135&VZ jte /?9'7 ANNUAL REPORT ON PORTFOLIO PERFORMANCE Fiscal 1993 Annual Report on Portfolio Per formance Fiscal 1993 The World Bank Washington, D.C. © 1994 The International Bank for Reconstruction and Development/ THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing September 1994 This report is a study by the World Bank's staff, and the judgments made herein do not necessarily reflect the views of the Board of Executive Directors or of the government they represent. ISBN 0-8213-2828-X iii Contents Abbreviations iv Findings and Conclusions I 1 Portfolio Size and Composition 4 2 Portfolio Rating Methodology 6 3 Portfolio Performance by Region 9 4 Portfolio Performance by Sector 13 5 Other Aspects of Portfolio Performance 14 6 Portfolio Performance Management 16 iv Acronyms and Abbreviations ARPP Annual Report on Portfolio Performance CPPR Country portfolio performance review CVPU Central Vice Presidential Unit DO Development objectives EAP East Asia and Pacific ECA Europe and Central Asia FSU Former Soviet Union GDP Gross domestic product IDF Industrial development and finance LAC Latin America and Caribbean MENA Middle East and North Africa NEAP National Environmental Action Plan NGO Non-governmental organization OED Operations Evaluation Department OPR Operations Policy Department OS Overall status PAR Project audit report PCR Project completion report PHN Population, health and nutrition PMTF Portfolio Management Task Force RM Resident Mission SAL Structural Adjustment Loan TA Technical assistance Findings and Conclusions This World Bank Annual Report on Portfolio Performance (ARPP) for fiscal 1993 has been prepared in accordance with the recommendations of the Portfolio Management Task Force (PMTF) report. It summarizes the results of a coordinated effort by the Bank's six Regional Offices, the Central Vice Presidential Units (CVPUs) and Development Economics to review performance of the operations portfolio in fiscal 1993 from the country and sector perspectives.' The main findings and conclusions of this review follow. Portfolio Performance Management * In fiscal 1993 the Bank intensified its efforts to improve the quality of new projects and the performance of ongoing ones. The main actions included preparing country assistance strategies and plans that emphasized effective portfolio performance management; encouraging participation of all stakeholders throughout the project cycle; enhancing borrower ownership, with emphasis on strengthening local institutional capacity for project preparation and implementation; ensuring that project designs were in keeping with the institutional, technical, and financial capacity of borrowers; integrating more effectively lessons from previous projects into new project design; improving financial accountability; restructuring poorly performing projects; and canceling or closing loans where prospects for improvement were poor. * Reflecting the stronger portfolio management efforts in fiscal 1993, the average supervision inputs per project rose to an all-time high of 15.4 staff-weeks, or 9 percent higher than the previous peak reached in fiscal 1992. Supervision inputs for the quarter of the Bank's portfolio that was intensively supervised in fiscal 1993 averaged 36 staff-weeks. Actual Regional allocations to portfolio management in fiscal 1993 exceeded the budget contract amounts by 13 percent. All Regions sought to increase the cost-effectiveness of supervision resources by devolving greater responsibilities for portfolio management to Resident Missions. * The above improvements in portfolio performance management are being strengthened further during the current fiscal year in the context of a Bankwide action program to implement the PMTF recommendations ("Next Steps").2 Since Board approval of this program in July 1993, significant progress has been made in various areas. Country portfolio performance reviews (CPPRs) are now widely accepted as an integral part of the Bank's core business practices and Regional business plans. A recently established Participation Fund to encourage and test new ideas for increasing beneficiary participation and improving project development results has been fully committed. New guidelines for country assistance strategy papers have been issued that, among other things, explicitly incorporate portfolio management concerns into the formulation of country assistance strategies and plans. Various task forces (cofinancing, financial reporting, and auditing and portfolio management information) have completed their work and have issued their recommendations for management review. An independent The World Bank consists of the International Bank for Reconstruction and Development (IBRD) and the International Development Association (IDA). The term "operations portfolio" refers to Bank operations under implementation and is used interchangeably with "portfolio" elsewhere in this report. The term "Regional" refers here to the Bank's Regional Offices. Dollar amounts are current U.S. dollars. 2 Getting Results: The World Bank's Agenda for Improving Development Effectiveness, World Bank, 1993 (available from the World Bank Bookstore). 2 inspection panel for the review of Bank-supported projects has been established. A Public Information Center is now open, making available previously restricted operational documents. Improvements in risk and sensitivity analysis and in monitoring and assessing progress toward project development objectives are being pursued. The identification of skills required for portfolio management is making good progress, and staff training programs in various areas of portfolio management are also being developed. A new personnel appraisal process that will recognize and reward portfolio performance management and evaluation work has been introduced. The improvements in portfolio performance management over the past two years, and further improvements in the future as "Next Steps" initiatives are implemented fully, would help enhance the development effectiveness of the Bank's portfolio in the medium term. The most visible impact thus' far of these efforts, particularly project restructurings and closing or cancellation of loans for poorly performing projects, has been the stabilization of overall portfolio performance. Beyond the immediate effects of portfolio cleanup, however, a definitive turnaround in overall performance in the next few years is not assured. Against the ongoing efforts to improve portfolio performance management must be reckoned the implementation risks inherent in the ambitious and complex development agenda of the Bank and its borrowers. The risks associated with Bank lending to borrowers that are at an early and difficult stage of transition to a market economy or of economic reform and adjustment are also significant. As experience shows, country economic and institutional factors ultimately have a considerable impact in determining the development outcomes of projects. Portfolio Performance Based on Overall Status Ratings * The Bank's portfolio of operations under implementation in fiscal 1993 comprised 1,897 operations in 126 country portfolios, with a total commitment value of $148.4 billion. The overall health of the Bank's portfolio, based on overall status ratings, was relatively stable between fiscal 1992 and fiscal 1993. The percentage of problem projects peaked at more than 20 percent in fiscal 1991 but declined to 18 percent in fiscal 1992 and 1993. Excluding countries that were in nonaccrual status, suffered serious civil conflict in fiscal 1993, or both, the percentage of problem projects declined from over 18 percent in fiscal 1991 to 16 percent in fiscal 1992 and 15 percent in fiscal 1993. * The total number of problem projects (based on overall status ratings) declined by only one, from 339 to 338. This negligible change masks significant, offsetting changes between fiscal 1992 and fiscal 1993 within the Bank portfolio: (a) a reduction of 129 problem projects because of loan cancellations and closings and improvements in performance; and (b) an increase of 128 problem projects, largely because of a deterioration in project performance and the effect of "aging."3 * Performance trends between fiscal 1992 and fiscal 1993 varied among Regional portfolios. The East European and Former Soviet Union (FSU) portfolio, in the Europe and Central Asia (ECA) Regional Office, registered a significant increase in the percentage of problem projects, largely reflecting the difficulties experienced by countries in transition in carrying out economic reforms. At the other end, the Latin America and Caribbean (LAC) portfolio experienced a 3 New projects are almost always given a nonproblem rating because implementation problems have not yet had time to emerge. As implementation gets underway, and problems begin to be encountered, a downgrading in performance ratings tends to occur. 