Policy, Research, and Extemal Affairs WORKING PAPERS Public Sector Management l and Private Sector Development Country Economics Department The World Bank March 1990 WPS 217 Public Sector Management Issues in Structural Adjustment Lending Barbara Nunberg Public sector management components of structural adjustment loans (SALs) progressed unevenly, and the outcomes varied with different political, administrative and economic condi- tions. Change was often incremental and som.etimes unsustain- able. Reforms linked to specific, actionable steps were more successfully implemented. The Policy, Research, and Extemal Affairs Complex distributes PRE Working Papers to disseminate the findings of work in progress and to encourage the exchange of ideas among Bank staff and all others interested in development issues. These papers carry the names of the authors, eflect only their views, and should be used and cited accordingly. The findings, interpreations, and conclusions are the authors! own. They should not be attributed to the World Bank, its Board of Directors, its management, or any of its member countries. Policy, Research, and External Affairs Publlc Sector Mlanagemnenl L and Private Sector Development This paper - a product of the Public Sector Management and Private Sector Development Division, Country Economics Department - is part of a larger effort mn PRE to analyze and provide policy guidance about the institutional aspects of macroeconomic policy relorm programs. Copies are available free from the World Bank, 1818 H Street NW, Washington DC 20433. Please contact Betsy Mitchell, room N9-065, extension 37644 (36 pages with tables). This paper reviews the Bank's experience in constraints must be weighed against the catalytic implementing public sector management re- benefits SALs offer to sr:ir action on institu- forms through structural adjustment lending. tional measures, however. The study focuses on those institutional aspects of adjustment that deal specifically with "macro- * Public management reforms through SALs management" issues related to improvements in are more successful when supported by specific the management performance of core central technical assistance projects. Otherwise, re- government institutions and to systemic changes forms may be one-shot events or may be subject in public administrations. The paper reaches the to slippage or reversal. This is recommended following broad conclusions: despite the poor performance record o, lnuch of the technical assistance provided for institutional P Public sector management components of reforms in conjunction with SALs. SALs progressed unevenly and outcomes varied with diverse political, administrative and eco- * The haste of SAL schedules and the lack of nomic conditions. Accordingly, no uniform dynamism and focus of traditional technical solutions to institutional problems were pre- assistance argues for the creation of a new scribed by SALs. Depending on circumstances, lending instrument for public sector manage- SALs sometimes fostered the creation of new in- ment reform which would adopt a longer time stitutions (often at the cost of promoting institu- horizon to ensure adequate implementation, but tional proliferation) or the reorganization or would contain conditionality specifically fo- even the elimination of old ones. Institutional cused on institutional reforms to ensure action. change was often incremental and sometimes unsustainable when Bank support was with- * P-eparatory country economic and sector drawn. work is crucial to successful public sector man- agement reforms undertaken through SALs. * Public sector management reforms for Suc}, analysis should become a routine task in which routinized methodologies and systems the design of any institutional reform program in were introduced and those that could be linked the context of structural adjustment lending. to actionable steps were more likely to be sustained over time. * Monitoring and supervision of institutional components of SALs needs to be systematized * Short time horizons of SALs posed severe and the quality of documentation improved. constraints on the effective implementation of Performance indicators need to be developed to public sector management reforms. These measure and/or evaluate progress. The PRE Working Paper Series disseminatcs the findings of work under way in the Bank's Policy, Research, and External | Affairs Complex. An objective of the series is to get these findings out quickly, even if presentations are less than fully polished. The findings, interpretations, and conclusions in these papers do not necessarily represent official Bank policy. L Produced at the PRE Dissemination Center iublic Sector Mana_emeat Issues In Structural Adiustment Lending TABLE OF CONTENTS SUMMARY . . . . . . .. . . . . . . . . i-iv I. INTRODUCTION . . . . . . . . . . . . . . . 1 Analytic Framework Methods II. OVERVIEW OF INSTITUTIONAL COMPONENTS IN SALS AND TALS . . . . . . . 3 A. ECONOMIC AND FINANCIAL MANAGEMENT . . . . . . . . . . . . . . 3 i. Public Sector Investment Program Management . . . . . . 4 ii Economic Policy Bodies . . . . . . . . . . . . . . . . . 6 iii. Debt Management Systems .............. 7 iv. Tax Administration.. . . . . . . 8 B. NON-FINANCIAL SECTOR REFORMS ...... .... ... 9 i. Agriculture Reforms ..... ... .... .. 9 ii. Industry Reforms . . . . . 10 . . . . . . 10 iii. Trade Administration Reforms . . . . . . . . . . . . . 11 C. PUBLIC ADMINISTRATION REFORMS ..... . . . . . . . . . . . 12 D. SUMMARY OF FINDINGS ........ .. .. ... .. .. . . 16 III. SELECTED OPERATIONAL ISSUES . . . . . . . . . . . . . . . . . . . . 17 A. TIMING ........................ .. . 17 B. TECHNICAL ASSISTANCE ....... .. . .. .. .. . .. . 18 C. INSTITUTIONALIZATION OF REFORMS . . . . . . . . . . . . . . . 19 D. COVERNMENT COMMITMENT TO REFORMS . . . . . . . . . . . . . . 20 E. DESIGN ISSUES. . . ......... 20 IV. CONCLUSIONS. ......... 22 V. ANNEXES SUIKARY The purpose of this paper is to provide an overview of public sector management reforms in structural adjustment lending and to assess the Bank's experience in implementing them. The review focuses on those institutional aspects of adjustment that deal specifically with macro-management* issues related to improvements in the management performance of core central government institutions and to systemic changes in public adpUnistration. The paper covers public sector management reforms in 59 Structural Adjustment Loans and 43 companion Technical Assistance Loans between 1979, when adjustment lending began, and June 1988. Although public sector management reforms were a part of Bank lending prior to structural adjustment, most earlier institutional interventions were through sector-specific, project-related lending. The implementation requirements of the adjustment process highlighted the need to enhance the capacity of core government economic policy making institutions to formulate and execute macroeconomic recovery programs. Public sector management reforms in SALs, therefore, have been focused and targeted on a few institutions that are key to the adjustment process. These iastitutional reforms have been attempted in relatively short time horizons, although they have sometimes been sequenced over serial SALs, and with the support of companion technical assistance loans. The following are the principal areas of public sector management reformt 1) institutional changes required for m4cropconomic and financial management (present in all 59 SALs); 2) non-financial sectoral institutional reform (found in 56); 3) trade policy and administration reforms (present in 50); and, 4) public administration reforms (present in 31 SALs). In the area of economic. and financial management, most SALs emphasized improvements in the public investment programming process. Often this involved strengthening the coordination of investment decision making among government entities, improvements in the institutional and procedural mechanisms for planning and investment, and strengthening of investment monitoring. Efforts to improve economic and financial management also involved the creation or reinforcement of economic policy bodies. These were often linked to immediate demands of SALs, but occasionally.became institutionalized as a technocratic cadre able to engage in policy dialogue with Bank SAL teams. Institutional support 'for debt management reform was also provided in SALs. Most reforms established or reinforced institutions to monitor and control debt and tried to develop systems to feed pertinent information on debt into budgeting and public investment programs. Tax administration reform was another aspect of econowic and financial management stressed by SALs. In general, tax administration improvements involved the streamlining of customs administration procedures, revision of tax codes, and the improvement of tax collection systems. Institutional measures in.most SALs went beyond the core economic and financial institutions to support sectoral restructuring and policy reforms as well. Sectoral institutional reforms have emphasized: 1) institutional support fo:; policy changes involving the liberalization of price and incentive structures.; 2) improving capacity for strategic planning in sectoral organs; 1' - ii- and, 3) reorganization of sectoral institutions. Sectoral lnstitutional components have been heavily concentrated in agriculture and industry. Energy was the third most prominent sector followed by health and social security, transportation, population, water and urban development, and housing. Institutional support for trade policy reforms also figured prominently in structural adjustment lending. Many reforms sought to streamline - institutions or procedural arrangements for trade administration and to revise regulatory codes to promote liberalization. Although non-financial sectoral reforms constituted an important aspect of many SAL programs, Bank staff were increasingly inclined to deal with these issues through Sectoral Adjustment Lending. SALs were the preferred instrument only if: the sectoral issues were important enough to the overall economy to justify a centralized SAL operation; administrative efficiencies could be gained by a centralized SAL effort and sectoral reforms could be linked to more central structural reforms without fear of the whole package being sabotaged due to lack of political commitment on one or another single item; or, structural adjustment could be used as a means to provide reinforcing cross-conditionality on sectoral reform issues. Increasingly, public administration reforms have become prominent institutional components of structural adjustment lending. These components generally comprise civil service management and/or reorganization reforms, mostly emphasizing structural changes in central government employment, pay and incentive systems. These reforms sought to correct excessive public wage bills, surplus numbers of civil servants, wage compression and overall salary erosion. WhiWe SALs often leveraged one shot policy reforms such as employment reduction. TALs provided instittutional support for longer term civil service management improvements. The following trends could be detected from the paper's findings: o In a few instances, improvements in public investment programing institutions and procedures may have led to or set the stage for budgetary reform. o New institutional mechanisms for dealing with debt management have been established in some countries; improvements in coordination between ministries and central banks and other financial authorities may have resulted. O A good number of economic policy institutions, responsible for developing coordinating and reviewing macroeconomic management in general, and adjustment operations in particular, have been established, with varying impact. O New agenqies have arisen to deal with trade regime reforms. There has been simplification and perhaps rationalization of trade administration procedures in some cases. O Organizational and managerial improvements in key sectoral - iii - institutions may have occurred, and the reduction of institutional proliferation in some crucial economic sectors may have led to improved performance. o A reduction in the number of civil servants and a redefinition of the role of the civil service may have occurred on a limited basis in some countries. Despite some success, movement on institutional reforms has been halting for the following reasons: 1) weaker than expected management capacity unable to absorb excessive reform demands; 2) ripid turnover of management and political leaders; 3) poorly executed technical assistance and inadequate training of local staff in skills necessary to implement and sustain reforms; 4) failure to install systems and routine procedures that can be institutionalized and internalized by borrowers over time; and, 5) failure to monitor and supervise reform programs to keep them on track. The dissonance between SAL schedules and the longer time horizons required for institutional development also constrained reform efforts. But some high speed eleme!nts of SALs act as a trigger for some public sector management reforms by exerting leverage for politically difficult actions and by focusing attention on the urgency of some reforms. Moreover, serial SALs may allow for more gradual development of institutional capacity. SALs may also be followed by sector adjustment loans to ensure a longer sequence of institutional reform. Technical assistance that is tightly coordinated with SAL reforms enhances the chances of institutional success. Most longer term public sector management issues have been addressed through TALs. But TAL schedules are often driven by SAL timing requirements, leading to inadequate preparation of T.A. projects. Technical assistance was often poorly delivered