Awma Economic Development Institute 000 of The World Bank The Civil Service System and Economic Development The Japanese Experience Hyung-Ki Kim EDI LEARNING RESOURCES SERIES EDI LEARNING RESOURCES SERIES The Civil Service System and Economic Development The Japanese Experience Report on an international colloquium held in Tokyo, March 22-25, 1994 Hyung-Ki Kim The World Bank Washington, D. C. Copyright C 1996 The International Bank for Reconstruction and Development / THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing July 1996 The Economic Development Institute (EDI) was established by the World Bankin 1955 to train officials concerned with development planning, policymaking, investment analysis, and project implementation in member developing countries. At present the substance of the EDI's work emphasizes macroeconomic and sectoral economic policy analysis. Through a variety of courses, seminars, and workshops, most of which are given overseas in cooperation with local institutions, the EDI seeks to sharpen analytical skills used in policy analysis and to broaden understanding of the experience of individual countries with economic development. In addition to furthering the EDI's pedagogical objectives, Policy Seminars provide forums for policymakers, academics, and Bank staff to exchange views on current development issues, proposals, and practices. Although the EDI's publications are designed to support its training activities, many are of interest to a much broader audience. EDI materials, including any findings, interpretations, and conclusions, are entirely those of the authors and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. Because of the informality of this series and to make the publication available with the least possible delay, the typescript has not been prepared and edited as fully as would be the case with a more formal document, and the World Bank accepts no responsibility for errors. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center, 27 Congress Street, Salem, Massachusetts 01970, U.S.A. The complete backlistof publications from the World Bank is shown in the annual Index ofPublications, which contains an alphabetical title list (with full ordering information) and indexes of subjects, authors, and countries and regions. The latest edition is available free of charge from the Distribution Unit, Office of the Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, Banque mondiale, 66, avenue d'I6na, 75116 Paris, France. At the time of writing, Hyung-Ki Kim was chief of the Studies and Training Design Division in the World Bank's Economic Development Institute. Library of Congress Cataloging-in-Publication Data Kim, Hyung-Ki, 1936- The civil service system and economic development : the Japanese experience : report on an international colloquium held in Tokyo, March 22-25, 1994 / Hyung-Ki Kim. p. cm.-(EDI learning resources series, ISSN 1020-3842) Includes bibliographical references. ISBN 0-8213-3170-1 1. Civil service-Japan-Congresses. 2. Japan-Economic policy-1945- -Congresses. 3. Civil service-Congresses. I. Title. II. Series JQ1646.K56 1996 354.52001-dc2O 94-48164 CIP Contents Foreword v 1. Introduction I 2. Report on the International Colloquium on "The Civil Service and Economic Development: The Japanese Experience" 3 Hyung-Ki Kim 3. Civil Service in Changing Times: The Japanese Bureaucracy and Its Reform 39 Toshiyuki Masujima 4. The Role of the Bureaucracy in Japan's Economic Development 53 Yutaka Kosai iii Foreword The experience of Japan and other East Asian economies in the postwar period suggests that the civil service, working in conjunction with markets, can make a strong contribu- tion to development. In this context, a colloquium on "The Civil Service and Economic Development: The Japanese Experience" was held in Tokyo in March of 1994. The three- pronged objective of the colloquium was to learn more about the Japanese civil service system, to compare it with that of other countries, and to draw some lessons for the devel- oping and transition economies. The colloquium also evaluated examples of successful practices and lessons learned from civil service reform efforts in other countries. This book presents a summary report on the colloquium along with two chapters on the Japanese experience with civil service reform and the role of the civil service in Japan's economic development. Vinod Thomas, Director Economic Development Institute V 1 Introduction The experience of Japan and other countries in East Asia in the period following World War II suggests that the civil service, working in conjunction with the private sector, can make a strong positive contribution to the process of economic development. In this con- text, a colloquium entitled "The Civil Service and Economic Development: The Japanese Experience" was organized jointly by the Economic Development Institute of the World Bank and the National Personnel Authority of Japan, and held in Tokyo on March 22-25, 1994. The three-pronged objective of the colloquium was to learn more about the Japa- nese civil service, to compare it with civil services in other countries, and to draw lessons from Japan's experience. To meet this objective, the colloquium also evaluated examples of successful practices and lessons learned from civil service reform efforts in other coun- tries, sought common denominators, analyzed differences, and proposed ideas for future reform. Briefly, the Japanese experience suggests that a disciplined, internally competitive, and reputable bureaucracy, which is capable of "crowding in" private sector participation, can play a valuable role in the process of late development. The Japanese experience also underscores that the context within which the bureaucracy operates is important. Key amongst the variables in this context are the commitment of the political leadership to development, and the relationship between the civil service and an elected political lead- ership. It is sometimes argued that the role of the state in Japanese economic development in the postwar period is unique to Japan and cannot be readily replicated in other countries, given differences in cultural and historical contexts. However, several countries in Asia have emulated Japan as a model for successful state intervention. It should be remem- bered also that Japan itself borrowed extensively from Europe in the late nineteenth cen- tury and from the United States in the postwar period. Institutional importation by the Japanese took the shape of a process of conscious choice and creative adaptation to the indigenous cultural and social context. The colloquium underscored that a key challenge facing policy makers, as they go about planning civil service reform, is to recognize that institutions such as the civil ser- vice operate as a system, and that cultural factors and historical contexts interact with for- mal rules and organizational structures in determining the efficiency of a civil service system. Recognizing the complexity of the institutional matrix is a key requirement for realistic planning of civil service reform, just as it is an important requirement in economic policy making. This report is structured as follows. Chapter 2 provides an overview of the collo- quium, with a discussion first in general terms of structural elements of civil services and I 2 Introduction the potential role the civil service can play in the process of economic development. This general discussion is followed by a detailed look at the Japanese experience of building and maintaining a competent bureaucracy, and an appraisal of the role the civil service has played in the process of that country's economic development. Chapter 3 and 4 pro- vide two interesting and divergent perspectives on the Japanese bureaucracy by key par- ticipants from within the system.Toshiyuki Masujima, in his capacity as Vice Minister of the Management and Coordination Agency, is currently a custodian of the civil service system in Japan. Yutaka Kosai was formerly a high-ranking official in the Economic Plan- ning Agency, and now works in the private sector. These insider perspectives on the evo- lution of the Japanese civil service system are meant both to reinforce and to complement the formal findings of the colloquium. 2 International Colloquium on "The Civil Service and Economic Development: The Japanese Experience" Hyung-Ki Kim Japan's rapid and sustained development in the postwar period has fascinated policy- makers and academics worldwide. It is generally agreed that the Japanese civil service played a critical role in the country's rise to economic prominence in the twentieth cen- tury. Policymakers thus look to the Japanese experience for answers to such questions as the following: * What is the appropriate mix of state intervention and market activity in economic development? * What can be done to ensure public sector efficiency and productivity? * How can the size of the bureaucracy be limited without damaging the quality and breadth of the services provided? * What are appropriate personnel policies and pay incentives? * How can political control of agencies be ensured without stifling the bureaucracy's creativity, initiative, and dynamism? It is often said that what Japan has done is unique to Japan and cannot be readily cop- ied, yet other countries in Asia have done many of the same things Japan has done. With this in mind, the colloquium's primary objective was to discuss and to learn from Japan's experience in developing and reforming a modem civil service system, and to debate the civil service's role in economic development. Right at the start an important caveat was raised: such things as fully importable models, standard prescriptions, or best practices are not available. Rather, all that can be offered are sound practices that can be adapted from the systems currently using them to the particular conditions and objectives of another country to solve the problems common to civil services everywhere. The colloquium discussion reported here is structured as follows. First, structural char- acteristics of civil services, the role of the civil service in economic development, and the dynamics of civil service reform are discussed at a general level. The second part of the report brings in the specifics of the Japanese case to build on this general discussion. The Role of the Civil Service For our purposes, civil servants are the employees of central and local governments and of public corporations. The fundamental goals of any civil service can be agreed upon by 3 4 The Civil Service and Economic Development: The Japanese Experience those with widely differing philosophies on specifics. The civil service must administer the laws of the country. Further, every country hopes its civil service system can make a positive contribution to its economy. In this regard, the first principle to which a civil ser- vice should adhere is, like the Hippocratic oath taken by doctors, first do no harm. How- ever, the reality is that poorly organized, inept, bloated, and inefficient civil services have greatly contributed to long-term economic stagnation in many countries. Many countries today thus find their civil services to be part of their development problem. However, a World Bank (1993, chapter 4) study of economic growth and public policy in East Asia underscores how a reputable civil service was instrumental in the success of the high-per- forming East Asian economies. Civil services can be beneficial, and models for potential emulation are numerous, but all involve tradeoffs. Civil Service Reform The participation of politicians in initiating and sustaining civil service reform is essential. Reform can be thought of as the upsetting of the status quo or the dissolution of an equilib- rium. Bureaucracies are good at restoring or keeping equilibria, but often are not good at dissolving them. This is the business of politics. Politicians must define the country's objectives and formulate the policies to achieve them. The fundamental job of bureaucrats is to carry out the government's policies and to provide services to the citizenry. In some countries the relationship between legislative organs and the civil service affects the speed and nature of reform. Thus, congressional oversight in the United States significantly affects what the leaders of the executive branch can do. Under a Westminister parliamentary system this is less of an issue. Reform needs to be carried out with two questions in mind. First, what is the develop- ment agenda of a country at a given historical juncture? Second, what is the role of gov- ernment consonant with this agenda? The business of the state in transitional economies is different from that in countries such as South Africa and India. The way to reconcile the conflicts inherent in the attempt to keep the government small while at the same time meeting demands on the system to grow is by carefully selecting the development agenda. Either through political process or bureaucratic discretion, the range of business to be conducted by the government, and therefore the size of the government, must be kept in check. The case of civil service reform in Africa provoked considerable discussion on the part of participants. Inherited from colonial times, public administrations in Africa tend to be extractive and to enforce compliance, rather than to serve civil society. It is clear that well- enforced systems of accountability need to be put in place and incentive structures need to be modified. However, as one participant stressed, emphasis must be placed on institu- tional adaptation rather than institutional transplantation when attempts are made to emulate other civil service models. Technical and Structural Problems in a Bureaucracy Bureaucracies everywhere face similar problems. Among these are problems of cost con- tainment; the level of centralization of services; coordination of different agencies; the The Civil Service and Economic Development: The Japanese Experience 5 recruitment, training, and retention of high quality personnel; and the political respon- siveness, administrative independence, and integrity of the civil service. A modem civil service is essentially a hierarchy with myriad branches. It is a relatively rigid structure, in accordance with Max Weber's classical definition of bureaucracy. It is, however, virtually impossible to divide the tasks of government without overlaps, so one of the inherent defects of a completely vertically structured administrative system is likely to take the form of turf wars; that is, arguments over domains or sectionalism. Further- more, if government agencies have a monopoly over decisions regarding their sphere of activities, the system can be opaque. Competition among agencies might mitigate the ill effects of structural rigidity and opaqueness. When quarrels arise between agencies, the legislature or the public must be able to intervene. Political Responsiveness and Administrative Independence Resolving the question of political ties and of how a civil service can maintain sensitivity to the political goals of a regime's leadership, while simultaneously making technocratric decisions based on its own skills and awareness of the problems, is not easy. The only real- istic way to approach this is to realize that the higher the level, the closer civil servants are to political decisions, and the more critical it is for them to balance technocratic demands with political demands. Top civil servants have to bridge the world of administration and the world of politics, and to monitor the traffic back and forth, ensuring that the people reporting to them understand that they have to meet certain political goals and needs. At the same time, they also have to protect the administrative bureaucracy from day-to-day interventions by politicians who want to, for example, make sure that the new bridge is built in their dis- trict or is constructed by their relatives' companies. For the administrator, the key question centers around keeping politics out of adminis- tration, so that the civil service can function effectively. While tradeoffs between policy objectives and administrative possibilities and capabilities exist, the fundamental task is to make the civil service sensitive to the political needs and objectives of elected officials without undermining its independence and competence. Indeed, reconciling the need for autonomy of the civil service and the role of the executive, the judiciary, and the legisla- tive branches in an ever-shifting political, social, and economic environment is a task that has often eluded even the most developed countries. Nations less far along the path of economic development may find the task even more daunting. The United States has a long tradition of putting political appointees at the top levels of agencies to ensure political loyalty, a practice called "layering in." Many other govern- ments and civil services, such as those in France, Germany, and Japan, are, by contrast, highly resistant to political oversight and to any interference with what they see as tech- nocratic decisionmaking. Yet within democracies, some mechanisms must exist to blend sensitivity to the civil service's need for technical competence free of vote-seeking politi- cians, while at the same time ensuring that the civil service is not totally insensitive to the fact that many decisions are political, and thus beyond its ken. In France, many politicians, including even prime ministers, are graduates of the National School of Administration who have moved into politics from senior civil service positions. France has three levels in the upper reaches of government. The first and high- est level is government ministers. The next level is the ministerial cabinets, a special insti- 6 The Civil Service and Economic Development: The Japanese Experience tution also found in some other European countries. These cabinets typically consist of ten to twenty personal advisors for each minister whom the minister appoints and discharges. The cabinets play an important role as organs of communication between the political and the administrative levels. Belgium and Italy have similar arrangements, and Germany has a class of political civil servants, as opposed to "real" civil servants, who are chosen at the minister's discretion. Many of them follow the minister to each new position, serving as personal advisors. The third level consists of directors of central administration and other discretionary positions, such as prefects, rectors, and heads of public bodies. The directors are mainly charged with three kinds of tasks: (a) to provide the minister with information on all issues within their jurisdiction; (b) to prepare and to implement political decisions; and (c) to supervise the daily functioning of units under them, including handling any difficulties that may arise. As such, the director's role is mainly administrative, but is connected to the upper political level through the activities of the ministerial cabinet, which forms strong links between the two levels. Directors are considered relatively politicized by some, but in the French tradition they are not all changed by an incoming government, or at least not immediately. A look at the Balladur government gives an idea of the degree to which appointments are political in the French bureaucracy. About 600 posts fall within the government's dis- cretion, of which 150 changed title with the inauguration of the new government. Among administrative directors, 64 of 130 were changed, as were 24 of 78 prefects of department with administrative responsibilities, 14 of 28 rectors (heads of the educational administra- tion), and 14 of 22 regional prefects. Fewer than half of these senior posts were changed, and fewer than a tenth of the others were changed. Germany's bureaucracy is much more politicized, a feature which Germans see as instrumental in the creation of a consensus. Around 60 percent of German civil servants are actually members of a political party. In the United Kingdom, by contrast, the civil ser- vice is relatively neutral. Centralization versus Decentralization of Government Services Most industrial countries have civil service management structures rooted in nineteenth century Weberian, centralized, control-oriented, rigidly hierarchical civil services with professionalized cadres. Despite quite different political systems and environments, bureaucracies in these countries have tended to be largely depoliticized and relatively free of corruption. Under rapidly developing technological and financial stress and increased demands most, however, have seen some major systemic changes. Approaches to administration characterized as the "new public sector management" or the "new managerialism" now exist. These seek to improve the efficiency and respon- siveness of government, variously focusing on inputs, as in staff cutbacks, or on outputs, as in performance management. A study (Nunberg 1992) that looked at the relevance for World Bank borrower countries of administrative practice and reform in twelve advanced industrial countries found several approaches to administrative reform, namely: Incremental changes at the margins of still rather tightly controlled, centralized, hierarchical systems, whose values remain more or less administratively determined (France, Germany, Japan, and the Republic of Korea). The Civil Service and Economic Development: The Japanese Experience 7 * More radical change, including a move to decentralize financial and personnel man- agement (Australia, New Zealand, Sweden, and to some extent the United King- dom). In theory, these reforms respond to market principles and performance criteria and use sophisticated information management systems to monitor behavior and to supplant tighter controls. * Canada, the Netherlands, and the United States have used a combination of these two approaches. * Singapore has carried out many innovations that are singular in nature. Thus, while some governments have increased centralization of their agencies, others have opted for decentralization and enhanced flexibility. Each course of action has its advantages and disadvantages, and it is hard to argue that one approach is inherently bet- ter than the other. For example, the decentralization entailed by New Zealand's and Aus- tralia's recent reforms is much lauded. These approaches have drawn on concepts of total quality management (TQM), the idea being to allow lower levels of government, that is, those presumably closer to immediate problems, to monopolize information and to act on the basis of their best judgment, with little need for clearance from higher levels. In contrast, in Japan, which has a centralized system, great effort goes into sharing important information across agencies and from higher to lower levels within the same agency. In part, this is made possible by the centralization of bureaucratic decisionmaking organs in and around the Kasumigaseki district of Tokyo. At the same time, however, this approach can be criticized as extremely time consuming. Its advocates counter this by not- ing that decisions are based on a widely shared information pool and can be effected more quickly. Critics suggest that time given to information sharing restricts the capacity of individual agencies to move quickly and creatively to deal with problems. The French experience underscores the need to exercise caution when contemplating decentralization. If local authorities are given recruitment powers, the number of local civil servants is likely to expand significantly because of patronage, and the likelihood of nepotism and corruption increases. In addition, elected governmental officials may use decentralized bodies as a means to advance their own political agendas. An intermediate stage may be what the French term "deconcentration," a delegation of authority and powers to a representative of the central government called a prefect. Pre- fects have strong links with the local authorities and negotiate and implement central gov- ernment objectives in accordance with local demands. This stage can, of course, eventually evolve through further decentralization. A tightly monitored, centralized system typically enforces nationally determined staff- ing levels by means of a coordinated effort of the central personnel authority and the min- istry of finance. Often routine functions are delegated to department-level and line agency personnel staff, who are then trained to carry out these tasks, but overall staffing levels are kept in check centrally through regulation of recruitment and periodic staff inspection. This type of system is good at controlling the size of the bureaucracy and ensuring stabil- ity, coordination, and continuity of policy (institutional memory), but less good at being flexible and responsive to staffing needs as perceived by line managers. A decentralized system uses running cost controls as key determinants of optimal staffing levels. This imposes across-the-board spending limits, but implies indifference to the level of staff