popular, both in the schools and throughout the community. This public-private partnership between government and a religiously-based non-government organization appears to be so successful that the government is studying it as a possible model for national education reform. Source: van der Gaag (1995) Recognizing their poor performance and fiscal constraints, govemments are also contracting out the delivery of social services to the private sector and especially NGOs. In Bolivia, for instance, contracting schooling to a local church is producing promising results (see Box 10). In Uganda, the government is contracting out not only the provision of curative health services to local NGOs, but it is also relying upon NGOs to deliver preventive health care which is normally within the public domain (van der Gaag 1995). Like regulatory contracts or the management of vouchers, contracting out is not a panacea. Contracting is particularly suited primarily for services whose outputs are relatively easy to specify and where there is a strong market so that the effectiveness of alternative suppliers can be readily judged. Contracting out of activities which are complex, individualized and non-routing will inevitably incur higher transactions costs than those which are straightforward and repetitive. Further contracting is prone to corruption, waste and mismanagement, much as internal contracts within the public sector are. For instance, the Inspector General of Uganda points to several instances of corruption and fraud in the administration of contracts with the private sector, including payments for roads that were not built or maintained. Once again, then, the capacity to effectively write and enforce contracts by the public service becomes inescapably important. Improving Delivery through the Broader Public Sector Even though markets and contracts with the private sector provide opportunities for improving delivery, a number of services will inevitably remain in the public sector - whether operated through the civil service or through state-owned enterprises. Among other things, this is because several services - such as defense and primary education - produce outputs that are difficult to specify and risky to contract out. Further, the strength 31 of markets varies across countries and sectors, and workable opportunities for using market-type mechanisms may not and have not been available in several cases. This implies the need for improving internal incentives for better delivery of activities that remain within the public sector. Where politicians have faced fiscal pressures and pressures from clients, they have tried to set up internal mechanisms to strengthen incentives for performance. Broadly, these can be grouped under four categories: (i) creating internal "exit" mechanisms through internal contestability (e.g., administrative decentralization, internal markets) and contracts with public-sector agencies; (ii) strengthening "compliance-loyalty" for better performance; and (iii) strengthening "voice" mechanisms. Each is discussed in turn. INTEPNAL CONTESTABILITY. Even for services for which external market competition has not been feasible, there have been recent attempts to create internal markets to provide stronger incentives for performance. Most budget-funded agencies have traditionally received services such as property and printing from a central body at no cost to themselves. Consequently, they have no incentives to economize on quantity or quality, while the service provider has been happy to lobby on behalf of its citizens. Several countries have introduced user charges for these services and changed the rules of the game dramatically by this one single reform. The supply of services is now determined by the amount that the consuming agencies are prepared to pay for. A number of countries have gone further in attempting to create internal markets. Consuming agencies had no choice but to take their services from the central supply bodies. This kind of requirement was aimed at achieving economies of scale and using combined purchasing power to negotiate favorable terms of purchases. To improve efficiency and quality, several countries have moved to allow the purchase of services from alternative suppliers. The public-sector provider must compete directly against the private sector for public-sector business. This is the case, for example, of information technology services in Scandinavian countries. The creation of internal markets is also being pursued as an instrument of sectoral policy. An example in some countries is the health sector, where arrangements have been made to implement managed markets allowing competition and choice. PERFORMANCE-BASED AGENCIES AND INTERATAL CONTRACTS. Called the new public management, some reforms in industrial countries have sought to break up the core public sector into a series of distinct business groups or special-purpose agencies. In general, these agencies have clarity of purpose and task, greater managerial flexibility in the allocation of financial and human resources, and greater accountability for results. Sweden and other Nordic countries have long separated cabinet ministries from agencies with specific purposes. In France, many state services have recently been established as special- purpose agencies, or centres de responsabilite. The United Kingdom has moved nearly two-thirds of its civil service into executive agencies charged with specific delivery functions. New Zealand has broken conglomerate ministries into focused business units and separated policy from operations, purchaser from provider, and regulator from operator. 32 Box 11: Internal Market in the UK Health Service - The Purchaser-Provider Split An internal market has been created within the U.K. health service by transforming local health authorities and general practitioner groups into purchasers of hospital services on behalf of their patients. Competition has been established among hospitals as a decentralized mechanism for reallocating resources and a market-type link has been created between the effectiveness of the hospital and its access to public funds. Incentives are therefore linked to the perfonnance of the hospital managers and staff. Hospitals are given a great deal of autonomy as regards the techniques and approach they adopt for the management of all their resources. Prices charged to customers reflect all costs, including capital charges, and act as the markers for reallocating business among different suppliers of hospital services. The purchaser-provider split has been most successful where there is real competition, in areas with several hospitals competing to provide services and many general practitioner findholders shopping for them. Source: Organisation for Economic Co-Operation and Development (1995) These changes have been accompanied by substantial devolution of managerial authority and accountability for results. In Australia, Denmark, Ireland, and Sweden, detailed and itemnized administrative costs have been consolidated into a single-line budget item, making it possible for managers to reallocate resources in accordance with changing priorities and needs. Managers are also getting greater authority over human resources management; in New Zealand and Sweden, managers have the right to hire, fire, and set pay within overall bounds. Along with devolution of resources has come more stringent performance requirements. New Zealand has replaced permanent secretaries with chief executives on fixed-term performance contracts linked to the cost-effective production of outputs. Among developing countries, Singapore had a head start in creating focused business units. As early as the 1970s, the Singaporean civil service was organized around the concept of statutory boards. Elsewhere, Jamaica has selected eleven pilot agencies for conversion into executive agencies. But countries with inadequate controls over inputs and weak capacity need to proceed with caution. The industrial countries that have relaxed detailed control over inputs developed credible restraints on arbitrary behavior over a long period of time. For the many countries that have not succeeded in instituting credible controls over the use of inputs, providing greater managerial flexibility will only increase arbitrariness and corruption without commensurate improvements in performance. Also, writing and enforcing contracts, particularly for complex outputs, requires specialized skills that are scarce in many developing countries. A recent study of public enterprises found the overall record of performance contracts in developing countries to be disappointing (World Bank 1995). In particular, there were few credible rewards or sanctions for performance. Managers operated under government-imposed constraints and political interference that seriously limited their autonomy. Governments not only controlled the prices of factors and products and decisions about wages and layoffs, they also strongly influenced whether an enterprise was paid by its customers. 