and expatriate-dominated, providing little transfer of skills to nationals. Institutionalization of public sector management reforms has been difficult, and some impruvements have unraveled or been reversed over time. It may be unrealistic, however, to view institutional change as monotonic or linear. The likelihood of sustained reform is also linked to government commitment. The findings of this review suggest that regimes' willingness to carry through on institutional reform vary with the state of the economy, the nature of political opposition, and the degree to which leaders agree with the substance of the reform measures. General design problems affecting SALs in general are also significant for institutional items. Thes" include the excessive number of institutional conditions attached to SALs. and the weak monitoring and supervision of institutional components. Tranching, though under criticism with regard to overall conditionality, may serve a useful institutional purpose, however, by prolonging Bank-country dialogue and encouraging ongoing monitoring of institutional components. Political factors are enormously important to institutional reform in - iv - SALs and not well analyzed in the Bank. Greater attention to these issues is needed in sector work. Preparatory sector work ln general was important in public sector management reforms in SALs. Most SALs seemed to promote the creation of new inatitutLons, raLsing concerns about institutional proliferation. Publlc sector management reforms that from the outset could be linked to actionable steps agreed upon by Bank and borrower stood a better chance of being successfully implemented. Those Lnstitutional reforms for which routinized methodologies were put in place had a greater chance of being sustalned over time. I. INTRODUCTION This review examines public sector management issues in structural adjustment lending. It focuses on those institutional aspects of structural adjustment that deal specifically with "macro-management" issues; that is, those issues related to improvements in the management performance of core central governmet institutions and to rstemlc changes in the public administration.1, The paper delineates the overall objectives of Bank public sector management interventions linked to adjustment programs, surveys the range of actions undertaken through structural adjustment loans and credits (SALs and SACs) and related 4echnical assistance loans and credits (TALs and TACs), identifies key implementational problems connected with these actions, and, to the extent possible, assesses their impact. Analytic Framework Two broad types of institutional reforms are considered. The first type consists of those items constituting institutional support for specific economic policy reforms contained in SALs (e.g., the simplification of administrative procedures for export incentives to support trade liberalization). The second type includes those institutional reforms that in and of themselves are considered critical to successful adjustment (e.g., governient wage bill reduction aud civil service retrencbaeiet). To a considerable degree, both types of interventions are fundamentally associated with the emrgence of structural adjustment lending from the late 19709 forward. Although Bank lending did support some institutional reforms prior to 1980, these were mostly limited, sector-specific interventions, undertaken either through project-related lending or through free-standing technical assistance. The increased implementation requirements imposed by major economic adjustment exercises in many borrower countries, however, highlighted the need to strengthen institutional capacity for policy formulation and program execution. Therefore, structural adjustment lending pursued a wider range of reforms of core economic policy and management institutions. In addition, deficiencies in the borrowers' overall public administration systems increasingly came to be acknowledged as contributing significantly to poor government performance as well as to distortions in the economy. Thus, SAL- related institutional approaches represented a more targeted institutional intervention, focused directly on specific policy outcomes. In this way they differed not only from institutional work done through Bank projects in individual sectors, but also from the more general, extensive exercises in institutional capacity building that characterized much institutional 1/ For a treatment of institutional issues related to state enterprise reform in adjustment lending, see John Hellis, OA Review of Public Enterprise Reform in Adjustment Lending, 11989. An assessment of institutional issues in sector adjustment lending is found in Samuel Paul, lnstitutional Reforms in Sector Adjustment Operations,' 1989. -2- develoruent activity carried out by various other donors prior to the advent of adjustment lending.2' To a degree, these earlier institutional reforms thus proceeded on a slow, sometimes meandering track, often prompting frustration and impatience at their limited impact. Structural adjustment lending attempted to carry out certain urgent institutional reforms in the 'fast lane," concentrating on only a few key items with relatively short time horizons. Is the present adjustment oriented approach to public sector management reform working? After ten years of adjustment lending what can be said about the achievements in institutional terms? What alternative models or instruments might work better? In its examination of the institutional experience with SALs to date, this review tries to address some of these questions. The analytic approach employed does not, by and large, emphasize the substantive content of individual components of the SAL reform package. Thus, conclusions about best practice with regard to specific reforms of the tax administration system, for example, are not drawn. The analysis focuses instead on generic institutional themes that transcend individual substantive components. These include issues such as the timing or sequencing of institutional reforms, the extent to which reforms are successfully institutionalized or sustained over time, and the degree of government commitment to change. Issues of internal Bank practice regarding analytic preparation, design, and supervision of structural adjustment lending are also treated. Hethods Sources for this paper included secondary materials, president and staff appraisal reports, supervision audits, projezt completion reviews, and, where available, OED ex post evaluations. File documentation was also consulted for some countries. For selected country cases, key informant interviews were conducted with Bank staff. (The country cases examined in depth includeds Kenya, Cote d'Ivoire, Panama, Jamaica, Ghana, Gambia, Senegal, Turkey, Togo and Thailand.) Both SALs and, where applicable, their companion TALs were analyzed. It is important to emphasize that the types of institutional questions posed by this study cannot be adequately addressed by a desk review. The impoverishment of institutional information in most internal Bank documents -- particularly with regard to implementation issues -- and the necessarily partial perspective of Bank personnel interviewed can at best support only inconclusive findings. These findings can provide working hypotheses to be expanded and verified through a more intensive analysis based on field visits and interviews with borrower country participants in SAL programs. These data constraints have meant that :he criteria for evaluating results of a given component were often limited to whetner inputs were 2/ Such extensive interventions typified much of the institutional work sponsored by the United Nations and by the U.S. Agency for International Development, two of the agencies most active in this area in the 1960s and 19709. -3- successfully put in place through a SAL or TAL. The degree to which the delivery of these inputs could be associated with particular outputs or outcomes was not always-possible to &nalyze. Finally, this paper should be viewed as a cumulative product as it draws heavily from a previous CECPS paper, "Institutional Reform: Some Lessons from Structural Adjustment Lending,* written by Geoffrey Lamb in 1984. The following section presents an overview of the types of institutional components incorporated into SAL programs. The third section identifies common institutional issues that have emerged in the implementation process. The final section draws some conclusions about this experience. II. OVERVIEW OF INSTITTUTIONAL COMPQOENTS IN SALS AND TALS As of June 1988, significant public sector management components were present in all fifty-nine structural adjustment loans. Of these, thirty-one were in Africa; seven in Asia; thirteen in Latin America and the Caribbean; and eight in EMENA. In all, forty-three Technical Assistance Loans have provided spocific institutional support for these SALs. Well over half of these T.A. projects were in Africa. The others were distributed between EMENA and LAC; Asia had none. (Annex 1 shows all approved SALs and TALs with institutional comDonents. Annex 2 gives a regional breakdown of SALs with institutional components as of June 1988. Annex 3 shows the distribution and relative importance of different institutional reform components for SALs and TALs. Annex 4 shows the frequency and regional distribution of institutional components in structural adjustment and related technical assistance loans as of June 1988.) Most institutional components of SALs &re directly linked to the economic policy reforms of the adjustment process itself. The following areas, identified by Lamb in his 1984 review of SALs, remain prominent: institutional changes required for macroeconomic and financial management (present in all 59 SALs), non-financial sectoral institutional reform (found in 56). and, trade policy and administration reforms (present in 50). In addition, public administration reforms, not strictly part of economic policy adjustment but considered crucial to public expenditure control, and government performance were part of 31 SALs. Within these broad categories, a diverse set of activities has been undertaken. Although results have varied, some common patterns of experience can be discerned. This experience is discussed in the sections below. A. ECONOMIC AND FINANCIAL MANAGEMENT The most common institutional feature in SALs has been their emphasis on improvements in governments' management of core economic instruments. As Annex 3 shows, this rubric encompasses a series of activities aimed at improving the functioning of relevant government agencies. Measures -4- undertaken through SALs include the following (in descending order of frequency): strengthening of muanagemext of the public investzent programming process; strengthening of institutional mechanisms-for national planning; creating and/or reinforcing economic policy-making bodies; improving systems for debt management; Improving institutional mechanisms for tax administration; ref6rming institutions involved in the budget process; reforming banking and financial institutions; and, reforming systems for accounting and auditing. Many actions taken to increase government efficiency in economic and financial management were designed to bring about short-term improvements, beginning with initial emergency steps (such as the creation of a SAL oversight committee to ensure loan conditionality compliance for tranche release) that would precede longer-term efforts requiring more complex changeq ;n economic management and public administration (such as the design and installation of new systems and procedures for customs administration). The types of elements that make up these economic and financial management components are elaborated below. i. Public Sector Investment Program Management Public expenditure .ntrol was viewed as key to most adjustment programs, and improving public invetsi-nt program management thus became a frequent reform item in SAL packages. Forty-eight SALs supported by 12 TALs (7 in Africa and 5 in LAC) contained measures to improve public investment programming (PIP) management and to strengthen its links to budgetary control. The actual measures varied from one country or one SAL to another, but in general, SALs promoted high-level coordination of investment decision making, improvement in planning and investment mechanisms, and strengthening of investment monitoring. Specific actions have commonly included the establishment of multi-year rolling public investment programs as was instituted for example, in Pakistan, or, more immediately, the inventory of existing investment projects and the installation of a computerized monitoring system for ongoing projects. In Ghana, for example, a central project monitoring unit, ultimately destined for the Ministry of Finance and Economic Planning, was temporarily housed in the World Bank resident mission during SAC I preparation because its activities were considered immediately critical to rationalization of the investment process. Efforts to strengthen financial and technical project preparation and evaluation capacity have been undertaken in planning and budgeting units in ministries of finance and planning as well as in sectoral ministries. Such efforts have sometimes involved increasing or enhancing staff capabilities of key entities, either through training or recr .'tment. Enhanced capacity has commonly been sought through the use of forei.s., consultants financed through technical assistance. In some cases, such consultants were charged with training local counterparts; in most instances, however, they were hired to carry out specific tasks themselves. In some instances, such as SAL II in Kenya, systems and methods were standardized in a handbook for project evaluation subsequently adopted by the