expenditures within recurrent spending categories. Typically decentrali- zation requires a high level of micromanagement of information inputs, with frequent, 8 The Civil Service and Economic Development: The Japanese Experience regular reporting to the finance authority in charge of overall system monitoring. Decen- tralization enhances flexibility, but provides less capacity for overall planning and requires a high level of technical and managerial capacity throughout the system. Facing choices among these imperfect options, what should countries building or ren- ovating civil services do? Colloquium participants suggested several stylized guidelines. Policymakers should not anguish over what kind of systems to introduce wholesale, but rather should concentrate on individual elements within systems and how these might be introduced. Traditional, mainly centralized models of civil service management seem to provide the best starting point for the typical developing country. Given that such countries will tend to have a relatively small pool of talent to draw upon, husbanding them at the center is logical. Agency (or decentralized) approaches, which require technological and human resource skills beyond the present capacity of many developing countries may, however, provide models toward which countries can strive. Certainly, some of the administrations in transition from central planning could successfully adopt some elements of decentral- ized reform. A second general principle is that strong, centralized institutions are necessary to sup- port fledgling civil service management functions in many developing countries. Shared power and good coordination among a limited number of organs at the center ensures checks and balances to the degree that capacity can be developed in line agencies for civil service management. However, some deconcentration is desirable for flexible and respon- sive personnel management. The function of the bureaucracy seems to be conditioned not only by the type of gov- ernment and its mission and objectives, but also by the state of development of local gov- ernment institutions. Many industrial countries assign a lot of functions to local governments, which helps keep the central bureaucracy lean and small. In many other countries, however, often for historical reasons, a local government system has not devel- oped. The central authority performs most functions through district and subdistrict out- posts. Even though having many individuals seeking high-prestige jobs in the national gov- ernment may be inevitable, and even desirable, in the early stages of development, one of the major tasks of any developing civil service is to help create local government institu- tions. In other words, helping local governments create their own civil services with greater responsiveness to regional problems than the central government is an inevitable consequence of economic development. In many respects this contradicts the notion of the Weberian civil service, where power tends to be concentrated in the national capital. It is a selfless task that the national civil service may find difficult to carry out, but it must. The Cost and Size of a Bureaucracy A difficult problem, endemic to any bureaucracy, is internal pressure to expand its size. It may also encounter external pressure to do so, and in many countries government employment has increased not so much because of the functions the government is carry- ing out, but because of patronage pressures. Often, expanding civil services and better compensation have combined to drive up the total cost of government at a rate far faster than that at which the national economy is growing. In this way a bigger, even a better, civil service can become a handicap to economic growth. Hence, a central question that The Civil Service and Economic Development: The Japanese Experience 9 governments have been forced to confront has been how to reduce the total costs of gov- ernment operations without also reducing the quantity and quality of services. In addition to domestic pressures, there are international forces to reduce the scope of government. International economic linkages have promoted deregulation to attract for- eign direct investment and to maintain the international competitiveness of domestic firms, even if this conflicts with the domestically driven goals of the civil service and the government. For instance, New Zealand and Sweden long had widely respected civil ser- vices that provided the public with a wide array of services, but the civil services' large size and elaborate programs had become burdens by the mid-1980s. As a result, govern- ments in both countries have sought to reduce the number of national civil servants and to privatize many government functions. The civil service as a whole, and the individuals in it, must recognize that times will arise when they will have to be willing to turn over some of their roles and responsibili- ties to the private sector. This is difficult. Too often civil services assume that the private sector simply is not ready yet, but people must be trained and institutions created that can take over tasks devolved by the government. Definitional problems sometimes make comparisons of civil services across countries, and even time, difficult. Most broadly civil servants are employees of central and local governments and of public corporations, and members of the military. Looking at the dis- tribution of civil servants between central and local administrations shows just how diverse the possible arrangements are. For example, the relative size of different govern- ment levels depends on how a country classifies its police force, teachers, and health workers. Thus, one speaker at the colloquium presented data showing that France had some 78 civil servants per 1,000 population and the United Kingdom had 84, while another speaker, using a different definition, reported figures of 112 and 73, respectively. Whatever their size, almost all European countries are attempting to reduce the size of their civil services, with varying levels of success. In France, the government had planned to reduce the number of civil servants by 0.5 percent per year, but no real reductions have actually occurred because of the increasing number of local agents tied to the decentrali- zation policies implemented since 1982. France has, however, been able to reduce the number of ministries. The government of Prime Minister Balladour has twenty-nine min- isters, compared to forty in the previous government. Coming up with standard prescriptions as to the size of the civil service is extremely difficult. Overstretched governments in Africa can point to too many employees in the civil service, but in Eastern Europe and the former Soviet Union, core government is sur- prisingly small in terms of numbers. Often it boils down to what governments can afford. Certainly in many African and Latin American countries, the issue is one of fiscal feasibil- ity. Thus, at one point, in Ecuador something like 90 percent of the budget was used to pay civil servants, whereas in Japan the figure is around 5 percent. One trend worth looking at in terms of affordability is keeping the core civil service small, and around this core build- ing a less permanent, more flexible cadre of people who can be brought in on a contractual or less than permanent basis as needed. Many in the least developed countries expect the government to play a principal role in early industrialization, and so a high proportion of the work force will work for the civil service in these countries. The Japanese experience, however, shows that a proactive government can be small. 20 The Civil Service and Economic Development: The Japanese Experience An important point to note is that transition, deregulation, and administrative reform in general may cause unemployment. However, a dynamic and prosperous private sector that can make use of the knowledge and experience of retiring civil servants is one way to keep a bureaucracy small, healthy, and even content with relatively lower salaries. Japan had a larger government during World War II, and the memory of the misery of that period has created a preference for small government. The transition from a con- trolled economy to a market economy was not easy. For example, in 1949, Japan National Railways, now partly privatized, discharged almost 100,000 employees to balance its books. This resulted in nationwide strikes, and the governor of Japan National Railways was either assassinated or committed suicide. Overall growth can absorb those becoming unemployed in declining or overstaffed industries. From a macroeconomic view, transi- tory unemployment is simply an investment, and short-term sacrifices are needed for long-term gains. Japan was able to override hardships by forging a national consensus, but there is no denying that hardships were a feature of its transition. Building a Civil Service Attracting and retaining a cadre of competent, well-motivated individuals, and ensuring that corruption and personal gain do not become the driving motivations for a civil ser- vant's career are not easy tasks. Cost-cutting measures exacerbate this problem. For gov- ernments with shrinking budgets to attract talented individuals in competition with the private sector and, in many instances, the very international agencies calling for bureau- cratic reforms, can be difficult. Recruitment and Promotion Broadly, two general approaches to building civil service systems exist, the job system and the career system. In the job system, used in the Netherlands and in the United States, for instance, an individual is a candidate for a job, and if recruited, individuals remain in this job for the rest of their lives. If they want a promotion or to change jobs, they must compete for the position and perhaps pass a competitive examination. There is lateral entry, and signifi- cant vertical and horizontal mobility. Thus, all personnel are eligible for promotion regardless of their entry level. Such open systems are often market oriented, have fewer age restrictions, and quite frequently use pay variability. They can sometimes be managed at the decentralized level of department managers or agency managers according to cen- tral guidelines. They often advertise to recruit personnel. In most European countries and in Japan, the civil service is organized along career lines rather than job lines. These are systems with closed entry, are rigidly hierarchical, permit limited interclass mobility, and employ highly selective entry requirements. Often there is a single feeder institution, as in France, which perpetuates the notion of a chosen elite and often inducts mainly generalists into the administrative class. Japan draws most of its elite bureaucrats from a small number of schools, the University of Tokyo in particu- lar, although entry is open to anyone who passes the nationally-administered examina- tions, but it is often said that much of the success of Japan's bureaucracy rests on its capacity to take individuals trained for civil service careers, and move them systemati- cally through many different jobs so that they gain a wide variety of experience and skills. The Civil Service and Economic Development: The Japanese Experience 11 Many career systems do their promoting using what is termed the "closed multitrack model," because lateral entry from outside is generally not permitted and the entry track (lower, intermediate, or elite) is what governs how far a civil servant can rise. Japan, Korea, and Taiwan (China) in principle use the closed multitrack system. In other words, under the career system, the path of a civil servant is in effect legally mandated, with the law and the applicant's educational qualifications determining whether and at what level they can enter. Once appointed, career civil servants pass through several grades and ech- elons based on criteria of efficiency, evaluation, and seniority. Once appointed, civil ser- vants know approximately where their career will begin and at what level they will finish. A problem with the career system is that individuals are appointed in a grade, which does not imply any specific function. They carry amorphous titles such as administrator. Administrators do many different things, and from one person and position to another with the same salary and the same place in the hierarchy, functions can vary significantly. More precise job descriptions could improve the career system. Although the closed model leaves little room for abuse, the system is rigid, and intro- ducing new blood is extremely difficult, if not impossible. More often than not, private sector skills brought in by lateral entry invigorate the public service. Furthermore, minis- tries frequently require special advisors in technically demanding fields. Advisors can be brought in from careers in the private sector where they have developed a great deal of specialized competence. However, a negative aspect of lateral entry is the difficulty of exercising objectivity in recruitment. A career civil service has an entrance examination that can be devised to be objective. In many cases, lateral entrants are selected for reasons other than competence. To get around this problem, Korea has left a small window for nonregular civil ser- vants to enter higher levels of the civil service through transparent eligibility requirements and selection procedures. Such a device may have some advantages for governments seeking to beef up their administrative core without compromising the merit-based bureaucracy. To become eligible to take special higher-level civil service examinations in Korea (the examination is not competitive, but is qualifying), one must have served a min- imum of three years in nonregular civil service (for example, as a special assistant to a minister) at the grade level at which entrance is being sought. The main principles governing the French system of recruitment and employment of civil servants are equality and employment security. Recruitment for an administrative position is through a competitive written examination called a concour. Most other Euro- pean countries prefer to leave matters of recruitment to the minister, but some countries such as Greece, Italy, and Luxembourg are trying to widen the use of competitive written examinations because they see them as more neutral and objective. However, implement- ing this kind of system is difficult. In the United Kingdom, the Civil Service Commission, which traditionally has been in charge of recruitment, has yielded some power to a recruitment office, which is an executive agency. Nevertheless, the commission maintains control over the criteria and general policy governing the recruitment of career civil ser- vants. Colloquium participants suggested that although a central authority may play an ini- tial role in screening recruits, it should not be responsible for recruiting personnel for gov- ernmentwide reassignment. Those who aspire to a public service career usually have strong preferences for the kind of job they desire, which is why they choose a particular 12 The Civil Service and Economic Development: The Japanese Experience ministry or agency. Thus, each ministry or agency, as well as local government, should be vested with the authority to appoint personnel. Training The issue of appropriate training systems is extremely important, particularly in transi- tional countries. In competitive models, such as in Canada or the United States, much of the training of civil servants takes place in the private sector, in the sense that this is where they gain the experience and skills they then bring to their government jobs. In career systems these are obtained within the system. In France, successful candidates for the higher civil service are admitted to the National School of Administration, and as trainee civil servants they are remunerated during their two and a half years of study. Their performance is continually assessed and graduates are ranked on the basis of performance. Before being appointed they must sign a promise to serve the state for at least ten years or refund the cost of their National School of Administration training. The United Kingdom uses a university training system, often called the Oxbridge sys- tem because of the predominance of graduates from Oxford and Cambridge in the civil service. The United Kingdom has also developed an in-service training system that is the responsibility of the Civil Service College, which enjoys a high reputation. Elitist Systems As the name implies, imperial China's civil service system is the archetypal mandarin sys- tem. France is a modern example. The grands corps are the elite recruited from the National School of Administration (ENA) for administrative positions and from engineering schools for technical ones. Politically neutral positions are traditionally reserved for them, such as general secretary to the government. Members of the grands corps enjoy almost unlimited secondment or "detachment" to the private sector or to other parts of the public sector, with the automatic right of reintegration. With all due respect to the successes of the French model at home, many other countries have tried it without much success. Although an appealing method, it appears difficult to implement outside the French con- text. Mandarin systems tend to be good at building quality leadership cores, but they have been criticized for a number of reasons, including their inegalitarian qualities; their emphasis on higher-level staff while ignoring the importance of supporting cadres; and their recruitment of too many generalists, and thus not enough technical specialists. Fur- ther, they are difficult to implement in places where education systems are weak. Despite their serious pitfalls, some aspects of these elitist systems offer advantages to developing countries trying to build a professional civil service. Elite recruitment may compensate for other system failings. However, governments should take measures to moderate the elitist characteristics of the mandarin approach, especially endorsing the flexibility to recruit scarce skills from outside the elite when necessary. Elitism among civil servants and politicians needs to be counterbalanced by other institutions and power bases. A strong and free press, think tanks and other intellectual centers where people of The Civil Service and Economic Development: The Japanese Experience 13 stature can criticize the policies of the civil service, a decentralization of some powers to local government, and a strong and vocal private sector can all serve this purpose. For more decentralized systems, one model for building elite cadres is the senior exec- utive service. The aim is to create a servicewide core of higher management professionals, to pay them at super scales, and to provide customized training and conditions of service. Although they have proved difficult to design and implement in industrial countries, many of their flaws can be corrected in adapting them to developing countries. Such pro- grams can complement recruitment rigor, but they need to be carefully designed to include career mobility and training as well as remuneration features, and to maintain their credibility they need to be very careful about excluding political appointees. Remuneration and Performance Remuneration, and incentives in general, are important issues in building a civil service. Two of the most important issues in this area are classification and performance. Two approaches are available with regard to classification, the unified pay and classification system, and the flexible, decentralized system with greater managerial discretion. The central question is whether performance should determine pay. Most countries, including those with some of the most successful civil services, con- tinue to rely heavily on the traditional approach of servicewide unified pay scales that, while theoretically built on merit, in practice usually operate on the basis of seniority. The problems with this method are that it often does not reflect real skill requirements, may be inflexible, tends to keep staff narrowly trained, and discourages creativity. It also encour- ages grade creep, the only means for increasing remuneration. However, it does tend to be fairly enforced across the system and to be relatively easy to manage and to administer. In contrast, many countries are introducing decentralized classification and pay arrangements. New Zealand is the leader of this group. The decentralized norms are gen- erally being introduced within guidelines set by central authorities. Managers determine the pay for certain jobs in certain agencies, and may even be paying rates determined by the market. Efforts also are being made to link salary to performance, and experimentation with broad job classifications and multiskilling as ways to increase salary flexibility is under way in some countries such as Australia. Others, such as the United Kingdom, have gone a step further, sharing savings generated by suggestions with those who make them, thereby providing a financial incentive for staff to come up with helpful ideas. However, such approaches to rewarding individuals have the potential to undermine the collective ethic of the civil service and to open loopholes for manipulators at the expense of the real contributors to the collective effort, who might be less visible team players. Performance is also an issue in promotion policies. Although all civil service manage- ment systems have some formal link between performance and promotion, the degree of consistency between these and actual practice varies considerably. New Zealand has introduced the tightest performance-promotion link, whereby the government measures performance against objective standards agreed upon at the beginning of a review cycle. Performance appraisal also constitutes the building blocks of performance assessments between departments and ministries in New Zealand, and is credited with strengthening accountability throughout the system. 