33 Box 12: A Performance-based Agency for Tax and Customs Management in Ghana A performance-based agency approach to tax and customs administration was first used in Sub- Saharan Africa in 1984, when Ghana established of the National Revenue Secretariat. The secretariat was established because tax revenues had dropped drastically, partly because a progressive depression of salaries had drained out high-caliber staff. Staff who remained lacked motivation and performed their tasks in a perfunctory manner. Incentive systems were poor, corruption and absenteeism were common. and the Public Service Commission's hold on recruitment and dismissal often constrained management. The first step in reforming tax and customs administration involved the creation of two new services: the Internal Revenue Service, responsible for direct taxes, and the Customs, Excise, and Preventive Service, responsible for most indirect taxes. Both agencies are headed by commissioners appointed by the cabinet, with governing boards consisting of a chairman and six cabinet appointees. The controller and accountant-general sits on both boards. The boards are responsible for the collection of all taxes, policy- related management, and all appointments, promotions, and disciplinary action. They draw up service schemes that prescribe the terms and conditions of service and guidelines for remuneration. A number of measures were adopted to sustain a performance-based culture. Most important, the revenue services were instructed to set targets (agreed with the Ministry of Finance), the attainment or non-attainment of which govern incentive packages. In line with the general belief that low pay feeds corruption and undercollection, the basic salary levels of all staff were raised to levels comparable to those in the private and banking sectors. Professional staff at the Ministry of Finance - alone in the civil service - now get an additional inducement allowance of 12 percent. Revenue officers in the National Revenue Secretariat get an additional 22 percent. In addition workers that achieve agreed targets get a 15 percent bonus on their basic salaries. The National Revenue Secretariat meets its operational expenses by retaining a small percentage (on average, 3 percent) of revenues collected and has been able to improve working conditions for its staff. At the same time more, than 250 corrupt officials have been dismissed, while others were retired. This approach appears to have paid off, but not without problems. Between 1984 and 1988, tax and customs revenues increased from 6.6 to 12.3 percent of GDP, mainly because of better collection. But the rest of the civil service chafed at the special treatment afforded tax collectors, and the Ministry of Finance objected to its loss of authority. The program could not have gone forward without strong support from the top. Source: Dia (1995) While performance contracts have not been successful, developing countries have sought to create performance-based agencies for easily specified and high-priority tasks such as road maintenance or tax collection. These agencies are typically set up as enclaves within the civil service, with greater managerial flexibility, better pay, and greater accountability for results. In Sub-Saharan Africa, for instance, performance-based agencies have been created to achieve tax collection targets in Ghana, Uganda, and Zambia (Dia 1995). Other countries appear ready to follow suit. Enclaving tax collection has been viewed as a prerequisite for improving a government's capacity to raise revenues (see Chapter 3) and improve incentives for the rest of the civil service. Results have been impressive, but the enclaves are not without problems (see Box 12). While tax revenues have increased, the rest of the civil service has bristled at the benefits provided to the tax collectors. 34 In aid-dependent countries, donors have set up project enclaves with their own systems of remuneration and accountability. Often, these enclaves have not been based on any systematic consideration of the nature of the services provided or of the optimal sequencing of institutional reforms. Such enclaves have often created disparities. Enclaves have often been designed as quick fixes. Though they have sometimes accomplished short-term goals, they can create obstacles to deeper institutional reform. Enclaves can be useful as an experimental stage of reforms to be progressively extended, and as a demonstration that reforms can be effective. Even then, it is important to ensure that systematic criteria are applied in hiving off agencies, such as the ease with which outputs can be monitored in exchange for greater managerial flexibility. Hiving off agencies that produce easily specified outputs, such as tax collection, can be a useful starting point in a sequence of long-term institutional reforms. But it is no substitute for the longer-term reform needed to create a motivated, capable civil service. Improving Delivery through the Core Public Sector Contracting out, setting up performance-based agencies, and ensuring formal accountability for results are not viable options for many services in countries with weak capacities. The challenge is particularly acute for street-level bureaucracies whose operators, such as police constables, irrigation patrollers, health workers, teachers, and extension workers, interact daily with citizen-clients, are geographically dispersed, have substantial discretion, and produce outputs that are difficult to monitor and are not subject to competitive pressures. Countries have adopted a variety of approaches to improve delivery of such services. Some have emphasized more clarity of purpose and task and greater predictability in the flow of resources, along with systems of monitoring compliance with rules or procedures (including accountability for the use of financial resources). To lower the costs of monitoring and enforcement, some systems have emphasized building loyalty and esprit de corps to achieve a better congruence between individual and organizational goals. In others, pressure from clients and beneficiaries has been seen as a cost-effective monitoring device for ensuring performance. Experience suggests that a combination of mechanisms can be set up to improve incentives for performance. A study comparing irrigation agencies in India and Korea revealed that the Indian organization provided few incentives for people to work conscientiously, while the Korean organization was full of them (Wade 1994). Korean irrigation patrollers had greater clarity of purpose and task and were subject to random monitoring from three separate agencies. Supervision techniques in India sought to discover grounds for punishment; in Korea, they sought to solve problems. With staff from all parts of the organization traveling frequently up and down the canals in Korea, there was more external pressures from farmers and stronger partnerships for better performance. The successful training and visit system in agricultural extension operates on a similar principle. 35 Performance-orientation and predictability in resource flows. Even where formal performance contracts are unfeasible or have proved unsuccessful, it is possible to introduce greater performance orientation in public-sector organizations. An important starting point is greater clarity of purpose and task. Colombia, Mexico, and Uganda are introducing performance measurement to orient managers to achieving desired results. While some countries (New Zealand is one) have emphasized goods and services produced (outputs) as performance measures, others (Australia, Colombia, and Uganda) are emphasizing the impact of outputs on beneficiaries (outcomes) and ex post evaluation. But even with greater clarity of purpose and task, public-sector managers will be unable to perform effectively if they face too much uncertainty about the flow of budgeted resources. In many developing countries, the budget as approved bears little resemblance to the budget as implemented. The problem starts in a policymaking process in which there is a mismatch between the resource requirements and availability. This leads to problems downstream, when overextended government departments and services are starved of necessary operations and maintenance funds. A credible medium-term expenditure framework - such as in Australia (see Box 5) - provides greater consistency between policies and resources and therefore more predictability in the flow of resources. Malawi and Uganda are in the initial stages of instituting such a framework. Several Latin American countries are planning or undertaking measures to leverage enhanced predictability during budget execution for improved performance. On a pilot basis Ecuador plans to guarantee greater reliability in resource flows. In return, managers will report on performance with the understanding that these incentives can be withdrawn if results are not forthcoming. Bolivia and Nicaragua are undertaking similar reforms. FINANCIAL AND IMNAGEMENT CONTROLS. Since public-sector outputs are often difficult to measure and monitor, financial control and accountability are needed to ensure probity, prevent the poor use or abuse of public resources, and improve service delivery. An expenditure tracking exercise in Uganda revealed that a significant portion of funds allocated for basic social services never reach health clinics or schools, particularly in rural areas. In many countries, financial accounts and audits are outdated and inadequate and therefore do not provide credible restraints. Furthermore, too many detailed controls can stifle managerial flexibility and tempt managers to circumvent controls and use nontransparent mechanisms, prompting the imposition of new centralized controls. Sound financial control and accountability start with transparency in the use of public resources. Examples of mechanisms that enhance transparency are accounting systems that meet specified functional and performance standards, audit trails to identify individuals using resources, competitive procurement of inputs, periodic reporting and auditing in accordance with specified requirements, and formal review by independent bodies such as an auditor-general and parliamentary committees (Reid 1996). To improve the transparency and quality of their financial accounting and auditing systems, countries are modernizing their financial information systems to reduce lags and 36 improve the reliability of financial accounts and audits. Some countries are setting into law the principles of accounting practices, backed by strong professional associations inside govemment and the private sector. . Argentina has modernized its financial information system and is posting quarterly financial reports on the Internet to demonstrate its commitment to transparency. * In China, new national accounting and auditing standards are being developed based on internationally accepted standards. * Standards and laws are being complemented by professional associations of accountants and auditors in Kazakstan and Moldova, which have active Accountants and Auditors Associations. * In Indonesia, the quality of university faculties in accounting is being strengthened through advanced degree programs abroad as well as in-country training. * Several countries in Latin America are adopting integrated financial management systems that improve the quality of financial management inforrnation by integrating information from budgeting, accounting, cash and debt management, and internal and external audits. Bolivia has established a fully integrated national system of accounting. A review found government audits to be of better quality than those of several private firms. Although systematic evidence about the effects of these reforms is not available, some useful policy lessons can