Treasury, and new regulations for regular review of projects were introduced. Improvements in the public investment progrmding process have involved reorganization of planning and budgeting structures. The organizational arrangements endorsed by SALs varied. Some reorganization efforts strengthened coordination among ministries of finance or planning and other - key agencies, such as t.i Ministry of Comerce and Industry in the Central African Republic. Others promoted the establishment or reinforcement of a planning agency separate from the Ministry of Finance. This occurred in Ghana and in Jamaica (in the latter case with the creation of the Planning Institute of Jamaica). In a few countries, reorganization strengthened centralized control of the budgeting process. In the Cote dl'voire, for example, the Ministry of Finance and Economic Planning was at the center of this effort. In general, Bank staff report that the direction and shape of these reorganization schemes were determined less by a technical prescription of the best organizational arrangemeuts for planning activities than by the institutional preferences and traditions of governments themselves. Public investment prograumming evolved in many different ways. In Togo, for example, PIP reform began with the introduction of macroeconomic planning linked to the establishment of procedures for investment project monitoring on a quarterly basis. The reform was carried out by expatriate consultants with minimal emphasis on training of nationals. In the Gambia, SAL T sought to strengthen core and sectoral institutions involved in the public investment programming process through technical assistance and training and the promotion of routinized review of projects. In Turkey, efforts to rationalize the PIP process with the ultimate aim of reducing the number of projects in the investment program were phased-in gradually over a period of five SALs. Key institutions, such as the Evaluation Unit for Special Projects, were upgraded through technical assistance and training. It is exceedingly difficult to assess the results of these diverse efforts at reform. Interviewed Bank staff conveyed partial, conflicting or inconclusive impressions as to the degree of progress achieved. They reported improvements in Togo, Jamaica, Gambia and Turkey, but in Kenya repeated delays in the preparation of a rolling public investment program (in conjunction with a forward budgeting system) led to a nine month delay in tranche release in SAL II. Poorly trained staff and high management turnover la the sectoral ministries exacerbated the weak PIP performsnce. In the Central African Republic, the management capacity of key agencies was overwhelmed by attempts in SAL I to introduce a series of simultaneous reforms, including PIP improvements, and the entire reform agenda slipped behind schedule. In Bolivia, SAL I efforts to strengthen expenditure control and investment planning in the Ministry of Planning and Coordinat'on were deemed unsuccessful because no systematic procedures were set up to ensure analysis and execution of future investment programs. Even the more positive Turkish experience was not an unqualified success. Despite significant cuts in the annual growth in public investment (frem 18 percent to 2 percent over the period of five SALS), the PIP still included some 6000 projects, and by the end of SAL V there remained a clear need to strengthen institutional planning capabilities in sector ministries and agencies. After the end of the SALs the growth in the Turkish PIP rose again, to about 1OZ per annum in the period 1985-1987. In Jamaica, SALs were used to develop PIPs that for the first time linked investments and the mid term macro-economic framework. Under Sal III a three year rolling PIP program established a direct connection with the government's Medium Term Economic Policy Framework for the period 1988-1991. The Senegalese dialogue with the Bank in SAL II led to a comprehensive study of investment programming and budgeting systems. This resulted in some improvements in SAL III, including the 1987 implemsntation of a computerized monitoring 3ystem for the physical and financial execution of public investments. In the Cote d'lvoire, PIP reform made some headway with support from a related technicaf assistance loan establishing a project monitoring unit in the Investment Department in the Ministry of Budget, but deficiencies in the budgeting process and in national accounts procedures persisted throughout the period of SAL I to III. Although some reforms did occur, these experiences suggest that the process has been difficult and the results decidedly mixed. Slow progress has been linked to the difficvlties borrowor government staff have in mastering the technical skills needed to manage the PIP process. This may be due to the - failure of technical assistance to provide adequate training for these personnel or perhaps to the inappropriateness of the systems to the peculiarities of individual country circumstances. Although rarely mentioned by Bank staff, political pressures brought to bear upon the investment planning process, either by external interest groups or by internal bureaucratic actors, may well have derailed some reforms. ii. Economic Policy Bodies Structural adjustment programs often responded to deficiencies in macroeconomic policy-making capacity by promoting the creation of new economic policy-making bodies or the strengthening cf already existing ones. Such entities were part of 26 SALs and 18 supporting TALs, of which about three- - fourths were in Africa. Examples include the National Development Planning Commission in Gtina, the Financial Coordination and Investment Control Committee in the Cote d'Ivoire, the National Economic Policy Steering Committee, created in Thailand to coordinate macroeconomic management with the National Economic and Social Development Board, and the National Economic Council in Mauritius. In some cases, these bodies were set up to carry out policy functions on a long-term basis, in the absence of established government institutions. In other instances, entities were put in place precisely to deal with immediate SAL requirements. The Comite de Suivi et Reflexion in the Cote d'Ivoire was conceived thusly. Its purpose was to monitor the progress of SAL activities and serve as an informed interlocutor with the Bank SAL team. Although it was hoped that this group would eventually become permanently institutionalized in the Ivoirian policy environment, its explicit brief was more directly linked to short-term SAL demands. -7- The record of such bodies has been mixed. When constituted primarily for SAL purposes, they proved 4ifficult to sustain in the absence of persistent Bank prodding and funding. Indeed, some officially constituted committees never materialized as real working groups. In Malawi, for example, the Investment Coordination Committee that was to review public investments never took form, and its mandate was eventually assumed by the Economic Planning Department in the Office of the President and Cabinet. Bodies that succeeded in becoming more institutionalized sometimes became policy enclaves, isolated from other parts of government and siphoning off managerial talent from central bureaucratic organs. Both the Thai NESDB and the Economic Liaison Unit established under SAC I in Ghana suffered such criticism. The overall impact of these groups on the adjustment process is also hard to assess, although it may be easier to detect the adverse effect on adjustment programs when such mechanisms do not exist. In Turkey, the absence of a strong central institution with responsibility for keeping the macroeconomic adjustment program on track led, according to involved Bank staff, to the aforementioned slippage in public sector investment program reforms. In Bolivia, the failure to establish a central coordinating body to follow up on policy measures put in place during preparation and execution qf SAL I partly explained, according to Bank staff, the diminished impact of SAL programs ont the economy. Occasionally, the above bodies were coincident with those special project units set up to administer a number of SALs. Such units sometimes acted as secretariats for the economic policy-making bodies described above, and they were oftejn staffed with the best and the brightest from local technocratic elites, many drawn directly from the central bureaucracy. Most Bank staff contended that these units constituted an essential mechanism for effective SAL management, however, and that this talent diin was justified by the urgent, short-term requirements of SAL program implementation. Rarely was explicit attention given to the effect of this drain on the ability of other parts of central government to carry out routine functions. In countries where skilled human capital was scarce, this often constituted a significant problem. iii. Debt Management Systems Reductions in international trade and declining commodity prices have underscored the importance of management of the foreign debt to successful adjustment. Thirty-three SALs accompanied by 7 TALs (about one-half in Africa) instituted changes in debt management arrangements. The reforms usually involved the establishment or reinforcement of institutions to monitor and control debt and the development of systems to feed pertinent information on debt into budgeting and public investment programs. In particular, reforms emphasized coordinating efforts to link line ministries and financial authorities (such as central banks) into debt management systems. Under the SAL programs Thailand developed a comprehensive three year rolling program for public foreign borrowing. Bank staff report, though, that the analytical capability to utilize fully the available computer facilities is still lacking. The Cote d'Ivoire, on the other hand, reportedly strengthened its system for debt manaBement under SAL I with the computerization of debt management information and close monitoring of guarantees and internal debt repayment schedules. SALs introduced computerized debt management systems into a number of countries, including Malawi, Mauritius, Niger and Panama. Some countries set up or reinforced special units to monitor debt. A debt monitoring unit was established in the Treasury in Malavi and became an important management tool. Despite this success, Bank staff commented that by the end of SAL II, a coherent strategy and plan for debt management still needed to be developed. This was taking much longer than the SAL program had anticipated. Indeed, the degree to which the introduction of these mechanisms has transformed debt management practices remains to be seen. Turkey, for example, introduced a SQries of institutional improvements for debt management over the course of five SALs. After extensive studies in SALs I and II, improvements in external debt reporting and computerization along with staff strengthening through training and outside recruitment were carried out in the Ministry of Finance in SAL III. Nonetheless, by SAL IV debt manaBLment was still thought to be unsatisfactory, despite the Government's stated overall commitment to the SAL policy dialogue. Bank staff attribute this disappointing performance, at least in part, to the Government's failure to reconcile its conflicting goals of decentralizing authority to line ministries with the need for the Central Bank to assert authority over external debt. In general, SAL implementation has yielded more transparent procedures and more streamlined mechanisms for debt management at least in the short term, but the institutionalization of real, meaningful controls over debt management processes may be more elusive and difficult than many SAL designers originally envisaged. iv. Tax Administration As part of the process of rationalizing the macroeconomic framework, efforts have also focused on the revenue side of government adjustment programs, addressing the need for reform of tax administration institutions and procedures. Institutional components dealing with tax administration were found in 33 SALs supported by 7 TALs (Africa accounted for one-half and the rest were distributed among LAC, EMENA and Asia). Measures to upgrade the efficiency of tax administration have been diverse and defy easy generalization. In several cases, such as Togo, efforts have been made to develop a new tax code, with good results. In Dominica, customs administration procedures were streamlined. Similar measures, involving the reduction of the export tax on cashews and improvements in tax collection procedures, were undertaken through SAL I in Guinea Bissau. In Thailand, fiscal policy and related tax administration procedures were a key feature of the SAL programs. A revenue projection model was developed for multi-year revenue planning and annual budgeting purposes, replacing the -9- purely ad hoc annual extrapolations previously utilized. Concomitant efforts were made to streamline and standardize tax administration proceduAres through organizational changes in the revenue, customs and excise departments. Turkey's SALs also aimed at better tax admin_.stration. While improvements occurred during SALs II and III, budget deficits as a share of GyP started to rise by SAL IV. One factor contributing to these mixed results was the- emergence of autonomous agencies (using extra-budgetary funds) that preempted some tax reserves while increasing the overall level of expenditure. B. NON-FINANCIAL SECTOR REFORMS Institutional measures in most SALs went beyond the core economic and financial institutions to support sectoral restructuring and policy reforms. All but three SALs contained such components. Sectoral institutional reforms have emphasized: 1) institutional support for policy changes involving the liberalization of price and incentive structures; 2) improving