14 The Civil Service and Economic Development: The Japanese Experience Performance-based remuneration and promotion have been criticized because perfor- mance is not always amenable to objective measurement, and it tends not to work as well in settings where group values are stronger than individual ones. In addition, it tends to place greater importance on the achievement of short-term goals to the detriment of long- term, strategic objectives. Furthermore, it can be a demotivator for those who are not rewarded as much as it can be a motivator for those who are rewarded. However, perfor- mance-based criteria do introduce transparency, and this can counter corruption, political- ization, and patronage. A uniform application of personnel remuneration is perhaps the best starting point for developing countries to simplify implementation, ensure impartial applicability, and root out patronage and corruption. Nevertheless, assuming that necessary minimal levels of establishment and budgetary control have been met, unified pay and classification sys- tems in developing countries could arguably be relaxed. The common tendency in many developing civil services toward circumvention of existing unified norms, along with the need to target scarce skills more directly with appropriate remuneration packages, rein- forces this view. The benefits of performance pay remain to be demonstrated in industrial country public sectors, however, and the management requirements and costs of install- ing such systems can be considerable, including dealing with resistance. Given these diffi- culties and the urgency of other civil service management tasks for developing country administrations, they should give a lower priority to this reform. Motivation, Ethics, and Corruption How can one motivate civil servants in a system characterized by substantial employment security? A general idea, but an important one, is ethics. France has a tradition of neutrality and administrative efficiency, and many civil ser- vants are proud to be of service to the state or to local authorities. This, in turn, provides a strong incentive to work hard and to put oneself at the service of the citizenry. A system to repress corruption is also, of course, embedded in statutes. In addition, the French govern- ment has tried several programs to improve the accountability and effectiveness of the civil service, such as the one Michel Roccard implemented in 1989. This program of pop- checks and accountability centers was influenced by total quality control strategies. In return, the civil service was given more autonomy in management, finance, and materials procurement. Civil servants, like all human beings, will serve their self-interests. To assume that they will be more self-sacrificing than the general population is unrealistic. Moreover, in situations of economic crisis and destruction of the social fabric, such as in many former socialist countries, and the former Soviet Union in particular, such self-interested behavior becomes exacerbated. In circumstances where people must concentrate on day-to-day sur- vival, civil servants are not preoccupied with public service. Indeed, part of the state func- tions to guarantee the interests of the civil servants and their families, but not those of society in general. How to organize public employees to serve their country and society in the absence of a cohesive national agenda and in the presence of social and ethnic frag- mentation is a difficult problem. Financial incentives are about the only positive motiva- tors in these cases, and they must be commensurate with fiscal realities. The Civil Service and Economic Development: The Japanese Experience 15 An Overview of the Japanese Bureaucracy Toshiyuki Masujima, vice minister of Japan's Management and Coordination Agency, listed the ten following characteristics of Japan's bureaucracy as especially deserving of attention: * The doors are kept open to anyone with proven ability to become a civil servant. * Within the civil service many measures are available to help candidates for executive positions acquire expertise and experience. These include job assignment leeway, interministry personnel exchanges, and systematic training under the National Per- sonnel Authority. * A high degree of integrity pervades the civil service. * Merit ratings of personnel are strict. Civil servants are continuously evaluated from the day they are recruited. These evaluations take into account appraisals by subor- dinates and colleagues. * There is a clear division of responsibility in the government as a whole and within individual organizations. * Within the government as a whole, as well as within individual organizations, a high regard is placed on decisionmaking by consensus in the planning stages of programs, which ensures coherence, speediness, and uniformity in program implementation. * Adjustment mechanisms to complement the market mechanism have been devised and introduced to prevent a proliferation of public administration. These force con- tinual reexamination of policies and programs and of the means available to imple- ment them. * High-level organs specializing in overall coordination exercise control through man- agement, as distinct from through the budget, for example, the Management and Coordination Agency and the Cabinet Legislation Bureau. They coordinate policies and programs through their authority over organizational structure, manpower, administrative processes, and legislative reviews. * Procedures for administrative reform are available. In the 1980s, a dramatic reform (usually called the Rincho administrative reform) took place against the backdrop of the governments tight fiscal position. One of the most notable reforms was the privatization of giant public corporations, including Japanese National Railways and Nippon Telegraph and Telephone. A drastic reexamination of existing policies and programs also took place that led to the amendment of relevant laws. During 1981-93 the government passed 172 administrative reform laws. Insulating the Bureaucracy An important contrast between the Japanese civil service and the public services of many other countries is their insulation from the political world and from societal interest groups. Although some cogent arguments were made during the colloquium as to areas where the Japanese civil service is not insulated, it is certainly not vulnerable to the kinds of policies prevalent in many developing countries, whereby there is almost total turnover of staff halfway down through ministries when the government changes. 16 The Civil Service and Economic Development: The Japanese Experience A World Bank study (1993, p. 168) discusses the technocratic insulation of the Japanese civil service, that is, its relative independence. At the same time, the Japanese system is often criticized as being part of an "iron triangle" which is comprised of the political lead- ership, the bureaucracy, and industry As far as the allocation of government resources is concerned, the Japanese bureau- cracy seems to be fairly well insulated from vested interests. However, as concerns the question of how to distribute the fruits of economic growth equitably, even Japanese civil servants have found it difficult to be aloof from the demands of political and other pres- sure groups. The typical example is agriculture. Price controls on agricultural products ostensibly exist to promote agricultural productivity. More often than not, however, they have constituted the means to redistribute income from urban to rural areas. On normative issues such as income distribution, decisions cannot be made free of the influence of political and other pressure groups. Inherently political issues have come to the fore in Japan; the values that underlay the bureaucracy's surface objectivity and neu- trality have been called into question. Recognizing that the bureaucracy cannot be truly insulated from politics, there has been a call by the bureaucracy for avoiding such opaque measures as regulation and sectoral protection, and instead making use of general distrib- utive methods, such as income tax and social security. The Structure and Size of Government in Japan Kim (1988, p. 1) has described Japan's public bureaucracy as "one of the smallest and least expensive systems in the world." This is seen in table 1-1, which compares public employ- ment in several countries; table 1-2, which shows that government expenditure as a per- centage of gross national product (GNP) in Japan has been smaller than in both the United States and Germany (the fact that Japan has kept defense expenditures below 1 percent of GNP since World War II has been an important element in keeping government expendi- tures down); and table 1-3, which shows the national burden ratio, which is the sum of tax revenues and social security contributions as a percentage of national income. Table 1-1: The Size of the Public Bureaucracy, Selected Countries, 1990 Number of government workers per 1,000 population Country 38 Japan 112 France 73 United Kingdom 69 United States Note: The figures include only administrative personnel in the national and local governments, plus workers in government corporations. Defense personnel are not included, nor are members of the judicial or legislative branches. In the United States, each of these is large relative to other coun- tries. In addition, the U.S. private sector provides much more medical care than in Europe, and the United States has few public corporations. This means that the U.S. numbers, in particular, are an understatement if one is looking at the size of government employment in relation to the size of the labor force. Source: Prime Minister's Office, Management and Coordination Agency, Japan. The Civil Service and Economic Development: The Japanese Experience 17 Table 1-2. Fiscal Expenditure as a Percentage of GNP, Selected Countries and Years General Capital Social Country Year government Consumption formation security Miscellaneous Federal Republic of Germany 1970 39.0 15.8 4.4 12.7 6.1 Germany 1989 45.9 18.3 2.3 15.2 10.1 Japan 1970 19.4 7.3 4.4 4.5 3.3 1989 32.3 9.1 5.2 11.0 7.1 United States 1970 32.6 19.1 2.5 7.7 3.4 1989 36.0 17.8 1.6 10.7 5.9 Source: OECD National Accounts. Table 1-3. National Burdens, Selected Countries, 1990 Item Germany Japan United States a Taxes 29.4 28.1 26.1 Social security contributions 21.9 11.5 10.4 National burden 51.3 39.6 36.5 Note: The national burden is the sum of tax revenues and social security revenues as a percentage of national income a. Data are for 1989. Source: OECD National Accounts. Table 1-4. Overall Government Deficits as a Percentage of GNP, 1990 Level of government Germany Japan United States a Central -2.1 -0.3 -3.8 Local -0.7 0.3 -0.4 Social security fund 0.9 3.5 1.1 Total -1.9 3.5 -3.1 a. U.S. data are for 1989. Source: OECD National Accounts. Japan's national burden is slightly higher than that of the United States, but substan- tially lower than Germany's. However, the United States is incurring large fiscal deficits. Therefore, the fact that government fiscal expenditure in Japan is smaller than in the United States, while tax and social security revenues are larger, suggests a small and effec- tive government. Finally, a comparison of overall government deficits (table 1-4) shows that even though Japan's central government ran a deficit in 1990, the government recorded a net surplus because of accumulations in the social security fund, while both the United States and Germany incurred deficits in their general accounts. 18 The Civil Service and Economic Development: The Japanese Experience Although low taxes do not necessarily mean a small government presence, much less crowding out of the private sector occurs in Japan than in developing countries, or even in most industrial countries, where large government deficits and debt serve as a damper on private sector borrowing. Comparing public and private sector employment in Japan in 1967, when national government employment peaked, and 1993, the number of national government workers in the executive branch (excluding the defense forces) has declined by more than 4 per- cent. In comparison, employment has increased 73 percent in the private sector and 40 percent in local government. A strong market economy is consistent with small government. If government becomes too large, it may harm the international competitiveness of domestic companies. In Japan a conscious effort has been made to keep government small. This can be achieved through fiscal constraints. For example, in the event of a budget deficit, a special law must be enacted to approve the issue of bonds to finance it. In other words, the government cannot issue bonds simply for the sake of consumption, although construction bonds for public investment are permitted. In addition, as early as 1960, the Government Tax Coun- cil advised placing a limit of 20 percent on the ratio of tax revenues to national income so as to maintain the vitality of the private sector. In the 1980s a long-run limit of 50 percent was set on the national burden ratio. Such measures provide an upper limit on govern- ment expenditures. Japan has a two-tier local government system, the prefectural level and the municipal level. There are 47 prefectures; 3,235 municipalities, including cities, towns, and villages; and also many local public enterprises, such as water supply entities, public hospitals, and police and fire departments. Including the employees of local public enterprises, more than 3.2 million people work for local governments. The central government level has about ninety public corporations either sponsored by or affiliated with the national government. The total number of national government employees, including those working for public corporations, the self-defense forces, and the executive organs of government, is almost 1.2 million. Of these, "nonindustrial" national government employees, which refers to personnel in the general ministries and agencies and to those employed by national schools, amount to more than 534,000. Japan also has more than ninety national universities, to which are attached high schools, junior high schools, and even elementary schools. In addition, the 200 or so national hospitals and sanitoriums employ some 188,000 people. The "industrial sectors," or government enterprises, consist of the postal service (including postal savings and postal life insurance activities), with more than 304,000 employees, and the national forests, printing entities, and mint, which together have some 22,000 staff. The largest national public corporation is Nippon Telephone and Telegraph, with more than 200,000 employees. However, before Japanese National Railways was divided into several passenger and freight companies, it was by far the biggest, at one time employing 400,000 people. Today, because of rationalization, streamlining, and privatization efforts, the total work force of Japanese National Railways is around 200,000. Note that although the government has initiated privatization programs for these public corporations, their employees are still categorized as belonging to the public sector, because they are still majority-owned by the government. The Civil Service and Economic Development: The Japanese Experience 19 The number of central government personnel has been decreasing since 1967, but has been increasing rapidly at the local level. This parallels the decentralization of govern- ment functions, although the local increases have not been completely a result of decen- tralization or burden shifting. Although the central government funds only half of education expenses, school teachers and clerical staff are counted as local government employees. One-third of the total increase is explained by that factor. In addition, the number of police officers increased by around 70,000, and the number of firefighters by 7,000. Moreover, many local public corporations were formed to provide services. "Hidden" Aspects of the Japanese Civil Service The Japanese government is widely believed to be effective, but at the same time it is com- paratively small. However, as Yutaka Kosai notes, it has "hidden" elements that magnify its power and influence. One such institution, unique to Japan, is the Fiscal Investment and Loan Program, directed by the Ministry of Finance's Trust Fund Bureau, which is dis- cussed later. In some cases regulation can substitute for government expenditure, thereby reducing the fiscal burden. For instance, agricultural protection can be carried out in two ways. One is to provide subsidies. However, subsidies involve direct government expenditure. The imposition of an import quota or a complete prohibition of imports, though it transfers the burden onto consumers in terms of higher prices for agricultural goods, requires minimal government expenditures. Thus, the cost of agricultural protection as a percentage of GDP in the United States is estimated at 1.5 percent, with 1.1 percent in the form of subsidies, so that taxpayers bear most of the burden. In Japan the cost of agricultural protection amounts to 2.5 percent of GDP, with only 0.5 percent borne by taxpayers and the brunt of the burden falling on the consumer. The use of administrative guidance by Japanese civil servants, which takes the form of ministerial directives and suggestions provided to the private sector without explicit legal power of coercion, is well known. It is by no means legally binding, but it can be an effec- tive instrument in pursuing public policy. Although certain prerequisites for the effective- ness of such moral suasion must be met, such as the creation of an intimate relationship of trust between the private sector and the government sector and a belief in the respectabil- ity of civil servants, guidance provides a way for government to influence the private sec- tor without direct expenditure. However, the recent passage of the Administrative Procedure Law limits the use of this procedure. Is administrative guidance and regulation good or bad? Take consumer taxation, that is, substituting rationing for taxation. A subsidy is more transparent, because it can be directly measured and its level debated, but the costs of regulation are harder to deter- mine, compounding the ill-effects of a subsidy with a lack of transparency Small size is important in terms of bureaucratic efficiency, but openness may be even more important in some cases than the actual size of government. Although administrative guidance is in some cases quite effective, it is open to bureaucratic abuse. Controlling the Size of Japan's Civil Service Of the methods available to limit the expansion of the civil service, the actions of the bud- getary authorities in Japan (Ministry of Finance) and Korea (Economic Planning Board) 20 The Civil Service and Economic Development: The Japanese Experience are noteworthy. These authorities put great stress on maintaining macroeconomic stability through fiscal balance and on avoiding any misdirection of resources away from produc- tive investment in the private sector to potentially "unproductive" government consump- tion. An across-the-board downsizing of ministries, such as the one the Sato government undertook in the late 1960s, is a crude but at times effective approach to reducing the size of the civil service. The Japanese government also deliberates such matters through blue ribbon commissions. In the 1980s two such commissions undertook administrative reform, with tangible outcomes (see Lincoln 1988, pp. 116-22). In Japan, the Government Organization Law has imposed a strict limit on the number of bureaus in the government, and the Total Staff Number Law sets ceilings on the num- ber of civil servants in each agency. The Government Organizational Manual, published annually for public dissemination, contains a list of the ceilings, which can only be changed with the approval of the cabinet and the legislative branch. An upper limit is placed on the organizational size of the Japanese government. At present, the Japanese national government has one office and twelve ministries under the cabinet. The office is the Prime Minister's Office, which contains eight agencies that are the equivalent of ministries. Their respective establishment laws delineate the functions of these ministries and agencies. The first-tier organizations within the ministries and agen- cies are the minister's secretariat and the bureaus. A governmentwide ceiling for the num- ber of such organizations was written into the National Government Organization Law. Since 1980 Japan has had 128 secretariats and bureaus, which is the prescribed ceiling. To add a bureau, the National Government Organization Law would have to be revised. The functions of the secretariats and bureaus are provided by cabinet order. A secretar- iat or a bureau may sometimes have a deputy director (icho) or a department (bu) within it, but the basic functional units are divisions (ka). The whole government has less than 1,500 divisions. Their names and functions are also specified by cabinet orders. A cabinet order is necessary to finalize the allocation of new functions to cope with a demand for new services. The enforcement of the "scrap and build" principle is another method by which the size of the Japanese bureaucracy is kept in check, in this case by preventing the unneces- sary proliferation of government organizations. The Management and Coordination Agency (MCA) is authorized to examine and decide whether or not to grant the request for the creation of organizations of any size within the executive branch. It grants permis- sion to create a new unit only if an organization of equivalent level is abolished from within the ranks of the executive organ requesting it. This principle has been rigorously upheld. The philosophy behind it is that a new organization is an additional decisionmak- ing unit within the government, and thus an increase in the range of government opera- tions. Scrap and build curbs the proliferation of the sphere of government activities. Prior to the enactment of the Total Staff Number Law, staffing levels were stipulated in the law establishing each ministry and agency, and thus the Diet's approval was necessary for any changes. This procedure was cumbersome and took time. In 1967 a new system was introduced, whereby the number of personnel was set by cabinet order, but the law set an upper limit on the government's total nonindustrial permanent work force. Only special personnel, such as the self-defense forces and the chamberlains and stewardesses of the Imperial Household Agency, are exempt from this law. The upper limit has not been changed since the law's enactment in 1967. The Civil Service and Economic Development: The Japanese Experience 21 Even before the law was enacted, the number of industrial permanent employees of each government enterprise was decided by Cabinet order, not by law, but because enter- prises require greater flexibility to respond to business conditions. Also, since their total expenditures on personnel were controlled, the Diet did not also need to control the num- ber of personnel by law. Personnel Reduction Plans During the high-growth era of the 1950s and 1960s, public programs expanded rapidly, as did the number of public employees. During 1957-67, the number of government employ- ees increased by more than 20,000 per year on average. How to stop this tendency and to control the size of the government work force, while still maintaining the responsiveness of the bureaucracy to the ever-changing demands of civil society, was an extremely impor- tant issue. The result was the 1967 Total Staff Number Reduction Law and the first person- nel reduction plan. The return of Okinawa in 1972, and other factors, such as creation of the Environmental Protection Agency, prevented much in the way of an overall reduction in government size (less than 0.1 percent) during the first fifteen years of the law and per- sonnel reduction plans, but beginning in the 1980s there has been a sustained shrinkage in staff. The control mechanism involves enforcing a personnel reduction plan under which each ministry and agency is assigned a target figure for personnel cuts to be achieved through attrition. This process proceeds on both an annual and on a long-term basis. An annual review of the staff requirements of each ministry is performed by MCA. Every year each ministry must submit a request to the MCA to change its staff size. To cope with increasing work loads, a ministry can ask for additional personnel. The MCA examines the requests on the basis of several criteria, including the jurisdiction of the min- istry concerned; the extent of overlap with other ministries' need for any new program, and the possibility of contracting out the work or reallocating existing staff. The MCA also takes into account projections of vacancies, turnover rates, future work loads, and the like. These reviews are conducted in the context of reduction targets for a certain period, usually of from one to four years, set by cabinet decision for each ministry. The MCA pre- pares a proposal for the cabinet and negotiates it with the ministries. The rate of personnel reduction is different for each ministry, and the responsibility for reaching the target num- ber is shared. The MCA keeps in close contact with the personnel offices of each ministry, which are responsible for implementing the reduction program. When an increase in staff is granted, it is done by drawing on the staff number account. As a result, staff reallocations can take place within the limits of the Total Staff Number Law. A mixture of pressure and initiative is characteristic of the Japanese system. The MCA presses ministries and agencies to rationalize or streamline unnecessary programs, while individual initiative is encouraged. A sense of balance and mutual respect contributes to the MCA's credibility in the eyes of the agencies, and to a continuing relationship that can produce results. Initially, the Personnel Reduction Plan faced stiff opposition from unions of govern- ment workers, as well as from the labor movement in general. Resistance was especially strong at the Ministry of Agriculture, which was assigned the highest ratio of staff reduc- tions. Personnel officials in each ministry first tried to explain to the staff the need to 22 The Civil Service and Economic Development: The Japanese Experience improve efficiency in the public sector. The argument they presented was that the private sector was making considerable efforts to improve its efficiency, so why should the same principle not apply in the public sector, where taxes from the general public are used to cover costs. Ultimately the unions agreed, but asked for an increase in employment in cer- tain programs, and for balanced reductions across ministries. Opposition was also reduced through the use of natural attrition rather than dismissal. Discussion of Japanese Staffing Management The colloquium participants were intrigued by Japan's experience which suggests that a country can decrease the size of its civil service without layoffs. In Lebanon, for instance, the civil service has been unable to get rid of inefficient staff through legislation that was copied from France at a very early stage of civil service development. Zero based budgeting, scrap and build, and personnel reduction plans are primarily attempts to create a mechanism conducive to a constant reexamination of policies, organi- zations, and so on. The notion of fixing the number of personnel, bureaus, and divisions in effect creates an important principle within agencies that goes completely against the con- cept of an ever-expanding bureaucracy, namely, if you want to do something new, then tell us what you are doing now that you can stop doing. In this way, notions of creativity and adjustment are subtly built into the Japanese civil service. The basic philosophy is that as the economy and society evolve, inevitably some of an agency's activities will decline in importance. As a result, there will always be room for