18993 January 1995 FOR OFFICIAL USE ONLY REVIEW OF PUBLIC EXPENDITURE WORK January 1995 The World Bank Office of the Vice President Development Economics t Table of Contents Page No. Acknowledgement Executive Summary ............ ............. i-iii I. INTRODUCTION. 1 1I. OTHER REVIEWS OF PE WORK . 2 Previous Reviews of PE Work. 2 De Melo (1988). 2 Kahnert (1987). 3 Ongoing Reviews of PE Work. 3 SPA Public Expenditure Working Group. 3 IMF Task Force on Public Expenditure. 3 Pradhan (1994). 4 Edwards et. al. (1994). 6 m. THE CURRENT REVIEW. 6 IV. MAJOR FINDINGS AND ISSUES .10 Process .................................... 11 Priorities of the Review ... ..................... 11 Timing ...................... 11 Staff Incentives. ..................................... 11 Bank-Fund Interaction ............................ 12 Political Economy ......................... 12 Analysis .12 Macroeconomic Framework .12 Public Sector Concepts and Definitions .13 Role of the State .13 Expenditure Priorities .13 Data ............................ 13 Management Capability and Incentives .13 Impact .14 Country Objectives .14 Bank Objectives .14 Private Sector .15 Donors .15 -2- V. PROPOSALS FOR FOLLOW-UP ACTION ....................... 15 Scope and Timing . .................................. 15 Analytical and Management Issues ......................... 16 Staffing Issues ..................................... 17 Evaluating Impact ................................... 17 Coordination with the IMP.............................. 17 BIBLIOGRAPHY APPENDICES A. Public Expenditure Reports B. Conceptual Analytical Framework C. Framework for Evaluating Public Expenditure Work D. Public Expenditure Reviews in Ghana Acknowledgement This report is part of an ongoing effort to respond to the need for improved Bank practices. It seeks to address some of the criticisms raised in the Wapenhans report by examining current practice and identifying areas for improvement in the general area of public expenditure work. It also seeks to address some areas of mutual interest to the Fund. Presently a Bank Working Group is pursuing the various issues raised here both with the Bank and with Fund counterparts. The present report was carried out by a team led by F. Desmond McCarthy and comprising Hjordis Bierman, Friedrich Kahnert, Satish Mannan and Sweder van Wijnbergen. Many others contributed to this effort. These included the task managers of the 21 reports considered, participants in an internal Bank review meeting and members of the Working Group on Bank Public Expenditure Work. Editorial support was provided by Milagros A. Divino. Executive Summary 1. This report is part of an ongoing effort to improve Bank practices. It examines current practice in the area of public expenditure (PE) work and identifies areas where improvements could be made. It also addresses some areas of mutual interest to the Fund. Presently a Bank working group is pursuing the various issues raised here both with Bank staff and Fund counterparts. 2. The Bank carries out public expenditure work in a variety of forms ranging from comprehensive reviews to short notes on specific issues. Over the period 1987 to 1993 there were 113 reports in this area. These were quite costly--on average they required about 100 staffweeks. For the present report, previous and ongoing work was reviewed and this was supplemented by a sample of 21 public expenditure reports completed over the period 1987-93 (plus a series on Ghana). Findings and Issues 3. The findings are divided into three general categories: process, analysis and impact. Process * Priorities. Priorities were not well stated ex ante in terms of scope, franework and audience. Much more interaction is needed with the client government especially at the time of identification of the objectives of the PE report. * Timing. Need to have reviews on a more regular basis taking into account aid coordination and Fund needs. * Staff Incentives. There is a need for a critical intellectual mass to ensure an adequate diagnosis. Most sectoral staff perceive the current incentive system as not conducive to placing any high priority on contributing to public expenditure work. * Bank-Fund Interaction. Relations often depended on personal relations (quite good in the Baltics). Institutional measures are needed to put this on a firmer basis. * Political Economy. Most reports failed to assess the likely political viability of policy proposals and the need to offer more palatable alternatives. Analysis * Macroeconomic Framework. While some reports do draw on the Fund's macro work this is not done in a consistent manner and in some instances the Fund's analysis remains a black box to Bank economists. - ii - * Public Sector Concepts, Definitions and Data. There is a wide variety of definitions and public sector concepts used. In addition the data suffers from timeliness, coverage and general quality. A major effort is needed to address these shortcomings probably in conjunction with current efforts by the Fund in this area. * Role of the State. Few reports addressed this. Even in areas where there is significant public investment it is important to consider the role of private investment and financing together with the appropriate legal, regulatory and supervisory framework. * Expenditure Policies. Two areas of analytical difficulty are: (a) analysis of whether sectoral shares are adequate to meet objectives and priorities; and (b) adequacy of recurrent expenditures. * Management Capability and Incentives. At a minimum this would include capability for budget planning to ensure adequate funding for all approved expenditures. Most reports did not address this area. Some did indicate obvious shortcomings but did not offer answers to suit the particular country situation. Impact Impact needs to be assessed at the country level and at different points in time. The present report was not able to assess this area adequately. There were some indications that the better ones in addressing country objectives were prepared in conjunction with country counterparts. It also appeared that a number of PE reports were quite useful in helping frame Bank lending strategy. Many reports failed to have any significant private sector involvement. The issue of the appropriate role for donors varied between countries. It seemed that where feasible it was desirable to include donor assistance in the overall budget envelope. Proposals for Follow-up Action 4. Scope and Timing. While large omnibus reports may have a role they are unlikely to be the primary instrument of PE work in the future. It would seem more appropriate to take a strategic view. This would include deciding on what type of review is appropriate for each country depending on country circumstances, role of the Bank, external factors and the present state of knowledge for that country. This would probably be a three-year program, discussed with the Government and with other key players such as the Fund and outlined in the CAS. 5. Analytical and Management Issues. A simple conceptual framework has been developed for further discussion. This would involve establishing: (a) boundaries of government; (b) budget envelope; (c) allocation of expenditures; (d) sustainability; and (e) implementation. 6. Bank-Fund Collaboration. The report made a number of recommendations in this area. The following actions have been taken to date: (a) Fund and Bank have established task forces to address many of the issues identified. Both institutions have shared their reports and pilot cases have been identified, both regional and country specific, to discuss priorities for and, issues in public expenditure reviews. The first regional meetings (AFR and ECA) have taken place, and the county pilot studies (Cote d'Ivoire, Lithuania, and Zambia) are underway. (b) A joint guidance note has been drafted for management to reemphasize the importance of public expenditure work and collaboration with the Fund in this area. (c) Joint Bank-Fund research efforts are underway to address the analytical issues. I. INTRODUCTION 1. Public expenditure analysis plays a central role in the Bank's work. It serves many objectives: supporting the policy dialogue, identifying areas for capacity building, and providing a framework for the Bank's lending program and aid coordination. However, there is also a perception that much of this work may be too expensive, especially at a time of substantial cuts in ESW resources, and of uneven quality. The objective of this paper is to review recent experience with public expenditure work in the Bank and to identify areas where present practice could be improved. 2. The Bank carries out public expenditure (PE) work in a variety of forms: from comprehensive reviews to short notes on specific issues. As shown in Appendix A, the Bank devotes considerable effort to PE work. Over the period 1987 to 1993 there were 113 reports in this area. These reports were quite costly--on average they required about 100 staffweeks. Table 1 indicates some major changes over time and between Regions. Most reports were produced in 1990 while the Africa Region was the most dominant. Presumably this was related to the major impetus from SPA-style programs at that time. Table 1 also indicates the IBRD and IDA commitments over the period. A quick perusal of the individual country listings in Appendix A shows some notable features. There does not appear to be any regular pattern in the frequency of these reports, although presently some countries are moving to a more regular schedule to meet demands such as to provide background papers for CG meetings. What is perhaps somewhat surprising is how few reports were done on a number of major Bank borrowers (e.g., India until recently and China). 3. In the present review, a two-track approach was adopted. First, previous and ongoing work was reviewed. This was then supplemented by a sample of 21 PE reports carried out over the period 1987-1993. A series of PE reports on Ghana over the period 1983-94 was also reviewed, to look at the cumulative impact of repeat exercises. 