Senegal - Public expenditure review

Report No. 11559-SE Senegal Public Expenditure Review May 27, 1993 Africa Region Sahelian Department Country Operations Division FOR OFFICIAL USE ONLY MICROFICHE COPY Report No.:11559-SE Type: (ECO) Iratle: PUBLIC EXPENDITURE REVIEW Author: OSEI, ALBERT Ext.:34952 Room:J 9105 Dept.:AF5CO Docuwent of the Worid Bank This document has a restricted distribution and may be used by recipients only in the perfornance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  FOR OMCIAL USE ONLY SENEGAL UBL EEmNDUREV TAKLE DECONIaN RXECIrUIVE SUMMARY ..... ................................ v L I?MRODUCTION 1........ ... . 1 X MACROECONOhC FRAMEWORK ......... 2 A. Recen Macroecononic Perfmance .............. 2 B. Exel Trends ........ . 3 C. T inknRwen ... ....... . . . . . . . 9 D. MediumTermMo utlook. ... .. . .. . 11 IL SECrORAL E NPEND1 RE PROGRAM, ... ..... . 18 Education .2. 22 A.grkuhi dend aReourc .................. 27 .. ... 27 Road - I ---atoure......... ...... ... ...... O.*. ...~.... 32 Eninrgy Sector ...................... . ....... 35 Summnafyof ExpWmdiurPrposais fr Plor1tySoctom .. ........3 Iv. EXPENDTUREPLANNING AND DIPLEMENTATION 44..... A. Imdesi ...................... 44 B. do ..................... 4S C. VeNI o ing and Execution.. 48 D. R.e...n.endatio.s ..S3 ANJNETX 1: DATALSSIUES ....................................... 55 ANNEX 2: DETALED SECrOR REPORIS Health .................... .. ... ..... 59 E....*...................... 74 A.grluiuhred Natral.eourcee.8 Road I ructe .. . .... .101 This document has a restricted distribudon and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  STATISTICAL ANNEX Table 1: Central Govenment Revenue by Source Table 2: Ceral Coverment Revenue in percent of GDP Table 3: Remrent Expenditure by Economic and Functional Claificaton (80/81) Table 4: Recurret Expenditure by Economic and Functional Classification (82/83) Table 5: Recurnt Expenditure by Economic and Funcdonal Classification (84/85) Table 6: Reurrent Expenditue by Economic and Functional Classiflcation (8687) Table 7: Recurrent Expenditure by Economic and Functional Classification (87/88) Table 8: Recurrent Expenditre by Economic and Function Classfcation (8/89) Table 9: Recurrent Expenditure by Economic and Functional Classification (89/90) Table 10: Recurren Expenditure by Econmic Classiflcation (80/81-89/90) Table 11: Recurrent Expenditure by Functional Classification (80/81-89/9 Table 12: Invesment Expenditure by Functional Classification (81/83-89/90) Table 13: Total Expenditure by Functional Classification (81/83-89J90) Table 14: Execution of Investment Expenditures (PTIP) 1988/89 Table 15: Execution of Investment Expenditures (PMIP) 1989M90 Table 16: Financing of Investment Expenditures 1988/89  This report is based on the fndiWn of a mission that visited Dakar in April/May 1991. Tho mission consisted of Mr. Albert Osei (Mission Leader), Mr. Heak Koppen (Consultant), Ms. victoria Kwakwa (Economist), and Mr. David Jones (Agricultnur Eonomist). Contributions were also received from Mr. Rend Vaurs (Senior Economist, AF5PE), Mr. Alassan Diawara (AFSE), and Ms. Brigitta Mitchell (Senor Economist, AFSN). Mr. Richard Wesn served as peer reviewer. Ms. Katherine Marshall is the Departme Director, Mr. Franois Laote, th Department Lead Economist, and Mr. Jean-Louis Sabib, the managing Divhision Chief. Secretarial support was provided by Ms. Myrina HaTris and Mr. Mather Pfiffenberger. This report doer nt reflect developmen since the end of the 1991 fiscal year. Since then, the financial siuaon in Senega has considerably worsened. The wage bill is rising again and total accumulat arrears are increasing at an alarming rate. The maroeconomic fiamework is seiously off track and specific conditions related to the disbursement of the last tramche of the ongoing strctural adjustment operation are becoming less meaningl. Neverheless, the Public Expenditure Review remai very relevan for fiuture opetional work on Sengal. It represents a first attempt at a retrospective analysis of public expenditures in Senegal and gives a detailed survey of the issues at the sectoral level. Should the macroeconomic siuation improve in the near futre, it would provide an invaluable guide to adjustment Issues following stabilization.  