rationalization, including the simplification and mechanization of some activities and the contracting out of others, and the agency can identify such areas. Japan has used this approach effectively for more than twenty-five years. Japanese-Style Personnel Management The three core features of Japanese personnel management have been lifetime employ- ment, a seniority-based wage and promotion system, and enterprise-based or in-company unions. These are often considered to have contributed significantly to the success of Japan's economy in both the private and public sectors. The economic realities facing con- temporary Japan have called the viability of the first two into question, but they worked well for at least the first thirty years or so after World War II. Interestingly, the government never made quite the same commitment to lifetime employment and seniority-based advancement that large private businesses did, at least as regards elite civil servants. Group work and participatory management, often referred to as the ringi system, which is essentially a bottom-up decisionmaking system, are also important features of Japanese-style management. Although the process is often criticized for being slow because of the need to reach consensus, implementation tends to be rapid because of this prior coordination. Recruitment and Promotion In general, Japan's civil service attracts the best and brightest members of the population. Candidates are screened through three levels of competitive examination and evaluation, although there has never been a formal educational requirement. Competitive examina- The Civil Service and Economic Development: The Japanese Experience 23 tions consist of a general culture test and a test in areas of specialization, such as public administration law, civil engineering, or forestry. The National Personnel Authority (NPA) administers the tests. This agency compiles rosters of successful candidates for each field of specialization, ranking them in order of scores earned. The ministries and agencies select people from these lists for interviews, but usually far fewer openings are available than the number of successful candidates. Koh (1989) provides a detailed description of recruitment (chapter 4) and promotion (chapter 5). The principles guiding recruitment are equal treatment of candidates, except that they must be twenty-one to thirty-two years old; open door, competitive examinations; and the merit system. As no lateral entry is permitted, Japan's civil service is often referred to as a closed sys- tem. However, as far as initial recruitment is concerned, it is completely open and compet- itive. Employees needing specialist licenses, such as doctors and teachers, are recruited through evaluation processes. Three kinds of entrance examinations exist, two for university graduates and the so- called Level 3 for senior high school graduates. Core administrators are only appointed and promoted from those passing the Level 1 or Level 2 examinations. Generally only those admitted at Level 1 reach the highest positions, such as administrative vice minister or director general. Those who pass the examinations are recruited individually by each agency. Although appointing authority rests with the ministers, in practice it is often delegated to lower lev- els within the ministry. Each agency provides new hires with an appropriate orientation to convey its transorganizational bonds and to inspire loyalty to the agency and its objec- tives. Although the emphasis on allegiance to one agency, or, more precisely, to one bureau, makes developing substantial expertise in the administrative elite easier, this tra- dition has bred so-called sectionalism, both at the interagency and intra-agency levels (Koh 1989, especially chapters 5 and 6). Common expressions such as "there are bureaus, but no ministries" aptly convey the intense preoccupation with bureaus in each agency of the Japanese government. The National Public Service Law stipulates that promotion should be based on com- petitive examinations, but few ministries or agencies conduct such examinations. Instead, promotion initially depends primarily on the length of service, although functional pro- motion depends on merit. With each promotion the level of responsibility and prestige increases, which is believed to provide a tremendous incentive. In contrast, in France, a promotion assures only the grade and the salary, not any type of position. The result is that people at the same grade-that is, having the same salary level-have widely differ- ing responsibilities and authority. Another method used in Japan is continuous competition. There are two cycles in the career of a Level 1 Japanese civil servant. The first lasts about ten years, from recruitment to becoming an assistant section chief. During this stage promotions are on a seniority basis. From section chief to administrative vice minister, the top level attainable, advance- ment is based on competition. What is fascinating about the system is the concept of nenji or "class" of recruits. All public servants recruited in the same year are in the same class. All will probably reach the level of assistant section chief. Those who do not often find a second career in the private sector or in government corporations, a process known as amakudari. 24 The Civil Service and Economic Development: The Japanese Experience The following demonstrates how keen the competition for director general positions among members of the same class becomes. Each year large ministries recruit roughly twenty Level 1 candidates from the law, economy, or administration examinations. Only three or four of these become director generals, and only one becomes administrative vice minister. Also, as administrative vice ministers often serve for two years, some classes may be denied that position completely. Incentives Competition among agencies for the right to undertake crucial tasks promotes intense loy- alty within each agency as well as a collective effort to gain the upper hand over other agencies. Because each agency has to compete for additional tasks and functions, agencies compile a great deal of information and conduct numerous analyses to this end, all of which the government-the ruling party-uses to arrive at policy decisions. This process, though prolonged, ensures that decisions are well-informed, and thus their implementa- tion faces fewer obstacles. The overriding test is whether the decision will enable an agency to promote growth in terms of the government's strategic objectives. In Japan's system, the bureaucracy is tightly controlled by the imperative of fiscal bal- ance. Hence, it is not subject to Parkinson's law, as are civil services in other countries, and does not have to prove its effectiveness and efficiency by its performance. This is possible in part because of the effective coordinating role performed by the Ministry of Finance in its budgeting processes and procedures. Furthermore, the prime minister and the ruling party can initiate administrative reforms; these reforms and their implementation serve as checks on bureaucratic expansion. Salary and Compensation Structure It is often said that if salaries in the public sector are not competitive with those in the pri- vate sector, the best people will not seek employment in the public sector. The Japanese case seems to indicate the exact opposite. While salaries in the public sector are lower than those in the priate sector, the cream of the crop desire employment in the Japanese civil service. The NPA is constantly comparing the salaries of public servants with those of similar positions and age groups in the private sector. The rule is to maintain equality with the salary scales of private companies with more than 100 employees. Pay raises in the private sector are usually followed by pay raises in the government, but with a lag of several years. According to the NPA, in 1993 the differential between sectors was less than 3 per- cent. Nonetheless, estimates indicate that top-level bureaucrats earn about 30 percent less than private company executives, and that many talented thirty-five year old public ser- vants would get 10 percent more for doing similar tasks in the private sector. However, other forms of compensation exist besides cash salary. For instance, although the fixed retirement age for public employees and at most major companies is sixty, only about 1 percent of public employees actually work until then. Government offi- cials reaching their mid-fifties are routinely introduced to new jobs in the private sector or in government corporations. This is called amakudari (descent from heaven), but would probably be more appropriately termed "ascension from hell," according to some observ- The Civil Service and Economic Development: The Japanese Experience 25 ers. In many cases, their new positions involve a higher prestige job, and possibly higher pay. Corruption Although Japan has no anticorruption laws, if civil servants engage in illegal actions, they can be subject to such disciplinary actions as dismissal, reprimands, reductions in pay, or probation. Because only the top candidates from a population with a very high literacy rate and a high proportion of university graduates are recruited into the civil service, pub- lic servants tend to be highly educated, highly motivated, competent, and disciplined peo- ple. Add to this the relatively good remuneration of Japanese civil servants, the fact that very few engage in activities for illicit gain is not surprising. Disciplinary power rests with each ministry and agency. The NPA also has disciplinary power over all public employees, but it has never been invoked. Basically, appointing officers make decisions about disciplinary actions regarding their respective employees. Although no investigation is undertaken prior to the disciplinary action, a process of appeal to the NPA exists if the employee objects. Amakudari raises the issue of compromising the independence and integrity of govern- ment bureaucrats and conferring unfair advantages on firms hiring former officials. How- ever, retiring officials are placed by the NPA, they do not select their own employers. This discourages collusion between civil servants and business firms. Moreover, top civil ser- vants are prohibited from accepting executive positions in a profit-making enterprise for two years after retirement, five years if any connection exists between the new job and their prior government position. Some 200 senior bureaucrats take executive positions in the private sector every year. What do retired bureaucrats who go into the private sector actually do? Aside from former technical employees with specific skills, retired public employees in high executive positions become links between the private enterprise and the public sector. They often use their connections to obtain public contracts or to loosen restrictions on licenses and permits. As other large firms will have their retired civil servants doing the same thing, arguing that the firm has acquired an additional advantage is hard. In the United States, retired government officials circulating into the private sector also often find jobs that relate to their influence networks or government contacts. In a scenario of deregulation, with no need for licenses, approvals, and so on, where the civil service will be less important in the decisionmaking of private firms, the net- working skills bureaucrats bring with them will be less highly prized. In other words, part of what public sector officials now bring to the private sector will become less valuable. The only political appointees in the Japanese civil service are the ministers themselves and the parliamentary vice ministers. They are rarely given appointments in the private sector after retirement. Some retiring bureaucrats enter the political arena, which conceivably can compromise their objectivity and neutrality in dealing with politicians. If one considers the prerequi- sites of a successful political career in Japan, including finding a political patron and a fac- tion to align with, the risk that civil servants may compromise on certain issues to ensure themselves a political career at a later point is clearly a possibility. Overall, however, the method of Japanese organizational decisionmaking centered on the formation of consensus reduces such risks. The process of consensus-building can 26 The Civil Service and Economic Development: The Japanese Experience neutralize the decision of any individual. In countries where hierarchy and decisions at the individual level are emphasized, a revolving-door system like that in the United States or an amakudari system like Japan's carries considerable risks. The Integrity of the NPA and Technocratic Independence The paradox of corruption scandals in the political realm and the apparent lack thereof in the bureaucratic one is riveting. The existence and integrity of the NPA seems to have had a great effect on the quality of Japanese civil service. In particular, it seems to have been able to protect the integrity of the recruitment of civil servants in a political environment that is not free of corruption. Four elements seem to safeguard the NPA's integrity, namely: * The three commissioners are appointed by the cabinet for four-year terms, and they must be approved by both houses of the Diet, which means that there is a legislative sanction in addition to the executive sanction. * The three commissioners' independence is assured in that once appointed, they can be removed from office only through impeachment. * Commissioners may be reappointed, but only for a maximum of three terms (twelve years). * To safeguard against nepotism and collusion, no two commissioners can belong to the same political party or be graduates from the same university. The NPA's monopoly over recruitment further ensures the integrity of the system. This is especially clear if we look at the African case, where ethnicity and nepotism often come into play in civil service recruitment. Monitoring, Inspection, and Auditing In the Japanese system, several organizations monitor, inspect, and audit the performance of administrative agencies and public corporations. The Management and Coordination Agency (MCA) conducts inspections and nearly all administrative agencies and public corporations have their own internal auditing divisions. These functions are executed within the framework of the executive branch or cabinet. As an external auditing organ, the Board of Audit examines and reviews the final accounts of administrative agencies and public corporations. Audit coverage for 1993 was only 9 percent, but for the more important head offices coverage was 43 percent. The Board of Audit is a constitutional organization, as opposed to the ministries and agencies created by laws passed in the Diet, and is therefore independent of the cabinet. Indepen- dence of auditing is vital to eliminate interference by the cabinet or the Diet. The Coure de Comptes in France, the Auditor General's Office in the United Kingdom, and the Board of Audit and Inspection in Korea perform similar functions. One function of the MCA is to recommend improvements in administrative systems and operations. In other words, the MCA is a kind of outside party monitoring the opera- tions of ministries and agencies, although it is actually a ministerial agency within the executive branch. The MCA's administrative inspections are especially important because of the lopsided concern of government officials, politicians, and the mass media over the planning as opposed to the evaluation of government programs and operations. Inspec- The Civil Service and Economic Development: The Japanese Experience 27 tion provides a means of self-criticism for the government as a whole, in that information about the actual performance and results of administrative systems and policies is fed back to be used for better planning. The scope of administrative inspection is broad. Virtually no limits are imposed on the matters that can be taken up or on the criteria to be employed in carrying out inspections. Common criteria include efficiency, effectiveness, fairness, and transparency. As resources are limited, inspections are carried out as systematically and efficiently as possible based on three-year plans. The reports and recommendations of investigations are made avail- able to the public after being issued to the relevant ministry or agency, and the more important ones are reported to the cabinet. The focus of operations of the MCA and its predecessors has shifted gradually. In the 1970s and 1980s the emphasis was on assisting administrative reform. Investigations were mostly ex post and involvement in the implementation of particular programs was rare. Since the late 1980s and into the 1990s, however, the MCA has been monitoring programs. Administrative Counseling Japan has a process for redress of citizen's grievances against the civil service similar to the ombudsmen and parliamentary commissioner systems in other countries. Operated by the Administrative Inspection Bureau (AIB) within the MCA, it receives grievances and uses its good offices to settle the matter with the ministries and agencies concerned. More than 5,000 administrative counselors-locally respected private citizens such as doc- tors, teachers, and business people, as well as housewives and former civil servants-vol- unteer their services without pay. They help receive and communicate the public's grievances to the AIB, the agency involved, or other concerned parties. The Administra- tive Law sets out the status and functions of administrative counselors. Raising the aware- ness of officials who are in daily contact with the public about their attitudes, pronouncements, and the general environment in which they provide services has also been a target of the AIB. Administrative counseling is an offshoot of administrative inspection. It was initiated in 1955 because the large number of grievances received during inspections was deemed to warrant separate treatment. A link has remained between inspection and counseling. Indeed, they are inseparable components of monitoring and evaluating the activities of government ministries, agencies, and, in some cases, local governments. They provide an important source of feedback to the central government, the cabinet, and the prime minis- ter about what is happening at the point where services are actually provided to the peo- ple. Sometimes grievances are the visible tip of a huge problem hidden within the bureaucracy. Thus, several administrative inspections often stem from a single grievance. The connection between administrative inspection and counseling provides an incentive to officials to take complaints seriously and to respond swiftly. Administrative counseling is a service provided free to all the people of Japan. When someone has a complaint, he or she can either go to the administrative inspection office or they can visit the local administrative counselors who report it to the regional office of the AIB. Bureau personnel then contact the agency involved and ask for an explana- tion. Although the AIB has no legal power to implement or even to recommend a solu- tion, 86 percent of the complaints filed with the AIB have been resolved in favor of the complainants. 28 The Civil Service and Economic Development: The Japanese Experience A system of litigation against the civil service is also available, but no separate admin- istrative court exists. People can file suits in the ordinary courts, but this is time consum- ing and costly, a problem common in other countries as well. Administrative counseling, in effect, complements the formal system of redress against the public administration. It is simple and inexpensive, yet perhaps more effective than the formal system. State Intervention and Economic Development in Japan This section will focus on the colloquium's conclusions regarding the broader theme of state intervention and economic development. The contributions of the civil service to economic development are implicit if not always explicit in the discussion of state inter- vention, since the civil service serves as a key structural element of government. An activist state in Japan has been justified on two grounds. The first is the goal of dynamic technological change, which requires an activist state in the face of market fail- ure. The second justification involves a coordinating role for the state in an effort to reduce transaction costs, which has been raised as a new institutional interpretation of how Japan has successfully intervened. The Japanese experience makes clear that it is the form rather than the degree of state intervention which is critical in the context of late development. In the case of the Japanese government's promotion of R&D associations, which constituted a violation of the coun- try's antimonopoly and fair trade laws, association activities were not heavily subsidized. Rather, the government acted as a catalyst in forming these consortiums. For instance, to catch up with IBM in the early 1970s, the government established three consortiums among six different firms, which the Ministry of International Trade and Industry had ear- lier unsuccessfully tried to merge into a single firm. Bringing these parties together was a much more effective form of government intervention than mere subsidies, which Minis- try of International Trade and Industry says amounted to Y160 billion over four years or US$700 million at the exchange rates prevalent at that time. Further, the Japanese experience suggests that an interactive relationship between the government and the private sector is essential. However, to assume that economic devel- opment is somehow going to be painless, costless, or easy is to make a major mistake. Maintaining or establishing equity is important, but by the same token, the Japanese expe- rience of the late 1800s points to a great deal of inequity Capital clearly was transferred from the rural sector into industrial development. In the postwar period immediately fol- lowing the occupation, the Japanese government severely restricted domestic consump- tion. Domestic production of consumer goods was limited and their import was severely restricted. Consumers wound up placing their money in savings, which were used for development. Policy decisions will always result in income transfers. These decisions need to be transparent so that their welfare costs are fully appreciated. Support for Small and Medium Firms Japanese policy toward small and medium enterprises illustrates the importance of the private sector. The World Bank conducted a survey of small and medium enterprises in Colombia, Indonesia, Japan, and Korea, asking what the relative importance of private and public support was in their success. In the Japanese case, companies were asked about financial, marketing, and technology support. The overwhelming response was that The Civil Service and Economic Development: The Japanese Experience 29 private support was more important than public support. This held even in the case of financial support, where public funding was relatively cheaper than private funding. Jap- anese government support to small and medium enterprises is embedded in a highly effective and efficient private financial system. Public support has acted in a countercycli- cal fashion, offsetting periods when the banking system was not ready to provide loans. One interesting finding was that local public support was more highly valued than central public support. All the successful entrepreneurs surveyed in all the countries and sectors, even includ- ing sectors with very small enterprises, such as carved furniture production in Indonesia and leather goods manufacturing in Colombia, had at least some secondary education. This implies that an important role for the government in promoting the growth of small and medium enterprises is the development of a strong educational system. The breakdown of education expenditures in East Asian countries, including Japan, averages two-thirds for basic education and one-third for higher education. By contrast, in many developing countries, particularly in Africa and the Middle East, this proportion is reversed, with one-third of expenditures going to basic education and two-thirds to higher education. This suggests a need to address some fundamental problems, and the civil service itself is not-and should not be-responsible for doing so. Rather, determin- ing the allocation is a political decision. Response to the Oil Crises Interventions by the Japanese civil service have succeeded in the sense that they have sup- ported market forces and competitiveness. This is especially well illustrated by the reac- tion of the Japanese government to the 1973-74 oil crisis. Four measures stand out. First, oil price increases were passed on to consumers, which promoted energy conservation. Second, the government introduced fiscal and monetary stringency, which kept the rate of inflation down, and thus kept wage demands in check. Third, nuclear power