3 significant drop in the percentage of problem projects as a result of improvements in project performance and loan cancellations and closings. Portfolio Performance Based on Development Objectives Ratings * The percentage of problem projects (based on development objectives ratings) increased from 12 percent in fiscal 1992 to 13 percent in fiscal 1993. This increase does not appear to indicate a real deterioration in the development performance of the portfolio. It reflected Regional efforts, within the limitations of the current rating methodology, to base development objectives ratings on more realistic expectations about the extent to which development objectives would be met. * The Regional percentages of problem projects (based on development objectives ratings) were all less than the corresponding percentages based on overall status ratings. This indicated a widely held expectation among Bank staff that projects would eventually achieve their development objectives once implementation problems are resolved. Portfolio Performance by Sector * The telecommunications, education, technical assistance (TA), population, health and nutrition (PHN), transportation, and industrial development and finance (IDF) sector portfolios had a significantly lower percentage of problem projects (based on overall status ratings) than average in fiscal 1993, while the industry, agriculture and power sector portfolios had a significantly higher .percentage of problem projects. However, portfolio performance between sectors differed to a much lesser degree than between Regions or countries, implying that performance problems tended to be more country-specific than sector-specific. * Between fiscal 1992 and fiscal 1993, the water supply and sewerage portfolio experienced the largest improvement in performance. The percentage of problem projects in that portfolio fell from 26 percent in FY92 (highest among the sectors) to 17 percent in fiscal 1993 (a little less than the Bankwide average) because of project restructurings and other actions taken by borrowers to strengthen water and sewerage authorities and improve their financial performance. 4 1. Portfolio Size and Composition Portfolio Size The Bank's portfolio of operations under implementation in fiscal 1993 comprised 1,897 operations in 126 country portfolios, with a total commitment value of $148.4 billion (Table 1) and an undisbursed balance at the beginning of the year of $77.0 billion (52 percent of commitments)'. The commitment value represented over 40 percent of the total investment cost at appraisal of the 1,897 operations ($368.7 billion). The active portfolio, which excludes countries in nonaccrual status on June 30, 1993, consisted of 1,886 operations with a commitment value of $147.5 billion. Between fiscal 1992 and fiscal 1993, the total commitment value of the Bank's portfolio in U.S. dollars increased by 4.4 percent in nominal terms and 1.7 percent in real terms, mainly because of the start of Bank lending to the FSU countries. TABLE 1 BANK PORTFOLIO IN NOMINAL AND REAL TERMS, FY91-FY93 Number of Projects Lendine Commitment Value (US$m) Average Loan Size FY91 FY92 FY93 FY91 FY92 FY93 FY91 FY92 FY93 Nominal Terms Adjustment a/ 143 142 136 22,690.2 23,415.9 23,651.6 158.7 164.9 173.9 Investment 1702 1710 1761 115,342.2 118,729.2 124,723.4 67.8 69.4 70.8 Total 1845 1852 1897 138,032.4 142,145.1 148,375.0 74.8 76.8 78.2 Real Terms b/ Adjustment a/ 22,690.2 22,815.8 22,446.3 158.7 160.7 165.0 Investment 115,342.2 115,686.6 118,367.1 67.8 67.7 67.2 Total 138,032.4 138,502.4 140,813.4 74.8 74.8 74.2 a. Adjustment includes both structural and sectoral adjustment and debt reduction loans. b. Deflated to FY91 prices using the following commitment deflators: FY91, 100; FY92, 102.63; FY93, 105.37. Regional Composition Reflecting the start of Bank lending to the FSU countries, the ECA Region increased its share in the total portfolio in fiscal 1993 but continued to have the smallest Regional portfolio in terms of number of projects (Table 2). The Africa portfolio continued to have the largest (about 35 percent of the total number of projects), followed by LAC with 18 percent and East Asia and Pacific (EAP) with 17 percent. The portfolio remains concentrated in a few countries. The five largest country portfolios in terms of commitment value-India, China, Indonesia, Mexico and Brazil-accounted for 22 percent of the total number of projects and about 42 percent of total commitment value in fiscal 1993. These countries had 54 percent of the population of Bank borrowing countries. Forty borrowing countries (including the five largest borrowers) each had more than $1 billion in lending commitments, at least 20 projects ("large country portfolios"), or both, and together they accounted for over 70 percent of the number of projects and 87 percent of total commitment value in fiscal 1993. These countries accounted for 87 percent of the population of Bank borrowing countries. 4. The number of portfolios includes three multicountry or regional portfolios. 5 TABLE 2 BANK PORTFOLIO DISTRIBUTION BY REGION, FY91-FY93 F91 F2 F193 No. % %d o tat % Q Nof f Comminent %of NQf " %(Y o nent %o Rgion Po. Total Valuie(EM) Total Poi Total Valae(US$M) Total Proi. Total Value(USM Total Africa 641 34.7 22123.3 16.0 644 34.8 23,238.7 16.3 656 34.6 23,669.1 16.0 East Asia 3(9 16.7 27,301.3 19.8 319 17.2 29,844.6 21.0 326 17.2 31,954.3 21.5 SouthAsia 277 15.0 31,077.3 225 275 14.8 30.862.3 21.7 277 14.6 3291.0 20.4 Etrope&Central Asia 10) 5.4 13,858.9 10.0 101 5.5 14,276.2 100 120 63 16,473.9 11.1 MidleEast&NorthAfrica 179 9.7 9,974.7 72 181 9.8 1l720.4 75 178 9.4 11,556.4 7.8 LafinAmenica 339 18.4 33,696.9 24.4 332 17.9 33,202.9 23.4 340 17.9 34,430.4 23.2 Banmkide'Ibtal 1845 100.0 138,032.4 100.0 1852 100.0 142,145.1 100.0 1897 100.0 148,375.0 100.0 Sectoral Composition The portfolio share of the agricultural sector continued to shrink in fiscal 1993 (see Table 3) but remained the largest among the sectors (over 25 percent of the total number of projects and 19 percent of total commitment value). Sectors that have also declined in recent years include industry and IDF. Meanwhile, the combined share of the education and the PHN sectors increased from about 15 percent the total number of projects in fiscal 1991 to over 17 percent in fiscal 1993 and from over 9 percent in total commitment value to almost 12 percent. From a relatively small base in fiscal 1991, the number of projects and the commitment value of freestanding TA and public sector management projects have also continued to grow rapidly, pushing their combined share of the total number of projects from about 6 percent in fiscal 1991 to over 7 percent in fiscal.1993, and their share of total commitment value from over 2 percent in fiscal 1991 to over 3 percent in fiscal 1993. TABLE 3 BANK PORTFOLIO DISTRIBUTION BY SECTOR, FY91-FY93 FY91 FY92 FY93 Na. at % of Commitment 2L( No. of Commitment %t No. of %jof Commitment % of Sector Prot. Total Yaue US$M) Ttal Prot Total Value (US$M) Total Pro. Total Value US$M Total Agriculture & Rural Devt 513 27.8 30,337.6 22.0 501 27.1 29,579.9 20.8 482 25.4 28,832.3 19.4 Transportation 219 11.9 17,325.5 12.6 213 11.5 17,265.3 12.1 220 11.6 19,166.5 12.9 Urbanization 104 5.6 7,354.5 5.3 112 6.0 7.746.4 5.4 116 6.1 8,283.6 5.6 Tourism 0 0.0 0.0 0.0 0 0.0 0.0 0.0 1 0.1 130.0 0.1 Water Supply and Sewerage 97 5.3 6,533.2 4.7 96 5.2 6,993.1 4.9 1() 5.3 7,789.8 5.3 Industrial Devel/Finance 158 8.6 17,528.5 12.7 