Report No. 16594 Adjustment Lending in Sub-Saharan Africa: An Update May 29, 1997 Operations Evaluation Department Document of the World Bank Abbreviations and Acronyms CAR Country Assistance Review CFA Communaute Financi&re Africaine (African Financial Community) CFAF African Financial Community Franc ERC Economic Recovery Credit ESW Economic Sector Work FDB Financial Data Base FSAC Financial Structural Adjustment Credit FY Fiscal Year GDP Gross Domestic Product GFS Government Financial Statistics IBRD International Bank of Reconstruction and Development ICR Implementation Completion Report IDA International Development Agency IMF International Monetary Fund M&E Monitoring and Evaluating OED Operations Evaluation Department PAR Performance Audit Report PER Public Expenditure Review PSD Private Sector Development PSM Public Sector Management SAC Structural Adjustment Credit SAL Structural Adjustment Loan SECAL Sectoral Adjustment Loan SIAO Social Impact of Adjustment Study SPA Special Program of Assistance (for SSA) SSA Sub-Saharan Africa Director-General, Operations Evaluation Mr. Robert Picciotto Acting Director, Operations Evaluation Dept. Mr. Roger Slade Acting Division Chief : Mr. Rene Vandendries Task Manager : Mr. Manuel Peialver The World Bank Washington, D.C. 20433 U.S.A. Director-General Operations Evaluation Hay 29, 1997 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Adjustment Lending in Sub-Saharan Africa: An Update Adjustment lending to SSA countries exceeds $15 billion, through 163 operations in 37 countries, of which 121 have already been evaluated at completion. A previous OED Report, prepared in 1993 (Adjustment Lending in Sub-Saharan Africa: Selected finding from OED Evaluations) reviewed the 68 operations completed at that time and found that only 57 percent had satisfactory outcomes. The present report has reviewed all 121 completed operations as well as the design and quality at entry of some ongoing operations. The main findings of the report include a higher percentage of satisfactory outcomes for adjustment operations approved since FY90 (70 percent satisfactory compared to 60 percent for the whole period). The report also finds significant differences among countries: ten of the countries had good compliance of the agreed conditionality, whereas eleven countries had weak compliance and another fourteen had poor or negligible compliance. Seven out of ten good compliers saw an increase in GDP growth and investment, and six of them saw a reduction in inflation, while most (7 out of 12) poor compliers saw economic performance deteriorate. But the cumulative progress of even the best compliers has been limited: only half of them managed to prevent poverty increases and a similar number kept public debt from continuing to grow towards unsustainable levels. While most weak and poor compliers did worse on both counts, the limited progress of the best compliers raises questions regarding the design of the reform programs. Poor borrower ownership continues to be the most frequent problem causing poor performance (and outcomes), although there have been significant improvements in the operations approved since FY90. Three other major issues identified in the report are excessive ambitiousness, poor sequencing of reforms and insufficient attention to macroeconomic stability. Finally, there is also need for greater attention to institutional reform, to supervision, and to monitoring and evaluation. The study finds that the areas of poorest compliance with conditionality are macroeconomic stabilization, public expenditure reforms and public enterprise reform. In addition, conditionality on civil service reform, financial sector reform and measures to alleviate the social impact of adjustment, is present in only some of the countries and also shows low compliance. Best compliance exists on price liberalization, trade reform and regulatory reform. The study makes a number of recommendations regarding the need for i) increased selectivity in adjustment lending, ii) improved design of adjustment operations; and iii) improved benchmarks and indicators of progress towards the objectives of poverty alleviation and fiscal sustainability. Moreover, the report makes process recommendations including the use of multi-tranche operations with pre-determined disbursement intervals, combined with greater flexibility with cancellations of tranches and preparation of new operations after a previous one is canceled. u Contents Summary 3 1. Introduction 7 2. Main Findings 9 Compliance with Conditionality 9 How much adjustment lending has gone to each of the three categories of countries 12 Conditionality and Performance 13 3. Measuring Progress Achieved 15 Poverty Alleviation 15 Fiscal Sustainability 19 4. What were the Causes of Poor Performance? 25 Lack of Selectivity 25 Initial Conditions and Design Issues 26 Variations in Compliance with Specific Conditions 29 5. Current Portfolio and Challenges Ahead 33 Quality at Entry 33 Social Impact of Adjustment 36 Recent Performance of Active Countries 38 6. Recommendations 41 Greater Selectivity 41 Improved Design 42 Social Impact of Adjustment 42 Performance Indicators, Monitoring and Evaluation 44 Tranching and Cancellations 44 2 Boxes: 2.1. Compliance with Conditionality 10 3.1. Poverty Alleviation 16 3.2. Fiscal Sustainability 21 4.1. Uganda: An Emerging Success Story 26 4.2. Kenya: An Early Adjuster with Policy Reversals 30 5.1. Adjustment Experience in the CFA Countries 34 Tables: 2.1. Adjustment Lending to Strong, Weak 13 2.2. Compliance with Conditionality and Economic Performance 14 3.1. Public Expenditure in Health and Education 18 3.2. External Debt to GDP Ratios 19 3.3. Fiscal Performance and Indebtedness 20 4.1. Design Issues in Three Approval Periods 28 4.2. Design Issues in Strong, Weak and Poor Compliers 29 4.3. Compliance with Specific Conditionality Areas 30 5.1. Performance and Progress in Active Countries 38 Annexes: 1. Adjustment Lending in SSA (FY80-96) 49 2. SALs/SECALs 50 3. Summary Results of Adjustment Operations in SSA (FY80-96) 51 4. Policy Reform/Implementation 52 5. Growth, Inflation, Gross Dom. Saving and Gross Dom. Investment 53 6. Real GDP, Population Growth, Real Per Capita Growth 54 7. Adjustment Type 55 8. Adjustment Type, Lending Commitment 56 9. Poverty Trends 57 10. Minimum Growth Rates in National Income Required 61 11. External Debt 62 12. Implementation of Structural Reforms in Broad Areas 63 13. Adjustment Periods 64 14. Active Operations as of FY96 65 15. Adjustment Lending in SSA (Master List) 66 16. Trends in Public Expenditures in SSA Countries 74 The report was prepared by Manuel Pefialver (Task Manager) and Poonam Gupta (Consultant). Research assistance was provided by Salman Anees. Barbara Yale provided administrative assistance. A special debt of gratitude is due to Ms. Maria Cristina Germany, Economist, Africa Live Data Base task manager, who always responded promptly and efficiently to the ever expanding requests for data. 3 Summary 1. This study reviews the Bank's experience with adjustment lending to Sub-Saharan Africa (SSA), from FY80 to FY96. During this period, a total of 163 adjustment operations, for $15 billion have been approved to 37 countries. Of these, a total of 121 have already been evaluated at completion. The study is based on information from all the operations, both completed and ongoing. 2. The operations evaluated so far have a satisfactory outcome ratio of 60 percent, which is below the Bank average of about 70 percent for all adjustment operations. But the trend is improving: adjustment operations approved since FY90 show an average satisfactory ratio of 70 percent. The lowest satisfactory outcomes are found among operations approved in the FY84-89 period. Country results, measured by the satisfactory outcome ratio, also show wide differences, ranging from 0 to 100 percent for all operations in a country. 3. This study focuses particularly on the performance, outcomes, and impact at the country level. Performance is measured first in terms of compliance with conditionality in adjustment lending, and then an overall compliance rating is constructed for each country. The compliance rating, which may differ from the results based on satisfactory outcome ratios (e.g. because of subsequent reversals or improved compliance), is used to classify the countries into three groups: those with good, weak and poor compliance records. 4. The study finds that ten countries (which received $4 billion of the total $15 billion) have had good compliance with the agreed upon policy conditionality, and fourteen countries (which received $4.3 billion) have had poor compliance. Another eleven countries received $6.3 billion and had weak compliance. The remaining two countries do not yet have any operations completed and were not rated. 5. An interesting finding of this study is that 7 out of 10 good compliers saw improvement in GDP growth after adjustment. In contrast, growth improved in only 9 out of 22 weak and poor compliers. Similar results are found for inflation and investment (inflation declined in 6 out of 10 good compliers and only 2 out of 21 weak and poor compliers; gross domestic investment increased in 7 out of 10 good compliers, and in only 7 out of 19 weak and poor compliers). 6. But even in the best performing countries the improvements have not been large enough to make substantial progress towards the objective of reduced poverty: only five of the ten good compliers have been able to prevent an increase in the number of poor in the most recent four years (Mauritius, Mozambique, Benin, Mauritania and Ghana), and only in Mauritius has poverty alleviation been substantial. Among the other two groups of countries the results are worse. Only three of the eleven weak compliers (Uganda, Guinea, and Guinea Bissau) and two among the fourteen poor compliers (Equatorial Guinea and Sudan) seem to have been able to prevent poverty increases. In the last two 4 countries, this conclusion is only based on statistical data showing substantial increases in per capita GDP. 7. Growth in the ten best compliers was not accompanied by sufficient fiscal adjustment to prevent substantial increases in the debt to GDP ratio: their external debt as a share of GDP was nearly 169 percent in 1994. Of the ten countries, only two (Mauritius and Ghana) have debt to GDP ratios below 100 percent. The weak and poor compliers do not fare any better. In only four countries among the remaining 25 did the external debt to GDP ratio not increase. These countries are Senegal, Zambia, Equatorial Guinea and Nigeria. The study also looked at the public debt trends for the last four years, by constructing a simple index based on each country's fiscal deficit and GDP growth rate. Based on this index, five of the ten strong compliers have been moving towards improved sustainability, as have three of the weak compliers (and one among the poor compliers). 8. Poor compliance by a majority of countries, and weak results for even the best compliers are the result of several factors, both endogenous and exogenous, some of which are within Bank control. The two most important ones among the latter have been lack of selectivity in lending and poor design of the operations. Lack of selectivity was most noticeable in the FY84-89 period when approvals grew rapidly. In the process, poorly prepared operations, in countries unwilling or unable to implement them, were approved. Since then, selectivity has increased, and the number of countries with ongoing adjustment operations has been reduced to 22 at the beginning of FY97. 9. Design issues, most notably insufficient attention to borrower ownership, were already identified in the 1993 OED report on Adjustment in SSA. The present study revisited and quantified the design issues identified in all of the 121 evaluations at completion. In the operations approved after FY90, poor borrower ownership continues to be the most important factor negatively affecting performance, although there have been substantial improvements from the previous period. Three other major issues identified in the evaluations are excessive ambitiousness, poor sequencing of reforms and insufficient attention to macroeconomic stability. The evaluations also raise, with increasing frequency, the need to pay greater attention to institutional reform and to supervision, monitoring, and evaluation. On the positive side, substantial improvements appears to have been made in the quality of Bank preparation and ESW, the realism of projections and the attention to supply side measures. 10. The detailed analysis of compliance with conditionality allows us to look at which types of policy reforms are more frequently included in adjustment operations, and which ones are more often complied with. Four of the nine categories of conditions, (macroeconomic stability, public expenditure, public enterprise reform, and price liberalization) appear in practically all countries (not necessarily all operations), whereas conditions on social impact of adjustment appear in only half the countries and on financial reform in about 60 percent of the countries. Compliance shows wide variations across conditions: it was particularly poor for macroeconomic stability, public expenditure reform, public enterprise reform and privatization. At the other extreme, 5 compliance was good for price liberalization, trade reform and regulatory reform. The three areas of poorest compliance, together with the two areas where conditionality was low (social impact of adjustment and financial sector reform) suggest an agenda of items that deserve attention in portfolio management and future lending. 11. The current portfolio (at the beginning of FY97) includes 31 operations in 22 countries. Some of these are "older vintage" operations that remain open for a variety of reasons and may deserve to be closed. It is worth noting, however, that the Africa Region carried out a significant streamlining of the adjustment portfolio in FY96, with 20 closings, compared to only ten new approvals. The most recently approved operations in the portfolio (20 operations approved in FY94 or later) show improvements in the quality of conditionality (greater precision, more targets and benchmarks) and in the attention paid to borrower ownership and participation by stakeholders. A number of issues, however, were also identified. The issues include insufficient attention to fiscal reform (revenue impact of proposed measures, links between public sector reform and fiscal performance, need for better monitoring and implementation as prerequisites to fiscal performance, etc.). Other such issues include insufficient attention to parastatal reform/privatization and, in other cases, excessive optimism regarding privatization prospects. 12. The recent operations also show increased attention to poverty alleviation issues, and to prevention, or correction, of negative impacts of adjustment on selected groups of the population. A major area where improvement has taken place is in the attention to public expenditure issues. In spite of these improvements, however, there seems to be a gap between the findings and recommendations of the growing number of Poverty Assessments, and the treatment of the same issues in the adjustment operations. Finally, gender issues are also conspicuously missing from even recently approved adjustment operations. These areas require increased attention. Recommendations 13. The main recommendations emerging from the findings in this report are the need for i) increased selectivity in adjustment lending, ii) improved design, and iii) improved benchmarks and indicators of progress towards the objectives of poverty alleviation and fiscal sustainability. The following paragraphs elaborate on each of these three groups of recommendations and discuss some implications for Bank processes, particularly tranching and cancellations. 14. Increased selectivity is required, to stop financing delays in the adoption of needed reforms. The poor results from the past show that lack of selectivity resulted in financing too many cases of low growth and increased indebtedness. This was an unanticipated effect of excessive willingness to support weak programs and/or reluctant reformers. The need for increased selectivity has been recognized by the Africa Region in its 1995 report on "Higher Impact Adjustment Lending" (as have most of the other recommendations in the present report). But greater selectivity will have to be exercised in the preparation, appraisal -- and approval -- of each proposed new operation. 6 15. The lackluster performance of even the best compliers highlights the need for better design of new adjustment operations. Improved design is also the second recommendation of the "Higher Impact" report. The present study has noted the specific design areas which continue to be a problem. In addition to the perennial ownership question, ambitiousness, sequencing of reforms and attention to macroeconomic stability are still cause for concern. Moreover, the need for more attention to institutional reform and to supervision, monitoring and evaluation are given prominence in recent evaluations. And as noted above, at the level of specific conditionality, five areas emerge as requiring additional efforts: macroeconomic stability, public expenditure reform, public enterprise reform and privatization, financial sector reform, and social impact of adjustment (including gender issues). 16. Attention to poverty issues and social impact of adjustment is higher in recently approved operations, but this report finds that the links to the recommendations in the Poverty Assessments are still weak. These links need greater emphasis and explicit attention in future operations. 17. This report puts emphasis on two specific areas where objectives, benchmarks and performance indicators need greater clarity and attention: the extent to which growth is resulting in poverty alleviation and the extent to which growth is achieved in a fiscally sustainable manner. 18. Finally, this report recommends that new approaches be tested for adjustment operations, in order to increase impact and minimize downside risks. While in some cases this may lead to single-tranche operations, at the other extreme, a larger number of smaller tranches, at pre-determined intervals, may be required. Some experimentation along similar lines was done in three FY96 operations. But this "IMF-like" approach would require an equally IMF-style culture that would make cancellations, and quick preparation of a revised operation, a more routine and less traumatic process. The increased flexibility would combine the advantages of single tranche operations based on already implemented reforms with an approach to adjustment lending that supports reforms in an explicit multi-year framework. This is essential to achieve the countries' growth and poverty alleviation objectives while minimizing the risks of non-compliance. 19. The proposed approach would require, however, a fresh look at the allocation process for IDA funds, which is now operated on a commitment basis (i.e. cancellations are ultimately lost to the borrowing countries). This makes the countries' reluctance to accept cancellations a rational decision and provides incentives for country managers to keep non-performing operations open in the portfolio. This approach contrasts with the IMF quota-based system in which an interrupted program provides the possibility of access to the non-utilized tranches of the quota, on the basis of the merit of a revised program. A similar approach should be adopted in the Bank to encourage earlier and more frequent cancellations that would not penalize the country if a revised program is adopted and implemented shortly thereafter. 