Senegal Public Expenditure Review Executive Summary i. This report sets out the findings of a Public Expenditure Review conducted in May/June 1991 and based on data for the 1981-1990 decade. It should be read in conjunction with the Macroeconomic Update Report appearing at the same time, which focusses on the broader issue of macroeconomic performance during the same period and which reaches the same broad conclusions on fiscal performance as this study. The findings contained herein represent the views of Bank staff; nevertheless, the basic task of reconstructing a data series was done by a joint Bank-Senegal team. The convergence of views between the Bank and Senegal on the data thus lends comfort to the expectation that the diagnoses are largely shared as well. ii. The major objectives of the study, the first formal review of this nature carried out by the Bank in Senegal, have been: a) to check the completeness and consistency of the data base on public expenditures; b) to seek insights into the nature and extent of possible expenditure restructuring plans that would more closely respond to Government's expressed development priorities; c) to identify institutional and procedural reforms that would improve the management of the process for establishing expenditure plans, implementing them and measuring the results of such spending; and d) to indicate public expenditure issues that require further study. iii. After a brief introductory chapter, the report sets the context for an examination of public expenditures by a brief look at the outcome of macroeconomic policy, and especially aggregate revenue and expenditure trends, during 1981-90 (Chapter II). The chapter also recalls the macrofinancial framework contained in the most recent Policy Framework Paper (PFP), which sets out the factors likely to condition expenditure restructuring efforts in the next 2-3 years. iv. Four major observations emerge from the retrospective review with respect to public expendituress a) There has been good progress towards meeting the stabilization objectives of the adjustment program. In particular, fiscal, monetary and external balances have improved.  - vi - b) The tools used to arriva at these results have been flawed, in the sense that on both revenues and expenditures, the decisions taken are either inimical to stated development objectives or are not sustainable. C) Neither the level nor the composition of expenditure has been adjusted enough to make a real and lasting impact on improving competitiveness, which has been the over-arching objective of adjustment efforts in the eighties. d) Overall, the judgement has to be that there has not been active, aggressive use of fiscal management as a policy instrumentl rather, Government seems to have consistently taken the course of least resistance. This is particularly troublesome given the central nature of fiscal adjustment for the strategy Senegal has chosen. v. Stabili&ation Gains. By all the usual summary measures, Senegal has made significant strides towards macro-economic stabilization during the eighties. In 1981/82, the overall fiscal deficit, on a commitment basis and excluding grants, was 8.0% of CDP; by 1990/91, there was an approximate balance between budgetary revenues and outlays (a budget surplus equivalent to 0.2% of GDP was recorded). On a cash basis (i.e., taking account of the changes in domestic and external arrears) and excluding the effects of non-continuing expenditures (for paying separation grants to civil servants and for meeting the Cobnroment'l obligations arising from banking reforms), fiscal performance has been consistently positive, with a declining deficit from 10.8% of GDP in 1981/82 to 1.7% in 1990/91. These results have in turn allowed improvements in the current account balance (a deficit of 24.3% of GDP at the beginning of the period had been reduced to 7.8% by the end) and in the overall resource gap (from CFAF 130.2 bn. to CFAF 63.4 bn). vi. Instruments of Fiscal Adiustment. The instruments used to attain these gains contain the seeds of future trouble, are generally non-sustainable, and have not infrequently worked against the achievement of stated development objectives. Broadly speaking, expenditure reduction rather than revenue mobilization has carried the fiscal adjustment: Revenues have gone from around 23% of GDP in 1980 to about 18.9% in 1991; in contrast, total expenditures, which were 33.8% of GDP on average during 1981-83, had fallen to 23% of GDP by 1989/90. There are several things wrong with this. vll. On the r"venue side, the low yields have been in spite of an lncrease in the avorage rate of taxation on the formal sector and tax revenues as a share of GDP has gone down even more (18.2% to 15.2 percent) than the overall revenue-to-GODP decline. This double weakness- -efforts to increase revenues have led to disincentive taxation rates on the formal private sector which has shrunk partly as a consequence,  - vii - leading to tax revenues growth that has not only not kept pace with CDP but has in fact been negative--dramatizes