capacity for strategic planning in sectoral organs; and 3) reorganization of sectoral institutions. Sectoral institutional components have been heavily concentrated in agriculture (45 SALs and 14 TALs) and industry (39 SALs and 16 TALs). Energy, an institutional component in 17 SALS and 6 TALs, lagged considerably behind as the third most prominent sector. In descending order of importance, operational components providing institutional support for policy reforms were distributed among the following additional sectors: health and social security (9 SALs and 6 TALs); transportation (7 SALs and 1 TAL); population (5 SALs and 2 TALs); water and urban development (2 SALs and 2 TALs); and housing (2 SALs and 1 TAL). i. Agriculture Reforms Institutional reforms in agricultural sectors frequently stressed reorganization of sectoral institutions. They typically emphasized improvements in policy analysis, strategic planning and monitoring capacity, and overall reduction of the size and number-of institutions operating in the sector. In the Gambia, the Ministry of Agriculture was streamlined (in conjunction with an overall public administration reform) and priority was given to functions of strategic planning and monitoring. In Senegal, large rural development agencies were restructured and their staffs reduced. Planning, investment programming, budgeting and statistical analysis were strengthened in the Ministry of Rural Development. In Cote d'Ivoire, an important additional element was a series of management studies to be carried out on principal agencies in the sector. In Kenya, SAL II supported a previously launched effort to strengthen the institutional capacity of the Ministry of Agriculture through improved budgeting, computerization of project monitoring and personnel management, the introduction of a system of timely expenditure reporting, and the quicker release of authority to incur expenditure. - 10 - SALs have sought to strengthen agricultural sector institutions in support of liberalization of pricing policies, removal of subsidies for inputs and the reduction of the role of the state generally in marketing and production. Institutional support for these reforms consisted of studies (e.g. in Togo, Cote d'Ivoire, Kenya, Senegal), creation of new institutional mechanisms such as the Price Analysis Unit of the Export Czip Marketing Board in Togo or the Task Force on Budget and Financial Management in the Ministry of Agriculture and Livestock Development in Kenya, or, in so'le instances, the total or partial elimination of institutions. A good number of organizational reforms sought to limir. the activities of government institutions in various aspects of agricultural marketing and production. In Jamaica, major commodity marketing boards were divested of all non-marketing functions, and private sector participation in external marketing of export crops was promoted. In Cote d'Ivoire, the sugar marketing board (SODESUCRE) was reorganized and various of its direct refining operations liquidated. SAL I in Panama generated some broad institutional changes in support of liberalization policies. This was effected mainly through the passage of legislation prompted by Bank-country policy dialogue. The Agricultural Incentives Law of 1986 abolished import substitution targets, reduced import duties on food crops,.and announced the end of crop purchasing by the National Rice Institute as well as the abolition of the crop-pledging system of the Agricultural Development Bank, a condition of second tranche release in SAL II. Each of these reforms resulted in the limiting or elimination of functions of the supervising institution. ii. Industry Reforms In industry, major institutional reforms were often linked to changes in the trade regime, usually involving rationalization of the structure of industrial regulatory institutions and procedures. This mainly involved efforts to reduce the proliferation of institutions and streamline bureaucratic processes in order to provide an institutional push to exports. A variety of measures were taken through SALs to achieve these ends. In Kenya's SAL program improvements in the administration of industrial investment incentives were sought through the constitution of an interministerial New Projects Committee. In an effort to apply standardized technical, financial and economic viability criteria to government investment decisions, this committee was charged with the review of all proposed projects. In Thailand, the emphasis was on reducing the case-by-case. discretionary basis of investment decisions by developing practices to clarify qualification criteria and procedures. In the Cote d'Ivoire, it proved difficult to establish a system of export financing and credit insurance. The unit to monitor the industrial reform that was to be located in the Ministry of Industry was never established, in part because the Ministry was reluctant to use technical assistance funds to set up the office. Bank staff cite the failure of government to liberalize the manufacturing sector in the Cote d'Ivoire as a - 11 - reflection of the overly ambit-ious and politically unrealistic nature of SALs II end III. Indeed, price liberalization measares taken under SAL II in the shoe end textile industries were rescinded in SAL III in response to heavy political pressure placed upon the President by the Ivoirian industrialist lobby. iii. Trade Administration Reforms In its policy dialogues on structural adjustment, the Bank has consistently pushed for more liberal trade regimes. Policy reforms have included moves to abolish quantitative restrictions, lower tariffs in concert with the maiW-qnance of a realistic exchange rate policy, end establish an incentive structure neutral between import substitution or export generating activities. Although most countries have taken only incremental steps in these directions, a number have undertaken a series of institutional changes in support of the kinds of trade reforms promoted through structural adjustment lending. Many institutional reforms in this area sought to replace complicated systems of import and *xport quotas and licenses with simplified tariff systems. Gabon, Burundi, and Central African Republic, among many others, followed this route. Countries streamlining institutions or proceduzal arrangements for trade administration included the Cote d'Ivoire (which created en export credit insurance system under SAL III) and the Gambia (which liberalized exchange control procedures so that commercial banks were not required to submit detailed documentation to the Central Bank when paying for commercial import transactions). Several countries carried out studies of various parts of their trade regimes under SALs. In Turkey, a SAL I study was undertaken to set up a joint public-private export promotion center. (It was revitalized under SAL II.) In Turkey's SAL V another study examined the possibility of introducing a system of credit for indirect exporters. Trade reform also prompted the creation of new agencies and the reinforcement of old ones. An Export Development Bank was created in Turkey under SAL II, and in Burundi the Bureau of Standards was created under SAL I to ensure the quality of the country's exports. In several countries (Central African Republic and the Gambia), existing organizations such as the Chamber of Commerce were strengthened. In Thailand, the Export Development Committee was created. In some countries, revision of existing regulatory and investment codes or new draft tariff codes were the main institutional steps taken toward trade liberalization. This occurred, for example, in Gabon and Senegal. Several countries put drawback systems in place. In others, export declaration systems replaced export authorization systems (Central African Republic), and in still others (Sao Tome) import license auction systems were developed for non-essential imports. - 12 - As the above sections suggest, sectoral reforms clearly constituted an important aspect of many SAL programs. At the same time, Bank staff reported an increasing inclination and/or preference for dealing with sectoral institutional issues through sector adjustment operations rather than through structural adjustment umbrellas. It would appear, nevertheless, that SALs are still the instrument of choice for sectoral reforms if: 1) the sectoral issue is important enough to the overall economy and essential to the adjustment process; 2) the country is small enough that administrative efficiencies can be gained by a centralized SAL effort and sectoral reforms can be linked to more central structural reforms without fear of the whole package being sabotaged because of lack of political comitment on one or another single item; or, 3) structural adjustment can be used as a means to provide reinforcing cross-conditionality on sectoral reform issues. In general, though, there does seem to be a move from the early use of SALs for sectoral reforms to SECALs in more recent cases. This stems, in part, from the prevailing belief among Bank staff that it is generally easier to get countries to adopt narrower, more specific reforms. Sectoral line agencies are more likely to comprehend the technical rationalization for specifl- reforms in thoer own sectors and therefore more likely to agree to them moxe readily. Furthermore, many sectoral reforms can proceed independently of necessary macroeconomic reforms and thus do not have to be linked to larger SAL programs. It has also been suggested, however, that the price for such reforms has come high: #t number of sectoral reforms once -'.cluded but not completed in earlier SALs (e.g., agricultural reforms in Kenya) have reappeared in later incarnations in more specific SECALs. C. PUBLIC ADMINISTRATION REFORMS Increasingly, public administration reforms have become prominent institutional components of structural adjustment lending. Such components generally comprise civil service management and/or reorganization reforms. mostly emphasizing structural changes in central government employment, pay and incentive systems. Of relatively recent origin, reforms in public administration were components in 23 SALs between 1981 and 1987. In addition, thirteen TALs supported these SAL-driven reforms. In general, these reforms were designed to address the following problems: excessive public sector wage bills, too many civil servants, gradual erosion of public service salaries, or wage compression between top and bottom grade levels in overall salary structures. SALs varied in the depth of their treatment of civil service reform issues. In reforming public pay and employment systems, SALs sometimes merely set global reduction targets for the aggregate wage bill (often determined in collaboration with the IMF) without specifying a strategy to reach these goals. Such gross prescriptions may actually have done more harm than good, since countries with highly compressed salary structures sometimes responded to this requirement with simple, across-the-board wage restraint, without adequately considering the complex trade-offs between employment and pay. Sometimes the inclusion in a SAL of such conditionalities was intended to - 13 - encourage the continuation of government actions already taken. In most cases, however, the Bank did specify the means by which the overall wage bill might be reduced. In these cases, conditionality was introduced to restrain salaries, freeze or reduce employment and pay in the civil service, or, where necessary, raise civil service salaries, usually in the context of an overall reduction in civil service numbers In general, the TAL approach to public employment and pay issues was to provide support for longer term, more comprehensive reform efforts. Nine TALs sought to strengthen institutions and procedures or to collect information that would provide the foundation for more radical employment and pay reforms at some later date. Several countries have embarked on rather ambitious programs of public administration reform that have been supported by SALs, including Jamaica, Ghana, the Gambia, and Thailand. An examination of these cases reveals some of the implementational issues arising in such programs. In Jamaica, administrative and budgeting reforms were instituted under SALs I and II through management audits of core ministries and initial moves toward performance budgeting. As the program of reform gaiqed momentum it developed into a $4.5 million public administration reform project (mainly comprising technical assistance), strengthened by crosscutting conditionality in SAL III. The Administrative Reform Project I (ARP I) focused mainly on reforming personnel management practices in central.government through strengthening and decentralizing the Ministry of the Public Service, the Public Service Commission and improving staff personnel management capabilities in the line ministries. In addition, it sought to promote and deepen budgeting reforms in the Ministry of Finance and relevant parts of sectoral ministries. Finally, ARP I was intended to support public pay reforms included in SAL III. Although progress may have taken place for some elements of ARP I, the program as a whole remained very dependent on the Prime Minister's backing, and it was continually constrained by the absence of high quality managerial staff in the civil service. The SAL program tried to ease this constraint -- with only limited success -- by encouraging pay reform to attract qualified personnel back into civil service posts. SAL IIT, for example, instituted conditionality that higher level management posts in the civil service be compensated at 85 percent of equivalent grades in statutory bodies over a three-phase pay hike. Although two raises for managers did occur, the opposition from unionized civil servants at lower skill levels presented obstacles