7 1. Introduction 1.1 In the 1993 study of adjustment in Sub-Saharan Africa (SSA)1 OED reviewed the 68 adjustment operations completed at that time. It found that only 57 percent had satisfactory outcomes. It also identified ten major factors causing the poor results: these were grouped into the two categories of "Program Design and Implementation" and "Borrower Ownership and Political Economy Issues." 1.2 Since then, many more adjustment operations have been approved and completed. This report is based on the 163 adjustment operations approved through the end of FY96, both ongoing and completed. A total of 121 operations have been evaluated at completion and the satisfactory outcome ratio has now increased to about 60 percent, still well below Bank average. The amount lent for adjustment in SSA exceeds $15 billion, reaching a total of 37 countries. The current portfolio is smaller: at the beginning of FY97 there were a total of 31 active operations, for a total commitment of $2.7 billion in 22 countries. 1.3 The share of completed operations with satisfactory outcome is a limited indicator of success, which does not tell us about trends over time or differences across countries. The trend and country results give us a more complex and evolving picture. First, the trend has improved over time; for operations approved since FY90, the satisfactory outcome ratio is over 70 percent (outcome ratings were lowest for operations approved in the FY84-89 period). Secondly, the differences among the 37 countries are very large, ranging from zero percent satisfactory outcomes for some countries to 100 percent for others. Of the 37 countries, 9 have a satisfactory outcome ratio of 100 percent (Benin, Gambia, Mali, Mauritania, Mauritius, Mozambique, Sao Tome, Sierra Leone, and Zimbabwe), although in some cases this is based on only one evaluated operation. Another 8 countries have satisfactory rates above 60 percent (but less than 100 percent) and 16 countries have less than 50 percent satisfactory outcomes. 1.4 This report does not look at the OED outcome ratings, but focuses on the differences among countries in terms of compliance with agreed reforms, the impact of compliance on the countries' economic performance, and the progress achieved to date. Thus, the main "unit of account" is the country, not the individual loan. Moreover, our ratings of compliance with agreed reforms, at the country level, may differ from the aggregation of compliance with individual loans, e.g. where there have been policy reversals. The report also reviews how the factors affecting outcome have changed over time, and the challenges still ahead. The report concludes with a list of recommendations to increase the positive impact of adjustment lending. Adjustment in Sub-Saharan Africa. Selected findings from OED Evaluations. Report No. 12155. OED. June 1993. 9 2. Main Findings 2.1 Compliance with Conditionality. Several other studies have looked at the 2 relationship between country policies and country performance in SSA2. The present study is concerned with the impact of Bank assistance for adjustment on economic performance and poverty alleviation. Its starting point is not "policy quality" but each 3 country's compliance with adjustment lending conditionality . To measure and compare this compliance across countries, we have classified the multiple reform measures sought by the adjustment operations into four groups (Macroeconomic Stabilization, Public Sector Management, Private Sector Development and Social Impact of Adjustment). We have rated compliance with measures in each of these groups from 1 (highest) to 4 (lowest) and by combining these results, we have constructed a "country rating." The country rating measures overall compliance with conditions and reform measures agreed under all adjustment operations in a given country. The detailed data and results, adjustment operations included, adjustment periods for each country and ratings for groups and subgroups of measures are presented in the Annex Tables. A summary of the methodology and results is presented in Box 2.1. 2.2 Our ratings show a strongly differentiated performance across countries: ten countries (Benin, Gambia, Ghana, Malawi, Mali, Mauritania, Mauritius, Mozambique, Sierra Leone and Tanzania) can be considered strong compliers with conditionality, eleven are weak compliers (Burkina, C6te d'Ivoire, Guinea, Guinea-Bissau, Madagascar, Niger, Senegal, Togo, Uganda, Zambia and Zimbabwe) and fourteen are poor compliers. Two other countries are not rated because no operations have yet been evaluated at completion. In other words, less than 30 percent of the countries have a good compliance record, another 30 percent have a weak record and 40 percent have a very poor one. 2 In contrast, two recent studies, "Adjustment in Africa: Reforms, Results, and the Road Ahead" (World Bank, 1994) and "Sub-Saharan Africa: Growth, Savings and Investment, 1986-93" (IMF Occasional Paper, 1995) ranked policy performance and results, providing a typology of countries according to policy stance and economic performance. The purpose of this evaluation methodology is to-asessilhequality of the conditionality instead of assuming that it was the right one. This is achieved by comparing compliance with conditionality with actual results: if the reforms agreed were implemented, but the results were not achieved, the explanation has to be sought in the quality of the conditionality as well as in other exogenous factors. At the same time, this approach allows us to construct our own control group: the countries who did not comply with the conditionality; their economic performance is compared to that of the good compliers. 10 Box 2.1: Compliance with Conditionality The starting point of the country rating in this study is compliance with conditionality. Policy reform measures agreed upon in the context of adjustment lending cover a large number of areas and are expressed in many different forms. For the purposes of this study we have constructed indices for each country, based on compliance with individual measures in the different operations. Eight separate headings were initially used and then grouped into three categories as follows: i) Macroeconomic Policy: includes all conditions regarding fiscal deficit reduction, fiscal revenues, public expenditure levels, exchange rate, etc. ii) Public Sector Management: all measures under this category were grouped into three main headings: Civil Service Reform, Public Expenditure Reform and Public Enterprise Restructuring and Privatization. The overall rating is an average of these three. iii) Private Sector Development: the measures here are first grouped under Financial Sector Reform, Trade Policy Reform, Pricing Polices and Incentives, and Regulatory Environment. The rating is an average of the rating for these four groups. The three categories were given equal weight in constructing the overall index of compliance. Moreover, compliance with conditions to alleviate negative social impact of adjustment on specific groups was also measured but is not included in the overall index, because these conditions were present in only a minority of countries. The results of these exercises are summarized below. Ten of the 37 countries have been rated as good compliers with the conditionality. Their average rating is 1.7. Macr PSM PSO Social impact TQW No. of Ops. Evaluated Benin 2 2.2 2 3 2.1 2 Gambia 1.5 1.5 1.8 1 1.6 2 Ghana 1.7 1.8 1.2 2 1.6 10 Malawi 2.5 1.7 2.3 3 2.2 6 Mail 1.7 2.3 1.3 3.5 1.8 3 Mauritania 1.5 1.5 2.3 3 1.8 3 Mauritius 1 1.5 1.4 - 1.3 3 Mozambique 1.5 2.3 1.7 - 1.8 2 Sierra Leone 1 1.3 1 2 1.1 1 Tanzania 21 21 11 - 2A 4 Group Average 1.7 1.8 1.7 2.5 1.7 3.6 Another eleven countries were rated as having had weak compliance with conditionality. Their individual ratings range from 2.5 to 2.9 and the group average is 2.7. These countries are: 11 Box 2.1: Compliance with Conditionality (cont.) Mar EM MSQ Social Impact nta No. of Ops. Evaluated Burkina 3 2.7 2.3 1.5 2.7 2 Cote d'ivolre 3.2 2.7 2.4 - 2.8 10 Guinea 3 2.7 1.8 3 2.5 4 Guinea-Bissau 3.5 3 2.3 - 2.9 3 Madagascar 3 3.3 2 - 2.8 4 Niger 3.5 3 1.3 - 2.6 2 Senegal 2.5 2.4 3 - 2.6 4 Togo 3.5 2.9 1.4 - 2.6 4 Uganda 3.5 2.4 1.8 - 2.6 2 Zambia 3 3.3 1.5 3 2.6 1 Zimbabwe a U 17 4 21 1 Group Average 3.2 2.9 2 2.9 2.7 3.4 Finally, fourteen countries are rated as having had poor compliance with conditionality. Their average ratings range from 3 to 4 (with 4 reflecting total lack of compliance). These countries received adjustment lending but failed to implement adjustment policies: As a result, they can be considered as the "control group" of the analysis. The comparison of their economic performance and progress in poverty alleviation with that of the other two groups is presented later in this chapter and in Chapter 3. Macm PM PSD Social impat tl No. of OpS. Evaluated Burundi 4 3 2.7 - 3.2 2 Cameroon 4 3.7 2.5 - 3.4 2 CAR 3.7 3.1 2.2 4 3 4 Chad 4 3.3 - - 3.7 1 Congo 4 2.3 3 - 3.1 1 Eq. Guinea 4 4 4 - 4 1 Gabon 4 3.7 2 3 3.2 1 Kenya 3.2 3.2 2.8 3 3.1 7 Nigeria 4 3 2.3 - 3.1 2 Rwanda 4 3.7 3 3 3.6 1 Sao Tome 4 3 2 - 3 1 Somalia 4 3.7 3.8 3 3.8 2 Sudan 4 - 3 - 3.5 2 Zall 4 4 4 - 4 2 Group Average 3.9 3.4 2.9 3.2 3.4 2.1 The ratings represent compliance, during a period of time covering in most cases more than five years and reaching up to fourteen years for some countries. Thus, among the group of poor compliers some countries had brief episodes of compliance followed by reversals (e.g. Kenya). Others may have had improved compliance in the most recent period (e.g. Cameroon) but the overall rating is still one of poor compliance. This is the relevant assessment to establish the possible association between compliance with the policy reform agreements and economic performance: a year of reform in the most recent past, or in the middle of a period of reversals, would not be expected to be associated with improved performance. 12 2.3 Any classification is bound to produce some anomalies. But the anomalies do not reduce the usefulness of the methodology; rather they pose many interesting additional questions. The major anomaly among the group of good compliers is Sierra Leone: it received a very high rating of 1.1, but the rating is based on only one operation completed and evaluated (out of three approved). Also, the period covered is the shortest of all countries in this group (three years) and policy performance deteriorated as civil unrest increased. But on the basis of compliance with conditionality agreed to during the relevant period, it still belongs in the "strong compliance group". 2.4 Some other country cases are more nuanced and deserve special attention. One example is Kenya, one of the earliest recipients of adjustment lending, with a total of nine approved operations over 16 years (accounting for 42 percent of total Bank lending to the country). In terms of ICR and PAR "satisfactory outcome ratings" Kenya shows a better than average performance (67 percent satisfactory for the seven operations evaluated). But on account of its repeated policy reversals, and applying our rating methodology it falls into the poor complier category, with an overall rating of 3.1. Because of the interest of this case it is included among the selected country case studies later in this report (Box 4.2). 2.5 At least one country experience is worth noting among the "weak compliers". In Uganda, the adjustment period considered starts in 1987 (with the Museveni Government) and excludes the poor compliance period of the early 1980s. The early attempts at adjustment (1983-86) were unsuccessful and ended with a political change and a fresh start at adjustment. Thus, little would be learnt by combining this early period with the following one. In the last decade, the overall compliance with adjustment lending conditions and reform measures was still weak but shows a clearly improving trend. Thus, if this trend continues, a similar exercise one or two years later could put Uganda in the strong compliers category. For these reasons, the case of Uganda is also selected for more attention later in this report (Box 4.1). 2.6 A group of countries that deserve separate treatment is that of the CFA member countries. The diverse performance within the group is reflected in their distribution among the compliance categories: there are some of these countries in each of the three categories. But at the same time, the commonality of some elements of reform (absence of exchange rate adjustment throughout the 1980s and early 1990s, the CFA franc devaluation in January 1994, and the accompanying reforms thereafter) as well as the coordinated assistance efforts from the Bank and other donors in the aftermath of the devaluation makes this group of countries worthy of special attention later in the study (Box 5.1). 2.7 How much adjustment lending has gone to each of the three categories of countries? The ten best compliers received a total of $4 billion in adjustment lending, slightly over one fourth of total adjustment lending to the region. The fourteen weakest compliers received nearly 30 percent ($4.3 billion) and the eleven weak compliers received the remaining 42 percent (Table 2.1). Most countries received multiple 13 adjustment operations, including those in the category of poor compliers. Only two countries in this last category, Equatorial Guinea and Rwanda, have received one single operation; the two countries not rated, Comoros and Ethiopia have received only one operation which is not yet closed. On average, however, the weak compliers have received the largest number of loans, with about six adjustment operations per country, followed by the best compliers (an average of 5.4 operations) and the poor compliers with only three operations per country. Table 2.1: Adjustment Lending to Strong, Weak and Poor Compliers Amount ($ billion) No. of Loans Strong Compliers (10) 4.0 54 Weak Compliers (11) 6.3 65 Poor Compliers (14) 4.3 42 Source: Annex Tables 7 and 8. 2.8 In US$ terms, weak and poor compliers received nearly three quarters of all adjustment lending to the Region. A large share of this lending was to Nigeria ($1.3 billion) and Kenya ($1.2 billion) among the poor compliers, and C8te d'lvoire ($2.4 billion) and Zambia ($1.1 billion) among the weak ones. Together these four countries account for 40 percent of total adjustment lending to Sub-Saharan Africa. 2.9 Most of the 42 loans received by the poorest compliers had more than one tranche. Interestingly, however, only four of the fourteen poor compliers had second tranches canceled (Burundi, Chad, Rwanda and Somalia). A fifth country (Cameroon) had a third tranche cancelled in the context of the follow-up to the 1994 CFA-frane devaluation. The tranche (of an IBRD loan) was replaced by an IDA credit. 2.10 Conditionality and Performance. Overall, countries that complied with the conditionality did better than countries that did not. For each country, we have measured GDP growth, inflation and investment for the 4-5 year periods before adjustment, after the initiation of adjustment and in the most recent years (ending in 1995). The results show that in 7 of 10 good compliers GDP growth improved in the most recent four years compared to the preadjustment period. In contrast, growth improved in only 9 of 22 weak and poor compliers (similar results are found for inflation and investment; inflation declined in 6 of 10 good compliers, 2 of 10 weak compliers and in none of the poor compliers; gross domestic investment increased in 7 of 10 good compliers, 5 of 10 weak and 2 of 9 poor compliers). It is worth noting that, when 4-5 year periods are considered, there is no evidence of a "growth pause" or even and investment pause during adjustment (for the good compliers)4 Similar exercises were attempted using savings and private investment which are, in principle, better indicators than total investment. The results confirm the findings above, but statistical data on these two indicators are poor and unreliable as they seem to be generated as residuals. 14 2.11 Within each of the three groups of countries there are significant differences in performance (Annex 5). For example, one of the countries in the good compliers category shows substantially lower growth in the most recent years than before adjustment (Sierra Leone where civil disturbances have affected performance). If Sierra Leone is excluded, the performance of the remaining group becomes much more robust. 2.12 The weak compliance group shows, not surprisingly, a mixed performance picture: four countries achieved increased growth and six had lower growth in the most recent four years. The countries showing increased growth are Burkina, Madagascar, Niger and Uganda. As noted earlier, Uganda is by far the strongest performer. Finally, the poor compliers show a generally negative picture. Only five of the twelve countries for which data after adjustment is available have increased growth: Gabon, Nigeria (both oil- producing countries) Sao Tome, Sudan (but the statistical data for Sudan may be unreliable) as well as Equatorial Guinea. The CAR growth rate was the same before adjustment and in the most recent four years. Four countries in this group had negative GDP growth in the most recent four year period (six if we included Zaire and Somalia, for which reliable data are not available). Table 2.2: Compliance with Conditionality and Economic Performance No. of Countries with Improvement in No. of Countries Growth Inflation GDI* with Data Strong Compliers 10 7 6 7 Weak Compliers 10 4 2 5 Poor Compliers 12 5 0 3 *Data during and after adjustment available for only 10 poor compliers. Source: Annex 5. 2.13 How much progress has been achieved in the countries that complied with the conditionality and followed through with the reform measures? The answer to this question focuses on poverty alleviation and sustainability of growth; it should help identify the challenges ahead, the main areas that future adjustment lending should focus on, and provide guidance on performance indicators to be monitored as part of the Bank's assistance. What were the causes ofpoor performance in so many countries? This question is addressed by looking at initial country conditions and design issues, to learn lessons that can be applied in future support for adjustment. 15 3. Measuring Progress Achieved 3.1 This chapter explores the extent to which improved performance has led to progress in poverty alleviation and financially sustainable growth. The first section assesses the impact on poverty. The second section discusses the financial sustainability by looking at the changes in debt and debt service, and measures the additional fiscal effort needed to achieve sustainability. Poverty Alleviation 3.2 Measuring progress in poverty alleviation is a difficult task because of the paucity of data. For the purposes of this study we have used the methodology developed in a previous OED study , combined with an updated data base and increased country coverage. Given the limited coverage of individual indicators we have combined three separate sources to assess progress on poverty reduction. The sources are per capita GDP growth, household surveys providing poverty headcounts at two points in time, and the "growth elasticities" estimated in the previous OED study, which measure the poverty alleviation impact of distribution-neutral growth. The specific question asked in this study is: which countries have managed to decrease the total number of people (not the share of the population) living in absolute poverty? An overview of the methodology and the detailed country data is presented in Box 3.1. 3.3 The two major findings are that i) poverty alleviation, or at least no increase in the absolute number of people living in poverty, appears to have been achieved in very few countries (10 out of 35) and ii) poverty alleviation was greater in the group of good compliers and lowest among the poor compliers. Countries with progress in poverty alleviation include five of the ten best compliers (Mauritius, Mozambique, Benin, Mauritania and Ghana), three among the eleven weak compliers (Uganda, Guinea, and Guinea-Bissau) and two among the fourteen poor compliers (Equatorial Guinea, and Sudan). 