150 8.1 17,107.2 12.0 135 7.1 14,822.4 10.0 Industry 66 3.6 6,149.9 4.5 60 3.2 6,120.9 4.3 57 3.0 6.105.4 4.1 Communications 34 1.8 1,978.8 1.4 32 1.7 2,340.2 1.6 35 1.8 2,620.0 1.8 Power 154 8.3 21,325.9 15.4 155 8.4 21,961.2 15.4 158 8.3 21,390.4 14.4 Energy 57 3.1 5,317.2 3.9 53 2.9 5,408.0 3.8 52 2.7 5,720.8 3.9 Education 170 9.2 8,617.9 6.2 179 9.7 9,655.9 6.8 197 10.4 11,039.1 7.4 Population 98 5.3 4,274.4 3.1 114 6.2 5,151.0 3.6 132 7.0 6,715.6 4.5 Technical Assistance 94 5.1 1,094.6 0.8 99 5.3 1,162.2 0.8 112 5.9 1,626.3 1.1 Public Sector Management 13 0.7 1,945.1 1.4 17 0.9 2,546.8 1.8 24 1.3 3,105.8 2.1 Program Lending/SAL 68 3.7 8.249.3 6.0 71 3.8 9,106.7 6.4 76 4.0 11,027.1 7,4 Bankwide Total: 1,845 100.0 138,032.4 100.0 1,852 10.0 142,145.1 100.0 1,897 100.0 148,375.0 100.0 6 2. Portfolio Rating Methodology Form 590 System The portfolio review for.the fiscal 1993 ARPP employed the same rating methodology as was used in prior reviews (the "Form 590" system, named for the form used to summarize the views of each supervision mission). Within this methodology, however, the review paid greater attention than in the past to assessing the expected development results of projects. Projects were first rated individually in the Form 590 on several performance categories, including (a) the extent to which project development objectives (DO) were expected to be met and (b) overall status (OS). For various aggregations of projects-i.e., country, Regional, sector, and Bankwide portfolios-the corresponding ratings on the various performance categories were then built up from individual project ratings. Under the Form 590 system, projects are rated on a four-point scale, with 1 representing the best rating and 4 the worst rating. A DO rating of 1 means that all project development objectives are expected to be substantially achieved. A rating of 2 means that major objectives are expected to be achieved; although some minor ones may not be met, the project continues to be justified. A rating of 3 indicates that major project objectives are likely to be achieved only partially and that the continued justification of the project is uncertain. A rating of 4 indicates that major objectives will probably not be achieved and that the project appears to be no longer justified. The DO rating is used in this report as a measure of the expected development results of a project. The average DO rating of projects comprising a portfolio, and the percentage of projects with a DO rating of 1 or 2, are used to indicate the expected development results of that portfolio.' Taking into account the DO rating and the ratings on the other performance categories (essentially the delivery of various project inputs and outputs), a project is given an OS rating.' An OS rating of 1 means no significant problems; 2, moderate problems; 3, major problems, with, however appropriate actions being taken to address these problems; and 4, major problems that are not being adequately addressed. The Form 590 guidelines call for DO ratings to be reflected in the OS ratings. However, a recent joint analysis by the Operations Policy Department (OPR) and the Operations Evaluation Department (OED) of the divergence between ARPP ratings and ratings on project completion reports (PCRs) and project audit reports (PARs) for a sample of 1,030 projects approved and implemented in the 1980s concluded that the OS ratings for the sample seemed to be dominated by short-term concerns about project implementation. Therefore, the OS ratings were not a fully reliable measure of the extent to which projects would eventually meet their development objectives. The analysis also found that for a large number of projects in the sample studied, the DO ratings were generally more optimistic than the OS ratings, reflecting longer term optimism s. Currently, the DO ratings are based on judgments about the expected development results of projects under implementation. Expected project development results are not always clearly or precisely defined, which adds to the problem of interpreting the DO ratings. Comparisons of DO ratings between projects, or the aggregation of project DO ratings into a portfolio index of development effectiveness, should therefore be interpreted with caution. 6. These project inputs and outputs include project management, availability of funds, and compliance with legal covenants, which are required to be rated, and procurement, training, TA, studies, environmental aspects and other categories, which are rated at the option of the Region. The overall status rating summarizes the staffs judgment on all the items rated, on physical progress in project implementation, and on the disbursement, closing date, and cost records. 7 that objectives will be achieved once implementation problems have been resolved. The DO ratings for a significant number of projects were also optimistic in relation to the eventual performance ratings given at completion (in PCRs and PARs).' In light of the above findings, and consistent with the PMTF recommendation to focus on the development results of Bank-assisted projects, the guidelines for the fiscal 1993 ARPP stressed the need for staff to improve the reliability of the DO ratings within the Form 590 system. Departing from past portfolio reviews that relied primarily on the OS ratings as a performance measure, the guidelines required staff to use the DO ratings as another principal measure of portfolio performance. Staff were also asked to pay special attention to the rating of projects in countries that were found to have had large differences between ratings during supervision (on OS and DO) and ratings at completion. In addition, any significant gap in assessments based on OS and DO was to be explained. As noted below, the greater attention of staff to the DO ratings for the fiscal 1993 ARPP has helped to improve the reliability of these ratings. However, key elements in further and substantial improvement in their reliability are the precise definition of the expected development results for each project and better means for gauging progress toward these results during the loan disbursement or implementation period. In principle, progress should be assessed using the same cost-benefit or cost-effectiveness criteria that are applied at project appraisal or at completion. If cost were not a constraint, this would involve repeating periodically during project implementation the original cost-benefit or cost-effectiveness analysis done at appraisal with revised or updated data or estimates. As a practical and much less costly alternative, the PMTF report recommended that progress toward expected development results be monitored on the basis of a limited number of variables ("performance indicators") that have been identified in the risk and sensitivity analysis at appraisal as critical to achieving the expected project development results. This requires improvements in risk and sensitivity analysis and performance indicator monitoring. In the context of "Next Steps" to the PMTF report, work is proceeding on, among other measures, developing guidance to staff on risk and sensitivity analysis and performance indicators. The project rating methodology and project performance rating system are expected to be revised on the basis of this work by June 1994. Training programs for staff in applying the new risk and sensitivity guidelines, performance indicators, and project rating methodology and performance rating system would be developed as soon as work in these areas has been completed. Full implementation of these improvements would realistically take a few years, given the sharp departure from current practices that is entailed and the large number of projects in the Bank portfolio. Meanwhile, the next ARPP (covering fiscal 1994) would seek to increase the emphasis on assessing the expected development results or effectiveness of the Bank portfolio by making the following changes: (a) the Form 590 rating system would be replaced with a dual rating system that would give separate overall ratings to the implementation status of a project and to its development effectiveness (that is, development objectives ratings would no longer be subsumed as part of the 7. It is worth noting that the OS, DO, and PCR ratings are all done by Regional staff. PAR ratings are done by OED staff. For the sample of 926 investment projects covered by an OPR/OED analysis, 30 percent have been audited by OED. Where it was possible to distinguish in OED's database between the PCR and PAR ratings for audited projects, PCR and PAR ratings differed in only 8 percent of the cases. 