was pro- moted. Fourth, distressed industries were assisted in adjusting, but they were not pro- tected. This was thus a case of easing adjustment, but not preventing it. This contrasts sharply with Latin America, where many countries reacted by borrow- ing heavily in international capital markets and delayed adjusting to international oil prices. Some countries such as Ecuador and Venezuela have still not come close to adopt- ing international oil prices in their domestic markets. Not only was the introduction of fis- cal and monetary stringency slow, it usually occurred only after a crisis situation had been reached. The introduction of protectionist measures in many countries stands in sharp contrast to the Japanese reaction of restructuring and diversifying, as well as of allowing some companies to perish where necessary. Creation and Allocation of Credit Since World War II the Japanese government has been actively involved in financial mar- kets, both as a collector of deposits through the postal savings system and as a supplier of funds through the Fiscal Investment and Loan Program (zaito). Zaito is a financial credit system, or resource allocation scheme, unique to Japan that has played a vital role in enhancing Japanese economic development and improving the quality of life. 30 The Civil Service and Economic Development: The Japanese Experience The program's priorities have changed from reconstruction (land conservation, disas- ter reconstruction, transportation, and communications) and modernization of key indus- tries in the immediate postwar period to construction (housing, roads, railways, and water supply installations), helping medium and small enterprises, and even overseas economic aid. In the 1990s, with economic stagnation and tax revenues falling short of expected levels, the role of zaito has perhaps become even more important. The thinking behind the system is that public policy objectives for economic develop- ment and enhancing the quality of life are more than the government can undertake with its limited tax revenues. Zaito allows the government to pursue a greater variety of objec- tives by extending loans. Although projects are expected to repay what they borrow, this does not mean that the program excludes ones that will not generate operating cash flows sufficient to amortize their debt. However, in such cases, the government need only spend tax revenues to the extent that a project cannot pay its own way from user fees and the like. A typical case is railway construction. Fares, the main source of revenue for debt repayment, are kept low as a matter of public policy, so subways and urban railways are not profitable on a direct cost-accounting basis. Tax revenues are used to subsidize the sys- tem. Individual house buyers are also subsidized under certain programs. The government sets plans for zaito in parallel with the formulation of the national budget to keep them consistent. Both are submitted to the Diet for approval. The zaito pro- gram is often called the "second national budget" because of its scale and role in the nation's economy. For fiscal 1994, the zaito budget is more than V47 trillion, which is nearly 10 percent of expected GNP and nearly two-thirds the size of the national budget. Zaito funds are disbursed through more than sixty public institutions, which borrow from the government at lower interest rates than they could obtain from private financial institutions. The average maturity period of such loans is more than fifteen years. Interest rates are fixed. For example, in March 1994 the long-term zaito rate was 3.65 percent, com- pared to the 4.40 percent long-term prime rate offered by private banks. Rates on new loans are reviewed monthly, taking full account of market rates, particularly the ten-year government bond, which is the bond market's benchmark. Some of these institutions, such as the subway system and airport authorities, use the funds directly on projects, while others lend to private sector businesses and individuals. When zaito began providing loans to the steel industry, private investment was crowded in, that is, private sources of capital considered the government funding a signal to invest in the industry. By contrast, zaito loans to the coal and machinery industries seem to have crowded out private investment, such that a net addition to investment may not have taken place. With most developing economies going through financial market liber- alization, funding strategic investment at below market rates becomes a near impossibil- ity. The modern phase of zaito seems to be trying to correct the earlier distortions it had, at least partly, created. The context in which zaito has operated is important. Yukio Noguchi noted that zaito was able to play such a valuable role because Japan's saving rate has been high and because the program has been administered with limited influence from interest groups. In addition, Japan has been fiscally prudent, and this fiscal stability helped the program. Finally, financial markets were administered, which included preventing development of a bond market, so that indirect financing was the only way to finance long-term projects. The Civil Service and Economic Development: The Japanese Experience 31 Zaito funds come primarily from postal savings and the national pension system. Since reforms took place in 1985, everyone aged twenty to sixty is required to join the National Pension Program, which provides a flat rate basic benefit. At the same time, a second tier was inaugurated called the Welfare Pension Program, which provides earnings-related benefits. The amount in the combined pension programs is more than V100 trillion (some US$960 billion). The fund currently has a large surplus because the ratio of contributors to pension recipients is quite high, 5.7 to 1, compared to 5.5 in the United States, 4.2 in the United Kingdom, 2.8 in Germany, and 2.1 in France. Over 100 years ago, when Japan had few bank branches, post offices started to accept deposits as a convenience to the public, and more than 20,000 post offices all over the country still collect deposits. The amount of postal savings is more than US$1.6 trillion, while the deposits of the 150 private banks total around US$3.8 trillion (September 1993). The Ministry of Posts and Telecommunications operates the system, and it is recognized as the largest financial institution in the world. The deposits it receives are redeposited with the Ministry of Finance, which makes them available to zaito. Postal deposits have generally offered higher effective interest rates than private financial institutions were allowed to (their rates were controlled), which meant private banks were denied funds. Although postal savings systems exist in many Asian and European countries, using the funds raised in a directed investment program seems to be unique to Japan. In Bang- ladesh, for example, postal savings are merged with the general account of the govern- ment. Taiwan (China), however, is also using postal savings as a tool of public policy to mobilize the pool of domestic capital for industrialization, but instead of creating a system like zaito, postal savings are placed in the central bank. The Japan Development Bank and Directed Investment In the wake of World War II, the government established a number of institutions to help reconstruct the war-torn Japanese economy. The Japan Development Bank (JDB) was established in 1951 as a government-owned financial institution when the Reconstruction Finance Corporation and U.S. Agency for International Development Counterpart Fund were abolished. Its role has been to provide low interest, long-term financing, which com- plements private financing, to the areas deemed most likely to improve the national econ- omy. Those areas have, of course, changed given economic growth, and with the changing needs of society. The JDB specializes in term lending rather than in providing working capital loans. Although the interest rates charged are below market, they are positive in real terms. The JDB's activities are confined to domestic projects, and it does not lend to the public sector. In 1993 it had outstanding loans equivalent to US$120 billion and a staff of 1,100. Japanese policy-based financial institutions such as the JDB are assured autonomy regarding their specific loan decisions, including which individual projects to fund and the specific amount to be funded, although the government formulates general policy and basic plans. The JDB carries out its own assessment of whether a project conforms to pol- icy, the creditworthiness of the target firm, and so on. Indeed, a condition imposed by the JDB's first governor was that no politician ever ask him to provide a loan to a nonviable project. With an initial assurance of independence, which has been honored since, the JDB proceeded to focus its efforts on project appraisal capabilities and monitoring. After pro- 32 The Civil Service and Economic Development: The Japanese Experience viding funds, the JDB verifies how the funds are being spent and the impact of its invest- ment. Monitoring has been a crucial facet of the JDB's lending procedure. For policy-based finance to be effective and efficient, the free market mechanism must be respected, and JDB operations are based on sound financial principles. This has con- tributed to the fact that, unlike the numerous failures of development banks worldwide, particularly those owned by governments, the JDB is considered a success. Conclusions The three-pronged objective of the colloquium was to find out more about the Japanese civil service system, to compare it with those of other countries, and to draw some lessons from the Japanese experience for developing countries. The most germane issues and con- clusions in terms of these objectives are summarized here. Keeping the size of the bureaucracy small is one way to make it reputable. If every- body can enter it, then it cannot be reputable. Some sort of selectivity in hiring has to be exercised. In this sense, the system of open and competitive examinations is a good approach. Legal and budgetary safety valves are also needed to guard against undue growth of the bureaucracy. The context in which the bureaucracy operates is important. Fiscal restraint must be exercised, inflation must be checked, and private investment must not be crowded out. Strict financial management needs to be ensured by independent auditing. Most countries are still in the rudimentary stage of auditing, concentrating only on financial audits and giving little or no attention to performance audits. Both are necessary. In many countries the state apparatus is overextended-the government is all-intru- sive and pervasive. This needs to be stressed because resources are not abundant, indeed, the essence of policymaking is that resources are scarce, and great care must be taken to allocate them efficiently. A stable and efficient government is indispensable for the devel- opment of a nation; therefore it is crucial to retain excellent public employees and to create small, but vigorous, governmental organizations. Meritocracy is desirable for building a reputable civil service, despite the problems of elitism. Many meritocratic civil services, however, tend to provide little opportunity for lateral entry, and thus cannot directly absorb talent from the private sector. When lateral entry is allowed, guarding against political patronage is crucial, perhaps by requiring lat- eral entrants to take rigorous entrance examinations. In many countries there is a conflict of interests and even a considerable degree of mis- trust between politicians and bureaucrats. In a democracy, the civil service cannot behave as if it were the sole arbiter of the public interest. In other words, it cannot be completely aloof from the political world. Boundaries between what governments should do and should not do need to be defined by politicians in consultation with bureaucrats, after which the foremost task of the civil service should be the loyal execution of the policies of the government of the day. Nevertheless, barriers against political influence in technical decisionmaking need to be devised, and political patronage must be effectively avoided. The system must also allow for local initiative and autonomy. In many developing countries, local government institutions are virtually nonexistent. Often, the narrow self- interest of the central civil service and of politicians does not allow local governments to develop. Where strong, viable local governments exist, questions concerning the alloca- The Civil Service and Economic Development: The Japanese Experience 33 tion of functions, authority, and resources arise. For countries that are political federa- tions, the question of decentralization is of special importance. The colloquium might more accurately be entitled as "The Civil Service as a Catalyst for Economic Development," because that was the spirit of many of the presentations. Given that many of its discussants are responsible for organizing civil services or dealing with problems of administration, it becomes easy to assume that somehow the civil ser- vice can solve all the developmental problems of a country. The Japanese experience, however, makes clear that underlying economic success are many variables including importantly: what the private sector is doing, how society is organized, and what the character of a country's political leadership is. While it can play an important role in long-term economic development, the civil ser- vice is only one actor in a complicated process. As a result, governments and societies wishing to advance economically have to be sensitive to how that civil service interacts with society's economic, social, and political segments. In other words, to emulate what Japan has done is to do more than copy programs supporting the private sector or repro- duce the characteristics of the civil service. It also means developing an environment that fosters private innovation, dynamism, and entrepreneurship. Japan's economic success since the late nineteenth century, and particularly since the early 1950s suggests that it has been doing something right. Most would agree that its civil service played a positive role in enhancing Japan's economic well-being. The Japanese civil service has certainly defined itself to be proactive in matters of economic growth. The ministries of Finance and of International Trade and Industry have been particularly strong in their roles as growth enhancers, although many feel that the government's role has been overstated. Whatever the case, Japan's has generally been a competent civil service, with a histor- ical appeal to the country's best and brightest. It contains built-in elements of competition and creativity, both in the national personnel system and in agency-to-agency relation- ships, so the temptation toward lethargy and routine that bureaucracies confront are sig- nificantly reduced. The scrap and build principle, among others, has installed a dynamism into the Japanese system that prevents it from becoming fossilized. Japan's system also offers lessons in keeping government small. It has made a long- standing series of efforts to check the size of the civil service. Indeed, the Japanese case provides a positive example of how to restrain the automatic tendency toward bureau- cratic expansion. As such, the Total Staff Number Law and the personnel reduction plans, which provide the framework for controlling the size of the national government, are worthy of close study in countries where a bloated service is a burden on the budget and on economic development. If skilled administrators are the scarcest commodity in a coun- try, economizing on those skills by keeping the civil service small but highly capable is wise. Collective decisionmaking, a practice firmly ingrained in Japan, not only contributes to the quality of decisions made, but also greatly facilitates their implementation. In addi- tion, Japanese civil servants have demonstrated their ability to work as a group. In many countries, people in general, and bureaucrats in particular, do not seem to work well in a group. If this type of group synergy can be emulated-if it is not cultural, but rather some- thing that can be learned-then many countries should seriously look into doing so. Another lesson seems to be that direct intervention in the private sector only works when it is of a catalytic nature, a phenomenon that has been called "participatory inter- 34 The Civil Service and Economic Development: The Japanese Experience vention." The government cannot substitute for the private sector, nor can it ignore the response of the private sector and of markets when it intervenes. Thus, in Japan, policy- mandated financial aid has, at the recipient level, been disbursed on the basis of objective economic criteria. Adding to the need to enlist the cooperation of the private sector is the fact that although the Japanese civil service has a broad mandate, it has only limited pow- ers of enforcement. Good policies and interventions compatible with the market worked in tandem in Japan. The private sector needs space to flourish. It is not an accident that Japan was success- ful and at the same time kept taxes and government deficits low. The partnership with the private sector, the sense of flexibility that is observed in Japanese policymaking, and the willingness to correct mistakes all represent a different approach to the private sector than that prevalent in many other countries. In addition, fiscal irresponsibility by local governments needs to be kept in check. In the case of Japan, local governments must obtain the approval of the Ministry of Home Affairs to issue bonds, which in turn must consult with the Ministry of Finance before a green light can be given. In many countries the home affairs ministry or the interior minis- try is really a bastion of police functions and the like, and rarely has any economic func- tions. Every civil service system has developed in a particular culture and history, and thus has its particular advantages and disadvantages. When one country tries to introduce another country's system, it is sometimes necessary to harmonize it to indigenous ways of thinking. Japan learned a great deal about creating a civil service system from Europe in the late nineteenth century and from the United Stated after World War II, but it adopted the principle of "pick and choose." Thus, Japan did not adopt the U.S. job system because it was not suited to Japanese work traditions, which are based on long-term employment and flexible job assignments. The fact that even emulation needs to be adaptive is a partic- ularly important lesson derived from Japan's experience. Even Roses Have Thorns Caveats must be given: not everything the Japanese civil service has done has been a suc- cess. Japan's rapid economic growth makes it all too easy to assume that everything some- how played a positive, and hence "copyable," role in that growth. If one looks at the Japanese civil service before World War II, it was influenced by both political parties and the military. The major architect of the modern Japanese civil service in the late nineteenth century, Ito Hirobumi, tried to make it a meritocracy, in part by mak- ing civil servants agents of the emperor so as to isolate them from political influence. However, competitive examinations were not given, and applications were simply sub- mitted. It is therefore important to look at the Japanese civil service system as an evolu- tionary continuum. If one looks only at its successful and desirable features, one is missing the point. The evolutionary patterns of the present system are the most important source of lessons. Serious challenges confront the Japanese civil service today. While the colloquium was being held, the Japanese media were heavily criticizing the bureaucracy for disobeying political orders and for being oblivious to public needs. They urged deregulation to rein- vigorate the economy, and decentralization to reinvigorate local communities. Indeed, this last point was borne out by the colloquium's visit to Kamakura City, where the efforts of The Civil Service and Economic Development: The Japanese Experience 35 the central government at delegating power to local government are thought to be too lit- tle, too late. Japanese officials enjoy anonymity, and much of what they do and how they do it is not readily discernible to outsiders. To counter the ill-effects of this, increased transpar- ency in administration has been called for. A Sense of Mission There is a strong tendency in Japan to build societal or national consensus. In the early development stage in Japan, catching up with advanced nations was a widely shared goal. An almost "up against the world" syndrome created a tremendous incentive for bureaucrats to be efficient. This contributed to a partnership between the private and pub- lic sectors in Japan that is especially striking, and that contrasts with the animosity toward the private sector that seems to be prevalent in a number of developing (and even indus- trial) countries. This sense of mission and partnership is an important reason for the effec- tiveness of the Japanese bureaucracy during the high-growth era. Now that Japanese per capita income has surpassed that of most other industrial nations, at least as measured by the current exchange rate, an agreed upon national goal no longer exists. Japanese society has divided and criticism of the bureaucracy has mounted. Just as the mobilization and commitment of the Japanese civil service may have been assisted by the sense of mission created by the "catch up" phenomenon, the current lack of such incentives may be promoting bureaucratic complacency. This is not merely an aca- demic question, as disquieting signs are apparent. Let us consider two of these, which may be isolated incidents, but could just as well be portents of things to come. The first is the arrest on March 11, 1994, of a member of the Diet's Lower House and former minister of construction, Ichiro Nakamura, for bribery and influence peddling at the Fair Trade Commission, which is the watchdog for antimonopoly violations. Nakamura is accused of trying to prevent the Fair Trade Commission from filing charges against a group of con- tractors allegedly involved in bid rigging. What is most disquieting is that the Fair Trade Commission seems to have given in, or at least appears to have been influenced. This raises the question of the commission's integrity and its ability to maintain independence from politics. The second is the arrest of former political king-maker Shin Kanemaru for tax evasion. These incidents seem to be harbingers of the weakening of the iron triangle at the political corner, and the extent to which that effects the civil service is a major issue. The Concept of Organizational Learning The importance of organizational learning as a key factor of competitive advantage has been widely recognized in the private sector. There is no reason why a similar degree of attention should not be paid to this phenomenon in the public sector, and a number of characteristics of Japan's civil service can be shown to be related to organizational learn- ing. Aspects of this-such as the acquisition of knowledge, the distribution and interpre- tation of information, and organizational memory-can provide clues to the adaptive efficiency of the Japanese civil service. The process by which any knowledge-based organization acquires, maintains, and enriches its knowledge base is important. Several aspects of the structure and functioning of the Japanese civil service seem to facilitate this process. The first is the ability of Japan's 36 The Civil Service and Economic Development: The Japanese Experience civil service to attract and retain highly educated individuals who are extremely receptive to the learning process. Second, strong competitive pressures exist throughout the career of the civil servant and among agencies, and this contributes to learning, as do the numer- ous opportunities for on-the-job training and interministerial exchange. Finally, the sys- tematic participation of the private sector in the formulation and implementation of economic policies forces civil servants to keep in touch with the external environment and to incorporate market realities into their decisionmaking. A number of factors facilitate information distribution within the civil service. The small size of the central bureaucracy itself facilitates the transfer of information. Numer- ous formal and informal coordinating mechanisms exist, both at the intra-agency and interagency levels. The mandate of individual agencies is typically very broad in Japan, but enforcement powers are rather weak. This, combined with a culture of consensus, cre- ates the conditions for using persuasion as an instrument of administrative guidance. Per- suasion requires dialogue and information sharing, thereby distributing knowledge across the civil service. The active dialogue between the central and local governments through the Ministry of Home Affairs also furthers information sharing. Information sharing in itself is inadequate to sustain organizational learning if differ- ent parts of the organization interpret information in a different way. Japanese civil ser- vants and Japanese society have, at least until recently, shown remarkable convergence in values and beliefs. Cultural homogeneity and the high value placed on consensus cer- tainlv constitute factors facilitating common interpretation of information. Also undoubt- edly important has been the pervasive pro-growth and pro-business ideology, the roots of which can be traced to the nineteenth century. For the bureaucracy, the stabilizing political influence of the Liberal Democratic Party majority rule was important, as has been the tendency among Japanese civil servants to consider themselves (and actually act as) guardians of the public interest. As far as institutional memory is concerned, at least two factors facilitate the retention of knowledge in the collective memory of the civil service. One is undoubtedly, the life- long career of the civil servant within one agency. Even the practice of amakudari is an instrument through which the knowledge and experience of senior civil servants are pre- served and transplanted into parastatal and occasionally private organizations. Note, however, that the same practices can produce different results in different orga- nizational and institutional settings. The very notion of lifelong careers and the practice of promotion from within can produce stagnation and an inward-looking attitude. The con- sultative process has a heavy cost in terms of the time required to arrive at decisions, although this time may be recouped by speedier implementation. An excessive focus on conflict avoidance and consensus building may result in what is known as "group think" or other aberrations of collective decisionmaking. Finally, there is always the risk, particu- larly in very homogeneous cultures, that past successes will be extrapolated well beyond the circumstances that made them possible, thereby creating complacency and other con- ditions incompatible with adaptive efficiency. The lesson to be drawn is not simply that one should try to replicate what has worked well in another country, although Japan itself has shown how effective creative imitation and adaptation of the institutions of other countries can be. Rather, the challenge faced in civil service reform, especially in developing countries, is that institutions operate as a system, and that cultural factors and historical traditions interact with formal rules and organizational structures in determining the final efficiency of a civil service system. In a The Civil Service and Economic Development: The Japanese Experience 37 sense, recognizing the complexity of the institutional matrix is a key requirement for real- istic planning of civil service reform, just as it is a requirement for good planning of eco- nomic policies. References Kim, Hyung-Ki, Michio Muramatsu, T. J. Pempel, and Kozo Yamamura. Forthcoming. Catalyst for Development: The Role of the Civil Service in Japan's Economic Development. New York: Oxford Uni- versity Press. Kim, Paul S. 1988. Japan's Civil Service System: Its Structure, Personnel, and Politics. New York: Green- wood Press. Koh, B. C. 1989. Japan's Administrative Elites. Berkeley, California: University of California Press. Lincoln, Edward J. 1988. Japan: Facing Economic Maturity. Washington, D.C.: The Brookings Institu- tion. Nunberg, Barbara. 