4. From the combination of these two tracks a set of findings was developed. These were grouped in three broadly-defined areas: process, analysis and impact. Finally a set of proposals for follow-up action are given. These include: (a) a proposal to move towards a more strategic approach, integrated with the CAS process, for public sector economic work; (b) proposed work with the Fund on a common analytical framework together with a uniform set of concepts and definitions together with an improved database; and (c) proposed follow-up on the impact of PE work, first by OED and then by regular, but selective, peer-group evaluations. A joint working group has been established with the Fund, to address common concerns on public expenditure issues and find ways to improve Bank/Fund collaboration in this area. Bank members of the working group will also pursue public expenditure issues of more direct concern to the Bank. - 2 - Table 1: NUMBER OF BANK REPORTS ON PUBLIC EXPENDITURE ISSUES a! Year EAP SAS LAC AFR MNA ECA Other Total 1987 2 0 4 4 3 1 0 14 1988 1 0 2 5 0 0 0 8 1989 0 1 7 8 1 1 0 18 1990 0 3 3 10 4 2 1 23 1991 1 0 4 7 2 1 0 15 1992 3 2 4 7 1 1 0 18 1993 2 4 3 6 2 0 0 17 Total 9 10 27 47 13 6 1 113 IBRD and IDA Commitments (US$ billion) 1987 3,71 4.29 5.21 1.82 1.76 1.42 0.00 18.21 1988 3.85 3.74 4.96 4.05 1.13 1.41 0.00 19.14 1989 4.81 4.77 5.96 3.32 1.98 1.49 0.00 22.33 1990 4.11 3.63 6.71 3.68 0.87 2.47 0.00 21.48 1991 5.25 3.66 5.17 3.93 2.32 3.85 0.00 24.18 1992 5.20 3.31 5.86 3.29 1.06 3.15 0.00 21.86 1993 4.80 2.51 5.30 2.18 2.32 3.33 0.00 20.44 Total 31.73 25.91 39.17 22.28 11.45 17.11 0.00 147.64 a. Details are given in Appendix A. II. OTIHER REVIEWS OF PE WORK Previous Reviews of PE Work 5. De Melo (1988). The last major report on Bank public expenditure reviews (including public investment reviews) was by de Melo in 1988. The basic information for that report was obtained from a sample of 22 PE reviews carried out during the period 1985/87. It offered four broad categories of recommendations: (a) Setting and Achieving Objectives. The primary goal should be giving objective advice to the government on the level and composition of expenditures; Bank staff should consider an explicit link between the PE review and the country program. (b) Coverage of Issues. Core report should include macroeconomic framework, review of the public investment program and recurrent expenditures; coverage should be selective and tailormade to country circumstances. - 3 - (c) Resource Use. The Bank should exercise more restraint in committing resources to PE reviews and use more upstream work to preclude the need for mounting large PE missions. (d) Organization and Management Support. Each Country Operations Division Chief should designate a staffmember to become familiar with the types of PE reviews which have been undertaken and the Bank's experience with them. The de Melo report also discussed the analytical approaches in use and concluded that the macro framework provided was often incomplete. 6. Kahnert (1987). A report by Kahnert (1987) on public expenditure analysis focussed on the EMENA Region. He paid particular attention to the need for a broader perspective than the virtual exclusive focus, at that time, on cutting public fixed investment. He indicated the need for internalizing PE work in Bank work and putting it on a much more routine basis. He also highlighted the need for broadening PE work over time to include all recurrent and nonrecurrent expenditures of the various tiers of government and relevant policy, institutional and procedural issues. Ongoing Reviews of PE Work 7. Presently there are a number of ongoing studies on public expenditures. These include the following. 8. SPA Public Expenditure Working Group.' The focus of this group is- on SPA countries. Their study, which resulted from a meeting in January 1994 managed by the European Commission, focuses on the lessons of past experience with PE work. They also plan to carry out ten country case studies to examine the impact of PE work at the country level. The primary focus of this group is on impact. The interim report raised three principal issues related to: (a) how to internalize the recommendations of PE reviews more effectively; (b) how to make the PE review more of an operation owned by the country; and (c) how to improve collaboration among donors. The case studies are expected to focus on budget management, poverty and gender issues. These are expected to be completed by April 1995 and help prepare guidelines for the future conduct of PE work in SPA countries. 9. IMF Task Force on Public Expenditure. This group was established in March 1994 to see how the Fund could make more effective use of PE work prepared by the World Bank in the Fund's policy dialogue with country authorities. The report of that Task Force& has made a number of recommendations for improvements in Bank/Fund collaboration on public expenditure issues. The report expresses a desire for better timing of reports and improved data, together with closer interaction with Bank colleagues on policy issues. Specific proposals See SPA Public Expenditure Working Group, Meeting on October 24, 1994, Chairman's Report. 2 Report of the Task Force on Public Expenditure Policy, December 1994. - 4 - include: (a) early agreement with Bank staff on priority countries and issues; (b) exchange of short notes from Bank and Fund staff on issues of mutual interest; and (c) more regular interaction in missions and review meetings. 10. Pradhan (1994). Recently within the Bank, Pradhan (1994) has conducted a study on public expenditures. His broad objective is "to present an analytical methodology and data framework for evaluating broad allocations of public expenditure." As background to his work he carried out a detailed review of 12 recent PE reviews. These are listed in Table 2 together with a summary of key areas and issues. In his review he identified the following problems: (a) most PE reviews focus primarily on inputs (i.e., expenditures) rather than outcomes (leave alone their valuation); (b) the role of government versus the private sector in service provision is rarely discussed; (c) the unit of analysis is usually inappropriate, with few PE reviews moving below the aggregate sectoral level; (d) judgment appears to be the dominant methodology employed; (e) there is inadequate emphasis on follow-up to ensure that the PE findings are incorporated into country spending decisions and Bank/donor lending strategies; and (f) few PE reviews focus on institutionalizing a more systematic decisionmaking process and improved data collection within client countries. 11. He also noted that many PE reviews do not even present the basic data on functional and economic composition, major programs, private sector provision, etc. that would be useful, if not indispensable, for decisionmaking. However, there are some instances of better-practice PE reviews (e.g., Tanzania, Indonesia, Uganda) which are identified in the paper. 12. In a follow-up work program, PRDPE staff are working with Regional staff on a number of "model" PE reviews as well as undertaking further research focussed on institutional issues related to public sector management. HRO staff are also working jointly with Fund staff to develop a manual to provide hands on guidance for practitioners, especially in the general area of budget management.3 Holmes, Malcolm (1994). - 5 - Table 2: COVERAGE OF PE ISSlTES IN PRADHAN'S SAMPLE I. Overview A. Macro Framework 58 B. Discussion of Role of Government 8 C. Tax and Revenue 42 D. Total Expenditure - Current 67 - Capital 67 E. Civil Service 42 F. O&M 8 G. Public Investment and Finance 50 H. Budgeting & Administration 75 I. Data Collection 8 ]I. Sectors A. Health 100 B. Education 100 C. Transport 83 Intersectoral Allocation 0 Note: Information gathered from Bangladesh Public Expenditure Review Public Resource Management During the Fourth Five-Year Plan, FY91-95, 3/89, #7545; Bolivia Public Sector Expenditure Review with a Special Emphasis on the Social Sectors, 9/89, #7746; Cote d'Ivoire Public Sector Expenditure Review, 10/15/87, #6980; Ecuador Public Expenditure Review: Changing the Role of the State, 8/93, #10541; Ghana Public Expenditure Review (90-92), 6/90, #8616; Kenya Public Expenditure Issues, 4/89, #7508; Malawi Public Expenditure Review, 4/90, #7281; Mozambique Second Public Expenditure Review, 12/92, #11524; Nigeria Public Expenditure Review (1986-90), 10/85, #5920; Paraguay Public Expenditure Review: The Social Sectors, 6/92, #10193; Philippines Public Expenditure and Their Finance, 9125/84, #4919; Sri Lanka Public Expenditure Review, 9/93, #12337. Source: Pradhan (1994). - 6 - 13. Edwards et. al. (1994). Recently there has been some further thinking on how best to carry out ESW, and within this, PE work in particular. Some of this "new thinking" is perhaps best reflected in a recent report of a Task Force headed by Edwards, "ESW Reform in LAC" (1994). The report finds a need to reform ESW work in LAC, largely because of changing external conditions in the global economy, and the concomitant changes in the way the Bank should approach its analytical work in the Region. The proposed outcome would result in a major redistribution of ESW work for the period FY96-98 compared with FY95-97. Traditional areas such as CEMs and PE reviews would be reduced while major increases are envisaged for ESW on environmental and urban issues, infrastructure, private sector development and public sector management. Formal reports would be reduced in favor of shorter more-focussed policy notes, and the Bank's work would be more closely coordinated with that of partner agencies (e.g., IMF, EDB, CEPAL). While this report applies specially to LAC, similar changes have been proposed by other Regions in the recent round of Prospects Summaries.4 In. THE CURRENT REVIEW 14. To provide a baseline for the present work, a sample of 21 reports were chosen (see Table 3).5 It is noted, in Table 3, that they were chosen to try and represent most areas of the Bank, but also to include a variety of report types, ranging from CEMs with a major PE theme, to comprehensive public expenditure reviews, to public investment reviews and also some that limited their efforts to particular policy or sector areas. Almost all included contributions from consultants; typically this was about 25 percent of the total. It is difficult to ascertain the contribution of local staff. Based on currently available data, local staff were only used to any extent in three instances: Bangladesh, India and Indonesia. 