8 overall status ratings); (b) newly approved projects would not be rated on development effectiveness, thus avoiding the optimistic bias of ratings at entry; (c) projects would be rated on development effectiveness only after the first year of implementation when a better basis for the rating would likely exist; (d) existing OED guidelines for evaluating the development results of completed projects (which are consistent with guidelines to be issued shortly to staff on the new Implementation Completion Reports or ICRs) would be applied to the assessment of development results of active projects; and (e) the last supervision missions for projects that are about to close would pay special attention to the assessment of the development effectiveness of these projects in order to minimize the gap between the last supervision ratings and the eventual ratings in ICRs at project completion. These changes are expected to improve further the reliability of the development objectives ratings for the fiscal 1994 ARPP. Performance Measures In this report, the percentages of projects rated 3 or 4 on OS and on DO (hereafter called "problem projects") are used as the principal performance measures. Because the percentages of problem projects can range, in principle, from 0 to 100, whereas the average OS or DO ratings have a more compressed range, from 1 to 4, the former lend themselves more readily to interregional, intercountry, or intersectoral comparison. In addition, they can be compared more easily with the percentage of unsatisfactory projects based on PCR and PAR ratings. 9 3. Portfolio Performance by Region Overall Status The percentage of problem projects (OS) for the Bank portfolio as a whole peaked at 20.3 percent in fiscal 1991 before falling to 18.3 percent in fiscal 92 and 17.8 percent in fiscal 1993 (Table 4). An even greater improvement is shown when countries that presented a completely different order of portfolio management problems in fiscal 1993 (that is, were in nonaccrual status, suffered serious civil conflict, or both) are netted out of the portfolio. On this adjusted basis, the percentage of problem projects declined from 18.4 percent in fiscal 1991 to 15.3 percent in fiscal 1993. TABLE 4 PORTFOLIO PERFORMANCE BY REGION, FY91-FY93 FY91 FY92 FY93 % Problem Proi. % Problem Proi. % Problem Proi. No. ot Overall Devel. No. o Overall Devel. No.o Overall Devel. ProL. Status b Pro. Status Obi. Status Ob. Africa 641 29.2 22.6 644 25.6 19.6 656 26.2 19.5 East Asia 309 3.9 3.2 319 3.1 1.6 326 4.3 2.1 South Asia 277 19.1 6.9 275 17.5 8.0 277 15.9 12.6 Europe & Central Asia 100 17.0 8.0 101 17.8 12.9 120 22.5 17.5 Middle East & North Africa 179 15.6 6.7 181 17.7 9.9 178 17.4 13.5 Latin America 339 22.7 12.1 332 19.9 10.5 340 14.7 9.1 Bankwide Total 1,845 20.3 12.7 1,852 18.3 11.8 1,897 17.8 13.0 al Others 135 43.7 28.9 133 45.1 32.3 124 53.2 42.7 Bankwide Total (net) 1,710 18.4 11.5 1,719 16.2 10.2 1,773 15.3 10.9 a. Countries that were in non-accrual status and/or suffered from civil conflict in FY93 (Angola, Chad, Congo, Liberia, Rwanda, Somalia, Sudan, Togo, and Zaire; former Yugoslavia; Haiti). Between fiscal 1992 and fiscal 1993, the overall number of problem projects in the Bank portfolio decreased by only one project, from 339 to 338 (Table 5). Excluding countries that were in nonaccrual status, suffered from civil conflict in fiscal 1993, or both, the number of problem projects decreased from 279 to 272. The negligible change in the total number of problem projects masks significant offsetting changes within the portfolio. This change can be decomposed into (a) the net effects of exits and entries and (b) upgradings and downgradings in ratings because of changes in performance and project age (Table 5). * Thirty-eight problem projects were canceled or closed in FY92, while entries into the portfolio in fiscal 1993 included only four problem projects, resulting in a net reduction of 34 problem projects. This alone would have reduced the percentage of problem projects to 16.1 percent of the fiscal 1993 portfolio, even if nothing else had changed. * The net effect of performance and aging factors was an increase in the number of problem projects in fiscal 1993 by 33, nearly offsetting the net reduction of 34 problem projects from exits and entries. This increase of 33 problem projects was the result of a downgrading in rating for 124 projects, mostly from 2 in fiscal 1992 to 3 in fiscal 1993, and an upgrading in ratings for 91 projects, mostly from 3 to 2. The upgradings included 19 projects that were restructured in fiscal 1993. In the absence of these restructurings, the overall percentage of 10 problem projects would have increased to 18.8 percent in fiscal 1993, instead of decreasing to the actual level of 17.8 percent. In addition, it is important to note that Regions were generally conservative in rating restructured projects. The 19 restructured projects noted above were a small part of some 140 projects that were being restructured in fiscal 1993. Ratings for most of these projects have not been upgraded, pending clear evidence of improved performance based on their restructured (and generally simpler) objectives and designs. Thus the full impact of project restructurings on quantitative measures of portfolio performance is likely to show up only in future years. TABLE 5 EFFECIS OF EXITS, ENTRIES, UPGRADINGS AND DOWNGRADINGS ON THE NUMBER OF PROBLEM PROJECIS (OS RATINGS) BY REGION Increases (+Decreases (-) in the Number of Problem Projects Exits fro Entries grading Downgradings Net FY92 FY92 into F193 Net Between Between Upgradings/ FY93 Actual Poafol 82rfolia Eisthem FY92F392 &Me& FM93 Downgradings Actual Africa 165 -16 0 -16 39 62 23 172 East Asia 10 0 0 0 b/ 8 12 4 14 South Asia 48 -2 2 0 c/ 18 14 4 44 Europe & Centml Asia 18 -2 0 -2 3 14 11 27 Middle East & North Africa 32 4 0 -4 5 8 3 31 Latin America 66 -14 2 -12 18 14 -4 50 Bankwide Total 339 -38 4 -34 91 124 33 338 Others at 60 -9 0 -9 2 17 15 66 Bankide Total (net) 279 -29 4 -25 89 107 18 272 a. Countries that were in non-accrual status and/or suffered from civil conflict in FY93 (Angola, Chad, Congo, Liberia, Rwanda, Somalia, Sudan, Togo, and Zaire; former Yugoslavia; Haiti). b. None of the new and cancelled/closed projects were rated problem projects. c. The number of problem projects among cancelled/closed projects was equal to the number among new projects. Performance trends (based on OS ratings) between fiscal 1992 and fiscal 1993 varied between Regional portfolios (see Table 4). The most notable changes were as follows: * The percentage of problem projects in the ECA Region increased significantly, reflecting the difficulties experienced by countries in transition in carrying out economic reforms and the aging of its relatively young country portfolios. Growth of gross domestic product (GDP) in the Region was -14 percent in 1992. As shown in Table 5, the increase in the number of problem projects resulting from a deterioration in performance and aging greatly exceeded the reduction due to exits and entries. * The percentage of problem projects in the LAC Region declined significantly as a result of the cancellation or closing of problem projects (exits). Portfolio performance also improved, reflecting the success of the Region in maintaining positive GDP growth for the second consecutive year despite the recession in Brazil in 1992, as well as the positive effect of project restructurings. This performance improvement added to the relatively large reduction in the number of problem projects resulting from exits (Table 5). The South Asia and Middle East and North Africa (MENA) portfolios experienced small improvements in performance in fiscal 1993, while the Africa and EAP portfolios showed a small deterioration in portfolio performance. 