1992. "Managing the Civil Service: What LDCs Can Learn from Developed Coun- try Reforms." Policy Research Working Paper. Washington, D.C.: World Bank. World Bank. 1993. The East Asian Miracle: Economic Growth and Public Policy. New York: Oxford Uni- versity Press. 3 A Civil Service in Changing Times: The Japanese Bureaucracy and Its Reform Toshiyuki Masujima Japan's bureaucracy is one of the forces that has sustained the development of the coun- try's economy and society. Ironically, however, Japan's mass media have been busy criti- cizing the bureaucracy virtually daily since the Liberal Democratic Party went out of power in August 1993. The principal points of the criticisms are that the Japanese bureau- cracy has too much power, disobeys political orders, favors particular interest groups and industries, and lacks an awareness of its public service responsibility. In 1959 I started my career in the Japanese government by entering the Administrative Management Agency (AMA) (which was reorganized into the Management and Coordi- nation Agency (MCA) in July 1984) in the Prime Minister's Office. Some years later, I par- ticipated in a training program as a tutor. This was the training program for new recruits to the civil service, all of whom had passed the Senior Entrance Examination (now the Level I Examination) for candidates for future top executive jobs. There I heard Yosanji Kobayashi make the keynote address to the program participants. Kobayashi was a former administrative vice minister in the Ministry of Home Affairs, and later became president of Nihon Television and chairman of the board and chief executive officer of Yomiuri Shimbun, the newspaper with the largest circulation in the world. In the address, Kobayashi talked about his experiences when the Ministry of the Inte- rior, for which he worked, was broken up in the aftermath of Japan's defeat in World War II. At this point he had to decide where in the government he should work. He said he had thought anywhere would do. He was interested in the Ministry of Construction, which is the public works department, and he had his eye on the Imperial Household Agency, because he was concerned with possible changes in the monarchy and the direction they might take. Finally, he concluded that after all, he was concerned not so much with decid- ing where to go as with choosing a mission of importance to which he could devote him- self. He added that if a department is without a mission that would attract promising young people or without the capability to accomplish such a mission properly it should be abolished. I was deeply impressed by the way he had been able to develop such uncon- strained thought at one of the most critical moments of his life. Affected by Kobayashi's remarks, I resolved to devote myself to carrying out the special mission of the Administra- tive Management Agency. I have been fortunate enough to be able to spend my whole professional career per- forming the kinds of duties that probably are not carried out in any other government 39 40 A Civil Service in Changing Times: The Japanese Bureaucracy and Its Reform department. I have been able to participate in developing most of the important laws and systems on which Japan's administrative management is based. They include the Total Staff Number Law and the personnel reduction plans that provide the framework for con- trolling the size of the government; the amendment of the National Government Organi- zation Law, which greatly expanded the government's ability to initiate reorganizations of its organizational structure in response to economic and social change; the Personal Infor- mation Protection Law, which protects individuals' privacy; and the Administrative Pro- cedure Law, which helps to protect people's rights and interests by ensuring fairness and transparency in the administrative process. Looking back, I cannot but think of myself a lucky fellow to have been able to find such an important mission, and from the bottom of my heart I believe that becoming a civil servant was one of the best things that happened in my life. I am not alone in this. Many others in the government feel the same way. Working in the Management and Coordination Agency in the center of the government provides an invaluable opportunity to become familiar with the people and jobs in other ministries and agencies. Taking advantage of that, I have made myself a self-appointed watcher of Japan's bureaucracy. From this vantage I can say that not just the mere existence, but the domination of the bureaucracy by officials committed and wholly devoted to their duties is one of the most salient features of Japan's civil service. The mass media's criticism of the Japanese bureaucracy, which has noticeably intensi- fied in recent months, is related to the events surrounding the Liberal Democratic Party's fall from power for the first time in thirty-eight years; the subsequent formation of a coali- tion government under Prime Minister Morihiro Hosokawa; and the new government's initiative to push through administrative reforms, including deregulation. Some of the criticisms are completely off the mark; others are justified. I will not dwell on them. Suffice it to say that such criticism has provided an opportunity for the many officials in the Japa- nese government to reflect on the bureaucracy and on themselves. I will discuss the char- acteristics, advantages, and disadvantages of Japan's bureaucracy and the direction of current reforms. The Nature of Public Service in Japan Masaharu Gotohda, the first director general (minister) of the Management and Coordina- tion Agency, is a distinguished and highly able politician. He served in the capacity of managing director of the cabinet under Prime Minister Yasuhiro Nakasone, whom many consider to have been one of the strongest prime ministers of postwar Japan. I worked for Gotohda as director of general affairs of the minister's secretariat in the agency. He had an unusual way of thinking: Always go back to the starting point. He frequently asked me such interesting questions as "What is politics?" or "What is administrative reform?" One day Gotohda summoned me to his office and fired off another question at me: "What is the nature of public service?" It was a tough question, but I managed to respond by drawing a table (see table 3-1) that described the characteristics of public service as compared with private sector activities. A Civil Service in Changing Times: The Japanese Bureaucracy and Its Reform 41 Table 3-1. A Comparison of Public Service and Private Sector Activities Private sector activities Public service (firms) Service to the whole (serving public Service to a part of society (serving private interests) interests) Possession of power exercised from a Nonauthoritative body (contract principle) position of authority Source of funds: taxes Source of funds: profits Existence of limits on the scope of Flexibility in the scope of activity responsibility Concept of bankruptcy does not apply Bankruptcies can occur Multiple and complex objectives of action: Market mechanism provides the means of difficulty in measuring results measuring results Procedures are stressed Results are stressed Activities directly contribute to bettering Activities indirectly contribute to bettering people's lives people's lives I had no idea why Gotohda had raised this question. However, it has a major bearing on the current discussion. All those concerned with public service, whether working in the bureaucracy or criticizing it, must have a clear understanding of the difference between public service and private activities. In other words, they need to comprehend the nature of public service. Without this, discussions about administrative reforms will not be constructive, nor will the government's activities be productive. Public service has its own unique character. At the same time, its problems are closely related to this unique nature. They are like the heads and tails of a coin. Administrative reforms aim at maintaining the basic premises underlying public service while eliminat- ing its deficiencies. Let us consider an example. As table 3-1 indicates, government offices possess public power, which they exercise from positions of superiority. Such power must, of course, be exercised properly, and people must have avenues available to inquire into any abuses of such power. This notion leads to the principle of clear division of function and responsi- bility among many offices of government. In Japan the government comprises one office and twelve ministries under the cabi- net. The office is the Prime Minister's Office, which has eight agencies that are the equiv- alent of ministries. The agencies' scope of functions is delineated by the laws that established them. The first-tier organizations within the ministries and agencies are the ministers' secretariats and the bureaus, which currently number 128 in all, which hap- pens to be the number written into law as a governmentwide ceiling for such organiza- tions (see figure 3-1 for trends in the number of secretariats and bureaus). Figure 3-1. Changes in the Number of Secretariats and Bureaus, 1956-91 '. - 131 130 .13 11711 1128 920 1956 1960 1965 1970 1975 1980 1985 1991 Notes: The figures presented pertain to the end of each fiscal year. The increase of fifteen bureaus in 1957 comprises four bureaus established in the Home Affairs Ministry, three in the Administrative Management Agency, and four in the Economic Planning Agency, and is attributable to the revision of the National Government Organization Law. Source: Data from the Management and Coordination Agency. A Civil Service in Changing Times: The Japanese Bureaucracy and Its Reform 43 A cabinet order sets down the scope of functions of a ministry's secretariat and bureaus. A secretariat or a bureau may sometimes have a deputy director or be subdi- vided into departments, but the basic units into which secretariats and bureaus are split by function are divisions. The total number of such divisions is currently almost 1,500. Their names and functions are also delineated by cabinet orders. If new functions are needed to cope with the demand for services, a cabinet order finalizes their allocation. Fig- ure 3-2 presents the typical internal structure of ministries. Figure 3-2. Typical Internal Structure of a Ministry Ministry Minister (usually member of Parliament) Parliamentary vice minister (member of Parliament) Administrative vice minister (career civil servant) Agency Director-general (career civil Research institutes Training institutes Councils Bureau Secretariat Bureau Department Divisions Divisions Divisions Divisions Divisions Localbranch offices Note: Most of the Agencies in the Prime Minister's Office have equal status as ministries. They are headed by Ministers of State and have bureaus as internal structure. Source: Masujima and O'uchi (1995(a), p. 20). The orderly configuration of divisions is the product of careful planning to ensure the proper exercise of power by government offices, but it sometimes means more than that. 44 A Civil Service in Changing Times: The Japanese Bureaucracy and Its Reform A former minister for the Administrative Management Agency, Masuo Araki, during his tenure as education minister, reportedly advised his aide when he was to be promoted to division director: "You have to keep in mind that becoming division director means that in your jurisdiction, you are the only person invested with authority, being responsible as the most authoritative expert concerning every affair under your charge." Indeed, this view is strongly held by division directors of the ministries and agencies of the Japanese government. Junichi Ky6goku (1983), professor emeritus of the University of Tokyo, writes in his book entitled Nihon no Seiji (Japanese Politics): "The divisions in the headquar- ters of the ministries and agencies are virtually the Government of Japan itself." (Ibid p. 347). There is some truth to this statement judging from the way government offices oper- ate and the mentality of their officials. I myself still remember the joy I felt when I was promoted to the division directorship after sixteen years in government service. Undoubtedly, the system provides the officials who work within it with a source of pride and a sense of mission. However, there is the other side of the coin: The system is likely to have inherent defects. A situation whereby the government's tasks are ultimately divided among 1,500 divi- sions without overlaps and gaps represents a completely vertically structured administra- tion, which inevitably results in some negative effects. A typical example of such an effect is bureaucratic turf wars, which can be both positive and negative. Things will also become difficult when a person with an inflexible way of thinking and lacking in ability to make decisions from a broader perspective becomes a division director. The appointment may impede the division's ability to respond to newly emerging problems, and thus ren- der the public administration incapable of achieving its expected role. Measures to eliminate such negative effects, both organizational and operational, have to be devised. One such measure is the mechanism for overall coordination. Japan's gov- ernment agencies are equipped with a sophisticated system for overall coordination at the level of divisions, bureaus, and the ministry or agency as a whole. In addition, the units for overall coordination are usually manned by "the best and the brightest." A personnel arrangement is in place to cope with the problems of a vertically struc- tured administration. The practice of regular personnel reshuffles, usually every two years, under which an official gains work experience in different divisions of a ministry or agency, is part of the solution. The practice of interministry personnel exchanges works along the same lines. Currently, more than 80 percent of those appointed to division direc- torships have experience working in another ministry or agency, including public corpo- rations and local public entities. Nevertheless, I believe that further efforts are needed to get rid of the negative effects of a vertically structured administration. This area is also the one most frequently pointed out as a target for bureaucratic reform. Let us consider another example. I noted earlier that in the case of business activities, their final objective is simple profit seeking, and their performance can easily be measured through the market mechanism. In the case of public service, however, activities have multiple and complex objectives, so that coming up with a yardstick for measuring per- formance can often be extremely difficult. While in some policy areas measuring the eco- nomic effects of a program may suffice, in others you have to decide to what extent a particular program contributes to the maintenance of public safety, provides the public with a sense of security, or increases other countries' trust of Japan. This is no simple mat- ter. The ability to adapt to the changing socioeconomic environment is another, albeit A Civil Service in Changing Times: The Japanese Bureaucracy and Its Reform 45 somewhat abstract, objective of any government program. The extent to which it has been attained can only be known by simultaneously taking account of many different values. In addition, the market mechanism enforces the survival of the fittest: Businesses that perform poorly risk bankruptcy, but the concept of bankruptcy is alien to the public ser- vice. Organizations or functions that would have vanished in the private sector tend to remain in the public sector. Japan's unique answer to this difficult issue of the lack in the government of a built-in adjustment mechanism like the market mechanism is the govern- mentwide manpower control mechanism and the principle of "scrap and build" applied to organizational structure. The manpower control mechanism starts by enforcing a person- nel reduction plan under which a ministry or an agency is assigned a target figure for per- sonnel cuts. These are usually made through attrition during the planning period, a period of five years or thereabouts. The number of staff dropped by the ministries and agencies is then recorded in the staff number account of the Management and Coordina- tion Agency and pooled for all entities. Meanwhile, to cope with increased work loads, ministries and agencies can request an increase in staff numbers. After rigorous examina- tion by the agency, such requests may be granted-by redistributing the pooled staff num- bers. Then what remains in the agency will become the total number of staff to be dropped from the government each year. This mechanism has been in effect since 1968 (see figure 3-3 for government manpower trends). The basic philosophy behind the mechanism is as follows: * Ministries and agencies will inevitably have declining areas of business correspond- ing to changes in the economy and in society. * There is always room for rationalization in the operation of ministries and agencies by improving efficiency, simplifying or mechanizing activities, contracting out, and the like. * The identification of declining areas of business and methods of rationalization and resulting reductions in staff number are best done by a special ministry or agency. * The manpower control mechanism will not result in any dismissals of personnel. All reductions in staff numbers will be achieved through attrition. * Because the mechanism is an artificial one, unlike the market mechanism, resistance to it is inevitable. The strong and firm resolve of the cabinet is always needed in put- ting the mechanism to work. In Japan we seldom hear criticisms of redundant staffing in the national government. More likely we hear calls for increased manpower in specific areas. An international com- parison of manpower in the government indicates that Japan has the lowest numbers of government employees per thousands of population and a much smaller number of employees in the public than in the private sector (see figures 3-4 and 3-5). The enforcement of the scrap and build principle relating to organizational structure is another artificial mechanism, this time to prevent the proliferation of government orga- nizations. By law, the Management and Coordination Agency is authorized to examine any request for creating organizations within the executive branch. In this connection, when a ministry or agency wishes to create a new bureau, department, or division, it has to abolish an organization at the equivalent level. Thus the requesting ministry or agency is obliged to reorganize its organizational set up. Reorganizations in the ministries and agencies were eased in 1984 following an amendment of the National Government Orga- nization Law, whereby the creation of bureaus or departments, previously allowed only 46 A Civil Service in Changing Times: The Japanese Bureaucracy and Its Reform Figure 3-3. Number of National Government Employees in Ministries and Agencies, 1957-93 Thousands of people Provisional 1,000 -Commission on Planned reduction Administrative start: the Total Staff A38 100 . Reform Number Law enacted 1 - 899,333 898,26586,3 800 Return of Okinawa A37,032 600 67 125 400- 200 1957 1967 1972 1981 1993 Fiscal year Note: The Self-Defense Forces are excluded. Source: Data of the Management and Coordination Agency. through changes in the law, became possible through issue of a cabinet order. This means that reorganization at the bureau level can be carried out solely at the initiative of the gov- ernment. In Japan, the creation of a new organization is considered to be the addition of a new decisionmaking unit within the government, resulting in an increased workload. The rig- orous examination of new organizations under the scrap and build principle is thus a means of curbing the proliferation of the sphere of government activities. Incidentally, in Japan, people view organizations as embodying authority and prestige. A slight change in an organization under their jurisdiction sometimes becomes an issue of major concern to top executives, including the minister of a ministry or an agency. To sum up, the two examples cited show that to preserve the positive characteristics of public service you have to get rid of problems that arise as a result of its nature by devis- ing effective measures. A Civil Service in Changing Times: The Japanese Bureaucracy and Its Refonn 47 Figure 3-4. International Comparison of Number of Government Workers, per 1,000 of Population. Selected Countries, 1990 Person 120 ns 120 - 110 - 112 persons 100 - 90 - -80 persons 81 rsons 8ppersoeo 60 - 50 - 40 ersons 40 - W 38 persons 30 - 20 - 10 Japan U.K. France US.A. Defense Personnel Administrative Personnel Notes: Government workers consist of national government workers, local government workers, and employees of government corporations. Source: Data of the Management and Coordination Agency. 48 A Civil Service in Changing Times: The Japanese Bureaucracy and [is Reform Figure 3-5. Number of Employed People by Public Sector and Private Sector, 1967-94 Ten thousand people 1,400- 1,367 1,300- Manufacturing /4 .1,272 1,200-.. / / 1 ,121 1,100 - / 1,000 1,000 .4* Wholesale and retail trade, bars, 900 - and restaurants 800- 700 - ,* 600- 40000-0 Local governments - . ................... ---- 325 Financing, insurance .. ** - 244 and real estate ,... 