15. Each of these reports was first reviewed. This was followed by an interview with the Task Managers (or the nearest available). A preliminary assessment is given in Table 4. The assessment should not be viewed in a normative way but rather as providing the basis for identifying a set of issues. In particular it was very difficult to assess impact. A number of those interviewed felt that Bank and country objectives coincide. The whole question of impact needs to be addressed separately. The involvement of private sector was weak in most reports. Only recently has the nonexpenditure role of government come to the fore. However it is somewhat surprising that some reports made little effort to elicit the views of the private sector. For details, see Memorandum from Mark Baird on Economic and Sector Work (January 18, 1995). This was extended to include a series of four more reports on Ghana done over the period 1983-94. - 7 - Table 3a: SAMPLE OF PE REPORTS a! Gap Country Name Cover Mission Date Report Date (Months) Report No. Argentina Pub. Fin. Rev Gray 11/91 2/93 15 10827 Bangladesh Pub. Exp. Rev. Gray 11/87 3/89 16 7545 Brazil Pub. Sector Inv. Rev. Green 11/85 1/87 14 6600 Colombia Pub. Sector Exp. Rev. Gray 10/88 1/91 27 7891 Egypt Pub. Sec. Inv. Rev. Gray 2/92 10/93 20 11064 Estonia Pub. Expend. Rev. Gray 10/93 7/94 9 12787 Ghana Pub. Exp. Rev. Green 11/89 6/90 7 8616 Hungary Ref. of Soc. Pol. & Gray 4/90 7/91 15 9349 Exp. India Pub. Exp. Rev. Yellow 12/92 12/93 12 12570 Indonesia Pub. Exp. Prices and Gray 1/92 8/93 19 11293 Poor . Jamaica Pub. Sec. Inv. Prog. Gray 11/91 5/92 6 10297 Kenya Re-Inv. in Stab. and Gray 7/91 1/92 6 9998 Growth. Latvia Pub. Exp. Rev. Gray 11/93 7/94 8 12973 Lithuania Pub. Exp. Rev. Grav 10/93 7/94 9 12792 Malawi Pub. Exp. Rev. Gray 11/87 4/90 29 7281 Mexico Mex. in Trans. Gray 3/90 5/91 14 8770 Towards .. Morocco Issues & Pros. in Pub. Gray 5/91 6/92 13 10157 Sec. _ _ _ _ _ _ _ _ _ _ _ _ Nepal Fis. Rest. & Pub. Res. Gray 2/93 3/94 13 12281 Met... Philippines Cer: Pub. Sec. Res. Gray 5/91 2/92 9 10056 Mob... I Tanzania Pub. Exp. Rev. Gray 10/87 5/89 19 7559 Zambia Pub. Exp. Rev. G 9/91 12/92 15 11420 a. In addition to the Ghana (1990) report four additional reports on Ghana were considered in assessing the merits of a series of reports in a particular country. - 8 - Table 3b: TRS STAFFWEEKS FOR PE REPORTS Local HL HL Consultant and Con. Total Country -FY rD Name Staffweeks Staffweeks Staffweeks Staffweeks Argentina 93 6ARGER044 Pub. Fin. Review 72.4 91.8 .0 164.3 Bangladesh 89 8BANERO25 Pub. ExP. 146.9 5.Q 37.3 189.2 Brazil 87 6BRAER016 1986 Pub. Sec Inv. 107.7 22.6 .0 130.3 Colombia 90 6CLMER034 Pub. Sec. ExP. Rev-C 61.7 10.1 .0 71.8 Eg,,ypt 94 SF.GTER027 Pub. Sec. Inv. Rev. 58.1 45.4 .0 103.6 Estonia 94 7ESTER003 Pub. ExD. Rev. 40.5 6.3 .0 46.9 Ghana 90 . 3GHAER29 Pub. Exp. Rev. 52.3 1.0 .0 53.4 Hungarv 92 7HUNER023 Social Pol. & Distr. 59.0 39.0 .0 98.0 India 93 8INDSR196 Agric. Public Expend 45.5 .0 46.0 91.5 8INDSR197 Infra-Pub. ExD. Rev. 52.4 33.2 32.9 118.5 Indonesia 93 4INSER038 Pub. ExP. for Povertv 89.9 11.0 22.0 122.9 Jamaica 92 6JAMER027 PSIP 25.4 24.3 .0 49.7 Kenva 92 3KENER040 CEM (PSM) 58.1 15.4 .0 73.5 Latvia 94 7LATER004 Pub. Exp. Rev. 40.7 16.5 .0 57,2 Lithuania .94 7LTHERQ04 Pub. ExP. Rev. 38.4 20.2 .0 58.7 Malawi 89 3MALER014 Pub. ExP. Rev. 71.4 17.4 .0 88.8 Mexico 91 6MXCER027 CEM 113.2 52.1 .0 165.3 Morocco 92 5MYCER020 Pub. Sector Study 52.4 33.9 .0 86.3 Nepal 94 8NEPER016 CEM 109.0 . I .0 109.1 Philippines 92 4PHI-ERO29 CEM 50.8 .0 .0 50.8 Tanzania 89 3TANERD07 *Rev. of Pub. Exp. 56.5 43.8 .0 100.3 Zambia 93 3ZAMER029 PER (FY92-93) 40.4 5.8 .0 46.2 Total 1442.9 495.1 138.1 2076.1 Source: MIS Database. -9- Table 4: SUMMARY ASSESSMENT OF PUBLIC EXPENDITURE REPORTS No. % No. of Judged Judged Reports As As Evaluated Good Good Country Codes for Reports Judged as Good a/ A. Process and Organization 1. Timing (scheduled on regular basis) 21 4 19.0 CLM, EGT, GHA, JAM 2. Link to sectoral work 21 12 57.1 EGT, GHA IND, JAM, LAT, LIT, MAL, MXC, MYC, NEP, TAN, ZA;W 3. Broad Political Economy (has 21 4 19.0 BAN, INS, LAT, NEP reviewed consid. who gains/loses?) 4. Priorities of the rev. considering 21 10 47.6 EST, HUN, INS, JAM, LAT, LIT, MAL, NEP, PHI, ZAM country priorities B. Definitions and Coverage 1. Clarity of Definitions Pub. sector concepts, expend. & fin. 20 5 25.0 INS, KEN, LAT, LIT, MAL def. 2. Objectives Stated 21 12 57.1 BAN EGT EST, IND, JAM, KEN, LIT, MAL, MYC, NEP, TAN:Zj1W Met in report 21 13 61.9 ARG, BAN, EST, HUN, IND, INS, KEN, LAT, LIT, MAL, MYC, NEP, TAN Outside the report 21 9 42.9 ARG, EGT, EST, GHA, IND, INS, KEN, LAT, LIT 3. Comprehensiveness Macroanal ysis 2 90 AG A,ML A Fiscal implications and analysis 21 7 33.3 ARG, BAN, EST, KEN, LIT, MAL, TAN Expenditure: cross-sectoral 21 7 33.3 ARG- EST, LAT, LIT, MAL NEP, TAN Expenditure: sectoral analysis 21 8 38.1 BAN, IND, LAT, LIT, MAL, NEP, TAN, ZAM PSIP or core PSIP 21 7 33.3 EST, GHA, JAM, LIT, MAL, NEP, TAN 4. Issues of management of 21 12 57.1 ARG, BAN, CLM, EST, GHA, IND, LAT, LIT, MXC, PHI, olannin, bud2et and expenditure TAN, ZAM C. Analysis 1. Background Inside the report 21 12 57.1 ARG, BAN, EGT, EST, IND, INS, KEN, LAT, LIT, MAL, TAN,' ZAM Outside the report 20 11 55.0 ARG, BAN, EGT, EST, GHA, IND, INS, KEN, LAT, TAN, ZAM 2. Framework Strength and relevance of 21 6 28.6 ARG, BAN, KEN, MXC, MYC, TAN macroanalysis 3. General fiscal analysis 21 8 38.1 ARG, BAN, EST, INS, KEN, LAT, MAL, TAN 4. Government or public expenditure Cross sectoral analysis 21 9 42.9 ARG, BAN, EST, LAT, LIT, MAL, NEP, TAN, ZAM Sectoral analysis 21 8 38.1 BAN, IND, LAT, LIT, MAL, NEP, TAN, ZAM PSIP 21 10 47.6 BAN, EST, EGT, GHA, LAT, LIT, MAL, MXC, NEP, TAN Integ. of PSIP in fiscal analysis 21 6 28.6 BAN, GHA, LIT, MAL, NEP, TAN 5. Management of Planning, Budgeting 21 13 61.9 ARG, BAN, CLM, EST, GHA, JAM, LAT, LIT, MXC, MYC, & Exp. NEP, PHL, ZAM 6. Data Appropriateness of data used 21 12 57.1 ARG, BAN, CLM, EST, HUN, IND, INS, JAM, KEN, LAT, MYC, ZAM Restrict.imposed by data available 21 7 33.3 ARG, BAN, CLM, IND, INS, KEN, MYC - 10 - No. % No. of Judged Judged Reports As As Evaluated Good Good Country Codes for Reports Judged as Good a/ D. Impact 1. Impact Furthering Bank objectives 21 14 66.7 ARG, BAN, EST, GHA, IND, INS, LAT, LIT, MAL, MYC, NEP, PHL, TAN, ZAM Furthering Country objectives 20 11 55.0 ARG, EST, GHA, INS, KEN, LAT, LIT, MXC, NEP, PHL, ZAM Influencing donor objectives 19 11 57.9 ARG, BAN, EST, JAM, KEN, LAT, LIT, MYC, NEP, PHL, ZAM 2. Broader involvement in PER/PIR Private sector: In review preparation 21 2 9.5 BAN, MXC Outside the report 21 5 23.8 ARG, BAN, IND, MXC, ZAM Donors: In review preparation 21 1 4.8 BAN At or after completion 21 12 57.1 ARG, BAN, EGT, EST, IND, JAM, KEN, LAT, LIT, NEP, PHL, ZAM 3. Implementation of general recommendations 21 5 23.8 ARG, JAM, LAT, LIT, ZAM of specific recommendations 20 4 20.0 JAM, LAT, NEP, PHL a. List of Country Codes: ARG - Argentina; BAN - Bangladesh; BRA - Brazil; CLM - Colombia; EGT - Egypt; EST - Estonia; GHA - Ghana; HUN - Hungary; IND - India; INS - Indonesia; JAM - Jamaica; KEN - Kenya; LAT - Latvia; LIT - Lithuania; MAL - Malawi; MXC - Mexico; MYC - Morocco; NEP - Nepal; PHI - Philippines; TAN - Tanzania; ZAM - Zambia Also surprising was the low level of political economic analysis. Even in instances where broad policy initiatives were proposed one finds little reflection on who the winners and losers might be. Again a number of issues arose that had been identified in previous reviews of PE work that seem to continue posing difficulties. These include the need for clearly stating objectives, delineating the bounds of the work and improving data. There was some improvement in incorporating sectoral work but in a significant number of reports it was felt that Bank incentives still mitigate against a stronger contribution here. Another broad area that seemed to have been largely ignored was the overall institutional setting together with an assessment of what changes were needed to be more effective. IV. MAJOR FINDINGS AND ISSUES 16. The reviews of PE work and follow-up interviews sought to address the following questions. What are the present objectives of PE work? To what extent are the objectives being accomplished? How are staff going about accomplishing the objectives? Rather surprisingly, the interview process revealed that objectives were often not elaborated at the outset. In a number of instances the objectives, as stated in the reports, were often modified (or even introduced) after the mission. Most of the reports fail to clearly state the review objectives. - 11 - Yet those interviewed (Bank staff) did not seem too concerned about this. Little effort is made to define and motivate the boundaries of the review properly. Three broad categories were considered in assessing these reviews: process, analysis and impact. Generally this review found that on process there was significant progress, while analysis still remained problematic. The report was not able to assess impact beyond some preliminary observations. Process 17. Priorities of the Work. These were generally not well stated ex ante in terms of the envisaged scope, framework and audience. This is similar to what de Melo reports. There usually should be a number of objectives for Bank PE work. These typically reflect areas of actual or potential Bank lending in addition to assessing the overall macro situation and its relation to public expenditure. There is also a broader need as an input to overall Bank portfolio management. Here an incisive review of public expenditure could provide a useful input into overall risk management. The reviews pointed to the need for incorporating country priorities into the overall design. Much more interaction is needed with the client government especially at the time of identification of the objectives of the PE report. This may range from a need to generate an overall strategy to helping with some specific problem areas.6 The more successful reviews (Indonesia and the Baltics) had good working relations with the authorities throughout the process and the analysis. Most reviews, especially the less successful ones, tended to consider the report itself as a major output and often did not place as much emphasis on interaction with the authorities. 