one dilemma in fiscal managements tax at higher rates in order to generate sufficient revenue although thli tends to cause the "informalization" of the tax base which triggers a new round of increased rates. A second implication of this manner of reverue mobilization has been the increased dependence on "extra-normal" tax measures, in this case the appropriation by the public exchequer of the "windfall" profits accruing from the fall from unusually high levels in the world petroleum price (and similar implicit duties on imported rice). These measures have had perverse effects, in the one case of increasing energy prices for local producers of tradeables and in the other case of keeping the price of the main wage good high with the predictable result that wages are subjected to upward pressue. To complete the picture of weaknesses in the revenue generation policy during the eighties, reference needs to be made to the increasing role of external sources in financing public expenditures. External finance rose from 31.8% of total public expenditures in 1981/82 to 41.0% in 1989/90 and, worse, budgetary (non-capital) external finance, insignificant in 1980/81, peaked at 59.3% of all external flance and almost 25% of total expenditures in 1987/88. viiL. On the expenditure side, the robust reduction would be irable if it were a managed outcome. In fact the sectoral and functional incidence of the reductions leave one with the inescapable conclusion that it happened by default rather than by design. ix. The first remark is that investments as a share of total expenditure fell from nearly 40% on average during the early part of the decade (1981-83) to 32% in the 1989/90 fiscal year. In real terms, there was a reduction of some 30% in investment spending by the public sector during the period. In this falling share, the big gainers in terms of economic classification were: General Public Services (which accounted for two-and-one-half times more proportionally of investment expenditures at the end of the period than at the beginning), and agriculture (88% more). Health (38 percent) and education (6 percent) also saw some increae but from a low base; there was also a high proportion of recurrent expenditure hidden in the education and health figures because of the convention of treating all project-related external assistance as capital expenditure. The high share of agriculture also highlights a cause of distortion, as described below (para. xi). X. Within recurrent expenditures, the trend was initially in the right direction until 1986/87, when they began to rise again, with the result that by the end of the decade, real recurrent expenditures had more or less caught up again with the beginning-of-period levels. The major cause of this inability to continue to contain recurrent expenditures was, of course, the large interest burden arising from the (largely foreign) borrowing used to forestall adjustment after commodity price and petroleum price shocks in the mid-to-late seventies. Within these broad developments, however, there were again sectoral and functional distributions that give pauses Compared to expenditures at the beginnLng of the period, recurrent expenditures (in constant prices) on health had gone down by 26 percent, whereas defense and general  - viii - pwalc sevices wero only, resptiv ly, S and 9% lower; and the wage and salary blll wnt down only 6.3% compred to 24% for xpenditures on operations ad maintnanc. xi. Fiscal oliCy and cutitiye. Senegals adjustment strategy depends for its success on sustained dmestic deflation to improve competitivenoe. Given the limited influence of national authorities on monetary (credit and interest rate) and exchange rate policies in the West African xonetary Union (uM0), management of the size and composition of public sector revenues and expenditures become the principal method by wh1ch the public authorities can affect overall demand and thus the general level of domestic prices as well as the relative prices of tradeables and nvan-tradeables. The conduct of fiscal policy, as briefly descrlbed above, has frequently had procisely tho opposite effect on the overall price level. To start with, the persistently high wage bill is the combined effect of both a buoyant demand for labor by the public sector (employment grew by about St in the 10-year period) and an average civil service salary which was abnormally high to begin with (on average Senegalese civil servants are paid 9 times per capita GDP compared to 3 times in a sample of non-CPA Franc countries). This eontitutes a major expenditure on a non- tradeableg further, the public sector wage level tends to drive the level of wages