3.4 Poverty alleviation is the result of growth but also of increased human capital (improved health and education) as well as adequate provision of public goods. Therefore, it is also important to look at public expenditure composition and particularly at the effort on social sector expenditures (education, health and social welfare) as a result of adjustment lending. OED's Social Impact of Adjustment (SIAO) report showed that performance in this area was mixed, both in SSA countries and Bankwide, although public expenditures in the social sectors did better than total discretionary expenditures. The Social Impact of Adjustment Operations: An Overview (Report No. 14776, June 30, 1995). 16 Box 3.1: Poverty Alleviation The first indicator we used to assess poverty alleviation is the per capita GDP growth (for the most recent four years). The Social Impact of Adjustment study (SIAO) showed that no progress in poverty reduction was achieved in countries without growth. Thus, where the period shows no per capita growth we conclude that poverty increased. This calculation is presented for all 35 countries. A second indicator is given by household surveys that provide direct measurements of poverty at two points in time. But the periods covered vary, and are not necessarily close to the two periods defined above. Moreover, the coverage of these surveys is uneven. They provide, however, a useful component to the GDP per capita measure. The third indicator uses "growth elasticities" to estimate the poverty alleviation impact of distribution-neutral growth, based on the initial income distribution. The SIAO study calculated the growth elasticities for eleven SSA countries, and we have applied them to the growth in the most recent four year period, to estimate a "poverty growth gap" defined as the difference between actual GDP growth and the minimum growth required to prevent the number of poor from increasing. The summary results are presented below. Per Capita GDP Growth Most Recent Direct Poverty Number Good Compliers 4.yeam Poverty Counts Growth Gap ofPoo Benin 1.2 n.a. n.a. stable Gambia - n.a. n.a. increase Ghana 1.4 decline 0.2 stable Malawi - n.a. 4.2 increase Mali 0.7 n.a. n.a. increase Mauritania 1.6 n.a. 0.2 stable Mauritius 3.4 decline n.a. decrease Mozambique 4.0 n.a. n.a. decrease Sierra Leone - n.a. n.a. increase Tanzania 0.1 increase 5.3 increase Among the ten best compliers, two shows clear positive results. They are Mauritius where both the per capita GDP growth and the direct poverty data show this, and Mozambique, where per capita GDP growth has been robust. Another three countries are estimated to be in the borderline. The growth elasticity for Mauritania, when combined with its recent growth shows that the country had a very small poverty growth gap (0.2 percent p.a.). In Ghana the two surveys available (1988 and 1992) show a small decline in the percentage of population living in poverty, but the poverty growth gap also at 0.2 percent p.a. for the most recent four years shows a small increase in the number of poor. In Benin, the per capita GDP growth is similar to Ghana's so we also estimate the result as stable. The remaining five countries show an unequivocal increase in both, the number and the percentage of poor in the total population. 17 Box 3.1: Poverty Alleviation (cont.) Per Capita GDP Growth Most Recent Direct Poverty Number Weak CompliM 4.yeam Povert Counts Growth Gap o Por Burkina n.a. n.a. increase Cote d'lvoire - increase 3.6 increase Guinea 1.3 n.a. n.a. stable Guinea-Bissau 1.1 n.a. n.a. stable Madagascar - increase n.a. increase Niger - increase n.a. increase Senegal 0.6 n.a. 2.8 increase Togo - n.a. n.a. increase Uganda 3.9 increase 0 decrease Zambia - n.a. n.a. increase Zimbabwe n.a. 4.6 increase Among the weak compliers, Uganda, appears to have achieved a decrease in the total number of poor. Although two poverty surveys, in 1989/90 and 1992/93 show a small increase in poverty, the strong per capita GDP growth, and the zero poverty growth gap support this conclusion. Two other countries in this group, Guinea and Guinea Bissau, show per capita GDP growth rates consistent with a stable or slightly declining number of poor. Per Capita GDP Growth Most Recent Direct Poverty Number Poor CompHers A.years Povey Counts Growth Gap othat Burundi - n.a. n.a. increase Cameroon - n.a. n.a. increase CAR - n.a. n.a. increase Chad - n.a. n.a. increase Congo - n.a. n.a. increase Eq. Guinea 6.7 n.a. n.a. decrease Gabon - increase n.a. increase Kenya - decrease 4.0 increase Nigeria - decrease 2.8 increase Rwanda - increase 11.0 increase Sao Tome - n.a. n.a. increase Somalia n.a. n.a. n.a. increase Sudan 3.5 n.a. n.a. decrease Zaire n.a. n.a. n.a. increase Only Equatorial Guinea, and Sudan show per capita GDP growth consistent with a stable or declining number of poor. But this progress has nothing to do with adjustment. The Kenya and Nigeria poverty surveys show a decline in poverty, but not for the two periods considered in this study. In Kenya the two surveys (1981 and 1992) show a very small decrease in the percentage of rural population living in poverty (from 47.9 to 46.4 percent). But per capita GDP growth was negative for the most recent four years and the poverty-growth gap is large. Thus, we conclude that the number of poor has increased. In Nigeria, the surveys show a decrease from 1985 to 1992 (from 43 to 34 percent), and an increase again, to 39 percent in 1995. Negative per capita growth, and a poverty growth gap for the latter period of 2.8 percent p.a. suggest that the number of poor has increased. 18 3.5 In the current study we have also attempted a more systematic evaluation of public expenditure changes in SSA countries that received adjustment lending, before and after adjustment started. The first, and disappointing, finding is the paucity of the data: in only 18 of the 35 countries was it feasible to compare public expenditures in health and education over time, and only by using several sources, most of them from outside the Bank6. Given the Bank's interest on public expenditures and the importance of tracking their changes, the lack of such performance indicators is worrisome and deserves increased attention. 3.6 What happened to public expenditure in health and education before and after adjustment? For the 18 countries together, the total share of these expenditures increased, from 4.6 percent of GDP to 4.9 percent of GDP (Table 3.1). This is an encouraging finding but it is flawed by the data limitations: there is no way to find out whether the absence of data in the remaining countries is related in a systematic way to the attention paid to these sectors (or to the quality of public management). At the same time the differences among good, weak and poor compliers, and between health and education, are worth exploring. Table 3.1: Public Expenditure in Health and Education (as percentage of GDP) Heal Education Total Country Group Before 2 most Before 2 most Before 2 most Adjust. recent yrs. Adjust. recent yrs. Adjust. recent yrs. Strong Compliers (7 out of 10) 1.3 1.3 2.3 3.2 3.6 4.5 Weak Compliers (9 out of 11)' 1.3 1.3 3.9 3.5 5.2 4.8 All Available (18 out of 35)" 1.3 1.4 3.3 3.5 4.6 4.9 "' The high share of education is affected by Zimbabwe's expenditure of 8 percent of GDP. b) Only 2 of the 14 poor compliers are included (Equatorial Guinea and Kenya). The high expenditure on education in Kenya (7 percent of GDP) also affects the total. 3.7 Health expenditures have remained flat at about 1.3 percent of GDP, and there are no significant differences among the strong, weak and poor compliers for which data are available. Education expenditures, however, increased after adjustment for the good compliers and declined for the weak ones. For all 18 countries they also increase. At the individual country level there are some more interesting differences (Annex 16). It is worth noting that education expenditures increased in each of the countries in the good compliers category, and in only three of the nine countries in the weak compliers category for which data is available. 6 A detailed description of the sources (GFS statistics for three countries, Public Expenditure Reviews for two, Poverty Assessments for one and IMF Recent Economic Developments for twelve countries) and their limitations is presented in Annex 16. Although in recent years the number of PER has increased (26 countries between FY94 and FY96) only a few of these (5) have sufficient data to allow calculation of total expenditures on health and education as a percentage of GDP. In the case of Poverty Assessments only in 3 of 26 reports was this data available. 19 3.8 As a counterpoint to the changes in social sector expenditures, it is worth looking at changes in military expenditures in SSA countries, again between the years before adjustment and the most recent years. Information is available for 18 of the 35 countries (but not the same 18 for which expenditures on health and education are available). For the group as a whole, military expenditures fell from 2.7 percent of GDP to 2.2 percent (Annex 16). The decline was stronger among the best compliers (from 3.4 percent to 2.5 percent of GDP. The high average for this group is mostly affected by Mozambique which has the highest percentage of military expenditures (7.6 in the most recent four years), but also by Mauritania and Sierra Leone. Military expenditures also fell, on average for the weak compliers (from 2.4 to 2.1) and for the poor compliers (2.4 to 1.9). Fiscal Sustainability 3.9 Is the growth achieved so far sustainable? Or, has it been financed through large fiscal deficits and growing public debt, both domestic and external? The findings of this study confirm the commonly held view that very few countries have achieved a combination of growth and fiscal deficit reductions that puts them on a sustainable debt path. But the study also finds that the group of best compliers has performed better than the weak and poor compliers. This section looks, first, at the changes in the external debt of all countries that received adjustment, between the year preceding the first adjustment operation and 1994. Then, the study traces in more detail the fiscal deficits, growth, and their combined effect on public debt, for each country, during the most recent 4 years (ending in 1995). 3.10 The external debt of all three groups of countries has increased between the year before the first adjustment operation and 1994 (Table 3.2)7. Their external debt to GDP ratio ranges now from one and a half times the GDP to twice the size of GDP. This is a clear indication that improved performance is needed in all three groups. But the debt increase among the good compliers has only been 38 percent, whereas among the poorest compliers it has been 127 percent (and 48 percent for the weak compliers). As a result, the poorest compliers, who started with the lowest debt to GDP ratio (87 percent) have now the highest of all three groups (197 percent). Table 3.2: External Debt to GDP Ratios Pre-Adjustment 1994 Increase Time Span Level Level (%) Years Strong Compliers 122.0 168.9 38 8.8 Weak Compliers 97.5 144.3 48 7.7 Poor Complers 87.0 197.5 127 8.6 Source: Annex 11. External debt is measured here as total stock outstanding in current prices. 20 3.11 Looking at the individual countries (Annex 11), the ten good compliers include two countries where the external debt was reduced (the Gambia and Mauritius), and two others where the increases were small (Mauritania and Sierra Leone), but also several countries where the increases were quite large: Malawi, Benin and, the highest of all, Tanzania. Annex 11 also shows that individual country performances worse among poor compliers. In particular, nine of the 14 countries among the latter group saw their external debt increase by more than five percent p.a. throughout the period. 3.12 A more complete, but still easy to calculate, indicator of the individual countries' trend towards increased or decreased sustainability of their fiscal performance and indebtedness is given by the relationship between the GDP growth rate, and the fiscal deficit as a share of GDP. In simple terms, the higher the fiscal deficit, the higher the resulting increase in the total public debt, both domestic and external, and the lower the GDP growth rate the faster the increase in the debt to GDP ratio. The recent performance of each of the SSA countries who have received adjustment lending, in terms of GDP growth and fiscal deficit, and the implications for sustainability are presented in Box 3.2, together with some methodology considerations, and are summarized below. Table 3.3: Fiscal Performance and Indebtedness: Trends 1991-95 Country Group Fiscal GDP Potential Fiscal Effort Actual Ext. Deficit Growth Debt/GDP to Reduce Debt (1994) Rate (%) D/GDP to 50% (%) Good Compliers 4.5 3.0 150 3.0 169 Weak Compliers 5.9 2.1 281 4.9 144 Poor Compliers 10.3 1.3 792 9.7 197 Source: Box 3.2. 3.13 The table above confirms the findings of the previous sections: a better performance of the group of good compliers, but a poor overall performance. For the good compliers, with recent growth rates of 3.0 percent p.a. and fiscal deficits of 4.5 percent of GDP, the public debt would tend to grow until it reaches 150 percent of GDP. For these countries, an additional fiscal effort of 3.0 percent of GDP (or substantially higher growth of GDP) would be required to put them in a path leading to a debt to GDP ratio of 50 percent. For the poor compliers the picture is much worse: with deficits of 10.3 percent and growth of only 1.3 percent p.a., the debt would increase (if it could be financed) until it would be nearly eight times the level of GDP, and the fiscal adjustment, required to move to a path leading to a 50 percent debt to GDP ratio is 9.7 percent p.a.. 3.14 Among individual countries (Box 3.2), five of the ten strong compliers have been moving, during the last four years, towards improved sustainability, although in some cases the progress has been small and the starting public debt levels extremely high. These five countries are Benin, Gambia, Mauritania, Mauritius and Mozambique. The remaining five countries, however, require further fiscal adjustment, increased growth, or both. In Malawi and Sierra Leone the "growth-fiscal deficit" gap that needs to be closed is very high. 21 3.15 Among the countries with weak and poor compliance, the results are again bleaker-as when we assessed progress in poverty alleviation. Only three of the weak adjusters (Guinea, Senegal and Uganda) and one among the poor adjusters have been on a path to sustainability during the most recent four years. In addition, data is missing for Somalia and Sudan. Thus, we are left with only four out of 23 countries where some improvements have taken place. The remaining 19 countries are on clearly unsustainable paths and their "growth-deficit gaps" indicate the size of the additional effort required in terms of increased growth/reduced deficit. 3.16 It is important to notice that the time required to reach a given sustainability target (debt to GDP ratio) will depend on both the rate of progress and the starting amount of debt. Therefore, in some cases, small progress, starting from a very high debt level will not be sustainable if substantial additional financing is not available for a long period of time. This study recommends that such calculations, for individual countries, become part of the standard "tool kit" of country economic analysis, and that the relevant projections be prepared in the context of adjustment lending. The "growth-deficit" gap should be monitored, together with its medium term financial implications, as one of the main performance indicators for country results (e.g. in the Country Assistance Strategy). This recommendation emerges also from two other recent reports from the IMF and OED . Box 3.2: Fiscal Sustainability The sustainability of fiscal deficits and of public debt levels and trends is a complex issue that depends, for each country, on many variables. These include, among others, the rates of growth of output and exports, the initial debt and debt service levels, the degree of openness of the economy (i.e. the shares of imports and exports in GDP) and, eventually some more intangible factors such as the willingness of world financial markets and aid donors to continue to lend to a country. Thus, fairly sophisticated models can, and have, been constructed to measure the debt sustainability of specific countries. On the other hand, simple "rule of thumb" models can also be built and be useful guides for policy makers and analysts. For the purposes of this report, we have applied one such simplified model to all 35 SSA countries for which adjustment operations have been completed. Previous OED studies (e.g. the second SAL/SECAL overview, and the Social Impact of Adjustment study) have used more sophisticated calculations of debt sustainability, in line with the approach noted above. Other Bank papers (notably "Solvency, Liquidity, and External Debt - Relief in Severely Indebted Countries" and "Fiscal Solvency and Sustainability in Creditworthiness Analysis" by H.T. Dinh) have refined the analysis further. But although detailed analysis is needed in country economic work and policy advice, the simplified approach used here is useful to grasp the overall picture for the 35 countries. 8 "Overview of Developments in Countries with Stand-by and Extended Arrangements Approved During 1988-91". Policy Development and Review Department, IMF, 1994 "Fiscal Management in Adjustment Lending" Report No. 16040, OED, 1996. 22 In this simplified model we need to look only at two easily available economic indicators (the rate of GDP growth and the fiscal deficit) and to their arithmetical relationship with total public debt. To put it simply, the fiscal deficit is covered, every year, by an increase in the public debt, either domestic or external. As the deficits continue, the debt increases and it may lead to insolvency unless GDP (and exports) grow sufficiently to repay the debt principal and interest coming due. This simplified relationship between fiscal deficits, GDP growth, and public debt can be expressed by a simple equation that represents the change over time of the debt to GDP ratio (D/GDP). In mathematical notation, this change over time is written as (D/GDP) and can be decomposed as follows: (D/GDP) = D/GDP - (D/GDP) GDP/GDP 0 0 where D is the change in the amount of debt, equal to the fiscal deficit. Therefore D/GDP is the usual measure of fiscal deficit as a share of GDP. In the second term, (D/GDP) is the debt to GDP ratio at the beginning of the period, and GDP/GDP is the increase in GDP divided by GDP, or the GDP growth rate. In simple economic terms, the equation tells us when the debt to GDP ratio will increase or decrease. It will increase when the fiscal deficit(as a share of GDP) is higher than the growth rate multiplied by the starting debt to GDP ratio. In the special case that the D/GDP ratio is I (i.e. the debt is equal to GDP) the relationship between fiscal deficit and GDP growth rate is even simpler: the debt to GDP ratio will increase when the fiscal deficit (as percentage of GDP) is higher than the growth rate. And in even simpler terms, the equation expresses the thought that fiscal deficits will lead to debt increases and to insolvency unless GDP growth increases, or the fiscal deficits are reduced or, preferably, both. How are the 35 SSA countries in the study doing when looked at from this perspective? Total public (domestic and external) debt data is not easily available (or reliable), but fiscal deficit and GDP growth rates are. Therefore, using available data it is easy to see which way these countries are headed if they continue with the average fiscal deficit and growth rates achieved in the most recent four years. To calculate this, we look at when will the debt to GDP ratio remain constant: for every level of fiscal deficit and GDP growth there is one such value, although the country may become insolvent well before the level is reached. In terms of the equation used above, the left hand side will be equal to 0 , and the right hand side will be D/GDP = (D/GDP) GDP/GDP or, rearranging D/GDP = (D/GDP)/(GDP/GDP), i.e., the debt to GDP ratio would grow until it equals the fiscal deficit divided by the growth rate. So, if the fiscal deficit is six percent of GDP and the growth rate is two percent, the debt would continue to increase until it reaches three times the size of GDP, unless, of course the country becomes insolvent because no willing lenders are to be found before this debt level is reached. 