nevertheless be gleaned. Modern, computer-based information systems can improve both transparency and strengthen aggregate controls while reducing the need for transaction-specific controls (Reid 1996). Experience from some Latin American countries suggests that automating data entry and retrieval can enhance timeliness and reliability, facilitate consolidation of accounts and public dissemination, and ease the task of generating reliable audit trails by building appropriate checks and balances into the software itself Input controls can be broadened sequentially as systems and trust are built up, moving away from detailed, ex ante line-item controls to broader budget categories, broader-band salary scales, and greater latitude in procurement and recruitment. The shift from ex ante to ex post transaction-specific input controls should proceed carefully and only as controls over budgetary aggregates are improved, as executing agencies demonstrate that they can be trusted with greater autonomy, and as ex post controls build up through strengthened accounting and auditing capacities. But experience also suggests that moving from a highly centralized, transaction- specific, control regime to a more decentralized one can encounter resistance. For instance, in Ecuador, a 1995 plan to devolve payment controls has yet to be implemented, largely because of central agencies' fears of fiscal lack of discipline (Reid 1996). Trust needs to be built up first by strengthening systems of performance measurement and ex 37 post input controls. As the systems become more credible, there will be less resistance to the changeover, and managers can be provided greater flexibility with stronger accountability for results. LOYALTY, MOTIVATION, AND COMPETENCE. Better systems of monitoring, accounting, and auditing alone may not improve the delivery of many services. Mechanisms for enhancing the loyalty, motivation, and competence of the civil service are also needed. Loyalty promotes staff identification with an organization's goals and a willingness to take a longer-term view of responsibilities. Loyalty is essential in the core public sector, where activities are not easily specifiable or able to be monitored, and exit has no meaning. The U.S. Forestry Service has worked to build a sense of commitment to shared objectives among its staff, a quality deemed critical to effective performance given the isolated nature of the job, the geographical dispersion of the forest rangers, and the lack of specificity of outputs. More generally, the civil services in France, Germany, Japan, and Singapore all sought to ensure that the loyalties of a small group of people were fundamentally aligned with the interests of the state. Job security in the civil service was meant to foster this alignment of interests. This notion has worked in some contexts but not in many others. As discussed in the next section, the experience of successful countries suggests that building this commitment and motivating and attracting capable staff require long-term career rewards, adequate pay, and mechanisms to instill an esprit de corps. Strengthening "Voice " Mechanisms: Pressures from Clients Where clients or beneficiaries cannot "exit" or opt for other market alternatives, they can exercise pressure or "voice" to improve performance. Voice is seldom exercised at the individual level; rather, it takes different forms of collective action, from simple activities organized by local NGOs to active community participation. But this is not without cost to the clients in terms of time and political capital or of the risk of voice being captured by special groups. Citizens have exercised "voice" through various mechanisms. In this section we identify three mechanisms which are potentially powerful in exerting pressures for better delivery: (i) community participation and management of service delivery; (ii) client surveys to obtain transparent feedback on the quality of services provided; (iii) benchmarking and citizen charters to bind the government to publicly stated performance targets for service delivery (Organisation for Economic Co-operation and Development 1995). At the same time, we highlight some of the risks of capture. COMMUNITY PARTICIPATION AND MANAGEMENT. If there is a way of effectively getting the community involved in the service provision, even indirectly, it will create a stronger incentive for pressing for better services. For example, household participation can help improve school performance (see Box 13). The social investment funds in Latin America foster community participation and induce community financing of social infrastructure. In Brazil, public funding is conditional upon a level of community financing. 38 Box 13: Household Participation to Improve School Performance Accountability and performance of schools can be enhanced if households are more closely involved in the schools that family members attend, Most households already contribute, directly or indirectly, to the costs of education, but they can also participate in school management and participation. In many countries, parents and communities are becoming more involved in the governance of their children's schools. Legislation enacted in 1993 in Sri Lanka established school development boards (SDBs) with the purpose of promoting community participation in school management. Each SDB consists of representatives from the school staff, parents, and past students, and it is chaired by the school principal. The elected boards of trustees that manage schools in New Zealand are drawn from parents of children at the school. In Mauritius, parent-teacher associations have been so successful that government funds are being used to stimulate the partnership further. Many countries have found that communities that participate in school management are more willing to assist in the financing of schooling. Jamaica has set up a major program to stimulate this tendency; Bangladesh's Social Mobilization Campaign, which involves the community in education, has been accompanied by a reactivation of school management committees throughout the country. El Salvador has started to involve communities in rural school management, with significant results in improved teacher attendance. Student achievement levels are comparable with those at traditional schools, though the students tend to come from poor backgrounds. But effective involvement in school governance does not come easily or without effort. New Zealand realized after it had embarked on its reform that intensive training was required for the newly elected parent trustees. Jamaica is training parents to help manage schools. Botswana found it very difficult to attract sufficiently qualified people to lower-secondary school board of governors, especially in rural areas. Training is often required and can be effective both in literate communities, such as New Zealand, and in relatively illiterate ones, as in parts of Uganda. Action Aid in Uganda is providing community training in two districts for parent-teacher associations and school management committees. Source: van der Gaag (1995) More generally, community or user delivery of services is most common for local, small-scale infrastructure - such as rural feeder roads, community water supply and sanitation, and maintenance of local drainage systems - and it often complements central or provincial services. Successful community delivery mechanisms requires user involvement in decision-making, especially to set priorities for expenditures and to ensure an equitable and agreed sharing of the benefits and costs of service delivery. When these elements are present, community self-help programs can succeed over long periods. A community organization in Ethiopia devoted to maintaining roads (the Gurage Roads Construction Organization) has worked well since 1962 because it sets its own priorities and allocates its own financial and in-kind resources. CLIENT SURvEYS. Feedback mechanisms such as client surveys have the advantage of providing information about the performance of the agency vis-a-vis its clients among the general public. Not only do they increase transparency by making a larger number of people aware of the agency's performance, but they also improve accountability since they implicitly exert pressure on the agency to perform. 39 Client surveys come in a variety of shapes and forms. But they all share a common feature. An appropriately structured questionnaire is constructed and distributed to individuals from a random sample of clients of an agency. The individuals in the sample are asked to complete the questionnaire and the individual responses are collated and analyzed statistically. The results are then summarized and made available publicly. Surveys are useful for obtaining citizen feedback on a particular service being delivered by the government, e.g. garbage collection, utilities provision, issuance of licenses. Surveys effectively address a collective action problem. Individual citizens can complain about poor service, but this generally makes no impression on the agency concerned unless the individual is highly influential. Citizens are cognizant of this and are dissuaded from coming forward -the benefits of doing so are minuscule, but the costs are high. The result is no one does anything, and the poor service continues. A client survey helps aggregate the complaints (or for that matter, the praises) of individuals in a cost-effective way. Individuals need not spend much time to respond to the questionnaire. Hence, for many, the benefits of "complaining" through the survey questionnaire are likely to outweigh the costs. Consequently, the clients' collective voice is brought to bear on the agency, making it more costly for the agency not to respond. Client surveys are of particular interest to developing countries since the required technology and human resource capability is not very demanding. Recent experience with client surveys in Bangalore, India, Nicaragua, and Uganda suggests that this can be a potentially powerful mechanism for increasing the pressures to improve service delivery in developing countries (see Box 14). A VOIDING CLIENT CAPTURE. The idea of focusing on citizens as customers, although attracting many adherents, is more complex than it first appears. The achievement of service quality objectives is but one aspect of performance measurement, and the recipient of a particular service is only one of many stakeholders. The interest of the general taxpayer has to be given due consideration. The pursuit of higher quality and responsiveness also must be weighed against its cost. More broadly, the public service is not directly accountable to its clients. Accountability to the public is indirect through government and the Parliament. Increased emphasis on performance as measured by client assessments may increase the influence of particular client groups. The risk is greater when interest groups are well organized in lobbying on their own behalf Consequently, a balance needs to be achieved between client responsiveness and client capture. In developing countries, however, with poor quality of service delivery and weak internal systems of accountability, client surveys and other systems of external accountability may well be a promising and powerful mechanism for reducing enforcement costs and improving performance. 