to diffe.rential wage increases favoring higher grade employees, and the third increase to bring up higher echelon civil servants to the desired level never took place. As a result of this delay and the continuing wage erosion through inflation, qualified professionals and managers are increasingly scarce in the central government. Technical analyses supporting the pay reform policy prescriptions (i.e. a pay comparator survey) was to be supplied through the Administrative Reform Project, but the work had not yet been completed at the project's end due both to internal project management - 14 - constraints and possibly also to the perception that this reform had fallen off the Prime Minister's agenda and should be given a low priority. Such reforms may indeed have slipped into low or even reverse gear in Jamaica in the post-adjustment climate. That Jamaica may have suffered a form of adjustment fatigue' can be seen in the slippage reported to have occurred with regard to civil service retrenchment. Although approximately 20,000 emplovees were laid off by central government over the last three years (at the urging of the IMF rather than the Bank), it is suspected that some portion of these workers have been rehired., The reasons for this reversal in Jamaica remain hazy. The vast majority of redundant workers came from local governments (counted officially, however, as part of the central civil service). The functions of local governments were virtually abolished for many parishes under the Seaga regime. Some of these functions appear to have been contracted out to the private sector, but others may have been reincorporated into central government, thus explaining, in part, some rehiring. In addition, the powerful position of unions with strong links to the dominant political parties may have influenced central government managers to accept old staff back into their jobs despite official sanctions against this practice. In Thailand, the broad program of SAL-supported reform included (as a preliminary step to future reforms) a study of civil service organization and compensation. Although Thai reluctance to borrow for technical assistance curtailed some potential reform activities, involved Bank staff stated that the process of change was institutionalized,* by being actively taken up by the Cabinet's administrative reforms committee. This miade the reform program an integral part of government decision making rather than an effort remote from the bureaucracy. The cost of this support might have been the risk of politically-inspired changes in the original reform concept, but this risk was substantially avoided, and the SAL program set in motion a process which could lead to institutional changes. These changes could include, eventually, the adoption of proposals currently before the Cabinet to restructure the civil service pay system in ordet to recruit higher quality staff more effectively in light of rising competition from the parastatals and private sector. In Ghana, SAC I included conditionality for the retrenchment of civil service cadres and the reduction, phased over a three year period, of personnel expenditures to between 5 and 6 percent of GDP. This fairly immediate reduction was considered critical to the overall success of the government's economic recovery program. But it was recognized that implementation of even these immediate steps would require enhanced institutional capacity. Longer-term institutional strengthening of both core economic policy agencies and those organs charged with the management of cadres in the civil service would necessitate additional support. This support was supplied through the Structural Adjustment Institutional Support technical assistance project, whose components included studies and assistance to improve the performance of the civil service in carrying out such tasks as formulation of salary and grading policy, functional reviews of central government agencies, redeployment of surplus staff, strengthening the - 15 - management services division, and training and career development. This project offered institutional support of a more focused nature than previous SAL-related T.A. projects. At the time of this writing, Bank staff view Ghana's pay and employment reforms as a qualified success. Approximately 11,000 civil servants were retrenched in 1987, falling somewhat short of the 15,000 target set by the Bank and the Fund. The government's ability to sustain the reform program in the face of potential political opposition to the retrenchment scheme remains to be demonstrated of course. And, despite the reiatively focused and coherently designed nature of the technical assistance for the overall civil service reform program, such aspects of implementation as lay-offs and redeployment, retraining, severance and possible credit or incentive schemes were not fully designed or costed out, resulting in some delays and, at times, some concern about program derailment. There are also indications that perhaps 3000 civil servants had slipped back onto the government payroll, reversing at least some of the positive financial effects of the employment cuts. Despite these qualifications, Ghana, nonetheless, made important strides in its reform program. In the Gambia, the SAL supported a series of clvil service reforms, including the retrenchment of approximately 15 percent of civil service employees (which proceeded ahead of schedule), preparation of a comprehensive program c. organizational changes, improvements in personnel management, manpower budgeting, pay and grading, and functional reviews. Although there were delays in implementation due to difficulties in assembling technical experts to assist the Gambians, the civil service reforms are thought to have gone well -- especially the retrenchment exercise -- partly due to the preparatory technical analysis. This analysis provided information for difficult policy decisions about where the redundancies were and how workers should be dismissed. Dismissals were carried out with considerable sensitivity to political and social factors. The Establishment Office, charged with the overall management of the retrenchment exercise, reportedly avoided firing people from the same family or tribe, for example. The stability of the regime and the popularity of the president are credited with the smooth execution of the program so far. It is assumed that many retrenched civil servants have been absorbed into airicultural activities in the countryside, but no monitoring has taken place. I Bank staff expressed concern.that urban unrest may increase with further cuts in public sector employment. The true test of the efficacy of these reforms triggered by the SAL in the Gambia as well as elsewhere is the degree to which an ongoing process of improvements in civil service management will have been set in motion. This will depend largely on the institutionalization of establishment control systems (including a computerized link between personnel records and 3I This positive view may need to be revised in light of a field evaluation which is to take place in the context of an Africa Department Regional Study of Cases in Pay and Employment Reform, scheduled for early calendar 1989. Indeed, interim reports on progress in the Gambian case have raised some doubts about the roseate picture presented here. -16 - the payroll), training in personnel policy and manggement techniques for Establishment Ministry, and revamping of career incentive structures for the, civil service. D. SUMMARY OF FINDINGS Despite the methodological limitations mentioned earlier, the data upon which the above sections wera based do suggest the following emerging trends that might be substantiated through more rigorous field investigation: o In a few instances, improvements in public investment programming institutions and procedures led to or set the stage for budgetary reform. Rolling investment programming was introduced and investment project monitoring was tied into macroeconomic planning. o Ne. institutional mechanisms for dealing with debt management have been established in some countries; improvements in coordination between ministries and central banks and other financial authorities have resulted. o A good number of economic policy institutions, responsible for developing, coordinating and reviewing macroeconomic management in general, and adjustment operations in particular, have been established, with varying impact. In some cases, these bodies have assumed an important role in economic policymaking and in Bank-country dialogue. In others, they have become enclaves, divesting scarce human resources from other government organs. In still others, such economic policy bodies have been short-lived, failing to become institutionalized in the country's bureaucratic environment. o New agencies have arisen to deal with trade regime reforms. Although this has sometimes raised the spectre of institutional proliferation, there has been simplification and rationalization of trade administration procedures in a number of cases. o Organizational and managerial improvements in key sectoral institutions has occurred, and the reduction of the number of institutions operating in some crucial economic sectors has led to improved performance. o A reduction in the number of civil servants and a redefinition of the role of the civil service has occurred on a limited basis in some countries; movement toward a more cost-effective use of civil servants has also taken place. Despite qualified successes in a number of areas, however, institutional achievement has been, for the most part, limited. Indeed, many initially successful reforms have been reversed subsequently. Some of the reasons for - 17 _ this halting progress have been catalogued above: 1) weaker than expected management capacity unable to absorb excessive reform demands; 2) rapid turnover of management and political leaders, thus breaking the continuity of the learning process as well as the commitment to reforms; 3) poorly executed technical assistance and inadequate training of local staff in skills necessary to implement and sustain reforms; 4) failure to install systems and routine procedures that can be institutionalized and internalized by borrowers over time; and, 5) failure to monitor and supervise reform programs to keep them on track. Finally, although staff training, reorganization, and introduction of new management techniqtxes may have *strengthened' procedures and institutions, a link between such improvements and better government performance has not been established in all cases. III. SELECTED OPERATIONAL ISSUES A number of operational issues emerge from the above discussion of experience. Some of these issues have been identified in earlier reviews of SALs, but their importance has been confirmed by the accumulation of more recent evidence. A. TIMING The dissonance between SAL schedules and the longer time horizons required for institutional reform has been discussed in previous analyses of SALs. The findings of the present review confirm this mismatch, especiully for many of the African cases where institutional structures are weak and new and may take a generation or more to develop or recover. Most institutional development activities in the SALs examined in this review were taking much longer than had.been foreseen in SAL design -- if they were being implemented at all. Although many Bank staff involved in puttiAg together SAL programs were well aware of this problem, schedules and deaC'lines contained in action plans and loan agreements continue to be unrealisti: about the time required to accomplish institutional reform. Despite these commonly expressed frustrations with the inadequacy of quick-disbursing SALs for instiputional reform, however, it may also be true that some of the high-speed elements of SALs can sometimes facilitate institutional change. SALs can act as a trigger for some public sector management reforms (such as pay and employment measures), both by exerting leverage for politically difficult actions and by simply focusing attention on the urgency of certain reforms. Moreover, it may be something of a caricature to depict SALs as relentlessly short term. For a number of countries (e.g., Turkey, Cote d'Ivoire, Jamaica, Togo, Senegal), SAL programs have been implemented and were conceived originally as serial processes, sequenced over periods as long as - 18 _ ten years. Th'is is probably ample time at least to begin the institutional development process. (Over a serial succession of SALs, studies can become actions and some *behaviors' get institutionalized; public investment programming methodologies get more comprehensive, for example, and spread to collaborating agencies). And for some types of sector-specific institutional reforms, SALs may be the first step in a more sustained process in which sector adjustment loans follow upon activities initiated under a SAL (e.g., agriculture in Kenya and industry in Ghana.) In addition, there are numerous examples of cross-sectoral adjustment programs of which the SAL is only one element. Finally, serial SALs not only provide a sustained infusion of resources to bolster policy reform, they may also furnish an ongoing context for policy dialogue. This, in the view of some Bank staff, may be the most important contribution Qf structural adjustment lending to development. In addition, an ongoing SAL lending program may encourage the establishment of a stable technocratic cadre in government, an independent policy elite with the capacity to critique Bank reform proposals and to produce its own reform program strategies. This was reportedly the result of successive SALs in Turkey, Thailand, and, to some extent, in the Cote d'Ivoire. B. TECHNICAL ASSISTANCE The crucial importance of technical assistance to carrying out reforms in the