23 Box 3.2: Fiscal Sustainability (cont.) For the 35 countries in the study, these calculations show, once again, interesting differences among the good, weak and poor compliers. For the good compliers (Table below) the average fiscal deficit during the last four years was 4.5 percent p.a. and the average GDP growth rate was 3.0 percent p.a. Therefore, their total debt would eventually reach 150 percent of GDP, which is loMr than their current level (their external debt alone is now 169 percent of GDP). As a group then, their debt situation is improving although this is not the case for all the countries. The table below shows the data for each of the countries in this group, and it also includes data on their external debt and on the additional fiscal adjustment, additional growth, or combination of the two, that would be needed to put them on a path leading to a maximum debt level of 50 percent of GDP (and the same for 100 percent of GDP). Fiscal GDP Equilibrium Actual External Additional Fiscal Adjust. Deficit Growth DebtlGDP Debt (1994) to 50% 100% Benin [0.5] 4.1 0 106 - - Gambia 0.2 0.9 22 111 - - Ghana 6.1 4.1 149 67 4.1 2.0 Malawi 10.0 1.4 714 151 9.3 8.6 Mali 3.8 3.7 103 149 2.0 0.1 Mauritania 1.9 4.2 45 227 - - Mauritius 2.7 4.9 55 40 0.3 Mozambique 5.9 7.0 84 374 2.4 - Sierra Leone 4.5 -3.7 Inf. 167 6.4 8.2 Tanzania 5.7 3.2 178 293 4.1 2.5 Average 4.5 3.0 150 169 3 1.5 The table shows that five countries (Benin, Gambia, Mauritania, Mauritius, and Mozambique) have been, for the most recent four years, on a path that, if continued, would lead them to reduced indebtedness and increased sustainability. In the case of Benin, the fiscal surplus would lead to reductions in the absolute level of debt, not only in the debt to GDP ratio. Mali is not on a path to increased sustainability, but the gap it would need to close is small. Finally, Ghana, Tanzania, Sierra Leone and Malawi are clearly off-track and need additional fiscal adjustment. In the case of Sierra Leone, with negative GDP growth, there is no maximum debt level: the debt to GDP ratio would continue to increase forever In the other two groups the recent performance has been much worse. For the group of weak compliers, the average deficit was 5.9 percent and the average growth only 2.1 percent p.a. If the trend continued at these levels, their combined public debt would increase until it would reach 281 percent of GDP (5.9/2.1 = 2.81) unless, of course, financing was not available and an insolvency crisis developed. And their current situation is already tight, with external debt averaging 144 percent of GDP. For the poor compliers the average deficit has been 10.3 percent and the GDP growth rate 1.3 percent. Finally, for several of the countries in this group, with large fiscal deficits and negative GDP growth, the "maximum" debt to GDP ratio cannot be calculated-i.e. it would continue to grow forever, if it were to be financed. The tables below show the data for the individual countries in the two groups. 24 Box 3.2: Fiscal Sustainability (cont.) Fiscal GDP Equilibrium Actual External Additional Fiscal Adjust. Deficit Growth DebtlGDP Debt (1994) to 50% 100% Burkina 3.3 1.9 174 40 2.4 1.4 Cote dIvolre 8.5 1.4 607 275 7.8 7.1 Guinea 3.4 4.0 85 89 1.4 - Guinea-Blssa 11.1 3.3 336 337 9.5 7.8 Madagascar 7.1 1.3 546 147 6.5 5.8 Niger 4.7 0.5 940 102 4.5 4.2 Senegal 1.9 3.3 58 95 0.3 - Togo 8.9 0.1 8900 150 8.9 8.8 Uganda 3 7.2 42 86 - - Zambia 4.1 -0.5 Inf. 189 4.4 4.6 Zimbabwe 8.4 0.5 1680 80 8.2 7.9 Average 5.9 2.1 281 144 4.9 3.8 The weak compliers group has only three countries on a sustainable debt path (Guinea, Senegal and Uganda). In Burkina Faso, the gap that needs to be closed is small. But all other countries in the group are on a clearly unsustainable path. Finally, the poor compliers show the following picture: Fiscal GDP Equilibrium Actual External Additional Fiscal Adjust. Deficit Growth Debt/GDP Debt (1994) to 50% 100% Burundi 3.7 -3.4 Inf. 124 5.4 7.1 Cameroon 6.9 -1.0 Inf. 97 7.4 7.9 CAR 6.5 1.9 342 91 5.6 4.6 Chad 9.2 1.7 541 90 8.4 7.5 Congo 12.1 -0.7 Inf. 334 12.5 12.8 Eq. Guinea 8.5 9.8 87 163 3.6 - Gabon 2.4 2.2 109 101 1.3 0.2 Kenya 4.7 1.7 276 105 3.9 3.0 Nigeria 7.5 2.2 341 95 6.4 5.3 Rwanda 7.5 -6.7 Inf. 61 10.9 14.2 Sao Tome 37.6 1.5 2507 871 36.9 36.1 Somalia n.a. n.a. n.a. 259 n.a. n.a. Sudan 17.1 6.3 271 269 14 10.8 Zaire n.a. n.a. n.a. 105 n.a. n.a. Average 10.3 1.3 792 197 9.7 9.0 As expected, these group of countries presents a bleak financial picture. Excluding two countries for which data is not available (Somalia and Zaire), only Equatorial Guinea shows a financial path which could be sustainable, while Gabon shows a very small gap. All other countries are on unsustainable financial paths and require substantial fiscal adjustment, higher growth, or both. 25 4. What Were the Causes of Poor Performance? 4.1 The previous chapters have shown that only half of the countries that received adjustment lending improved their economic performance, and fewer made substantial progress towards poverty alleviation and sustainable growth as a result of their improved performance. Why was this so? The answers are also outlined in the previous chapters: many countries did not achieve the expected improvements because they did not comply with the conditions they had agreed to (in other words: they did not reform). But how about the countries that did comply but did not achieve major improvements? And Vwhy did so many countries fail to comply with the conditionality? The first question is linked to the quality of the conditionality, i.e. to the quality of the reform programs that adjustment lending helped to design and implement. The second question requires an examination of county "ownership" of the reforms, and of the Bank's decision making process in adjustment lending. Lack of Selectivity 4.2 The main answer to the question of why so many Sub-Saharan African countries received adjustment lending but failed to adjust is the lack of selectivity in Bank adjustment lending, particularly in the second half of the 1980s. In that period, the number of countries with active adjustment operations increased rapidly, from 2 in FY80 to 16 in FY85 and 32 in FY91. In the process, poorly prepared operations, in countries unwilling or unable to implement them, were approved by the Bank. This was followed by a high failure rate, and it misfired into a generalized skepticism about the usefulness of the instrument. Since FY91, however, the number of SSA countries receiving adjustment lending has been reduced, to a total of 22 at the end of FY96. Both the lowering of standards in the FY84-89 period, and the improvements since around FY90 are reflected in the changing ratios of satisfactory outcomes mentioned above (para. 1.3). Has this increased selectivity addressed satisfactorily the problems of the past? A look at the 22 countries with active operations countries shows improvements but also some remaining problems. Seven of the active countries had good compliance in the past, but seven had weak compliance and another six had poor compliance : while this is an improved mix, and some countries may be more ready now than they were in the past, the current portfolio deserves continuing attention. This question will be taken up again later. 9The other two (Comoros and Ethiopia) are not rated because their ongoing adjustment operations are the first ones. 26 Initial Conditions and Design Issues 4.3 The reasons for the poor achievements among the best compliers have to be sought in the quality of the adjustment programs agreed with the countries (and in the impact of external shocks). Thus, the presence of design flaws, and shortcomings, needs to be explored. This study revisited the ten major factors, identified in the 1993 OED report, as negatively affecting outcome, and quantified their level and relative importance, as identified in the individual project evaluations at completion, for the 121 operations for which such evaluations exist (PCRs/ICRs and OED Audits). The data was then divided into three subperiods: projects approved in FY80-84, in FY85-89 and in FY90 and later (Table 4.1). Not surprisingly, the overall level of these negative factors is much higher in the middle period, which is the one with the lowest satisfactory outcome rates. Box 4.1: Uganda: An Emerging Success Story? After reviewing the performance and progress of the three groups of countries that received adjustment lending in a "comparative statics" fashion, it is useful to focus on a specific country story to illustrate the dynamic aspects of the process we described. Uganda provides an interesting example in that, although compliance in the past has been weak and spotty (particularly in terms of macroeconomic stabilization), the compliance trend shows a clear improvement over time. As a result, progress towards the ultimate objectives of poverty alleviation and sustainable growth is already observable. Uganda, a naturally rich country (it was called "the pearl of Africa" at the time of independence) emerged in 1979 from the Idi Amin dictatorship an utterly devastated country. Reconstruction started in the early 1980s after Milton Obote was reinstated as president (after the Tanzanian army routed Amin's forces). The donor community provided abundant support and the Bank approved three adjustment operations (in FY82, 83 and 84). But the economic and political situation deteriorated rapidly. Obote was replaced by two military governments, and the civil war intensified until the forces of Yoweri Museveni took over in January 1986. In the event, the outcome of the three adjustment operations approved in the early 1980s was rated unsatisfactory. Adjustment had not taken place. When President Museveni took office the economic situation was critical. Throughout 1986, the new government attempted an interventionist policy, including extensive price controls, an overvalued exchange rate and large increases in public expenditures. This worsened the crisis. But in 1987 the Museveni government changed its policy stance: it initiated a new reform program based on improved functioning of markets, price liberalization, public enterprise reform and privatization, as well as an attempt at reducing in the huge budget deficit. This policy change was partly the result of intensive policy advice from the Bank, the IMF and other donors. And, after the policy change, the donor support was also strong: total aid levels hovered around six percent of GDP between 1985 and 1989 but increased rapidly thereafter to 18-19 percent of GDP in 1992-93. 27 Box 4.1: Uganda: An Emerging Success Story? (cont.) Bank assistance in the form of adjustment lending was large in the last decade: six adjustment operations were approved in FY88, 90, 91, 92, 93 and 94. This is the period used in this study to compute compliance and performance (i.e., the three operations in the early 1980's are not included, and the period through 1987 is considered as the "pre-adjustment" period). Counting the early 1980's as an adjustment period would be misleading. Only three of the six adjustment operations approved since FY88 have been evaluated at completion, and their outcomes have been rated as unsatisfactory in all three cases. But a closer, detailed look at compliance shows a more mixed picture (not captured by the "outcome" rating). First, when similar areas of conditionality exist in the three evaluated operations, the trend in compliance is always improving, even if the level of compliance is not high enough. Second, the improvements over time in the Private Sector Development group of conditions are very strong and compliance at the time of the third adjustment operation was good in this group. Finally, the evaluation of the third operation (conducted in 1995) shows clear improvements in compliance (and in performance of policy variables), over the previous evaluations (conducted in 1990 and 1993), even if the outcome rating was still "marginally unsatisfactory". With several other adjustment operations completed (but not yet evaluated at completion by the Region) it appears likely that if the better compliance trend continue, the outcome of the later operations might be rated as satisfactory and Uganda may move up from the "weak" compliance category to the "strong" one. In terms of performance and progress achieved, the analysis in the report shows improved country results: GDP growth increased from zero in the four years before adjustment to about six percent p.a. during the first four years of adjustment (1988-91) and a high seven percent p.a. in the most recent four years (1992-95). Inflation has been reduced, from 117 percent p.a. before adjustment, to 23 percent p.a. in the last four years (and was on five percent in mid-1996). And most importantly, there are indications that poverty has decreased: the strong growth in per capita GDP growth, combined with the "growth elasticity" calculated in OED's Social Impact of Adjustment study shows a zero poverty growth gap. Two poverty surveys, in 1989/90 and 1992/93 showed a small increase in poverty (Annex 8) but the data for the longer period, and the recent increase in growth suggest that the small deterioration was a temporary phenomenon that has been overturned. Financial sustainability also appears to be on the right track: with growth of 7.2 percent p.a. and a fiscal deficit of 3.0 percent, the ultimate level of debt consistent with these two rates, if they were sustained would be 42 percent of GDP, below the current level of the extemal debt alone, which was 86 percent of GDP in 1994. The fiscal deficit level of 3.0 percent, however, includes a very high level of grants; without the grants the deficit is much higher. Were these very substantial grants to decline, or GDP growth to fall below its current level (or worse still, both at the same time), the sustainability of Uganda's recovery could be threatened. This calls for further improvement of the fiscal performance. Overall, however, Uganda's performance is an example of how sustained policy improvements, supported by Bank adjustment lending and other donor's support can produce measurable improvements in growth and poverty alleviation. 28 4.4 More important than the level (i.e. the percentage of loans on which each of these issues was identified, at completion, as a factor negatively affecting outcome), is the change in their relative importance over time: As shown in Table 4.1, in the most recent approval period, poor borrower ownership continues to be identified, in PCRs/ICRs as well as OED audits, as the most important negative factor, but some other factors have increased in importance while others have decreased. Thus, limited success in achieving macroeconomic stability remains a substantial problem, while sequencing of reforms and excessive ambitiousness seem to have become a smaller problem but are still important. At the same time, poor attention to institutional reforms has appeared as a more important issue than in the past, and problems in supervision, M&E remain stable. Table 4.1: Design Issues in Three Approval Periods Identified as a Problem (%) Approval Years Assessment of Conditions FY80-84 FY85-8 FY90-94 Borrower Ownership 23 42 33 Bank Prep. and ESW Quality 18 10 5 Realism of Projections 32 19 5 Excessive Ambitiousness 14 42 28 Risk Analysis n.a. n.a. n.a. Content of Conditionality Macroeconomic Stability 18 27 23 Sequencing of Reforms 14 32 26 Supply Side Measures 9 8 0 Institutional Reforms 9 5 13 Implementation Supervision, M&E 14 15 15 4.5 In the recent years, poor borrower ownership continues to be listed as the most common factor negatively affecting outcomes (in one third of the cases). While this constitutes an improvement over the 42 percent level in the FY84-89 approvals, it indicates a still serious problem. Three other factors are identified as substantial in the evaluations of adjustment operations approved in the most recent period. They are excessive ambitiousness, poor sequencing of reforms, and insufficient attention to macroeconomic stability. All three factors show declines from their previous levels, but remain high. Finally, two other factors with substantial negative impact are poor supervision, monitoring and evaluation and insufficient attention to institutional reforms. On the positive side, major improvements appear to have been made in the realism of projections, preparation and ESW, and attention to supply side measures. 4.6 How are these design issues and initial conditions related to compliance and, therefore, performance and results? Mapping the presence of the above factors across the three groups of countries used throughout this study shows that poor assessment of initial conditions is strongly linked to poor compliance, but the design issues are not. In other words, the countries with poor compliance are also the ones with poor borrower ownership and where adjustment loans were excessively ambitious (Table 4.2). When it 29 comes to the content of conditionality, however, the problems are more randomly distributed among countries and in some cases (poor sequencing of reforms, insufficient attention to institutional reforms) the problems appear more frequently in the best complying countries. This paradoxical finding-emerging from individual evaluations of adjustment operations-reflects the fact that, in the non-complying countries, problems with sequencing or institutional reform shortcomings had no opportunity to arise because the prior policy measures were not taken in the first place. Finally, the importance of supervision, monitoring and evaluation on the results is also confirmed: the poor compliers had a higher incidence of poor supervision. Table 4.2: Design Issues in Strong, Weak and Poor Compliers Assessment of Conditions Strong Weak Poor Compliers (%) Compliers (%) Compliers (%) Poor Borrower Ownership 19 39 49 Bank Preparation and ESW 14 4 14 Realism of Projection 8 24 14 Excessive Ambitiousness 14 33 51 Content of Conditionality Macroeconomic Stability 11 35 23 Sequencing of Reforms 43 22 14 Supply Side Measures 8 10 9 Institutional Reforms 16 6 3 Implementation Supervision, M&E 11 16 17 Variations in Compliance with Specific Conditions 4.7 In which areas of economic policy has Bank adjustment lending been most and least successful in encouraging reform? The ratings of compliance with conditionality presented in Chapter 2 (Box 2.1) and Annexes 4 and 12 allow us to identify the strong and weak areas of adjustment lending. Three of the nine categories of conditionality show particularly weak results: they are Macroeconomic Stability and Public Finance, Public Expenditure Reform and Public Enterprise Reform and Privatization. For all three categories, compliance in 22 countries was rated weak or poor, and in 10-11 countries it was rated good (Table 4.3). At the other extreme, compliance was good with Price Liberalization and Incentives Reform conditions (22 countries), Trade Reform (19 countries) and Regulatory Reform (16 countries). It is worth noting that specific conditions regarding the social impact of adjustment appeared in only half of the countries (17 out of 34), and compliance was good in only five of these countries. 