40 Box 14: Client Surveys to Motivate Service Improvements - India, Uganda, and Nicaragua In several countries, client surveys have helped motivate processes that have resulted in improvements in public-sector performance. By tapping the real experience of citizens, who are the "clients" of public services, and involving them in the process of monitoring and evaluating the services, priority problems have been agreed on and innovative solutions have been designed and implemented. Examples from India, Uganda, and Nicaragua, although facing quite different contexts, illustrate how client surveys have worked. In Bangalore, India, "report cards" ask citizens and businesses to grade the public agencies they had to deal with to solve a problem or to get a service. The report cards, undertaken by the NGO The Public Affairs Centre Bangalore, address not only the quality of citizens' interactions with public agencies but also the "costs" of that interaction. In the first report card, the Bangalore Development Authority (BDA), responsible for such services as housing, scored the lowest in several categories, including staff behavior, quality of service, and information provided. Only 1 of 100 respondents considered BDA's services satisfactory. PAC held seminars to disseminate the findings. Instead of viewing this as a threat, the head of BDA took this as an opportunity. He launched a citizen-govermment initiative, Swabhimana, to address the delivery problems. Other agencies in Bangalore have also taken action based upon the report. Groups in five other Indian cities, including Bombay, have started their own report cards. In Uganda, it is the government itself - rather than an NGO as in India -that has undertaken to carry out Service Delivery Surveys (SDS). But it is carrying out the task in collaboration with NGOs and communities. A local NGO worked with government counterparts to design and implement the surveys. The first Uganda SDS showed, for instance, that only 11 percent of households in rural areas had ever been visited by an agricultural extension worker. Several districts have incorporated the SDS findings into their official district plans. One district has instituted further training for extension workers and is actively lobbying central government for permission to spend more of its budget on additional extension workers. The Nicaragua SDS, as in Uganda, is "owned" and initiated by the government. As in India, Nicaragua is an example where the use of the press has helped build consensus on the problem. For instance, the first SDS in 1995 revealed that 14 percent of bus riders suffered from an assault on the bus. It also showed that 90 percent of the bus drivers did not respect the official fare of 85 cordobas by not returning the 15 cordobas of change to riders. The survey also revealed that people were willing to pay the additional 15 cordobas. Based upon these findings, the Vice Minister took the problem to the people, arguing that users were being robbed. In addition, the fare was raised to $1.00. A follow-up survey a year later revealed that 90 percent of the users said that the official rate is respected. Source: Langan (1996) VI. Capable, Motivated Staff Whether it is policy making or delivery, whether it is direct provision or the administration of contracts - capable, motivated staff constitute the life blood of an effective public sector. Discussions on how to build a motivated, competent civil service usually focus almost exclusively on pay. Adequate pay is indeed crucial and indispensable. But also vital are some other elements, such as mechanisms to recruit and promote staff based upon merit and creating a sense of commitment for a shared purpose or mission. Indeed, as the experience of successful countries suggests, building competency requires three interrelated mechanisms: merit-based recruitment and promotion, adequate pay, and creating an esprit de corps. 41 Box 15: Citizen's Charter in Malaysia and the United Kingdom In Malaysia, Client's Charters have been developed for over 300 public-sector agencies. They are exhibited at strategic public locations in government offices and distributed in booklets. Agencies are responsible for setting their own service standards, with awards provided to encourage improvements. Charters set out the standards of service, explain how performance will be assessed against those standards, indicate how the public can provide feedback against those standards, and tell the public how to complain in the event of dissatisfaction. The UK Citizen's Charter, published in 1991, seeks to raise the standards of service in the Civil Service, the wider public service, and the privatized utilities. The Charter sets out principles and mechanisms to improve public services and make them more responsive to the needs of users. These include setting, monitoring, and publication of explicit standards; providing full, readily available information about services; providing choice where practicable and consultation with service users; a courteous and helpful service from accountable public servants; an apology, a swift and effective remedy if things go wrong; and value for money through efficient and economical delivery of public services. The Citizen's Charter has resulted in more privatization and contracting out, wider competition, and tougher, more-independent inspectorates. Initial measures included guaranteed maximum waiting times for certain National Health Services (NHS) treatments, fixed appointments for hospital out-patients, publication of league tables for school results, and new powers to protect consumers of public services. There are over 40 Charters covering the main public services. These include published Charters for parents, patients, rail passengers, court users and benefit claimants. The Charters have been criticized as rhetorical devices offering little redress in law. However, they have served to focus attention on service quality and have provided benchmarks against which managers must test the outputs for which they are responsible. Source: Organisation for Economic Co-operation and Development (1995), Manning (1996) In contrast, recruitment, promotion, and transfers in several countries are often motivated by political patronage and ties to the clientele. The civil service has grown rapidly to levels that cannot be afforded, while average pay has declined and the salary structure has become compressed. Consequently, corruption and malaise has set in, eroding esprit de corps and at times even a modicum of public-sector ethos. Merit-based Recruitment and Promotion Building a competent bureaucracy requires, among others, the establishment of a merit- based recruitment system. "Merit" is the operative word, for it is through meritorious criteria that competent applicants can be identified, filtering out the better ones from the applicant pool and acting as a signaling device. Such a system helps create prestige for civil service positions and thus makes it attractive to talented individuals. The obverse is a recruitment system based on favoritism. The latter creates a "lemons" problem for the civil service that in the long run leads to a personnel pool that is, on average, of poor quality, particularly relative to personnel in the private sector (Campos and Root 1996). In many of the industrialized countries as well as in the more economically successful developing countries, the basic merit system used is a national civil service entrance exam whose standards are tough enough to be able to delineate more skilled from 42 less skilled applicants (e.g., Japan and Korea).2 But this need not be the only way. In some countries, the primary filter is performance in school. For instance, in Singapore recruitment is based on high standards of academic performance and rigorous personal interviews. In others, it is the completion of a graduate degree from a reputable university (primarily from the West). In still others, it is some combination of exams, school performance, and graduate degrees. Whatever the approach used, the basic objective is to filter out more competent from less competent applicants. Introducing a merit-based (internal) promotion system is also essential. In order to motivate recruits, the civil service needs to establish objective, meritorious criteria for upward mobility. If promotions are personalized and - worse - politicized, civil servants will focus more on doing personal favors for their superiors and/or influential politicians. Thus, their talents, however exceptional they are, get diverted to activities that diverge from the objectives set for the civil service. Moreover, any progress towards building prestige via tough recruitment standards gets diluted. In Korea, for instance, promotion is based on a formula that combines seniority with merit-based elements. In India, for instance, promotion is based almost entirely on seniority. Internal promotions based on merit also tend to attract more individuals into the bureaucracy who have strong preferences for "having an impact" on the government's task of providing collective goods, i.e. general welfare. Merit-based recruitment and promotion can therefore serve as mutually reinforcing mechanisms to build commitment towards the goals of the public service. Many other countries, however, have had difficulty instituting meritocratic recruitment and promotion. In Latin America, Peru and Mexico, for instance, have been unable to institute a meritocratic recruitment process except in very small enclaves. Unusually extensive powers of political appointments complement lack of meritocratic recruitment. Consequently, the State becomes a massive source of jobs (cabide de emprego) recruited on the basis of connection rather than merit. In East Asia, political appointments are deeper and more extensive in the Philippine bureaucracy than in the neighboring, high-performing East Asian economies, suggesting that recruitment and promotion are less likely to be merit-based (see Figure 7). In Kenya, political appointments based upon tribal allegiances under different regimes has contributed to expanding employment and managerial tiers in Kenya Post and Telecommunications Corporation. 