11 Development Objectives Unlike the OS ratings, the DO ratings deteriorated in fiscal 1993, as indicated by an increase in the percentage of problem projects (DO) in the Bank portfolio, from about 12 percent in fiscal 1992 to 13 percent in fiscal 1993 (Table 4). At the Regional level, the percentage of problem projects (DO) in the South Asia and MENA portfolios increased despite a relatively stable or decreasing percentage of problem projects as measured by overall status. The percentage of problem projects (DO) also increased in the ECA Region, accompanying an increase in the percentage of problem projects as measured by overall status. The Regional percentages of problem projects as measured by development objectives were all less than the corresponding percentages as measured by overall status. The Africa, LAC, and ECA Regions had the largest gaps, ranging from 5 to 7 percent of their respective portfolios. The gaps reflect a widely held expectation that once implementation problems are resolved, projects would eventually be able to meet their development objectives. This expectation is anchored on the various actions being carried out by the Regions to address implementation issues. These actions are discussed in the Regional chapters and summarized in the discussion of portfolio performance management below. The above increase in the overall percentage of problem projects (DO) does not necessarily indicate a real deterioration in the development performance of the Bank portfolio. Rather, it reflects Regional efforts, within the limitations of the current rating methodology, to base DO ratings on more realistic expectations about the extent to which project development objectives would be achieved. In making these judgments, Regional staff took into account, whenever applicable, the findings of the OPR/OED study on the disconnect between supervision and completion ratings for projects implemented in the 1980s. The increased attention to the DO ratings is further shown by the increasing trend over fiscal 1991-93 in the percentage of problem projects in the age group nearing completion (six to eight years), from 13 percent in fiscal 1991 to about 18 percent in fiscal 1993. The ratings for this particular age group, rather than the ratings for all age groups, are better suited for comparison with the ratings given to projects at completion.! For projects in this age group, the percentage of problem projects (DO) in fiscal 1993 was 12 percentage points lower than the percentage of projects rated unsatisfactory in PCRs and PARs for completed projects approved mostly during the first half of the 1980s. This gap in performance between these two different sets of projects is significant. However, because PCRs and PARs for projects nearing completion in fiscal 1993 have yet to be written, it remains to be seen whether the gap between supervision DO ratings for these projects and the ratings in their eventual PCRs and PARs will be larger or smaller than 12 percentage points." At the Regional level, the gap between the percentage of problem projects in fiscal 1993 for the six- to-eight-year age group and the percentage of completed projects rated unsatisfactory was minimal 8. The average age of a project at completion is almost seven years. 9 OED preliminary projection, based on a small sample of projects approved after 1983 (which may or may not be representative of the total population) suggest that ratings for at least two or three subsequent approval years could deteriorate before showing signs of a modest improvement ("Annual Review of Evaluation Results 1992, World Bank", October 13, 1993). 12 or relatively small for ECA, MENA, and Africa but large for LAC, South Asia, and EAP. While the implementation environment for projects in many countries in the last three Regions has improved in recent years because of better economic management and successful adjustment to the external shocks of the first half of the 1980s, it is not possible to conclude on the basis of current data that the failure rates for these Regions in the future would be much less than the respective historical failure rates. In this context, it should be noted that applying the historical failure rates to the six-to-eight-year age group in the South Asia and EAP portfolios would result in relatively small increases in the numbers of problem projects beyond those indicated by the percentage of problem projects in fiscal 1993. 13 4. Portfolio Performance by Sector The performance in fiscal 1993 of the portfolio by sector, as measured by the percentage of problem projects, appears mixed (Table 6). Based on overall status, there was a further slight improvement in fiscal 1993. This was the net result of improvements in agriculture, water supply and sewerage, telecommunications, power, PHN, TA, and public sector management, and deteriorations in transportation, urbanization, IDF, industry, energy (oil and gas), education, program lending and structural adjustment lending (SAL). Performance of most sectors is now better than in fiscal 1991, except for urbanization, industry, energy, PHN, public sector management, and SAL. TABLE 6 PORTFOLIO PERFORMANCE BY SECTOR, FY91-FY93 FY91 FY92 FY93 % Problem Pro. % Problem Proi. % Problem Proj. No._of Overall Devel. No.o Overall Devel. No.o Overall Devel. Proi. Status Obi Pro Status Ob. Proj Status Ob. Agriculture & Rural Devt 513 26.1 17.9 502 22.4 13.6 482 21.8 14.9 Transportation 220 16.9 9.6 213 14.6 10.3 220 15.5 10.0 Urbanization 103 17.3 11.5 111 17.0 11.6 116 18.1 12.9 Tourism 0 0.0 0.0 0 0.0 0.0 2 0.0 0.0 WaterSupply and Sewerage 98 22.7 10.3 96 26.0 10.4 101 17.0 13.0 Industrial Devel/Finance 159 17.1 13.9 150 14.0 12.7 135 15.6 11.1 Industry 66 16.7 10.6 59 18.3 16.7 57 22.8 22.8 Communications 33 17.6 5.9 31 12.5 6.3 34 11.4 2.9 Power 153 22.1 9.1 156 20.6 9.0 157 20.3 14.6 Energy 57 10.5 5.3 54 17.0 9.4 51 19.2 11.5 Education 170 17.6 7.1 180 13.4 11.2 197 14.2 9.6 Population Health & Nutrition 98 13.3 6.1 115 18.4 7.9 133 15.2 9.8 Technical Assistance 94 27.7 25.5 98 16.2 14.1 112 14.3 15.2 Public Sector Management 13 15.4 15.4 17 17.6 17.6 25 16.7 16.7 ProgramLending/SAL 68 11.8 11.8 70 15.5 14.1 76 17.1 17.1 Bankwide Total: 1,845 20.3 12.7 1,852 18.3 11.8 1,897 17.8 13.0 Based on development objectives, the indicators show some deterioration in performance overall, notably in water supply and sewerage, industry, power, energy, PHN, and SAL. The gap between performance, as measured by overall status and by development objectives, has closed somewhat but is still wide. The overall status and development objectives indicators continue to give a more favorable picture of performance of the portfolio than the historical PCR and PAR ratings. Taking only the performance of the portfolio in the later years (six to eight years), which correspond more closely to the age of projects covered by PCRs and PARs, reduces but does not eliminate the overall gap between the ratings. Gaps remain for agriculture and rural development, urbanization, water and sewerage, telecommunications, power, education and PHN, and TA. Some sectors performed better than others in fiscal 1993. Telecommunications, education, TA, PHN, transportation and IDF all have a significantly lower percentage of problem projects than average (in terms of overall status), and agriculture, industry, and power a significantly higher percentage. But the variation in performance by sector is much less than that by Region, and a fortiori that by country. This suggests that performance problems tend to be country-specific rather than sector-specific. 