30 Table 4.3: Compliance with Specific Conditionality Areas (Number of Countries) Type of Condition Good Weak & Poor Total Compliance Compliance Compliance Macro Stability, Public Finance 11 22 33 Civil Service Reform 7 15 22 Public Expenditure Reform 10 22 33 Public Enterprise Reform, Privatiz. 10 22 33 Financial Sector 8 11 19 Trade Liberalization 19 11 30 Price Liberalization 22 11 33 Regulatory Reform 16 9 25 Social Impact 5 12 17 Source: Annex Table 12. Box 4.2: Kenya: An Early Adjuster with Policy Reversals Kenya was one of the first two SSA countries to receive adjustment lending in FY80 (the other country was Sudan). Kenya had been growing strongly since independence, but three factors combined to create a crisis situation around the end of the 1970s: first was the loss of the Ugandan and Tanzanian markets as a result of the problems of these countries; second, the end of the 1976-77 coffee boom; and third, the second oil price increase of 1979. With a good track record in the previous decade Kenya looked like a good candidate for adjustment lending in 1980. Since then, it has received a total of nine operations (three SALs and six SECALs), for a total of nearly $1.2 billion. Yet, its economic performance has been poor, with GDP growth declining from nearly eight percent p.a. in the four years before adjustment to about three percent in the first five years of the adjustment period, and to 1.7 percent in the most recent four year period(1992-95). With population growth exceeding three percent per year, per capita GDP has barely increased. Seven of the nine adjustment operations have already been evaluated at completion, and the outcomes of five out of the seven were rated satisfactory. Why then the poor results, including reduced growth, no progress in poverty alleviation and increased indebtedness? To examine this question it is useful to distinguish several sub-periods and to look at adjustment lending and performance in each sub-period. There were two operations in the first half of the 80s (SAC I in FY80 and SAC II in FY83). Both operations focused on a similar group of four policy components: reduce protection (through elimination of import restrictions and lowering tariffs), promote exports, improve public expenditures and improve agricultural marketing. Initial progress was made towards these objectives and the outcome of SAC I was rated satisfactory. But by 1984-85 there were substantial policy reversals in all areas, and the outcome of SAC II, approved in FY83, was rated unsatisfactory. In the process, the first half of the decade had passed without adjustment, but indebtedness had increased, and the decline in GDP growth, combined with high population growth led to increased poverty. 31 Box 4.2: Kenya: An Early Adjuster with Policy Reversals (cont.) The second half of the 1980s and the early 1990s saw a change in the Bank's assistance strategy, turning to sectoral adjustment lending (six SECALs between FY86 and FY92). But this was also a period of stop and go reforms, with frequent reversals and increasingly unstable macroeconomic performance. The poor macro performance, insufficiently captured by the SECAL instrument, and the increasing governance problems led to the well-publicized break down of the donor community's support to Kenya, which took place in the context of the 1991 Consultative Group Meeting. This signaled the end of a trend of increasing aid levels, which had increased from six percent of GDP in 1980 to nine percent in 1985 and 11-12 percent in 1991-92, years which saw negative GDP growth. Economic reforms in Kenya started again in 1993, together with some progress on governance and some steps towards political reform. Donor assistance was restarted in 1993 (Consultative Group Meeting of November 1993) and the Bank approved another adjustment operation in FY96. Ironically, this most recent adjustment operation (still ongoing) was named SAC I which is also the name of the first adjustment operation to Kenya approved 16 years earlier. The impact on Kenya's economy and population of the decade and a half of (mostly failed) attempts at adjustment is reported in different parts of the study: per capita GDP growth has been barely positive for the full period and has been negative in the most recent four year period. The result has been increased poverty although two surveys conducted in 1988 and 1992 show somewhat stable trends for this sub-period. In terms of financial sustainability, Kenya's external debt increased from 44 percent of GDP in 1979 to 105 percent in 1994. There are some signs of improved performance in the most recent years. For example, there has been liberalization of the exchange and payments system, of the trade regime, and of agricultural marketing and price controls. Moreover, GDP growth recovered since 1994 and, there are indications that the debt to GDP ratios are beginning to fall. But the challenges ahead are substantial and the legacy of more than fifteen years of postponed adjustment are difficult to break. The 1996 SAC operation addresses up-front the issue of government commitment as the main risk associated with the program and focuses on reform of public expenditures, civil service reform and public enterprise reform and privatization. 4.8 The small number of countries in which adjustment operations include conditions relating to the alleviation of negative effects of adjustment on some population groups, and on financial sector reform, cannot possibly be taken as an indication that these were not important issues in the reform process of the countries. Their low level of compliance is also worrying. These two items, together with civil service reform stand out in Table 4.3 as an indicator of areas on which further efforts are needed, both in the management of the current portfolio and in the design and implementation of future lending. A close second is the group composed of macroeconomic stabilization, public expenditure reform and public enterprise reform and privatization. This agenda is a good starting point to look at the current portfolio and future needs in the next chapter. 33 5. Current Portfolio and Challenges Ahead 5.1 The findings in the previous three chapters show the size of the challenges ahead. In simple terms, the limited economic growth that has been achieved has only reduced poverty in a few countries, and it has often been at the cost of continuing-non- sustainable-increases in external debt. This chapter assesses to what extent the ongoing portfolio of adjustment lending is dealing with these challenges. It looks at design issues from a "quality at entry" perspective and puts them into the framework of the economic outlook of the 22 countries currently receiving adjustment lending. 5.2 There were 31 active adjustment operations in 22 SSA countries at the beginning of FY97 (Annex Table 14). These operations, with total commitments of $ 2.7 billion were located in seven (out of ten) strong complying countries, as well as in six weak compliers (out of eleven) and six poor compliers (out of fourteen). Two operations are in the two non-rated countries (Comoros and Ethiopia). The country distribution of ongoing operations is, therefore, better than the distribution of completed projects, with a higher concentration on good compliance countries. Moreover, a look at closings and approvals during FY96 shows an active program to improve portfolio quality: A total of 20 operations were closed and only ten new operations were approved. 5.3 What is the likelihood that ongoing operations will perform better than completed ones, particularly in countries with poor compliance record? Two reasons for moderate optimism include the higher concentration on good compliance countries, and the fact that some of the new operations are in previously weak or poor complier countries which appear to have improved their environment for reform (Cameroon, Kenya) and have reasonable chances of success. But the risks in these countries are still high and their portfolio should be monitored closely. Finally, other operations are holdovers from the past and deserve a closer look at the prospects for restructuring or cancellation. In three of the countries with active operations (Burundi, Nigeria and Sao Tome) the loans date back to FY89 and FY90 and country performance has been poor. These are candidates for closing. Quality at Entry 5.4 To gauge the expected impact of the current portfolio, it is useful to look at the recently approved operations, and to assess to what extent they address issues identified in the past as causing poor results. For this purpose, we have reviewed 20 ongoing operations, all of them approved in FY94 or later. 34 Box 5.1: Adjustment Experience in the CFA Countries: Common Features and Policy Differences Three years ago (in January 1994) the CFA countries took a momentous decision: they agreed on a coordinated devaluation of 50 percent in the CFA franc's parity with the French franc. The previous parity of 50 CFA franc = 1 French franc had been maintained since independence and had been the basis of a long, and eventually unsuccessful, attempt at coordinated economic policies by the member countries, based solely on internal (fiscal) adjustment for the correction of disequilibria, with the exchange rate as a nominal anchor to achieve price stability, and an automatic monetary adjustment (outside of the individual countries' control). The Bank provided adjustment lending through Economic Recovery Credits (ERCs), in the months following the devaluation, to nine of the thirteen CFA countries which had already received adjustment lending earlier years (the other four countries are the CAR, Equatorial Guinea, Benin, and Togo-the last two received new adjustment operations in FY95 and FY96 respectively). The nine ERCs have already been closed and the Africa Region has prepared evaluations at completion (ICRs - Implementation Completion Reports) for all of them in recent months. These evaluations, together with a review of follow-up operations in some of the countries, permit a rare glimpse at a cross-country group of adjustment operations prepared under similar conditions. The first finding is that, beyond the similarities of the devaluation and of the attempts to control the immediate inflationary effects and the impact on the poorest groups of the population, there were substantial differences in the policies adopted by the countries and in the impact of the ERCs. A second finding is that the different post-devaluation policy stances of the CFA countries is, to some extent, a reflection of differences that had existed throughout the 1980s and early 1990s. These differences should not be hidden by the common problem they faced in attempting adjustment without using the exchange rate instrument until 1994 (the practical impossibility of this attempt, increasingly clear to many observers since the late 1980s, has been reported in numerous OED audits of the adjustment operations approved at that time). For example, CMte d'Ivoire and Senegal had initiated adjustment programs already in the early 1980s, followed by Togo (in FY83). These three countries are among the group of weak compliers as a result of their performance during the years prior to the devaluation (with nine loans in C8te d'Ivoire, five in Senegal and Togo). By and large, they made serious attempts at reform, and achieved some positive results during the 1980s, which were gradually eroded in the early 1990s when the overvaluation of the CFA franc made the situation unsustainable. 35 Box 5.1: Adjustment Experience in the CFA Countries: Common Features and Policy Differences (cont.) The remaining ten countries received adjustment lending in the second half of the 1980s, when practically all SSA countries received these operations. But their performance varied enormously. Little needs to be said about the CAR (four operations from FY87 to FY90) and Equatorial Guinea (one in FY89). These two countries did not adjust and, in retrospect, the loans should not have been approved. Cameroon, Congo, Gabon, Chad and Niger received only one or two adjustment operations in the late 1980s and early 90s and, again, showed weak or poor compliance: as a result of this performance no new operations were approved and adjustment lending had a long hiatus of several years until it was resumed with the ERCs in FY94. The three other countries, Benin, Mali and Burkina also started late but kept a steady pace and continuity in policy reform before and after the 1994 devaluation. They are some of the most likely beneficiaries of the Bank's adjustment support. Benin and Mali are clearly within the "good compliance" category (the only two countries in the CFA contingent). Their growth has been modest, but increasing. Benin seems to have managed to stop the increase in poverty and to have achieved strong growth with fiscal equilibrium. But Mali's growth is not strong enough to reduce poverty, and the country is walking a tight line regarding fiscal sustainability. Burkina is falling behind in compliance and all performance indicators are weaker, as expected. What does this mean for the countries outlook and for the Bank's current portfolio of adjustment operations? In addition to Benin and Mali, other countries that have received additional adjustment lending aftu the ERCs include Cte d'Ivoire (two operations in FY96), Senegal (two in FY95), Cameroon (one in FY96) and Chad (also one in FY96). In all of these countries the attempts at reform continue but the risks are high and performance post- devaluation has been uneven (particularly in Cameroon). Thus, all of these ongoing operations require careful attention and monitoring. 5.5 A major finding from the recently approved ongoing operations is that the quality of the conditionality has improved in several ways: it is more precise than in the past, and it incorporates more targets and benchmarks to record progress attained towards achievement of objectives. For example, the Malawi Fiscal Restructuring and Deregulation Program (FY96) has a good discussion of experience with prior adjustment operations, clear conditionality, good risk analysis and a well laid out supervision plan. Moreover, it has an explicit discussion of its poverty alleviation objectives, instruments and poverty indicators, and its approach is consistent with the findings and recommendations of the Malawi Poverty Assessment. At the same time, however, there are still cases of too many, highly detailed conditions (e.g. in the FY95 FSAC to Guinea). Similarly, performance indicators are often related to inputs rather to outputs or impact (e.g. in the FY95 Mali EDSAC) or are not clearly linked to the objectives (e.g. the FY96 Mauritania PRMC) 5.6 Another positive finding emerging from the review of recent operations is the improved discussion of ownership issues, and the efforts to ensure participation by multiple stakeholders. This increased emphasis is noticeable in countries where borrower 36 ownership has been weak in the past, such as C6te d'Ivoire, Cameroon and Senegal, and should improve the chances for future success. 5.7 A number of design issues that could reduce future success have also been found in these twenty operations. For example, the attention to fiscal reform could have been greater in the Guinea FSAC (FY95), where the revenue impact of tax reform in the financial sector was not analyzed, and in the Sierra Leone SAC (FY94) where the discussion of the mining sector, and its contribution to revenues was limited. This was also the case in the Uganda SAC II (FY94) where there was no discussion of parastatal reform, although it had been identified by an OED audit as critical for expenditure control, and in the Zambia Economic and Social Adjustment (FY94) where improvements in budgetary planning monitoring and implementation appeared to be prerequisites to improved fiscal performance. 5.8 In the area of private sector development and privatization, some sequencing issues included liberalization of interest rates prior to the establishment of a sound regulatory framework (Guinea FSAC), lack of attention to the regulatory framework (not taken up by Benin SAC III in spite of a waiver in SAC II) and excessive optimism about privatization prospects (in Ghana PSAC, FY95) given prior experiences in the country. Social Impact of Adjustment 5.9 Finally, this study also reviewed the attention given to poverty alleviation objectives, (as well as instruments and performance indicators) in the recently approved adjustment operations, and the attention to social impact of adjustment issues. The main finding is that a more systematic consideration of these issues has started to appear (e.g. in the Malawi operation mentioned in para. 5.5 above). 5.10 The number of SALs/SECALs with poverty-focused components (other than their emphasis on growth) increased from one of four operations approved in FY93 to one half of the eight operations approved in FY95. The main poverty focus was on the reform of public expenditures and on addressing distortions. Targeted programs, support for the collection of data on poverty and monitoring the impact of adjustment on the poor, and development of poverty policy were not included in any of the four operations. 5.11 On public expenditures, the major indicator used was the share of public expenditures on the social sectors. None of the operations used unskilled wages, CPI (lower income or food only) or rural terms of trade, to ascertain the impact on the poor. While Bank conditionality is increasingly focusing on efficiency and equity of public expenditures, the review of these operations shows that there is still limited use of indicators to measure performance and progress in these areas, particularly the access of the poor to primary education and health care. Conditionality is frequently on allocations rather than actual expenditures, but actual expenditures may bear little resemblance to allocations. For example, the Kenya SAC (1996) includes a provision to increase allocations for scholarships at the secondary level, but the Kenya Poverty Assessment, 37 notes that actual disbursements for 1994 were only half the amount allocated. This discrepancy between allocations and actual expenditures is not addressed by the SAC, but ongoing work on public expenditures is addressing the issue of monitoring actual expenditures. 5.12 In several cases, the actions proposed seem inadequate or insufficient to achieve the objectives pursued. For example, the Senegal Private Sector Adjustment and Competition credit (FY95) notes that private sector activities in rural and informal sectors employ 90 percent of the country's labor force but the credit does not include any specific measures addressed to these sectors. Similarly, in C8te d'Ivoire, the impact of the Agriculture Sector Adjustment Credit (FY96) on the poor will be limited unless the issue of export taxation on cocoa is resolved. In Mauritania, the Public Resource Management Credit (FY96) seeks to enhance the efficiency of public outlays but the capacity building program focuses on five ministries none of which deal with the social sectors. In the Chad SAC (FY96), proposed improvements in cotton pricing are a useful first step, but they can be of limited value to producers unless the cotton parastatal, COTONCHAD is reformed/divested. 