18. Timing. This is one of the concerns raised by the IMF. There appears to be little uniformity in this across the Bank. If a country is in crisis then one component of -the Bank response may involve a major PE review. In principle this provides the background for various policy discussions and subsequent lending operations. Otherwise the PE review tends to be a more routine operation, undertaken as a standard tool for Bank staff, or in response to a request for help from a country focusing either on general or some specific problems. In a few countries PE reviews are currently done on a regular basis, perhaps every other year, to provide background for CG meetings for instance. However, in most instances, the rationale is rather obscure, sometimes there "happened" to be money left over in the budget or it was thought appropriate to provide an opportunity for a junior staff member to be a task manager. In other instances a public expenditure review was used to provide some degree of rationale for a lending program that was essentially agreed ex ante. 19. Staff Incentives. The staffing requires a critical intellectual mass to ensure the capacity to make an adequate diagnosis. There are two areas of concern: the quality of the mission leader and the role of sectoral staff. This varies a great deal between PE reports. A necessary, but not sufficient, requirement to do a good job is that the leader be an experienced staff 6 In some instances this may involve a Ministry of Finance, for example, which would like to use the Bank to gain access to areas where they are not adequately informed. Sometimes there are important companies (oil, power, railways) which tend to keep their operations relatively secret although they are public or semi-public companies. This use of the Bank for internal policy purposes is probably not a good idea. - 12 - member, preferably with a good grasp of both macro and sectoral issues and with some background in policy implementation. Certainly not an inexperienced, albeit bright staffmember. The limited role played by sectors was also identified as a problem by de Melo. Some progress has been made here. This requires a review of previous sector work together with a grasp of overall sectoral strategy. This would include an understanding of which sectors the country has been relatively successful in and where are the problem sectors and why. This requires participation from the sectoral staff. In spite of rhetoric to the contrary, most sectoral staff perceive the current incentive system as not conducive to placing any high priority on contributing to public expenditure ESW. However, when sectoral staff were able to bring their expertise to bear it had a favorable impact on the outcome. One approach (used in the Baltics) was to have the Department Director convey the importance of the work to the sector chiefs at an early stage. If this is indeed followed up it may be useful. However, it is evident that in most Regions of the Bank, further change is needed to encourage sector staff to gain an understanding of sectoral priorities and strategies and frame lending in the context of overall sectoral priorities. This may involve some retraining of sector staff as well as breaking the perceived link between career prospects and lending. 20. Bank-Fund Interaction. It is notable that the reviews of the Baltics also benefitted from good working relations with the IMF. Presently, Bank-Fund cooperation in PE work often seems to reflect a strong element of personal relations. This could be put on firmer footing by institutional support measures. The Bank-Fund working group (discussed later in this report) is expected to address this issue. 21. Political Economy. Some reports sought to integrate the views of diverse groups: opposition, private sector, unions, other creditors. Most were weak here and failed-to assess the likely political viability of policy proposals and the need to suggest more palatable alternatives; in an extreme case the review was done against the wishes of the government. Analysis 22. Macroeconomic Framework. The findings were similar to those of de Melo and Pradhan. The framework of analysis needs to be adequate to answer the questions posed within the review process. Most reports treat the macroaspects of the reviews in a very cursory fashion. Because of the superficial treatment, one cannot assess the macroeconomic framework. This gives rise to two concerns: * While some PE reports do draw on the Fund's macro work (e.g., fiscal deficits in the Baltics), this is not done in a consistent manner and, for many countries, the Fund's analysis remains a "black box" to Bank economists. * While the RMSM framework can be extended to incorporate the public sector, there are few policy links to the rest of the economy (e.g., impact of export taxes, used to close the fiscal accounts on export performance). It is easily manipulated. As a result the RMSM model is often used to justify macroprojections which may not be consistent with the fiscal projections. - 13 - 23. Public Sector Concepts and Definitions. There is a wide variety of definitions and public sector concepts used. Most reports did not indicate what constituted the public sector and where its boundaries should be. Some PE reports focus on the Central Government, while others attempt to incorporate state/local levels of government and public enterprises. In some reports, there is also ambiguity over which concept of the deficit is appropriate, while in others there is little uniformity on what constitutes capital and current expenditures. The precise definitions and concepts of public sector used are not reasoned and defended based on the objective of the review and the issues to be analyzed. 24. Role of the State. Few PE reports address this issue. However, it is crucial for establishing spending priorities. In some areas (e.g., industry), there may be little role for state action beyond maintaining sound macro and structural policies. In others (e.g., telecommunications), it may be more important to think about privatization and the regulatory framework for private sector activity, than to assess new public investments. Even in areas where public investment will continue to be important (e.g., primary education), it is important to consider the appropriate role of private investment and financing. 25. Expenditure Priorities. Two general areas of analytical difficulty are: (a) analysis of whether sectoral shares are adequate to meet the objectives and priorities, and in this context the question of intersectoral choices; (b) analysis of the adequacy of recurrent expenditure, given past investment, and the integration of recurrent cost as an integral part of the determination of the composition and the size of the PSIP. Intersectoral allocations are inherently difficult to evaluate. However, at a minimum, significant sectoral shifts may emerge from consideration of the role of the state (see above) and elimination of obvious white elephants. At one extreme some suggest the use of rules of thumb. The IMF has offered this suggestion for unproductive (military) expenditures and the UNDP for social expenditures. However, they should be used only as a last resort and then simply to flag glaring imbalances. Before moving towards policy proposals it is essential to consider country-specific expenditure priorities. 26. Data. In most countries there are significant data problems. Precise data are often not readily available; in some cases they are simply preliminary budgetary estimates; in others cash rather than accrual are used so that the perceived situation may be quite misleading. In a number of countries, data restrictions may impose restrictions on the objectives. For instance lack of knowledge on inter sector arrears may limit the ability to sort out which firms are viable or not and which are simply suffering liquidity problems. The IMF has made some progress in addressing data needs by collecting some government statistics. These are published in the GFS. While these provide a useful first step they could benefit from improvement in terms of their coverage, timeliness and general quality. In the PE work considered here, the data was generally considered adequate. However, this necessitated a major effort especially in such areas as public enterprises and quasi-fiscal deficits. 