short term and to sustaining them in the longer run was repeatedly stressed by Bank staff. Technical assistance in support of the institutional. aspects of SALs has evolved over the history of structural adjustment lending in the Bank. Early SALs tended to use TALs somewhat as relatively unspecified 'slush funds. dipping into them as needed. This provided flexible institutional resources for SAL requirements, but it did not ensure focused, continuous, tightly designed support for key SAL activities. Increasingly, however, the longer term institutional issues in structural adjustment are being addressed through companion technical assistance loans, carefully crafted to provide more direct institutional support for SAL measures. These companion TALs are most prominent in Africa - - especially among the more recent loans -- where overall capacity of most institutions is weakest. TALs designed in conjunction with SALs keep institutional concerns on the agenda in a way that previous ad hoc technical assistance did not. But TALs accompanying SALs are not in the driver's seat; their design often suffers as a result of the sometimes break-neck speed applied in taking a SAL to the Board. Often they are under-prepared -- both in terms of the prior sector work necessary to ensure good design and of the time required to allow full participation and internalization by governments. In Africa, there is an enduring lament that technical assistance means expatriates who, despite statements to the contrary, provide little training for nationals. In the case of Togo, for example, the technical assintance for SALs I and II was all foreign, with little internalization of results. Bank staff complained that although the Togolese complied to the letter with most - 19 - of the requirementr of SAL I, country officials lacked real understanding of the meanin8 of the adjustment process. Instead, their compliance was motivated by fiscal desperation and by what one Bank economist saw as an apparent desire to belong to the 'SAL club,' membership in which was becoming a sine qua non of successful development. The tendency to rely heavily on foreign technical assistance for SAL implementation is a common Bank response to the urgent need to carry out critical SAL activities in a timely manner. It is also a reflection of the inclination of countries and the Bank to concentrate on discrete and formal aspects of institutional change. As Lamb pointed out in his earlier review, 'since SALs are short and bureaucratic life is long -- and since in any case attention is often quite properly focused on near-term policy changes -- there may be implicit incent$ves for governments to substitutp formal compliance for substantive change.' 41 C. INSTITUTIONALIZATION OF REFORMS The degree to which public sector management reforms can be stitained over time is also an important issue. In many cases it is far too early to determine if institutional changes are taking hold, but there is, nonetheless, reason for concern. Some institutional improvements appear to have been implemented but then unravel or are eventually reversed. Rationalization of the public investment programming process over the first several Turkish SALs, for example, resulted in a significant reduction in the number of investment projects. As noted earlier, however, the trend seems recently to have turned around. In a number of SALs institutional solutions have involved setting up committees and working groups, with varying degrees of ad hoc-ness, to deal with assorted aspects of SAL oversight and economic and financial policy management in general. In a number of cases (e.g.. the Comites de Suivi et Reflexion in the Cote d'Ivoire), such bodies have been assembled and then dissolved -- often shortly after tranche release -- or they may only be officially constituted to show formal compliance but never actually meet.. When such bodies do acquire a life of their own, they offer the possibility, suggested earlier, of generating a thoughtful, national counterpart with whom the Bank can maintain policy dialogue, ideally leading to some internalization of the reform process. This appears to have occurred in Thailand with the creation of the Thai Development Research Institute and the National Economic Social Development Board. But there is also the danger that such groups may become policy enclaves -- further debilitating traditional bureaucratic organs -- or that they may contribute to already rampant institutional proliferation and fragmentation. Other examples of slippage of institutiqnal reform are found in the area of government employment reforms, where cadres are reduced under pressure but 4/ Lamb, 1984 - 20 - then civll service recruitment is resumed -- as * function of all the contextual forcei that made it rise in the first place. (This seems to have occurred in Jamaica and Ghana.) The one possible consolation with regard to the slippage of reform is that institutional change should not be regarded as linear or monotonic. Expectations that fail to allow for some backsliding are probably unrealistic. D. GOVERNMENT COMMITHENT TO REFORMS The degree to which reforms may be sustained over time is related to the question of government commitment to change. The need for governments to make and sustain commltments to the reform process and the conditions under which this commitment takes place have been considered extensively in reviews of the adjustment experience. Still, several findings of this review may contribute to the ongoing debate. Staff perceive a correlation between government comitment to institutional reform and the overall state of the economy. In some cases, improvements in the economy reinforced government commitment to reform, as the fruits of past labor could increasingly be demonstrated to constituents. (Turkey and Jamaica are examples.) In other countries 'adjustment fatigue' set in, and economic improvements -- even when linked conspicuously to structural adjustment programs -- tended to defuse government fervor for change. As the objective need to impose reform at all costs relaxed, political leaders were more inclined to heed discontent among supporters and take more seriously the organized opposition to unpopular policy measures. (Kenya and perhaps Ghana during some periods could fit this explanation.) Thus governments appear to commit to or 'buy' institutional reforms under varying circumstances. First, governments often 'agree' to reforms contained in SALs if they represent measures already initiated or even carried out by the country itself. In this case, SAL provisions simply reward actions already taken. Second, governments may have decided on their own to undertake reforms with which the Bank is in agreement. Under these circumstances, the country's *internalization' of reform measures proposed by the Bank is something of a fiction. Third, the greater the economic need and the more dire the state of the economy as a whole, the greater the government's 'commitments to reform. (Most of Sub-Saharan Africa Conforms to this pattern, including Togo, Ghana, and the Gambia.) Fourth, as is widely acknowledged, countries which internalize the rationale for reforms or come to own them, usually after sustained dialogue with Bank staff, demonstrate a hig4er commitment to seeing them through. (This may have occurred in Turkey, Thailrnd and more recently in Ghana for some issues.) E. DESIGN ISSUES The degree to which governments commit to reforms, and the extent to which reform programs can be sustained over time, are obviously affected by -21- the way in which SALs are designed and by the Bank's own role in implementation. Several issues are important. The way in which SAL conditionality for public sector management reform is designed and implemnted may affect the institutionalization of the reform program. A iignificant point of discussion among those Bank staff involved in SAL design was the use of tranches in promoting compliance wit:h institutional conditions in adjustment. Many staff believed that tranching provided only limited leverage in enforcing conditionality. S'nce tranche release appears to have occurred in the overwhelming majority of SAL operations, the Bank's real objective in tranching SALs therefore requires clarification. Some staff maintained that, in light of virtually automatic tranche release, conditionality would be better linked to loan effectiveness. And indeed, proponents of such 'front-loaded" conditionality -- which appears to be increasingly common in adjustment lending -- maintain that at least it ensures compliance on a few important issues though it may not be as comprehensive an approach as a loan tranched over a longer peried. Indeed, some countries, such as Thailand, have themselves objected on principle to the use of tranches. It may be argued, on the other hand, that although tranches may or may not be a useful mechanism to enforce SAL conditionality, their utilization may serve a worthwhile institutional purpose -- to prolong the period of policy dialogue -- as effective a feature in some instances of SALs as the transfer of resources. In addition, tranching, according to some Bank staff, also provides a convenient mechanism for monitoring 'fuzzy' institutional components. The weak monitoring and supervision of SALs was another design concern frequently raised by Bank staf'- Institutional components of SALs have not been well monitored, partly because the activities that comprise them are not always specified in step-by-step action plans. This sometimes makes it difficult as well for countries to know what steps are necessary to carry out obligations or to know how to get from point A to B. And, as has been repeatedly underlined, supervision of institutional components. (often accorded A low priority in relation to other SAL elements) has been spotty, haphazard and indifferent. Indeed SAL supervision reports written by Bank staff uniformly contain only fleeting references to institutional components, and implementation difficulties that arise are discussed so superficially that it is often impossible for the unfamiliar reader to grasp the nature of the problem. It has frequently been asserted about adjustment lending in general that the number of conditions contained in most SALs is excessive. (For example, the Cote d'Ivoire's SAL III included 80 conditions.) This burden takes on institutional dimensions to the extent that already weak managerial capacity becomes strained, and effective implementation and adequate monitoring by the Bank is made virtually impossible. As a result, an informal process of priority-setting takes place. Thus, enforcement is sought for those selected conditions explicitly linked to tranche release, whereas a more lenient position is assumed for other conditions. A more realistic design would limit conditions to those few which could be reasonably enforced by Bank and country. - 22 - The number of conditions is one reflection of the degree of complexity in SAL design. The other persistent question is related to the number of institutional components which may be coherently combined to form a SAL institutional package. The much discussed Ochristmas tree^ problem still constrains effective Implementation of many SAL institutional programs and exacerbates the strain on already weak monitoring capacity. :ne nature of conditionality attached to institutional components contained in SALs varies significantly from one loan or country to another and oveor time. Many early SALs (and, later, the first group of SALs in Sub- Saharan Africa) tended to have *soft' conditionality tied to the execution of studies. Some staff believed that studies were an inappropriate subject for Bank conditionality. But for many African countries the deficit of adequate, reliable information to feed into the policy process wAs in itself a key developmental constraint, making the financing of studies reasonable. Whether SALs should have been the lending mechanism, however, needs to be assessed. A final point concerns analytic and sector work on basic political and administrative issues. Where SALs have foundered, some of the root causes have obviously been political or 'bureaucratic-political' in the Allisonian sense.51 (SAL 1I in Panama,. SAL II in Kenya, and SAL I in Senegal are cases in point.) Preparatory CESW -- emphasizing analysis of political-economy issues and, more simply, just laying out the public administrative landscape for policy reform -- is still uncommon in the Bank, although there are notable exceptions, such as Mauritania and Thailand. It is also worth pointing out that even with this basic groundwosc, unpredictable institutional or political shifts (e.g. ministerial changes or electoral transitions) may derail institutional reform. Moreover, the degree to which Bank staff are informed about and able to analyze these changes is too limited. This may be an area Where resident missions need an enhanced role, or where institutionalized, systematic briefings by out-of-Bank country experts might become standard Bank policy. IV. CONCLUSIONS The enormously wide range of public sector management components in structural adjustment and technical assistance loans since 1979 poses daunting obstacles to an assessment of their record. The findings of this review do not form the basis for definitive judgments, but they do suggest the following tentative conclusions: 5/ See, Graham Allison, The Essence of Decision: Explaining the Cuban Missile Crisis, (Boston: Little, Brown, 1971). Allison portrays the policy process as one in which the major actors are bureaucrats whose predominant political instinct is to defend the interests of their agencies qua agencies. - 23 - o Public sector management components of SALs progressed unevenly, and outcomes varied vith diverse political, administrative, and economic conditions. In general, SALs