14 5. Other Aspects of Portfolio Performance Problem Projects Problem projects in fiscal 1993 numbered 338, including 60 that were rated 4 on overall status (that is, projects with major problems that were not being adequately addressed). Of the projects rated 4, 29 were in countries that were in nonaccrual status, suffered civil conflict in fiscal 1993, or both, and hence were effectively inactive in that year. Formal actions to close, cancel, or, in some cases, restructure the projects will be taken once relationships with the Bank are normalized or country conditions improve to the point where normal Bank supervision of projects can resume. Thirty-one of the projects rated 4 are in other, mainly problematic, country portfolios; of these, 17 have already been closed, and the rest are expected to be closed or canceled soon. In fiscal 1993 the Regions intensified their efforts to restructure problem projects and to cancel or close projects for which the prospects for improvement were poor. By the end of fiscal 1993, the restructuring had been completed or was under way for 104 projects and was being planned for 36 others. About 77 problem projects (including 13 in countries that were in nonaccrual status, suffered civil conflict in fiscal 1993, or both) were closed or canceled in fiscal 1993, compared with 38 in fiscal 1992. The project restructurings, closings and cancellations in the past two years have had a positive impact on the age structure of problem projects, with the greatest improvement in the proportion of projects that have been rated as problems for two consecutive years. However, over two-fifths of problem projects in fiscal 1993 have been in problem status for three or more consecutive years, indicating the Bank's willingness and persistence to remain engaged with borrowers in restructuring problem projects and helping to make them work (examples include power projects in India and Turkey). As already noted, the Regions have been conservative in rating projects, and pending clear evidence of improved performance, most projects that have been restructured have retained their problem project rating. Since July 1993 the Regions have adopted a more structured approach to reviewing the performance of projects that have been in problem status for over 12 months. For such projects, departmental managers have been required to agree with borrowers on action plans that specify the basis on which Bank support for the projects would continue. If the action plan is not implemented within the subsequent 12 months, the project involved will be canceled or restructured. Implementation of this rule in fiscal 1994 is expected to lead to further significant improvements in the age structure of problem projects. Disbursement Performance The disbursement performance of the Bank portfolio improved in fiscal 1993. Total disbursements for both investment and adjustment operations in fiscal 1993 increased to over $17.8 billion, an increase of $1.3 billion over the fiscal 1992 level. After declining by more than $0.2 billion in fiscal 1992, disbursements for investment operations increased by more than $0.8 billion in fiscal 1993, resulting in an increase in the disbursement ratio from 17.1 to 17.6 percent.' This increase was the result of a large increase in disbursements combined with a less than proportionate increase in The disbursement ratio is the ratio of disbursements of investment loans during the fiscal year to the undisbursed balance of investment loans at the beginning of the fiscal year. 15 undisbursed balances. The MENA Region experienced the largest increase in the disbursement ratio, from 14.1 percent in fiscal 1992 to 19.1 percent in fiscal 1993. The disbursement ratio for the Africa and EAP and Pacific Regions remained stable at 15.0 and 25.8 percent, respectively. The LAC ratio decreased from 18.7 to 18.2 percent, while the ECA disbursement ratio decreased from 11.5 percent in fiscal 1992 to 10.6 percent in fiscal 1993. Disbursement performance also varied across sectors. The urban sector experienced the largest increase in the disbursement ratio (from 12.6 percent in fiscal 1992 to 21.2 percent in fiscal 1993). The agriculture, industry, industrial development and finance and energy sectors experienced a small decline in the disbursement ratio. Although registering a small increase, the disbursement ratios for PHN and public sector management sectors (10.8 and 7.3 percent, respectively) remained well below the Bank average. Financial Accountability The Audit Reports Compliance System shows that the Bank eventually receives about 93 percent of audit reports required, with the balance accounted for by problem country portfolios and problem projects that are eventually terminated. However, while the number of reports received from most countries has steadily increased, majority of reports are persistently delayed. A review of audit compliance in fiscal 1993 showed that about 40 percent of reports due were received within the grace perioda given to project implementing agencies of six to nine months after the end of each financial year for the submission of audit reports. More importantly, report quality varied widely from borrower to borrower and from project to project. Reports were more often narrowly oriented to comply with audit covenants than to assist project implementation and management. Between 15 and 20 percent of audit reports received had a qualified or adverse opinion or a disclaimer. About 50 percent of all audits were carried out by government audit agencies (from a low of 23 percent in the Africa Region to a high of over 80 percent in the EAP Region), many of whom lack independence and skilled staff, particularly for the audit of commercial-type entities. Actions taken in fiscal 1993 to strengthen borrower financial accountability are summarized in the next chapter. 16 6. Portfolio Performance Management On the basis of an analysis of 545 completed operations, the 1992 OED evaluation report confirmed the PMTF findings about the importance of quality at entry and sound portfolio management to ensure satisfactory outcomes at completion. Various practices in these directions have been carried out by the Regions in fiscal 1993. These changes have been facilitated by greater organizational support for portfolio performance management. In addition, the amount of staff time devoted to supervision has increased, and Resident Missions have assumed greater supervision responsibilities. Improving Project Quality at Entry Measures taken by the Regions to achieve this objective include the following. * In fiscal 1993, many departments sought to ensure that key measures critical for successful implementation were taken prior to Board presentation. Detailed supervision plans were also prepared and agreed on at negotiations. In a number of cases, the processing of several major operations was halted pending up-front action on policy reforms, procurement, and project management. * Special launching missions and project launch workshops were used with increasing frequency to get projects off to a sound start. In the case of decentralized projects in some countries, a special effort was made to conduct project launch workshops at the state or local level. * The reorganization of the Regional Technical Departments in mid-1993 resulted in the allocation of more resources to upstream quality enhancement. * Simplifying the design of new projects was a major priority. The Regions carried out departmental reviews of all projects in the pipeline to ensure that designs were simple and in keeping with the institutional, technical, and financial capacity of the borrowers. * All Regions intensified their efforts to encourage participation of all stakeholders throughout the project cycle. Thirty percent of the projects approved by the Board in fiscal 1993 included participation by non-governmental organizations (NGOs). In some countries, seminars were held with opposition politicians, legislative bodies, business groups, industrialists, local economists, and NGOs. A Participation Fund was established in June 1993 to encourage the development of new ideas to improve development impact and business practices through increased beneficiary participation. * Greater efforts were made to enhance borrower ownership, with emphasis on strengthening local institutional capacity for project implementation. Expatriate technical assistance in the Africa Region, for example, is being replaced by local technical assistance wherever possible, and training programs for local project managers have been carried out in a number of countries. * There was greater effort to link new lending to portfolio performance and economic and sector work. Country assistance strategies