27. Management Capability and Incentives. Policymaking in relation to public expenditure--macro and micro--and the capacity to implement these policies efficiently and effectively requires a strong management capability and a set of core systems and consistent institutional, organizational, and individual incentives. At a minimum this would include capability for budget planning to ensure adequate funding for all approved expenditures; typically investment initiatives require allocation of significant current and maintenance expenditures for - 14 - their success. It is also important to be able to assess and program all investment expenditures. This can vary from a project evaluation unit to ongoing monitoring of all public expenditures. Examination of these issues should be part of any strategic approach to our public expenditure work. 28. In analyzing this general area, some reports did indicate obvious shortcomings but did not provide answers to suit the particular country situation. With limited resources available how should management be modified or what should be the appropriate incentive structure? What structures and incentives are appropriate at various stages of development to ensure efficient functioning of revenue and expenditure patterns? Most reviews did not address these issues in depth but rather identified institutional shortcomings and highlighted the need for further action. Impact 29. The general question of impact may be best assessed at the country level and then at different points in time: during the work, shortly after and then after some elapsed period. As with all policy work it is often misleading to make a hasty judgement on any new initiatives. In some instances some of the useful gains were made during the actual process of preparation, discussions and through informal meetings which often are not adequately reflected in any formal report. However, these often may produce beneficial effects. In evaluating impact, there are two broad issues: to what extent the public sector work was useful in furthering objectives of the various interested parties; and perhaps even more important whether these were the right objectives. 30. It was felt that the present effort was not adequate to assess impact and alternate approaches need to be considered. This is raised as an issue in the next section. The SPA report pays a lot of attention to impact issues and when its case study reports become available there should be some potential insights available. At this stage the following observations may be noted. 31. Country Objectives. A number of reports were quite successful in furthering country objectives. The better ones in this respect had been carefully prepared in conjunction with country counterparts. In some instances a particular Ministry, often a Ministry of Finance asked a Bank mission to help increase information on a particular sphere of public sector operations such as a public enterprise. Sometimes the Bank was able to bring to bear comparative experience to give some insights on problems being faced. At the other end of the spectrum one finds reports without full government cooperation; in one instance, a govemment agreed only after pressure was brought to bear. Not surprisingly, this last report had little impact. 32. Bank Objectives. Public sector ESW is also quite useful for Bank objectives. It helps provide a basis for an ongoing dialogue with the country, allows different sector groups to focus their efforts in a mutually agreeable way and brings to bear cross-country experience. It can also underpin the Bank's own lending operations. With the shift toward more sector lending, especially in Africa, public sector work will become even more important. One would envisage that continued satisfactory performance by the public sector could lead to a further shift towards general budgetary support. Particular attention will need to be given to domestic resource - 15 - mobilization efforts, as well as overall expenditure priorities (including for recurrent expenditure). The series of public sector reports on Ghana (see Appendix D) provides a number of lessons. In some respects, it was quite successful. The Ghanaians reached a point where they could carry out their own public sector reviews. However, a number of concerns were also noted. Some in government felt the need for an independent outside view. Some thought should be given to how this could be incorporated in future work. Much more worrisome is the seeming increase in aid dependency, for salary supplements, that has accompanied this effort. 33. Private Sector. In recent years the private sector/public sector nexus has grown steadily in importance. For most countries the public sector needs to play a more effective role in facilitating the private sector or at least minimizing the obstacles to its development. The appropriate role will vary among countries and also over time. However, it is important to have some private sector involvement throughout the process. This may typically involve chambers of commerce and trade associations. At a minimum the country should have a well-articulated strategy on what the present and envisaged role is for the public sector. There is some agreement that most productive activities are best left to the private sector. However, it is essential to have a clear understanding of what the private sector believes needs to be done by the public sector. Thus a broader review of the public sector should include an assessment of the adequacy of the legal and regulatory framewv-)rk, and which institutional and infrastructure bottlenecks need to be addressed to facilitate a tin ely supply side response. Given the relatively recent emphasis on this area it is not very surpr sing that the reviews in the sample paid little attention to it. However, without attention to -hese issues, the supply side response during stabilization is likely to be disappointing. 34. Donors. The level and modality for donm r involvement requires some thought. There seems to be general agreement that the typical -eport on public expenditure should be made available to donors at an early stage. There are some divergent views on whether donors might best be involved during review preparation. The experience has been mixed. Some have contributed useful mission members. However, in many cases the participation by donor groups has generated some difficulties either because they placed undue importance on their own goals, tended to overload busy (and scarce) key government officials and often were seen as failing to produce inputs for the report of sufficient quality in a timely manner. Certainly, donor interests raise new, and often complex, analytical issues (e.g., the gender impact of public expenditures). In some countries the potential donor impact can be improved by at least sounding them out ahead of time about their overall objectives and how much flexibility they may be willing to entertain. In a few countries, fragmentation of donor assistance has been addressed by including it in the overall budget framework. V. PROPOSALS FOR FOLLOW-UP ACTION 35. Scope and Timing. While large omnibus reports may still have a role, they are unlikely to be the primary instrument of public expenditure work in the future. Not only are they very expensive, but they also have a very short shelf life and are difficult to update or internalize within the country. We therefore need to take a much more strategic view of our PE work. It would seem appropriate to first reach a decision on what type of review is appropriate for each country. This will depend on country circumstances, role of the Bank, external factors and the - 16 - present state of knowledge for that country. It should also be linked to project and sector work. It may make sense to start with a broad review--either a quick public sector assessment or an in-depth public expenditure review--and then move on to a series of papers focussed on selected expenditure issues. This needs to be defined as a three to five year program, rather than a once- off exercise, with increasing local involvement and capacity building over time. This program, along with other ESW proposals, should be discussed with the Government and outlined in the Country Assistance Strategy (CAS). The CAS should bring out clearly the links between PE work and: (a) the Bank's overall country strategy; and (b) the requirements of other development agencies, including the Fund. 36. Analytical and Management Issues. A simple conceptual framework for PE work, developed by Van Wijnbergen, is set out in Appendix B. This warrants further discussion. In addition, this review has identified a number of analytical issues which need to be addressed: * Macroeconomic Framework. All PE work, even if it is dealing with selected sectoral issues, should be placed within a consistent macroeconomic framework. Every effort should be made to develop a joint macro view with the Fund. In many cases, the Fund's macro analysis may provide the best starting point, provided it is properly reviewed and understood by Bank staff. Several iterations may be required to develop a consistent set of projections on macro prospects, resource mobilization, public expenditure levels, spending priorities and overall sustainability. - Public Sector: Concepts and Defiritions. PE work is moving toward a broader definition of the public sector and public finance issues. Yet concepts and definitions vary widely between countries and PE reports. Some variation may well be appropriate, based on country conditions and data availability. But these variations should always be well-defined relative to some standard benchmarks. This is an area where joint work between the Bank and the Fund may well lead to more consistent analysis (see below). * Data Quality and Timeliness. At the moment, the best source of cross-country data on public finance is the IMF's Government Financial Statistics (GFS). However, GFS data are not particularly useful for PE work because: (a) they are usually several years out of date; (b) coverage of local government and public enterprises is incomplete; and (c) only limited disaggregations by expenditure categories are available. The Task Force needs to discuss how to improve GFS to overcome some of these constraints. * Expenditure Composition. There are major, perhaps intractable, analytical problems in trying to compare expenditures across sectors. In some instances, analysts have resorted to rules of thumb to set intersectoral spending priorities. However, the use of rules of thumb is not satisfactory. It may overlook significant differences between countries and may distract attention from important considerations of expenditure effectiveness and intrasectoral priorities. More work in this area is needed in relation to both policymaking and efficient and effective resource use to define sensible ground rules for using different criteria for intersectoral allocations in different situations. This would focus the efficiency and effectiveness of current resource use and in clarifying the role of government particularly vis-a-vis the private sector. - 17 - Public Sector Management Incentives. One of the primary objectives of PE work is to build local capacity to manage planning, budgets and expenditures. In this context, particular attention needs to be given to the core systems, such as accounting and the institutional aspects, as these embody powerful incentives. This is an area where considerable experience is now becoming available in many OECD countries. PRDPE is proposing a research program to learn from this experience and develop practices relevant to developing countries. This research warrants high priority. 