Report No. 32479-SN Senegal Enhancing the Efficiency of Public Investment Public Expenditure Review June 1, 2005 PREM 4 Africa Region Document of the World Bank TABLE OF CONTENTS EXECUTIVESUMMARY ........................................................................................................................ vi 1 ENHANCINGTHEEFFICIENCYOFPUBLICINVESTMENT . PUBLICEXPENDITUREREVIEW .................................................................................................. 1 2.ATHE .Introduction....................................................................................................................................... ROLEOFPUBLICINVESTMENTINSENEGAL: TRENDS AND OUTLOOK ..............13 A Definitionandstatisticalsources B A look backwards:A temporaland internationalcomparisonof public investment .................34 C Public investment:An opportunityfor the future? .. . ...................................................................................................... ........................................................................ 7 C.1 Debt sustainability ........................................................................................................................ 7 C.2 Preservingprice stability .............................................................................................................. 9 .3 Fiscalpolicy and external equilibrium....................................................................................... I O 11 3.DPUBLICINVESTMENTAND .CConclusion........................................................................................................................................ ECONOMICGROWTH 12 A PublicInvestmentandEconomicGrowth: Not a CertainRelationship 12 B Composition ofInvestment:I sitOptimal? .. .............................................................. .................................... ....................................................................................... ................................................................................... 15 C Capitalexpenditureand recurrentcosts 22 D Privatesector participation .. ............................................................................................................ D.1 Crowding out ofprivate investment is minimal .......................................................................... 24 25 . 2 Increasingprivate sector privatization by maximizing complementarities ................................ 25 27 4.EPUBLIC .0Conclusionand recommendations ................................................................................................. INVESTMENTAND BUDGETARYINSTRUMENTS ......................................................................... ................................................. 30 A Strategicchoices in Senegaland budgetarytools 30 B Strengtheningmedium-termplanning .. 32 C InvestmentsapprovedandImplementedinaccordance with PRSPpriorities . .......................................................................................... ......................... 34 C.I TheBCI as approved and the objectives of the PRSP ................................................................ 36 BCI as executed and PRSP objectives ........................................................................................ 38 39 5.DINVESTMENTPROJECT .C.2 ConclusionandRecommendations ................................................................................................ EXECUTION RATES 42 A InvestmentExpenditureExecutionRates 42 B Determinantsof the VariationinExecutionRates .. ..................................................................................... ........................................................................ ....................................................................... 44 B.1 Statistical analysis of the determinants....................................................................................... 44 Level of centralization of the implementing agency.................................................................... 46 47 6.CINVESTMENTBUDGETMANAGEMENT .B.2 Recommendations ............................................................................................................................ SYSTEM ................................................................. 49 A Preparationofthe Government'sInvestmentBudget ................................................................. 49 B InvestmentExpenditureExecutionProcedures .. ........................................................................... 52 B 1 Procedurefor investment expendituresfinanced with national resources ................................. . 54 56 C Government procurementprocedures D Recentinitiativesaimedat improvingthe management of investmentexpenditures .. B.2 Investment expendituresfinanced with external resources......................................................... .......................................................................................... ...............60 58 D.I New accounting andfinancialframework for investment expenditures ..................................... 60 62 E Recommendations .0. 2 Decentralization of public investment expenditure management ............................................... ............................................................................................................................ 65 ... 111 ANNEXES Annex A : Rrecent Developments inGovernment................................................................................... 69 Annex B: Public Investment and Economic Growth: a Simple MacroeconomicModel........................ Annex C: Overview o f Senegal .............................................................................................................. 78 76 Statistical Appendix ................................................................................................................................. 81 Table 3.1: Economic growth and investment. 1966-2000 ...................................................................... 14 Table 3.3: Infrastructure investments and impact on growth................................................................. Table 3.2: Sectoral allocation o fpublic investment, 2000-03 ................................................................. 16 -19 Table 3-5: Central Government equity interests inthe public enterprises ............................................. Table 3.4: The "r" coefficients by sector ................................................................................................ 24 Table 4.1: Priority action plan (2003-2005) ............................................................................................ -26 31 33 36 Table 4.4: Project amounts includedinthe BCI, 2003-2005 .................................................................. Table 4.3: Share o f the BCIallocated to combating poverty.................................................................. Table 4.2: Three-year forecast o f investmentbudget, by sector ............................................................. 39 Table 5.1: Execution rate by type o f expenditure ................................................................................... Table 4.5: Coverage o f strategic objectives and BCI as executed, 2003 ................................................ 40 43 Table 5.2: A statistical explanation o f executionrates, 2002.................................................................. Table 6.1: Expenditure circuit for investment expenditures financed with national resources...............45 55 Table 6.2: Quality of procurement procedures inAfrica......................................................................... 59 FIGURES Figure 2.1: Sustained growth o fpublic investment .................................................................................. Figure 2.3: Major sources o f BCI growth, 2000-2005 .............................................................................. Figure 2.2: Public investment ratio (2001-2002): Senegal i s close to the average for Africa .................456 Figure 2.4: Debt/GDP ratio (net present value) ........................................................................................ Figure 2.5: Debt service/exports ratio....................................................................................................... 8 8 Figure 2.6: Composition o f financing sources for the fiscal deficit, 2000-2005 ....................................... 9 10 Figure 3.1: Infrastructure stocks are low in Senegal............................................................................... Figure 2.7: Relationshipbetween public sector and external sector, 1994-2004.................................... 17 37 Figure 5.1: Investment expenditure execution rates by sector, 2000-2003 ............................................. Figure 4.1: Alignment of the BCIwith the pillars of the PRSP.............................................................. 43 Figure 6.1: Time frame for preparing the budget law............................................................................. 50 Box 3.1: Evaluating investment inthe social sectors.............................................................................. 21 Box 3.2: Measuringrecurrent costs ........................................................................................................ 23 Box 4.1: A Methodology for evaluatingthe regional allocation o f investment the example Box 4.2: The Ugandan Experience ......................................................................................................... of the health sector inBrazil ..................................................................................................... 35 41 Box 5.1: Characteristics o f the sample.................................................................................................... 44 Box 5.2: Urban development crdit and decenralization program .......................................................... -46 Box 6.1: Government agencies and specialized funds............................................................................ 53 iv ACKNOWLEDGMENTS 1. This report was prepared by a team comprising Jacques Morisset (Task Team Leader, AFTP4), Julien Bandiaky (AFTP4), Alex Segura-Ubiergo (IMF), and Emile Finateu (AFTFM). Within the World Bank, we benefited from the support and comments of Fily Sissoko, Bourama Diaite, Meskerem Mulatu, Iradj Alikhani, Julie Van Domelen, Eric D e Roodenbeke, Robert Blake, Antonio Estache, Jan Walliser, and Waly Wane. Logistical support was provided by Judite Femandes (AFTP4). We wish to thank the International Monetary Fund (IMF) for its participation, in particular for the analysis o f public debt sustainability presented in Chapter 2. Magueye Dia contributed to the chapter on execution o f the Consolidated Investment Budget. Finally, the Govemment o f Senegal supported the preparation o f this report from the outset, in particular by creating a task force headed by the Chief o f Staff o f the Minister-Delegate for the Budget, Ibrahima Sar, in which the major stakeholder units o f the Ministry of Economy and Finance participated. We were also supported by the Poverty Reduction Program Monitoring Unit,amember ofthe afore-mentionedtask force headedbyThiemo Niane. 2. This report is part o f the technical assistance effort on improving public finance management which the World Bank has undertaken to provide to the Senegalese authorities on request. This effort began in 2003 with the production o f the Country Financial Accountability Assessment (CFAA), the Country Procurement Assessment Review (CPAR), and was continued inthe first Public ExpenditureReview in2004. V EXECUTIVE SUMMARY 1. Senegal compares favorably with other African countries, which would be pleased to report an economic growth rate o f over 5 percent for the past decade, with inflation and fiscal and external balances under control. It is precisely this encouraging performance that gives rise to the new ambitions o f the Senegalese authorities, who are well aware o f the fact that the progress made i s not yet sufficient to enable them to meet the objectives that they themselves set intheir development strategy, namely a significant reduction in the number of people living in poverty and a narrowing o f the gap that still sets it behind the world's industrial economies or even emerging economies such as Tunisia, the Asian tigers, and Chile. At the current annual growth rate o f 5 percent, it would take Senegal more than 12 years to reach a per capita income of US$l,OOO (the level recorded today by countries such as the Philippines or Paraguay) or nearly 25 years to move ahead o f Tunisia and Brazil (assuming that these countries were to remain at their present levels). To date, economic growth in Senegal is explainedprimarily by thegrowth of thepublic sector 2. The basic question that Senegalese policy-makers must address i s how to move from a satisfactory performance level to one o f sustained accelerated growth at 8-10 percent a year, which could produce a significant and equitable increase in living standards for the population as a whole. The solution identified i s predicated on the leading role o f the private sector, following the successful examples o f countries like Chile and Mauritius. For this reason, the first pillar o f the development strategy initiated in 2002 is wealth creation, and this vision will be further strengthened through the new concept o f accelerated growth which emphasizes five priority sectors in which Senegal is perceived to enjoy comparative advantages. The role accorded the private sector i s crucial, but i s not to be exaggerated. Indeed, a rapid overview o f the economic expansion observed since 2000 shows that the stagnation o f both private investment and export levels undercuts the argument that the private sector is the primary driver o f the recent expansion. In contrast, public investment has increased by more than 80 percent; it now represents almost half o f all public expenditure and involume terms almost matches the level o f private investment. 3. Two lessons may be drawn from this observation. First, the Government i s correct to draw the private sector into its growth strategy, as this sector's role has been limited, suggesting that its potential i s still great. It is therefore imperative to improve the business climate, which is subject to numerous constraints in Senegal. The second lesson which emerges i s that fiscal policy, in particular as regards public investment, has been the major source o f growth in Senegal in recent years, and it is advisable to question whether it can continue to play this leading role inthe short andmediumterm. vi Increasing public investment isfeasible, but its impact on growth is not automatic 4. The decision o f the Senegalese authorities to continue using public investment as an instrument to promote economic growth i s clear under the 2005 budget, in which this variable increased by almost 30 percent, with a particular focus on infrastructure expenditure. Macroeconomically, such a policy i s sustainable, though it must remain cautious owing to the pressures that would bebrought to bear on the current account balance. 5. This said, it would be incorrect to believe that expanded public investment will automatically trigger accelerated growth in Senegal. Experience over the past two decades shows that public investment and economic growth are not necessarily correlated, or at least that this pairing i s less robust than in the case o f private investment. While there are multiple reasons for this weak correlation, the major challenge facing the Senegalese authorities is to increase the effectiveness o f this instrument, above and beyond its quantitative impact, that is, its impact on aggregate demand, by optimizing its impact on the productivity o f the economy. An efficient fiscal policy must contribute not only to increasing the stock o f physical capital (infrastructure), but also to improving the stock o f human capital (education and health), which in and of themselves should stimulate the growth o fprivate activity. 6. This is no small issue for Senegal, as it is at the core o f its development strategy. Inthis study, we have identified four major challenges that the Senegalese authorities must meet in order to optimize their investment policy. Thefirst challenge is to optimize the strategic choices underlying an optimal investment policy 7. Globally, the lack o f productivity associated with public investment i s explained by three phenomena in Senegal. The first relates to the composition o f public investment, which tends to favor sectors that do not have a direct impact on economic growth. The second stems from the confusion between expenditure on physical capital and the recurrent costs associated with investmentprojects, which pose an obstacle to maximizing the profitability o f public investment over time. The third involves the lack o f complementarity between public investment and private investment.The three phenomena together are tied to the problem o f the Senegalese authorities' limited capacity to reach the strategic decisions required and thereby to optimize their investment policy. A number o frecommendations inthis regard are set forth below: Optimize the composition of public investment. The authorities should improve their capacity to assess the sectoral allocation o f their investment projects. They should also accord greater importance to infrastructure-the share o f which i s still less than optimal by international standards-because o f its unquestionably positive impact on growth. Estimates indicate that by modemizing the road network and electricity grid to the levels achieved by Botswana, Senegal could increase its annual growth rate by almost 3 percentage points. Furthermore, it i s urgent to introduce an infrastructure expenditure management system that will make it possible to assess the economic and social returns o f such spending. This additional focus on infrastructure should not come at the expense o f the social sectors, which fulfill other longer-term objectives; however, it is essential to introduce a system for assessing such expenditure, as is proposed in this study. vii Furthermore, it i s essential to examine the justification for investment expenditure in other sectors, in particular the administrative equipment sector which has expanded by a factor o f four since 2000. The ongoing process o f adopting budget programming by objective should help optimize the efficiency o f allocating public investment. Program the recurrent costs associated with public investment projects. The recent surge in investment expenditure-with its cumulative increase o f CFAF 2,500 billion since 2000-is expected to lead to an increase in recurrent outlays o f CFAF 150 billion to CFAF 200 billion a year. However, the Consolidated Investment Budget (BCI) fails to draw a clear distinction between these two forms o f expenditure, which i s an impediment to properly evaluating Senegal's physical stock and to programming maintenance expenditure over time. This confision also results in a significant decline inthe return on investment projects, even when they are opposite, because o f their chronic lack o f maintenance. Accordingly, it i s recommended that the B C Ibe provided with an economic classification that clearly distinguishes between these two forms o f expenditure, which will subsequently make it possible to harmonize them with the recurrent costs included in the current budget, and to plan them more effectively over time, both globally and by sector. Maximize complementarities with the private sector. The authorities are cognizant o f the linkages between private and public investment. They have therefore managed to minimize the effect o f crowding out the private sector by maintaining a cautious policy toward borrowing on the domestic markets. However, the complementarities have yet to be fully exploited. It i s therefore recommended to continue-indeed, to accelerate-the privatization policy, as the Senegalese Government still holds equity in over 20 enterprises valued at almost CFAF 178 billion. At the same time, following the establishment o f a suitable legislative framework, concluded with the adoption o f the new law on Build-Operate-Transfer (BOT) contracts in 2004, projects with private sector participation should be encouraged, in particular in the infrastructure sector. It i s essential, however, that the Government take account o f the risks associated with such mixed capital projects so as to avoid jeopardizing its financial equilibriums in the long term. The second challenge involves improving the alignment of budgetary instruments 8. Once the aforementioned strategic questions have been resolved by the Senegalese authorities, they must still ensure that budgetary instruments are properly aligned on the priorities defined in their development strategy. The first observation to be made in the case of Senegal i s that such alignment has been observed over the 2003-2005 period, inthat the amounts allocated under the B C I globally cover the three pillars defined in the PRSP. However, this overall assessment masks enormous variations between one subsector and another, in particular inthe pillar on wealth creation and vulnerable groups. Moreover, it would appear that the share o f the B C I devoted to the priority objectives o f the PRSP has not increased over time, holding at about 40 percent between 2003 and 2005. More serious still, the implementation o f priority projects i s inadequate overall-the rate i s lower than the average observed for the B C I as a ... V l l l whole-contributing to reducing the BCI's effective coverage o f PRSP objectives to less than 55 percent interms o f the objectives defined inthe Priority Action Plan o f the PRSP. 9. To improve the extent to which budgetary instruments are matched with the strategic objectives o f the PRSP, the study proposes two complementary approaches. The first stresses the missing link in the Senegalese planning system, namely the Three-Year Public Investment Program (PTIP). It i s essential to improve the forecasting capacity o f the PTIP, which appears to be quite weak-a forecasting error o f 40 percent from 2004 to 2005 has been noted. Such an improvement presupposes capacity buildinginthe unit responsible for this forecasting, as well as improved cooperation between the Ministry o f Economy, the sectoral ministries, and donors. In the interest o f efficiency, it i s fbrther recommended that there be a focus on the health and administrative equipment sectors, which account for the lion's share o f this forecasting gap. The PTIP should also be used to greater advantage to optimize the regional distribution o f investment, which is an integral part o f the decentralization strategy pursuedby the Senegalese authorities. From this perspective, it i s proposed that information be published in order to enhance transparency, and that evaluation system be introduced which takes account of the needs and capacities o f each region. A suitable model for the education sector, based on the example o f Brazil, i s elaborated on inthe report. 10. The second approach i s aimed at systematizing the link between the Priority Action Plan (PAP) and the approved and implemented BCI, using a monitoring system described in detail in the study. This system has already been shared with the Senegalese authorities, who have begun to make use o f it in their PRSP assessment. It should make it possible to improve the budgetary monitoring o f Senegal's development strategy. Third challenge: Ensuring theproper execution and substance of the investmentprogram 11. Identifying the strategic choices and properly aligning them with budgetary instruments are both preconditions. It is then essential to ensure that execution o f the investment program i s in keeping with the objectives set in the budget. The fact is, however, that there are enormous variations in the implementation ratios o f investment programs in Senegal. This variation has been a phenomenon occurring over time-increasing from 61 percent in 2001 to 67 percent in 2003, but with a peak o f 82 percent in 2002-and across sectors, negatively affecting expectations as regards proper investment project implementation over time. 12. A number o f proposals are made with a view to gaining a better understanding o f the implementation rate variations observed in Senegal. First, there i s a need to improve the availability o f information, which i s not necessarily reliable and tends to become available belatedly. This situation is an obstacle to proper monitoring o f the system and to the rapid identification o fbottlenecks. Inthis regard, we support the initiative o f the Senegalese authorities aimed at introducing a monitoring system that will serve to track disbursement trends in real time. 13. Second, the authorities are strongly advised to devote particular attention to the ongoing projects in the infrastructure and social services sectors, which report implementation rates that are routinely lower than the average observed in Senegal. These lags may be explained inpart by the size o f the projects in these sectors-which require more complex procedures-but also by i x shortcomings noted inthe technical ministries concerned. A training and information effort must be directed at the financial specialists in these ministries, and coordination with the Ministry o f Economy and Finance mustbe improved. 14. Third, disbursements for externally financed projects appear to be slower than for domestically financed projects. This observation fails to reflect the fact that there are variations between donors, inasmuch as projects with grant financing generally have higher implementation rates than those financed by loans or subsidies. It would therefore appear that there are significant differences between donors, and these should be identified more clearly. More generally, this finding illustrates the urgent need to improve the management o f investment budget execution, which continues to show disparities depending on the financing source. 15. Finally, the extent to which the executing agency i s centralized appears to have an influence that should not be disregarded. Although the evidence remains partial, close attention to the World Bank portfolio does show that: (i) projects managed by decentralized agencies (PAC, PNIR) have relatively high implementation rates; and (ii) the components managed by units decentralized to the regional and community levels have more rapid disbursement rates than those under the auspices o f the Central Government. It bears noting, however, that these results have been obtained in part thanks to the creation o f implementing agencies that obviate the need to follow established procedures at the local government level. Moreover, although some local governments have demonstrated the technical and financial capacity to manage relatively smaller projects, the very size o f the B C Ijustifies the gradualist and prudent approach that the Senegalese authorities have taken to decentralizing the management o f this budget. Fourth challenge: Establishing an efficient and transparent management system 16. The Senegalese Government must continue its efforts to improve the public expenditure management system, especially as regards investment expenditure. Despiterecent progress, there are several weaknesses that needto be corrected as regards preparation and execution (including procurement procedures) and as regards internal and external controls. Our major recommendations may be summarized as follows: 0 Preparation o f the B C I needs to be improved by: (i) buildingthe capacities o f the DCEF and the technical ministries for investment project and program identification and preparation; (ii)improving the forecasting capabilities o f the PTP, establishing a systematic linkage with the PRSP objectives and the regional distribution o f investment; and (iii)improving the programming o f the recurrent costs associated with investment projects, inparticular by publishingthe B C Ibased on an economic classification, making it possible to draw appropriate distinctions between capital expenditure and current expenditure. 