2 example, see Kim, et al. (1995) on Japan, Organisation for Economic Co-operation and Development (1997), Quah on Singapore (1996), Wade on Taiwan (1990), and Valencia on Chile (1996). 43 Figure 7: Meritocracy in Korea's Civil Service Depth of political appointments Philippines Korea AssistaIr=t mir~ster Index of bureaucratic capability Ass.58alt Assistant minister Directorgeneral 1.0 -- ~~~~~~~~~~~~~~~~~~~~~~~~0.7 Service directOffceorreto LDnvision chief Ofie direct_or 0.5 Ii] Political appointees _ I BE :i:E :E:RI ~~~ ~ ~~~~~~~0.3_ Ratio of public to private wages Percent 0 Korea 80 63 0.0 Philippines Korea 60 40 25 20 . Philippines Korea Source: Adapted from Campos and Root (1996); Commander, Davoodi, and Lee (1996) Even where merit-based recruitment has been put in place and political appointments are not ostensible, political interference can be rampant through transfers, and the absence of merit-based promotion can erode incentives for performance. The Indian civil service provides a clear example: senior civil servants are transferred frequently, and seniority is the only basis for promotion (see Box 16). The result is that the legendary Indian Civil Service has been reduced to a shadow of its former self in terms of its perceived efficiency, effectiveness and integrity. Merit-based recruitment and promotion systems are of two broad but not mutually exclusive types: mandarin systems and open recruitment (Nunberg 1995). Mandarin systems, such as those found in France, Germany, and Japan, are closed-entry, hierarchical systems with highly competitive entrance requirements (see Box 17). Open recruitment systems such as in New Zealand, the United Kingdom, and the United States, provide a more flexible, decentralized, and increasingly market-driven system of civil service recruitment. While offering more flexibility in allowing managers to match job requirements with skills, including hard-to-find technical expertise, such systems can make it difficult to maintain professional standards and esprit de corps. 44 Box 16: India - An Elite Civil Service Gasping In The Political Winds The Indian Administrative Service (IAS), successor to the Indian Civil Service (ICS), the legendary "steel frame" of the British Raj, is battling against onslaughts to its relevance. The strengths of the Service lie in its extraordinary pool of skills and talents, its field experience, its extensive networking, its appreciation and overview of the functioning of govermnent at the cutting edge, its understanding of delivery systems for development, awareness of the formal and informal socio- economic networks in the field, its "can deliver" attitude, its role in national integration, its ready adaptability to new and unfamiliar situations and tasks, and its social orientation, bolstered by intense competition among the officers. Among its weaknesses are the rancor it has generated with other services at Central and State levels; the flourishing black sheep among its numbers; the increasing rarity of impartial policy advice; political and caste alignments; incompetence going unpunished, and initiative and integrity unrewarded; frequent transfers and the "transfer industry" in some States (where the average tenure of field officers may be as low as 8 months); and the demoralizing impact of judicial activism, especially stay orders which stynie development projects and contempt cases against individual officers. While the tiny ICS was an elite service entrusted with the administration of the British Empire in India, the IAS has a much wider and more challenging role as an agent of change. The IAS today is still elite, numbering about 3,500, of whom one-third are promoted from State civil services and the rest recruited centrally through a rigorous annual competitive examination process designed to select about seventy recruits from about 200,000 applicants. The recruitment process is perhaps one of the most objective in the world, and the quality of recruits is excellent, coming from across the social and educational spectrum. Critics have however questioned the efficacy of the system on the grounds that the LAS is staffed by generalists and therefore less skilled compared with the technical specialists who are recruited into East Asian bureaucracies. Put through a rigorous 2-year training process encompassing field training with all departments of the government, attachments with the Central and State legislatures, the judiciary, the armed forces, and industries, officers are allocated to different States where they will spend a large part of their career. They have to learn the regional language and the laws of the allocated State. After typically serving as field officers in successively higher capacities for about 7 to 10 years, culminating with the post of District Officer (also known as District Magistrate/Collector/Deputy Commissioner), officers generally serve alternately with the Center and the State. Officers' intimate knowledge of field conditions in the State enables them to render expert policy advice to the State and Central governments in successively senior positions. It is doubtful if any other service in the world exposes such young professionals (typically between 26 and 40 years old) to such a variety of challenges and with such decision-making opportunities. For instance, as District Magistrates (a district is part of a State and may typically have a population of three to four million and sometimes as much as seven to eight million), they are, in the last resort, responsible for everything in their jurisdiction, from conducting elections to urban and rural development, public health, school education, industrial promotion, calamity administration, law and order, protocol, prosecutorial and police supervision, land reforms, settlement of industrial disputes, supervision and control of the executive magistracy, jails and welfare, Treasury management, and issuing of arms licenses. It is not surprising, therefore, that the Harvard Business School referred to the Indian District Magistrate's job as one of the most challenging responsibilities in the either the public or the private sectors. In the States, almost all Secretaries are from the IAS. At the Center, about 70 percent of Secretaries to Government are from the IAS. Other Services are naturally bitter about this state of affairs and complain, "Nothing grows in the shade of the Banyan tree." Over-recruitment to the IAS and other services in the '70s and '80s has resulted in massive cadre congestion, unnecessary creation of posts to accommodate officers, and frustration among officers, compounded by a sharp fall in living standards relative to the private sector and in real terms. There has been a sharp reduction in intake, and almost 50 percent of seats are reserved for Scheduled Castes, Scheduled Tribes, and Other Backward Classes (SC/ST/OBC). The erosion of pay and standards, combined with the far-reaching economic reforms 45 ushered in 1991, are gradually making the private sector and not the IAS the coveted destination of young graduates. Notwithstanding such vicissitudes, the IAS today must play a crucial role in a leading edge of change in the country, ushering in reforms in different sectors. The question is, does it have the vision to reform itself into a more efficient, more transparent and leaner agent of change? Source: Adapted from Schiavo-Campo et al. (1996) Box 17: Recruiting and Grooming the Best and Brightest - Mandarin and Open Systems Countries have adopted two broad approaches to meritocratic recruitment: "mandarin," or closed- entry, hierarchical systems and open recruitment systems, which have lateral entry and more flexible entry mechanisms. Mandarin systems traditionally referred to upper echelons in certain East Asian bureaucracies, but in their modem usage, they encompass broader corps-career systems that includes middle and lower levels as well. Highly selective recruitment takes place on a centralized basis, generally ensured by rigorous entrance examinations and high test scores. Upon entering the service, elite cadres, who are mainly generalists, are placed on a fast track into the best jobs in government. But they are mainly hired into a career stream or corps rather than specific jobs. France and Japan best exemplify the mandarin recruitment system. At France's National School of Administration, elite civil servants are put through a one-year professional internship followed by fifteen months of coursework for high-profile careers in government. Japan's Tokyo University produces Japan's administrative elite, most of whom have law- generalist backgrounds that may be supplemented through in-service technical training. Variants on this elite grooming system are found in Singapore, where two-year "cadetships" rotate promising recruits, or in Germany, where a "practicum" offers a practical, on-the-job internship for potential elite cadre candidates. Open recruitment systems emphasize more flexible, decentralized, and increasingly market-driven civil service recruitment. The U.S system stands in stark contrast to elite models, permitting horizontal entry without age restrictions and providing considerable interclass (though not necessarily interagency) mobility. It has replaced its centralized competitive entrance exams with profession-specific exams and granted more autonomy to individual managers in hiring. The United States, like Australia, counters its horizontal recruitment mechanisms with a Senior Executive Service aimed at building an elite group from within the civil service. The boldest approach to open recruitment and career development is found in countries pursuing the new public management reforms that have devolved recruitment responsibilities even