20 ... National government - - - - - -- - - 116 100 .. - - * - - - -85.9 Executive branch of the national government 0 1 I _ 1967 1970 1975 1980 1985 1990 1994 Source: National government and private sector: Management and Coordination Agency; local governments: Ministry of Home Affairs. The Salient Features of the Japanese Bureaucracy Having touched upon the nature of Japan's public service in general, I will now move on to its specific characteristics and the question of what has made Japanese bureaucracy what it is today. The following, some of which we have already discussed, stand out in this respect: * The doors are kept open for anyone with proven ability to become a civil servant. The fairness and impartiality of the examination and recruitment system appear to be unparalleled. Each year, a large number of bright and promising young people aspire to a career in the public service. A Civil Service in Changing Times: The Japanese Bureaucrac and Its Reform 49 * Within the service, measures are provided to help equip candidates for executive positions with an abundance of expertise and experience. They include job rotation through a broad range of different sections of an organization, interministry person- nel exchanges, and systematic training programs under the auspices of the Training Institute of the National Personnel Authority. * Members of the civil service are endowed with a high degree of integrity. Japanese officials are known to be incorruptible, although occasional charges of bribery do surface. Officials seem to be content with honest poverty. * Merit ratings of personnel are strict. An individual's ability and character are contin- uously evaluated for a surprisingly long period. * Responsibilities both within the government as a whole and within individual orga- nizations are clearly delineated. * Overall coordination mechanisms are firmly in place. * Within the government as a whole, as well as within individual organizations, a high value is placed on decisionmaking by consensus in the planning stages of pro- grams, which in turn ensures coherence, speediness, and uniformity in program implementation. The annual budget compilation process, which is by custom the process of coordinating and finalizing the important policy decisions of the govern- ment, is typical of this emphasis on consensus. * Artificial adjustment mechanisms to replace the market mechanism have been devised and introduced to prevent the proliferation of public administration. Such mechanisms force continual re-examination of policies and programs and the means to implement them. Examples are the manpower control mechanism, the system of controlling organizational structures under the scrap and build principle, and the ceiling of budget requests employed in the budget compilation procedure. * Entities exist at the highest level of the government that specialize in overall coordi- nation, some armed with the philosophy of control through management as distinct from control through budget. The Management and Coordination Agency and the Cabinet Legislation Bureau are examples of such bodies that coordinate policies and programs through their authority over organizational structure, manpower, admin- istrative processes, and legislative review. * An established procedure exists for administrative reform. In the 1980s, there was a dramatic development in administrative reform (usually called the RINCHO admin- istrative reform) in Japan, against the backdrop of the government's tight fiscal situ- ation. One of the most notable reforms was the privatization of certain giant public corporations, including Japanese National Railways and Nippon Telegraph and Telephone Public Corporation. Existing policies and programs also were subject to scrutiny and many of them were revised by amending relevant laws. Since 1981 the government has passed 172 administrative reform laws (as of March, 1994). As the government submits around 100 bills to the Diet each year, this is a large number of laws. The success of the RINCHO administrative reform owes much to the adminis- trative reform procedure introduced under Prime Minister Yasuhiro Nakasone in 1981 (for a detailed discussion of the procedure of administrative reform see Masujima 1995(a), pp. 188-213). 50 A Civil Service in Changing Times: The Japanese Bureaucracy and Its Reform Reforms for Maintaining a Good Bureaucracy For Japan to maintain a bureaucracy staffed by civil servants whose morale is high while getting rid of the problems that spring up is not easy. More efforts are necessary in the face of such challenges, and also to respond to the legitimate criticisms mentioned earlier. I would like to touch upon the following six topics in relation to what is immediately needed in this respect. First, the government should aggressively pursue deregulation and decentralization. Deregulation is necessary to reinvigorate the economy and to limit the government's role to the essentials. Decentralization is needed to change the centralized system of adminis- tration to a system in which citizens' voices can be heard more readily and to build a sys- tem that can help promote the reinvigoration of local communities. Both of these are among the top priorities agreed on by the cabinet of Prime Minister Morihiro Hosokawa. Second, both government officials and the public must become knowledgeable about the Administrative Procedure Law, which was enacted in November 1993 and took effect in October 1994. This law, one of whose objectives is to eliminate the ill effects of adminis- trative guidance, must take root in the society and administration as quickly as possible. Third, and related to the second point, increased transparency in public administration must be further promoted. In my view, the criticism of the lack of transparency in the way Japanese government offices operate is legitimate. It is also true that Japanese officials tend to be timid in increasing this transparency. However, people's access to public infor- mation should be drastically expanded so that the public can understand Japan's public administration more easily and can direct more articulate criticism to the public adminis- tration. Transparency will not affect the stability of Japanese bureaucracy in any way. On the contrary, it will lead the public to have a better appreciation of the real strengths of Japan's public administration. Fourth, overall coordination capabilities at the cabinet level should be reinforced. While overall coordination is performed fairly adequately, particularly within individual ministries and agencies, some improvement is needed, especially at the cabinet level. Some have argued that coordination would be improved if an arrangement that facili- tated better exercise of political leadership were put in place. This argument is justified. It is time for the Japanese government to declare to its people and to the international com- munity a series of clear-cut policies in the face of volatile world politics, explosive growth of world population, and Japan's rapidly aging population. It is the role of politics to define the country's objectives and to formulate the policies to carry them out. The role of a country's administration, by contrast, is to develop and present a wide range of alterna- tives to the political leadership. Reform or innovation always involves the destruction of a certain equilibrium or of the status quo. The policy innovations urgently required at this time are no exception. It is important in this context to provide the cabinet with the orga- nization and staffing arrangements like strengthening the cabinet secretariat or having some high-ranking cabinet advisers that can sustain the exercise of political leadership. Fifth, personnel exchanges should take place on a far larger scale. A frequent sugges- tion for eliminating the negative effects of a vertically structured administration is that staff recruitment should be centralized rather than continuing the system of individual ministries and agencies recruiting their own staff. In my view, however, the suggestion is unworkable, and even harmful to the preservation of the strengths of the Japanese bureaucracy for the following reasons: A Civil Service in Changing Times: The Japanese Bureaucracy and Its Reform 51 * Those who aspire to a public service career usually have a strong preference for the kind of job they desire, and therefore choose to work for a particular ministry or agency. This is one of the reasons for the high morale of Japanese officials. * The selection, education, and other decisions concerning the treatment of employees are delicate decisions that are made based on performance within a particular minis- try or agency. These decisions cannot be left to somebody else outside that entity. Thus the most practical solution would be to promote a large-scale, regular, and sys- tematic exchange of personnel between ministries and agencies (including embassies and consulates) and between the national government and local governments or international organizations. If a system can be designed with the result that most senior officials have the experience of working in Japan's foreign establishment or in international organiza- tions as well as in local government, this would be extremely useful. Not only would it help Japan to contribute to the international community more effectively and to promote domestic decentralization, it would also help eliminate the negative effects of a vertically structured administration. Finally, another task that has so far been ignored in the context of administrative reform, but which I believe belongs in the reform agenda, is to promote the advanced use of information systems, or in other words, to make the public administration more sophis- ticated in its access to and use of information. Traditionally administrative reform has been concerned with the economy, efficiency, deregulation, and adaptation to changes in the international environment, but has seldom sought to improve the services it provides to the public. The thinking has been that improving such services is the purview of indi- vidual ministries and agencies. However, a gap has developed between the use of com- puters and communication devices in the government and in the private sector. For instance, the private sector has one personal computer for every four persons, while the government has one computer for every seven persons, according to the survey taken by the Management and Coordination Agency in 1994. The proper use of technology in the government would help to upgrade services provided by both the national and local gov- ernments. Conclusion I have noted that in Japan, continuing to correct the negative effects of the bureaucracy without losing sight of its better features is extremely important. However, a good bureaucracy cannot be realized without furnishing its environment. I would like to raise the following two points in this connection. First, many in Japan realize that political reform is highly desirable. Critics occasion- ally pointed out that the link between the Liberal Democratic Party, the bureaucracy, and big business was so close and strong that it amounted to a built-in structure of collusion. The formulation of a coalition government under Prime Minister Hosokawa and the demise of the Liberal Democratic Party appear to have gradually been affecting this rela- tionship. The suspicion of collusion, however, will not disappear without political reforms aimed at eradicating so-called money politics or interest politics. I hope that a political system under which an occasional change of regime is possible will be brought about by political reform. Such a system is the external environment in which the Japanese bureau- 52 A Civil Service in Changing Times: The Japanese Bureaucracy and Its Reform cracy may be able to identify its role and perform its functions while maintaining its better features. Second, communication should be established between the world of academia and the world of public administration. Every time I was involved in planning and developing a fundamental system of public administration I noticed that many academics have little knowledge of, or are indifferent to, the practice of public administration. Without such knowledge, academics cannot legitimately criticize the public administration or persua- sively propose bureaucratic reforms. One of the reasons for such a lack of communication may be that public officials in Japan think too highly of the virtue of anonymity in regard to their duties to willingly explain their actions or rebut criticisms. They should now discard this mode of thinking. Communications between the two groups should be aggressively pursued so that they may share their knowledge and experience. The fact that this international colloquium was held in Tokyo is significant in this respect, as in addition to its original objectives, it provides an opportunity for the sharing of information between academics and practitio- ners in Japan. References Ky6goku, Junichi. 1983. Nihon no Seiji (Japanese Politics). Tokyo: University of Tokyo Press. Masujima, Toshiyuki. 1995(a). "The RINCHO Administrative Reform." In Toshiyuki Masujima and Minoro O'uchi, eds., The Management and Reform of Japanese Government (second edition). Tokyo: The Institute of Administrative Management. Masujima Toshiyuki. 1995(b). "Organization Management, Staff Number Control, and Coordination Mechanism." In Masujima and Minoro O'uchi, eds., 1995. The Management and Coordination of jap- anese Government (second edition). Tokyo: The Institute of Administrative Management. 4 The Role of the Bureaucracy in Japan's Economic Development Yutaka Kosai I will begin by quoting some conflicting views on Japan's bureaucracy. First, Japan is known as a country where the government is strong. Many people believe that Japan's economic development was a product, at least in part, of government guidance and sup- port. However, compared to the size of governments in other industrial countries, Japan's government is small. How does one reconcile these two apparently conflicting views? Second, a World Bank report (1993, pp. 167-70, 174-80) contains some passages on the importance of technocratic insulation in achieving growth with equity, citing the Japanese case as an example. By contrast, the U.S. President Bill Clinton harshly attacked the Minis- try of Finance and the Ministry of International Trade and Industry (MITI) for preventing the opening of Japanese domestic markets, and claimed that this worked against the inter- ests of the Japanese people. This occurred immediately after the failure of the framework talks between Japan and the United States in February 1994. Japan's so-called iron trian- gles, which consist of the bureaucracy, political party bosses, and interest groups and stand in the way of liberalization and social progress, are now internationally known and criticized. These two views are in direct contradiction. Is the Japanese bureaucracy insulated from or intertwined with vested interests? Are vested interests promoting or obstructing the nation's economic progress? I will try in this paper to present my interpretation of the role of the bureaucracy in Japan's economic development by resolving the conflicts inherent in the opposing views noted earlier. The intent is not to present an authoritative discourse on the subject, but to give you enough information so that you can reach your own conclusions. Before starting this discussion, let me give you some background about myself. I entered the service of the Japanese government immediately after I graduated from uni- versity in 1958, and worked in the Economic Planning Agency for about twenty years. I left the government fifteen years ago to become a professor of economics at the Tokyo Institute of Technology, and later took my current position as director of a nongovernmen- tal, nonprofit research institute on economics. Thus I have changed my job three times; a rare career path in a country of lifetime employment. I have given you this account of my life to explain the sources of my interest and ability to judge the subject: the role of the bureaucracy in Japan's economic development. Con- trary to the previous speaker (Mr. Masujima), I am an ordinary citizen, and able to criti- 53 54 The Role of the Bureaucracy in Japan's Economic Development cize the public administration freely. As an economist, I have a duty to assess the role of the bureaucracy objectively, and having personally served as a government official, I have some firsthand knowledge of the workings of the bureaucracy. In short, my background provides me with a comparative advantage in discussing the topics presented in this paper. Note, however, that these are my own personal views, and you are, of course, free to disagree. Small but Strong Government The government's size can be measured using several different scales. One such scale is the ratio of general government expenditures to gross national product (GNP). In 1991 this ratio was 32.3 percent in Japan, 36.0 percent in the United States, and 45.9 percent in Germany. The difference was more pronounced in 1970, when the ratio stood at 19.4 per- cent in Japan, 32.6 percent in the United States, and 39.0 percent in Germany (table 4-1). Another measure is the national burden ratio [(taxes + cost of social security)/national income]. In 1990 this was 39.6 percent in Japan, and 51.3 percent in Germany and in the United States in 1989 was 36.5 percent (table 4-2). While the ratio was lower in the United States than in Japan, the former recorded larger deficits in government budgets (table 4-3). Finally, one can look at the number of government employees per 1,000 population. In 1991 this figure was 40 in Japan, 81 in the United States, and 120 in France. By any of these measures, the Japanese government is smaller than those of other industrial nations. Table 4-1. Fiscal Expenditure as a Percentage of GNP, Selected Countries, 1970 and 1989 General Capital Country government Consumption formation Social security Miscellaneous Germany 1970 39.0 15.8 4.4 12.7 6.1 1989 45.9 18.3 2.3 15.2 10.1 Japan 1970 19.4 7.3 4.4 4.5 3.3 1989 32.3 9.1 5.2 11.0 7.1 United States 1970 32.6 19.1 2.5 7.7 3.4 1989 36.0 17.8 1.6 10.7 5.9 Source: EPA (1992); OECD National Accounts. Table 4-2. The National Burden as a Percentage of National Income, Selected Countries, 1989 or 1990 United States Germany Japan Component of national burden (1989) (1990) (1990) Taxes 26.1 29.4 28.1 Social security contribution 10.4 21.9 11.5 Total 36.5 51.3 39.6 Source: EPA (1992); OECD National Accounts. The Role of the Bureaucracy in Japan's Economic Development 55 Table 4-3. General Government Deficits as a Percentage of GNP, Selected Countries, 1989 or 1990 United States Germany Japan Component ofgovernment deficit (1989) (1990) (1990) Central government -3.8 -2.1 -0.3 Local governments -0.4 -0.7 0.3 Social security fund 1.1 0.9 3.5 General government -3.1 -1.9 3.5 Source: EPA (1992); OECD National Accounts. Table 4-4. Number of Government Employees per 1,000 Population Country Including armed forces Excluding armed forces, 1991 France 120 112 Japan 40 38 United States 81 69 Source: MCA (1992). Why and how does the Japanese government remain small? The answer is that basi- cally, this is what both the public and government prefer. People understand that the only way for Japan to flourish is to keep its market economy free, vigorous, and competitive, and that too heavy a burden of government activities may suffocate the workings of the market economy. Smaller government has been necessitated and made possible by the existence of Japan's vigorous market economy. This simple fact cannot be overempha- sized. The belief in small government stems from lessons Japan learned during World War II. Militarist expansion was naturally accompanied by a big government, heavy taxes, bureaucratic control, inflation, devastation of the economy, and pauperization of the masses. More fundamentally, the caution against big government originates from Japan's circumstances as "a fragile blossom" (Brezinski 1972) because of its weak resource base and dependency on foreign supplies of foodstuffs, energy, and basic raw materials. Japan has to export its manufactured products to be able to pay for its imports, and must there- fore maintain its competitiveness in world markets by avoiding placing too heavy a bur- den on its industries. Given this basic premise, several institutional arrangements developed that con- strained government expansion. For example, the government's budget must be balanced and government bonds to finance deficits are disallowed by law. In other words, special legislation is required to issue a deficit bond. In 1960 the Government Tax Council advised that the ratio of tax revenues to national income should not exceed 20 percent to retain the vitality of the private sector. In the 1980s the Administrative Reform Council also set the long-term upper limit of the national burden ratio so that Japan would not end up as a European type welfare state in the face of its aging population in the coming decades. The 56 The Role of the Bureaucracy in Japan's Economic Development total number of central government employees is determined by law and decreased every year as far as possible.' As the economy grew vigorously, the government could remain small. Also to date Japan's demographics have allowed the authorities to keep social security payments rather small. The size of Japan's government when the country's population ages remains to be seen. How can a small government also be a strong one as is widely believed to be the case with respect to Japan? Perhaps the measures of government size noted earlier are defi- cient, and hide some important elements of government activities? Indeed, some deficien- cies are easy to identify. Fiscal Investment and Loan Program The Japanese government can rely on post office savings funds and the public pension fund to finance both government and private projects. The Fiscal Investment and Loan Program (FILP) amounts to 50 to 60 percent of the central government's general account. The FILP requires that projects are profitable so that they can pay interest and repay the principal. In that respect the FILP cannot be a perfect substitute for government expendi- ture. Nevertheless, the government exerts its influence through the FILP. Among FILP- related government financial institutions are the Japan Development Bank, the Japan Export and Import Bank, The Overseas Economic Cooperation Fund (OECF), and the Small and Medium-size Financial Corporation (SMFC). Imposed Obligations Employers are obliged to withhold taxes on their employees' wages and salaries, which saves the government the cost of tax collection. Other similar examples can be found. Regulation by Law The cost of agricultural protection in Japan was $95.4 billion in 1993, which was higher than $88 billion of the United States though lower than $137 billion of the European Union (table 4-5). The unusual feature of Japanese protection is its low dependence on subsidies (or on sacrifices by taxpayers) and heavy dependence on regulation (or sacrifices by con- sumers). While protection by means of subsidies is lower in Japan than in the United States and European Union, protection through regulation is at almost the same high level as in the EU. Thus regulation in some cases substitutes for fiscal expenditures. According to a study by the Economic Planning Agency, 40.8 percent of Japanese value added indus- tries are government regulated as concerns entry or prices of goods and services (EPA1994). 