0 Improving investment budget execution involves a number o f steps, including: simplifying and fully computerizing the expenditure circuit, and clearly delineating the responsibilities o f all participants in that circuit, in particular between the DDI and the Treasury as regards externally financed expenditure; (ii)annually preparing a report which describes and assesses the investment expenditure financed against domestic resources, and annexing it to the draft final budget accounting law; (iii) increasing the X transparency o f investment project execution by the autonomous agencies and specialized funds, whose role has expanded rapidly inrecent years; (iv) improving and implementing the new accounting and financial framework for investments by incorporating a mechanism for the deconcentration o f such expenditure; (v) at the WAEMU level, initiating the reform o f the framework for executing externally financed expenditure; and (vi) streamlining the procurement procedures as required by the revision o f the Code (in progress), limiting the use o f waiver procedures, and improving the oversight o f the decentralized implementing agencies and their practices. 0 As regards internal and external controls, it is proposed that a simplified internal control mechanism for the commitment o f such expenditure be introduced, accompanied by the following measures: (i) improved budgeting by expenditure line; (ii) connection to the SIGFIP computer system; (iii)increased independence o f the accounting officers for projects, via reports to project leaders; and (iv) initiation o f discussions with the BCEAO on opening special accounts on its books, or the inventory o f bank accounts with the commercial banks and the streamlining thereof. These measures should be supplemented by the effective application o f a reliable and independent external control system for investmentprojects-tantamount to strengthening the roles o f the Audit Office and o f the external auditors. x i 1. ENHANCINGTHE EFFICIENCY OF PUBLIC INVESTMENT PUBLIC EXPENDITURE REVIEW A. INTRODUCTION 1.1 In recent years, Senegal's macroeconomic performance has been encouraging, with a growth rate exceeding 5 percent on average since 1994 and improvement in its major fiscal and external balances. Thanks to this performance, the proportion o f households living in extreme poverty declined by 10 percentage points over the last decade. This performance i s encouraging, to be sure, but more must be done ifthe objectives the Government set for itself in its medium- term development strategy are to be met, namely reducing by half the population living in poverty and eliminating the gap between the average standard o f living in Senegal and that observed inthe more developed countries o f the globe. 1.2 With these objectives in view, the Government has launched a strategy with the acknowledged objective o f achieving, or even exceeding, annual growth o f 8 percent, while adopting an appropriate redistribution policy which makes it possible to improve the living conditions o fthe vulnerable groups that have reaped no benefits from growth. Properly following the example o f countries like Chile or Mauritius, the Senegalese authorities have endeavored to accord priority to development o f the private sector, which has long been overlooked inthis part o f the world. At the same time, they have stressed the role that should be played by the State, in particular in delivering basic social services, such as education and health, and in providing the infrastructure services which are still sorely lacking in a country like Senegal. As summarized in Annex A, the Govemment has introduced a number o f reforms aimed at improving its performance in the education and health sectors and achieving the ambitious objectives defined both in its poverty reduction strategy and in the Millennium Development Goals (MDGs). Alongside these efforts, a reform o f the Central Government's budgetary and financial procedures has been initiated with a view to enhancing the transparency o f public expenditure. Work is currently in progress in a number o f areas with the technical and financial support o f development partners, including the World Bank. 1.3 However, notwithstanding these efforts, few studies have been focused on the role of public investment in the growth strategy for the Senegalese economy. This role ought not to be overlooked, as public investment currently accounts for almost half the Central Government budget and i s almost as sizable as private investment. It is thus o f the utmost importance to understand the reasons behind this growth and to determine whether it will allow for boosting the growth rate to the level to which the Senegalese authorities aspire. 1.4 The primary aim o f this study i s to assist the Senegalese authorities with determining whether public investment expenditure can serve as a chief instrument o f economic policy in Senegal, inparticular by addressing the five questions set forth below: 0 Can the Senegalese Government give consideration to expanding public investment without jeopardizing its fiscal and external balances inthe short and medium terms? 1 0 To what extent are the Senegalese authorities able to determine the optimal composition o f investment, to judiciously program the recurrent expenditure associated with investment projects, and to incorporate possible complementarities with the private sector? 0 Do the budgetary instruments used by the Senegalese Government for programming its investment expenditure incorporate the strategic priorities defined in its development strategy? 0 A r e the authorities in a position to execute the investment projects called for under the investment budget, and what are the principal factors that have an impact on implementationrates? 0 How can the public investment management system be optimized both as regards preparation and execution (including procurement procedures) and as regards intemal and external controls? 1.5 This study i s part o f the overall assessment o f the efficiency o f public expenditure initiated by the Senegalese authorities, and should be regarded as a first stage that i s focused on investmentexpenditure. O fcourse, the authorities will have to incorporatethe other categories o f expenditure into the analysis, as there i s no question that the efficiency o f public expenditure depends upon the productivity o f the civil service and not just on the return on physical capital. Inlike manner, the investmentpolicy itself cannot bejudged separately from other expenditure: i s there a need, for example, to raise teacher salaries and/or to build more schools? These are important issues that merit beingtreated on a prioritybasis inthe near hture because, inthe final analysis, it i s such questions that will determine the success o f the development strategy being pursuedby Senegal. 2 2. THE ROLEOF PUBLICINVESTMENTINSENEGAL: TRENDS AND OUTLOOK 2.1 Since independence, public investment has played an important role in the Senegalese economy, almost matching the levels o f private investment and increasing over time. In this context, two questions are explored in this chapter: (i) has public investment in Senegal how evolved over time as compared to the investment pattems in other countries? and (ii) what i s the outlook for the future as regards the role o f public investment as a major economic policy instrumentinSenegal? 2.2 This chapter opens with an overview o f the major statistical sources used in Senegal for purposes o f measuring public investment, which will place in the proper context the efforts recently initiated by the authorities to revise the current series and correct the existing limitations. We will then propose a temporal description and an intemational comparison o f the public investment trend over the past two decades. Finally, we will conclude by showing that Senegal has a historic opportunity to use public investment as an instrumento f economic growth, this because o f the "good health" o f public finances and the relatively low debt levels. However, there is also an evident risk owing to the significant shortcomings o f the public finance management system, which could lead to suboptimal utilization o fpublic resources in Senegal. A. DEFINITIONAND STATISTICAL SOURCES 2.3 The main statistical source traditionally used in Senegal for measuring public investment i s the Table of Govemment Finance Operations (TOFE), which defines the public investment variable as total capital expenditure (financial and physical) by the public sector (see Chart 2-1 and Table 1 o f the Statistical Annex). Although this series i s still usedby the Government and by intemational financial institutions, it has a number o f shortcomings. For example, it covers only the investment expenditure o fthe Central Govemment, and consequently overlooks the spending o f the local governments, social security administrations (Social Security Fund, Retirement Pension Institution o f Senegal), the miscellaneous central government agencies (ODACs), and the public enterprises. In addition, the definition used i s not compatible with the intemational definition, which considers investment solely as a physical flow, because it includes a portion o f the operating expenditure associated with the investments made, sowing confusion as to the economic nature o fpublic expenditure inSenegal.' 2.4 For these reasons, the authorities have begun to revise the public investment series, endeavoring to incorporate the other components o f the State and bring the series into line with the `intemational definition. This revised series covers the period 1980-2003, but i s still somewhat approximate, and hence has yet to be officially approved by the authorities. Indeed, it i s calculated on the basis o f the physical capital content o f investment expenditure o f the central administration for 1999 alone, which was then used as a reference point for the other years. 'See Table 1o f Annex A for a description o f the TOFE for the 2000-2005 period. 3 Indeed, the structure of local government investment by unit (chamber o f commerce, trade councils, municipalities, rural communities) i s available only for 1998. Despite these shortcomings, much effort has been made to improve the understanding o f public investment, and should be continued. 2.5 While the statistical problem ought not to be underestimated, its importance should be assessed in relative terms for purposes o f preparing a macroeconomic evaluation. Based on the preliminary data, even though the revised series and non-revised series are divergent, the gap between the two i s reasonable, as public investmentinthe new series i s equivalent on average to 79 percent o f public investment measured in the old series over the period 1980-2003. The fact that this gap i s not larger i s explained in part by the preponderance of the central administration inpublic investment, where it accounts for 93 percent ofthe total inthe new series (see Table 2 of the Statistical Annex).2 It i s nonetheless true that it i s urgent for the authorities to finalize their revision o f the national accounts. Figure2.1: Sustainedgrowth of public investment +Q0 +!$+e* +QO 7Q+ 7Q+ +*77Q*v +* +*>+Q0 +Q$ +QQ 7Q0J %+ -Public Investment -Private investment1 Source: National accounts. B. A LOOK BACKWARDS: A TEMPORAL AND INTERNATIONAL COMPARISON OF PUBLIC INVESTMENT 2.6 Public investment has played a preponderant role in the Senegalese economy in recent decades, as it has in the majority o f African countries (see Chart 2.2 and Tables 3 and 4 o f the Statistical Annex). While it remained relatively stable until the late 1970s, public investment took off spectacularly in the early 1980s and especially in 1998, reaching the peak o f almost 10percent o f GDP in2003. This increase has continued, as the ConsolidatedInvestment Budget The major missing component continues to be the absence of investmentdata frompublic enterprises. 4 (BCI) has risen by 61 percent between2000 and 2005, an increase o f CFAF 225 billion. This i s explained at a sectoral level by the increased focus on surface transportation (contributing 39 percent to the growth o f the B C I over this period), followed by administrative equipment (23.4 percent) and health (13.5 percent) (see Chart 2.3). It bears noting that the increase observed for road transport i s heavily concentrated in 2004 and 2005 (when such investment grew by a factor o f two), whereas the increases in administrative equipment and health have been more gradual over time. Figure 2.2: Public investment ratio (2001-2002): Senegal is close to the average for Africa - Mozambque hgena I zamo a I Guinea-8ssad i Botswana Cape Verde Ethiopia Ghana Chad Mauntania Namibia Congo Rep of Malawi Mali Burklna FaSO Senegal Central Guinea Afnque Benin Niger Gambia The Rwanda Uganda Madagascar Gabon surundi Tanzania Sudan Kenya Togo Camemon 0 0 2 0 4 0 BO 8 0 I O 0 120 14 0 Public Investmentratio Source: World Bank. 2.7 In addition, it must be stressed that public investment accounts for a more sizable proportion o f the Central Government budget in Senegal than it does in most African countries. For each franc spent under the Senegalese budget in 2002, about 38 centimes was associated with investment expenditure, as compared to less than 20 centimes for Africa on average. The proportion o f public investment expenditure in Senegal's Central Government budget i s now close to the proportions observed in countries like Mali, Burkina Faso, or Uganda. In fact, Senegal i s the African country which, after Zambia, has made the greatest reallocation o f its budget in favor o f investment expenditure between the 1985-94 period and the 1995-2002 period. 5 Figure 2.3: Major sources of BCI growth, 2000-2005 Roadtransport Administrative equipment Healthandnutrition Housingandtown planning Railtransport Educationand training 0 0.1 0.2 0.3 0.4 0.5 % of increase in total BCI Source: BCI. 2.8 The trend o f public investment in the African economies, including Senegal, differs significantly from the one observed in Latin America or in the OECD countries, which has tended downward over the past decade. 2.9 This divergence may be explained by at least four factors in the case o f Senegal. First is surely the impact o f the heritage o f the French colonial period, during which the Central Government played an important role in the development o f infrastructure as well as other activities regarded as strategic. The second explanation o f the magnitude o f public investment i s the weakness of private investment in Senegal (11.9 percent o f GDP over the 1995-2002 period), which i s below the average for Africa (13.5 percent), which itself i s low in comparison with other regions o f Africa (16.1 percent inNorth Africa, for example). Of course, there i s causality working bothways, but it i s virtually certain that the Central Government has had to andwill still have to stand in for the private sector in certain strategic sectors, in particular infrastructure. Third, Senegal has not experienced fiscal pressures such as those observed in other countries, which might have compelled it to reduce its investment, which tended to be the first thing to go under the adjustment measures adopted in other countries. Senegal reported a public deficit o f 0.6 percent o f GDP on average for the 1995-2002 period, as compared to over 2.8 percent inthe rest o f Sub-Saharan Africa. Finally, a sizable proportion o f public investment i s financed by extemal resources, which have been sufficiently stable to ensure the continuation of public investment programs in Senegal. 6 c. PUBLICINVESTMENT:A N OPPORTUNITY FORTHE FUTURE? 2.10 The question that policy makers must address in Senegal i s whether public investment can continue or increase the fundamental role it plays in Senegal's development strategy. The answer to this question calls for analysis that must be carried out at multiple levels. O f necessity, it requires determining whether an increase inpublic investment is possible without jeopardizing the fiscal balances o f the Central Government. Only within the context o f adhering to the macroeconomic framework can the questions about the allocative and operational efficiency o f public investment merit being raised. To do so, we present below a simple public expenditure sustainability analysis which examines, in tum, whether an increase in public investment can occur while preserving: (i) the sustainability o f the public debt; (ii) control of inflation; and the (iii) maintenanceofequilibriuminthecurrentaccountbalance. Thereisafourthfactorthat the i s generally used to assess the relevance o f a fiscal policy-the crowding out o f the private sector. This fourth component will be examined inthe next chapter. C.1 Debt sustainability 2.11 When the ratio o f public debt to GDP increases too rapidly, the government's indebtedness position becomes unsustainable, which may prevent it from repaying its future debt. Public sector debt becomes unsustainable if: (i)it i s projected to grow indefinitely over time, beyond an acceptable limit; and/or (ii)the cost associated with repaying this debt (amortization and interest payments) absorbs an ever larger amount o f public resources. In this context, a country which reports a high level o f indebtedness has less o f an opportunity to increase its fiscal deficit and conduct an ambitious investmentexpenditure policy. 2.12 This case does not correspond to the situation o f Senegal, which has external and domestic debt levels estimated at about 28 percent o f GDP (in net present value terms) and 7 percent o f GDP, respectively, in 2004. In addition, it should be stressed, the majority o f this debt i s on concessional terms, which eases the debt service burden. The relatively low debt level o f the Senegalese public sector i s the result not only o f the extemal debt reduction which occurred in April 2004, but also o f the long period o f fiscal discipline. Indeed, Senegal now belongs to the small group o f African companies which report sound fiscal performance, along with Equatorial Guinea, Nigeria, Mauritania, and Gabon. If the oil producing countries are excluded, Senegal may be regarded as the country with the best fiscal performance inthe region inrecent years. 2.13 Based on the macroeconomic projections agreed on by the IMF and the Senegalese authorities (see Charts 2.4 and 2.5), Senegal's indebtedness i s expected to remain acceptable in the medium term. The net present value o f the debt as a proportion o f GDP is projected to decline from 31.4 percent in 2004 to 13.8 percent by 2024, while the net present value o f the debt as a proportion o f exports drops from 113.6 percent in2004 to 73.5 percent in2024. Itbears noting that these projections, which are slightly less favorable than those used when the completion point under the HIPC Initiative was reached, are based on the following macroeconomic assumptions for the 2005-2024 period: (i)average GDP growth rate o f 5.2 percent, about half a percentage point lower than the rate envisaged in the HIPC baseline scenario; (ii)an increase inexport volume o f 4.9 percent, as compared to 5.6 percent inthe HIPC baseline scenario; and (iii) a doubling o f the fiscal deficit (2.8 percent as compared to 1.5 percent 7 in the HIPC baseline scenario). Senegal's external indebtedness level would stay within acceptable limits even in the event o f relatively sizable exogenous shocks affecting the terms o f trade, GDP growth, and/or export growth (such as a climatic shock, for example). The latter scenario i s illustrated in the charts below, inwhich one can observe only a slight deterioration in the ratios inthe short term. The only shock that would have a lasting effect on debt sustainability would be an increase inthe cost o f external financing. For example, a 3 percentage point increase inthe nominal interest rate on external loans-or a 20 percent decrease inthe grant element- would result ina deterioration inthe debt/export ratio and the debt service/exports ratio, butboth would nevertheless remain below the ratios defined by the IMF as being sustainable for low- income countries. Figure 2.4: Debt/GDP ratio (net presentvalue) 60 55 8 5 0 9$ - 4 5 - 0 L, - zf 4 0 9 3 5 - ----____ c $ 3 0 Ei - Historic scenario (withno debt reduction) - 2 5 25 20 - 15 - 2004 2006 2008 2010 2012 2014 2016 2018 2020 2022 2024 20 - Historic scenario (with no debt reduction) -, x - . e -- 15 P I C I/ / '28 V 10 Scenario with export shock 5 Baseline scenario (with debt reduction) \ 2004 2006 2008 2010 2012 2014 2016 2018 2020 2022 2024 8 2.14 More revealing still i s the fact that the public sector indebtedness level would remain within sustainable limits, even assuming an increase in the public deficit, including through an increase ininvestment. Thus, a permanent 3 percent increase inpublic expenditure (as compared to the baseline scenario) shows that if the public deficit were to reach its highest level o f 5.8 percent o f GDP in 2015 (and then drop gradually to 5 percent by 2024), the debt/GDP ratio would stabilize at about 32.6 percent in2024 even though its path over time would not o f course be as favorable as inthe baseline scenario. C.2 Preserving price stability 2.15 Inaddition to the sustainability of the external debt, a second reason is often put forward to justify a prudent fiscal policy (including investment policy), name1 ,the inflationary pressures on prices that monetary financing o f the public deficit could exert. Fortunately, this situation Y would hardly appear to be a concern in Senegal, because the authorities have only marginally had recourse to monetary financing to cover the public deficit-which, moreover, has stayed within reasonable limits (see Chart 2.6). Since 2000, the main source o f financing has beenfrom abroad (grants and loans), while domestic bank credit has played only a marginal role. Finally, Senegal's membership in the regional monetary union limits and will continue to limit the Government's room for maneuver as regardsthe monetary financing o f its budget. Figure2.6: Compositionoffinancing sourcesfor the fiscal deficit,2000-2005 7 12 6 10 8 6 4 1 - 2 0 ~ 0 - ~ -1 ~ - -2 - -2 -3 -4 Grants Domestic Financing I External Financing +Overall balance (excluding grants) Of course, inflation is also determined by factors other than the monetary financing of the public deficit, in particular exogenous shocks such as surges in oil prices or a sharp depreciation of the CFA franc, which would contribute to boosting import prices. 9 C.3 Fiscal policy and externalequilibrium 2.16 The last major reason which might justify a cautious fiscal policy in Senegal refers to maintaining extemal equilibrium, namely, its current account balance. When a country spends more than the value o f its output, for example by increasing its public investment expenditure, the deficit in the current account balance i s increased, which may be difficult to overcome with extemal financing and hence engender the risk o f a disruption inthe future. 2.17 At present, our analysis suggests, Senegal is not confronted by exaggerated pressure on its external accounts. The recent deterioration in its balance of payments i s explained above all by the decline inthe terms of trade and the lack o fbuoyancy inthe export sector (see Chart 2.7)' although the increase in the public deficit did play a growing role in the increase in the current account deficit beginning in 2003. While the Senegalese Government appears to have appreciable room for maneuver, the fragility o f the private sector (inparticular the export sector) i s such that an exaggerated increase in the budget deficit could have negative implications on Senegal's extemal equilibrium ifnot offset by an increase inexternal financing. Figure 2.7: Relationship between public sector and external sector, 1994-2004 I O 0 0 -I0 Es-2 0 t5-30 k -4 0 -5 0 -6 0 -7 0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004Proj 1--CSoldeduSecteurPublique --CSoldeduSecteurPrive Balance en compte courant1 Note: The equilibrium o f the public sector is measured by the budget deficit, while that o f the private sector is measured as the residual between saving andprivate investment. 10 D. CONCLUSION 2.18 Senegal i s at a crossroads as regards the role o f public investment. Indeed, it enjoys a historical opportunity to increase this role owing to the sound public management o f the Central Government's financial equilibriums in recent years, strengthened by the recent external debt relief. Such an increase should be within reasonable limits in order to avoid exaggerated pressures on the current account balance, which could deteriorate in the event o f an overly ambitious fiscal policy. This opportunity has not gone unnoticed, and the PRSP places special emphasis on the contribution o fpublic investment. 2.19 The initial assessment o f the role o f public investment is difficult, however, owing to statistical deficiencies. It i s strongly recommended that the Senegalese authorities continue their effort to revise the existing series on public investment, which does not meet international standards. Continuation o f this effort i s valuable not only in a statistical context but also, as the next chapter will show, to gain a better understanding o f the relationship between capital expenditure and maintedance expenditure over time. For the same reason it i s urgent to broaden the coverage o f public investment and public debt to include the public enterprises, which at present are not part o f the official statistics. Giventhe role that these enterprises continue to play inthe Senegaleseeconomy, this extension is essential. 11 3. PUBLICINVESTMENTAND ECONOMIC GROWTH 3.1 The Senegalese Govemment has the unique opportunity o f using public investment expenditure to achieve its economic policy objectives. However, to make the most o f this opportunity it i s essential that the authorities be assured that public investment i s contributing effectively to the objectives defined in their development strategy, in particular in terms o f economic growth. A brief review will demonstrate that the efficiency o f public investment expenditure in Senegal i s far from guaranteed, thus proving that it i s not enough just to increase public investmentinorder to automatically trigger a higheconomic growth rate. 3.2 While there are several reasons which might explain the low productivity o f public investment, intemational experience indicates that the efficiency o f investment policy i s closely associated with: (i) the composition o f investment; (ii) assumption over time o f the recurrent the costs associated with investmentprojects; and (iii) the interaction betweenpublic investment and private investment. 