5.13 The recommendations from poverty assessments are not always taken into account. For example, the Benin Poverty Assessment notes that cotton cannot be the engine of post-devaluation growth, but the Benin SAC II focuses on cotton as the main source of agriculture growth. The Senegal Poverty Assessment considers wage regulations a key issue but the Senegal Private Sector Adjustment and Competition does not address this matter. In Ghana, the Private Sector Adjustment Credit (FY95) seeks increased allocations to non-wage components of education, health and rural infrastructure, but the Poverty Assessment notes that the key problem in health services is that they are centralized and urban biased. The Kenya SAC (FY96) seeks to improve the efficiency of expenditure allocations, but the program does not address the most inequitable aspect of public expenditure which is the relatively high share devoted to university education. 5.14 A continuing area of weakness is the lack of attention to gender issues. Adjustment policies have been found to create different opportunities and constraints for men and women, and evidence suggests that Bank adjustment strategies will not be as effective as they could be if women's potential productivity is not taken into account. Gender considerations must, therefore, be an integral part of design of policies and programso. The Bank issued OP 4.20 (The Gender Dimensions of Development) in April 1994 and is in the process of revising OD 8.60 (Adjustment Lending Policy) to reflect gender concerns. 10 SPA Workshop on Gender and Economic Adjustment in Sub-Saharan Africa: Summary Report of Workshop Conclusions (April 8, 1994). 38 5.15 A follow-up SPA workshop on Gender and Economic Reform in Africa (October 1995), led to the adoption by the Bank of a four-point approach, including: * Attention to gender-related issues in the macro-policy dialogue (e.g. protection of core expenditures, legal reform and girls' education) * Pilot adjustment operations including gender-related issues in three countries in 1996 (Burkina Faso, Mali and Mozambique) * Process improvements (effective representation of women in policy dialogue, gender analysis in preparation of SALs) * Gender-related actions in complementary investments. Of the three pilot countries, gender specific constraints (such as access to land and credit) were only addressed, so far, in one adjustment operation: the Mali Economic Management Credit. The Current Portfolio: Recent Performance of Active Countries 5.16 To conclude this chapter it is useful to have an overview of the recent performance of the SSA countries where the Bank has an active portfolio of adjustment operations at the beginning of FY97. For this purpose, we bring together, for the active countries, the analyses presented in the previous chapters of this study. 5.17 Excluding the two countries not rated, and the three with only older operations, the remaining 17 active countries have better than average performance. A summary of their compliance record, performance and progress to date is as follows: Table 5.1: Performance and Progress in Active Countries Good Compliers Recent Growth" Poverty Sustainability (per capita) 92-95 Increase/Decrease Path (92-95) Benin 1.2 stable yes Ghana 1.4 stable no Malawi - increase no Mali 0.7 increase no (but improving) Mauritania 1.6 stable yes Mozambique 4.0 decrease yes Sierra Leone - increase no Weak Compliers C6te d'Ivoire - increase no Guinea 1.3 stable yes Senegal 0.6 increase yes Togo - increase no Uganda 3.9 decrease yes Za,boa - increase no Zimbabwe - increase no Poor Compliers Cameroon - increase no Chad - increase no Kenya - increase no (-) means zero or negative growth 39 5.18 The table above is a snapshot-as of early FY97-of the recent performance of this group of countries. As a static representation of a dynamic situation, it is bound to be quickly outdated by changing events in the countries. But it is important keep track of it for at least three different reasons. First, it d= provide a synthetic overview of trends in the early 1990s. Second, it summarizes an analytical approach, and a set of easily monitorable variables that should be collected and used as performance indicators, with a greater degree of detail and accuracy than has been possible in this study, by the Bank staff dealing with each of these countries. In the following and last Chapter, we present a set of recommendations to improve the outcome, and impact of adjustment lending in Sub-Saharan Africa. I 41 6. Recommendations 6.1 The findings of this report suggest a number of recommendations to increase the positive impact of future adjustment lending in sub-Saharan Africa". We have grouped these recommendations under four major headings: (i) greater selectivity; (ii) improved design; (iii) greater attention to social impact; and (iv) process recommendations. A summary of the specific recommendations follows. 6.2 Greater Selectivity. The rapid increase in adjustment lending in the second half of the 1980s led to plummeting success rates in a large number of countries. By the mid 1990s, greater selectivity, meaning (a) a more rigorous assessment of country willingness and ability to adjust (ownership and capacity); (b) improved quality of Bank preparation and ESW; and (c) improved analysis of risk factors, both endogenous and exogenous, has been sought and initiated. But the limited achievements so far, both in the ten countries with strong compliance in the past, and in the 22 countries with adjustment lending underway, suggest that the risks of overextending the Bank's lending for adjustment, and of falling short of the ultimate objectives of poverty alleviation and self-sustaining growth still exist. 6.3 Lest the pressure to lend be erroneously interpreted as a good thing to help borrowing countries even when the chances of success are limited, (i.e. to "take risks" in order to help)it is worthwhile to look at the impact of failed adjustment lending. The analysis in Chapter 2 shows that the 14 countries identified as "poor compliers" received a total of US$ 4.3 billion, in 42 adjustment loans during the last decade and a half. They also received a large amount of cofinancing associated with this Bank lending. Today, their GDP is barely higher than it was before the attempts at adjustment and their external indebtedness is much higher. Their per capita GDP fell in all three periods (before and during adjustment, as well as in the most recent four years) and their total external indebtedness has increased from 87 percent of GDP in the year before adjustment lending, to 197 percent of GDP in 1994. Similarly, the weak compliers received $6.3 billion in Bank adjustment lending, achieved very little progress in terms of GDP growth (2.1 percent p.a., negative in per capita terms), and saw their external indebtedness increase from 98 percent of GDP, to 144 percent in 1994. 6.4 Even in the ten countries that complied best with conditionality and achieved the highest benefits of adjustment, the results fall short of the ultimate objectives. The shortcomings can be identified, and measured, by the two gaps we have presented in Chapter 3: the first is the poverty growth gap, which indicates how short the countries are falling from the minimum growth rate that would keep constant the number of people The recommendations in this chapter, although derived from the analysis and finding of this report, coincide substantially with (and have been influenced by) those of a 1995 report prepared by the Africa Region: Higher Impact Adjustment Lending. Report of the Working Group to the SPA Plenary, Office of the Chief Economist, Africa Region, 1995. 42 living in poverty, and the second is the "growth fiscal deficit gap" that could be defined as the excess of fiscal deficit over the level that would keep the debt to GDP ratio constant given their growth rate. The importance of these two gaps, and the need to monitor them as major performance indicators in adjustment programs is taken up again below. 6.5 Improved Design. The poor results among the best compliers suggest that the conditionality of adjustment lending needs improvement. The discussion of design issues in Chapter 4 has noted the changing trend in these issues over time. On the positive side, the total incidence of design problems has gone down, and some of the problems that were found to be important in the 1993 OED study seem to have been taken care of (this includes, notably, the realism of projections, and the attention to supply-side measures). On the negative side, in addition to the old staple of "poor borrower ownership" (see paras on selectivity above), three major design issues continue to cause problems: excessive ambitiousness (in the number and complexity of measures, not in the "toughness" of the reform), inadequate attention to macroeconomic stability (which confirms the findings of the 1996 OED Report on Fiscal Management) and sequencing of reforms. 6.6 Two other design and implementation issues require continued attention: the quality of supervision, monitoring, evaluation and, and the attention to institutional reforms. On supervision, the apparent slight decline in quality over time may be due as much to the increased attention that evaluations put on this issue as to a "real" worsening, but the continuing appearance of the problem deserves attention. 6.7 Poor attention to institutional reforms has also been reported increasingly over time in evaluations. As in the case of supervision, however, this does not mean necessarily that the Bank is paying less attention to this issue now than it was in the late 1980s. It is indeed possible that self-evaluation and independent evaluation are focusing more closely on this issue because it is now recognized as a major determinant of success or failure. Be it as it may, it deserves to be strengthened in future design of adjustment operations and complementary programs. Finally, two other important issues, social impact of adjustment, and monitoring and evaluation, are treated in their own right in the next sections. 6.8 Social Impact ofAdjustment. This report reviewed the explicit attention given to poverty issues, and to the potential negative effects of some adjustment measures on specific subgroups of the population, including women. For the completed operations the findings are not good: only half of the countries (17) had conditions related to the social impact of adjustment, and compliance with these conditions was good in only five countries. The review of adjustment operations approved in the 1990s, however, shows an improving trend: the discussion of these issues in President's Reports and other documents at entry increases over time, both in terms of quantity and quality. But a number of shortcomings are also apparent and require urgent attention. The following paragraphs discuss four specific aspects that need improvement in future operations: i) 43 explicit linkages to Poverty Assessments; ii) the expected poverty alleviation impact of growth; iii) the role of public expenditure reforms; and iv) the need for, and adequacy of, safety nets. 6.9 The design of adjustment operations, including content and sequencing of reforms should be informed by, and linked to, the findings of the Poverty Assessments, including gender issues. As these Poverty Assessments become more abundant and more complete, their importance, and the influence they should exert on the adjustment operations cannot be overemphasized: an otherwise successful adjustment can result in negative social impact. This was, for example, the case for Zimbabwe's first adjustment credit (SAL/SAC1). This operation, approved in 1992, was successful in supporting trade and domestic regulatory reform, which created the basis for increased growth. But the program did not reduce poverty and unemployment as its architects had hoped, because it did not reach the majoril of Zimbabweans, who work predominantly in the informal sector and in rural areas. 6.10 The growth objective represents only an intermediate goal through which to achieve poverty alleviation. But for growth to have a strong, positive impact on poverty alleviation, both its level and its composition are important. For each pattern of income (and wealth) distribution, the minimum amount of growth needed to achieve specific poverty reduction targets can be estimated and should be monitored. In addition, a better understanding of the impact of alternative sectoral growth performance, can help understand the impact of this growth on poverty alleviation. Finally, the linkages between policy reforms, growth, poverty alleviation and income (and wealth) distribution need to be addressed in each adjustment operation. This is an area where Poverty Assessments can also help. 6.11 The importance of improvements in public expenditure allocations, and their influence on the social impact of adjustment programs and adjustment lending needs little reminding. The OED reports on Social Impact of Adjustment Operations and on Fiscal Management in Adjustment Lending have analyzed the achievements and shortcomings bankwide. The sub-Saharan Africa experience is consistent with those findings. On the positive side, the number of public expenditure reviews available for sub-Saharan African countries is considerable, and should be used to improve the quality of this aspect of adjustment lending. On the other hand, this study found that the majority of public expenditure reviews did not provide sufficient information to trace the changes in public expenditures in the social sectors. Thus, improved data collection is an important prerequisite for good analysis and policy recommendation in this area. 6.12 Lastly, the importance of safety nets, to complement the impact of growth and public expenditure reform needs to be emphasized. This does not mean that all adjustment operations need to include safety net mechanisms as part of the conditionality. In fact, many such activities can be better done outside adjustment lending, for example 12 See "Structural Adjustment and Zimbabwe's Poor". OED Prdcis No 105, February 1996. 44 through specific, investment-type projects. Some such operations have been prepared in SSA (for example the "AGETIP" type projects). But adjustment lending and safety nets need to be closely coordinated, and adjustment operations should discuss in all cases whether the need for special safety net mechanisms exist, and how are they to be tackled, within or outside the adjustment lending operation. 6.13 Performance Indicators, Monitoring and Evaluation. The findings of this study underline two major shortcomings in the area of monitoring and evaluation: the first is that the availability of data regarding the ultimate objectives of Bank assistance through adjustment is extremely limited: for example, only in eight SSA countries do we have data for poverty levels at two points in time. Similarly, reliable data on income distribution, and on the links between growth and poverty alleviation (the "growth elasticities") is only available for a handful of countries. The second shortcoming is the fact that the little information and data available are often not used effectively even when available: this is but one indication of the need to improve the identification and definition of objectives of Bank assistance, including both ultimate objectives and intermediate outputs. 6.14 A major recommendation of this study is that all adjustment operations should include a clear definition of the main objectives of the government's reform program including poverty alleviation objectives, and a financially sustainable growth path. The operations should also include quantifiable measures, or performance indicators, that can gauge to what extent there is progress towards these objectives. Two major performance indicators to measure progress against these objectives are the poverty-growth gap and the growth-fiscal deficit gap, estimated in this report. The indicators would allow the Bank to track progress (or lack thereof) in poverty alleviation and fiscal (debt) sustainability. Starting from these two indicators a larger number of simpler indicators for a variety of components (or building blocks) should be regularly included in the President's Reports, Supervision Reports and Implementation Completion Reports of adjustment operations. 6.15 Tranching and Cancellations. The findings of this study suggest also the need for modifications in the Bank's adjustment lending instruments to (i) increase compliance with conditionality; and (ii) increase the positive impact of adjustment in countries that comply with the conditionality. The attempt to address the first of these two objectives has led to the increased interest in, and reliance on, single tranche operations with the loan following the policy reforms instead of the reforms following the loan. This approach was strongly advocated by the Africa Region's study on Higher Impact Adjustment Lending. As noted earlier, a large crop of such one-tranche operations was approved in the CFA countries in 1994. The OED reviews have noted that, indeed, single tranche adjustment operations are appropriate under certain circumstances. 6.16 But although single-tranche operations with 'already met' conditionality are useful to ensure compliance (by definition), they do not lend themselves easily to ensure that the achievement of the second objective above: to increase the impact of the 45 conditionality. This is because adjustment is a long term process which requires a careful definition of objectives, instruments and intermediate outputs in a multi-year framework, in order to achieve the ultimate objectives over time. In other words, single tranche operations attempt to present a "slice" of the program but fail to present an overview of such program. As a result, several one tranche operations could be prepared and implemented without real progress being made towards the ultimate objectives. At the opposite extreme, however, the traditional alternative of multiple year, multiple tranche operations has also shown severe limitations, partly due to rigidities in the program agreed upon (which does not allow easily for adaptation to exogenous changes) and partly due to institutional constraints. The latter have led, at times, to the disbursement of tranches in cases where some of the conditions were not really complied with. Is there a solution to this apparent dilemma? 6.17 The answer proposed by this report is to adapt one of the well established practices in the IMF programs: programs with multiple reviews (or tranches) at clearly specified points in time, together with an established procedure to cancel the program when the targets are not being met. This can be followed by discussions on a new program. This approach could be used by the Bank (with suitable adaptation but also with changes in the Bank's "culture") to solve the apparent inconsistency between the recognition that adjustment takes time, and the need to proceed one step at a time. 6.18 Today, however, cancellation of Bank adjustment operation is considered to be an extreme decision only to be taken in the most unusual circumstances (para. 2.9 above). This is partly "cultural" but it has also a rational element in the case of IDA-eligible countries (i.e. practically all SSA countries). Under current IDA allocation rules, a cancellation results in a net reduction in IDA resources going to the country, because total approvals or commitments, irrespective of cancellations are used, to calculate the country's use of IDA funds, subject to a maximum level determined by an allocation system that "rations" scarce IDA funds. Thus, the incentive system makes it rational for governments to resist cancellations and for Bank staff to go along with requests to keep the credits open and, eventually, release second tranches. This approach contrasts with the IMF quota-based system, where the percentage of quota used is affected only by the amounts disbursed and outstanding, not affected by the interruption of the program. 