37. Staffmg Issues. A critical mass of experienced staff is required to undertake PE work. Within the Regions, proper incentives--possibly supplemented by training--are needed to get the full commitment of sector staff. Central capacity in this area is limited, and split between two Divisions: PRDPE (which handles public economics issues) and HROVP (which handles public sector management issues). PRDPE is currently working with the Africa Region on several model PE reviews and HROVP is working jointly with the Fund on a manual on budget management. Should this central capacity for PE work be strengthened and/or better coordinated? Alternatively, should the Regional TDs maintain a core capacity to support PE work?7 38. Evaluating Impact. In a time of tightening budget constraints, PE work will have to be both selective and cost effective. Proposed tasks should be justified by their potential impact, not just in terms of the Bank's lending program, but more importantly in terms of their contribution to policy decisions and capacity building in the client country. Budget and staff resources should be allocated accordingly. The SPA country case studies should provide some useful feedback on impact. More generally, a two-phased approach is proposed. First, OED should review the impact of the Bank's recent PE work, to identify examples of best practice and propose ways to make our PE work more effective. This review should also cover process and analytical issues, possibly drawing on the framework developed by Pradhan for evaluating PB work (Appendix C). The second phase would involve more regular, but selective, peer- group evaluations of future PE work. This could include representatives from the country team, Central Vice Presidencies and outside experts. 39. Coordination with the IMF. In countries where the IMF requires Bank inputs on public expenditure issues, the program of work outlined in the CAS should also be discussed with the Fund. In addition, it may be useful in some Regions (e.g., AFR, ECA) to have annual reviews with the Fund on the program of PE work for the coming year or two. It clearly will not be possible to meet all of the Fund's needs at any point in time. But the objective should be to cover the priority issues over a three- to five-year period. To cover the gaps, and to distill information from a variety of sources, the Fund's proposal to exchange short notes on issues of mutual interest makes a lot of sense. It may be of interest to review the experience of addressing the debt issue in a number of countries. Here a broad group was formed for each of the relevant countries to bring to bear a wide array of Bank expertise. - 18 - 40. We have agreed to establish a joint working group to follow up on the issues raised in the IMF's report.8 From the Bank's side, this working group will include representatives from the Regions, HRO and DEC. As a first task, the workingb group will look jointly at the experience with PER work in two African countries. It will also review the issues raised in this paper and identify possible areas for joint work on analytical issues (including a consistent set of concepts, definitions, data, etc). One outcome may be a joint letter from Messrs. Preston and Camdessus, outlining the importance of PER work and proposals for improving Bank/Fund coordination. Recent work on Bank/Fund relations in this area is covered in a separate note by Mark Baird, January 6, 1995. - 19 - BIBLIOGRAPHY Baird, Mark. (1995). Memo on Economic and Sector Work: Implications of Budget Cuts. The World Bank. Chu, Ke-young, Hemming Richard, eds. (1994). Public Expenditure Handbook, IMF. de Melo, Martha. (1988). "Public Investment/Expenditure Reviews: The Bank's Experience." Country Economics Department, World Bank, mimeo. Edwards, Sebastian, et. al. (1994). "ESW Reform in LAC." Task Force Report. World Bank, mimeo. Holmes, Malcolm. (1994). "Handbook on Budgetary and Fiscal Management." In preparation. Kahnert, Friedrich. (1987). "A Long-Term Strategy for Public Sector Analysis in EMENA." World Bank, mimeo. Pradhan, Sanjay. (1994). "Evaluating Broad Allocations of Public Expenditure." Preliminary Working Draft. Premchand, A. (1986). "Changing Patterns in Public Expenditure Management: An Overview." IM4F Working Paper. Rajapatirana, Sarath. (1986). "Review of Economic and Sector Work." Economics And Research Staff. World Bank, mimeo. SPA Public Expenditure Working Group, Meeting on October 24, 1994. Chairman's Report. Task Force on Public Expenditure Policy. IMF. December 1994. Yue, Sanliong. (1994). "Public Expenditure and Poverty." Draft. - 20 - Appendix A Page 1 of 10 PUBLIC EXPENDITURE REPORTS This Appendix provides some general background information on public expenditure reports. Table A. 1 lists Bank reports on public expenditure issues at both the national and sector level but excluded social and infrastructre reports. Table A.2 shows the distibution of PE reports by year and country. Table A.3 shows time required for processing of ESW reports Bankwide. It is noted that the sample selected for this report is somewhat similar. Figure A. 1 gives a quick review of the number of reports and Bank commitments during the period 1987- 93. T'lable A.l: LIST OF BANK REPORTS ON PUBLIC EXPENDITURE ISSUFS (1987-93) Country Title Staff Weck Budget Dlate Code Clhialk ludggdary lpolicy & intergovernlicntal fiscal iclations 7/93 11094 Iodoncsia llospital cxpcndiaurc in Indonesia 6/87 PlIN8717 I nlonesia 'ublic cxpentdittircs. pricc and thlc poor 8/93 11293 Korea Manating licalilt expendittirc under tile nalional helll insurancc: casc of Korca 10/91 WtYPl56 Malaysia Fiscal rcoirmn for stabic growili 4/92 10120 P'Apa N4cw Gui,ct I'lic costs a*id financing of educticlio 6/87 6767 I'hilippines Selccicd issues in public zcsourcc managecmnti 4/88 6887 I'llilippines Country economic rcport: public sector rcsourcc mobilization and expecnditure managemnci 2/92 10056 Vict Nam Rcstrucituring public financc and public enicrpriscs 4/92 10134 Anclatidesli Piblic cxpcnidilurc rcvicw: public rcsourec manageciic iluring Ihe 411t fivc ycar plaii FY91-95 3189 7545 Dlingladesli Povcrly and public cxpcnditures: an cvaluatlin of Ihc impact of seliccte govcrnimient prograins 1/90 7946 Indti A llcalt secior financing: coping witl adIjustmcnt opportunilics for rcorm 6/92 10859 IiuliA Agriculturc public cxpcn(diture review 5/93 12196 Indiia Public cxpcnditure review 12/93 12570 Nepal ipcnditurc in Ihc road scclor 1/93 10988 Necral Maintaiiniig structural rcrorm and manacinig public resources 3/90 8352 Pakistin Public expendilurc rcview 10/92 10770 Sri Lanka Public cxpcnditurc rcvicw 12/93 12337 Asia Iealth finanticing in Asia 4/90 8553 Argenlina Reallocaling resource for Ilic improvcmcnl of educalion 1/91 PUD7471 r Artgenina Public finance rcview: from insolvency to growlhi 2/93 10827 t) p Argcntina Military expcnditure review V92 10611 ° lloliviA ubilic seclor cxlxn(lidtirc revicw willi a spccial cmiplissis on llhc socinl sector 9/89 7746 llolivia A review of ilic public invcslenwl program and fiiiariciig eliuuircmcilts, 87-90 6/88 7248 lir(ail Piblic espenditure subsidly policics anid budgctary rcforn 12/89 7738 lirazil Ptublic scclor invcsiiien( review 1/87 6600 Colombia Iublic sector expcntliture rcvicw 1191 7891 Colombia lducalion scclor rinancc and efliciciicy 4187 6587 Costa Rica Public sector expendilurc review 9/89 7877 Domninicmn Rcp. liuc 89-91 puiblic seclor inivcstmcnt program 3/90 7600 Ecuador Public cxpendiltire rcvicw: chaniginig tIhc rolc of tlic sualc 8/93 10541 utisemala 'Public sector expenditure review 5/89 7478 llonduras l'ublic seclor expcnditurc review 12189 7861 JamAica Public sector investnicnl program atnalysis, crforrnancc ansd prospects 11187 7020 Jamaica The FIY90/9 I PY92/93 ptmblic seclor invesitimcin program 2190 8404 Janmaica Public scclor invcsimenti progrim: 92/93 to 94/95 5192 10297 Mcxico llcalih sector expenditurc and finanicing epoilt 5/89 7802 M(cxico Mcxico in transition: towards a new rolc for ilic public seclor 5191 8770 Nicaragua Pub lic sector expeiidilurc review 9192 10785 Plaragumy Public txpenditure rcview: Ilic social sectors 6/92 10193 Pcr Public expcndilurc review 1/81 652B uhuguay Povcty Assessmcnt: public social cxpcndiiltire and their imnpact on Ilic income distribulion 5/93 9663 lUrugtuay Piblic sector investll1cils revicw 5/88 7030 VeeztuelA A review of ilie 90-93 public seclor iumvcsitimciut 12/90 8588 LAC Trends in government cxpctdliturc antd rcvenue in [atiii America, 75-88 11/91 10149 LAC Nutrition and licalilt progrmns in latlin Auiterica: largceinig social cxpcn(lilurc 6/89 11593 VI.YLAfricc 47 Angola I'ublic expendliturc issues and prioritics during tranisition to a maiket cconomy 6/93 11649 llenin Review of ptiblic expenditure. 