3.3 The Senegalese authorities have incomplete knowledge about the composition o f public investment and, in particular, conduct no systematic analysis o f the linkages between capital investment and recurrent costs, because o f the deficiencies in the planning and ongoing management o f public expenditure. This latter shortcoming i s serious at a time when infrastructure investment projects are multiplying and the central government i s considering the decentralization o f investment budget execution to the local government level. Finally, if increased private sector participation in public investment projects i s one o f the Government's declared objectives, special attention must be paid to the risks engendered for the public sector, inparticular as regards the sustainability o fits fiscal position. A. PUBLIC INVESTMENT AND ECONOMIC GROWTH: NOT A CERTAIN RELATIONSHIP 3.4 Intemational experience reveals that several objectives are generally associated with an increase inpublic investment. First, the authorities are seeking to boost the economic growth rate through an expansionary fiscal policy and through its impact on the amount and quality o f the stock o f physical and human capital. There are a number o f other concerns, such as the desire to reduce inequalities, to comply with law and the enforcement authorities, and to protect the most vulnerable groups. In Senegal, the authorities have also expressed their interest in ensuring balanced and diversified growth interms o f outputs and regional distribution. 3.5 Despite the profusion o f objectives, the success o f an investment policy i s oftenjudged in terms o f its capacity to increase the economic growth rate. While this criterion i s manifestly restrictive, it i s difficult to argue that the investment policy will be optimal in the absence o f a significant increase in the economic growth rate. Furthermore, the latest household survey (ESAM 11) conducted in Senegal in 2001 concluded that 1 point in economic growth leads on 12 average to a decrease o f 0.9 point in the poverty index.4 It i s in this context that we devote special attention below to the possible relationship between public investment and economic growth. 3.6 Economic literature has failed to provide clear-cut conclusions as to the statistical relationship between public investment and economic growth. There are a number o f reasons for this. First, it i s difficult to control all the factors other than public investment that also affect economic growth in the short and long terms. Second, an appreciable share o f public investment i s justified by objectives other than economic growth, such as the redistribution and provision o f public services, and the maintenance o f public order and the rule o f law, which have no direct linkage to increasing production. Finally, the aggregated nature o f public investment expenditure causes its impact to be spread over time and makes it difficult to identify statistically. 3.7 Despite these limits, we have examined the contribution o f public investment to economic growth by using a simple macroeconomic model (described in Annex B) which emphasizes: (i) the linkage between growth and physical capital, distinguishingbetween private investment and public investment; and (ii)the existence o f a possible relationship between private investment and public investment, through the availability o f financing on domestic markets. Estimation o f this model for the 1966-2000 period enabled us to obtain the results presented inTable 3.1,which all in all are satisfactory and robust to changes in specifications. It i s important to clarify that the results presented are indicative and not necessarily conclusive owing not only to the technical limits cited previously, but also to problems with the statistical series usedinthe regression^.^ 3.8 The most pertinent finding for purposes o f our analysis i s that public investment does not appear to exert a significant impact on the rate o f economic growth, in any event during the current year, during the period under review. On the other hand, there i s a positive correlation betweenprivate investment and economic growth inSenegal. The lack o f productivity associated with public investment in Senegal is not really surprising in light o f the results found in the economic literature (see Khan and Kumar (1997) or Devarajan, Easterly, and Pack (2000)).6 This result should be interpretedwith caution, as it does not necessarily mean that public investment has no positive effect on economic growth in Senegal. It does, however, make it more difficult to argue that such an effect i s significant, and above all that it i s superior to that associated with private investment. 3.9 The marginal impact o f public investmenton economic growth i s confirmed, even if one endeavors to include its possible effect lagged over time. The introduction o f lagging structures Poverty Reduction Strategy Paper Progress Report, March 2004. The regression presented i s intended to identify the possible existence o f a statistical linkage between public investment and economic growth in Senegal. It is not aimed at explaining economic growth per se, which o f course i s influenced by many other factors on the demand side as well as that o f aggregate supply. The explanation o f growth sources in Senegal i s a subject which is beyond the analytical scope o f this report, but i s an issue that the Senegalese authorities should pursue. For a recent study on this subject, see for example Gaye Daffe, "Profil de la croissance au Shegal" [Senegal's growth profile], January 2005. 6 M. Khan and M. Kumar, "Public and Private Investment and the Growth Process in Developing Countries," Oxford Bulletin of Economics and Statistics 59 (l), William R. Easterly, Shantayanan Devarajan, and Howard 1997; Pack, "Is Investment in Africa Too L o w or Too High? Macro and Micro Evidence," World Bank Working Paper No. 2519, January2001. 13 did not enable us to identify a significant impact after three years, even an unequal one over time. The statistical results appear to indicate that the effect of public investment i s lagged over time and that its positive impact appears to show through after two years. It is important to note, however, that this positive effect follows a negative impact during the first year, suggesting that inthe final analysis, the total impact ofpublic investment on economic growth is neutral after a three-year period. 3.10 The lack o f productivity associated with public investment seems hardly to have changed over the past 20 years. To test this hypothesis, we carried out two series of statistical tests whereby we endeavored to determine whether the productivity o f public investment vaned: (i) in accordance with changes in the incumbent Prime Minister; and (ii) following the devaluation o f December 1994. The results o f these two tests did not identify any profound changes in the productivity o f public investment, which i s generally unproductive regardless o f the period or the institutional framework considered.' Table 3.1: Economicgrowth andinvestment, 1966-2000 (t-statisticsbetweenparentheses) , A Y A Y A Y Constant -0.11 -.174 -0.120 (-2.24)* (-3.05)* (-2.22)* Ig .21 .597 (.23) (0.64) [ISI-I -2.427 (-1.98) 2.479 (1.58) IP .92 1.322 .910 (2.14)* (2.93)* (2.09) Ab .199 .331 .195 (1-72) (2.41)* (1.63) AY-1 -.327 (-2.02) R2 0.20 0.35 0.29 F Statistic 2.61 3.69 2.91 Notes: Variable A Y is defined as the rate o f growth o f real gross domestic product (GDP), I,private investment, I,public investment, Y gross domestic product, and Ab the ratio of private bank financing to GDP. The principal statistical sources used are the national accounts provided by the Senegalese Government and the World Bank's databases. The (*) sign means that the coefficients are statistically different from 0 in a confidence interval greater than 95 percent. For the first test we used a dummy variable associated with each Prime Minister, while for the second we used a dummy variable which takes the value o f unity after the devaluation. These dummy variables appear not to have any significant impact on the elasticity o f public investment, regardless o f whether they are introduced into the regressions additively or multiplicatively. It should be noted, however, that a trend change may have occurred in recent years, in that the capital expenditure o f the Central Government increasedby more than 16 percent a year on average over the 2000-2004 period, but the composition o f this investment as presented in Table 3.2 fails to presage any relationship whatever between growth and public investment, at least inthe short term. 14 B. COMPOSITIONOF INVESTMENT:I S IT OPTIMAL? 3.11 The relatively scant contribution o f public investment to economic growth in Senegal could be because its composition is suboptimal. This argument triggered a wave o f empirical studies in the economic literature, especially in the 1990s and also in the developing countries, which demonstrated that the linkage between public investment and economic growth was more firmly established when only the investment expenditure inthe productive sectors was taken into account. In particular, a number o f authors favored investment on infrastructure which can not only contribute to the accumulation o f physical capital but also have a positive impact on the activities o f the private sector, thereby engendering a twofold positive effect on economic growth (for example, see Ashauer (1989), S. Devarajan et al. (1996), or more recently, Milboume, Otto, and Voss, 2003).' 3.12 To what extent can this argument on the composition o f public investment be applied to Senegal? This i s by no means easy to answer owing to statistical problems and the lack o f an evaluation and monitoring systemg For this reason, we will confine ourselves below to establishing the frame o f reference for a more complete analysis that the authorities should pursue. 3.13 As a starting point, we broke down investment expenditure into spending that should have a short-term impact on economic growth and spending that would only have an impact in the long term. This breakdown is, to be sure, arbitrary, as it is difficult to avoid ambiguity when classifying the components o f public investment by their potential effect on growth, but we used the classification proposed by Clemens, Radelet, andBhavnani (2004)." Investment expected to have a short-term impact on the economic growth rate is defined as investment inthe following sectors: Infrastructure (transport and communications), energy, banks, agriculture, and industry. Investmentexpected to have an impact inthe medium and longterms is concentrated inthe other sectors, such as education, health, water supply, and defense. * D. Ashauer, "Is Government Spending Productive," Journal of Monetary Economics 23, 1989; S. Devarajan et al. (1996), V. Swaroop, and H. Zou, "The Composition o f Public Expenditure and Economic Growth," Journal of Monetary Economics 37, 1996; Milbourne Ross, G. Otto, and G. Voss, "Public Investment and Economic Growth," Applied Economics 35, March 2003. It bears noting that this result has not always been verified empirically, depending on the technique and the sample used. 9 One possible approach would have been to examine the rates o f return associated with each project and then to compare them across sectors. This approach i s not possible in Senegal owing to the lack o f systematic project evaluation and monitoring (except perhaps for donor-financed projects). One alternative would have been to gather sufficiently long time series at the sector level to statistically test the relationship between the composition o f public investment and economic growth inSenegal. lo "Counting Chickens when they Hatch: The short-term effect o f aid and growth," Center for Global Development, Working Paper No. 44, July 2004. 15 Table 3.2: Sectoral allocation of public investment, 2000-03 Percentage of total investment executed Short-termimpact on Medium-termimpact on Sector economic growth economic growth PRIMARY Agriculture 8.4% Livestock 0.8% Water and forests 2.6% Fisheries 2.8% Rural and agricultural water supply 1.9% Studies and researcldinstitutionbuilding 5.2% SECONDARY Mining 0.3% Industry 0.5% Crafts 0.2% Conventional and renewable energy 5.9% Industrial studies and researcldinstitutional support 2.3% TERTIARY Trade 0.1% Tourism 0.2% Transport 13.4% Post and Telecommunications 0.2% Studies and researcldinstitutional support 1.2% QUATERNARY Urbanwater supply and sanitation 5.O% Culture, youth, sports 1.3% Housing and town planning 5.5% Health and nutrition 13.4% Education and traininghocial development 11.1% Administrative equipment 11.9% Studies and researcldinstitutional support 5.7% TOTAL 37.7% 62.3% Source: Government o f Senegal. 3.14 The results o f this classification for Senegal show that the investment with a short-term impact accounted for less than 40 percent of total public investment over the 2000-03 period (first column o f Table 3.2), while investment expected to have a longer-term impact exceeded 60 percent o f the total (including almost a third in education, health, and the construction o f administrative buildings). Under these conditions, it i s of course not surprisingto find that the " relationship betweenpublic investment and economic growth was weak, inany event inthe short term in Senegal, and that the authorities are taking stock o f the optimal composition o f public investment. It bears noting that the above breakdown is intended to prompt the authorities' assessment of the optimal composition o f their investment budget, but does not necessarily advocate a realignment in favor o f the sectors that have an impact on economic growth in the "Foradetaileddescription oftheConsolidatedInvestmentBudgets approvedandimplementedover the 2000-05 period, see Tables 5 and 6 inthe Statistical Annex. 16 short term. This assessment should be conducted in a more general framework, which should include the linkages between the investment budget and the current budget and the operational efficiency o f each sector. It would be inappropriate, for example, to promote the allocation o f expenditure toward a sector where the projects have low social and economic rates o f return owing to shortcomings inthe management, evaluation, andmonitoring procedures. 3.15 Below we propose an examination o f expenditure trends in three specific sectors o f Senegal's Consolidated Investment Budget, namely infrastructure, the social sector, and the administrative equipment sector. We focus on these three sectors because they account for almost all the expansion o f the B C I in the past three years, but this does not mean that the other sectors should be disregarded. B.1 Shouldinfrastructureexpenditurebe increased? 3.16 Total investment on infrastructure (water, electricity, and transport) represented more than a fourth o f the B C I implemented over the 2000-03 period, or about 2.0 percent o f GDP, which appears to fall short o f international standards given that the 1 Figure 3.1: Infrastructure stocks are low in Senegal World Bank advocates investment 1 rates inexcess o f 2.0 percent o f GDP, O X C I only for the transport sector. OW04 Although an international comparison C N C 3 i s difficult because o f the lack o f oow3 statistical data, a comparison with the Ow02 countries o f Latin America indicates ooc02 that public spending on infrastructure oow in Senegal is at approximately the 0131 same level as in Brazil and Chile c o w (1.0 percent and 1.7 percent o f GDP, 1 2 respectively, during the 1996-2001 Eknrlcty Paved mads (megawattsperthousad inhabtants) (km per thourand mhabdants) period), although those countries IEIWorld .Sub-Saharan Africa OSenegal I 3.17 Not only is the infrastructure investment rate low in Senegal, but the lack o f infrastructure there i s flagrant. Electricity costs five times more than it does in C8te d'Ivoire. There are effectively no good paved roads available for travel to Mali. The Gambia and Mauritania can be reached only by using ferries, resulting in slow-moving waiting lines at the borders. Dakar i s not immune to infrastructure problems: the sanitation network i s obsolete and the roads are so crammed that it i s almost impossible to move in the center city during peak traffic periods. In short, there i s a genuine need for investment in the support infrastructure in Senegal (see Chart 3.1 for an international andregional comparison). l2As a result o f which the total infrastructure investment ratio inChile exceeded 5.5 percent o f GDP. l3It isdifficult to find international data o n investment ininfrastructure. For Latin America, we use data developed by Luis Serven for a World Bank Working Paper. 17 3.18 This rapid comparison suggests that Senegal should consider increasing its investment in infrastructure. To gain a better sense o f what i s really at stake, simulations can be performed which indicate that Senegal would have an annual growth rate that i s 3.0 percentage oints higherifit managed to increase its stock o finfrastructureto the level attained by Botswana (see p4 Table 3.3). Although this estimate should be interpreted with caution, it does illustrate the benefits that increased investment in infrastructure could produce for the Senegalese economy. If the benefits associated with an increase in public investment in infrastructure are potentially enormous in terms of economic growth, the effort called for i s commensurate. Indeed, the stock o f electricity infrastructure in Senegal is, comparatively speaking, smaller than that o f Botswana by a factor o f 4.2. The same deficits are apparent in Table 3.3 for the stocks o f road infrastructure and communications infrastructure. 3.19 The authorities know what i s at stake, and have sharply increased public investment in infrastructure, primarily for roads, in the B C I for 2005. The share allocated to infrastructure sectors increased to almost 40 percent of the BCI, as compared to only 25 percent during the 2000-03 period. The principal beneficiary sector is the landtransport sector, the share o f which i s boosted to 23 percent in 2005 as against only 14 percent duringthe four preceding years. This i s an increase equivalent to CFAF 70 billion, or a doubling o f the investment budget in this sector. This approach corresponds to a new aspect of the Government's strategy which consists in improving living and infrastructure conditions in the Dakar urbanized area. To this end, the Government decided to initiate a development program aimed simultaneously at modernizing the city o f Dakar and deconcentrating a number o f activities by sending them to other cities. A sizable share o f the investment budget (12.8 percent, or 1.7 percent o f GDP) i s devoted to improving urban mobility within Dakar so as to meet infrastructure needs and improving the living conditions o f the people living inthe capital city and its vicinity. l4 Cesar Calderon and Luis Semen, "The Effects of Infrastructure Development on Growth and Income Distribution," World Bank Policy ResearchPaper No. 2400, September 2004. 18 Table3.3: Infrastructureinvestmentsandimpact on growth Investment effort to be made Anticipated increase (N.B.Factorby which Senegalmustincrease its inannualrateof stock of infrastructure) growth Number of phone Numberof lines per 1,000 Megawatts per kmper km2 inhabitants 1,000 inhabitants of land area Botswana 4.2 5.2 0.2 3.0% CGte d'Ivoire 0.9 3.4 2.0 2.6% Ghana 0.4 2.4 2.0 0.7% South Africa 9.7 38.2 2.6 9.0% Zimbabwe 1.2 6.3 2.1 3.5% Mauritius 14.3 12.4 11.8 10.1% Morocco 4.0 6.1 1.7 3.9% Tunisia 6.3 8.7 2.1 6.4% Costa Rica 15.1 15.6 8.9 9.6% Peru 5.4 9.0 0.7 4.6% Thailand 4.2 7.4 4.2 6.4% Memo:(stock 1995-1999) Senegal 30.24 0.06 0.08 Note: Serven and Calderon (2004). The effort to be made is measured by estimating the number o f times that Senegal's stock o f infrastructure should be multiplied in order to achieve the level o f the corresponding country. For example, the number o f telephone lines in Senegal should be increased by a factor o f 4.2 to achieve the number o f lines in existence in Botswana. Then, the impact o n the annual growth rate in Senegal i s simulated by using the coefficients estimated by Serven and Calderon for a sample o f over 100 countries during the past two decades. For example, if Senegal had stocks o f infrastructure at the level o f those in CBte d'Ivoire, its annual growth rate could be 2.6 percent higher. 3.20 While increasing public investment in infrastructure appears overall to be justified in Senegal, it i s essential that this focus be based on a systematic evaluation o f the projects concerned, interms o f both their economic and their social rate o f retum. It i s also necessary that their implementationbe monitored and overseen by intemal and external bodies. Inother words, while an analysis o f the allocative efficiency o f public investment does suggest an increase in favor o f infrastructure, such expenditure must still pass the operational efficiency test. In the final chapter o f this report we will see that this latter condition i s met only imperfectly inSenegal because o f the current shortcomings in the system for managing public finances, and investment inparticular, inSenegal. B.2 How should investmentexpenditurein the social sectors be evaluated? 3.21 While an increase inpublic investment in favor o f infrastructure i s justified, it would not be advisable to decide to do so ifit were to come at the expense o f investment expenditure inthe social sectors. It i s important not to lose sight o f the fact that such investment can have an impact on economic growth in the medium term. Thus, an increase in public investment in the health sector can bring about improvement in the quality o f the labor force, which is less frequently ill and thus becomes more productive. The same indirect link potentially holds true for investments ineducation. Inaddition, it must be recalled that these social sector investments also respondto 19 other objectives set by the authorities, such as the reduction o f social injustice and increased protection o f the poorest population groups. 3.22 A readingo f the BCIshows that the Senegalese authorities have devoted special attention to investment in the social sectors, in particular health, where spending has increased by more than 120 percent since 2000. As a result, investment in the health and education sectors represents about 16 percent o f the B C I approved in 2005 (or 21 percent o f the B C I actually executed over the 2000-2003 period). These ratios are equivalent to those observed in Mali (14 percent o f the approved budget over the 2000-2004 period) and generally higher than those observed in the other countries o f the subregion. They also reflect the global effort made by the Senegalese authorities in favor o f the social sectors, as total such Central Government expenditure represents about 6 percent o f GDP in Senegal, equivalent to the average for Africa (5.9 percent) and the ratio observed in Uganda (5.9 percent), but exceeding that o f Mali (4.7 percent). 3.23 This rapid expansion o f investment expenditure in favor o f the social sectors is justified inview of the shortfalls observed in Senegal's health and education indicators. At the present stage, it nevertheless requires that the authorities exercise particular vigilance as regards the selection o f projects and their impact on the Senegalese people. There are a number o f different methodologies which endeavor to evaluate the impact o f public investmentsinthe social sectors, the most widely known o f which are studies o f marginal benefits (benefit incidence studies) and o f behaviors (behavioral approaches).l5The major difficulty involved in applying these two approaches, apart from their methodological limits, i s their sizable requirements in terms o f statistical information. Indeed, it i s necessary to conduct detailed periodic surveys to provide precise information about the end users o f these services. Senegal has nevertheless initiated this type o f evaluation process by using the data provided by the household survey completed in 2002. The results will make it possible to develop an initial response, but one which cannot but be dated. 3.24 Owing to these weaknesses, we recommend the adoption o f a supplementary approach which might help the authorities evaluate the efficiency o f investment inthe social sectors inthe short term. It must be acknowledged, however, that there i s no one approach with unanimous support and the efforts to address this problem area continue in the scientific community, including by researchers within the World Bank. The first option i s to use simple ratios, such as the trend inprimary and secondary school enrollment ratios as compared to public expenditure in the education sector. The same approach can be followed in the health area (using, for example, the ratio o f vaccination coverage as compared to public expenditure on health). These ratios can then be compared over time as well as in terms o f the performance inother countries. Box 3.1 below presents a more sophisticated methodology which endeavors to measure the efficiency o f such expenditure. The Senegalese authorities should devote serious attention to identifying a methodology which better serves the aim o f measuring the efficiency o f their expenditure in the social sectors. l5For an overview of these two methodologies, see, for example, Public Spending and the Poor. Theoy and Evidence (eds. V a n de Walle and N e d (1995)). 20 Box 3.1: Evaluating investment inthe social sectors The Senegaleseauthorities should make reference to the economic literature with a view to identifying a number o f intermediary variables for focusing in on the efficiency o f public expenditure inthe area o f health and education. According to Herrera and Pang (2005),16 the most relevant intermediary variables are: (i) life expectancy at birth; (ii) the DPT vaccination coverage rate; (iii) measles vaccination coverage rate; (iv) the Disability Adjusted Life the Expectancy (DALE) index; (v) primary and secondary school enrollment ratios; (vi) the number o f students having completed the primary and secondary cycles; (vii) the average number o f years spent inthe school system; and (viii) the illiteracy rate among youth. Some o f these variables, it bears noting, are already included inthe quantitative monitoring o fthe PRSP. The methodology i s then to assess the extent to which the public expenditure occurring inthe education sector and health sector have an impact on the intermediary variables proposed above. Two techniques are generally proposed: the first defines efficiency in terms o f inputs (how much public expenditure must be mobilized to achieve a given level for the intermediary variables?); the second defines efficiency in terms o f outputs (what is the level for the intermediary variables corresponding to a given amount o f public expenditure?). The empirical application o f these two techniques is predicated on determining the efficiency frontier, that is, the maximum efficiency achievable in terms o f inputs and outputs. Then, each country i s evaluated in terms o f its distance from that efficiency frontier. The main advantage o f this approach is that it makes it possible to evaluate the efficiency o f public expenditure in the social sectors with a relatively short lag, as the information needed is generally available without too much delay. The drawback i s that the results are relatively sensitive to the statistical technique used andto changes inthe sample, and that the analysis remains quite static. Table: Efficiency of public expenditure in the social sectors (1996-2002) Senegal 0.66 0.51 0.66 0.17 0.64 0.58 Africa 0.69 0.62 0.65 0.23 0.65 0.62 East Asia 0.74 0.79 0.69 0.50 0.73 0.83 I Latin 0.74 0.82 0.69 0.61 0.68 0.87 Note: The efficiency of each country is measuredinterms of its distance from the efficiency frontier. The less efficient a country is, the farther its value i s fromunity. l6Efficiency of Public Spending in Developing Countries: An Eflcient Frontier Approach, World Bank, January 2005. This study, as well as a detailed explanation o f the techniques described in this box, may be found on the Intemetsite: http://intranet.worldbank.org/WB SITE/INTRANET/SECTORS/INTECONOMICPOLICY/INTQFNO, ' menuPK:3 69265%7EpagePK: 151716%7EpiPK: 176772%7EtheSitePK:369258,OO.html 21 B.3 I s investmentin administrativeequipment always justified? 