6.19 The above argues for the consideration of a more flexible multi-tranche approach, together with a change in the IDA allocation rules and a more open and proactive approach to cancellations. A modified version of the first half of this proposal has been introduced by the Africa Region in three FY96 operations (in Cbte d'Ivoire, Mali and Mauritania). But these 'floating' tranche operations leave open both the timing and sequencing of measures. Moreover, they do not deal squarely with the question of cancellation. The willingness and ability to cancel and renegotiate the program, without a "stigma" and without an inevitable loss of resources should offer the necessary opportunities to achieve the higher impact adjustment lending that SSA countries need and the Africa Region is searching for. I Annexes 49 Annex 1: Adjustment Lending in Sub-Saharan Africa, FY1980-96 Other Adjustment FY Adjustment Lending Total ($ m) Lending as a Lending ($ m) Commitment ($ m) % of Total 1980 120.0 1,426.6 1,546.6 8% 1981 170.0 1,642.4 1,812.4 9% 1982 220.0 1,589.3 1,809.3 12% 1983 406.5 1,367.5 1,774.0 23% 1984 818.2 1,570.5 2,388.7 34% 1985 162.8 1,434.5 1,597.3 10% 1986 1,040.3 1,550.0 2,590.3 40% 1987 1,240.9 1,278.2 2,519.1 49% 1988 1,070.6 1,970.8 3,041.4 35% 1989 1,491.3 2,461.3 3,952.6 38% 1990 1,361.6 2,598.1 3,959.7 34% 1991 1,118.1 2,276.1 3,924.2 33% 1992 1,923.3 2,050.4 3,973.7 48% 1993 677.2 2,140.3 2,817.5 24% 1994 1,323.1 1,484.8 2,807.9 47% 1995 576.3 1,708.0 2,284.3 25% 1996 1,137.9 1,602.2 2,740.1 42% Total 14,858.1 30,151.0 45,009.1 33% Source: Financial Database (FDB) 50 Annex 2: SALs & SECALs in the Africa Region, 1980-96 Total SALs and SECALs 163 % of Total SALs 86 53% SECALs 77 47% % of Total SECALs by Sectors/Areas Secals Agriculture 20 26% Export Rehab/import Rehab 7 9% Fertilizer Cr. 2 3% Educ.Sector 7 9% Fin. Sector/ Ind. Sector/Trade/Private Sector Adj. 26 34% Public Enterprise Adjustment 4 5% Energy Sector 1 1% Water Supply 1 1% Pop/Health 1 1% Human Resource Adj. 1 1% Regulatory Reform 1 1% Recovery Credit 2 3% Public Resource Management 1 1% Others 3 4% 77 100% 51 Annex 3:SUMMARY RESULTS OF ADJUSTMENT OPERATIONS IN SUB-SAHARAN AFRICA, FY 1980-1996 # of Loans % of Loans # of Loans % Satis. to FY # of Loans Evaluated Eval. to Total Rated Satis. total Eval. 1980 2 2 100% 2 100% 1981 4 4 100% 3 75% 1982 2 2 100% 1 50% 1983 7 7 100% 43% 1984 8 7 88% 3 43% 1985 5 5 100% 3 60% 1986 15 15 100% 7 47% 1987 14 14 100% 10 71% 1988 14 14 100% 6 43% 1989 13 11 85% 6 55% 1990 17 13 76% 9 69% 1991 12 8 67% 5 63% 1992 15 9 60% 6 67% 1993 4 0 0% - - 1994 13 9 69% 7 78% 1995 8 1 13% 1 100% 1996 10 0% 0 TOTAL 163 121 72 59.5% Annex 4: POLICY REFORM I IMPLEMENTATION (35 COUNTRIES) Av. Macro & Civil Public Public Public Pvt. Rating OED Macro & Pub.Fin. Service Inv. Enterp. Sector Fin. Trade Sector Social (Excl. %S to Country Pub. Fin. Average Reform Prog. & Pvtn Av. Sector Policy P & Incent. Reg. Dev. Impact S.I.) Tot.Rat. BENIN 3,1 2 2,2 3 2,2 2.2 2,2 1 2 2,3 2 3 2.1 100 BURKINA 3 3 2 3 3 2.7 3 2 2 2 2.3 2,1 2.7 UR BURUNDI 4 4 - 3 3 3 2 3 3 - 2.7 - 3.2 0 CAMER 4 4 3 4 4 3.7 3 2 2 3 2.5 - 3.4 0 CAR 4,3,4 3.7 3,3,4 4,3,3 2,2,4 3.1 - see next 2,2,1 3,4,1 2.2 4 3.0 0 CHAD 4 4 4 4 2 3.3 - - Cotton - - - 3.7 0 CONGO 4 4 3 1 3 2.3 4 see next 2 3 3 - 3.1 0 COTE* 3,4,3,4,3,2 3.2 2,3,3 2,3,3,3 2.7 2 4,2.2 2,3,3,3,2,2 2,2 2.4 2.8 63 EQ. GUIN 4 4 - 4 - 4 4 - - 4 - 4.0 0 GAMBIA 1,2 1.5 1,2 1,2 1,2 1.5 3 - 2,1 1 1.8 1 1.6 100 GHANA 1,1,3 1.7 2,2 1,2 2,2 1.8 1 1 1,1,2 1 1.2 3,1 1.6 80 GABON 4 4 4 4 3 3.7 - 2 1 3 2 3 3.2 0 GUINEA 3,3 3 2,3 3,3 3,2 2.7 2 1 3,1 2,2 1.8 3 2.5 75 GUIN-BIS 4,3 3.5 3,2 4,2 4,4,2 3 4,4,1 1 3,3,1 1 2.3 - 2.9 33 KENYA 3,3,3,3,4 3.2 - 3,3 3,4 3.2 2,4 4,4,2,2 3,3,3 1 2.8 3 3.1 67 MADAGAS 2,4 3 4 2,4,1,4 4,4 3.3 1,3 1,3,1,1 3,2,1 4 2 - 2.8 75 MALAWI 2,3,2,3 2.5 - 1,1,1,2 2,2,3 1.7 1,1 2,2 3,3,1,3 4,2,3 2.3 3 2.2 83 MALI 1,3,1 1.7 3 1,3 2 2.3 1 1,2 1,1 2 1.3 4,3 1.8 100 MAURITANI 1,2 1.5 2 2,1 1.5 3 2 2 2 2.3 3 1.8 100 MAURITIUS 1 1 - 1 2 1.5 - 1,1 2,2 1 1.4 - 1.3 100 MOZAMB 1,2 1.5 - 2 3,2 2.3 - 2 1,2 - 1.7 - 1.8 100 NIGER 3,4 3.5 - 3 3 3 - 1 2 1 1.3 - 2.6 0 NIGERIA 4 4 - 3 - 3 - 3 3 1 2.3 - 3.1 50 RWANDA 4 4 4 4 3 3.7 - 3 3 3 3 3 3.6 0 SAOTOME 4 4 3 3 3 3 - 3 1 - 2 n.a 3.0 100 SENEGAL 1,4 2.5 4 2,1 3,2 2.4 3,3 4 2,3 3,3 3 1,1 2.6 80 SIERRA L 1 1 1 1 2 1.3 - 1 1 - 1 2 1.1 100 SUDAN 4,4 4 - ? - - - - 3 - 3 - 3.5 50 SOMALIA 4,4 4 - 4,4 3 3.7 4 4 4,3 - 3.8 3 3.8 0 TANZANIA 2,2,3 2.3 - - 1,3,3 2.3 3 1,1 1,2,2 2 1.7 - 2.1 75 TOGO 2,4,4,4 3.5 2 3,3,4,3 2,3,4,2 2.9 - 1 2,1,2 1 1.4 - 2.6 50 UGANDA 4,3 3.5 1,2 3 3,3 2.4 - 1 3,1 3,1 1.8 - 2.6 0 ZAIRE 4,4 4 4 4 4 4 - 4 4 - 4 - 4.0 0 ZAMBIA 3 3 4 3 3 3.3 - 1 2 - 1.5 3 2.6 S ZIMBABWE 3 3 3 3 4 3.3 - 3 1 1 1.7 4 2.7 100 AVERAGE 3.0 2.8 2.2 2.7 Note: 1=Full Implementation;2=Substantial Implementation;3=Weak Implementation;4=No implementation I * If 1995 ERC is excluded the average macro rating is 3.4. 1'Information on macro & Public finance very fragmentary. Аппех 8: GROWTH, 1NFLATIOM, GROSS DOMESTIC SAVINGS, AND GROS3 DOMESTIC 1NVESTMENT(35 countries) MR U1I 1895 lnflation Groиrth GDS(°k of GDpi Dt(°� of GDP1 GDpvtl(°А°of GDP) 1mnlementвtioп (1.5-2.2)- $�S S?�l°1s$l С-� � A�1L�J[[п�! �l� � 9s(Тi5Уг�1 Mhl�4,![[� Pl� 9S1i5.KC�! ML4.X[� Рг� �dICSYLyC�11 МВ9-1[[t Pl'� dS(1G'�УГВ1 �B�J[С� BEN1N 2.1 1оо Х -1.s 1.а 1а.о з.о zs а.1 г.з а.а 7.1 12.0 1з.7 15.а в.7 а.а s.з � GAMBIA 1.6 1оо г9.з ю.о 5.2 1.1 а.а о.9 5.7 а.з а.з 17.2 18.в 21.0 6.з 12.9 1з.а GHANA 1.6 80 47.4 52.0 26.1 -2.6 3.8 4.1 4.8 4.8 3,9 5.0 8.0 15.5 - 3.6 4.8 MALAWI 2.2 83 9.7 12.9 39.4 2.9 1.2 1 4 15.1 1а.0 0.8 26.2 18.7 15 4 8.5 5.0 6.9 MAL1 1.в 1оо Х 8.1 1.z 1z.s г.з з.s _ з.7 -з.7 s.z ь.s 19.5 z1.s za.1 1o.z 11.7 1з.о AAAURITANIA 1,8 100 9.9 6.8 6.2 0.1 2.4 а.2 -0.5 10.0 8,8 29.7 25.2 18.6 18.5 18.1 13.3 MAURITIUS 1.з 1оо 1г.9 в.в s.2 з.s а.в a.s 16.з п5 2а.1 гб.о г1.з i 2s.a n.a 1з.а 1s.6 MOZAMBIQUE ta 1оо � si.z t as.e а8.7 1.7 а.�.о о.2 -з.о а.а гг.з a7.s _� 57 7 1а.г zs.o з5.5 SIERRA LEONE 1_1 � 100 � 61.5 51.7 33.5 3 4 0,3 3.7 8.ь 8.6 3.7 11.8 9.9 � � 7.а 5.5 6.2 TANZANIA � г.1 � 7s ! fi-t ао.з � z3.s ' 1.е�а.г � з.2 I s.1 I а.1 о.а j 1а.з зо.7 i з1.s 1г.з го.9 18.а Average 1.7 'i _ гв.s zзл г�.е , 1.7 зл 1� э.о � 5.в � ze е.а 1в.в г�.в zэ.9 �t.э �2.6 ц.1 l�p/eп►eпtatip�(2.3-2.9) _ _ _ ! _ BURKINA г.7_ uR Х г.б а.а е.в 1.з г.2 1.s a.s 7.з 76 2о.7 г1.5 zz.z 1з.а 1з.9 1з.а СОТЕ 2.8 63 Х 9.5 6.9 14.0 1.7 0.4 1.4 23.2 22.2 14.9 27.б 15.� 9 8 14.2 6.4 5.4 CiU1NEA 2.5 75 - - 8.6 - - 4.0 16.6 10.5 16.6 15.3 - 8.5 8.5 GU1N-B1S 2.s 3з __1 2s.a 77.о а7.в 7.г a.z з.з -1.9 -б.з 2.1 гв.б zs.7 з1.о - - в.в MADAGASCAR -� г.а � 7s � 2te 1ь.з 2в.а -г.з г.а 1.з t1 ь.а г.7 9.2 12.s 11.1 - з.9 а.з NIGER , � 2.6 � 0 �,� 72�0 � 9.3 -4.0 2.1 0.5 5.0 9.5 � 1.5 j 12.4 12.8 5.6 - - - SENEGAL � 2.6 � 80 � Х 10.1 b 9.3 �4.3�- 3.5 3.3 ' 0.1 � 6.3 � 7 З ! 11.2 I 12.2 � 14 1 8.Э � 8.6 9.4 70G0 � _ z.s so Х �s.1 о� 1 а � о i г�r 1 а.з зз.s гз.5 � 1о 7 , в.а _ 1z.s в.5 "' UGANDA z.ь о 11s 5�аа s 1 2z.ь i о.о ; i - --f---� ---У- -- --- -- - - - -п-- ---- --- - - 5�8- -- I- - 7.2 1 4.0 1i -0.1 'I 3.3 � 8.9 � 13 2 ---- 15 6 9 1 -�-- ZAMBIA � j 2.s i � 7о.ь 117.s i1o.o г.о п-о.з -o.s 1а.1 1о з s з 1з о 1ta 10.г а.5 а.7 5 а � _ _ __ _ � � _ ZIMBABWE ' 2 7 � 1оо � �+в 7� гs-з 27.z s i � tь о 5 � 2з.в � 1e. ��a s �22.5 га.о � 2з а 16.о 2о.б zo.2 AV@Гд9Е � � 2.7 +� -- i 1 у9.3 I ц.0 'I 2Q.9 ! 1.Q ! 2.4 j 2Т � 9.6 9.8 А.7 f8.8 f7.5 15.4 , 10.д 8.9 9.4 1т lementaUon 3-41 j _ I �� �� � � - -- � � �� ���_ � - rz ( -- �- - -�- - - --� � QURUNDI 3.2 0 ' ��� 8�.8 �3.3 9 7 1 3 7 3.7 -3.4 3.4 1.4 6 � 17.4 16.1 1` 13.9 � 2.2 ( 2.7 2.1 CAMEROON з.а о �Х , з.1 ' о.г ь.7 о.а -а.а �-1 о 2а.а 1s 7 18.е _' 2a.s 1 1s.1 1s о � 1з.а� 1г.а 1з.z CAR з о � Х! ь.9 � о.9 ` 1о.5 � 1.s i о.з -�- 1.s '-o.s � -а.7 � а.ь i 12.� �2.з 1z.з г.а_j 1.s j 1.s � - _ т -.- � - � �� --- �- - CHAD � �� з.7 � о �� Х_�� з.�7 -о.г 11_з 1 7.з г.о � 17 i-1в.6 -1з.а j -11.а , a.s L 8.s i_a.s i-'�- - CONGO ' _ - �� з.1 I о ,` Х I -1.7 �-о.з s_ 1-о.г : 2.z _� -о.� i zs.o аг.7 �� г1.7 ' 27.s 'i 17.з з1.s _ i -�-- -г---+--,---; - --�--- EQ. GU1N а о !_ Х� I - i 1з.а �-�� г.в 9.в , '� -1а г а.9 1s.9 га.1 _� - j -_ GABON � з.г � о � в��' z.s , 1е.з �-з.� s.з _'�i.z � з в зв.о � а1.а � зs.ь 1 z6.s _ 2з.7�! 27.5 ; г5.а 18.з т--; KENYA з.i �_ а7_ '�_ I а.ь '�� 1о.7 � 1а.а �i 9 � z.s � 1.7 , 21.1 19.1 _� 18.з i za.o _2а.1 _ 1в.1 �� 12.7 � 11.s 1о.з N1GERIA , з.i " 5о ; в.s _2в.ё аз.7 о.7 � s.s _! z.z � 11.s Э 2о s 2а а 1г.1 1а.о 21 г ' a.z i s.a , 1г.1 - --~- -3 10.0 '� �- --1 .1 -6.7 '� . � � -1 1�6 ' 11 3 � �� - �- - RWANDA з.ь _'' о з 2з.s .з г_ � z 2 -1а.з _ s.z в, _ 5АО ТОМЕ з' ioo � 18.7 i зз.7 sз.з 1 s� o.s � 1.s -1з.2 �_-1z.a '_-zz.z t-�- - - �-- 50MAL1A � з.а �' о ' � ао.а '� as.a - з s� i.2 -11.г; -о.з - � гб.а � zs.6 - - - ' - -- �-�---� - - - - - ---т SUDAN i з.s , о _ i zо.з 2s.s ae.s � 1.1 � 2.s � б.з � s.2 � а.1 - , 17 ' 1в.а _- 7s , а.а - ZA1RE , а о - � --� - i.7 ' - ' - ! 9.s ! - i - 1о.г'� - - ' - - - �---�� � } � +- Averege � �, з.а i 1, 9.а '� 1zг � г6.7 1 1.в � 1.о ,i.з fi8.2 � а.7 � 7.6 � 19.х »s i8.o i 1о.о 9.9 � 9.6 AVERAGE FOR SAMPLE , 2.7 1 1 20.8 ; ц.3 � 24.9 ', 1.6 2.1 � 2.1 i 7.9 7.2 6.9 i 18.9 � 18.8 19.0 10.8 i 10.9 1 11.2 54 Annex 6: REAL PER CAPITA GROWTH.(35 counbies, 1992-95) Implementation (1.5-2.9)1 Rating OED(%S QEA Real Per Capita BENIN 2.1 100 X 1.2 GAMBIA 1.6 100 -2.7 GHANA 1.6 80 1.4 MALAWI 2.2 83 -1.3 MALI 1.8 100 X 0.7 MAURITANIA 1.8 100 1.6 MAURITIUS 1.3 100 3.4 MOZAMBIQUE 1.8 100 4.0 SIERRA LEONE 1.1 100 -4.5 TANZANIA 2.1 75 0.1 Average 1.7 0.4 Implementation(2.3-2.2 BURKINA 2.7 UR X -1 COTE 2.8 63 X -1.6 GUINEA 2.5 75 1.3 GUIN-BIS 2.9 33 1.1 MADAGASCAR 2.8 75 -1.8 NIGER 2.6 0 X -2.8 SENEGAL 2.6 80 X 0.6 TOGO 2.6 50 X -2.9 UGANDA 2.6 0 3.9 ZAMBIA, 2.6 -3.4 ZIMBABWE 2.7 100 -1.9 Average 2.7 -0.8 Implementation(3=1 BURUNDI 3.2 0 -5.9 CAMEROON 3.4 0 X -3.8 CAR 3 0 X -0.3 CHAD 3.7 0 X -0.8 CONGO 3.1 0 X -3.5 EQ. GUIN 4 0 X 6.7 GABON 3.2 0 X -0.4 KENYA 3.1 67 -1 NIGERIA 3.1 50 -0.7 RWANDA 3.6 0 -9 SAO TOME 3 100 -1 SOMALIA 3.8 0 - SUDAN 3.5 0 3.5 ZAIRE 4 0 - Average 3.4 -1.4 AVERAGE FOR SAMPLE 2.7 -0.6 55 Annex 7: Adjuster Type, No. of Loans and Adjustment Lending as a Percentage of Total (FY96) Adj. Lending as % of No. of Adj. Adj.Loans as% of Total STRONG ADJUSTERS (10) Total Lending to Ctry Loans Adj. Loans BENIN 22 3 1.8% GAMBIA 27 2 1.2% GHANA 36 13 8.0% MALAWI 40 8 4.9% MALI 35 7 4.3% MAURITANIA 39 6 3.7% MAURITIUS 25 3 1.8% MOZAMBIQUE 37 4 2.5% SIERRA LEONE 31 3 1.8% TANZANIA 32 5 3.1% 54 33.1% WEAK ADJUSTERS(1 1) BURKINA FASO 19 3 1.8% COTE d'IVOIRE 64 12 7.4% GUINEA 21 5 3.1% GUINEA-BISSAU 25 3 1.8% MADAGASCAR 27 4 2.5% NIGER 28 3 1.8% SENEGAL 36 8 4.9% TOGO3 42 6 3.7% UGANDA 38 9 5.5% ZAMBIA 64 9 5.5% ZIMBABWE 26 3 1.8% 65 39.9% POOR ADJUSTERS(14) BURUNDI 28 4 2.5% CAMEROON 34 4 2.5% CAR 33 4 2.5% CHAD 19 3 1.8% CONGO 56 2 1.2% EQ. GUINEA 21 1 0.6% GABON 56 2 1.2% KENYA 42 9 5.5% NIGERIA 22 4 2.5% RWANDA 14 1 0.6% SAO TOME 27 2 1.2% SOMALIA 34 2 1.2% SUDAN 11 2 1.2% ZAIRE 20 2 1.2% 42 25.8% OTHER ADJUSTERS COMOROS 12 1 0.6% ETHIOPIA 14 1 0.6% TOTAL 163 100.0% 56 Annex 8: Adjuster Type, Lending Commitments Adj.Lending Adj. Lending Commitments as STRONG ADJUSTERS (10) Commit.($ mill) % of Total Adj. Comm. BENIN 140 0.9% GAMBIA 39.5 0.3% GHANA 1109.7 7.5% MALAWI 628.6 4.2% MALI 324 2.2% MAURITANIA 188.2 1.3% MAURITIUS 80 0.5% MOZAMBIQUE 558.6 3.8% SIERRA LEONE 116 0.8% TANZANIA 798.7 5.4% 3983.3 26.8% WEAK ADJUSTERS(1 1) BURKINA FASO 133 0.9% COTE d'IVOIRE' 2347.2 15.8% GUINEA 200.2 1.3% GUINEA-BISSAU 53.4 0.4% MADAGASCAR 328 2.2% NIGER 165 1.1% SENEGAL 471.7 3.2% TOGO 3 242.3 1.6% UGANDA 856.7 5.8% ZAMBIA 1132.7 7.6% ZIMBABWE 370.6 2.5% 6300.8 42.4% POOR ADJUSTERS(14) BURUNDI 184.3 1.2% CAMEROON 539.1 3.6% CAR 130 0.9% CHAD 94.1 0.6% CONGO 170 1.1% EQ. GUINEA 10 0.1% GABON 80 0.5% KENYA 1188.2 8.0% NIGERIA 1322 8.9% RWANDA 90 0.6% SAO TOME 16.8 0.1% SOMALIA 132.6 0.9% SUDAN 115 0.8% ZAIRE 243.3 1.6% 4315.4 29.0% OTHER ADJUSTERS COMOROS 8 0.1% ETHIOPIA 250.7 1.7% TOTAL 14858.2 100.0% 57 Annex 9: Poverty Trends STRONG ADJUSTERS Ghana: Poverty Trends(adjustment period: 1983-95) 1988 1992 Headcount 36.9 31.4 Poverty gap 11.9 8.1 Source: Poverty Assessment, June 1995 (Based on household total expenditures on constant Accra May 1992 prices, GLSS1, GLSS2 and GLSS3) Ghana: Social Indicators 1988 1993 Infant mortality (per 1000 births) 77 66 Stunted (% of age group) 31 26 Wasted (% of age group) 8 11 Source: Poverty Assessment, June 1995 Mauritius: Poverty Trends(adjustment period: 1981-) 1981 1986/87 Headcount (Upper Poverty Line) 28.4 19.5 Headcount(Lower Poverty Line) 18.6 11.2 Source: Poverty and the World Bank, 1996 Tanzania: Poverty Trends (Rural)(adjustment period: 1987-95) 1983 1991 1993 1995 Headcount 65 50.5 49.7 58.6 Source: Tanzania - The Challenge of Reforms: Growth, Incomes and Welfare, May 1996 (1983 and 1991 based on household budget surveys, 1993 based on HRD survey and 1995 based on PPA data on expenditures) WEAK ADJUSTERS C6te d'Ivoire: Poverty Trends(adjustment period: 1982-95) 1985 1988 1995 Headcount 11.1 17.8 36.8 Poverty gap 2.9 4.5 10.4 Severity 1.3 1.7 3.2 Source: Poverty Assessment, May 1996.(1985-88 LSMS and 1993 and 1995 Priority Surveys) 58 Madagascar: Food Expenditure, 1980-1993 1980 1993 62.6 71.2 Source: Madagascar: Poverty Assessment, Report no. 14044-MAG, November 8, 1995. Niger: Poverty Indicators (Adjustment period: 1986-95) Year 1990 1995 Headcount 56.6 63.5 Poverty gap 18.5 21.9 Source: A Resilient People in a Harsh Environment. Niger Poverty Assessment, draft, February 2, 1996.(One Household Income and Expenditure Survey, 1989-93, and projections) Togo: Comparison of Per Capita Farm Incomes, 1989-94(Adjustment period: 1983-95) (1994 with 1989 = 100) Farmer incomes Cost of living index Maritime 107 136 Plateaux 113 133 Centrale 71 125 Kara 88 112 Savannas 104 113 Note: Farm incomes and cost-of-living index provides a proxy measure to estimate changes in people's standard of living. Source: Togo: Poverty Assessment, draft, March 1996. Uganda: Poverty Trends (adjustment period: 1988-95) 1989/90 1992/93 Headcount 39 40 ' Two thirds of mean household exp. Excludes in 1992/93 seven districts not surveyed in 1989/90.(Data from Household Budget Survey of 1989/90 and 1992/93 Integrated Household Survey) Source: Uganda: The Challenge of Growth and Poverty Reduction, January 1996. 59 POOR ADJUSTERS Cameroon: Per Capita Consumption by Type of Household, 1983-1993, Yaounde (Adjustment Period: 1989-95) Growth rate 83/93 Type of Household Consumption per head ('OOOCFAF) (% mean) 1983 1993(CFAF83) Salaried 455 316 -30.5 Non-salaried 368 230 -37.5 Unemployed/inactive 373 106 -61.6 Total 416 231 -44.5 Source: Cameroon: Diversity, Growth, and Poverty Reduction, April 1995.(Household Budget Survey, 1983/84 and 1993 survey in Yaounde) Note: Per capita consumption in 1993 adjusted for changes in consumer prices. Note the steep decline for inactive/unemployed. Cameroon: Urban Poverty Trends, Head Count, 1983-1993 Region 1983/84 1993 Douala 2 >30 Yaounde 1 >20 Source: Cameron: Diversity, Growth, and Poverty Reduction, April 1995. Note: 1983/84 Household Budget Survey and estimates for 1993. Gabon: Poverty Trends, 1985-1990(Adjustment period: 1988-95) 1985 1990 Minimum wage used as poverty line Headcount 85.00 87.00 Poverty gap 53.00 55.00 Severity 38.00 39.00 Poverty line set at a dollar a day Headcount 15.00 23.00 Poverty gap 4.00 7.00 Severity 1.00 3.00 Source: Evaluation de la Pauvrete au Gabon, draft, October 31, 1995.(No national household survey data available. The trends are simulated on the basis of the 1975 Lorenz curve and the assumption that growth has been distributionally neutral) 60 Kenya: Poverty Indicators(Adjustment period: 1980-95) Rural Urban 1981/82 1992 1992 Food poverty (2250 calories) Headcount 34.5 37.4 n.a. Poverty gap 28.0 37.0 n.a. Absolute poverty line Headcount 47.9 46.4 29.3 Poverty gap 30.0 40.0 30.7 Relative poverty line (1/3 of mean) Headcount 11.4 20.0 24.8 Poverty gap 21.0 39.0 36.0 Gini coefficient of expenditures 0.4 0.49 0.45 Source: Kenya: Poverty Assessment, March 15, 1995.(1981-82 expenditure survey and Welfare Monitoring Survey of 1992). Nigeria: Poverty Indicators(Adjustment Period: 1987-95) 1985 1992 1995 Headcount 43 34 39 Poverty gap 16 14 17 Severity 8 8 10 Source: Nigeria: Poverty Assessment, February 1996 (Consumer Expenditure Surveys in 1985 and 1992. Rwanda: Poverty Estimates, 1985-1992 (Adjustment period: 1991-95) Item 1985 1992 Headcount 40.0 53.7 Poverty gap 8.4 14.2 Severity 2.4 4.9 Source: Rwanda: Poverty Reduction and Sustainable Growth, Report no. 12465-RW, May 16, 1994 (Household budget and consumption survey, 1983-85 for 1985 and Bank staff estimates for subsequent years) 61 Annex 10: Minimum Growth Rates in National Income Required (Above Which the No. of Poor Will Decline)a Avg Pop Gr. 1992- Growth 1/growth Mininum Actual Gr Actual-Min Country 1995 Elasticityb elasticity Gr.ReqdO (1992-95) Req. Cote d'Ivoire 3.5 -2.3 -0.43 5.0 1.4 -3.6 Ghana 2.7 -1.7 -0.59 4.3 4.1 -0.2 Kenya 2.7 -0.9 -1.11 5.7 1.7 -4.0 Malawi 2.5 -0.8 -1.25 5.6 1.4 -4.2 Mauritania 2.5 -1.3 -0.77 4.4 4.2 -0.2 Nigeria 3.0 -1.5 -0.67 5.0 2.2 -2.8 Rwanda 2.6 -1.5 -0.67 4.3 -6.7 -11.0 Senegal 2.7 -0.8 -1.25 6.1 3.3 -2.8 Tanzania 3.2 -0.6 -1.67 8.5 3.2 -5.3 Uganda 3.2 -0.8 -1.25 7.2 7.2 0.0 Zimbabwe 2.4 -0.9 -1.11 5.1 0.5 -4.6 'Assuming neutral distribution of growth of income, population growth at an average of 1992-95 and growth elasticity which is a reflection of the extent to which the pattern of growth is "pro-poor". b Proportional change in the headcount index as a ratio of the proportional change in national income growth. c Growth of National Income = Population Growth Rate (1-1/Growth Elasticity). See OED Report (May 1995) 62 Annex 11: External DebtlGDP (%) Pre Adjust. 1994 Level Change in Time Span Interest Level (1) Levels (Years) rate (r) Strong Compliers Benin 65.5 106.4 40.9 6 0.08 Gambia, The 132.8 110.8 -22.0 8 -0.02 Ghana 36.4 67.3 30.9 12 0.05 Malawi 66.3 154.8 88.4 14 0.06 Mali 105.2 148.7 43.5 7 0.05 Mauritania 217.4 226.5 9.1 9 0.00 Mauritius 41.3 40.3 -1.0 14 0.00 Mozambique 290.9 374.2 83.4 7 0.04 Sierra Leone 165.6 166.9 1.3 3 0.00 Tanzania 99.1 293.4 194.3 8 0.15 Group Average 122.0 168.9 46.9 8.8 Weak Compliers Burkina Faso 32.3 39.6 7.3 4 0.05 Cote d'voire 96.3 274.7 178.4 13 0.08 Guinea 87.9 88.6 0.8 4 0.00 Guinea-Bissau 177.7 336.6 158.9 10 0.07 Madagascar 72.8 146.8 73.9 10 0.07 Niger 83.9 101.9 18.0 9 0.02 Senegal 99.4 94.8 -4.7 9 -0.01 Togo 116.3 149.5 33.2 12 0.02 Uganda 30.6 86.1 55.5 7 0.16 Zambia 220.1 188.8 -31.3 4 -0.04 Zimbabwe 54.7 80.4 25.7 3 0.14 Group Average 97.5 144.3 46.9 j 7.7 Poor Compliers Burundi 39.6 123.8 84.2 9 0.14 Cameroon 34.0 97.4 63.4 6 0.19 Central African Repu 46.8 90.9 44.1 8 0.09 Chad 34.7 89.7 55.0 6 0.17 Congo 187.5 334.4 146.9 7 0.09 Equatorial Guinea 166.1 163.2 -2.9 9 0.00 Gabon 76.9 100.6 23.6 7 0.04 Kenya 43.6 104.6 61.0 15 0.06 Nigeria 107.0 95.1 -11.9 8 -0.01 Rwanda 31.9 61.1 29.2 4 0.18 Sao Tome and Princi 122.7 870.9 748.1 8 0.28 Somalia 187.0 258.5 71.4 9 0.04 Sudan 53.7 268.6 214.8 15 0.11 Zaire 85.8 105.4 19.6 9 0.02 Group Average 87.0 197.4 110.5 8.6 1. Level in 1994 or the latest year available; Figures in italics were from a year other than 1994. Annex 12: IMPLEMENTATION OF STRUCTURAL REFORMS IN BROAD AREAS:COUNTRY PERFORMANCE Implem. Macro& Pub. Civil Service Public Inv. Pub Enter. Financial Average Fin Reform Program Reform sector Trade P & Incentive Regulation Social 1 to 2 Benin Benin Congo Benin Benin Benin Benin Burkina Burkina Gambia Burkina Gambia Chad Burundi Burkina Burkina Cote D'Ivoire Gambia Ghana Gambia Ghana Gambia Cote D'Ivoire Cameroon Cameroon Gambia Ghana Malawi Ghana Malawi Ghana Ghana Gabon CAR Ghana Senegal Mali Sierra L Mali Malawi Guinea Ghana Congo Guinea Sierra L Mauritania Togo Mauritania Mali Madagascar Guinea Gabon Guinea-Biss Mauritius Uganda Mauritius Mauritania Malawi Guinea-Biss Gambia Kenya Mozambique Mozambique Mauritius Mali Madagascar Ghana Mali Senegal Senegal Sierra Leone Malawi Guinea Mauritania Sierra L Sierra Leone Tanzania Mali Madagascar Mauritius Tanzania Mauritania Mali Nigeria Mauritius Mauritania Tanzania Mozambique Mauritius Togo Niger Mozambique Uganda Sierra Leone Niger Zimbabwe Tanzania Sao Tome Niger Togo Sierra Leone Uganda Tanzania Zambia Togo Uganda Zambia Zimbabwe 3 to 4 Burkina Cameroon Benin Burkina Burkina Burundi Burundi Benin Benin Burundi CAR Burkina Burundi Cameroon Cameroon Cote D'lvoire Cameroon CAR Cameroon Chad Burundi Cameroon Congo Cote d'voire Guin-Biss CAR Gabon CAR Congo Cameroon CAR Eq. Guinea Kenya Kenya Congo Guinea Chad Gabon CAR Congo Gambia Nigeria Malawi Gabon Kenya Congo Guinea Chad Cote D'lvoire Guin-Biss Rwanda Nigeria Madagascar Malawi Cote div Guin-Bis Cote d'voire Gabon Kenya Sao Tome Rwanda Malawi Mali Eq. Guinea Madagasar Eq. Guinea Guinea Mauritania Senegal Senegal Senegal Mauritania Gabon Mali Gabon Guinea-Biss Senegal Somalia Somalia Rwanda Rwanda Guinea Rwanda Guin-Biss Kenya Somalia Zaire Sudan Somalia Guin-Bis Sao Tome Guinea Madagascar Tanzania Zimbabwe Zaire Zambia Kenya Senegal Kenya Mozambique Zimbabwe Madagascar Zaire Madagascar Niger Niger Zambia Niger Rwanda Nigeria Zimbabwe Nigeria Sao Tome Rwanda Rwanda Senegal Sao Tome Sao Tome Somalia Sudan Somalia Togo Somalia Sudan Uganda Togo Togo Zaire Uganda Uganda Zambia Zaire Zaire Zimbabwe Zambia Zambia Zimbabwe Zimbabwe 64 Annex 13: Adjustment Period Country Pre-Adj. Adi.