did not prescribe uniform solutions to institutional problems. Depending on the circumstances, SALs at times fostered the creation of new institutions or the reorgan4zation or sometimes the elimination of old ones. They often called for the streamlining and rationalization of procedures and commonly introduced new management systems and mechanisms. Many SALs sought improved coordination among agencies charged with carrying out reforms. For the most part, where such prescriptions did effect successful institutional change, however, it was often incremental and not necessarily sustainable in the absence of Bank support. O Despite the heterogeneity of institutional responses described above, a large number of SALs did promote the establishment of new institutions -- either with long- or short-term mandates, even when bureaucratic rationalization and streamlining were being prescribed as desirable policy goals. Although evidence about the impact of these institutions is not readily available in most cases, the dangers of institutional proliferation raise concerns about the advisability of this practice. Moreover, as perceptions of the proper role of the state in the economy change in many developing countries, a crucial adjustment task may increasingly become the dismantling rather than the creation of institutions. O Those public sector management reforms for which routinized methodologies and systems are introduced are more likely to be sustained over time than those for which no clear set of procedures was put in place. Public sector management reforms that from the outset can be linked to actionable steps agreed upon by Bank and borrower stand a better chance of being successfully implemented. o Calculations made by Bank staff about appropriate time allocations for public sector management reforms were almost always in error. The timing requirements of SALs impose serious, sometimes fatal, constraints on the effectiveness of public sector management reforms, although the catalytic benefits of SALs can often he significant. O Public management reforms promoted through SALs are more successful when supported by specific technical assistance projects. Without targeted institutional support, reforms are more likely to end as one-shot events subject to reversal or slippage. o Targeted technical assistance directly linked to SAL public sector management components work better than loosely structured, conventional technical assistance projects in supporting institutional reforms in SALs. But, in general, technical _ 24 - assistance has been sloppily delivered, expatriate-dominated, and often only minimally effective. Evldence of real transfer of skills and institutional capacity through technical assistance is not abundant. Nonetheless, SALs should be accompanied by tightly designed companion TALs to support implementation of public sector management reform components. These companion TALs may, at some later point, be succeeded by freestanding projects that may or may not be directly linked to policy reforms in SALS, however, as new areas of institutional reform are identified. Real progress on the institutional front may well depend, however, on the development of new lending instruments which are more suited to the time requirements of institutional reform than SALs, but do not have the enervating effect on borrowers of most traditional technical assistance. O One new instrument might take the form of a policy-based type sector adjustment loan specifically designed to address public sector management issues. Such an operation might contain conditionality for institutional reforms, but would be prepared independently of and perhaps on a slower track than a conventional structural adjustment loan, (although it would be closely coordinated with overall country strategy objectives, as would any SECAL). A public sector adjustment loan might also adopt a longer time horizon by several years than do SALs at present to complete project activities. Ideally, it would not focus exclusively on policy reforms, but would also involve medium-term efforts to improve the management of key public sector institutions through training, technical assistance andlor investment programs. o Implementation of public sector management reforms in SALs receive scant attention, either through monitoring, supervision or evaluation. When such assessments do take place, the analysis rarely goes beyond the finding that progress was either satisfactory or unsatisfactory. As a result, documentation about what worked, what failed, and why these results occurred is almost non-existent. The Bank's own internal system for keeping such reforms on track needs to be revised and more stringent information requirements should be imposed on supervisory and evaluative staff. The current system for monitoring the implementation of institutional components could be improved through the introduction of real bureaucratic incentives for .competent supervision. This might involve subjecting supervision reports to more public scrutiny within the Bank, and tightening up the analytic reporting framework so that key issues pertaining to the execution of public management reforms must be addressed routinely and systematically. In the same way that adherence to disbursement schedules is taken as an indicator of progress in the physical and financial implementation of projects, a clearer notion of where institutional components should be at each stage of SAL implementation should be developed, programmed and monitored. Mid-term assessments of institutional elements in SALs -25- might be carried out by disinterested evaluators with appropriate ezpertise within the Bank to head off institutional disasters during the life of the loan. Investigation of the application of such evaluation system. by other donors such as USAID and the UNDP mighpt be a useful starting point. O Preparatory country economic and sector work is crucial to public sector management reforms undertaken through SALs Such analyses can probe institutional issues that may be the focus of future reforms, identify actors and institutions that are critical to the reform process, or, at the very least, explain how relevant aspeots of the administrative and political systems work. Sector work may be undertaken independently or may be embodied in studies prepared as part of SALs, providing analysis for institutional components in succeeding SALs. At present, there are nio systematic requirements that such preparatory country economic and .sector work be performed, and no clear quality standards have been established for this type of analysis. Indeed, this work is largely carried out at the discretion of individual country divisions. Such analysis should become a routine, obligatory task in the preparation of any public sector management program undertaken in the context of structural adjustment. - 26 - ANNEX I SALs A TALs WITH INSTITUTIOUL CW*fNMrTS ................................................. (APesy As oP JUN 1i9u LN/C sOA LN/CN caorty PROECf NUIU1 APPRWAM AL UT (US") ................................................................. ......................................... olivia SAL I Is" 06/05/60 SO.0O bujuid SAL I 1m 05s/22/86 1.00 SAL It 1919 06/14/88 90.00 Efcnmic P.S. MgM. TAL 179 OS/14/87 ?.so NWirga Ar. TAL 18,7 12/10/8? 10.00 C.A.A. SAL I 13 09/09/8 14.00 SAL It 1916 06/09/' 40.00 TAL 11 1581 04/16/8 8.00 Chil SAL I 26 lolm22/s 250.00 SAL II 2767 11/20/86 250.00 SAL 11 2892 12/15/6? 250.00 TAL 1 2504 03/14/U 11.00 Congo SAL I 286 or/21/8? 70.00 tAL 1 22am 05S19/85 11.00 rAL 11 2m 09/16O86 4.00 P.1. Intt. Owt. 7AL, 286 07/21/8? 13.20 Costs Mfca SAL 1 2518 04/16/85 80.00 TAL 1 2519 04/16/85 3.50 Expot Oewt. TAL 2m 2 5osios/ 25.20 Coto dIvsire SAL I 2018 11/24/81 SO.00 SAL 11 2m2 07/05/5 250.70 SAL.-I 2711 06/03/86 250.00 TAL 1 205 11/24/81 16.00 Daoinica SAL I l81r 06/09/4? 3.00 Gabon SAL 1 293 04/21/8 50.00 Oebi - SAL I 1730 08/26/86 5.00 Ohaw SAL I 1Fm 04/14/87 34.00 Accra OiStr. Rabab. tAIL 154 03/21/81 22.00 Export Rehab. TAIL 1436 01/03104 17.10 hnstt. Suppwt TAL 1770 04/14/87 10.80 Guine SAL 1 1659 02/11/86 25.00 SAL It 1926 06/16/88 65.00 Econe mic ast. TAL 1 1559 03/14/5 9.50 Powr II Ins. & TAL 1595 05/1685 8.00 Guina Issu SAL I 17Y6 05/21/8? 10.00 TAL I 15r1 09/11/64 6.00 Guyun SAL t 1096 02/03/81 22.00 T?L It 1949 02/03/81 1.50 Jmica SAL 1 2105 03/23/182 76.20 SAL It 2315 06/14/85 60.20 SAL III 2476 11/20/84 55.00 TAL 1 2104 03/23/2 6.10 TAL II 250I 03/19/0 9.00 Water Suply & Sur. tAL 2412 05/22/64 9.00 PA* AdL Eeotr TAL 2423 05/22/4 4.50 Kama SAL I o099 03/25/80 SS.00 SAL I1 2t90 07101/82 130.90 Expowt PromotIon T*AL l145 06/12/0 4.SO Ar. TAL 1 r 07/01/12 6.00 Kore SAL I 20? 12/17t/61 250.00 SAL It 2a 11/08/85 300.00 alawi WSAL I 2028 06/25/81 '5.00 SAL 11 142? 12/20/85 5S.00 SAL 11 161 12/19/81 30.00 TAL 11 1418 12/20/85 1.50 "owl tonfc SAL I 1812 06/02/6? 1S.00 P.E. Rehab. TAL 56 03/26/ 16.40 Inetl. , Artisuet Govt. TAL 1 03/26/5 S.25 - 27 - SALs & TALs WITH INSTItUTIONAL CWMIEMTS ................................................. (APPROVED AS Of JUNE 1986) LN/C3t BOARD NC COUNTRY PROJECT NluI APPROAL AMOUNT (USIA) ........................"t......................................0,....................'/B'l..............is... 0..... Nagritiu SAL I W010 06/02/61 15.00 SAL tt 2361 12/08/83 40.00 TAL 1 23a" 12/08/83 5.00 Nepa SAL I 1769 03/24/87 50.00 Niger SAL I 1660 02/18/" 20.00 Ec. & Fin. Nagt. tarYt. TAL 1493 06/07/84 11.70 P.o. Inatt. 0evT. TAL 1836 07/O/67 5.50 Pakistan SAL I 216 06/01/82 140.00 TAL I 12% 06101/U2 7.00 P: an_ SAL 1 2357 11/15/83 60.20 SAL It 27" 122111U 100.00 TAL t 2358 11/15/83 5.00 Phitippines SAL 1 1903 09/16/60 200.00 SAL It 226 0"/26/3 302.30 S-a TOs SAL I 1815 06123/47 4.00 Seaml SAL I 108 12/18/00 60.00 SAL'It 1656 02/04/6 20.00 SAL ti1 1802 05/21/07 45.00 Ec. & fin. Pltw. TAL 1061 08/26/80 5.30 irrigation TAL 1632 10129/65 4.90 Parubtic It TAL 1390 07/07/83 11.00 Powr Eng. TAL s0o6 05/22/60 3.30 Urba Nagt. & Rehab. TAL 1458 04/17/84 305 Thailad SAL 1 2097 03/02/82 150.00 SAL tI 2256 03/31/83 175.50 togq SAL t 135 03/17/83 40.00 SAL It 159 05/30/65 27.80 SAL 1I 189 03/31/86 45.00 TAL 1 0930 06/12/t9 2.20 TAL 11 1270 06/24/82 3.50 TAL III 1600 05/30/15 6.20 Power Ing. TAL 1190 11/10/81 *.98 tunisia SAL 1 2962 06/16/86 150.00 turkey SAL I 1818 03/25/80 200.00 & suppl.ant 1915 11/13/80 75.00 SAL It 1987 05/12/81 300.00 SAL ItI 2158 05/27/82 304.50 SAL IV 2321 06/23/83 300.80 SAL V 2441 06/14/84 376.00 Uruguaw SAL 1 2836 06116/17 80.00 TAL for Pub. Sac. 1a"g. 2843 06/16/8? 1.00 Yugoslavia SAL t 2326 06/21/83 27n.00 Zaire SAL I 1831 06/25,/6 55.00 le. & Fin. Nagt. Institutions TAL 182 06/25/87 12.00 - 28 - ANNEX 2 SALS WITN INSTITUtZIL COPUNTS .......................................... ( of JUne 19U1 NE.!,m APPIROW PROPI ................................................................... AFRICA hanid I Ceosroen F*lv ~~~~~It Cha II I5 II C.A.E. I Ctis It Cow Uhuo I I Coto d'lvoire I Guinn u.s. It III Newitanla it Gabon Niger 11 ohm" Zalto iI Guinea I Kenya I 'I Ntwi I tat Nmwwiti I gr Iti Togs I It Zaire ASIA KOa I Saladea _egl philippines t tml.tam I [I thUul I II WNEM Pakistan cco Turkey I lv V Yugoslavia LAC Bolivia Costs fica It Chlile I Ucugs II Costa Rice Deminies Jmica I It III - 29 - tALS WITH INSTITUTIONAL C INTS .......................................... (as of June 1966) *lGION APPIOV P*OPOS .................................................................... Pmi I IX UfiJU .. . .. . . .. . . ......... o ...... *wf ..... ...... ....... ..... s ..... ..... ... ... .. .......... s .. . . . . .. .... .. .. . . ... . ...... .. s . ..... ..... . ... a&s I I La IG a¶a***w-B*e*w * vew @ o @v*was*e *oo o w *@ w ooo@ee sese @s o*@@* I'oSes j e eew Ive*h*@z** w5w o@ @v *Awbos 11S - z ¢ w + +e@wo I II * a I's I* 'S W)SIN +w l-|l ~~a@. '." I': a*. *'. ae. I'. I". a'. Z .................... ...... ... ...... .... ....... ..... . ..... .. - . ';'... ..... .... ; '';''.. .... .......... .... ......... ........... . ..... .......... ........ ..... ......... . ........... ..... -- . .... . .. ....^' V& i's i's 'IAM4 1 ll 63* I1* *-5 .1l .5s .'I .I a l v a I's i 15Is I' m I'I mI" I'" I 'I" u 3 I's l0s w o** i' s s' I'M *I*s SS IM a i a'w@w~hew * eDBw>s U@B S'S* )' *3eb@ U~~ I'M*P ae>@**@@si B9@*P@*@oe@ a,¢ .. . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .M .3 *3 *M . 3 *'. .. .. ... .'3 .3 I'mai's * au a' I's ..- I'm aa I's 0 a's 3., I'm a~~~~~~~~~~~~~~~~~~~~~~~~~~~~' I'm ' a~*' I O's O's~~~~~~e 3 .3 .3s .1 .3v .3 .w .3" *'3 .3m .3 .'3 It I'mI' I'm 's I' , l's '3' ls I' I 1 I'm I .14 .11 "a~~~~'3I' 1 S*3100 e WAIm A mm 'is mu re-a 'm 'm ~~~~~si mS*ui uiin smazam. mmmii masis. 153mM wi~~~~~~~~ai 'slI'ss IIVIW3W 115 'd Iai a'* 'Is .'' . Ml., IW. .'s Om .m . a.n .61 SI .NE It w- * l§" ls l" l" l" ll le eX §§F l@" l*" l*" ,- l 3 .................... ...... ... ...... . ...... ....... ....... ..... .... ..... ...... ..... ........... .. .......... .. .............. ... .... .......... ...... ........ .......... .. ........ . .......... ...... .... . ....;;.... ;;.....,.. ImV I'I'm I' aam- aaam sn.Bmemia I'm1usd jmWI'm I'maI'm- 3amI 4*hUV, ._........ ...... .. ...... . ....... . ....... . ..... . ...... ...... . ..... ...... .... ....... .... ......... ...... ... ..... ......... ..... ....... ..... ...... .... .... ..... ... ........ .. ......... .. ........... .............. ... .. .......... .. I's I'm I'am Im'mm s-nI.,m V Ar a Mvb .................... ......... ...... .... ... ...... ....................... .............. .............. ....... .......... .... ........ .... .. ...... ........ ......... . ..... .... sia t s.. . u ... . .u . |g08 _u oen _ ~~~~~~-lw t ' s _ W VW ls_ _o 04 ga- *_s im s MIM ' e sa > 1 us '31 U111111111111111 ls a~ 4w o 0101 GS l 111111 1111ox WMUi1I m1w d "AMUJ a SaMJIf£ia _a an Wa _ua I 13 lIam lt S Am "it D5U $&Ut1}_'#|0tm l Z1X . . . . . . . . . . .. . . . ...... . ..... ... .. . .. . . . ......... .............. ..... ....... I ......... ........... . ....... ............. .... ., ..... . .. 