3.25 Analysis o f the composition o f the B C I demonstrates that there i s a relatively large share o f investment in sectors which in principle have no immediate impact on economic growth. Among the sectors where investment has risen the most rapidly inrecent years is "administrative equipment." After infrastructure, the largest increase observed in the amounts of public investment i s found in this category, where amounts grew by almost a factor o f four to over CFAF 70 billion (12 percent o f the BCI) in2005. This rapid expansion merits a close look by the authorities. 3.26 At a time when there should be a faster increase in infrastructure investments, assessing the justification o f the projects included inthis category o f the B C I is essential. First, the number o fprojects classified inthis category i s sizable: almost 60 in2005. Second, these projects cover a multitude o f different activities, ranging from the financing o fmilitary equipment to the purchase and construction o f buildings in Senegal (CFAF 15 billion), the purchase o f buildings in Paris (CFAF 2 billion in2005), improvements at the site o f the Organization o f the Islamic Conference (OIC) (CFAF 2 billion), construction o f a market in Touba (CFAF 2.0 billion), and purchases by diplomatic posts abroad (CFAF 2.1 billion). Finally, to the extent that all these projects are financed against domestic resources, there are no systematic evaluation systems that would make it possible, for example, to calculate the economic and social rate o f return associated with them. At this stage, and we will returnto this issue inthe final chapter, our recommendation is that the authorities more closely examine the nature o f these expenditures and reach decisions on their appropriateness, in particular in relation to the needs expressed in the other sectors o f the economy and the objectives o f the PRSP. c. CAPITAL EXPENDITUREAND RECURRENT COSTS 3.27 The inefficiency o f public investment in Senegal also stems from the confusion between capital expenditure (physical capital) and the recurrent costs associated with investmentprojects (see Box 3.2 for a definition). These outlays are not clearly identified inthe budget, which makes it impossible to properly evaluate the stock o f public physical capital in Senegal, as it would be far from accurate to suggest that an increase in public investment in Senegal i s automatically reflected in an accumulation o f physical capital. This confusion goes part o f the way toward explaining the lack o f a strong correlation betweenpublic investmentexpenditure and economic growth. On an operational level, it has serious consequences, as it contributes to the approximation involved in identifying the maintenance and upkeep requirements necessary for keepinginvestments ingood condition over time. 3.28 In order to determine the magnitude o f this problem in Senegal, we made use o f data providedby the Directorate o fForecasting and Statistics on public investment in 1999 (see Table 7 inthe StatisticalAnnex for details on the breakdown). This examination reveals that 41 percent o f the investment expenditure included in the B C I in that year went to covering recurrent costs, such as the payment o f consultant^,'^ fees, and operating costs (fuel, administrative overheads). This average masks significant variations among the sectors, projects, and financing sources l7Consultants' fees are not necessarily tied to the lifespan of the project. Nonetheless, they are not expenditure on physical capital. 22 considered. The inclusion o f recurrent costs i s a widespread practice in many sectors, covering the social sectors such as health and nutrition (65 percent) and social development (78 percent), and sectors like tourism (100 percent), energy (78 percent), and livestock fanning (89 percent). Inaddition, it is more important inthe case of donor-financedprojects, where it accounts for 43 percent as against only 32 percent for projects financed exclusively with domestic finds. Box 3.2: Measuringrecurrentcosts Recurrent costs are defined as the costs associated with investment projects that are necessary for maintaining those projects at an adequate utilization level during their lifespan. For example, for a school, the recurrent costs might include teachers' salaries, instructional materials, as well as needs for electricity and the other costs required to maintain and operate the school. They constitute a significant share o f an investment project, as according to World Bank estimates they average 7 percent o f total capital expenditure in sectors like education and health, and about 2 percent in the infrastructure sectors. In Senegal, the National Directorate o f Forecasting and Statistics uses a 6 percent ratio for purposes o f projecting public expenditure over time. Table 3.4 summarizes the rates obtained for 144 World Bank projects and44 African Development Bank projects. The share o f recurrent costs varies enormously across sectors, from 7.4 percent ineducation to 1.7 percent in urban development in the World Bank sample. The same variations are reflected in the African Development Bank projects. These variations themselves mask wide disparities within each sector, as in education, for example, where projects focused on primary education tend generally to have rates that are higher than those for secondary or tertiary education. The conclusion is that while orders o f magnitude may be deducted at a general or sectoral level, it remains essential to identify and quantify recurrent costs at the level o f each project, as these may vary enormously depending upon the project's nature and financing. 3.29 The inclusion o f recurrent costs in the investment budget i s based on an administrative rather than economic justification. It reflects the administrative requirement o f including in the investment budget all expenditure managed by the units responsible for the projects, which is especially the case for donor-financed projects. Moreover, it i s often explained by the greater flexibility possible in terms o f the practices followed under investment budgets as compared to operating budgets. The practice o f including a sizable share o f recurrent costs in the investment budgetis thus not unusual inmost developing countries, inparticular inAfrica. Among the many possible examples, mention may be made o f Uganda, where recurrent costs amounted to almost 40 percent o f the investment budget over the 1996-2002 period. In Zambia, the World Bank estimates that recurrent costs are equivalent to about 10-20percent o f the expenditure included in the investment budget. * 3.30 The first stage for the Senegalese authorities is thus to properly classify expenditure, distinguishing between capital expenditure and the recurrent costs associated with investment projects, thus presentingthe B C I in accordance with an economic classification. This exercise i s laborious, but essential. Fortunately, it has already been begun inthe effort to revise the national accounts, described in the preceding chapter. Our recommendation i s therefore to finalize this revision as soon as possible. It will also be necessary to expose the amounts o f recurrent costs includedinthe operating budget, which at present are not available globally or by sector. Once l8 Reinikka and Svenson, "How inadequate provision o f public infrastructure and services affect private investment," World Bank Working Paper, December 1999. 23 these efforts have beencompleted and the information i s available, it will thenbe easy to identify and plan for the recurrent costs associated with investment programs, in order to streamline the use o f public resources and thereby maximize their impact on economic growth over time. Inasmuch as this capacity must be part and parcel o f the overall system for public finance management in Senegal, our recommendations will be set forth inthe final chapter of this study. 3.31 Inthe meantime, inorder to assess the amount of recurrent costs interms of investment expenditure in Senegal, the economic literature suggests utilization o f the "r" coefficient, which expresses the annual amount o f recurrent costs as a proportion o f the total amount o f the investment project." This coefficient could help steer the Senegalese authorities inthe recurrent expenditure laming exercise. Using the indicative coefficients provided by a recent World Bank study?'we can thus illustrate that the cumulative total public investment for the 2000-04 period, or over CFAF 2,500 billion, should have resulted in an increase in recurrent costs o f CFAF 150-200 billion a year. This is not a negligible amount, as it i s equal to about 35 percent o f Senegal's B C I and could be even higher taking into account the increased investment that has been partially completed in the social sectors. Unfortunately, because o f the chronic lack o f information about the scope o f the recurrent costs included in the Senegalesebudget, it has been impossible to determine whether the amount referred to above i s close to the one envisaged by the authorities. Table 3.4: T h e "r" Coefficientsby sector African World Development Bank Bank Projects Projects Sector Number o f rcoefficient r Number r r projects (average) coefficient o f Coefficient coefficient (median) projects (average) (median) Agriculture 7 0.023 0.010 . 22 0.047 0.019 Education 3 0.029 0.011 17 0.074 0.032 Energy 7 0.047 0.037 14 0.013 0.002 Environment 5 0.074 0.056 12 0.017 0.014 Health 5 0.073 0.020 15 0.030 0.029 Telecommunications 4 0.043 0.027 3 0.003 0.000 Transport 6 0.019 0.010 15 0.025 0.009 Urban development 2 0.016 0.016 11 0.017 0.013 Water 5 0.054 0.063 12 0.044 0.02 1 Average (nonweighted) 44 0.042 0.028 123 0.030 0.014 Average (weighted) 44 0.043 0.028 123 0.035 0.017 Source: Hood, Husband, and Yu (2002). D. PRIVATE SECTOR PARTICIPATION 3.32 To enhance the efficiency o f public investment, the authorities know that complementarities with the private sector must be identified, first by minimizing the crowding out o f private investment by pursuing a prudent financing policy on the domestic financial See P. Heller, "Under financing o f Recurrent Development Costs," Finance & Development, March 1979. 2o See Ron Hood, David Husband, and Fei Yu, "Recurrent Expenditure Requirements o f Capital Projects," World Bank Policy ResearchPaper No. 2938, December 2002. 24 markets, and second by encouraging private sector participation inpublic investment projects, in particular ininfrastructure. D.l Crowdingout of privateinvestmentis minimal 3.33 To what extent i s private investment crowded out by public investment in Senegal? The results o f our econometric analysis reveal that this effect i s certainly marginal, as the coefficient associated with the domestic financing o f private investment i s relatively high (see Table 3.1). This result reflects the authorities' clear desire to avoid significant recourse to domestic bank credit-the stock o f (net) credit to the Government actually declined by 60 percent between 2000 and 2003, and came to only 2 percent o f GDP in 2003.21This decline is explained in part by holding the public deficit within reasonable limits, and in part by the availability of external financing, which covered almost 60 percent o f public investment expenditure over the 1994- 2003 period. 3.34 The marginality of the crowding out effect i s verified not only by the limited amount o f public borrowing on the domestic market, but also by the relatively low level o f bank interest rates charged in Senegal. Large enterprises-those which intheory would be competing with the Central Government for obtaining credit-are able to borrow at a rate below 10 percent. The joint assessment conducted by the IMF and World Bank concluded that the Senegalese banking system had excess liquidity, a problem stemming more from the poor quality of the private projects to be financed and the risk-averse behavior o f the banks than from the crowding out o f private projects by the public sector. D.2 Increasingprivatesector privatizationby maximizingcomplementarities 3.35 The Senegalese authorities have indicated that they intend to encourage private sector participation in the financing and implementation o f public investment. The new BOT law promulgated in February 2004 i s considered to be in harmony with international best practices. The gradual introduction o f the legal and institutional framework explains the existence o f several projects carried out inpartnerships with the private sector, including the Dakar-Thies toll highway, the new Blaise Diagne Intemational Airport, the West Africa Business Park, and the standard gauge railroad. This list should be increased by adding the thermal power plant project and the industrial platform project. For most o f these projects, it i s still too early to program the total costs, andhence the scope o fprivate sector participation.22 3.36 It would, however, be a mistake to believe that mixedinvestment projects are a novelty in Senegal. Indeed, there i s a relatively lengthy list o f enterprises taking advantage o f joint participation by the public and private sectors. The list set forth below makes it clear that this is no marginal phenomenon, despite the privatization process, and that public sector capital inthese 2' Source: IMF, Statistical Annex, February 2005. This amount does not take public enterprise borrowing into account, which i s a sizable flaw inthe Senegalese monitoring system. 22 For the highway, the first financial simulations performed have made it possible to identify a Central Government contribution o f about CFAF 60 billion, or roughly half the total project cost. Inthe case o f the airport, it i s estimated that public sector participation would be minimal, between 8 percent and 40 percent o f the total cost, largely for strategic reasons rather than financial ones. 25 enterprises amounts to at least CFAF 178 billion (see Table 3.6). This amount excludes those enterprises in which the Senegalese State does not have a majority interest, some o f which are among the largest in Senegal, such as the chemical company (ICs), Sonatel, the MCridienHotel, and Air SCnCgal. This profusion o f enterprises with State participation o f course raises the issue o f the privatization process, which i s far from complete in Senegal. It also raises the issue of taking these enterprises' financial performance into account inthe Central Government accounts. The chronic lack o f information precludes precise evaluation o f the (implicit and explicit) risks being runby the Government inthe event o f their excessive indebtednessor bankruptcy. Inthis context, our recommendation i s to enhance the capacity o f the Audit Office to audit the financial statements of the public enterprises. Table 3.5: CentralGovernmentequityinterestsin the public enterprises Government equity Capital (in participation (in millions o f Government equity millions o f CFA CFA francs) participation (%) francs) Electricity (SENELEC) 119,433 100 119,433 Groundnuts (SONACOS) 22,627 87 19,696 Radio and television (RTS) 7,000 100 7,000 Public housing (HLM) 6,000 100 6,000 Port (SN PAD) 5,000 100 5,000 Water (SONES) 3,927 100 3,927 Postal services (La Poste) 2,900 100 2,900 Housing (SICAP) 2,743 90 2,468 Agricultural services (SAED) 2,500 100 2,500 Ship repair (SIRN) 1,928 99 1,909 Urban transport (DAKARDEMDIK) 1,500 77 1,155 Engineering (CEREEQ) 1,239 100 1,239 Sanitation (ONA) 1,200 100 1,200 Tourism (SAPCO) 1,200 99 1,185 Oil (PETROSEN) 1,200 99 1,187 Agricultural services (ISRA) 411 100 411 Vocational training (CNQP) 393 100 393 Vocational training (ONFP) 228 100 228 Auto insurance (FGA) 200 50 100 Trade promotion (CICES) 140 59 82 Agricultural services (SODAGRI) 120 54 65 Lottery (LONASE) 110 100 110 AgriculturaVrural training (ANCAR) 91 51 46 Press (SSPP Soleil) 27 55 15 Loan collection (SNR) 25 100 25 Press agency (APS) 23 100 23 Automobile manufacture (MSAD) 17 100 17 Food technology (ITA) 5 100 5 Total 182,188 178,321 Source: Government o f Senegal. 26 3.37 Private sector participation has a number o f advantages, such as the availability o f financing and the opportunity to enhance the management and utilization o f resources. The Senegalese Government i s well aware o f these advantages, which are summarized in a recent IMF They must still be evaluated in terms o f the potential risks to public finances represented by private sector participation. The main unresolved issue i s that there i s not yet any universally accepted accounting methodology that makes it possible, without ambiguity, to include projects with private sector participation in the Central Government budget. Hence, assuming the extreme case inwhich the project i s fully financed by the private sector, which also contributes all o f its operating costs, there i s inprinciple no need for any financial transaction to appear inthe Central Government budget. However, it would be manifestly incorrect to maintain that such a project entailed no risk for the Central Government, particularly ifthe latter provided financial or policy guarantees to the private producer. For this reason, efforts are being made to incorporate this risk incurrent accountingpractices, but there are differences from one country to another. In some countries, only those payments explicitly included in contracts are recorded on the Central Government's accounts. Inothers, as inthe case o fthe UnitedKingdom andthe State o f Victoria in Australia, the project i s regarded as a public investment in its entirety in the Government's accounts, regardless o f its financing structure. 3.38 Pending international agreement on an accounting methodology, Senegal, which has a program arrangement with the IMF, i s required to include all liabilities associated with a participation or concession contract in their entirety in the Government's accounts so that the public debt can be properly assessed.24This involves two parallel actions: 0 First, the net present value of the payments scheduled by the Government duringthe life o f the project (adjusted by the amount o f gains received) must be included as a liability in the initial public debt stock. This approach makes it possible to treat such payments as debt service over time. Second, the explicit guarantees offered by the government in each contract must be made public and the risks that they entail must be assigned a value by multiplyingthe stock o f guarantees by the probability o f loss occurrence. 3.39 The Senegalese authorities must improve their capacity to correctly assess the risks associated with their participation in public and mixed capital enterprises (whether existing or potential), and have this valuation audited by an independent institution in order to avoid conflicts o f interest. With the above, it i s strongly recommended that contracts which commit the Central Government be made public and disseminated in order to reduce the risks o f lack o f transparency. E. CONCLUSIONAND RECOMMENDATIONS 3.40 The Senegalese authorities, following the example o f other countries, highlight the relationship between public investment and economic growth as one o f the major arguments supporting their expansionary fiscal policy in recent years. This relationship is important, 23PublicInvestment and Fiscal Policy, March 12, 2004. 24International Monetary Fund,No. 12274,001850, August 2000. 27 because a country will not achieve economic success without improving its growth rate and per capita income. Moreover, it is now established that an increase in this rate leads to poverty reduction in Senegal. Unfortunately, historical experience has shown that while there i s a relationship between public investment and growth, it i s not pronounced, and in any event i s not as great as the relationship associated with private investment. The low productivity o f public investment in Senegal comes as no surprise owing to three factors. First, only 37.7 percent o f public investment in Senegal is in sectors that have a short-term impact on economic growth. Second, there i s some confusion between investment representing capital expenditure proper and the recurrent costs associated with it, which makes it impossible to evaluate the stock o f physical capital in the country and to maximize the retum on public investment projects over time. Finally, there i s a lack o f complementarity between the public andprivate sectors. 3.41 The analysis set forth in this chapter identifies a number o f recommendations which should guide the authorities in their efforts to optimize their investment policy, the most important o fwhich may be summarized as follows: 3.42 Composition of investment: It i s essential that the authorities introduce a system whereby they can evaluate and monitor the components o f investment, inparticular as regards the impact o f investment on the objectives defined in the PRSP, especially that o f achieving an economic growth rate exceeding 7-8 percent. To this end, it i s proposed to: 0 Increase the proportion o f investment devoted to the infrastructure sectors, for which the amounts are still relatively low by international standards and also low in light of requirements inSenegal, which is somewhat behindinthis regard. 0 Maintain investment levels in the social sectors while improving the analysis o f the impact o f such spending, inparticular by examining intermediary variables for evaluating andmonitoring the productivity o f such expenditure over time. 0 Take stock o f the justification o f investmentexpenditure insectors not directly associated with economic growth, with special emphasis on projects classified as "administrative equipment," which represent a growing share o f the B C I and encompass a wide range o f activities, without any genuine economic analysis being perfonned to justify them and to evaluate their economic and social rates o fretum. 3.43 Investment expenditure and recurrent costs. The relationship between these two categories o f expenditure needs to be improved, as it contributes to skewing the allocation o f resources and reducing the productivity o f public investment over time. These investments deteriorate rapidly because they are not maintained in good condition. For example, the massive investments inthe social sectors will not be optimized unless the number o f teachers i s increased intandem withthe growthinthe number ofschools. Basedonthe above, we recommend: 0 Producing the investment budget using an economic classification, which will make it possible clearly to identify those expenditures that contribute to the accumulation o f physical capital and those which constitute recurrent costs. At the same time, there should be a clear identification o f the recurrent costs included in the operating budget, which cannot be discerned either globally or by sector through a reading o f the existingbudget. 28 Adopting a planning system that will ,make it possible to anticipate the recurrent costs associated with the investment projects in each sector and to compare them with international standards. 3.44 Private sector participation. The authorities are contemplating private sector participation in public investment projects through partnership arrangements. A number o f projects-especially in the infrastructure area-are already programmed, but there are already many enterprises inwhich the Central Government has a majority or minority equity interest. To maximize the advantages o f such arrangements while minimizing the fiscal risks, it i s recommended to: Continue the privatization policy, particularly as it applies to enterprises engaged in productive activities such as automobile manufacturing (MSDA), the press (le Soleil), or ship repairs (SIRN). Finalize the implementation of the institutional and legal framework, particularly the regulatory framework. 0 Adopt an accounting framework that will make it possible to reflect in the Central Government's accounts the risks associated with projects with private sector participation. 29 4. PUBLICINVESTMENTAND BUDGETARYINSTRUMENTS 4.1 In the earlier chapters, we showed that the level and composition of investment both played an essential role in achieving the development objectives set by the Government in its PRSP, in particular as regards economic growth. Although there have been several recommendations made as to the composition o f the link between capital expenditure and operating expenditure, this chapter focuses primarily on the strategic objectives as expressed by the Government inthe Priority Action Program (PAP) for 2003-05. The relevant question then i s whether these choices are properly carried over into the budgetary instruments available to the Central Government. More specifically, this chapter will endeavor to answer the following two questions: 0 To what extent do investmentbudget allocations reflect the strategic choices expressed in the PAP? Do the investments made correspond to the expectations o fthe strategic choices? 4.2 The answers to these questions should enable us to identify the current problems o f the system used by the Senegalese authorities and to make a number o f specific recommendations aimed in particular at improving the Government's capacity to plan investments in the medium term and to establish a methodology that provides a systematic linkage between the strategic objectives and the investment budget approved and implemented. It i s worth noting that this effort to improve the investment budget should be carried out within the context o f an integrated budget, as emphasized in Chapter VI, as the interdependencies between the two budgets cannot be disregarded, particularly as regards the linkages between capital expenditures and the associated recurrent costs. A. STRATEGIC CHOICES IN SENEGAL AND BUDGETARYTOOLS 4.3 Senegal's development strategy for the medium term i s delineated in its PRSP, which proposes an allocation o f the investments to be made as part o f a Priority Action Plan (PAP) covering the 2003-05 period. It provides a list o f activities reflecting their financing requirements and relationship to the three pillars o f the PRSP, namely wealth creation, capacity building in and the promotion o f basic social services, and the protection o f vulnerable groups (see Table 4.1). A briefreview indicates that the allocation under the PAP i s focused on the first two pillars, with a limited share o f the total devoted to the vulnerable groups. At a sectoral level, the PAP emphasizes agriculture, the environment, health, education, and the supporting infrastructure. 4.4 The PAP should be regarded as a planning instrument which orients policy-makers in their preparation o f the annual budget. The Senegalese Government has two main budgetary instruments for planning such investment expenditure, namely: (i) the Consolidated Investment Budget (BCI), which encompasses the investment expenditure projected by sector and by project for the current year and sets forth the nature o f the external or domestic financing thereof; and (ii) Three-Year PublicInvestmentProgram(PTIP),whichincludesathree-year forecast for the the investment to be financedbydomestic and external finds. 30 Table 4.1: Priorityactionplan(2003-2005) Safe drinkingwater Natural resourcesand Source: PRSP (2002). 