(5 yrs) MR 4 yrs BENIN 1985-88 1989-93 1992-95 BURKINA FASO 1987-90 1991-95 1992-95 BURUNDI 1982-85 1986-90 1992-95 CAMEROON 1985-88 1989-93 1992-95 CAR 1983-86 1987-91 1992-95 CHAD 1985-88 1989-93 1992-95 CONGO 1984-87 1988-92 1992-95 COTE d'IVOIRE 1978-81 1982-86 1992-95 EQ. GUINEA 1982-85 1986-90 1992-95 GABON 1984-87 1988-92 1992-95 GAMBIA 1983-86 1987-91 1992-95 GHANA 1979-82 1983-87 1992-95 GUINEA 1982-85 1986-90 1992-95 GUINEA-BISSAU 1981-84 1985-89 1992-95 KENYA 1976-79 1980-84 1992-95 MADAGASCAR 1981-84 1985-89 1992-95 MALAWI 1977-80 1981-85 1992-95 MALI 1984-87 1988-92 1992-95 MAURITANIA 1982-85 1986-90 1992-95 MAURITIUS 1977-80 1981-85 1992-95 MOZAMBIQUE 1984-87 1988-92 1992-95 NIGER 1982-85 1986-90 1992-95 NIGERIA 1983-86 1987-91 1992-95 RWANDA 1987-90 1991-95 1992-95 SAO TOME 1983-86 1987-91 1992-95 SENEGAL 1982-85 1986-90 1992-95 SIERRA LEONE 1988-91 1992-95 1992-95 SOMALIA 1982-85 1986-90 1992-95 SUDAN 1976-79 1980-84 1992-95 TANZANIA 1983-86 1987-91 1992-95 TOGO 1979-82 1983-87 1992-95 UGANDA 1984-87 1988-92 1992-95 ZAIRE 1982-85 1986-90 1992-95 ZAMBIA 1987-90 1991-95 1992-95 ZIMBABW E 1988-91 1992-95 65 Annex 14: Active Operations as of FY96 Lending Form Country Instru. FY Dev. Obj. Imp. Prog Date Commitments BENIN SACIII 95 S U Jun-96 40,000,000 BURUNDI Ag. Serv 89 U U Oct-96 33,100,000 CAMEROON SAC 96 180,300,000 COMOROS Macro reform 91 S S May-96 8,000,000 CHAD SAC 96 30,000,000 COTE PvtSector 96 180,000,000 COTE AgSector 96 223,600,000 ETHIOPIA SAC I 93 S S Jul-96 250,670,000 GHANA FINSAC 92 S S Jun-96 106,600,000 GHANA Pvt. Sctr Adj 95 S S Jun-95 74,800,000 GUINEA Fin Sectr 95 U U Jun-96 23,000,000 KENYA SAC 96 126,800,000 MALI Ag. Secal 90 HS S Jun-96 53,000,000 MALI Edu. Secal 95 HS S Jun-96 50,000,000 MALI Eco.Mgmt 96 60,000,000 MALAWI Fiscal Restr. 96 106,400,000 MAURITANIA Ag. Secal 90 HS S Jun-96 25,000,000 MAURITANIA Fin&Pvt 95 S S Jun-96 30,800,000 MAURITANIA Pub.Res Mgm 20,000,000 MOZAMBIQUE SERC 94 S S Jun-96 200,000,000 NIGERIA* Edu. Univ. 90 S S Jun-96 120,000,000 SAO TOME SAC II 90 S S Jun-96 9,750,000 SENEGAL Ag. Secal 95 S S Jun-96 47,800,000 SIERRA L Ag. Sectr 84 S S Jun-96 21,500,000 SIERRA L SAC 94 S U Jun-96 50,490,000 TOGO ERAC 96 50,000,000 UGANDA Ag. Sectr 91 S S Jun-96 100,000,000 UGANDA Fin Sectr 93 S U Jun-96 101,050,000 ZAMBIA PIRC II 93 S S Jun-96 110,000,000 ZAMBIA Ecp.Rec&I 96 152,100,000 ZIMBABWE SAC II 93 S U Jun-96 125,000,000 2,709,760,000 ANNEX 15: ADJUSTMENT LENDING IN SUB-SAHARAN AFRICA, FY 1980-1996 MASTER LIST (COUNTRY) Date of Number Lending Loan Report Latest OED of Loans FY Country Instrument Credit Project Name No. PCR/ICR PAR Evaluation Rating per Ctry 1989 BENIN 2SAL 2023 SAL I 1844 1105 Jun-93 S* 1991 BENIN SAL C2283 ¯ SAL 11 P5570 15061 RENE Nov-95 S 1995 BENIN SAL C2727 SAC IlIl P6573 X X - 3 1991 BURKINA __SAL 02281 SALI P5562 15264 x Feb-96 S 1992 BURKINA SAL C2381 Agricuture Sec. P5721 15581 X Apr-96 s 1994 BURKINA SAL C2590 Economic Recov. P6296 15263 X Feb- 96 5 3 1986 BURUNDI SAL C1705 SAC 10685 May-92 U* 1988 BURUNDI SAL C1919 SALII 10685 May-92 U* 1989 BURUNDI SAL C2024 Agr. Services 7557 X X - 1992 BURUNDI SAL 02376 SALiI P5806 15581 x Apr-96 U 4 1989 CA--ERON SAL L_089 SL P5079 15167 SICAT Dec-95 U 1994 CAMEROON SAL C2576 SAC 15167 SICAT Dec-95 u 1994 CAMEROON SAL C2627 ERC P6359 15682 X Jun-96 S 1996 CAMEROON SAC SAC P6773 X- 4 1987 CAR SAL C1732 SAL *9744 Jun-91 U 1988 CAR SECAL C1836 Cotton Sector 10848 Jun-92 S* 1988 CAR SAL 01916 P4813SALII P 10 810848 Jun-92 U* 1990 CAR SAL 02162 5ÃLIlI P5339 15237 RENE - -Feb-96 u 4 Date of No. of Lending Loan Report Latest OED Loans FY Country Instrument Credit Project Name No. PCRIlCR PAR Evaluation Rating per Ctry 1989 CHAD SECAL C1945 Fin. Rehab Cr P4806 14601 Jun-95 U 1994 CHAD SAL C2589 Eco ecov P6289 15665 Jun-96 S 1996 CHAD SAL SAC 785 X X - - 1991 COMOROS SAL C2270 Macro Reform 5 X- 988 CONGO SAL -L2866 SALI P4544 *745 Jun 1994 CONGO SAL 02635 ERC P6372 15765 X Jun-96 S 2 1982 COTE IV SAL L2058 SAL I 5741 Jun-85 S* 1984 COTE IV SAL L23 ¯ 8872 Ju 1986 COTE IV SAL L2711 SALIII 8872 Jun-90 U* 1990 COTE IV SECAL L3127 Ag Sect. Adj 13120 X Jul-94 S 1990 COTE IV SECAL L3150 Energy Sector 11532 + 13290 Jun-94 S* 1990 COTE IV SECAL L3240 Water Supply 12635 13283 Jun-94 S C2303 1992 COTE IV SECAL L3408 Fin Sector Adj P5632 15314 RENE S YC323 1992 COTE IV SECAL L3428 Human Res. Adj P5653 14841 X Jun-95 U _ - ~ _ - 1C2324 1992 COTE IV SECAL L3429 Regulatory Ref. P5691 14715 RENE Jun-95 S 1995 COTE IV SAL C2656 ERC 15234 RENE Jan-96 S 196 COTEIV SECAL --_ Ag. Sec.Adj P6671 X X 196 COTEIV SECAL Pvt. Sec. Dev P646 x ___X 12 1986 EQ. GUNEA SAL 01690 !Rehab.Import P4288 X 1063 Jun-92 U* 4-1 Date of No. of Lending Loan Report Latest OED Loans FY Country Instrument Credit Project Name No. PCRIICR PAR Evaluation Rating per Ctry 1993 ETHIOPIA SAL C2526 SACI P6066 X XX--- 1988 GABON 1 SAL L2933 SALI -- P4742 X 12075 Jun-93 U* 1994 GABON SAL L3759 ERL P6361 15695 X S 2 1987 THE GAMBIA I SAL C1730 SAL 8827 Jun-90 S* 1989 THE GAMBIA SAL C2032 SALII 12936 X Apr-94 S 2 1983 GHANA SECAL C1393 Recon.Imp.Cr. 9580_May-910S* 1984 GHANA SECAL C1435 Export Rehab 9580 May-91 U* 1985 GHANA SECAL C1573 Recon.Imp.Cr. 9580 May-91 S* 1986 GHANA SECAL C1672 Ind.Sector.Adj. 11844 12137 Jun-93 S* 1987 GHANA SECAL C1744 Edu.Sect.Adj. 12622 13575 Sep-94 S 1987 GHANA SAL C1777 SAL 10686 May-92 S* 4_T _H N C1 7 __________*I - . 1988 GHANA SECAL C1911 Fin.Sector Adj. 14158 X Mar-95 S f____ I 4 j S * 1989 GHANA SAL C2005 SACII 10686 May-92 S* 1990 GHANA SECAL C2140 Edu. SACII P5244 14833 X Jun-95 S 1991 GHANA SAL C2236 Pvt.Invest.Prom. P5499 14827 X Jul-95 U 1992 GHANA SECAL C2318 FINSAC P5659 X X - 1992 GHANA SECAL C2345 Agr.Secal P5523 Closed No ICR - 1995 GHANA I SAL C2718 Priv.Sec.Adj. P6543 X X 13 1986 GUINEA SAL _C1659 SALI P4162 8826 Jun-90 S* 1988 GUINEA SAL IC1926 SALII P4805 15164 SHAH Dec-95 U 1990 GUINEA SECAL C2148 Pve.Sector.Pro. P5207 13302 X Jul-94 S 1990 GUINEA SECAL C2155 __ Edu.Sec.Adj P5288 14617 X Jun-95 S 1995 GUINEA --SAL C2653 Fin.Sector P6339 X X - - Date of No. of Lending Loan Report Latest OED Loans FY Country Instrument Credit Project Name No. PCRICR PAR Evaluation Rating per Ctry 1985 GUIN-BISSAU SECAL C1531 Rehab.imports P3901 10824 Jun-92 U* 1987 GUIN-BISSAU _ SAL C1798 SAL P4524 10824 Jun-92 U* 1989 - GUIN-BISSAU SAL C2019 SACII P4980 14685 JOHNSON Jun-95 S 3 1980 KENYA SAL C0999 SAC--- 4934 Feb-84 S* C1276 1983 KENYA SAL L2190 SACIl 5682 May-85 U* 1986 KENYA SECAL C1717 Ag.Sector 8797 8862 Jun-90 S* 1988 KENYA SECAL -C1927 Ind.Sector.Opr 9746 Jun-91 S* 1989 KENYA SECAL C2049 Fin.Sec.Op 11422 12126 Jun-93 U* 1991 KENYA SECAL C2197 Export Dev. 13886 14698 Jun-95 _ __S 1991 KENYA SECAL C2204 Ag.Sec.Adj.ll P5415 Closed No ICR - - 1992 KENYA SECAL C2295 Edu.Sec.Adj. P5571 15065 X Nov-95 S 1996 KENYA SAL SAC P6837 X X 9 1985 MADAGASCAR SECAL C1541 Ind.Sect.Lnll P4488 10405 Feb-92 S* 1986 MADAGASCAR SECAL C1691 Ag.Sec.Adj.Cr 12544 13081 I May-94 I S 1987 MADAGASCAR SECAL C1834 Ind.&Trade Pol. 10540 14774 Jun-95 S* 1988 MADAGASCAR SECAL C1941 Pub.Sec.Adj. P4780 ___14774 Jun-95- U 4 1981 MALAW SAL L2026 SAL 6833 Jun-87 S* 1983 MALAWI SECAL C1352 Fertilizer Credit 9689 Jun-91 -S*7- 1984 MALAWI SAL C1427 SALIl 6833 Jun-87 S* 1986 MALAWI SAL C1644 SALII 8886 Jun-90 S* 1988 MALAWI SECAL 01920 Ind/Trade Adj. 12156 Jun-93 S* 1990 MALAWI SECAL C2121 Agr.Sector 13603 Oct-94 U 1992 MALAWI SAL C2396 Entr.Dev&Droug P5461 Closed NoICR - - 1996 MALAW1I SAL I Fis. Restr. DeregI P6799 X x 8 Date of No. of Lending Loan Report Latest OED Loans FY Country Instrument Credit - Project Name No. PCRIICR PAR Evaluation Rating per Ctry 1988 MALI SECAL C1937 P.E.Sec.Adj P4734 11994 13272 Jun-94 S* 1989 MALI SECAL C2054 Educ.Sect.Cons P5010 15720 X Jun-96 NOT REV. 1990 MALI SECAL C2163 Ag. Secal P5324 X X - - 1991 MALI SAL C2188 SALI P5318 15644 X May-96 s 1994 MALI SAL 02580 Econ.Recov.Cr, P6274 15819 X Jun-96 s 1995 MALI SECAL C2673 Edu.Secal P6484 X x-_- 1996 MALI SAL Eco. Mgmt.Cr P6909 X X - - 1986 MAURITANIA SECAL L2643 SNIM Rehab. P4139 9487 10784 Jun-92 S* 1987 MAURITANIA SAL C1812 SAL P4550 ____ 9682 Jun-91 S* 1990 MAURITANIA SECAL C2093 Ag.Sec.lrrig. P4988 X X - - 1990 MAURITANIA SECAL C2166 P.E.Sec. Adj. P5293 13929 Jan-95 s 1995 MAURITANIA SECAL C2726 Fin&Pvt P6574 X X - - 1996 MAURITANIA SECAL Pub. Res.Mgmt P6873-- X X - - 1981 MAURITIUS SAL -2 SAL 6870 Jun-87 1984 MAURITIUS SAL L2361 1 SALII 6870 Jun-87 S* 1987 MAURITIUS SECAL LM791 Ind.SAL 10653 1 May-92 _ 1~ 3 1988 MOZAMBIQUE SAL C1841 Rehab.II 11767 12128 Jun-93 s* 1989 MOZAMBIQUE SAL C2021 Rehab.Ill 13866 12128 Feb-95 S* 1992 MOZAMBIQUE SAL C2384 Econ.Recov.Cr P5775 Closed No ICR - - 1994 MOZAMBIQUE SECAL C2628 Econ.Rec.Cr.I1 P6337 X X - 4 1986 NIGER SAL C1660 SAC P4185 10845 12104 Jun-93 u* 1987 NIGER SECAL C1833 P.E. Sector Adj. P4522 11366 12104 Jun-93 U* 1994 NIGER SAL C2581 Econ.Recov. P6264 15772 X - U Date of No. of Lending Loan Report Latest OED Loans per FY Country Instrument Credit Project Name No. PCR/ICR PAR Evaluation Rating Ctry 1984 NIGERIA SECAL L2345 Fert.Imports P3574 X 9673 Jun-91 i U* 1987 _ NIGERIA SECAL L2758 Trade Policy I P4369 12611 14820 Jun-95 S 1989 NIGERIA SECAL L3011 Trade&inv.Policy P4779 12611 14820 Jun-95 U 1990 NIGERIA SECAL C2139 Educ Univ.DEv. P5172 X X - - 1991 RWANDA SAL C2271 SALI P5544 14873 -NICHOLAS Jul-95 u 1987 SAOTOME SAL 1825 SAL P4586 13303 X JuI-94 S 1990 SAO TOME SAL C2165 SACIIITA P5308 X X - 2 C1084 1981 SENEGAL SAL L1931 SAL 5637 May-85 S* 1986 SENEGAL SAL _C1656 SACII 8695 May-90 S* 1987 - SENEGAL SAL C1802 SALIII 8695 May-90 S* 1_____0 T__ _____18 13- __ ay9 - 1990 SENEGAL SECAL C2077 Bank/Fin SE P5183 13724 14818 Jun-95 S 1990 l SENEGAL SAL C2090 SALIV P5183 13723 14818 Jun-95 U 1994 SENEGAL SAL 02582 Econ.Recov.Cre P6273 15874 X Jun-96 S 1995 SENEGAL SECAL -C2681 P.S.Adj.&Comp P6522 Closed No ICR - - 1995 SENEGAL - SECA L 1C2738 - - Ag.Secal P6610 X X - - T- 8 1984 SIERRA LEONE SECAL -C1501 i Ag.Sector I P3694 X X - - 1992 SIERRA LEONE - SAL C2352 Import Support P5731 14556 X Jun-95 S 1994 SIERRA LEONE SAL C2546 SAC P6157 X X - - 3 Date of No. of Lending Loan Report Latest OED Loans FY Country Instrument Credit Project Name No. PCR/ICR PAR Evaluation Rating per Ctry 1 9 8 6 S O M A L I A S E C A L C 1 7 1 1 A g . S e c . A d j . P r o g P 4 2 6 5 1 0 1 6 2 X D e c - 9 1 U * 1989 SOMALIA SECAL C2030 SAL/ASAPII P4995 12435 X Oct-93 U 2 1980 SUDAN SECAL C1000 Agr.Rehab. P2731 6170 May-86 S* 1983 SUDAN SECAL C1389 Agr.Rehab.1l P3591 7342 Jun-88 U* 2 1981 TANZANIA SECAL C1133 Export Rehab. 7347 Jun-8 U* 1987 TANZANIA SECAL C1741 Multi-Sec.RHI 10796 Jun-92 S* 1989 TANZANIA SECAL C1969 Ind.&Trade Adj. P4944 12139 Jun-93 S* 1990 TANZANIA SECAL C2116 Agr.Adj.Cr. 13741 X Dec-94 S 1992 TANZANIA SECAL C2308 Fin. Sector i PS629 14828 NICOLAS Jul-95 U 5 1983 TOGO SAL C1365 SAC 1 7361 Jun-88 I S 1985 TOGO SAL C1599 SACII 8832 Jun-90 S 1988 TOGO SAL C1892 SALIII 12086 Jun-93 U 1991 TOGO SECAL C2211 PoplHealth P5395 Closed No ICR - - 1991 TOGO SAL C2194 SALIV P5397 14723 NICOLAS Jun-95 U 1996 TOGO SAL ERAC P6859 X X 1982 UGANDA SAL C1252 Recons.Cr.II 8751 Jun-90 U* 1983 UGANDA i SECAL C1328 Agr.Rehab. 12634 Dec-93 U 1984 UGANDA SAL C1474 Recons.III 8751 Jun-90 U* 1988 UGANDA SAL C1844 Econ.Recov.Cr 10415 12076 Jun-93 U* 1990 UGANDA SAL C2087 Eco.Rec.Cr.II 13687 14758 Jun-95 __ U 1991 UGANDA SECAL C2190 Ag.Sector.Adj. P5431 X X - 1992 UGANDA SAL C2314 SACI Closed No ICR - - 1993 UGANDA SECAL C2496 Fin.Sector Adj. P5954 X X - - 1994 UGANDA SAL ]C2608 SACII Closed No ICR - - 9 Date of No. of Lending Loan Report Latest OED Loans per FY Country Instrument Credit Project Name No. PCRICR PAR Evaluation Rating ctry 1986 ZAIRE I SECAL C1708 Ind.Sector Adj. P4227 8838 Jun-90 U* 1987 ZAIRE SAL C1831 SAL P4596 11832 Apr-93 U* 1984 ZAMIà SECAL L2391 Export Rehab. 4624 _1 10847 Jun-92 1985 ZAMBIA SECAL 01545 Agr.Rehab 10341 May-93 _U* 1986 ZAMBIA SECAL C1630 IndOrientation ____10846 Jun-92 U* 198 ZAMBIA SECAL C1720 Recov.Credit 11555 12100 Jun-93 U* 1991 ZAMBIA SAL C2214 Recov.Credit P5483 12437 13238 JuSn-94 5 ____ 1992 i ZAMBIA SECAL C2405 Pvtnlind Closed No ICR - 1993 ZAMBIA SECAL C2523 PIRCII P5786 X X - - 1994 ZAMBIA SAL C2577 ýEcon&Social Adj P6248 15837 FY97 rev Jul-96 NOT REV 1996 ZAMBIA SECAL l- Ec.Rec&Inv.Pro P659 X X 1983 ZIMBABWE SECAL L2239 Manu.Sectxp. 8888 Jun-90 S* 1992 ZIMBABWE SAL L3430 SAP P5684 13842 14751 Jun-95 - S 1993 ZIMBABWE SAL C2527 SACII P6085 X x -_- 3 (Rene) 3__ _ _ _ _____ _ _ _ _ _ _ _ _ _ _ _ _ _____ _____ _ _ _ _____ ____ TOTAL (Ren 163 TOTAL RATED 121 closed but no ICR 9 closed but ICR dated JUL 96 2 Active operations (including new approvals in FY(96) 31 74 Annex 16: Trends in Public Expenditures in SSA Countries Increased spending on social sectors, in particular health and education was intended to be an important component of the African SALs/SECALs, in the mid to late-80s. Trends in education and health expenditures over time can measure the success achieved in this area. Analysis of trends is, however, complicated by the paucity of data on these sectors. Data on health and education expenditures are available for 18 countries, in the Government Finance Statistics (GFS) database, but the data ends in the 80s, for 12 of the 18 countries, thereby providing no information about the current situation. For six countries with data in the 90s, 1995 data is available for only three. Attempts to use data from PER documents met with limited success. Between July 1, 1993 and June 30, 1996 formal or informal public expenditure reviews were completed for 26 of the 35 countries under consideration (excluding C6te d'lvoire, Gambia, Mauritius, Sao Tome & Principle, Senegal, Somalia, Sudan, Togo and Zaire). For only five countries sufficient data were available to permit calculations of total expenditures on education and health as percent of GDP. For several other countries the information was partial or not in usable form. (For example, in Guinea-Bissau, only recurrent expenditures were found in the document, in Mauritania, only budgeted expenditures buy not actual expenditures, in Mali the expenditures were only in constant 1987 CFAF, in Burundi only expenditures on an aggregative category "social services"). A third potential source, the Poverty Assessments were also explored. Draft or completed Poverty Assessments for 26 of the 35 countries in the sample were available in various stages (gray, green, white, and yellow). From these assessments, actual health and education expenditures were available for only three countries. For other countries, the data sometimes reflected programmed by not actual expenditures (for example, Guinea-Bissau) or were partial (for example, for Madagascar health expenditures from 1988-92 were available, bur for education, only for 1985-1990). A fourth data source used was the IMF Recent Economic Development Reports (RED). The study combined data from the four sources to provide data on education, health, and total expenditures as a percentage of GDP for 18 countries. Specific sources are the GFS (Mauritius, Madagascar, Zambia), PER (Guinea, Equatorial Guinea), Poverty Assessment (Tanzania) and REDs (Gambia, Ghana, Malawi, Mozambique, Sierra Leone, Burkina Faso, C6te d'lvoire, Senegal, Togo, Uganda, Zimbabwe, and Kenya). REDs were preferred for some countries over GFS even though data was also available in GFS (example Kenya) because RED data was more up to date. Since most adjustment programs included conditionality on public expenditures only since the mid-80's, pre-adjustment period for almost all countries is an average of two years, between 85-88 depending on availability of data and the adjustment period. Lack of data prevented consideration of a longer time-period average. The pre-adjustment period was compared to the most recent two years. Of the 14 poor adjusters, data was available for only two, so comparison is restricted between the strong and weak adjusters. While more extensive data on total public expenditures was available than for social expenditures, for comparison purposes similar adjustment periods and countries were used for total expenditures. 75 Edu.Exp as % of GDP Hith Exp. as % of GDP STRONG COMPLIERS Pre-adj MR2 yrs Pre MR Pre MR BENIN GAMBIA' 90-91 94-95 0.4 2.0 0.8 0.3 GHANA 87-88 93-94 3.4 4.9 1.2 1.5 MALAWI' 85-86 93-94 2.5 2.6 2.0 1.5 MALI MAURITANIA MAURITIUS' 85-86 94-95 3.4 3.8 1.9 2.1 MOZAMBIQUE 84-85 90 3.8 5.1 1.6 1.4 SIERRA LEONE' 89-90 94 1.2 2.3 0.5 0.9 TANZANIA4 86-87 93-94 1.7 2.1 1.2 1.8 AVERAGE 2.3 3.2 1.3 1.3 WEAK COMPLIERS BURKINA' 88-89 92-93 2.8 3.0 1.0 1.2 COTE' 87-88 94-95 6.9 5.1 1.4 1.4 GUINEA2 88 93-94 2.0 2.5 1.3 0.6 GUIN-BIS MADAGASCAR3 88-89 94-95 2.3 2.6 0.9 1.3 NIGER SENEGAL' 85-86 94-95 3.9 3.4 0.7 0.6 TOGO' 85-86 93-94 4.1 3.2 1.2 0.9 UGANDA' 90 94 1.1 1.1 0.3 0.4 ZAMBIA' 87-88 94-95 2.8 2.7 2.0 2.4 ZIMBABWEI 87-88 96 9.2 7.7 3.1 3 AVERAGE 3.9 3.5 1.3 1.3 POOR COMPLIERS BURUNDI CAMEROON CAR CHAD CONGO EQ. GUIN2 88 92 1.9 1.5 1.4 1.2 GABON KENYA' 87-88 94-95 6.7 7.1 2.0 2.2 NIGERIA' RWANDA SAO TOME SOMALIA SUDAN ZAIRE SAMPLE AVERAGE 3.3 3.5 1.3 1.4 Notes: RED 2 PER GFS 4Tanzania:The Challenge of Reforms:Growth, Incomes and Welfare,1996 Countries in bold are those for which data are available, 76 Military Expenditures as Percent of GDP Miltary Exo Implementation (1.5-2.2)1 Rating OED(%S) CFA Proai, All MRAyr BENIN 2.1 100 X 2.1 1.5 1.4 GAMBIA 1.6 100 1 GHANA 1.6 80 0.6 0.8 0.7 MALAWI 2.2 83 3.2 2.1 1 MALI 1.8 100 X 2.2 2.5 2.5 MAURITANIA 1.8 100 6.1 3.8 3.8 MAURITIUS 1.3 100 0.3 0.3 0.4 MOZAMBIQUE 1.8 100 10.8 9.3 7.6 SIERRA LEONE 1.1 100 2 3.8 3.8 TANZANIA 2.1 75 3.3 1.9 Average 1.7 3.4 2.9 2.5 EXCL. MOZAMBIQUE 2.5 2.1 1.9 Implementation(2.3-2.9) BURKINA 2.7 UR X 2.3 2 2 COTE 2.8 63 X 1.1 1.1 1.4 GUINEA 2.5 75 GUIN-BIS 2.9 33 5.2 2.3 MADAGASCAR 2.8 75 2.5 1.5 1.1 NIGER 2.6 0 X 0.7 1 SENEGAL 2.6 80 X 2.6 2.1 TOGO 2.6 50 X 2.2 2.8 2.8 UGANDA 2.6 0 2.9 2.8 2.4 ZAMBIA, 2.6 2.3 3.4 3.1 ZIMBABWE 2.7 100 Average 2.7 2.4 2.1 2.1 Implementation(3-4) BURUNDI 3.2 0 3.2 3.1 CAMEROON 3.4 0 X 1.2 1.5 1.5 CAR 3 0 X 2 CHAD 3.7 0 X CONGO 3.1 0 X 3.3 EQ. GUIN 4 0 X GABON 3.2 0 X 3.1 KENYA 3.1 67 3.6 3.5 1.9 NIGERIA 3.1 50 1.5 0.9 0.7 RWANDA 3.6 0 2.2 7.2 SAO TOME 3 100 SOMALIA 3.8 0 2.9 1.9 SUDAN 3.5 0 2.2 2.5 ZAIRE 4 0 1.5 2.2 3.3 Average 3.4 2.4 2.9 1.9 AVERAGE FOR SAMPLE 2.7 2.7 2.6 2.2 EXCL. MOZAMBIQUE 2.4 2.3 1.9 Sources: SIPRI Yearbook, 1996, Armaments, Disarmament Peace Reaserch Institute, and staff estimates.and International Security; Stockholm International International Financial Statistics, IMF 1996. IMAGING Report No.: 16594 Type: OES