further. In New Zealand, for example, agency managers can hire staff at market salaries. Mandarin, elitist models have drawbacks when applied in countries lacking high-quality general education systems. But countries with critical shortages of well-qualified human resources may find mandarin systems useful as the basis for a selective rather than a comprehensive approach to personnel development. Moreover, a prestigious elite can have spillover effects in motivating other parts of the civil service. More open recruitment systems give managers greater flexibility in matching job recruitments with skills, including hard-to-find expertise. And open systems discourage civil service insularity by bringing in staff with fresh perspectives and ideas. The downside is that professional standards are less easy to maintain across the service, as in the core set of common public service values or esprit de corps. Source: Nunberg (1995) 46 ADEQUATE COMPENSATION. Building prestige through merit-based recruitment and promotion makes employment in the civil service attractive. But if compensation in the public sector trails far behind that in the private sector, prestige alone will not make up for the difference. A rough benchmark for evaluating the adequacy of public-sector compensation is the relative differential between public and private-sector compensation, with some discount for the greater security of civil service employment. Comparing public and private salaries is made difficult by the need to take into account differences in non-salary allowances, job requirements, and the like. In general, though, civil servants nearly everywhere are paid less than their private-sector counterparts. Singapore stands at one extreme, with public-sector salaries that average 114 percent of private-sector salaries and are higher than equivalent salaries of senior officers in the United States. In the Philippines, public pay averages 25 percent of private wages; in Somalia, 11 percent. In Kenya, the disparity between public wages grew by 3 percent a year during 1982-92. Contributing to the erosion of public-sector wages were fiscal austerity measures during the 1980s that tended to lower real wages rather than employment. A study of twenty- seven countries over 1972-92 found that government wage bills have been reduced primarily by lowering real wages rather than by cutting employment (Lindauer and Nunberg 1994). In addition to the overall erosion of public-sector wages relative to the private sector, a central problem in many developing countries is the low level of remuneration for higher cadres, which makes it hard to attract and retain high-quality staff. Wage compression - whereby the wages of higher-level civil servants are allowed to erode more than the wages of the lower echelons, often for political reasons - is thus a key problem in many countries. A study of ten African countries found that the ratio of the salaries of the highest to the lowest-grade civil servants declined from 13:1 to 9:1 (Lindauer and Nunberg 1994). Across-the-board increases do not solve this problem. In many developing countries in Africa and elsewhere, average wage erosion has been coupled with expanding employment, especially at lower skill levels (see Figure 8). Governments have become employers of last resort and dispensers of political patronage, offering jobs to family, friends, and supporters. As a result, individual wages have often been low even though the overall wage bill has been high. Moreover, growth in the wage bill has often outpaced expenditures on operations and maintenance, leading to the familiar story of teachers without textbooks and extension workers without bicycles. As public employment has risen, wages have fallen. Prompted by a desire to improve public-sector salaries and the need to correct aggregate fiscal imbalances, countries have embarked on civil service reform strategies to reduce employment, decompress the wage structure, and raise average pay. These efforts have met with mixed success. A study of civil service reform efforts between 1981 and 1991 found that wage 47 Figure 8: Pay and Employment in Twenty African Countries, 1972-93 13 si I I 71 2.5 3 3.5 4 4.5 5 5.5 6 6-5 Source. Data from Kraay and van Rijckeghem (1995) bill reduction and salary decompression were achieved in less than half the cases (World Bank 1991). Employment was reduced in more than half the cases, but reversals were sometimes reported later, and the size of the cuts was rarely enough to finance substantial salary increases for higher-level staff. In Peru, for instance, some 250,000 workers were cut from the civil service over three years, but 163,000 of them were then rehired; in addition, poor targeting of the cuts resulted in the departure of the most qualified staff. This mixed and often disappointing experience with civil-service reform nevertheless provides some lessons for future efforts. First, strategies for civil service reforms have focused exclusively on pay and employment issues, and within that, on reducing numbers (wage bill and employment). These are important, but so are other complementary elements - reorienting strategic policy priorities, merit-based recruitment and promotion, performance measurement and orientation, mechanisms to improve accountability, and building esprit de corps. Pay and employment reforms can, however, be embedded in a larger set of public-sector reforms that recognize that civil service reform is an ongoing process (see Box 18). In Uganda, civil service reform has dramatically reduced the numbers of central government employees (from 320,000 in 1990 to 134,000 in 1996), increased average pay, and decompressed pay scales. To improve performance, however, the focus has shifted to complementary mechanisms that strengthen the performance orientation of civil servants and increase their accountability. These include greater clarity of purpose and task, measurement of outcomes, and client surveys for key services. 48 Box 18: Public-sector Reform in China China's public sector is composed of four broad entities: (i) a core government administration of some 10 million civil servants, responsible for macro-managing the economy and financed by the Budget. This includes the government administration, permanent members of various political organs, the judicial forces (police, prosecutors, judges), and permanent members of mass organizations (e.g. the All China Federation of Trade Unions); (ii) a large non-profit sector of about 24 million, including education and health establishments partly financed by the budget, which are intended over time to become self- sufficient; (iii) a large state-owned enterprise sector, which are comprised of some 100,000 units and employing some 75 million; and (iv) the military and defense forces. China has been undertaking a reform of its public administration since the early 1950s . From the start, this was understood as a need to improve the organization and management of its public administrative structures and an upgrading of the quality and competence of its civil service. In the early 1980s, efforts mainly consisted of (i) streamlining the bureaucracy; (ii) retiring older cadres; and (iii) reforming the pay system. A first step in this direction consisted in segmenting the civil service by giving state-enterprise managers more powers over daily enterprise management, particularly pay and incentives. Subsequent efforts focused on rejuvenating the civil service and increasing overall qualifications. In 1991, the government initiated a new phase of administrative reform. Known as the Three Fixes, this program consisted of (i) fixing the role of each government agency; (ii) reorganizing government agencies through the shedding of micro-management units and functional re-structuring; and (iii) resizing the staff of these agencies and units. At present, a first rationalization of the core government group has been completed, with the retirement of some 2 million surplus civil servants. By carrying out a major reform of the civil service, the government expects to improve career management and provide incentives linked to performance and market conditions. An interim regulatory framework, promulgated in 1993, established the legal basis of the new system. It includes the following elements: competitive entry - clear rules of behavior and a system to redress abuses - a personnel evaluation system more specifically based on performance - a pay system linked to wage determination in the enterprise sector. The coming decade's agenda for public administration reform is certain to be driven by the need to effectively macro-manage a large and diversified economy as well as to improve the delivery of cost- effective, quality services in certain critical sectors where government presence will continue to be essential. This will require a sustained effort to strengthen government agencies and their staff and to rationalize the role of the State in public service delivery. Key issues to be examined by policy makers will include (i) a strengthening of the skills for the core civil service; (ii) redefining the role of the State in the delivery of public goods; (iii) improving incentives; and (iv) retrenching surplus staff. The government is preparing a new wage reform in accordance with provisions of the Interim Regulations of the Civil Service System (1993) The new system will link wages in the civil service (and by extension in the non-profit sector) with wages of administrators in manufacturing state enterprises. The linkage will be effected through annual wage comparison surveys carried out by the Ministry of Personnel. A pilot wage survey was completed in 1996 with a grant from the World Banks Institutional Development Fund . An analysis of these findings will provide valuable indications of the role of in-kind benefits, housing and pensions in the total compensation package for Chinese civil servants. It will also assist in suggesting ways for reforming the personnel management system to create incentives for attracting and retaining qualified personnel in the more critical sectors of China's economy. Source: Salem (1996) 49 A more careful sequencing of reforms is called for, starting with wage decompression. Wages at the top of the scale can be raised to attract higher-quality people and concentrate scarce skills in strategic areas, even if wages at lower grades are reduced, making employment less attractive