4.5 The B C I and the PTIP supplement the PAP in the sense that they should provide complete and detailed information on the investmentprogram for the coming year (the BCI) and over a medium-term horizon (the PTIP). Without question, however, there are several shortcomings which are harmful to the clarity o f budgetary planning in Senegal. For example, the forecasting capacity o fthe PTIPis minimal, giventhat the PTIP formulated in2004 projected total investment of CFAF 415 billion in 2005 while the B C I for 2005 authorizes expenditure o f CFAF 575.9 billion, a difference o f almost 39 percent. More seriously, it is difficult to interlink the B C I and the PAP owing to the absence o f any budgetpresentation by program. At present it i s difficult to evaluate whether the budgetary instruments are actually in line with the strategic objectives ofthe PRSP. 4.6 In view of this finding, the action we propose is predicated on two measures: first, improving the PTIP as a medium-term planning. instrument; and second, implementing a methodology that makes it possible to establish the extent to which budgetary allocations for investmentare inkeepingwith the strategic choices ofthe PRSP. 31 B. STRENGTHENING MEDIUM-TERM PLANNING 4.7 The weakness o f medium-term investment planning is illustrated in Table 4.2 below, which shows that the PTIP systematically underestimates investment amounts from one year to the next. The gaps between the amount anticipated in the PTIP and the amount effectively allocated the following year may approach 40 percent.25The major forecasting disparities relate to the quaternary sector and, to a lesser extent, the primary sector. The sole exception for 2005 i s the spike in investment in the tertiary sector (infrastructure), which explains half the forecasting gap.26 A more detailed analysis o f the quaternary sector shows, in tum, that the gaps stem primarily from the health and nutrition subsectors (accounting for almost 10 percent o f the gaps for 2005 between the 2004 PTIP and the 2005 PTIP) and from administrative equipment. The share o f the latter subsector inthe explanation of the gap cannot be overlooked, as it amounts to almost 30 percent in 2005 after accounting for nearly half the gap in 2004. The forecasting gaps inthis subsector arethus systematic, which shoulddraw the attentionofthe authorities. 4.8 The sectoral breakdown o f PTIP forecasting gaps makes it possible to provide a preliminary response to the remark frequently heard in Senegal to the effect that the gaps are principally attributable to the difficulty o f forecasting donor-financed projects. While this argument does apply to the health sector (where virtually 100 percent o f the projects are cofinanced with external funds), it does not extend to the administrative equipment sector, where projects are exclusively financed with domestic resources. The explanation o f the forecasting gaps i s found more in the large number o f projects (over 60 in the 2005 PTIP), the nature o f these projects focusing on highly disparate activities, and the absence o f monitoring and evaluation. It remains true, however, that coordination with donors is essential in order to improve the planning o f investment expenditure inSenegal. 4.9 The gaps between the investment expenditure forecasts and the approved budgets are systematic, and inpart reflect the excessive caution o f the authorities, who do not wish to include projects that are still uncertain. However, the gaps also reflect serious deficiencies inthe current planning system, especially the difficulty o f anticipating investment project execution over time.27The size o f the forecasting gaps is such that the PTIP cannot play its proper role, that of guiding budget allocations not only in light o f the strategic objectives but also in terms o f the macroeconomic framework. 25 See Table 8 o f the Statistical Annex for a detailed breakdown o f the PTIP for the 2002-2007 period. 26 In2005, there is a major gap inthe tertiary sector owing to the increase ininvestments inurbanmobility, butthis gap should be regarded as isolated inthat it may be traced to a shift inthe Government's policy. 27 Inthe next chapter, we will see that BCI implementation ratios vary not only over time but also across sectors. The authorities are unable to predict these ratios, making it difficult to plan investment expenditure over the coming years. 32 Table 4.2: Three-year forecast ofinvestmentbudget,by sector (Inbillionsof CFA francs) PTIP PTIP PTIP PTIP (2002-2004) (2003-2005) (2004-2006) (2005-2007) 2002 408.1 Primary 97.2 Secondary 30.5 Tertiary 67.9 Quaternary 212.5 2003 396.4 449.9 Primary 87.2 111.3 Secondary 21.4 28.7 Tertiary 80.5 79.8 Quaternary 207.3 230.1 2004 377.1 408.2 468.6 Primary 101.9 93.5 110.6 Secondary 10.5 23.5 34.9 Tertiary 100 92.3 80.9 Quatemary 164.7 198.9 242.2 2005 334.5 415.0 591.7 Primary 93.1 102.1 121 Secondary 18.8 20.2 21 Tertiary 80.3 95.5 162.2 Quaternary 142.3 197.2 287.5 2006 310.1 483.2 Primary 87.7 96 Secondary 17.8 25.1 Tertiary 52.9 135.1 Quaternary 151.7 227 2007 373.9 Primary 94.3 Secondary 22.9 Tertiary 87.7 Quaternary 169 Source: Senegalese authorities. 4.10 Given the scope o f the problem, the Senegalese authorities have not sat idly by, and have already made efforts to improve their capacity to forecast medium-term investment expenditure. For example, they are now using a flexible approach for planning such expenditure over time, making it possible for the amounts to be revised from one year to the next. They have also devoted particular attention to the state o f progress o f projects, drawing distinctions between those for which the financing i s still beingnegotiated, those for which it i s lined up, andthose for which it i s still being sought. An effort has also been initiated to present investment expenditure byministryinthe newbudget law for 2005.28 4.11 For this effort to continue, it also appears necessary to strengthen the staff concerned and their coordination, in particular in the social sectors and the administrative equipment sector, 28However, achieving a breakdown o f investment expenditure by ministry should not lead to the elimination o f the BCI document, as it includes a great deal o f information. Moreover, the breakdown o f the investment budget by sector becomes impossible since gaps remain between the amount o f the PTIP and the amount o f the BCI for the current year, even though the PTIP i s broken down by sector. 33 which appear to be the primary sources for the forecasting gap in the PTIP in recent years. To this end, it is recommended to: 0 Improve information flows between the Ministry o f Economy and Finance-the central planner-and the sectoral ministries. 0 Improve coordination with donors in order to identify in advance in each sector, and especially the social sectors o f health and education, fbture projects and the likelihood o f their beingimplemented. 0 Buildplanning capacities inthe Ministryo fEconomy and Finance. 4.12 Inaddition to implementing these practical recommendations, the Senegalese authorities should stress two conceptual aspects which could improve their budgetary instruments and thereby enhance the efficiency o ftheir investment expenditure. First, the PTIP should distinguish clearly between capital expenditure and the recurrent costs associated with investment projects. Such a distinction i s essential, as we argued in the preceding chapter, not only to evaluating the stock o f physical capital in Senegal, but also to maintaining the rate o f return o f investment projects inSenegal. 4.13 Second, we would like to encourage the authorities to include medium-terminvestment planning that would provide information on the regional allocation thereof. This inclusion i s desirable at a time when the authorities are focusing on regional development and decentralization. The first stage would be to present a regional classification of the investment projects in the PTIP, which would provide clarity and transparency for all local elected officials. It would then make it possible to assess whether the current regional allocation is the optimal one. This i s a question o f concern to many governments, and international experience suggests that the best practice i s to minimize arbitrary choices by using allocation criteria. Along these lines, the Senegalese authorities could considering following the example o f Brazil, where the regional allocation o f investment in the health sector i s guided by the following criteria: (i) size o f the region, as determined by its population; and (ii)the requirements expressed by each region. c. INVESTMENTS APPROVED AND IMPLEMENTEDINACCORDANCE WITH PRSPPRIORITIES 4.14 While the Senegalese authorities state that the annual budget set forth inthe Budget Law is prepared explicitly while taking account of the objectives defined in the PRSPY2'there is no detailed methodology whereby they can actually ensure a systematic linkage between the PAP o f the PRSP and budget programming. The methodology currently being used is crude-the authorities determine ratios interms o f allocations inthe operating budget for the priority sectors 29The 2005 Budgetlaw states that "the 2005 draft budget is consistent with observing the commitments made inthe PRSP" (2005 BudgetLaw, p. 1). 34 such as education and health-and is partial in that it does not incorporate investment expenditure.30 Box 4.1: A Methodology for evaluating the regional allocation of investment the example of the healthsector inBrazil The allocation of investment expenditure by region i s a question o f concern to the Senegalese authorities. To assist them in this exercise, we propose below a methodology developed by the Govemment of Brazil for the health sector, which should be regarded as food for thought. In order to avoid determining the regional allocation o f projects solely on the basis o f local governments' capacity to formulate requests and by arbitrary decision-making by the central authorities, the Brazilian authorities established a system which sets ceilings byregion. These ceilings are determined on the basis of: (i) the population o f each region; and (ii) indicator o f need in each region. The an ,needindicator factor was defined as the amount o f current expenditure in the health sector in each region, as this was deemed to be a good indicator o f health needs inthe region and because the data were readily available and checked by the Central Government. This expenditure indicator was preferred to performance indicators such as the vaccination coverage ratio, on which statistics are not always reliable or available. The formula used i s as follows: Ii/I=a(pop,/pop) (1-a)(Xmax -Xi)/( + resents public investment in the health sector, Iirepresents public investment in the health sector in region i,pop is total population, popi i s the population inregion i,X is the average amount o f operating expenditure inthe health sector-used A as an indicator o f investment requirements for that sector, Xmax is the maximum amount o f operating expenditure attained in a region, and Xi the amount o f operating expenditure inregion i.In accordance with this formula, the allocation o f health investment by region has a positive correlation with the share o f total population represented by the region's population, and with the relative gap betweenthe operating expenditure inthe region which spends the most and the operating expenditure o fthe regionat issue. Source: Philip Musgrove, "Equitable Allocation o f Ceilings on Public Investment," HCD Working Paper, World Bank, August 1996. 4.15 At this stage, it is essential for the Senegalese Government to be able to assess whether the investment expenditure set forth in the BCI is properly aligned with the objectives o f the PRSP. Such alignment i s important not only for achieving those objectives, but also in connectionwith maintaining dialogue with donors. 4.16 Incollaboration with a number of offices inthe Ministryof Economy and Finance, we endeavored to develop a methodology for evaluating the extent to which budgetary instruments are in keeping with the strategic objectives o f the PRSP. Before presenting the results for the 2003-2005 period, it should be noted that while the approach is delineated in detail in Tables 9 and 10 o f the Statistical Annex, its broad outlines may be summarized as follows: we began by aligning the various objectives delineated in the PAP with the related priority actions, and then compared these with the projects identified in the B C I which could be considered to be contributing to achieving these objectives and priority actions. This classification o f investment projects according to PRSP objectives was sometimes difficult, as some projects may be 30 The classification o f investment expenditure i s currently performed by means o f a coding system which assigns the letter P to those expenditures under the BCI which are in principle associated with the general objective of reducing poverty. However, even the authorities are o f the opinion that this crude classification i s not really o f value for assessingthe extent to which the BCI is inline with the strategic objectives o f the PRSP. 35 pursuing several objectives at once or may have objectives which evolve over time.3'Despite these difficulties, this leveling o f information enabled us to compare the budgeting o f the Priority Action Plan (PAP) for the 2003-2005 period first with the budget allocations inthe B C I and then with BCIimplementation. C.l The BCI as approvedandthe objectivesof the PRSP 4.17 The first finding to emerge from our methodology is that about 40 percent o f the B C I as approved is devoted to combating poverty during the 2003-05 period (see Table 4.3). This constant share may at first glance be disappointing, but it does correspond to amounts that rise in absolute value terms from CFAF 183 billion to CFAF 225 billion between 2003 and 2005, bearing witness to the desire o f the Senegalese authorities to support poverty reduction. It should also be noted that expenditures on urban infrastructure i s not included-this by agreement with the Senegalese authorities-as these were not given priority status under the PRSP. Table 4.3: Share of the BCI allocatedto combatingpoverty (in millionsof CFA francs) 2003 2004 2005 2003-2005 Total PAP amount 197,l 201,9 210,6 609,6 Poverty projects inBCI 183,3 193,6 225,l 602,l Total BCI 461,O 461,6 591,7 153,4 Share o f BCI allocated to PRSP 39.8% 41.9% 38.1% 39.8% Source: Statistical Appendix, Table 9. 4.18 The second finding is that the allocation o f the B C I i s perfectly in line with the three pillars defined in the PRSP, inasmuch as the amounts approved are virtually the same as those anticipated in the PAP (see Chart 4.1). The averages nevertheless show sizable disparities because the coverage for 13 o f the 21 activities defined in the PAP are lower than 80 percent, and for 9 the coverage i s below 50 percent for the 2003-2005 period (Table 4.4). This uneven coverage o f the PAP by the B C I may be identified by means o f a synthetic indicator.32We find that this dispersion indicator i s equal to 1.53 in 2003, but increased to 3.10 in 2004 before dropping back to 2.06 in 2005. In other words, the best coverage i s achieved in 2003 and the worst in2004, with some reversal in2005. This lack o f alignment at the sectoral level highlights the lack o f a systematic approach to aligning BCI allocation and the PAP objective^.^^ 31Moreover, any budget allocation intended for an investment project includes costs that are unrelated to physical investment and do not directly address the poor. Hence in the absence o f budgeting by objectives, it is difficult to determine what is being devoted to the achievement o f a given objective. 32We use the ratio o f the sum o f coverage variances for each sector divided by the average coverage for all sectors. A perfect coverage for all sectors should yield a ratio equal to 0. 33 This lack o f alignment may also be explained, in particular for 2005, by the modification o f the strategic objectives o f the PRSP. Thus, the authorities accorded a considerably greater relative importance to investments in urban infrastructure and urban mobility, which were not explicitly identified in the initial PAP prepared in 2002. In this case, it is essential that the authorities make the necessary adjustments intheir PRSP and their PAP to justify their new objectives ina transparent and participatory manner. 36 4.19 While the pillar devoted to wealth creation i s covered overall by the B C I (exceeding the allocation projected by the PAP by 1 percent over the 2003-05 period), there are enormous variations from one sector to the next. Agriculture i s by far the sector benefiting the most, followed by support infrastructures and energy in terms o f B C I allocations. Conversely, the private sector, crafts, employment, and the tertiary sector fare relatively less well under the BCI. Figure 4.1: Alignment of the BCIwith the pillars of the PRSP Investmentbudget allocations in terms of the strategic Pillars of the PRSPfor 2003-2005 700,000 600,000 500,000 I 400,000 300,000 200,000 100,000 0 Overall Wealth Capacity Vulnerable strategy creation building groups The investment expenditure devoted to poverty reduction in the agricultural sector exceeds the expenditure called for under the PAP by 50 percent. Objectives such as the promotion o f a sound marketing and distribution policy, the development o f irrigated agriculture, and the promotion o f agroindustry, are indeed taken into account in the budget. However, the activities aimed at intensifying and modemizing agricultural production and at enhancing the role o f small farmer organizations should be strengthened. The major projects contributing to the revitalization o f this sector are the PSAOP, the Agricultural Projects Guarantee Fund, and the rehabilitation and 1,700 hectare expansion o f irrigated perimeters inthe MCdina PCtC zone. 0 Investment expenditure aimed at poverty reduction by supporting basic infrastructure in rural areas covers only 73 percent o f the PAP forecasts. The main projects contributing to meeting the objectives o f strengthening support infrastructures and reducing the isolation o f rural areas are the National Rural Infrastructure Program and the Rural Equipment Program. 0 The objectives pursued under the PRSP in the energy sector, namely improving and securing the public's access to household fuels and enhancing rural electrification, have only 78 percent coverage under the B C I for the 2003-2005 period. The principal projects 37 contributing to meeting these objectives are the traditional energy management project and the project on rural electrification and solar energy. 4.20 The B C I allocations effectively cover all the PAP forecasts for the second pillar on capacity building and the promotion o f social services (99 percent), though exceptions must be made for environmental activities and, to a lesser extent, education (especially elementary education). The objectives inthe health sector, to wit to improve the quality and supply o f health services and to improve efforts to prevent and combat HIV/AIDS and malaria, are suitably covered by the Integrated Health Development Program and the Strategic Plan for Combating AIDS. Inaddition, the Government has filly reflected the objectives o f increasingaccess to safe drinkingwater and improving access to adequatewaste water and trash removal systems through sanitation programs and the construction o f tube wells and conventional wells in Dakar and the other cities. The sanitation component o fthe long-term water sector project supplements all these initiatives. 4.21 The amounts recorded in the B C I to benefit the protection o f vulnerable groups amount overall to 90 percent o f those called for under the PAP for the 2003-2005 period. Women and the disabled receive more financial support than projected in the PAP. Support for women i s channeled through the project to combat female poverty, the small appropriation for women's groups, and the appropriation to improve the economic and social standing o f women. The promotion o f education and training for the disabled i s filly addressed through the creation o f centers for the disabled and the mentally ill.The least favored groups are the elderly and children, and the projects associated with developing housing for the vulnerable groups also do not fare well. The actions in favor o f children are focused instead on addressing their needs in community nutrition centers under the program to enhance nutrition. Further efforts must be made to combat the worst forms o f child labor, to protect children at risk, and to improve the living conditions o f the talibk child beggars in the daara Koranic schools. The objective o f extending the vulnerable groups' access to improved lots and adequate housing i s only 80 percent covered by the B C I as compared to PAP projections. C.2 BCI as executed and PRSPobjectives 4.22 While the B C I as voted appears relatively well aligned with the amounts set forth inthe PAP, the same cannot be said when the comparison i s based on the B C I as executed. Thus, in 2003 (the only year for which data are available), coverage works out to just 53 percent, versus 93 percent (see Table 4.5). This disappointing result highlightsthe weaknesses o f the Senegalese system as regards the execution o f investment projects. Even worse, it shows that the priority projects o f the PRSP have execution rates lower than the average for the B C I (53 percent versus 67 percent), although the opposite should have beenexpected. 38 Table 4.4: Project amounts includedinthe BCI, 2003-2005 (As a percentageof the PAP amounts) Creation of wealth 100.9% Agriculture 139.9% Infrastructure 73.0% Livestock 425.1% Fisheries 60.5% Handicrafts 54.0% Industry 30.0% Energy 78.2% Mining 47.6% Tertiary sector 40.2% Private sector 0.0% SME support 186.3% Employment 14.3% Social services & capacity bldg 98.9% Educatiodtraining 74.2% Health 146.7% Drinkingwater 105.6% Environment 36.8% Sanitation 406.6% Vulnerable groups 90.2% Children 32.8% Women 417.6% Disabled 222.6% Elderly 0.0% Housing for vulnerable groups 80.0% Source: Statistical appendix, Table 9 4.23 The strategic category "creation o fwealth" shows an average executionrate of 65 percent inrelation to budget allocations, with the result that BCIcoverage o fthe PAP is only 61 percent in2003. This is due to the fact that sectors such as agriculture, infrastructure, and energy show execution rates o f 51 percent, 79 percent, and 83 percent respectively. The rate o f coverage o f the second category, focusing on capacity-building and the promotion o f basic social services, i s only 42 percent in 2003 because the sectors that appear to be relatively privileged in the BCI allocation, such as health, drinking water, and sanitation, show extremely low rates o f execution o f their projects (59 percent, 32 percent, and 3 percent respectively). As for the category "vulnerable groups," the rate o f coverage decreases to 62 percent, mainly because o f the relatively low execution rate o fprojects targeting children. D. CONCLUSIONAND RECOMMENDATIONS 4.24 Senegal, to its credit, possesses a development strategy inwhich its priority objectives for poverty reduction are set forth, and which serves as a frame o f reference for its budget programming activities and its dialogue with donors. However, the management systemneeds to be further improved in order to better align budget expenditures with the strategic priorities 39 established for the long term. This chapter highlights two areas o f action that should help the Senegalese Government achieve this objective. Table 4.5: Coverage of strategic objectives and BCI as executed, 2003 BCIas BCI as executed voted/PAp BCIas executed/PAP BCI as voted TOTAL 93% 53% 57% Creation of wealth 94yo 61% 65% Agriculture 134% 68% 51% Infrastructure 61% 48% 79% Livestock 463% 274% 59% Fisheries 53% 28% 52% Handicrafts 24% 1% 6% Industry 0% 0% 0% Energy 73% 60% 82% Mining 57% 57% 00% Tertiary sector 30% 30% 00% Privatesector 0% 0% 0% SME support 163% 254% 55% Employment 0% 0% 0% Services and capacity-building 90% 42% 41% Educatiodtraining 75% 62% 84% Health 144% 86% 59% Drinking water 143% 45% 32% Environment 27% 9% 32% Sanitation 209% 7% 3% Vulnerable groups 100% 62% 62% Children 52% 17% 32% Women 237% 220% 93% Disabled 442% 0% 0% Elderly 0% 0% 0% Housingfor vulnerable groups 46% 23% 53% Source: Statisticalappendix, Table 10 4.25 The first area o f action focuses on the need to improve the PTIP forecasting capacity, which i s deficient in several respects. The recommendations are to build the capacities o f the relevant departments o f the Ministry o f Economy and Finance, promote a better flow o f information with sectoral ministries and local governments, and strengthen coordination with donors. In addition, officials should attempt to expand the PTIP in two directions in order to better grasp and optimize the evolution and allocation of investment over time, namely: (i) establishment o f the link between capital expenditures and recurrent costs; and (ii)the geographic or regional breakdown o f investment projects. 4.26 The second area o f action focuses on the needto implement a monitoring system so as to be able to assess the alignment o f expenditures listed in the BCI in relation to the strategic objectives o f the PRSP. The methodology presented inthis chapter has already been shared with the government. This methodology uncovered a number o f gaps which officials will need to address: 40 0 The percentage o f the B C I as voted that targets the PRSP objectives has remained constant at roughly 40 percent, which means an increase in absolute value o f nearly CFAF 50 billion from 2003 to 2005. Nevertheless, the monitoring and evaluation o f these expenditures remains weak and, as a result, officials have very little grasp o f their impact, leading many observers to question the grounds and effectiveness o f these expenditures. One possibility i s to adapt the approach usedinUganda, where attention i s paid to a set o f sectors and projects definedin advance (see box). 0 PAP coverage by the B C I is uneven at the sectoral level, and the sectors most heavily penalized are the environment, employment, the private sector, children, the elderly, and housing for vulnerable groups. This record illustrates that the government must align its B C I not only in relation to the overall objectives o f the PAP but also on the basis o f its projected commitments in the subsectors deemed essential in its poverty reduction strategy. As such, it is strongly encouraged that two initiatives now under way be finalized. First, the CSLP unit is inthe process o f harmonizing regional operations plans and sectoral operations plans, which should result in a better alignment, including at the local level. Second, the Senegalese Government plans to finalize its Medium-Term Expenditure Framework on a programmatic basis for at least four sectors by the end o f 2005 (education, justice, environment, and health), and this represents the best way to optimize the link between the investment budget and the PRSP objectives. 