explicitly took into account and reflected the outcomes of recently completed portfolio performance reviews and set performance benchmarks for portfolio 17 management. Sectoral strategies have improved greatly in several sectors, and this has contributed to better project technical design. Efforts to integrate lessons from previous projects into new project design intensified in fiscal 1993. The practice of requiring every initial project proposal (as summarized in an Initial Executive Project Summary) to show how lessons from previous projects have been incorporated into the project concept was continued. Peer reviews and brainstorming sessions were successfully used to exchange information on approaches that had been useful in other contexts, as well as to avoid repetition of mistakes. In the Africa Region several organizations were formed to facilitate the exchange of information on implementation problems and solutions. These included project managers clubs (The Gambia, Mali, Senegal, and Uganda) and an intercountry association of agencies of the Agence d' Execution de Travaux d'Int6r&t Publique type." Coordinators of National Environmental Action Plans (NEAPs) also created an organization to provide support to countries newly involved in NEAPs. Strengthening Portfolio Management Practices In addition to focusing on improving the quality of new projects, a variety of measures were taken to improve the performance of the existing portfolio. * Borrower ownership was fostered by encouraging joint supervision of projects and economic sector work. In some countries, borrowers drafted supervision mission aide-memoires, including progress reports and updates on covenant compliance and previously recorded agreements. In other countries, borrowers worked with the Bank in designing and organizing economic and sector work. These measures have also helped to reduce costs. * The Regions intensified their efforts to restructure problem projects in fiscal 1993, and to cancel or close projects where prospects for improvement were poor.12 In fiscal 1993, the restructuring of 79 projects was completed. In addition, at the end of the fiscal year a further 25 projects were in the process of being restructured, while 36 other restructurings were being planned. The amount of loan cancellation that is expected to result from completed and planned restructurings is estimated to total some $2.6 billion, distributed as follows: South Asia Region ($1.2 billion largely due to cancellations in the India portfolio); LAC and Africa Regions (about $0.5 billion each); ECA Region ($0.4 billion); and EAP ($37 million). Restructuring of projects in the MENA Region did not involve any cancellations in fiscal 1993, but some cancellations are expected in fiscal 1994. Restructurings and cancellations have helped to emphasize to borrowers the importance of effective implementation. In some Regions this message has been reinforced by a policy of not making new loans in sectors or countries with outstanding problems in the existing portfolio. * AGETIPs are autonomous nongovernmental project executing agencies responsible for project engineering, tendering, and contact administration. The first AGETIP was created in Senegal in 1990 in the context of the Public Works and Employment Project. Similar arrangements have been introduced in a number of African countries. 12. IDA and IBRD cancellations amounted to $4.5 billion in fiscal 1993 and involved total or partial cancellation of 277 projects, compared with cancellations totaling $3.2 billion and involving 231 projects in fiscal 1992. Country reasons or request by the borrower accounted for almost half of the total amount canceled in fiscal 1993. Undisbursed balances at closing date and cost savings, accounting for cancellations of over $1.0 billion, constituted the second most important reason for cancellation followed by noncompliance with project conditions ($0.4 billion) and restructuring ($0.4 billion). 18 * Efforts have also focused on reducing delays in project completion. This has largely involved tightening closing date management. Project extensions are no longer granted as a matter of course. To receive an extension, the borrower must demonstrate convincingly that the project will be completed during the agreed extension period, and the executing agency must have a record of satisfactory performance. * Counterpart funding has continued to be problematic and was addressed frequently in CPPRs with borrowers. In some countries, workshops were held to draw attention to the importance of this issue. In other countries, economic management and institutional development projects have been launched to improve the government's capacity to mobilize and allocate resources. Individual countries, particularly in Africa, have been encouraged to adopt innovative measures. For example, in Ghana the possibility is being explored of pooling counterpart funds in a centrally administered account to reduce the incidence of funds that are committed but unused. * Procurement has continued to be a major constraint on effective implementation. To deal with this problem, the Regions have continued many of the measures they adopted in the past fiscal year and have also initiated new efforts. Thematic supervision focusing on procurement has proved a cost-effective way of dealing with procurement problems and is increasingly being tried in a number of Regions (for example, LAC and Africa). The use of standard bidding documents is now common practice in all Regions. In addition, all Regions have hired additional staff to improve procurement processing and to conduct training in borrowing countries. Procurement workshops were held in a number of countries (23 in Africa alone). Country Procurement Assessment Reviews continue to be given priority and have formed the basis of dialogue on procurement reform with many governments. Some projects in Africa included procurement reform components, and a number of projects funded by the Institutional Development Fund are directed at improved procurement practices. * Inadequate borrower accountability caused by weak financial management, including project accounting, has also constrained project implementation. Measures taken in fiscal 1993 to address this bottleneck are wide-ranging. In LAC the increased use of audit letters has produced good results. The MENA Region issued departmental audit action plans and was quite successful in improving audit compliance. In a number of countries (for example, China, Indonesia, Nepal and Pakistan), the Bank is providing technical assistance, through freestanding loans or the Institutional Development Fund, for improving accounting and auditing capacity and practices. Where appropriate, borrowers have been encouraged to use private auditors, particularly for commercial-type enterprises. There has also been increased emphasis on enforcing audit covenants. Regions have resorted to suspending disbursement and to making receipt of audit reports a condition of effectiveness of new loans or of extension of loan closing dates. Several divisions have appointed audit coordinators. A Task Force on Financial Reporting and Auditing has been established to strengthen accountability in project performance. * In addition to the initiatives being undertaken in the context of strengthening project implementation, the Bank continued to provide financial support for public sector management projects. In fiscal 1993 the Board approved ten public sector management operations, an 19 increase from the six approved in fiscal 1992." These projects are designed to help strengthen the institutional framework and create a better implementation environment for future projects. * A number of measures were taken to improve the effectiveness of technical assistance. Conditionality has been strengthened for the appointment of critical consultants; consultant terms of reference now routinely include provisions requiring them to transfer skills to their counterparts; supervision of TA components has increased; and, in some Regions, a separate write-up on TA in supervision and Implementation Completion Reports (which have replaced the PCRs) is now routine practice. All