in the formal sector as a whole. Finally, at the aggregate level, the large size of the public sector as a share of GDP reduces resources and increases costs for the private sector. The high energy price is a case ln point; another source of high costs for the private sector derlves from the persistence of public sector deficits which has meant reduced access to credit and higher interest rates for the private sector. In the eighties, in spite of the effort of deflation, the real exchange rate is still significantly overvalued. xii. Actilve Manaeant of Flscal Policy. Given these consequences of fiscal policy in the eighties, it would be legltimate to ask whether tho Senegalese authorities have consciously chosen these outcome or, as other comentatorc have recently asserted, have beon allowed to postpone adjustment* by the unu.sually high amounts of external finance that was available to the loverument. The major indicators of the failure to make expenditur* shifts are worth citing againa Zxpendltures on General Public Services (lncluding Defense and Law and Order) have been protected in a period when outlays on Uconomic Services were being reduced; Salaries and Wages have decllned but not nearly as rapidly as have the operating funds needed to ensure that civil servants can perform their functions; - Covernment consumption has been favored over PublLe Investment.  - 13s - ThS Main recommendations of this report, relating to expenditure reStructuring have been made notwithstanding thls hlitory of falled or incomplete adjustment because the external resources that contributed to making it possible are not likely to be available to Senegal in the future. The argument made ln tho report is that the full impact of the years of lneffectlve fiscal management has perhaps never been fully conveyed to the authoritles, and that faced with the evldence, the necesasty for declsive action can no longer be denied. xili. The review has examined expenditure policy ln the key sectors of Health, Education, Agriculture and Natural Resources, road transport and energy (Chapter 11). The choice of sectors has been influenced by the apparent consensus that they represent areas wher Government lntervention is necessary. For each of these sectors there exists a clear statement of sectoral development policy agaLnst whlch the sufficiency and effLieency of actual expendltures in the recent past may be judged. The conclusions may be summarLsed as follows: In Health, the priority needs are to shlft resources (especially staff) from Dakar, Lncrease allocations for materials and maintenance, and much more for drugs (effectlve cost sharlng programs could and should pay for a large percentage of drug expendLtures). It is recommeAded that the shares of salaries, operations and maintenance, an drugs change from 74S24:S in 1989/90 to 48:30s22 by 1995 and that, in real terms, expenditure increase by some 54% above 1989/90 levels so that health sector expenditures would be 6.7% of total expendltures for 1995 compared to 4.1% in 1989/90. For Education ln order to meet the target enrollment ratio of 6S% by 1995, to improve quality, and to make slgnlficant internal effLeLency gains especially at the universLty level, an increase ln the share of the total recurrent budget going to the sector would be necedlary, of the order of 5 more percentage points of total expenditure in 1995 compared to the 1986-89 average. Intra-sectoral realignments in the recurrent budget would also be required, so that the prLmaryssecondarysuniversity/other shares will go from 41:25s34 to 53:28:19. For the investment budget, the major need is for greater coordination of external aid donors' programs so that exLsting and planned funds can be redirected to better respond to national priorities. Itt AarLculture and Natural Resources, better oversight of expenditure would require consolidating all sectoral programs ln a aingle budget (instead of the present split between mlinitry-accounted services and the expenditures of parastatals such as SED). Among other things, the gross imbalances between sub-sectors and among functions would be more readily apparent. At a omnimum, to begin to correct these imbalances, a major reduct$on in the servicee for Lrrigated crop production would be esskntial. Fortunately, thli should be feasible glven the demonotrated  inefficiencies in the sub-sector. In real terms, economies of some 33% and over 50% are possible in the cost of services for, respectively, rainfed and irrigated crop productlon. in the cae of rainfed crop production services, the economies will derive in large part from reductions in personnel for e.g. cooperative development and defunct reional