0 The overall coverage o f the PAP isjust 53 percent, based on the B C Ias executed. This is wholly insufficient and illustrates one o f the fundamental weaknesses o f the government finance system in Senegal. Urgent action to improve the execution rates i s critical to the success o f PRSP implementation in Senegal and, accordingly, we shall focus our attentionon this important matter inthe next chapter. Box 4.2: The Ugandan Experience Uganda i s a trailblazing country where a deliberate effort has been made to align the budget instruments with the strategic objectives set forth inthe development strategy. In 1998 the concept o f a Poverty Action Fund(PAF) was adopted, allowing the government to set a volume o f priority spending in its budget and objectives to be reached over time. After a few years o f operations, the results are favorable eventhough certain weaknesses have emerged. The principal advantage lies in the government's formal commitment to this spending from a medium-term perspective. This has enabled the govemment to make internal financing more secure and also, most o f all, receive donor support. Priority projects have also received heightened attention from the government, leading to increased rates o f execution. The government had wisely set aside 5 percent o f the total PAF amount to finance the monitoring and evaluation o f priority projects. This attention has served to increase the productivity o f this spending, especial11 insocial sectors suchas education andhealth. On the negative side, one weakness has beenthe rigidity of the system. While the volume of spending allocated to the PAF has increased over time (reaching approximately 35 percent o f the total budget in 2002, versus just 18 percent in 1998), this has tended to supplant other expenditures that, inhindsight, have turned out to be important, particularly the country's needs interms o f financing for infrastructure, which has been strongly penalizedbecause it was not incorporated into the PAF. This lack o f flexibility i s partly the result o f the administrative rigidity o f the govemment and donors, but it is also partly due to the lack o f a truly integrated system for evaluating public 41 5. INVESTMENT PROJECTEXECUTIONRATES 5.1 While Senegalese officials have demonstrated their capacity to align their investment budget inrelation to the overall objectives o f their poverty reduction strategy, the same cannot be said o f the execution o f priority projects. Planned schools, roads, and health centers are not built or are only built after considerable delays. Inthe end, as shown inthe preceding chapter, budget coverage o f the Priority Action Plan i s partial, barely exceeding 50 percent, and uneven. This partial and uneven performance underscores the need to improve the execution rates so that the priorities selected by the government actually take material form through the delivery o f capital assets to the Senegalesepopulation. 5.2 This chapter begins with a description o f the execution rates associated with investment expenditures in recent years in order to bring to light the degree o f variation that may be observed not only over time but also across different sectors. The second section is devoted to an examination o f the principal factors that influence the execution rates associated with investment projects in Senegal. To that end, we will use a sample o f 240 projects provided by the Senegalese Government and information collected on the portfolio o f World Bank projects. The final section will then conclude with a number o frecommendations. A. INVESTMENT EXPENDITUREEXECUTIONRATES 5.3 Investment expenditure in Senegal i s characterized by a relatively low level o f execution incomparisonto other public spendingincluded inthe government's budget. Duringthe period 2000-2003, the investment expenditure execution rate averaged just 67 percent, versus 90 percent for all other spending (see Table 5-1). This result i s not surprising inasmuch as capital expenditures are subject to authorization and disbursement procedures that are generally more complex than for other spending, not only because o f their relatively high amounts, but also due to their economic and social implications and their private sector subcontracting arrangements. 5.4 Interestinglyenough, it is difficult to detect a trend inexecution rates over time. The rate observed in 2003 is higher than the rate for 2002, but lower than the rate for 2001. The early indications for 2004 are that the average disbursement rate may have slightly increased. Infact, it is riskyto interpret the annual rates because investment projects are spread over several years, which makes it preferable to examine their execution rates over a relatively long period o f time. The disbursement rates vary tremendously by sector, reaching nearly 100 percent in certain sectors such as health and fisheries, but falling below 50 percent in others such as livestock, social development, and tourism (see Figure5-1 and Table 7 o f the statistical appendix). 42 Table 5.1: Executionrateby type of expenditure 2000 2001 2002 2003 2000-03 Labor costs 98% 95% 104% 98% 99% Goods and services 104% 73% 95% 94% 91% Transfers 85% 101% 118% 108% 103% Investment 59% 61% 82% 67% 67% Debt service 84% 113% 84% 101% 95% Total budget 79% 81% 92% 86% 84% Figure5.1: Investmentexpenditure execution rates by sector, 2000-2003 0.0 20.0 40 0 60.0 80.0 1000 120.0 Execution rate I Source: DCEF andWorld Bank 5.5 The low execution rates associated with investment expenditures and their temporal and sectoral variations have caught the attention o f Senegalese officials. Their objective is in fact to increase these rates and minimize their variations so that actual spending i s as close as possible to the strategic objectives and the initial allocation o f the investment budget. From this perspective, we shall attempt to identify below the main determinants underlying the variations inthe executionrates observed inSenegal. 43 B. DETERMINANTSTHEVARIATIONINEXECUTIONRATES OF B.l Statistical analysis of the determinants 5.6 For the purposes o f this analysis, we use a data base collected from the Senegalese Government covering 240 projects during the period 2000-2002. In addition to the execution rates for each project, this data base provides information on the size o f the projects in CFAF million, the sector and subsector o f activity, execution projections during the year in progress, and sources o f financing with a distinction made between the government and external resources (loans and grants). For this sample, the average execution rate works out to 74.3 percent, roughly the same as for the total BCI. However, there i s great variation in the rates, ranging from a low o f 0 percent to a higho f 200 percent. Box 5.1: Characteristics of the sample education andhealth. followed by transport projects and fishery projects with investment levels o f CFAF 15 billion and CFAF 8 billion 5.7 The variations noted in the execution rates o f investment projects are explained by a number o f variables which we have identified in the statistical approach presented in Table 5.2 below.34 The principal results may be summarized as follows: 0 Projects in the transport and social services sectors have execution rates systematically lower than those inother sectors because the coefficients associated with these sectors are negative and significantly different from zero. In contrast, projects classified as administrative equipment projects are executed at a higher rate than the rest o f the sample, even allowing for the fact that they are exclusively financed from internal resources. 34The results reported in Table 5-2 are only for 2002 because it is easier to interpret a transversal regression than one that uses cross-cutting "panel data." However, it should be noted that the results are nearly identical for all three years 2000, 2001, and 2002 and are available uponrequest. 44 The size o f the pro'ect influences positively, but only slightly, the execution rate o f investment pr0je~ts.l~ This result i s perhaps a bit surprising at first glance, since the procurement procedures are generally more complicated for large projects which are required to obtain CNCA authorization. It may however reflect the special attention which projects o f large scope receive because they are more visible and, perhaps, strategic. 0 The sources of financing have a significant impact on disbursement rates. The explanation often given, namely that projects financed with external funding are disbursed more slowly than those financed with intemal fbnds, appears to be partially verified by our analysis, particularly for projects financed through credit (in contrast, projects financed through grantshbsidies show relatively high execution rates). These variations seem to indicate the existence o f profound differences between donors, a matter which deserves closer examination when the data become available. They also suggest the existence of possible bottlenecks in certain sectors favored by donors, which exceed the absorptive capacities of the SenegaleseGovernment. Table 5.2: A statistical explanation of execution rates, 2002 Execution rate Executionrate Execution rate Executionrate Constant 0.061 0.105 0.162 0.488(*) (0.35) (0.60) (0.73) (2.66) Transport 0.249(*) -0.282(*) -0.175 -0.175 (-2.20) (-2.46) (-1.57) (-1.57) Admin.infrastr. 0.233(*) 0.196(*) 0.394(*) 0.394(*) (3.02) (2.45) (4.54) (4.54) Social services -0.153 -0.202(*) -0.202(*) (-1.72) (-2.40) (-2.40) Size (log) 0.057(*) 0.055(*) 0.042 0.042 (2.86) (2.79) (2.07) (2.07) Government 0.582(*) 0.590(*) 0.642(*) 0.317(*) financing (7.03) (7.08) (4.93) (3.43) External financing: -0.325(*) credit (-2.64) External financing: 0.326(*) grantstsubsidies (2.64) Other financing -0.304 -0.629( *) (-2.08) (-5.51) R2 0.203 0.205 0.296 0.296 No. observations 240 240 240 240 Note: An asterisk (*) means that the coefficients are significantly different from 0 within a confidence interval above 95 percent. 5.8 The limits of the statistical analysis are evident. They center first of all on the quality o f the data since the analysis focused on only a limited sample of projects. Inaddition, the data are also relatively old, stopping in 2002, which illustrates the relatively weak capacity o f the 35While project size correlates positively with the execution rate, this positive correlation declines with the size o f the project. Inother words, we find a positive but declining effect based on the size o f the project. 45 Senegalese Government to compile budget information on the basis o f execution and not just allocation within a relatively short time frame (a lag o f nearly two years is to be expected). Finally, the number o f characteristics associated with each project should be expanded in order to better identify the determinants that influence execution rates, notably the geographic implementation o f the project, its implementing agency, and the identity o f the donors. Nevertheless, even if they should be interpreted with caution, the results given above match intuition to a certain extent and should not surprise Senegalese officials. B.2 Level of centralizationof the implementingagency 5.9 The Senegalese Government is aware o f the need to increase the execution rates of investmentprojects, particularly those included on the priority list o fprojects insupport o fPRSP objectives. One important question dominates this debate, namely the level o f decentralization o f the implementing agency. It can in fact be argued that the responsibility for B C I execution is too centralized in Senegal and that it should be decentralized because local officials are closer to the beneficiaries and thus have greater motivation to execute and monitor the progress o f investment projects. 5.10 The government has taken the initiative to decentralize management o f its B C I infavor of local governments. Although this initiative is described in the next chapter, we would like to examine here the recent experience o f local governments in this regard. The absence o f data makes an overall assessment difficult, but an initial assessment can be drawn from projects that are part o f the World Bank portfolio in Senegal. The main lessons can be summarized as follows: 0 In cases where execution was Box 5.2: Urban development credit and decentralization program decentralized, The credit for urbandevelopment and the decentralizationprogramI(Cr. investment projects 3006, approved in 1997 and closed in December 2004) introduced the showed relatively concept o f municipal contracts in support o f the decentralization strategy high execution rates pursued in Senegal. This project involved the participation of 67 overall. Examples municipalities that signed contracts in excess of CFAF 46 billion in the include the National form of more than 400 investmentprojects, including the construction o f Rural Investment markets, bus stations, schools, health centers, and roads, thus contributing to the country's local andurban development. Project (PNIR), with a rate o f more than 83 To execute this project, implementation was placed under the percent as o f M a y 31, responsibility o f the Municipal Development Agency, which coordinated 2004, and the Urban the project and assisted local authorities with their preparation. Development Credit Concurrently, the contract management agency was incharge o f executing investment projects, in coordination with the Municipal Development with a rate o f 97 Agency and in accordance with the "delegated ownership agreement" percent in December appended to each contract with a municipality. This institutional 2004 (see box). arrangement proved effective because the project was almost entirely disbursed, reaching an execution rate o f 97 percent inDecember 2004. In the Nutritional Improvement Program (PRN), the component that falls under the responsibility of regional and departmental bodies i s executed more rapidly (with execution rates of 56 46 percent and 73 percent respectively) than the component under the responsibility o f the central government (45 percent). 5.11 It i s understood that the results given below are partial, but they still demonstrate that decentralizing the execution o f investmentprojects could lead to higher execution rates. But it should also be noted that execution o f a project such as the UrbanDevelopment Credit relies on ad hoc mechanisms (the Municipal Development Agency) created specifically to carry out these projects and thus act as a substitute for local government. These ad hoc structures, regardless o f how effective their performance may be, underscore the existing deficiencies in terms o f local government, which does not always have the technical and financial capacity to manage the projects. But it would be false to conclude that local governments have no capacity at all, since capacity varies greatly by region, and implementation o f the World Bank's PNIR and Social Fundprojects has demonstrated that this capacity exists for projects o f lesser scope and with an immediate impact on the community. 5.12 The Senegalese Government is aware o f the stakes because, even though early indications show that decentralized projects can be executed in a relatively satisfactory manner by local governments, the size o f the World Bank projects mentioned above remains limited in relation to the full dimension o f the B C I (for example, the Urban Development Credit i s equivalent to roughly 2 percent o f the total B C I on an annual basis). This i s why the government rightly anticipates prudent and gradual decentralizationofBCIexecutionover the coming years. C. RECOMMENDATIONS 5.13 Analysis o f the execution rates associated with investment projects reveals several limitations o f the current system in Senegal. First o f all, it i s clear that the information i s only partial and, worse still, i s only available very late. The information provided by DDI i s not available on a detailed level until more than a year after the accounts are closed. It i s thus imperative that this process be accelerated within the Ministry o f Economy and Finance. Such an effort i s necessary because the timely availability o f information on execution rates would help to better identify projects that are sources o f delays and their bottlenecks. This i s our first recommendation. 5.14 The second recommendation supports a recent initiative o f the Senegalese Government which involves developing a data base, project by project, to monitor in real time the execution o f priority PRSP projects. This initiative has begun to be developed within the Ministry o f Economy and Finance, at the instigation o f the unit incharge o fpoverty reduction, but it needs to be expanded to include a sufficient number o f projects. Over time, the goal is to deliver an on- line system that could be managed in real time and which would permit any partner to freely access information on project execution in a transparent fashion. Finalization and implementation o f this system are thus urgently needed and should constitute a priority for the Senegalese Government in view o f the results o f the preceding chapter, which highlighted the lack o f coverage o fPRSP objectives inthe B C I as executed. 47 5.15 The third recommendation focuses on the necessity o f trying to accelerate the execution o f investment projects, particularly in line with two characteristics that emerged as determinants inour statistical analysis. Efforts shouldbeconcentrated onprojects that: 0 Take place in the social and infrastructure sectors where execution appears to be systematically slower than inother sectors. There i s a need to buildthe capacities o f the relevant ministriesand promote better coordination at both the central and local levels o f government. 0 Are financed by external resources (particularly in credit form) and by end users, suggesting greater harmonization between the Senegalese Government and donors (see details inthe next chapter). 5.16 Finally, our last recommendation focuses on the level o f centralization o f investment project execution, which plays a determining role in Senegal. The results obtained from a limited sample o f World Bank projects indicate that decentralization o f the process could be promising for certain projects and for certain regions. However, the good results that were obtained can be partly explained by the recourse to ad hoc institutional mechanisms that offset the current deficiencies o f local governments, which have only limited capacity to implement investment projects o f large scope. The issue that arises for the government is how to determine the speed at which it expects to decentralize the BCI, and how to ensure the success o f this undertaking without jeopardizing the sustainability o f the system. The following chapter will provide an initial response to this question when we review the recent B C I decentralization initiative proposedby the SenegaleseGovernment. 48 6. INVESTMENTBUDGETMANAGEMENTSYSTEM 6.1 We have so far attempted to identify and address the issues o f economic policy which the Senegalese Government needs to tackle inorder to increase the efficiency o f public investments. This overview has led us to examine the composition o f the investment, identification o f recurrent costs, the relationship between public and private investment, the adequacy o f budget planning in relation to the strategic objectives o f the PRSP, and the factors underlying the variations ininvestment project execution rates. We will no longer dwell on why these decisions were made, but instead take a look at how to implement them through a rational and efficient management system. 6.2 The investment budget management system in Senegal i s part o f the general public expenditure system. It has both the virtues and the vices o f this system. On the positive side, there i s an iterative process between the Ministry o f Economy and Finance and the sectoral ministries, allowing officials to match investment expenditures to macroeconomic objectives. On the negative side, the preparation process appears to be excessively centralized, with no real project analysis performed ex ante by the departments involved. This gap i s particularly evident inthe case o fprojects financed with internal hnds and is a root cause of the strategic problems raised inthe preceding chapters, such as the limited capacity o f Senegalese officials to judge the sectoral allocation o f resources and connect investment expenditures with recurrent costs. 6.3 This chapter will first attempt to describe the required procedures for preparing the investment budget, which will help to highlight the deficiencies o f the current system. Next, special attention will be given to the two main procedures used to execute investment budget expenditures in Senegal, including the procurement procedures. To correct some o f the problems identified, the government has undertaken two recent initiatives which will be reviewed and analyzed. In conclusion, a number o f recommendations will be proposed for improving the transparency and quality o f the investment expenditure management system in Senegal. A. PREPARATIONOF THE GOVERNMENT'S INVESTMENTBUDGET 6.4 Preparationo f the investment budget i s part o f the general process o f budget preparation, which i s coordinated by the Budget D i r e ~ t o r a t eand ~ ~which follows the time frame shown inthe figure below (Figure 6.1). Beginning in April, the sectoral ministries prepare their annual investment expenditure proposals, which they then present to the Directorate o f Economic and Financial Cooperation (DCEF) o f the Ministry o f Economy and Finance. Budget conferences are organized in August in order to finalize the list o f projects by ministry. Finally, guidelines for the Three-Year Public Investment Program (PTIP) are developed from the project proposals 36The Budget Directorate has a staff o f 83, including 15 Level A managers and 12 Level B managers. Of this total, 44 employees work at the nine regional finance control offices, located in the administrative regions. The budget revenues and expendituresdivisions each have six employees. 49 presented by the technical ministries to DCEF under the national planning system. DCEF takes the first year o fthe PTIP to serve as Senegal's Consolidated InvestmentBudget (BCI). Figure 6.1: Time frame for preparing the budget law Vote on budget law I Start the work ofpreparingthe / budget law (revenues) Submit the draft budget law to the National Assembly Memo on the government's economic Adoption ofthe draft /November budget law by the Council of Ministers October Finish arbitrage (MEF-PM) May j > : L ; p F w who draw up alist ofpriority Inclusionin 1 1 PTIF' and BCI Start expenditure arbitrage ( D W August Set projections of Budgetconferences July receipts Sectoral ministries and DCEF Assessmentofprojects 6.5 From a practical perspective, the sectoral ministries' reduced capacity to identify, assess, and monitor public investment projects means that they are unable to develop accurate projections. These ministries have not called upon other entities, nor built in-house capacity, to perform the duties o f planning offices which were part o f their structure until the early 1990s. Furthermore, ministry departments are little involved in this work o f preparing investment projects, despite its importance for total mastery of their area o f activity. The technical shortcomings and the weakness o f both human and financial resources cry out at many o f the ministries, where no analyses (or only very incomplete analyses) are performed to determine the economic and/or social rate o freturn o f investment projects that will be submitted to DCEF. 6.6 DCEF itself i s not adequately equipped to carry out or support properly this work of project identification and monitoring/evaluation. It has a staff o f 34, including 24 operational manager^.^' Its budget programming capacities are thus weak. 37DCEFhas a staffo f34, including 24 operationalmanagers. 50 6.7 It is also recommended that coordination and collaboration be strengthened between the sectoral ministries and DCEF. The dialogue i s not always optimal and should be spread over the entire preparation period. Inthis regard, a greater role in this dialogue should be given to local governments, which cannot presently intervene in B C I preparation except through the sectoral ministries. The role o f local governments inpreparing the investment budget is heightenedat a time when the government anticipates deconcentrating BCIexecution at the regional level. 6.8 Apart from capacity-building and coordination, preparation o f the investment budget suffers from four major weaknesses which should command the fill attention o f Senegalese officials. 6.9 The first weakness concerns the lack o f any systematic analysis o f the optimal composition o f the investment budgeta3*This weakness i s mainly apparent inthe infrastructure, administrative equipment, and social service sectors, which represent a preponderant share o f the investment budget but have no evaluation systems for selecting the most viable projects. Inthis regard, there is an urgentneed to: 0 Set inplace a better system for selecting projects, first o f all at the relevant ministries but also at DCEF. 0 Establish systematic links betweenbudget preparation, on the one hand, and the PAP and the PRSP objectives, on the other. 6.10 The second weakness i s that, similar to other budget expenditures, the projections o f investment expenditures follow a logic o f resources, not o f results. Departments still pay greatest attention to the volume o f budget appropriations listed in the government budget, rather than to the objectives of these appropriations. This approach to budget preparation entails a certain rigidity that is not conducive to modern and dynamic management o f government finance; it does not encourage a critical look at budget appropriations granted earlier, but tends instead to result in the nearly automatic renewal o f these appropriations. As a whole, this system hinders the development o f initiatives to reform the technical ministries, and can even constitute an obstacle to the formulation o f sector policies. Officials have begun to address this problem by instituting budget programming in a few key ministries (Health, Education, Justice, and Environment), which should help better define the objectives to be pur~ued.~'To consolidate these efforts, however, there i s a need to address the constraints imposed by the current system o f financial management because, without a more thorough reform o f this system, the benefits o f preparing "budgets by objective" will remain very limited. 6.11 The third weakness lies in the forecasting inadequacies o f the PTP, which provides limitedand relatively unreliable information interms o f medium-term projections (see Chapter 4 for greater detail). It is also clear that the links with the priority objectives o f the PRSP, as 38 This weakness, rooted in inaccuracies in assessing the costs o f these projects, very often results in the underestimation, or more typically the overestimation, of these projects. No data on these projects are analyzed and taken into consideration when assessing future projects. There are practically no data on the unit costs o f certain expenditures, which would have facilitated the financial programming o f the projects. 39With the exception o f a few ministries, including the Ministries o f Health and Education, which prepare "budgets by objective." But internalization o f this new process by the financial departments o f these ministries has not yet been achieved. 51 established in the PAP, are not systematically taken into account and, in particular, are not properly assessed inthe preparation and programming o f the investment budget. Inaddition, the PTIP does not give any information on allocation o f the investmentby region, although regional development i s one o f the pillars o f government strategy. This shortcoming hinders the monitoring and evaluation o f the investment program and, in general, o f the PRSP as a whole. Several technical suggestions were proposed inChapter 4 for correcting this deficiency. 