Regions carried out a review of their overall approach to TA in fiscal 1993. * Midterm reviews of projects were carried out with increased frequency in all Regions and have been a useful forum for discussing the restructuring of projects or cancellation of unproductive components. Some Regions have systematically required midterm reviews in all loan agreements. * CPPRs and other implementation reviews were carried out for 39 country portfolios. * Supervision focused on specific themes or issues has worked quite effectively in fiscal 1993. Thematic reviews were used primarily for the supervision of procurement activities. However, it was also used to *deal with auditing, environmental, and technical assistance issues. The South Asia Region used sectorwide supervision strategies as the basis for allocating supervision resources. Strengthening Organizational Support for Portfolio Management Portfolio performance is a function not only of sound management and quality project design but also of a supportive organizational framework. The following measures aimed at strengthening organizational support for portfolio management were taken in fiscal 1993. * Efforts to enhance the role of Bank Resident Missions (RMs) continued. In several departments task management and supervision responsibilities for a number of projects were transferred to RMs. Some RMs now have primary responsibility for procurement, or donor coordination, or both. The staff profile of many RMs also changed considerably in fiscal 1993, with an increased focus on the recruitment of local higher-level staff. The number of headquarters staff transferred to RMs has also increased. Emphasis has been placed on the training of RM staff. Operational communication with RMs has greatly improved with the expansion of the electronic mail system. * Management involvement in implementation issues continued to increase. Regional Vice Presidents, department directors, project advisers, and division chiefs are regularly involved in reviews of portfolios and in the resolution of specific problems as they arise. * Staff involvement in implementation has increased. The Africa Region held a three-day workshop, "Getting Results in the Field," aimed at soliciting staff recommendations for 1 The numbers refer only to projects that were classified under the public sector management sector. However, a number of public sector management operations are also classified under the TA sector. Thus, the numbers reported above underestimate the extent to which the Bank is supporting institutional development in its lending operations. 20 improving quality at entry and project implementation. In addition, the Region also carried out surveys on ways to streamline procedures and established a departmental quality advisory group. Regions have also redesigned staff work programs and allocated more resources to implementation support tasks. * Steps have also been taken to increase the incentives for better implementation results. Performance reviews and merit increases more and more take implementation performance into consideration. * Some Regions have taken steps to strengthen the skills mix of country teams. Efforts to strengthen implementation skills through well-focused training of existing staff and recruitment of much-needed specialists continued in fiscal 1993. * Portfolio manager positions were created in a number of divisions in the Bank. Some units adopted the policy that the same staff who have appraised and negotiated a loan will also supervise it. Many units have made an effort to define and monitor "mileposts" to improve monitoring of implementation progress. Portfolio Management Resources The trends in fiscal 1993 in the amount and deployment of supervision resources are summarized below. * Staff resources allocated to supervision increased in fiscal 1993, continuing the upward trend of the previous two years. Average supervision inputs per project increased by over 9 percent, from 14.1 staff-weeks in fiscal 1992 to 15.4 staff-weeks in fiscal 1993. Supervision inputs for the one-fourth of the portfolio that was intensively supervised in fiscal 1993 averaged 36 staff- weeks. Actual Regional allocations to portfolio management in the fiscal 1993 administrative budget exceeded the budget contract amount by 13 percent. These allocations amounted to $94.3 million in fiscal 1993 (excluding staff benefits), an increase of 11 percent in real terms over the fiscal 1992 level. The allocation of supervision inputs by age of project reflects greater emphasis on preventive rather than curative care in the implementation of the Bank's portfolio. An average of 19 staff-weeks per project was allocated for projects in the one-to- three-year age group, compared with 11 staff-weeks in the six-to-eight-year age group. * To deal with portfolio problems, the level of (direct) supervision has been increased in the past few years in nearly all sectors. Supervision inputs per project are now at an all-time high, exceeding by more than 9 percent the previous peak reached in fiscal 1992. The increase has been especially sharp in TA, energy, PHN, public sector management, and education. The fiscal 1993 sectoral allocation of supervision inputs by Region, however, corresponds only very loosely to the overall allocation by sector. For example, the supervision input for PHN, relative to regional average inputs, is high in all Regions, but for SAL it is low in several Regions, especially the EAP Region. Some of the regional and sectoral average inputs are based, however, on very small numbers of projects, especially in ECA and MENA. * Emergency reconstruction loans required the most supervision-30.4 staff-weeks per operation in fiscal 1993, up from 21.6 staff-weeks in fiscal 1992. Adjustment lending also required considerable supervision inputs-23.2 staff-weeks for SALs and 18 staff-weeks for sectoral adjustment loans. Supervision of investment lending increased in fiscal 1993 to 15.5 staff- 21 weeks from the fiscal 1992 level of 14.1 staff-weeks. The increased attention given to improving the effectiveness of TA across the regions is evident in the increased staff time allocated to the supervision of Technical Assistance Loans from 12.3 staff-weeks in fiscal 1992 to 14.2 staff-weeks in fiscal 1993. * Increased allocations for supervision were recorded in all Regions. South Asia allocated the largest amount of resources for supervision-18.7 staff-weeks per project, followed by ECA and Africa at 17.5 staff-weeks and 17.0 staff-weeks per project, respectively. Reflecting the lowest incidence of problem projects in its portfolio, the EAP Region spent an average of 11.1 staff- weeks on supervision in fiscal 1993. * Resident Missions have assumed greater responsibility for portfolio management. In fiscal 1993, Resident Missions were allocated 15.5 percent of total supervision resources, up from 13.6 percent of a smaller supervision program in fiscal 1992. Actions Planned in Fiscal 1994 The various actions that the Regions plan to carry out in fiscal 1994 aim at intensifying the efforts made in fiscal 1993 to improve quality at entry and strengthen portfolio performance management. Emphasis is expected to be placed on integrating portfolio management with country assistance strategy, encouraging participation of all stakeholders in all phases of the project cycle, and simplifying and tailoring project design to local institutional, technical, and financial capacities. Emphasis will also continue to be placed on project restructuring and, as appropriate, loan cancellation. The Regional action plans will be carried out within the context of a Bankwide program to implement the recommendations of the PMTF report. The World Bank Headquarters European Office Tokyo Office 1818 H Street, N.W. 66, avenue d'I6na Kokusai Building Washington, D.C. 20433, U.S.A. 75116 Paris, France 1-1, Marunouchi 3-chome Chiyoda-ku, Tokyo 100, Japan Telephone: (202) 477-1234 Telephone: (1) 40.69.30.00 Facsimile: (202) 477-6391 Facsimile: (1) 40.69.30.66 Telephone: (3) 3214-5001 Telex: wul 64145 WORLDBANK Telex: 640651 Facsimile: (3) 3214-3657 RCA 248423 WORLDBK Telex: 26838 Cable Address: INTBAFRAD WASHINGTONDC ISBN 0-8213-2828-X