for lower cadres, where the bulk of overstaffing has taken place. In countries that are considerably overstaffed, reforms have been too modest to really downsize governments to sustainable levels, and they have tended to be of the one-shot variety rather than a steady program to redimension government over the longer term. Inevitably, pay and employment reforms will face political obstacles, although fears of political backlash have often been exaggerated. Some countries have viewed civil servants as partners in reform and have consulted extensively with them to find politically acceptable solutions. For instance, in the province of Santa Fe, Argentina, a close dialogue between the governor and the local civil service union helped in agreeing on measures for modernizing provincial public administration, including expenditure cuts of some 10 percent. In addition, the experience with civil service reform has helped develop a good set of technical tools -civil service censuses, functional reviews, the design of severance packages - for managing and implementing the process more effectively. But civil service reforms will generate losers who can be important constituencies of the political leadership and therefore a force to be reckoned with. Building An Esprit de Corps Effective and capable bureaucracies have been characterized by an esprit de corps, a shared commitment toward an organization's objectives. An esprit de corps among members of a particular group refers to a common understanding among its members about what is desirable and undesirable behavior, manifested in norms both formal and informal and grounded on some set of objectives, and a devotion to upholding the honor of the group based on this common understanding. It creates a high degree of coherence among the group's members, gives them a strong sense of purpose and belonging, and effectively creates a self-discipline among members that guides them towards achieving the group's objectives. King Arthur's Knights of the Round Table, the samurai in Japan, and even the Mafioso of past generations all embodied some form of esprit de corps. A few of today's civil services are said to do so as well, e.g. Japan, South Korea, the U.K., France, Germany, and Chile. Most of today's bureaucracies do not have this feature. Of those who do, many years, even centuries, of tradition have helped create, nurture, and institutionalize it. The esprit in Japan's civil service for instance goes all the way back to the traditions of the samurai (the emperor's select warriors). However, it is not at all impossible to develop an esprit among civil servants. For example, today, the Singaporean civil service is well known for its "coherence and sense of purpose." It was not too long ago (early '60s) when the country did not have much of a civil service to speak off. Within two decades, the country's civil service had evolved from what was at best a mediocre bureaucracy to one of the world's most reputable. The process of getting there has been difficult, but 50 nevertheless, the steps have been straightforward. Prospective recruits to the civil service are taken from the top 200 (less than 5 percent) of the graduating class at the National University of Singapore (and more recently, the Nanyang Technological University) and the selected candidates are put through a common one-year training program (Campos and Root 1996). This recruitment procedure creates the foundation for nurturing an esprit de corps. The educational background and the training programn bring the new recruits towards a common understanding of what is expected of them as a civil servant and create the basis for building trust among them. The internal promotion system, which tends to restrict the pool of potential candidates for promotion, together with the single- mindedness of leadership and its continuous efforts at imbibing the Service with its desired values, have helped strengthen the bond between civil servants. Together, all these factors - the recruitment process, meritorious internal promotion, focused efforts of leadership - have helped create an esprit de corps in the Singaporean civil service and turn it into the civilian equivalent of a well-oiled, highly focused fighting machine.3 Some of these lessons from Singapore are being transferred to Botswana using twinning arrangements. These initiatives are emphasizing team work and group performance - as opposed to the pursuit of individual or personal agendas - as key ingredients of successful performance. Indeed, worker dedication and commitment need not be confined to the usual success stories from developed or high-growth East Asian economies. In Brazil's poor northeast region of Ceara, worker commitment to the job helped make dramatic improvements in the quality of public services delivered (see Box 19). This resulted from a sense of "calling" and prestige for the tasks created by the state government, reinforced by innovative practices such as worker participation and self-managed worker teams, multiskilling of workers and multitask jobs, and flexibly organized or "specialized" production (Tendler 1996). Almost all these involved greater worker discretion and autonomy, greater cooperation between labor and management, and greater trust between workers and their customers. These are similar to the Japanese "lean" production methods, which introduced grounded models of how to change organizational practices and management. 3In Japan, the promotion system essentially induces civil servants to remain with the agency they start with and discourages ently at higher levels. In contrast, in Singapore, civil servants are encouraged to move around within the public sector, and to a limited extent, exceptionally good private sector managers are recruited into the service at higher levels. The two approaches create different types of bonding among civil servants, with the Japanese being more agency specific. The two have implications for other institutional arrangements. In the Japanese case, for example, there is a greater need for formal interministerial coordination, which it has. 51 Box 19: Building Worker Dedication to the Job -Good Government in Ceara, Brazil In 1987, the state government of Ceara, Brazil, confronted a crippling fiscal crisis and a legacy of mediocre performance. Yet, in four years, the fiscal crisis had been overcome, and the quality of services improved dramatically. For instance, vaccination coverage for measles and polio tripled from a low of 25 percent to 90 percent of the child population. The public works construction program employed over 1 rnillion unemployed rural farmers during droughts. The business extension and public procurement program for small firms saved over 30 percent of its previous purchases. A detailed research into the sources of this success found that increased worker dedication to the job was a key factor. The state government contributed in an unusual and sometimes inadvertent way to the new sense of recognition. It created a strong sense of "calling" and mission around key programs and their workers. It did this through public information campaigns, prizes for good performance, public screening methods for new recruits, orientation prograrns, and sheer boasting through the media about its successes. At the same time, workers carried out a larger variety of tasks than usual, and often voluntarily. They did this in response to what their clients needed. Workers were able to provide this more customized service because they had greater autonomy and discretion. The greater discretion and responsibilities inherent in the self-enlarged jobs did not result in greater opportunities for rent-seeking. This was because of two other mechanisms - hemmning in with new pressures to be accountable - worked in the opposite direction. On the one hand, workers wanted to perforn better in order to live up to the new trust placed in them by their clients. This, in turn, was the result of the more customized arrangements of their work and the public messages of respect from the state. On the other hand, the communities where these public servants worked watched over them more closely. The state's publicity campaigns and similar messages had armed citizens with new information about their rights to better government and about how public services were to work. Government played a powerful role in monitoring, then, but it did so indirectly. All of these created a cycle wherein workers in Ceara reported feeling more appreciated and recognized, not necessarily by superiors but by their clients and the communities where they vworked, reinforcing their dedication to the job. Source: Tendler (1996) An esprit de corps helps reduce transactions costs among civil servants because it produces clear norms that bind civil servants to acceptable forms of behavior. These norms thus become a basis for judging whether someone is straying from the objectives to which the group is committed and lays the foundation for building and nurturing trust between and among members. In many ways, an esprit de corps is a manifestation of a considerable level of social capital. Its implications for policy implementation are substantial: less paper work is required, fewer meetings, and in general, fewer bureaucratic procedures.4 We note, however, that this need not necessarily be good. A lot depends on the objectives of the group. The Nazi SS, for instance, is known to have had very tight relationships among its members, but it was also used to carry out one of the world's most horrifying objectives. 