6.12 The fourth and final weakness stems from the lack o f a link between capital expenditures and the recurrent costs that are sure to emerge over time. The investment budget should be prepared by highlighting the quantitative implications in terms o f maintenance and recurrent costs which the public sector will need to cover incoming years. From an institutional perspective, international experience suggests that the link between capital expenditures and recurrent costs should be managed in a relatively centralized manner within DGFe4'Today, in fact, in most industrialized countries the responsibility has fallen to a single department within the Ministry o f Finance and Economy that i s charged with identifyingand analyzing investment expenditures and recurrent costs. Such unification i s useful because it allows officials to effectively link these two types o f expenditures in budget estimates. It is not advisable to separate estimates o f capital expenditures and recurrent costs because this approach generally results in overestimation o f the capital expenditures, which are perceived as more "worthy," to the detriment o f operating costs. Separating these expenditures also leads to administrative complications and difficult coordination within government services. If these two types of expenditures are determined jointly during the analytic process o f budget preparation, it i s imperative that they be clearly distinguished in the economic presentation o f the budget, for the reasons given inChapter 3. B. INVESTMENT EXPENDITUREEXECUTIONPROCEDURES 6.13 Execution o f these investment expenditures does not follow one single procedure, as noted inthe recent report assessingthe management o f government finance in Senegal (CFAA - June 2003), but rather two procedures: investment expenditures financed with national resources follow the procedure mandated by the General Rules o f Public Accounting, which constitute common law for the execution o f budget expenditures; meanwhile, budget expenditures financed with external resources follow the rules established by the financing agreements concluded between Senegal and its technical and financial partners. This duality o fprocedures for executing budget expenditures once again makes implementation o f the government budget seem complex. This complexity is, moreover, heightened by the procedural features specific to each external financing agreement, and it causes difficulties in the planning and execution o f expenditures, which remain incomplete, as shown in the preceding chapters. Finally, the complexity o f the procedures for preparing and executing the investment budget contrasts with the nearly total absence o f ex post controls, except for those performed by donors. 40For greater detail, see Michael Spackman, "Multi-Year Perspective in Budgeting and Public Investment Planning," OECD, April 24-26, 2002. 52 Box 6.1: Government agencies and specializedfunds The Senegalese Government transfers a significant part of the execution of its investment budget to a number of autonomous government agencies and specialized funds. The former receive budget transfers but enjoy relative management autonomy, the legal nature o f which remains imprecise inthe statutes. The principal agencies are the State Property Construction and Rehabilitation Program (PCRPE) agency, the Investment and Major Works Promotion Agency (APIX), and the Road Works Agency (ATR). In 2005, the PCRPE agency received a government transfer inthe amount o f CFAF 46 billion and has a budget of roughly CFAF 90 billion. Also in2005, a new agency was created to manage funds related to organizing the Islamic Con The specialized funds do not generally have the status of legalentities, but the inbank accounts distinct from those of the Treasury. For investment credits, the 2005 budget lists the following funds: ------Roadfund Local government equipment fund Water supply fund Rural collateral fbnd Rural improvement fund Rural disaster fund - Nationalsolidarity fund -- Numerical solidarity fund Fundfor the advancement ofyouth - Urbantransport development fund - National fund for employment action It shouldbe notedthat the roadfund, the localgovernment procurementfund,andthe water supply fundare budge line items that retain the names given to the special Treasury accounts. These lines are not executed through b a d accounts. Reading the budget law is not sufficient to grasp incomprehensive fashion the total amount o fpublic investment fo which execution competence is transferred to these agencies and funds. It i s however clear that the number o f sucl agencies has increased inrecent years, and they no doubt account for a growing proportion o f B C I execution. FO example, the strong increase in infrastructure investments observed in 2005 should serve to expand the role o PCRPE andthe road fund inthe execution o f such projects. Apart from the lack o f information on the amount o f investment executed by these agencies and specialized funds there i s also a lack o f transparency concerning their procedures and the implementation o f projects for which the: are responsible. Their procurement procedures escape the supervision o f CNCA, and independent ex post control are rare and often incomplete. A recent evaluation o f ATR raises serious questions about its management, interm o fboth the execution andthe monitoring o fprojects for which this agency holds responsibility. On this basis, it is recommended that Senegalese officials focus some o f their attention on achieving greate transparency conceming the amounts and the procedures usedby the autonomous agencies and specialized funds ti execute the investments for which they are responsible. This effort should begin with accurate information on th amounts and projects so transferred in the budget law, as well as on the supervision and monitoring of thei operations. 6.14 This section reviews the efforts recently carried out under the auspices o f the Senegalese Government and the World Bank with the involvement o f technical and financial partners. For greater detail, the reader should refer to the abovementioned CFAA and the report assessing the government procurement system in Senegal (CPAR - 2003). 53 6.15 The execution o f investment budget expenditures in Senegal thus follows two distinct procedures, although it should be noted that a growing portion o f investment expenditures are executed by autonomous agencies and specialized funds which are not governed by the budget procedures described below, pointing to a series o fproblems that are described inBox 6.1. B.1 Procedurefor investmentexpendituresfinancedwith nationalresources 6.16 These expenditures are executed in accordance with the same rules as those that apply to the government's o erating expenses, which adhere to Directive 06/97 adopted December 16, 1997 by WAEMU! This common law o f public spending follows a procedure consisting of four phases: an administrative phase including expenditure commitment, validation, and payment order issue, and an accounting phase for actual payment. The responsibilities o f account managers and financial operations auditors (COF) come into play during the first phase, while public accountants are responsible for execution o f the second phase. 6.17 The table 6.1 below describes the circuit followed by investment expenditures financed with national resources. 6.18 This expenditure circuit reveals two major problems: the length o f its administrative phase and the absence o f independentcontrols during its accounting phase. 6.19 Ithas been noted that the lengtho fthe administrative phase results ina loss of efficiency and very slow execution of investment expenditures. The tests conducted for the studies cited above indicated that execution of investment expenditures takes an average o f four months. This situation i s partly due to the many redundancies identified, particularly in the back-and-forth betweenthe departments o f DBDDIand COF. 6.20 During the accounting phase, problems o f transparency and reliability arise in the absence of independent controls and timely information. Similar to extemal controls, which currently lack efficiency in Senegal, intemal management controls at the Treasury remain nonexistent, and there are no audits of the operations o f Treasury departments in the absence o f an entity dedicated to this mission. 6.21 Furthermore, the Senegalese legislative and regulatory mechanism for paying government debts, except for those resulting from the execution o f government contracts, sets no term o f payment for government departments and does not provide for the payment o f interest on overdue payments. This situation is hardly conducive to rapid execution o f investment expenditures financed with national resources. 4'This directive was adapted to Senegalese law by Decree 2002-101 o f March 13, 2003. 54 Table 6.1: Expenditurecircuitfor investmentexpendituresfinancedwith nationalresources STAKEHOLDERS ACTIONS I-Administrativephase A. COMMITMENT CREDIT OFFICER DATA ROOM 2. Retneves the BE data in the software, thus reserving budget appropnabons; sends the forms to the COF auditor. COF AUDITOR 3. Verifies the legality of the expenditure prior to its commitment; sends the forms to the COF manager. COF MANAGER 4. Signs the BEs and sends the forms to the IssuingAgent (DGF). 5. Confirms the reservationof appropriations inthe software. ISSUING AGENT 6. Issues the Titles of Credit and Certification(TCCs). (DGF) 7. Sends the TCCs to COF for distributionto suppliers andcredit officers. B. VAI,IDATION CREDIT OFFICER DATA ROOM COF AUDITOR COF MANAGER DBDDIMAILROOM OVERSIGHT DIVISION,DB or OFF. 13. Verifies the legality andconformity of the forms and sends them to the director. OF INTERNAL REGULATIONS/DDI Director of Budget 14. Signs the BEs, the Titles of Credit and Certification to confirm the validation and or DDI payabilityof the debt. 15. Issues the visa of conformity through the software to permit issuance of vouchers and pay slips. ISSUING AGENT 16.Issues vouchers andpay slips and sends the forms to DB or DDI. DBDDI 17. Signs the vouchers andpay slips. DBDDIMAILROOM Sends the forms to COF to endorsethe vouchers. cur DBDDIMAILROOM 11- Accounting phase EXPENDITUREOFFICEof DGCPT Recordsthe payment forms and checks offthe vouchers. (MAIL ROOM) Refers the matterto PGTand sends the forms to the ExpenditureOffice for verification. DGCPT EXPEKDITURE OFFICE 19. Venfies the forms and proposesapprovalor rejection, then sends the forms to PGT. (VERIFICATION) PGT 20. Endorses the vouchers or signs a note of rejectionand sends the payment forms to the ExpenditureOffice. PGT EXPENDITURE OFFICE 21. Records (i)rejections, which it sends back to the account manager, (ii)the endorsed vouchers, which it sends to the PGT AccountingOffice. PGT ACCOUNTING OFFICE 22. Takes charge of the vouchers and verifies any objections to settlement (cf. VAT, Third Holder Notice). 23. Issues the payment order (OP) andcredit notice(AC) to execute the transfer or issue a delivery slip bound for RGT, then sends the OPs and ACs to BCEAO to execute the bank transfer. 24. Keeps a recordof accounts. 42Inaccordance withthe procurement rules analyzedbythe CPAR. 55 6.22 Lastly, so far as these investment expenditures are concerned, there is remarkably no description of their execution, nor monitoring, much less any specific a posteriori evaluation. This absence o f oversight i s inconsistent with effective implementation o f stock accounting, which constitutes a major weakness o f the investmentbudget management system. No periodic audit and no report on the execution o f these expenditures is issued or i s made available to national officials. This situation ultimately undermines the transparency o f the end uses o f these expenditures which, it should be repeated, often involve large sums. The lack o f communication among officials on this subject i s a serious shortcoming that serves as an obstacle to the transparency o f government funds management. B.2 Investmentexpendituresfinancedwith externalresources 6.23 Investment expenditures financed with external resources have accounted since 2001 for an average o f 25 percent o f the government's total budget expenditures and roughly 60 percent o f its investmentbudget. External financing i s thus a preponderant element in the financing o f development operations in Senegal; this situation i s not specific to this country inasmuch as Senegal shares these circumstances with all Sub-Sahelian countries in which external loans and grants received still represent the bulk o f all financing o f investment expenditure^.^^ 6.24 It should be noted first o f all that Senegal's failure to comply with the WAEMU directive concerning the execution o f investment expenditures financed with external resources i s a characteristic common to all WAEMU members; thus, the case o f Senegal i s not isolated. In view o f the progress o f discussions undertaken in this area at the national level (see below the initiative taken by Senegalese officials), the Senegalese Government has an opportunity to advance proposals at this regional body. 6.25 In Senegal, investment expenditures financed with external resources are executed in accordance with an extra budgetary procedure, i.e. one that i s not integrated into the expenditure circuit described above. The managers o f investment projects are the key element o f the procedure, under the supervision o f the Debt and Investment Directorate (DDI) which intervenes in the capacity of delegated account manager as well as payer from project accounts receiving advances o f funds from certain donors. 6.26 These donors, similar to the Senegalese Government, sought a procedure that permits more rapid disbursement o f expenditures. The result i s a simplified expenditure circuit: 0 payments are made directly from a special bank account held by DDI, without recording the budget expenditures, even though they are listed in the budget law, nor the account transactions inthe government's general accounts kept by Treasury officials. 0 certain expenditures are executed from bank accounts held by project managers, which increases the risks associated with investment expenditure management. 0 DDI oversight focuses on the legality and payability o f a debt and is performed on the basis o f expenditure records submitted by the project managers; on-site oversight o f these 43 As much as 90 percent o f the financing of investment expenditures in Niger, where technical and financial partners account for 60 percent o f the financing o f government finance operations. 56 expenditures does not fall under the responsibility o f DDI, but falls instead to oversight bodies such as the General Inspectorate o f State, the General Inspectorate o f Finances, and the Audit Office. accounting records for these investment expenditures are kept by the project managers and DDI; concomitantly, DDI handles financial monitoring o f drawdowns o f hnds granted by certain donors. 6.27 These investment expenditures financed with external resources are then retranscribed each quarter in a statistical status report prepared by DDI, which is then submitted specifically to the Directorate-General o f Government Accounting and the Treasury (DGCPT), which in tum incorporates it into the Table o f Government Finance Operations. 6.28 This execution procedure poses certain problems in relation to national and regional (WAEMU) budgetary law. It does not in fact adhere to the principle o f separation between the account manager and the accountant since DDI handles oversight o f expenditure commitments and payment orders, on the one hand, andpayment o f the same expenditures, on the other. This confusion of functions without efficient external controls creates an embryonic risk for the execution o f this category o f expenditure. It should encourage Senegalese officials to review the respective roles o f DDI and the Treasury, and the Treasury should in time become responsible for all payments made in the budget, including investment expenditures financed with external funds. 6.29 In addition, the analyses recently performed have detected a number of problems in the execution o f investment expenditures financed with external resources: oversight o f the materiality o f the expenditure i s not systematically performed before the payment order i s issued, making knowledge o f the actual effectiveness of the expenditure uncertain; however, donors have set in place their own system of oversight and monitoring inan effort to circumvent this deficiency. information on payments made directly by donors i s not always known to DDI on a timelybasis, thus preventingthe accurate monitoring ofpublic debt. the consolidated balance o f government accounts does not reflect execution o f these expenditures. the accounting records of the project managers and DDIare not periodically reconciled, which raises the issue o f the reliability o ftheir accounts. payments made from the projects' special accounts are not entered in Treasury records, so DGCPT i s unable to produce comprehensive information on the status o f the public treasury. DDI, which functions as both account manager and accountant, is also responsible for monitoring the audit reports o f projects financed with external resources. This situation does not guarantee independent and reliable oversight o f projects financed with external resources. 57 0 expenditures financed with external resources are not connected to the integrated management information system for public expenditures. 0 stock accounting o f expenditures financed with external resources i s not regularly kept and monitored, particularly at the closing o fprojects. 6.30 Finally, the reliability o f the statistical information provided by DDI is fragile due to manual data collection, as DDI's technical capacities are deficient, particularly incomputing. As such, it i s strongly recommended that the SIGFIP information system, which has begun to function with promising results, be expanded to include investment expenditures financed with external funds inclose collaboration with the donor community. c. GOVERNMENT PROCUREMENTPROCEDURES 6.31 Carrying out public investments generally requires access to goods and services supplied by the private sector. Government contracts are thus awarded inaccordance with the procedures set forth in the procurement code. In practice, however, these procedures are often a source o f delays, thus explaining the relatively low execution rates for investment expenditures, and/or they are circumvented, which reduces the quality o fpublic investments. 6.32 InSenegal, there is no lack o fassessmentsregardingthe procurementprocedures, starting with the recent CPAR and the AAP report on government finance management, so we can be relatively brief on this subject. These reports concluded that Senegal was still very far from the best international practices, even inAfrica (see Table 6.2). Thus, despite some progress, Senegal falls roughly at the same level as countries such as Kenya, Mali, and Burkina, but behind Tanzania, Ghana, and Uganda. 6.33 Over the last two years, the Senegalese Government has reacted by adopting an action plan pursuant to the Interministerial Council o f July 22, 2003 and based on the recent CPAR study, prepared jointly with the World Bank,44 which emphasizes: (i) revision o f the institutional framework, particularly by strengthening the mechanisms o f independent advice and recourse; (ii)implementation o f procedures aimed at increasing capacity and supervision at the central and local levels; and (iii)adoption o f the new procurement code and revision o f the Code o f Government Obligations before September 2005. Next, to supplement these legal and institutional measures which will take time to influence mentalities and behaviors, the government imposed, inagreement with the World Bank and IMF, the condition that less than 20 percent o f the contracts submitted for the approval o f the National Commission on Government Contracts (CNCA) be approved under a scheme o f direct contracting. In early 2005, the government also imposed rigorous planning o f procurement at six spending ministries, which should enable CNCA to see to less frequent use o f the waivers granted by the procurement code. It is urgent and fundamentally important that the SenegaleseGovernment continue its efforts and complete the actions described above as rapidly as possible. 44Inthis regard, the recent aide-mkmoire of the World Bank (March 2005) notes the absence of a recent evaluation report on implementation of the action plan derived from the CPAR and that the composition of the steering committee createdto supervise implementationof its action plan has not yet been finalized. 58 Table 6.2: Qualityof procurement proceduresin Africa (From 1to 6, with amaximumscore of 1) Country 06/31/01 12/31/01 06/30/02 06/30/03 07/15/04 Angola 4.9 4.8 4.8 4.4 4.3 Benin 4.1 4.1 4.74 4.5 4.1 Botswana NA NA NA NA NA BurkinaFaso 4.7 4.6 4.6 4.1 3.5 Burundi 5.0 4.3 4.3 4.25 4.25 Cameroon 4.6 4.6 4.4 4.3 4.2 Cape Verde NA NA NA NA 4.3 CAR NA NA NA NA NA Chad 4.2 3.9 3.9 3.8 3.7 Comoros NA NA NA NA NA Congo(Rep.) NA NA NA NA NA Congo(RDC) NA 4.9 4.9 4.8 4.7 Cate d'Ivoire 4.8 4.8 3.8 3.4 4.0 Eq. Guinea NA NA NA NA NA Eritrea NA 4.6 4.6 5.0 5.0 Ethiopia 4.2 4.2 4.2 4.2 3.9 Gabon NA NA NA NA NA Gambia 4.7 4.6 4.5 4.45 4.3 Ghana 3.8 3.3 3.3 3.O 2.6 Guinea 4.2 4.2 4.2 3.8 4.1 Guinea-Bissau 4.8 4.0 4.0 3.6 3.6 Kenya 3.4 3.4 3.4 3.1 3.5 Lesotho 4.4 4.4 4.4 4.4 4.0 Liberia NA NA NA NA 4.7 Madagascar 3.6 3.6 3.6 3.2 3.0 Malawi 4.0 4.0 4.0 3.9 3 Mali 3.4 3.4 3.4 3.5 3.4 Mauritania 3.5 3.5 3.5 3.5 3.4 Mozambique 4.0 4.0 3.96 3.9 3.8 Namibia NA NA NA NA NA Niger 3.7 3.7 3.7 3.6 3.7 Nigeria 4.0 3.8 3.8 3.8 3.7 Rwanda 3.4 4.0 4.0 3.95 3.95 Sao Tome & Principe NA NA NA NA NA Senegal 3.8 4.0 4.0 3.5 3.5 Seychelles NA NA NA NA NA Sierra Leone NA 4.9 4.96 4.9 4.3 Somalia NA NA NA NA NA SouthAfrica NA NA 2.8 2.1 2.1 Sudan NA NA NA NA NA Swaziland NA NA NA NA NA Tanzania 3.4 3.4 3.4 3.0 2.8 Togo 5.o 5.0 5 4.9 4.9 Uganda 3.3 3.3 3.3 2.8 2.4 Zambia 3.7 3.7 3.7 3.7 3.7 Zimbabwe 4.1 4.1 4.1 4.6 4.6 Source: WorldBank (NA=not available) 59 6.34 These actions are welcome and their implementation i s indispensable because they contribute to promoting the use o f competitive and transparent procedures. However, they remain partial and should be completed by adopting a series o f supplementary actions: 0 First, it is essential that officials be able to identify the contracts that do not go through CNCA approval, the total amount o f which appears to exceed 60 percent o f expenditures included inthe B C I and the expenditures ingoods and services. At this stage, information i s lacking to assess the scope o f the contracts whose reduced amounts do not require going through CNCA and those that escape CNCA oversight because they go through alternative channels, for example implementing agencies.45Inthis regard, officials should make sure that contracts that pass through the latter channels are first submitted for CNCA approval. 0 Second, special attention should be given to efforts to limit the use o f the emergency clause which provides a waiver to competitive and transparent procurement procedures. Although this clause i s permittedby law, its use should remain exceptional and should be submitted to CNCA approval in accordance with the contents o f Article 76.6 o f the procurement code, which shouldbe limitedto natural disasters alone. D. RECENT INITIATIVES AIMED AT IMPROVINGTHE MANAGEMENT OF INVESTMENTEXPENDITURES 6.35 Aware o f the necessity o f increasing the efficiency o f public investment expenditures, Senegalese officials have recently undertaken a variety o f initiatives, stemming for the most part from the recommendations set forth inthe CFAA and CPAR action plans, approved on July 22, 2003 by an Interministerial Council devoted to budget and finance reforms supported by technical and financial partners. This review will examine two o f these initiatives concerning: (i) adoption o f a new accounting and financial framework for retranscribing public investment expenditures; and (ii) decentralization o f the management o f these expenditures in conjunction with the more general political process o fdecentralization. D.l New accountingandfinancialframework for investmentexpenditures 6.36 Designedin 1996 to support the political process o f decentralization, this accounting and financial framework was not introduced. The Senegalese Government therefore proceeded to update it, so as to take into account the regulatory changes that had taken place over the last eight years, particularly the directives adopted in 1997-1998 by WAEMU and adapted to Senegalese law in 2001-2003. This Senegalese initiative now appears to be particularly welcome because it seeks to correct the existing shortcomings in the statutes applicable to public investment expenditures. 6.37 Adoption o f this accounting and financial framework would serve to harmonize the various procedures resulting from implementation o f the Senegalese Accounting and Budget Nomenclature Plan, which has been fully operational since January 1, 2004. Setting accounting and financial standards should improve the monitoring o f public investment expenditures and ~ 45 Ithas infact been found that only one quarter o f the projects financed with external resources go through CNCA. 60 provide the government with a homogeneous frame o f reference that corrects the shortcomings o f the current system for monitoring these expenditures. 6.38 However, this new accounting and financial framework raises a few fundamental issues: 0 The proposed system for managingpublic investment expenditures remains concentrated. 0 The project management scheme i s incomplete. 0 This system institutes co-responsibility between the project director and the revolving fund manager concerning funds management, which does not clarify their respective responsibilities and creates a confusion o f functions. 6.39 First o f all, this framework retains a concentrated system for managing public investments. It thus sends a signal that contradicts the government's initiative aimed at deconcentrating the management o f public investment expenditures (see the analysis o f this initiative below). Clearly there i s a need for greater overall consistency in the strategy followed by the Senegalese Government, a strategy that can only move in the direction o f the deconcentration o f powers toward sectoral officials, or even deconcentrated administrative entities, which lie closer not only to actual execution o fpublic expenditures but also to citizens. 6.40 By moving toward this deconcentration o f public expenditures, the Senegalese Government would fall in line with the measure recently taken by the Minister o f Finance aimed at placing representatives o f Financial Operations Oversight (COF) within the sectoral ministries; this welcome action, which represents just the first step that should soon be followed by other steps toward deconcentration o f the administrative phase o f the expenditure circuit in the sectoral ministries, i s a way to avoid the systematic referral o f expenditure commitment and payment order matters to the Ministryo f Finance. This initiative thus effectively helps to reduce the term o f execution o fbudget expenditures. 6.41 In addition, this accounting and financial framework ignores the a priori financial oversight o f commitments and payment orders made by the project director, except in the case Of: 0 The allocation of budget appropriations andplacement o f credits. Certain expenditures regulated by the govemment procurement code in view o f their nature and amount.46 0 The refinancing o frevolving funds. 0 The mobilization o f externalresources. 6.42 On the other hand - and this is a positive feature - the framework favors a posteriori oversight: 46Without these expendituresevenbeingdefined. 61 0 o f expenditures paid from revolving funds: this oversight i s provided by the national oversight entities; 0 o f the project's financial execution: this system makes it appear that the function o f a posteriori financial oversight i s shared between national and external stakeholders. It i s to be feared that this sharing o f functions could lead in practice to the exclusion o f the former who will defer to the expertise o f the latter, which constitutes a demobilizing factor for the national oversight entities. 