development agencies. In irrigated agriculture, the bulk of the savings will come from autring back on "zehabilitation 0, maintenance and mechanization services all provided by SAND. In contrast, services and programe that have been underfunded (Research, Forestry and the Environment) will need to be allocated more resources than in the past. Overall, the suggested restructuring could mean savings of some CFAP 27.0 bn (28 percent) in 1995 compared to planned expenditures in 1990/91. For the and sTans subsector, no significant reallocation is required principally because the current expenditure program has been arrived at in the context of a multi-donor funded sector adjustment operation which has correctly Identified priorities. In the Enerow sector, the only substantive recommendation is to rationalize the constitution, funding and management of the Energy Fund. This Fund is theoretically sourced from the "excess profit tax on petroleum imports. In fact, the bulk of funds from thls source (more than 80% in 1988-90) support the general budget. Further, Its mission, which includes the promotion of renewable energy sources and rural electrification, has been dominated by another of its original objectives, to subsidize the energy consumption of the phosphate companies to compensate for the competitive disadvantage c operating in a high energy cost environment. The recommendation seeks to provide the Fund with a predictable source of revenue, to limit eligible projects to those that involve Oalternative domestic renewable" energy sources, and to streamline its operations. xiv. The overall effect of the recommended expenditure roalignments, taking the 1995 budget as the target horizon, would be that the priorlty sectors discussed above will require only one percentage point more of GDP in 1995 than was allocated in 1989/90. In constant prices, the increase in 1995 over 1989/90 would be 33 percent. Given this level of funding for the priority sectors; given the PFP projectLon of tax and non-tax revenues and of the target deficit; and given Interest obligations falling due in 1995I allocations to the non- priority sectors would need to be some 28% lower. Thus, although target expenditures on priority sectors are not disproportionately higher as a share of GDP in 1995 than five years earlier, accommodating this plan would require major restraint of expenditure elsewheres the 28% real reduction is a fair measure of the challenge of expenditure management in the nineties that Senegal is faced with.  - xi - xv. Any serious effort to come to terme with the problem will need to confront and deal with tne myriad sources of inefficiency and lack of transparency that characterize expenditure planning.and implementation processes in Senegal (Chapter IV). The principal problems highlighted area inflexible and non-rooponsive resource allocation proceduresl lack of coherence between capital and recurrent budgetsa poor organization and incompleteness of data on public expenditures; implementation procedures that are redundant, cumbersome and not geared to seeking efficiency in resource use so much as to establish the legality of transactions. Investment planning, which has benefitted from several years of technical assistance, is still merely a formal exercise and does not serve to haelp choose between competing alternative weso. It iL neither comprehensive nor always technically well-done. xvi. Detailed proposals for reform have to await further discussions between the Bank and the Senegalese authorities, which are scheduled to take place before the end of 1993.  1. This rport presents the flndings of the first fomal Public uxpeniture Review for Senegallf. as an Laltial attmpt, the aim ha bee limited to "xmbling, collating and checking baLc expendLture data for consistency and cohernce and examining recent fLical uianajement in OrL d to provild insights Lato the nature ad thb extent of such expenditure rstructurLng as may be required to meet medium-trm objectives in the key sectors of health, eduoation, agriculture, natural resources, transportation and energy. 2. The flrst chapter dLicusses rocent trand, broad xpendlture and revenue pattrns and the projected macroframeork through 1994/95. Chapter rI provLdes a detailed -eview of expendlture programs and Lisues in priority sectors and proposals for improving both the quality and the volume of the recurrent and investment expenditure in these sectors. The chapter ends with a broad estimate of desirable expenditure allocations between prlority sectors and non-priority sectors, glven the PIP based fLscal deficit targets. The flnal chapter discuss key Lasues in expendlture planning, Progracig and