52 VII. Conclusion In both past and present, countries have struggled with the problem of creating and managing a public bureaucracy. From kings in medieval Europe to emperors in China, rulers have in one way or another attempted to assemble a coherent body of public servants to help carry out their mandates. The increased complexities of the modem era have only made this task more compelling. Yet up till recently much of the emphasis of discussions within development circles has been on the kinds of policies that make for rapid socio-economic development. Today these discussions have indeed come to include debates on the role of institutions, i.e. the rules of the game, but these discussions have not placed sufficient emphasis on the inner workings of the public sector, the structure of incentives, how they affect policy outcomes, and, most especially, how they might be replicated or adapted in different settings. If history is any guide, problems of developing an effective public sector lie at the core of socio-economic development. While the types of policies a country chooses to adopt are critical, the kind of machinery that ultimately helps to formulate and implement these policies is equally important. For unless the machine works to produce the desired outputs, correspondingly appropriate policies will fail to materialize. This paper has attempted to develop a coherent framework within which to elaborate on the myriad of problems countries face in developing an effective public sector, to present best practice solutions to these problems, and to a lesser extent to evaluate the potential adaptability of best practices to other settings. Much more work needs to be done in this area and it is our hope that the paper stimulates scholars from various fields and persuasions to pursue the many unanswered questions that it has identified. 53 Bibliography Aoki, Masahiko, Hyung-ki Kim, and Masahiro Okuno-Fujiwara (eds.). 1997. The Role of Government in East Asian Economic Development, Oxford, Clarendon Press. Andic, Fuat, and Suphan Andic. 1996. The Last of the Ottoman Grandees: The Life and Political Testament of Ali Pasha. Istanbul: Istanbul's Press. Bates, Robert. 1996. Background Note, 1997 World Development Report. Brautigam, Deborah. 1996. "State Capacity and Effective Governance." In Benno Ndulu and Nicholas van de Walle, eds., Agenda for Africa's Economic Renewal. Washington, D.C.: Overseas Development Council. Campos, J. Edgardo, Margaret Levi, and Richard Sherman. 1994. "Rationalized Bureaucracy and rational Compliance," IRIS Working Paper. Campos, J. Edgardo, and Sanjay Pradhan. Forthcoming. "Building Institutions for a More Effective Public Sector." Background paper for 1997 World Development Report. Campos, J. Edgardo, and Sanjay Pradhan. 1996. "Budgetary Institutions and Expenditure Outcomes: Binding Governments to Fiscal Performance." Policy Research Working Paper No. 1646. Policy Research Department, World Bank, Washington, D.C. Campos, J. Edgardo, and Hilton L. Root. 1996. The Key to the Asian Miracle: Making Shared Growth Credible. Washington, D.C.: Brookings Institution. Cox, Gary and Matthew McCubbins. 1993. Legislative Leviathan: Party Government in the House, Berkeley: University of California Press. Chung, Lee. 1992. "The Government, Financial System, and Large Enterprises in the Economic Development of South Korea," World Development, 20: 187-197. Commander, Simon, Hamid Davoodi, and Une J. Lee. 1996. "The Causes and Consequences of Government for Growth and Well-Being." Background paper for 1997 World Development Report. Clague, Christopher, Philip Keefer, Stephen Knack, and Mancur Olson. 1996. "Property and Contract Rights under Democracy and Dictatorship." Journal of Economic Growth 1(2): 243-76. Dia, Mamadou. 1996. Africa's Management in the 1990s and Beyond: Reconciling Indigenous and Transplanted Institutions. Directions in Development Series. Washington, D.C.: World Bank. 54 Dixon, Geoff. 1993. "Managing Budget Outlays 1983-84 to 1992-93." In Brian Galligan, ed., Federalism and the Economy: International, National and State Issues. Canberra: Federalism Research Centre, Australian National University. Evans, Peter B. 1995. Embedded Autonomy: States and Industrial Transformation. Princeton, N.J.: Princeton University Press. Evans, Peter B., and James Rauch. 1996. "Bureaucratic Structure and Economic Growth: Some Preliminary Analysis of Data on 35 Developing Countries." University of California, Berkeley, Calif Fiorina, Morris. 1987. "Alternative Rationales for restrictive Procedures," Journal of Law, Economics, and Organization, 3:337-345. Geddes, Barbara. 1994. Politician's Dilemma: Building State Capacity in Latin America. University of California Press: Berkeley, California. Governance. Special Issue on Managerial Reform 3 (2): 1990. Heyneman, Stephen P. Forthcoming. "Education Choice in Eastern Europe and the former Soviet Union: A Review Essay." Education Economics. Hirschman, Albert 0. 1970. Exit, Voice, and Loyalty Responses to Declines in Firms, Organization and State. Cambridge, Mass.: Harvard University Press. Israel, Arturo. 1996. "A Guide for the Perplexed: Institutional Aspects of Social Programs." Transportation, Water, and Urban Development Department, World Bank, Washington, D.C. Keefer, Philip and Steve Knack. 1995. "Institutions and Economic Performance:Cross Country Tests Using Alternative Institutional Measures," Economics and Politics, Vol. 7: 207-227. Kim, Hyung-ki, Michio Muramatsu, T.J. Pempel, and Kozo Yamamura. 1995. The Japanese Civil Service and Economic Development, Oxford, Clarendon Press. Klitgaard, Robert. 1996. "Cleaning Up and Invigorating the Civil Service." Background paper prepared for the Civil Service Reform Study. Operations Evaluation Department, World Bank, Washington, D.C. Kraay, Aart, and Caroline van Rijckeghem. 1995. "Employment and Wages in the Public Sector: A Cross-Country Study." IMF Working Paper WP/95/70. Fiscal Affairs Department, International Monetary Fund, Washington, D.C. Langan, Patricia. 1996. Background Note 1997 World Development Report. Laver, Michael and Kenneth Shesple. 1994. "Cabinet Governmkent in Theoretical Perspective," in M. Laver and K. Shepsle (eds.), Cabinet Ministers and Parliamentary Government, Cambridge University Press. 55 Leipziger, Danny, and Kim Kihwan. 1992. "Korea: A Case of Effective Government-Led Growth." Paper presented at the World Bank workshop on the Role of Government and East Asian Success, East-West Center, Honolulu. Lindauer, David, and Barbara Nunberg, eds. 1994. Rehabilitating Government: Pay and Employment Reform in Africa. Washington, D.C.: World Bank. Manning, Nick. 1996. Background Note 1997 World Development Report. Nunberg, Barbara. Forthcoming. The State After Communism: Administrative Transitions in Central and Eastern Europe. publisher Nunberg, Barbara. 1995. Managing the Civil Service: Reform Lessons from Advanced Industrialized Countries. World Bank Discussion Paper No. 204. Washington, D.C.: World Bank. Organisation for Economic Co-operation and Development. 1997. Issues and Developments in Public Management, OECD, Paris. Organisation for Economic Co-operation and Development. 1995. Governance in Transition: Public Management Reforms in OECD Countries. OECD, Paris. Pempel, T.J., and M. Muramatsu. "Structuring a Proactive Civil Service." The Japanese Civil Service and Economic Development, Hyung-Ki Kim, M. Muramatsu, T.J. Pempel, and K. Yamamura, eds. Oxford University Press: 1995. Quah, Jon. 1996. "Public Administration in Singapore: Managing Success in a Multi- Racial City State," in A. S. Huque, J. T. M. Lam, and Jane C. Y. Lee (eds.), Public Administration in the NICs: Challenges and Accomplishments, London: Macmillan Press. Reid, Gary. 1996. Background Note 1997 World Development Report. Rose-Ackerman, Susan. forthcoming. "When Is Corruption Harmful?" Background paper for 1997 World Development Report. Rose-Ackerman, Susan. Corruption: A Study in Political Economy. New York, N.Y.: Academic Press. Salem, Claude. 1996. Background note to the World Development Report 1997 - The State in a Changing World. World Bank, Washington, D.C. Schiavo-Campo, Salvatore, Giulio de Tommaso, and Amitabha Mukherjee. "An Intemational Statistical Survey of Government Employment and Wages." Background paper for 1997 World Development Report. Shepsle, Kenneth, and Barry Weingast. 1994. "Positive Theories of Congressional Institutions," Legislative Studies Quarterly, 19: 149-179. 56 Stevens, Mike. 1996. Background Note 1997 World Development Report. Tendler, Judith. 1997. Good Government in the Tropics. Baltimore, Md.: Johns Hopkins University Press. Valencia, Sofia. 1996. "Public Sector Reforms in a Developing Country: An Analysis of the Chilean Experience." mimeo. van der Gaag, Jacques. 1995. Private and Public Initiatives: Working Together for Health and Education. Washington, D.C.: World Bank. van Rijckeghem, Caroline, and Beatrice Weder. 1996. "Corruption and Rate of Temptation: Do Low Wages in the Civil Service Cause Corruption?" Background paper for 1997 World Development Report. Wade, Robert. 1990. Governing the Market. Princeton, Princeton University Press. Wade, Robert. 1994. "The Governance of Infrastructure: Organizational Issues in the Operation and Maintenance of Irrigation Canals." World Bank, Washington, D.C . Webb, Douglas. Forthcorning. "The Judiciary: The Arbiter of Rules and Resolver of Disputes." Background paper for 1997 World Development Report. World Bank. 1991. "The Reform of Public Sector Management: Lessons of Experience." Policy Research Series No. 18. World Bank, Washington, D.C. World Bank. 1993. The East Asian Miracle: Economic Growth and Public Policy. A World Bank Policy Research Report. New York, N.Y.: Oxford University Press. World Bank. 1994. World Development Report 1994: Infrastructure for Development. New York, N.Y.: Oxford University Press. World Bank. 1995. Bureaucrats in Business: The Economics and Politics of Government Ownership. A World Bank Policy Research Report. New York, N.Y.: Oxford University Press. World Bank. 1996a. "Bangladesh: Government That Works: Reforming the Public Sector." Private Sector Development and Finance Division, Country Department I, South Asia Region, World Bank, Washington, D.C. . World Bank. 1996b. "Republic of Guinea: Public Expenditure Review." Western Africa Department, World Bank, Washington, D.C. World Bank. 1997a. "Ukraine Public Sector Reform Loan, Preparation Mission Report, Public Administration." World Bank, Washington, D.C. World Bank. 1997b. "World Development Report 1997: The State in a Changing World." New York, N.Y.: Oxford University Press. 57 The mission of WBI is to help World Bank clients and staff acquire new development knowledge and skills through a variety ot courses, seminars, and other learning events. It designs programs on topics related to economic and social development for governments, nongovernmental organiza- tions, and other stakeholders. The Institute produces and dis- semiiinates publications and electronic informiatiorn products that support these objectives. For information on WBI pLublications write to. Publications WBI The World Bank 1818 H Street, N.W. Washington. D.C. 20433 Tel: (202) 473-6349 Fax: (202) 522-1492 Visit us on the World Wide Web at: http://www.worldbank.org/wbi