1si aButr wa mxl So so *e MSu¢ll £ X3NNV - oc - I'S I's m*l Il X; * ;I'm- I'm S £ *t I'-;ImaImI'm g'm I'm - .. I'm I'm I's ~~~~~~~~I's gI I's *I's I'mI163 II * I I*'s I's i's I*s 1 I's i's I i vaa , t 6; )A . . . .. . . . . . .. . . . . . . . . . . . . . . . .. . . . . . . e @ * o v wo vs ; ; o ** o ; ; * ;ws * e @ Bww D .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . .. . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . .. . . . . . . O's -a ..a -a O'" .,11~~~~~ ub .0 - .* ..* .-M -a .* . -I' e . a * *I' m 's la I's VIsA * a I* Iw I'm a s*;;e*a***o** * ; ; ;- Bou*w** w zw ; **¢+@* a@w* 11*h;w*00z --;-- so@@@;@*b*-* w ;e @ *e@@ *;;;- I., .1.1 a -*- l;muS oP@*sein;s* t eo*0w @ 6* *@*me Pr*u .Uw w *; ; - m * *m; ; *u @*u oow@ - I's I's 1.4~~~~~~~~~ ss * *w I q I., au aIn'*%* I31 U GIxI I'm I'm I., B.,,*a I'm I'm I's I'm a I ia" IIm Vs I 33MB0 *;.w inm.0 aiwn ~* IiM Sf~.~55~ia iu*5 .'ma bsmwa a,~~~uu ~~~~ DLUS 51413 IaaLUiI~~~~~I` *1 V seamI'sSI'mUI'm£21 am 'mama 53ui561.3m ~~~~~~~ ~~~~I'Is i's I'sAlibiI).I a6esb**ee;ere*ossew I'o*w*es a'*@s I'll "s . VSaS IUlUW WU *s 5 iAa * Iu t III*ICISI o@s**e*s;@ sese w@vos*s*f **~~as * oe w w oa'@* *S;# t-*4 oo- e@@- * Bess * *@*w a's@@ 01 VwSsw as;*- I'm e --- *--;-t Vs it- .............. . ......... ..... ..... .... ....... .... ...... .... ........... .......... ............. .............. ....... ......... ..... ....... ........ ....... .......... ............ . ....... wow*oooZit***sow~~~~~~~~~~~~~ w¢VQ * w I., e o I'as@ es*@ XVe* z@s V,* *P.,.e.. VW*e* Ie@@so IN-- -w-* ;;--;-*G;A- *1;vP- +w@.0> IOU _ _ ta ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Wi "en mm1* e bl3g Vt" a b ll swiwm in*"sAsees _F16~~~~~~~~~~~~~~~~~~~~~~~~~~~~~d Aw __ "Ie '3_ 3 IOW "1_ 1Ill ....................... .. .. ............. ......... ...... ... .... . ..... ...... .. ............... ......... ..... ....... ....... ...... ............ . ... ...... .. ..... 7... ....... .... .......... ..... ......... .... ,........... . _~~~~~~~~~~~~~~~~~$m n _ M AI a_ Ue tai | 1w1t t* _h191 m 11 1 o" ......... ...... __..... ......... ................................... ....... ..... ...................... ................................ ..................... .................... .............. ......... ................... ... ................................................... ..................... ................................................. .................. .. ...... ... .-au Asmara wouatis *AD saa§.bw walfwonalsof ..iw : : . :::: ::::: ::::: ..:: ::::: . ....:::::........ ... ..... ........... .......... .............. .............. ....... .......... ........ ........ ......... .........-:: ......... ........ .............. .:::: ls @ l^. lt~~~~~~~~$I Ila a's Wa 6f I'm lS. $'a Cis M eoowwo*--.... ...*-- ...... ...... .... .... 7*@@ T ;z* .es ..... .**o s ;. ~@*v@ .....e@~e ;*PW ...vot ....e ;; ...S .. .... ..... .. . .. ...... ....... . .......... ;. I0 l. 6- ... 1'W . ........ ...... . IaI .wz**@*zo* .vs*@ .....e .s* .@o .....w ....... ....... ........ .. ... .. ....................... ...... ..B@ @* ......... ....** ** .v**X*@ ...*@ ...... ....... ..eBow .......... I'sl' I 's I's ,N I'la OE *i~~~~~~~~~~~~~~s * I'm* I's I's - Va z ewww- oe@v;;e ev@w*> oe @****sw aev*@ I'm Iv@ oU;-v6wOo;n @Bw*wSe Xws | . .. ...... . ...... ...... ....... ....... ..... ...... ..!. ....... ....... ........ ....... ... ..... ....... . ..... ....... ....... .. .... ........ ........ ......... .. .s~~~~~~~~~~~~~V .s I' .' I' I's I, . s, Il IW.O ~~~~~~._ j... ..... ...... ..... .. .. .... ..'';;;' ...';;.';; .......... . .......... .... ....... ..;;. ....... ....... ...... ........ .. ........ . ....... .. ... .... ........... .......... . ............. ... ............. .. o-;e~~~~~~~~~~~~~~~~~~~~~l V& ;; " ; i ;; ; I ';m . . ... ... ..... ......@- .....*o* *O* ....... ..o ..e@ ...... ..... w .......... ..........¢ ............. ........... .......o ......... ....... ...... ..... ...... ...... ....... ............. e~~~~~~~~~~~~~~~~~~~' i's I's V& Il I's '"M 'n 'so. 10 . ."| l . . .. . ._ ......... ...... ...... ...... ....... ..... ...... ..... ........... .......... .............. .............. ...... ......... ...... ........ ......... ........... .......... ............. ............ -_---------- ---- ... ............... ............... ....... .... I "_ ..... .............. .............. ....... . ..... .."v... .......... .... 3w.... ..."a... ......... ...... ........ .......... ............. ...................................................... ... ......... ...... .. ... ....... ...... ..... ........... .......... .............. .............. ....... .. . . .... .......................................................................................................... ............. SOON~~~~~~~ NWAM asstw MMs Dw ml lae" nua Wam low 1. ~_ z1 g*.~~~~~~~' IsI& la I's Is . 3 , , ..... .~~.... .. ... ........... . .... ................... ........... .............. .... ........ ...... ......... ......... 1------~~~~~~~~~~ . ,* a a' -I's I'm s I's I*.I' Slu I' wmStmI s *~~~~~~~~ iv. mu mu mu~~~~~~~I' 's$0 's10 ' III I; sa gi e*I'm Is to~r Va V*~~~~~~~~~~~~~' Ila I' %Ima a em . * . . ........ .. .. ....... ..u .... ..... ....... I'm I s's m*u **, ~~~ ~ ~~~~~ ~~~ ~ ~~~~~a's a ua's a' me0 ~~~ us aw * I Slim~~~~~~~~~~~laIl I'm I'm I'mm of ammidam V. ... .......3M .......... .... . ..... ............... ... ..... ........ .... ........ ...... .. m....... ........... ummsaI'sai umi #aINa I' ..... . .... .. . ..a....... .......................... ..... .............. ..... ..... ................ 38.31 ufawS Uma~~~as A rnuU ima11ft mi um VS* SOI (AIm*a ' ' Al S*u a U IU U U VM I'm U III Vto 1-8 I'm I'm n~ ~~~~~~~~~~~~~ 3` I AAUI's ,*A 5*j S~~~~~~~~~~~~~~~~~~' A i 'm I A S*S I*I I', IS VI~~~~~~~~~~~~~~~~.. 's . .0.. .1 *U .1 .1 *U~~~~ ~~~~~~~ .U *u*.* ** II all. . '. I's I*lSM 'II'Mt* S's S's 5M I'M S's I 'U ~~~~~~~~'s 's I's UCO* It I'm a's I'm I'm ~ ~ ~ ~ " 3I "I., 'mII in . mm umm~~~~~~~~~~~~~~~'sI m I' I'm m Ill*j I'm I' I' 135 3 353SEU 13-0 I .. . ............ ... ... ... ....... ...... . ...... ..... ....... ......... .... ..... .................... ............ .........1 n i's. 1r I'sEgAI 11* *jgjI'm Veo Voo MZAI'mIali *g.. ........... ...... .....I..I .... ..... I ....... ........... ............ ............. ..... ... u.... ....... ...... ........ ...... ....... ......... ... mum WUmI I M .j n sBUN mvi ................ ............ ................. ......... ..... ........ ........ ...................... ..... ........3...... ...... ................ ....... .......... ....... 3 -W eV PU OWs 66 0 "". *.. .. aiOi403 OsW JO * |l| 4OU saodsI *e AOI -(*"l W; "mv"oqm of&@ ttlA am t 40,0s Os t : sai" a -aa.v ('3 S 3 a1*S1 QW01 *W§ AW ii &iemji* oh . -s1 p .8& md..d agism1 JO amid mooi 'meSas. 063a5lh0 Iindd so i3IdS 1pAiift s5 a'1IhIiVI .................... ....... ..... ..... ....... ....... ..... ...... ..... ........... ..........' ........... .............. ..... ......... ''.;;' ';;' ';''....... . ..... ...... . ....... ...... . ........ .. . .. I's 'I --- V& I's 55 I's I's 'Sisa 'lo -I; -S- ................. ....... ........ .. ...... . .....l ....... ....... ...... ...... .... ........ ... .......... .. ............ .. ..... ....... ....... ...... .... ... ..... ...... .. ......... ........... . ......... .. ........... .......... . ,,_.... .. U I'm i's I'm I'm I'm I'm I'm I'm I'll V I' ml I'm I'm I'm m v .................... ....... .. ....... .. .... .. ...... ....... . .... ...... ..... ...... ..... ........... ... .......... . .............. ... ..... ......... .. ....... . ......... .. ........ . .......... .......... .. ........... ... ......... ........ Va I~~~ ~~~~~~'m I'm I'U I'm I'm I II*Ual*SA ilA,ii ,'. ......... .! .......... .. ... ... . ..... ....... ....... ..... ..... .... ........... .. ........ . ..... ... .............. ... .... .......... .. ...... ........ .. ... ....."... .... ..... .. ' ..... 56Y 'S *as .'l .f*.l i'S ~I_ 'UlW W "1 "zml_ {@ffi"}} ewla S div u a :i a I .................... ......... ...... ... ... ...... ...... . ...... . ...... . .... ..... ...... .......... ... .......... .. ............. ... ....... ....... ... ... ...... . ....... .. ... .... ....... .. . .... ...... . .. .. .. .. .... . ... . .. .. ..... ....... .... is I's Is I'lls I' i '_ ' ms I. I a' ., anla m VW 'l 1'1 5 1613 id Miami Now *1 iim SVA marm WsIU ..iii; iA3WAUi 411 m1 is K1 3sa UNa -.. . . . . . . . . . . . . . . . . . . . . . . . . . ! . . . . . . mSC. -m eeS.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - 36 - Annex 4 . ..................................................................... tsmIRIN As Or Am 'ma IN1IMINIC cuw missmaa awn1 mm rnaImc M MI wA nsuj. Amiu i_W 1 m u amnunse m unu An". _ auf" ss.Cr.a njMam oas . . cmYL m cs a... amz iuumt not" sucas town . iNsM a man _ . eta MaMAs.c. ggfI S -uto eImM aS C. ML usm mam. IasAt mm. t1nS"l sas. sm mom llma Mm A. man m U.N.10. mIm M M. os " uc am M .inmam _Ms. _._ M_um. -. VA M. ___now amu_. ua_ Us... __rm _ _m _ -m"__ _ W a as 2 a to I W as as M . ma a 2a a 2 21 sib tam is IS I a a~ as a. a a a M-?*M 5 a asa so am as to at 't W 41 W_ Is2Ma a _ 2 a ._.. ...... ............... ........_ _. ._.._.. .. ._ .. .......... ... .............. .......... .. ..... ._ ______.... ______._.____.... ... ____ .... .... ____ .__ _____. .............. ......___ __ ________ MSL I a a s a a . a a a a a a s a a 9 * a a as aMUS20 a s I 0 2 0 a a a I a a a a a Ss sesg a P a a a a a a : a a a 4 a a a a a a a a a, ._. ____ ....... _..___.... ........._._ .......... ......... ....... _.__.___.__ _.___.._..... ...... .............._ .............. ._ ..... ....... ...___....... ....... .................. .. ..._aa a a a a as ' 1 'I'- ' -- ' - '-' 'I - - ' * - ', - ' ' - ' ' ' -' ' ' ,. . .......... E ...... ......- - - - -------- ------- ----- ---------- 1-- ------ ------------ - - - --- ------ ---- ---:1 ------- -1 ------ --:1 ----- -------- - - -1 a a ~~~as a a a g a a a aa as, a_ a W a a a a a a a a a a a a a a * 2 648 mamaI a St a. . aa a a a a a a So am a5 a1 ay a.I a _- a W*4 - -G a9 - ., - j -l . -2 .. -. * -U mma I 4 ? S *, _ 5 * _a a . ' 5 a sa a _ *2 55 a m - - - ~ 4 CC C - . - - .. PRE Working Paper Series Contact iua Autbor DAQ for paper WPS191 Openness, Outward Orientation, Sebastian Edwards June 1989 M. Ameal Trade Liberalization, and Economic 61466 Performance in Developing Countries WPS192 Inflation, Price Controls, and A'ay Chhibber April 1989 A. Bhalla Fiscal Adjustment in Zimbabwe Joaquin Cottani 37699 Reza Firuzabadi Michael Walton WPS193 Voluntary and Involuntary Peter Nunnenkamp June 1989 S. King-'.WVatson Lending: A Test of Major Hypotheses 31047 WPS1 94 Efficient Debt Reduction Jeffrey Sachs May 1989 S. King-Watson 31047 WPS195 How Has the Debt Crisis Harry Huizinga May 1989 S. Kinr-Watson Affected Commercial Banks? 31047 WPS196 A Review of Alternative Debt Eugene L Versluysen May 1989 E. Hadjigeor- Strategies galis 33729 WPS197 Differentiating Cyclical and Femando Clavijo May 1989 K. Cabana Long-Term Income Elasticities Riccardo Faini 37946 of Import Demand WPS198 Equity in Unequal Deductions: Chad Leechor May 1989 A. Bhalla Implications of Income Tax Rules Robert Warner 37699 in Ghana and Nigeria WPS199 Private Sector Assessment: A Samuel Paul May 1989 E. Madrona Pilot Exercise in Ghana 37489 WPS200 Women and Development: T. Paul Schultz April 1989 J. Klous Objectives, Frameworks, and Policy 33745 Interventions WPS201 How Much Fiscal Adjustment Is William R. Easterly June 1989 R. Luz Enough? The Case of Colombia 61588 WPS202 Financial Policies, Growth, and Alan H. Geib June 1989 M. Raggambi Efficiency 37475 WPS203 Optimal Commodity Taxes Under Nanak Kakwani June 1989 M. Felix Rationing Ranjan Ray 33724 WPS204 The Impact of Labor Costs on Luis A. Riveros May 1989 R. Luz Manufactured Exports in Developing 61588 Countries: An Econometric Analysis PRE Working Pager Series Contact i Aulhor Dfr WPS205 What Determines National Saving? Sang-Woo Nam May 1989 R. Luz A Case Study of Korea and the 61588 Philippines WPS206 The Effects of Single-Sex Valerie E. Lee May 1989 C. Cdstobal Schooling on Student Achieve- Marlaine E. Lockheed 33640 ment and Attitudes in Nigeria WPS207 Occupational Training Among Ana-Maria Arriagada May 1989 C. Cristobal Peruvian Men: Does it Make a 33640 Difference? WPS208 Effective Primary Level Science Marlaine E. Locxheed May 1989 C. Cristobal Teaching in the Philippines Josefina Fonacier 33640 Leonard J. Bianchi WPS209 Can the Industrial Countries International Economics August 1989 J. Queen Return to Rapid Growth? Department and International 33740 Economic Analysis and Prospects Division WPS210 Notes on Cash-Flow Taxation Roger H. Gordon June 1989 A. Bhalia 37699 WPS211 Coffee Pricing Policies in the Panos Varangis May 1989 D. Gustafson Dominican Republic 33714 WPS212 How Private Investment Reacts to Andres Solimano December 1989 E. Khine Changing Macroeconomic Conditions: 37469 The Case of Chile in the 1980s WPS213 Conditionality and Debt Relief Stijn Claessens June 1989 S. King-Watson Ishac Diwan 31047 WPS214 Adjustment and the Labor Market Peter R. Fallon June 1989 R. Luz Luis A. Riveros 61588 WPS215 Adjustment and Income Distribution: Francois Bourguignon May 1989 M. Ameal A Counterfactual Analysis William H. Branson 61466 Jaime de Melo WPS216 Price and Quality Effects of Jaime de Melo June 1989 M. Ameal VERS - Revisited: A Case Study L Alan Winters 61466 of Korean Footwear Exports WPS217 Public Sector Management Issues Barbara Nunberg March 1990 B. Mitchell in Structural Adjustment Lending 37644 WPS218 Public Finance, Trade, and Vittorio Corbo July 1989 A. Oropesa Development: The Chilean 39077 Experience