85-90 1/88 6951 flulkinat Patso Public expendilurc review 5/93 11901 llurundi Piblic cxpenditurc review 2/92 8590 Cameroon Public investimicnt review 6/90 8866 Clhad Public expclditurc revicw 6/90 8424 Comoros A review of tlie public investimeni programss, 91-9. 10/91 10016 Cote d'lvoirc Public seclor expenditure review 10/87 6980 Cote d'lvoirc Public expenihiilur rcview 7/90 8748 t ;,. Elqualorial Guinca Public cxjcndiltaur issues 10/93 12367 Eiluiopia A sludy of e-alth financing: issue and oplions 4/87 6624 Eliliopla Public invesimenl program review 2188 6859 Ghtnt Public cxlcnditurc revicw, 90-92 6/90 8616 oCD Glana r'ulili: ex penditurc rcvicw, 89-91 10/89 7673 Kenya Investing in stabilization aind growbh lIhrougih public seclor a(djustmcnt 1192 9998 Kenya Public expenditure issucs 4/89 7508 Kenys lReview of cxpcndilure isstues st(d onptionis in hcaltih finanicinig 2/88 6963 1Icsoltio lPublic expcnditure ptiorities 9/88 7243 Madagascar Public expcndiure, a(ljusitcico & growvlt 12/89 7803 Madagascar Public expendilure review 90-92 7/90 8919 Malawi Public expcnditure review 4/90 7281 Maurilanis Review of rlic 89-91 public invcstiicni prograrn 6/89 7783 Mozambiquic Public cxpcnditurc rcview 9/89 7615 Mozambique Second public cxpendiittrc review 12/92 11524 Mozmnbiqtic Education sector expentliture, managemicnlt said rinancinig review 9/92 11000 Namibia Povcry Allevitklion with sustainablc growili 10/91 9510 Nigeria Public expenditure review 6/88 7175 Nigeria Public cxpcnditure managcmcnt 11/91 9865 Nigeria liallit care coils, fiAianeing and utilization 10/91 8382 lwandla Public expenditure program: an insirtment of economic stralegy 10/89 7717 Senegkl Public expenditure rcview 5/93 11559 Senegal Issues in be alth care finaiciirg 10/92 11059 Sencgal A review of lthe Ihrcc year public iinveslimenici progrAm, 87/88-89/90 2/87 6450 Senegal Parspublic scclor rcvicw 5/89 7774 Seychellcs Public sector investimenil program: an insiruincitt of policy 6/93 12010 Somalia Crisis in public expenditurc management 3/91 8727 SwazilAind Public expenditure revicw 7/93 11583 Tanzanis Public cxpenditure review 5/89 7559 tJgandA Public choices for privale iniltiaives: prortlizing public excpenditures for sustainable and e1uitabic growth 2/91 9203 UgAnds Managinig public expendliture 6/92 10512 Zambia Public expenditure rcview 1(0/87 6438 Zambia Public expenditure review 12/92 11420 7ambis Public invesitmcn program 4/90 8721 Zimbabwe Issucs in ilie inmncinlg or icaltils services 7/90 8100 Zailc A review or govcnmiicint cxpcndiiturc T) 8995 Africa thiblio sector decentralizalion: cconoolsic policy rcforn, and sector invcstimicnt progrtalis 5/90 l It Afiica Poverty-conscious restrucluring of public expenditure 12/90 9ijol 'X - Algeria Public in4esimcnt rcview 5/90 8629 O lahrain Public cxpcnditurc rceiew 2/89 9404 -8ypt Public sector investimen review 10/93 11064 lordan Public expcnndiure review 2191 8986 forocco Agriculturc expcnditure rcvicw 6/90 8878 Mtorocco DcenlrialiAtion in Morocco: local govcrnitict)t cxpcnditturc *nd unanagcmicn 8/90 8782 Morocco [leaching itc disadvantag:ed social expcndliturc prioritics in (lic 1990s 9/90 7903 Morocco Issucs and prospects in rtue public sector 6/92 10157 Mlorocco Costs, Financing and cfficicncy of ulc cducatlion syscim 6/93 11931 lTil;sia Adapting public expenditures to cliangitng resource availability 1/87 6604 Tunisia l!xpcnditure review of ilic agrictilturc seclor 6/81 6815 Tunisia Agriculture expenditure review 5/91 9511 Ycmen Arab Republic Domncstic resource mobiliztlion, tuscr charges, antl recurrent expentli(tire 6/87 6851 YLiPCA: 6 Ilungary Reform of social policy and cxpcndilure 7/91 9349 IPolsond Social sectors expcnditure rcview 1/92 10158 Turkey Adjusting public invesimclt 3/87 6603 Tutkey Costs and financing of primary and secondtary educalion 11190 9097 1riskcy Issucs and oplions in hcalith financinig 9/90 8042 \) Tulkey Inflation and ihc financing of govcninient cxpcndiltirc: an inlroductory analysis wilhl an application to Turkey 1189 8746 Yll.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 9,J. MultIRelonal: QI ALL Capital cxpcndilure for clectric powcr in ilic dcveloping counlrics in the 1990s 2190 9321 Source: World Blank Incmal Documcnt Unit, IDR Tlhis table covcrs only ilic rcporls whosc Itcmcs arc public cxpcndi(itlrc review at bolh national atdl soctor lcvel. All iic social and infrasiructurc seclor reporis (cducational, hicalil, tniusportation, watcr and sanitation, iringtion, power, ctc.) arc excldedl in tlic table aliltougii altmost all tlose sector reports contiin public cxpendilure analysis parts. - 25 - Appendix A Page 6 of 10 Table A.2: NUMBER OF BANK REPORTS ON PUBLIC EXPENDITURE ISSUES (1987-1993) 1987 1988 1989 1990 1991 1992 1993 TOTAL Africa Angola = = = = = 1 1 Benin 1 1 Burkina Faso = = = = = = 1 1 Burundi 1 1 Cameroon = = = 1 = = = 1 Chad 1 1 Comoros l 1 Cote d'Ivoire 1 1 2 Equatorial Guinea = = = = = = 1 Ethiopia 1 1 2 Ghana 1 1 2 Kenya 3 1 1 = === 3 Lesotho 1 1 Madagascar 1 1 2 Malawi 1 1 Mauritania = = 1 = = = = 1 Mozambique 1 2 3 Namibia == ==== 1 1 Nigeria 1 2 3 Rwanda 1 1 Senegal 1 = 1 = = 1 1 4 Seychelles 1 1 Somalia 1 1 Swaziland 1 1 Tanzania 1 . 1 Uganda 1 1 2 Zaire I = _=_ = 1 =I= 1 Zambia . 1 l 1 3 Zimbabwe I 1 Africa = _= =_2 = = = 2 Subtotal 4 5 8 10 7 7 6 | 47 - 26 - Appendix A Page 7 of 10 1 1987 | 1988 1989 .1990 | 1991 1992 1993| TOTAL MNA .__ -_ _ ___ __ Al zeria 1 } 1 Bahrain . 1 _ 1 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 1 1 Jordan 1 1 Mslorocco 3 _ I 1 5 Tunisia 2 | = 1 3 Yemen A.R. 1 _ T 1 Subtotal 13 O 11 4 2 1 T 213 Europe and Central Asia .',ungarV _ 1_ I __ 7_ 1 1 I 1 Poiand ______ r _ _ _ _____ Iurkev I 1 _2 _- I T 4 Subtotal 11101 1 2 11 I1 01 6 East Asia and Pacific _____ _ China I _ _ _ I _ | 17| 1 T Indonesia 1I 1 1 1 1 2 Korea I _ I _ 1 I 1 1 Malaysia _ 1 1 Paoua New Guinea I _I = __=_ 1 Philiopines _ _ 1 _ ____ 1_| | 2 Viet Nam 0 0 r __-r Subtotal 2 I | O O_1 3 | _2 | _ 9 South Asia Banzladesh 2 _ 1 1 I 1 _ ______ India I I _ __| _ | 1 | 2 | _ 3 N-eoal C _ __ | 1 2 | |_1_|_ Pakistan f____ 1 1 Sarn Lanka . | | | | | | | 1| Asia (Rezioinal) 1 1 Suibtotal o ° I 3 0 2 4 10 - 27 - Appendix A Page 9 of 10 1987 198 1989 1990 1991 1992 1993 TOTAL Latin America and Caribbean Arzentina 1 1 1 .. 3 Bolivia l 1 2 Brzl I 2 Colombia 1 _ . 2 Costa Rica I . 1 Dominican Republic 1 I Ecuador I 1 Glitemala 1 1 Honduras 1 1 Jamaica 1 1 1 _ 3 Mexico 21 1 T 1 1 2 I r_ _ _ _ - _ _ _ _ _ _ _ Nicaragua |1 _ I _ I | | 1 i 1 Parazav I _ i _ 1 1 1 11 Peru1! I I 1 Uruzuav 1 _ I _ _ ______1 __2 e V nezuela I__ I | 1 | 1 _ __ LLAC (Rezional) | | 1 _ _ _ 1_|_r _ 2 Subtotal 14 2 7 3 4 |4 3 7 Multi-Rezional I 1 | | l_|___ GRAND TOTAL 14 8 18 23 [15 18 17 113 - 28 - Appendix A Page 9 of 10 Table A.3: ELAPSED TIME BETWEEN FIRST MISSION AND FINAL REPORT Elapsed Time (in Months) 1987-93 ESW Reports (Bankwide) Yellow 10 Green 12 Gray 13 21 Selected Public Expenditure Reports Yellow 12 Green 11 Gray 15 Appendix A Pave 10 of 10 Figure A.1: NUASER OF PUBLIC EXTENDITURE REPORTS AND BANK LENDIG (1987 to 1993) so ~40 30 ~2O SX Z~~~~~~~~~~~~ RE GI ON- IMNO. OF REPORTS E COMMITMENTS (S BIL.) - 30 - APPENDIX B Page 1 of 2 CONCEPTUAL ANALYTICAL FRAMNEWORK - SUMMARY 1. This Appendix is summarized from a paper by Sweder van Wijnbergen. The full paper will be available in February. 2. An obvious prelirminary, omitted in most reports, is a defuintion of what actually constitutes the government. The only defensible definition is a comprehensive one including all layers of govemnment: on and off-budget; state, regional and local; areas of direct control and spending, and ones where the impact is indirect, through regulation or SOEs run at armr's length. 3. The next step should be an a priori determination of what the boundaries of the government should be: a priori to determining what the budget constraint is. The reason for such an a priori determination of the boundaries is that one can by now arguably accept the view that the Government should only be active in an area if there is a positive reason for doing so. To avoid being pedantic, the report should only cover areas where this is a non-trivial issue; there is no need to argue the case for a judiciary, say, or police protection. On the other hand, areas covered should not be confined to those where the government makes its presence felt through actual spending. Is there is a need to expand, modify or cut back regulahon? One would expect the report to focus in particular on SOEs and functions of government that could, even if they may not be right now, be run on a corporate basis (like the post office); how far should privatization go? Is there a need to accompany privatization by a proper regulatory framework to avoid the creation of monopolies? In fact are there obvious caps or excesses in the existing regulatory framework? 4. Once these a priori boundaries are set, the budget envelope becomes crucial. Given an agreement on where the govemment should be present, under what kind of budget constraint does it operate? In the short run, this is JIiMF territory; but if sustainability is at all an issue, a medium-term framework is more appropriate, and this is typically not the focus of DIfF programming. The issue is of course key, although not covered by manv of the reports surveyed. Such an exercise would require a look at the tax structure, at non-tax revenues and at the amount of inflation tax (or more generally seigniorage) that can be expectedc given the macroeconomic objectives the government has (i.e., growth and inflation rates) and the financial structure of the country. 