1. A series of measures were taken to reduce the number of central government employees by 46,531 between 1967 and 1993, roughly a 5 percent reduction. The Role of the Bureaucracy in Japan's Economic Development 57 Table 4-5. Agricultural Protection Selected Countries (billions of US dollars, 1993) Country Total Taxpayer Consumer European Union 137.0 59.0 77.1 Japan 95.4 20.8 74.6 United States 88.0 68.6 19.4 Source: OECD (1994). Administrative Guidance Japan's administrative guidance is well known and notorious among foreign observers. Administrative guidance takes the form of notes, written or unwritten, issued by senior government officials, that direct the activities of the private sector. My understanding is that administrative guidance is a means of moral suasion. The government's actions in this respect are not written into the law, but the government enters into administrative guidance in a spirit of public good. For administrative guidance to be effective, adminis- trative offices must have some moral authority, which derives from the honorable behav- ior of civil service staff. My impression is that the effectiveness of administrative guidance is sometimes exaggerated. For example, in a famous case of the 1960s, Sumitomo Metal finally gave in to administrative guidance by MITI officials, but MITI had to be more pru- dent in exercising administrative guidance after this event. Administrative guidance is effective when those guided consent and support it. This is another example of a government measure to exercise its power without relying on fiscal resources. However, the whole administrative guidance process is now facing changes. Outsiders as well as some within Japan are urging deregulation, together with the opening of the domestic market. The resulting enactment of the Administrative Proce- dure Law will make hidden forms of administrative guidance impossible. The most important deficiency in measuring the size of the government still remains. All the activities cited in the preceding paragraphs are concerned with government inputs. The strength of a government, however, depends not only on its inputs, but also on its outputs, or more accurately, on the relationship between inputs and outputs, that is, on its efficiency. In turn, the efficiency of government activities is related to the govern- ment's mode of operation and to the working of the market mechanism and the behavior of the private sector. Thus we will turn to the second set of conflicting views on the Japa- nese bureaucracy mentioned earlier. The Bureaucracy's Promotive Role The second issue is whether or not the Japanese bureaucracy is insulated from vested interests and whether or not the role of the Japanese bureaucracy is conducive for sound economic development. The existence of conflicting views on this issue suggests that the bureaucracy's role changes over time and from sector to sector. The bureaucracy plays a promotive role as well as a distributive one. Japan has long passed the stage of catching-up with other 58 The Role of the Bureaucracy in Japan's Economic Development industrial countries and moved into the stage of economic maturity. Thus we will focus on the bureaucracy's promotive activities during the catching up era and its distributive activities later. The basic conditions for the success of Japan's economy and economic management, such as the existence of a vigorous market economy and the importance of maintaining competitiveness, applied more forcibly during the catching up period. At this time, most people agreed with the national goal of the pursuit of rapid growth. Bureaucrats benefit- ted greatly from this consensus. Furthermore, a study of countries that had industrialized sooner made the intermediate goals of an industrial structure obvious, and the authorities assumed that the Japanese industrial structure would eventually be more or less similar to that in the industrial countries. For example, they expected that Japan could have a strong automobile industry in the future when the economy had grown and domestic markets had expanded, and once industries' technology had reached an advanced level. Industrial targeting, or picking winners, was an easy task when forecasts could be based on the pre- vious example of other countries. The government's ability to promote industrialization was initially constrained by the basic condition of Japan's economy during the catching up period. Because of its depen- dence on the world market, Japan had to maintain the market mechanism on the domestic front. The postwar dissolution of the zaibatsu and an improved opportunity for technical progress through imports of foreign technology resulted in fierce competition in domestic markets. Fortunately, the bureaucracy could take advantage of this market mechanism. However, the ceiling of the balance of international payments was rather low because of the competitiveness of industries in the process of catching up with those in the industrial nations and because of the need to spend any surpluses for accelerating the rate of eco- nomic growth still further. This constraint imposed a severe macroeconomic discipline on Japan, and the tight budget situation prevented the use of subsidies for promotive pur- poses. The promotive tools used during 1955-70 consisted mainly of (a) special tax treat- ments; (b) public investment in infrastructure; (c) government loans through the FILP; (d) tariffs, import quotas, and permission to import technological licenses; (e) the authoriza- tion of recession and rationalization cartels;2 (f) a special law for promoting specific indus- tries that authorized the administration to use the tools mentioned above; and (g) administrative guidance. The relative importance of these different tools changed over time. Trade liberalization in the 1960s and 1970s rendered tariffs ineffective, for example. Let us consider a few examples of promotive measures. A typical case of a special law was the enactment of the Machine Industry Promotion Law, which aimed at promoting industries that supported the automobile and other machine industries, such as parts sup- pliers, die-cast factories, businesses manufacturing gauges, and so on. These were small industries, and the government assisted them through FILP loans, through standardiza- tion, through presenting rationalization targets and encouraging the formation of ratio- 2. Cartels are illegal under the anti-monopolistic law in Japan. However, there are a few excep- tions to the law: firms in severely depressed industries are allowed to form "recession cartels" in order to restrict output or sellers' prices, and firms may form "rationalization cartels" to reach a con- sensus with common rationalization measures such as standardization of products and parts. Both these cases require special permission from the Fair Trade Commission. Other types of cartels (for example, those aimed at scrapping excess capacities) are also allowed under special laws. The Role of the Bureaucracy in Japan's Economic Development 59 nalization cartels, through the reduction of tariffs of imported rationalization equipment, and through accelerated depreciation of newly acquired plants. The current strength of the Japanese automobile and other machine industries lies in the existence of rationalized parts suppliers and the cooperative relationship between parts suppliers and assemblers. The Machine Industry Promotion Law laid the foundations for the competitiveness of Japan's machine industries later on. Another promotive measure was the protection of infant industries from foreign com- petition. Note, however, that this protection was only allowed on a temporary basis to strengthen the competitiveness of a particular industry. Protection of the domestic coal industry was abolished in the mid-1950s, which led to the loss of some 300,000 jobs when the cost of coal could not be reduced sufficiently and eventually lost out to imports of for- eign oil. From 1960 on, the authorities announced a trade liberalization schedule and encouraged industries to rationalize themselves so that they would survive after protec- tion was abolished. Japan achieved industrialization not in violation of, but fundamentally in accordance with, the law of comparative advantages. In the mid-1950s Japan was a country lacking in capital and land, but abundant in labor. The promotion of heavy industry appeared to be in contradiction to the relative abundance in its factor endowments, but looking into the situation a little more deeply will solve this apparent contradiction. The key is the quality of the factor input. Capital in the sense of accumulated stock was scarce, but the savings rate was high. This meant that the "vintage" of the capital was young, and the incorpora- tion of new technology was swift. Even though it lacked natural resources, Japan could rely on imports of energy and raw materials, taking advantage of the rapid reduction of marine transportation costs and its geographical location, which allowed it to build plants along the coast line. In other words, even though land was scarce, it was conveniently located. In addition, the abundant labor was educated and disciplined, and could be employed in such skilled industries as automobile manufacturing and iron and steel pro- duction rather than in cotton spinning, for example. Given the national consensus that Japan's goal was to catch up with the industrial nations, the vigorous competition in domestic markets, the constraints imposed by the final competitive test on the world market, and the ceilings imposed by balance of pay- ment and balance of budget disciplines, the administration did not have a free hand with the promotive measures during the catching up period. This lack of freedom, however, implied that opportunities for abuse of power were severely limited. This let the bureau- cracy focus on forecasting the market situation and fostering growing industries. These tasks were congruent with the bureaucracy's neutrality, objectivity, and technocratic role. The success of promotive measures during the catching up period can be attributed to this situation. If promotive measures follow market trends in the long run, the question may arise as to why the function cannot be performed through the capital market. The capital market's role is to allocate resources for the future, which is similar to the government's promotive measures. The question is legitimate, and little room exists for industrial policy if the cap- ital market is perfect and efficient. In reality, however, the capital market often suffers from inefficiencies, such as information asymmetry, and is not perfect. During the stage of catching up, because of the still low level of income, saving often took the form of riskless capital. This may provide a partial justification of government intervention in industries during the catching up stage. 60 The Role of the Bureaucracy in Japan's Economic Development In determining the role the bureaucracy played in engaging in promotive measures during the catching up stage, the honesty, skill, and reputation of the civil service mat- tered a great deal. At the minimum, its creditworthiness had to be equal to or greater than that of capital market institutions. How did the bureaucracy establish its excellent reputa- tion? I hope others will address this question. In passing I would like to comment on the "market-friendly policy" concept invented by the World Bank. My feeling is that this concept has missed some point if it applies to the Japanese case. The concept seems to be a product of compromise between neoclassical economics and the reality of government intervention in the "East Asian miracle." The concept contrasts the experience of high-performing Asian economies with the market- suffocating government intervention in planned economies. However, the Japanese authorities appear to have been more aggressive in intervening heavily in businesses' behavior. Government action can work as a catalyst, eliciting positive reactions from the private sector. I propose the concept of participatory interaction to explain the govern- ment's most successful interventions. We must remember that not every intervention will succeed. Note also that intervention can be likened to a drug. Used in the right amount, it can be effective, overused, it can be a poison. The market-friendly policy is a lukewarm expression not appropriate for reality, which is full of opportunities and dangers. The Bureaucracy's Distributive Role Sound economic development requires not only growth, but also equity. The bureaucracy is involved in ensuring this equity. In the case of promotive measures, the bureaucracy could serve as a neutral specialist that knows how to promote economic growth and industrial competitiveness, and during the catching up period in particular, promotive measures were left mostly in the hands of the bureaucracy. With respect to distributive measures, however, the situation differs. A single measure of performance, such as efficiency or competitiveness as is the case with promotive measures, does not exist. The concept of fair distribution is both value- and interest-loaded, and reaching agreement on it is difficult. The solution must be political. Thus there is room for actors other than bureaucrats to both participate and to meddle. The division of labor should be such that promotive measures are assigned to bureaucrats while distributive measures are decided through consultation among politicians, bureau- crats, and interest groups. The so-called iron triangles of bureaucrats, political bosses, and interest groups typically apply with respect to distributive measures. The government's distributive policy comprises taxation, social security, and other social legislation, but some industrial policy, in particular agricultural policy and policy toward small and medium-size industries, may have distributive purposes rather than promotive ones. I will be concerned mainly with this category of distributive measures. Distributive measures were important during the catching up stage. The Basic Law of Agriculture, enacted in 1960, declared that making farmers' incomes' equitable with those of other actors in the economy was a principal aim of agricultural policy. Until 1970, how- ever, Japan's level of agricultural protection remained relatively low compared with that in other industrial countries. After 1970, when Japanese industries caught up with those in the industrial countries and the country no longer suffered from balance of payment con- straints, the officially set price of rice began to rise sharply, with a result being rapidly The Role of the Bureaucracy in Japan's Economic Development 61 growing agricultural protection. This example demonstrates the increased importance of distributive measures during the stage of economic maturity. The situation was similar for small- and medium-size industries. The Large-Scale Retail Store Law, which allows the establishment of large-scale retail stores only with the agreement of small retailers in the neighborhood, was first proposed by politicians. MITI was initially opposed to this law, and instead proposed strengthening administrative guidance to solve disputes that arose concerning the establishment of new large-scale retail stores. When this idea was rejected, MITI gave in to some extent, by agreeing to a compromise that softened the need for agreement by small retailers to the notification of such retailers. However, it became customary that large-scale retail stores were neverthe- less not established until local retailers had agreed, and it was at this point they received the official notification. Protective measures initiated for distributive purposes differ from those implemented for promotive purposes. The latter are transitory, while the former are permanent, and are not expected to be withdrawn in the foreseeable future. The temporary protection can accelerate the process of rationalization, but protection for the long-term can cause entre- preneurship to stagnate and result in sluggish productivity gains, a loss of competitive- ness, and a need for increasing protection. Around 1980, the government started to move toward deregulation. Because of the worsening budgetary situation, the government had to reduce the fiscal costs of protec- tion. The Administrative Reform Council advocated deregulation as a way to achieve smaller government. From the international viewpoint, the opening of Japanese markets was requested by trade partners, particularly the United States. The Large-Scale Retail Store Law was to be amended as a result of the structural impediments talks, while agri- cultural policy had to be reviewed because of Japan's agreement at the Uruguay Round negotiations of the General Agreement on Tariffs and Trade to import a minimum quan- tity of rice. Furthermore, a wide disparity between domestic prices and international prices converted to yen at current exchange rates apparently heightened public awareness of the need to stimulate competition in the low productivity sectors by abolishing regula- tion. The fault for the distortion in resource allocation does not lie entirely with the distribu- tive measures. Some certainly were well intended and could have been justified under certain circumstances. In many cases, however, they degenerated into a distortion of resource allocation and ended up obstructing the adjustment needed in a particular sector. To avoid the spread of inadequate regulation in the name of distributive justice, the fol- lowing ideas may be worth considering: * The authorities should initiate redistribution not by sector-specific measures, but by general, open methods, such as income taxes and social security. Sector-specific dis- tributive measures are more likely to result in allocative inefficiency. It is also easy to hide them from the public, which is conducive to the formation of interest groups. * The organization of the government should be restructured. Sectionalism in bureau- cracy apparently tends to strengthen sectoral protective measures. Moreover, the ministries, departments, and divisions of the government are often so concerned with their territorial interests and power struggles that the good of the general pub- lic is sometimes overlooked. Yet, sectionalism is in a sense essential in a bureaucratic hierarchy. The solution may be either to promote competition among the branches of 62 The Role of the Bureaucracy in Japan's Economic Development government in such a way that the supervising entity can know what is happening and intervene appropriately, or reorganize the government in such a way that the same bureau looks after a variety of conflicting interests. In a democracy such as Japan, the final outcome depends on the will and wisdom of the people and their political leaders. Conclusion Japan remains as a country with a smaller than average government compared to other industrial countries because of its vigorous market economy and strong fiscal discipline. However, a set of tools has developed through which the government can exert influence on the private sector in a somewhat hidden way, such as through regulation. Promotive measures worked during Japan's catching up stage given the particular objectives and constrained circumstances. The situation at that time presented few risks of power being abused and allowed the bureaucracy to concentrate on the technical aspects of catching up. Distributive measures, however, particularly once Japan reached economic maturity, often led to excessive protection, a decline in entrepreneurship, and the forma- tion of interest groups because of the lack of established criteria for equity. The movement toward deregulation since the 1980s has given rise to some hopes for improvement. References Brezinski, Zbigniew. 1972. The Fragile Blossom: Crisis and Change in Japan. Harper and Row. New York. EPA (Economic Planning Agency). 1992. System of National Accounts. _. 1994. Nenji Keizai Hokoku (Annual Report on Japan's Economy). MCA (Management and Coordination Agency). 1992. Annual Report. Government Printing Office. Tokyo. OECD (Organization for Economic Cooperation and Development). 1994. Agricultural Policies, Mar- kets, and Trade. Paris. World Bank. 1993. The East Asian Miracle. Oxford University Press. New York. Wolferen, Karel G. 1986/87. "The Japan Problem." Foreign Affairs (Winter). Distributors of CZECH REPUBLIC R.O.Y. International Oxford University Press Mundi-Prensa Barcelona World Bank National Information Center PO Box 13056 5 Bangalore Town Consell do Cent, 391 Publications prodejna, Konviktska 5 Tel Aviv 61130 Shares Faisal 06009 Barcelona Prices and credit terms CS - 113 57 Prague 1 PO Box 13033 vary from country to DENMARK Palestinian Authority/Middle Karachi-75350 SRI LANKA country. Consult your SamfundsLitteratur East THE MALDIVES local distributor before Rosenoems All6 11 Index Information Services PERU Lake House Bookshop placing an order. DK-1970 Frederiksberg C PO.B. 19502 Jerusalem Edhorial Desarrollo SA PO. Box 244 Apartado 3824 100, Sir Chittampalam A. ALBANIA EGYPT, ARAB REPUBLIC ITALY Lima 1 Gardiner Mawatha Adrion Ltd. OF Licosa Commissionaria Colombo 2 Perlat Rexhepi Str. Al Ahram, Al Galaa Street, Sansoni SPA PHILIPPINES Pall. 9, Shk. 1, Ap. 4 Cairo Via Duca Di Calabria, 1/1 Intemational Booksource SWEDEN Tirana Casella Postale 552 Center Inc. Fritzes Customer Service The Middle East Observer 50125 Firenze Suite 720, Cityland 10 Regeringsgaton 12 ARGENTINA 41, Sherif Street, Cairo Condominium Tower 2 S-106 47 Oficina del Libro Internacional JAMAICA H.V dela Costa, comer Stockholm Av. Cordoba 1877 FINLAND Ian Randle Publishers Ltd. Valero St., Makati, Metro 1120 Buenos Aires Akateeminen Kirjakauppa 206 Old Hope Road, Manila Wennergren-Williams AB PO. Box 23 Kingston 6 P 0. Box 1305 AUSTRALIA, FIJI, PAPUA FIN-00371 Helsinki POLAND S-171 25 NEW GUINEA, SOLOMON JAPAN Intemational Publishing Solna ISLANDS, VANUATU, AND FRANCE Eastern Book Service Service WESTERN SAMOA World Bank Publications Hongo 3-Chome, Ut. Piekna 31/37 SWITZERLAND D.A. Information Services 66, avenue d'I6na Bunkyo-ku 113, Tokyo 00-577 Warzawa Librairie Payot 648 Whitehorse Road 75116 Paris Service Institutionnel Mitcham 3132 KENYA PORTUGAL C6tes-de-Montbenon 30 Victoria GERMANY Africa Book Service (E.A.) Livraria Portugal 1002 Lausanne UNO-Verlag Ltd. Rua Do Carmo 70-74 AUSTRIA Poppelsdorfer Allee 55 Quaran House, Mfangano 1200 Lisbon Van Diermen Editions Gerold and Co. 53115 Bonn Street Techniques Graben 31 PO. Box 45245, Nairobi ROMANIA Ch. de Lacuez 41 A-1011 Wien GREECE Compni Do Librarii CH1807 Blonay A-OlPapasotiriou S.A. KOREA, REPUBLIC OF Bucuresti S.A. BANGLADESH 35, Stournara Str., 106 82 Daejon Trading Co. Ltd. Str Lpscani no. 26, sector 3 TANZANIA Micro Industries Development Athens PO. Box 34, Yeoeida Bucharest Oxford University Press Assistance Society (MIDAS) Seoul Maktaba Street House 5, Road 16 HONG KONG, MACAO RUSSIAN FEDERATION PO Box 5299 Dhanmondi R/Area Asia 2000 Ltd. MALAYSIA ysdatelstvo Dar es Salaam Dhakaw 1209 Sales & Circulation University of Malaya 9a, Kolpachniy Pereulok Department CooperativeBookshop, Moscow 101831 THAILAND BELGIUM Seabird House, unit 1101-02 Limited, P0O. Box 1127CeteBosDtrbin uon o Lnoy16 22-n8 Hodham Street, pCeaupraNDIRBA AA ntra Books Distribution Jea D Lano 2228Wynhan Sret, Jalan Pantai Baru SAUDI ARABIA, OATAR 306 Silom Road Av. du Roi 202, 1 D60 Central. Hong Kong 59700 Kuala Lumpur Jadir Book Store Bangkok Brussels HUNGARY MEXICO P.O. Box 3196, Riyadh 11471 Foundation for Market INFOTEC TRINIDAD & TOBAGO, BRAZIL E SINGAPORE, TAIWAN, JAMAICA Publcacites Tecnicas Donomy 71 Apartado Postal 22-860 MYANMAR, BRUNEI Systematils Studies Unit Intemacionsis do, Dombovari ut 7-19 14060 Tlalpan, Mexico D.F Asahgate Publishing Asia #9 Watts Street Rue Peixoto Gomido, 209 H-1117 Budapest Pacific PIe. Ltd. Curepe 01409 Sao Paulo, SP INDIA NETHERLANDS 41 Kallang Pudding Road Trinidad CANADA Allied Publishers Ltd. De LndeboonInOr- #04-03 West Indies CANDAPublikaies Golden Wheel Building Renouf Publishing Co. Ltd. 751 Mount R 4 0 ksegeoe Whee uGdnDA 1294 Algoma Rced Madras - 600 002 74O0 Box 202brgnGstoLd Ontaa KB 3W8 INDONESIA SLOVAK REPUBLIC Madhvani Building Pt. Indira Limited EBSCO NZ Ltd Sloven G.TG. Ltd. P Box 9997 CHINA Jalan Borobudur 20 EBSCO NZ Bag Krupinska 4 Plot 16/4 Jinja Rd. China Financial & Economic P.O. Box 181 New Mail 80 BxB152 Kampala PulsigHueJakarta 10320 Newkarke 852 99 Bratislava 5 8, Da Fo SI Dong Jie UNITED KINGDOM Bejn DFoSIRAN SOUTH AFRICA, Microinto Ltd. Kowkab Publishers NIGERIA BOTSWANA PO. Box 3 COLOMBIA PO. Box 19575-511, Tehran University Press Limited For single litles: Alton, Three Crowns Building Oxford University Press Hampshire GU34 2PG Infoonlace Ltda.Jeih Apartado Aereo 34270 Ketab Sara Co. Publishers Pricho Southern Africa England BogotA D.E. Khaled Eslamboli Ave.,6th St. Private Mail Bag 50 PO. Box 1141 Kusheh Delafrooz No. 8 lbadan Cape Town 8000 ZAMBIA COTE D'IVOIRE Tehran NORWAY University Bookshop Centre d'Edition at de Narvesen Inormation Center For subscdpllon orders: Great East Road Campus Difluslon Africaines (CEDA) IRELAND International Subscription PO. Box 32379 04 B.P. 541 Government Supplies Agency Book Department Svc. Lusaka Abidjan 04 Plateau Oifig an tSolAthaIr PO. Box 6125 Etterstad PO. Box 41095 4-5 Harcourt Road N-0602 Oslo 8 Cralghall ZIMBABWE CYPRUS Dublin 2 PAKISTAN Johannesburg 2024 Longman Zimbabwe Center of Applied Research (Pte.) Ltd. Cyprus Coll e ISRAEL Mirza Book Agency SPAIN Toude Road, Ardbennie 6, DIogenestrest, Engomi Yozmot Literature Ltd. 65, Shahrah--Quald-e-Azam Mundl-Prensa Libros, S.A. P.O. Box ST125 P.O. Box 2006 P.O. Box 56055 P.O. Box No. 729 Castello 37 Southerton Nicosia Tel Aviv 61580 Lahore 54000 28001 Madrid The World Bank Headquarters rm Learning Resouirces are desig ned for 1818 H Street, N.W. use in vi courses and seminars. TheY Washington, D.C. 20433, U.S.A. discuss issues in economic development policy and lessons from experience in a Telephone: (202) 477-1234 way that can be understood b7y persons Facsimile: (202) 477-6391 without extensive background knowledge Telex: MCI h4145 WORLDBANK or technical expertise. Thcy 'will be of MCI 248423 WORLDBANK particular inter'st to readers concerned Cable Address: INTBAFRAD with 1public policy. WASHINGTONDC World Wide Web: http:/ /www.worldbank.org E-mail: books@worldbank.org tri1 European Office 66, avenue d'I6na 75116 Paris, France Telephone: (1) 40. 69. 30. 00 Facsimile: (1) 40. 69. 30. 66 Telex: 640651 Tokyo Office Kokusai Building 1-1, Marunouchi 3-chome Chiyoda-ku, Tokyo 100, Japan Telephone: (3) 3214-5001 Facsimile: (3) 3214-3657 Telex: 26838 III I II III 4001128 E9224 ISBN 0-8213-3170-1