5. Given the boundaries of a government activity and the medium-term revenue base, overall spending levels follow. This in turn raises the issue of the allocation of expenditure, both over sectors and over current and capital expenditure. Here there is a nisk of becoming theoretically elegant but not very useful in practice. Elaborate shadow pricing, deriving government objective functions, etc., are all very nice but in practice infeasible without making - 31 - APPENDIX B Page 2 of 2 up so much crucially needed information out of thin air that the whole exercise becomes dubious. An alternative approach would try to define medium-term objectives of the government in key areas; and seek to derive the necessary spending levels from there; if it turns out the overall budget constraint is violated as a consecuence, the issue of priorities would need to be raised with the government, either on the spending or on the taxation side. This section would also have to consider whether existing spending programs are in fact efficient given stated objectives and budget envelopes. This section would presumably be the main body of the text. Of course an iteration may be necessary if derived expenditure levels are wildly out of line with the budget envelope. 6. Sustainability is a key issue, to be covered next. Here there are two elements. One--given stated objectives and assuming efficient spending programs--are the current spending levels adequate in the face of developments to be expected in the future? Examples are an assessment of the adecuacy of pension arran-ements and health and education programs given likely demographic developments; recurrent expenditure in the case of investment programs; and so on--mundane matters, but obviously of crucial irnportance. A second aspect is more political and only relevant afte. major reform programs. Sustainability of a program of reform will critically depend on whether enough broad-based support will be forthcoming. In this context the issue of adjustment assistance will need to be raised. If powerful groups will be hurt in the reform process, sustainabilitv of the entire prograrn may well be at risk unless their main concerns are addressed one way or the other. The key issue is, of course, how to do this without derailing the reform program that starLed the whole problem. Recommending drastic reforms without addressing these issues is simply irresponsible and gratuitous. 7. Finally, once boundaries, budget constrants, and specific spending programs have been covered, there is the crucially important issue of implementation. Are the government institutions adequate, too extensive or not extensive enough? For example, is there a unit equipped to assess public investment programs and determine priorties? Is there a well functioning treasury function? Does the current organization of government lead to conflicts of interest (e.g., is competition policy managed, from the Ministry of Commerce)? Are administrative procedures excessively cumbersome for the private sector? Are budgeting procedures set up to handle recurrent costs of investment projects, i.e., is there an attempt at multi-year budgeting? And so on. Reforms in this area may in fact do more to help the non- government sector/private sector than anything else. - 32 - APPENDIX C Page 1 of 2 FRAMEWORK FOR EVALUATING PUBLIC EXPENDITURE WORK This Appendix gives a framework proposed by Sanjay Pradhan. Further details on his work on intersectoral allocations are given in Pradhan (1994). FIRAMEWORKI FOR EVALUATING iPUBLIC EXI'ENDITUREI WORK Improving Instituitional Capacity Idlentifying and Correcting Expendiltire Improving Donor Lending Priorities and( Processes Imbalances P'rocess * Close paritllerslilp witih * Slatfing (experiencedt tIlsk manager) * Donor involvemilent in mission governiment * lncentives/coiiuitiminii t of sector staff * Sl'A dlonors-Baniik task force in * Follow-vip on PERs * Mission size? Africa * Series of anntial PERs in Africa * Consulting interest grouips * Stalling (maltire task imnager) * Clar-ity of objectives * Stal-ing (institutional specialists) * Country versuis Banik priorities * Workshops for government Analysis * Presence of priority-setting * Comprelhensiveness of public sector * Analysis of dlonor-financed process * ltole of governilielit vs. pirivate sector Projects in PII * Reltlive au1ltoniomly of cential 0 Aggregate level and macro Iramilewoik * Iteview of Woilil i3ank len(ling agencies * I'll' review; while clephanlts, core portfolio * Incentlives andl( (legree of investmient; cnplilitl-curl-rclit cenridifivtion in budget * Civil service pay alndl emniploymlient preparation * Nonwage O&M; p)liysical input mix * Use of economic analysis in * Intrasectoral allocations--progranis (e.g., dlecision-miiakinig primary/terliary) * Bud(get implementation * Poverty incidenice * Procurement * Intersectoral allocations * Accountling * Methodology: iiptlts vs. * Atditinig outputs/outicomies; trend(ls; international * Evalatnlioll colllplll-isoIIs; cost-effectiveness, cost- bellnefi, jumdgllel)t Recommentlalioins * ReconImmendations for budget * Reconmmended restruic(wstitig of * Recolmlllended clhaliges in prepa-aftion, implementation, expeitu(lirtie allocaittiolns wittiill andl actross dolnor len(linig striategies accounlinig and auditing sectors, as %yelt tls econloiic caltegor ies Imnlpact * Actiual improvements in budget * AcItial reallocafions of spending in * Actual chianges in dtonor preparation, impllementation, bud(gets and( tihe pl)an, across andl witlinii lentding strategies, including accounling antd aud(liing sectors anid economicii calegories World Bank country assistance strategy t n - 34 - Appendix D Page I of 2 PUBLIC EXPENDITURE REVIEWS IN GHANA 1. This Appendix provides a brief review of the Ghana experience with PE reviews. This group was chosen to provide some insight on the merits of pursuing a series of PE reviews in one country. 2. Since the beginning of the Economic Recovery Program in 1983, the Bank and DMF have had an extensive dialogue with the Ghanaian govemment on public expenditure issues. Yearly consultations on public expenditure issues began in 1985 and five PE reviews have been written since that time.' The initial review in 1985 brought a very large team to Ghana and played a major role in assisting the Govemment in preparing its development plan. This review provided a base for those that followed. Later reviews used smaller teams and were able to expand their focus to issues of budget process and implementation. Through the series of PE reviews, a degree of expertise was built up among Ghanaians so that the most recent PE review was prepared by the Government itself. This has had imDortant advantages in terms of ownership, but has also meant that some important areas of analysis were excluded from the recent document. 3. The 1985 PE review set out estimates for the resources likely to be available to the Government over the following years (1986-88) and then considered issues of inter and intra- sectoral development. It set out a possible recurrent expenditure program and a possible capital expenditure program placing emphasis on health and education, as well as on the importance of restoring the efficiency of public administration in Ghana. A capital expenditure prograrm was built from "the bottom up". Sectoral strategies and programs were reviewed. Economic rates of return and domestic resource costs for some 159 projects were calculated and an investment program based on those projects with the highest rate of return. 4. Each of the following three PE reviews carried out a similar analysis--first discussing the macroeconomic frarnework, then considering allocation and implementation issues and finally looldng closely at sectoral strategies. The benefit of such regular reviews was that it allowed the Bank and the Government to set objectives and assess progress on a number of themes. For example, on resource allocation, each review discussed progress in improvinc the system of civil service compensation with the objective of improving the efficiency of public administration. Issues concerning the balance of recurrent expenditure between wage and non-wage expenditures were also discussed in each review. Norms for recurrent expenditures in certain sectors were set out early on and later reviews assessed progress in achieving these norms. Progress in See Priorities for Public Expenditures, 1986-1988 (1985), Ghana: PubLic Expenditure Review, 1989-91 (1989); Public Expenditure Review: 1990-92 (1990); Public Expenditure Review: 1992-94 (1992); Republic of Ghana: Public Expenditure Review 1993 (1994). - 35 - Appendix D Page 2 of 2 implementing the desired changes in the public investment program were also assessed. Over the years more attenton was focused on implementation issues, ranging from difficulties with the budgeting process, to diversions of fujnds, to basic difficulties in collecting information on actual expenditures. 5. The most recent PE review was carried out by the Ghanaian government. It presented a useful discussion of issues concerning budgetary process and implementation (an important indication that the government recognized the seriousness of these problems). However, it did not consider the overall macro framework nor any forward-looling plans for moving beyond the difficult fiscal situation that resulted from the 1992 election. 6. The experience with expenditure reviews in Ghana shows that regularly scheduled PE reviews can help the government develop expertise in dealing with public expenditure issues. Over time, they may be able to take over the expenditure review function. This will help to build ownership and better understanding of public expenditure issues. At the saame time, the Bank must encourage governments to maintain a high standard of analysis and a forward looking focus. 7. Presently there is some concern that this approach is generating significant aid depending on the fact that low-paid government workers do not feel motivated to carry out their normal tasks unless they can benefit from some of the largesse available for "outside" consultants.