6.43 In addition, the project management scheme contained in this accounting and financial framework i s not described infull. The scheme should attempt to describe incomplete detail the chain o f responsibilities and controls. This would more clearly show the risk contained in the dual function o f account manager and accountant played by the project director, already left with no independent internal oversight (see above). A complete scheme should describe not only the management structure for public investment expenditures, but also the oversight and supervisory bodies. 6.44 Finally, a public expenditure accounting and financial framework cannot institute co- responsibility for funds maintenance between project directors and revolving fund managers. Such shared responsibility i s inpractice not very operational and it brings the risk o f dilutingthis responsibility. It contains the seeds for weakened financial management, undermining the rigor requiredfor protectingpublic funds. D.2 Decentralizationof publicinvestmentexpendituremanagement 6.45 Once the CFAA and CPAR action plans were adopted in July 2003, the Senegalese Government set up an ad hoc committee in charge o f the study on decentralization o f the management o f Consolidated Investment Budget (BCI) resources. This committee, which welcomed the contributions o f government institutions and entities as well as representatives o f associations of locally elected officials, produced a report which was finalized in December 2004.47 After presenting the context o f decentralization in Senegal and its assessment o f currently centralized B C I execution, the committee proposes a reform aimed at decentralizing public investment expenditures at the local government level. This initiative falls within the framework o f the national poverty reduction strategy. 6.46 While the committee does not hide the advantages currently provided by centralized B C I management, it clearly highlights the disadvantages o f centralized management o f investment expenditures and, inparticular, the negative consequences for expenditure beneficiaries resulting from slow expenditure execution. In its report, the committee also emphasizes the obstructive nature o f the rule o f a single principal account manager (the Minister o f Finance), which in Senegal concentrates under this official's authority "full powers for managing the government budget." 6.47 Based on this assessment, the committee proposes that the political and administrative decentralization instituted in 1996 be accompanied by decentralization o f the financial 47 Decentralizationof BCI execution, December 2004. 62 management o f the BCI. It proposes that this decentralizationbe carried out by transferring the responsibility for investment expenditure management to local governments in order to achieve more efficient and transparent management. 6.48 Although it i s based on clearly articulated, general and specific objectives, the proposed reform suffers from a lack o f resolution o f preliminary issues which, in some cases, the committee itself identifiedY4*and, in particular, a lack o f precision on certain measures to be implemented, such as building the technical capacities o f deconcentrated government entities, which requires short- and medium-term training actions, and upgrading the human resources o f local governments. 6.49 Inview o fthe measures that shouldprecede this reform and the actions to be undertaken as soon as possible to establish the right conditions for its success, the time frame for its implementation, compressed to a period o f under two years, seems unrealistic, although the idea o f starting with two pilot projects in the health and education sectors appears reasonable to us. The document gives few indications o f the government's intentions in the two pilot projects in education and health. There i s no precise schedule, no information on the number o f projects or the amounts for which execution would be decentralized, and no identification of the regions or local governments that would participate inthese pilot projects. 6.50 Inasmuch as the B C I decentralization proposal is in line with the policy objectives pursued by the government, it is recommended that, as early as possible, the preliminary regulatory and legislative steps be taken one by one, before proceeding with actual decentralization o f investment expenditure management. For this reason, this initiative logically falls within the scope o f a medium-term objective, allowing verification at each stage that the process o f internalizing the management o f these expenditures has been properly assimilated. In this regard, Senegal could draw upon the experience o f neighboring countries, such as Mali, which first deconcentrated public expenditure management at the sectoral level, then at the local level, before carefully proceeding with decentralization. 6.5 1 The initial challenge for the Senegalese Government concerns the central authorities, which must first move forward with changes inthe legislative framework that will allow them to transfer responsibility for B C I execution to local governments. Apart from changes in the statutes, it will next be necessary to ensure that the appropriations associated with decentralized projects are indeed transferred to the local authorities and that such transfers occur on time. In practice, these transfers will not be automatic, and current experience with FECL and FDD transfers does not justify great optimism. The delays are substantial, sometimes as long as six months, which will constitute an initial obstacle to the execution o f investment projects inwhich disbursement rates would be lowered, rather than increased, thus working against the stated objective. 6.52 Inthis regard, the issue of the level of decentralization has not really been addressed by the Senegalese Government: is it in fact appropriate to transfer the management o f investment expenditures listed inthe government budget directly to municipalities or first to regions? There 48 Such as adaptation o f the accounting and budget nomenclatures or adoption of a law empowering local governments to execute investment expenditures. 63 are presently 441 local governments in Senegal: 11regions, 110 municipalities (67 common law municipalities and 43 district municipalities), and 320 rural communities. A first step could be to "regionalize" the BCI; this was a major recommendation o f the CFAA. 6.53 In fact, with the past increases4' and especially the coming increases in official development assistance for local governments in Senegal, the elected officials o f local governments face growing responsibilities, and no entity is currently able to conduct a thorough analysis o f all existing expenditures and programs within a local government's area o f jurisdiction, much less an analysis o f the budget execution rates o f municipalities and particularly rural communities. There i s currently little information available on the relevance o f investment projects and their possible overlap with sectoral programs (water supply, transport, social infrastructure, neighborhood restructuring, social development, etc.) implemented at the central level by the relevant technical ministries which have their own channels o f programming and financing. In this regard, as suggested earlier, the participation o f local officials in B C I preparation should be enhanced so as to raise their awareness and involve them in the projects from the very start. Decentralization o f B C I execution will o f course raise the problem of recurrent costs, particularly maintenance, which investment projects will obviously entail. Who will then be responsible for these costs? If they remain the responsibility o f central officials, there i s a great risk o f creating a dichotomy in expenditure management. If they become the responsibility o f local officials, these officials don't have sufficient resources to ensure their financing. The problem thus becomes a larger issue concerning the system o f decentralization that the Senegalese Government wishes to set in place, So far, this strategy i s not totally clear and many questions remain. 6.54 Furthermore, the oversight and monitoring/evaluation o f expenditures, as currently performed at the decentralized level, are deficient at both the national (problems o f monitoring and analysis at the central level, problem o f a posteriori oversight by the Audit Office) and local levels (few instruments are sent to the relevant government entities and, as practiced, oversight o f the legality o f expenditures i s more like recording the expenditures than providing real budget oversight). 6.55 Surprisingly, the committee limits this decentralization to the internal resources o f the BCI, which excludes a very large majority o f B C I expenditures and does not lend itself to coherent project management." This partial approach seems to stem from the concern about not overly complicating the task o f local governments. Indeed, due attention must be given to the great variety o f intervention procedures followed by donors, who hesitate to practice budgetary aid or accept local community ownership in areas o f expertise defined by law. This de facto situation raises a few fundamental questions to which answers are not always forthcoming: What conditions need to be met for this situation (budgetary aid and community ownership) to be achieved? Must these communities possess experienced staff, satisfactory selection and allocation criteria, and transparent procedures for executing their expenditures? 49From a very small amount in 1995-1996 (CFAF 1 billion per year), this assistance has grown to some CFAF 30 billion per year. The volume o f financial flows resulting from decentralized cooperation remains very difficult to evaluate because currently the interventions are not always recorded. 50Within the same project, expenditures financed withinternalresources would then be managed ina decentralized manner while those financed with external resources would continue to be managed at the central level. 64 6.56 The risks o f total decentralization toward intraregional communities (municipalities, rural communities) thus appear to be the following: These communities lack the capacity to mobilize and analyze information on execution of their expenditures, which makes the current system o f budget programming and allocation risky and inefficient (how should decisions about priority expenditures rationally be made, how should allocations rationally be appropriated, etc.). 0 There is no real separation between decentralized personnel and deconcentrated per~onnel,~~and this, added to the shortage o f qualified staff, gives little credibility to the legality controls. 0 Locally elected officials are sometimes illiterate and lack training; monitoring and oversight by citizens i s virtually nonexistent, which means that local notables could seize power. 6.57 It thus emerges inthis situation that the process o f decentralizing investment expenditure management should go forth gradually, starting with satisfaction o f the prior conditions and regionalization o f the BCI, based on a time frame that is not quite so tight. This conclusion should not be interpreted as a brake on decentralization but rather as a nod to the existing constraints, which should push officials to initiate the process as rapidly as possible because it will necessarily beplayed out over time. E. RECOMMENDATIONS 6.58 In view of the investment expenditure procedures followed in other countries, those followed in Senegal can hardly be ranked among the best practices, but they are not among the least appropriate either. At the international level, the situation in Senegal cannot really be compared to the situation in countries such as Brazil, India, or even Jordan, which all have an established tradition o f investment expenditure planning and execution, and in which the resource allocation mechanism for capital equipment maintenance and rehabilitation runs smoothly, i s homogeneous, i s subject to proper controls, and is integrated into the government budget execution system. 6.59 The Senegalese system o f public investment expenditure planning and execution is more comparable to the system in countries such as Peru and Colombia where, even though there are sectoral approaches to investment expenditures in priority sectors and the Ministry o f Finance works in concert with the sectoral ministries to take expressed needs into account, significant deficiencies are found in the investment project monitoring/evaluation system, resulting in a low rate o f returnon investments. 6.60 InSenegal, the problems relating to public investmentexpenditures mainly center on the weaknesses o f the system for programming such expenditures, the deficiencies in project 5 'The CFAA demonstrated that staff in deconcentrated accounting positions inmunicipalities, including large cities such as Thihs and Dakar, were basically o f municipal origin. This could have an effect on the oversight o f municipal finance management. 65 coordination, the lack o f analyses on project sustainability, the risks incurred, and the cost- effectiveness o f the expenditures. This situation makes Senegal more like countries such as Ghana and Ethiopia where there i s a need to strengthen the mechanism for evaluating public investmentprojects and develop national capacity inthis areas5* 6.61 The principal recommendations derived from the analysis presented in this chapter are summarized below. Recommendations(preparationof the Povernment'sinvestmentbudget): 0 Build the capacities o f the technical ministries as regards the identification and preparation o f investment projects and programs. 0 Build the capacities o f DCEF as regards the planning and development o f investment projects and programs and the monitoringlevaluation o f these projects and programs, including project selection in the three following sectors: infrastructure, social sectors, and administrative equipment, which have witnessed strong increases inrecent years and now account for nearly halfo fthe BCI. 0 Improve the PTIP forecasting capacity and establish a systematic link with PRSP objectives, so as to take into account the implications in terms o f recurrent costs and regional distribution o fthe investment. Recommendations(execution of investmentexpendituresfinanced with nationalresources): 0 Simplify the expenditure circuit and clearly define the responsibilities of all actors involved inthe expenditurecircuit. 0 Set a term o f payment for government debts (excluding contracts) and provide for the payment o f interest on overdue payments. 0 Prepare annually a descriptive and evaluative report on investment expenditures financed with national resources and append it to the draft budget executionreport. 0 Increase the transparency o f investment project execution by autonomous agencies and specialized funds, the role o fwhich has rapidly expanded inrecent years. Recommendations(executionof investmentexpendituresfinanced with externalresources): 0 Improve and implement the new accounting and financial framework for investments and integrate into the framework a mechanism to deconcentrate these expenditures. 0 Strengthenthis new framework by: 52In Ethiopia, many project evaluations are conducted according to standard international practices. The World Bank requires, for example, a minimumrate of return on investments of 12 percent in Ethiopia. 66 k adopting a streamlined mechanism of intemal oversight of these expenditures at the commitment stage, to be accompanied by the following measures: (i)improve budgeting by expenditure line; (ii)tie into the SIGFIP information system; (iii) ensure greater independence o f project accountants in relation to project managers; and (iv) engage in discussions with BCEAO on opening special accounts on their books or take an inventory o f bank accounts at commercial banks with an eye to their rationalization; k strengthening the Audit Office to ensure reliable and independentextemal oversight o f investment projects; k redefining the role of DDI and clarifying its functions as account manager, accountant, and monitor o f extemal audits; P implementing a system for evaluatingpublic investmentprojects; k settinginplaceandmonitoringthe stock accountingofprojects; and k initiating at the WAEMU level a reform of the framework for executing investment expenditures financed with extemal resources. Recommendations(procurementprocedures): 0 Accelerate implementation o f the new institutions andrevise the procurement code inline with the CPAR action planapproved bythe government. 0 Finalize procurement plans in six ministries, covering at least 60 percent o f their investmentexpendituresand operating costs. 0 Limit the use o f waivers in direct contracting to under 20 percent o f the contracts approved by CNCA, with extra attention to the use o f the emergency clause which should inall cases besubmitted for prior approvalbyCNCA. 0 Identify the altemative channels at CNCA that are used for purposes o f procurement, and that account for more than half the B C I expenditures and expenditures in goods and services, and ensure that competitive and transparent procedures are followed. Recommendations(decentralizationof investmentexpendituremanagement): 0 Decentralize in tandem investment expenditures financed with intemal and extemal resources. 0 Clearly define and determine prior actions to be taken before decentralizing management o f the government's investment expenditures (with special attention to the legal, financial, and accounting framework). 0 Set a complete and realistic schedule for the measures to be implementedinthe short and mediumterm. 0 Initially, regionalize the consolidatedinvestment budget. 67 0 Proceed concomitantly with: > regionalizationo f information and coordination of donor support; > stronger legality controls and budget oversight of local governments (principally regions and municipalities); > establishment o f a new system of staff distribution; and > greater inspection o f territorial and local government entities. 68 ANNEX A: RECENTDEVELOPMENTS INGOVERNMENT FINANCE INSENEGAL 1. This annex summarizes: (i) budgetposition inSenegal in2003-2004; (ii) principal the the elements o f the 2005 budget law and the program to which agreement was reached with the IMF; and(iii) progress inimplementingthe CFANCPAR action plans.The description i s intentionally brief because detailed information can be found in documents recently produced by the World Bank and the International Monetary Fund, especially the document o f the first Poverty Reduction Support Credit (Report 28332-SN, November 12, 2004), the Public Expenditure Review (Report 29357-SN, December 28, 2004), and the report on the second PRGF review (EBS/05/28, February 16,2005). A. BUDGETEVOLUTION OF THE 2003-2004 2. Over the last two years, the budget position in Senegal has continued to be satisfactory and has remained in line with the objectives set under the program with the International Monetary Fund. While the overall budget deficit rose from 0.1 percent o f GDP in 2002 to 2.7 percent o f GDP in 2004, this deterioration is mainly due to an increase in spending for priority sectors financed with HIPC finds, the higher than expected costs o f structural reforms, and the additional spending brought on by the invasion o f migratory locusts and the infrastructure works inthe city of Thibs. As a result ofthe availability of external financing, the government's (net) debt toward the domestic financial market fell substantially in2004 (see Table 1). 3. Success inmaintaining the balance in the basic budget - which excludes structural costs and other temporary measures - at roughly CFAF 50 billion between 2003 and 2004 indicates that the government's budget position has remained stable. It should also be noted that officials managed to offset the additional spending in the city o f Thies by reducing a number o f non- priority expenditures in the amended law passed in September 2004. They also complied with the budgetpolicy convergence criteria established within WAEMU. 4. Total revenues increased from CFAF 726 billion in 2002 to CFAF 839 billion in 2004 even though the implementation o f several reforms aimed at driving down rates but expanding the tax base might have been expected to result in stagnation, at least in the short term. The Senegalese Government has indeed lowered the tax rate for companies from 35 percent to 33 percent, adopted a new investment code which streamlines the incentives offered to enterprises, and simplified the property tax and trade licenses. This increase in revenues reflects the good performance o f the tax authorities, especially the unit in charge o f large enterprises, and improved collection o f the VAT. 69 Table 1: Table of GovernmentFinanceOperations, 2000-2005 (CFAF billion) 2000 2001 2002 2003 2004 2005 Total revenuesand subsidies 626.3 664.4 726.7 797.8 839.7 922.4 Revenues 562.3 602.7 664.6 720.1 771.6 841.0 Fiscal revenue 537.3 576.8 629.2 677.0 736.2 802.1 Non-fiscal revenues 25.0 25.9 35.4 43.1 35.4 38.9 Subsidies 64.0 61.7 62.1 77.7 68.1 81.4 Budgetary aid 14.1 0.0 1.9 17.6 21.8 14.8 Development projects 49.9 61.7 60.2 60.1 46.3 66.6 Total spendingand loans (net) 623.1 748.1 730.3 850.1 948.7 1049.8 Current expenditures 411.0 516.6 478.2 529.5 559.5 600.4 Wages and salaries 175.8 177.3 199.4 203.7 223.9 247.3 Interest payments 45.3 30.3 39.8 44.6 44.6 43.2 domestic debt 5.7 6.6 4.4 4.6 5.7 5.4 external debt 39.6 23.7 35.4 40.0 38.9 37.9 Other current expenditures 189.9 309.0 239.0 281.2 291.0 309.9 Transfers and subsidies 79.4 Goods and services 97.0 Other 13.5 0.0 0.0 0.0 6.2 10.0 Capital expenditures 193.2 232.3 275.9 338.5 370.8 422.9 From internalresources 106.6 133.6 147.9 190.3 211.6 279.0 HIPC financing 15.1 4.1 28.4 38.7 49.6 Non-HIPC financing 118.5 143.8 161.9 172.9 229.4 From external resources 86.6 98.7 128.0 148.2 159.2 143.9 SpecialTreasury and correspondingaccounts 14.0 3.8 -18.2 -11.1 0.8 0.0 (net) Loans (net) 4.9 -4.6 -5.6 -6.8 6.3 8.0 Loans 11.4 5.7 2.7 2.7 16.0 16.0 Repayment -6.5 -10.3 -8.3 -9.5 -9.7 -8.0 Temporary costs of structural reforms 0.0 0.0 0.0 0.0 11.3 18.5 Surplus or deficit (-), grantsincluded 3.2 -83.7 -3.6 -52.3-I09.0 -12 7.5 Surplus or deficit (-), excludinggrants -60.8 -145.4 -65.7 -130.0 -177.1 -208.8 Primary surplus or deficit (-) 37.2 -25.9 69.1 49.3 54.3 29.2 Financing -3.2 83.7 3.6 52.3 109.0 127.5 ExternalJinancing 17.1 54.9 68.4 60.9 129.I I 27.I Drawdowns 78.1 103.3 112.7 90.8 128.9 148.6 Treasury 37.1 60.6 42.2 0.0 0.0 55.3 Project credit 41.0 42.7 70.5 90.8 128.9 93.3 Other 0.0 0.0 0.0 0.0 0.0 0.0 Depreciation -65.2 -64.2 -81.O -73.9 -119.2 -127.1 HIPC 4.2 15.8 36.7 44.0 119.4 105.7 Domesticfinancing -21.1 17.4 -70.7 -8.5 -20.1 0.3 banking system -28.3 21.8 -75.4 -42.3 -31.5 2.5 outside the banking system 7.2 -4.4 4.7 33.8 11.4 -2.2 Errors and omissions 0.8 11.4 5.9 -0.1 0.0 0.I Source: Government of Senegal and IMF 70 B. ANALYSISOFTHE 2005 PROGRAM 5. The 2005 budget law and the program established with the IMF call for an increase inthe budget deficit, which should reach 3 percent o f GDP in 2005. As a result o f the favorable macroeconomic conditions and the availability o f external financing, this growth in the deficit should not carry major macroeconomic risks for Senegal, although due caution i s still advised. Inthis context, the public sector debt level should decline by CFAF 8.4 billion inrelation to the situation at the end o f 2004, even ifofficials plan to issue bonds on the sub-regional market. Expenditureanalysis 6. A review o f the TOFE indicates that total spending (labor costs, operating expenses, transfers, investment expenditures, special Treasury accounts, and public debt service) i s estimated at CFAF 1049 billion for 2005, i.e. an increase o f nearly 11percent over 2004. This increase i s largely due to: (i) costs (+10.5 percent) and (ii) labor investment expenditures (+14.1 percent), particularly from internal resources (+32.9 percent). The increase in labor costs i s attributed to the implementation o f new recruitment policies and an increase in the wages o f government workers decided in October 2004. It will however be offset by a drop in current expenditures on transfers and interest payments. The public sector wage bill should reach 19 percent o f fiscal revenue, which i s below the convergence criterion inWAEMU countries. 7. Investment expenditures are projected to increase by approximately 0.5 percent o f GDP once they are adjusted for the execution rate, which should not exceed 75 percent according to the government and the IMF. The budget law includes ambitious infrastructure projects, especially in urban infrastructure. A major portion o f the investment budget, some 12.8 percent (1.7 percent o f GDP), i s earmarked to improve urban mobility within the city o f Dakar in order to meet infrastructure requirements and improve the living conditions o f populations in and around the capital. The other sectors slated for substantial investment are the other priority sectors o f the PRSP (46 percent o f the total BCI), namely health, education, drinking water, sanitation, electrification, agriculture, fisheries, livestock, mining, and industry. 8. The alignment o f the budget law with PRSP objectives is shown by the growth in spending on social sectors such as health and education (see Table 2). The operating budgets for education and health increase respectively to 40 percent and 10 percent o f the total operating budget,versus 37 percent and 9.8 percent in2004. Within education, the portiono fthe operating budget earmarked for basic education increases from 45 percent in2004 to 46.4 percent in2005, which i s consistent overall with sector policy. In the area o f health, primary health care shows an increase of 14 percent in line with the strategic goals for this sector. The budget also reflects an increase in pay for contract workers inthe health field, as well as a build-up o f the incentive fund for healthworkers, particularly those working inpoor areas. 71 Table 2: Sectoralbreakdownof publicspending, 2000-2005 (% of GDP) 2000 2001 2002 2003 2004 2005 2000- Share 2005 oftotal Government 0.5 0.6 0.4 0.5 0.4 0.4 0.5 1.6 President o f the Republic 0.2 0.3 0.2 0.3 0.2 0.2 0.2 0.8 National Assembly 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.6 Court o f Appeal 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Audit Office 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 Other 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.2 General administrative action 3.4 3.6 3.4 3.4 3.3 3.2 3.4 11.9 Min.ofForeignAffairs, African 0.5 0.6 0.6 0.6 0.6 0.6 0.6 2.1 Union, and Min.o f Expatriates Armed Forces 1.5 1.6 1.5 1.5 1.4 1.3 1.5 5.1 Interior and Local Governments 1.o 1.o 0.9 0.9 0.9 1.o 0.9 3.3 Justice 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.8 Other (Prime Minister, Civil Service, 0.2 0.2 0.2 0.2 0.1 0.2 0.2 0.6 Labor, Employment, and Professional Organizations) Economic action 10.2 9.8 9.3 9.8 9.3 11.3 9.9 34.8 Fisheries 0.4 0.3 0.2 0.3 0.3 0.2 0.3 1.o Infrastructure, Public Works, and 2.4 2.1 3.O 3.2 3.0 5.3 3.2 11.1 Transportation Agriculture and Animal Husbandry 1.9 2.4 1.5 1.4 1.3 1.5 1.7 5.9 Economy and Finance and Planning 0.4 0.4 0.4 0.4 0.3 0.3 0.4 1.3 SMEs, Commerce, and Microfinance 0.0 0.0 0.0 0.0 0.1 0.1 0.0 0.2 Urban and RegionalPlanning, Housing 0.6 0.7 0.7 0.6 0.7 1.o 0.7 2.5 Miningand Energy, Water Supply, 2.8 2.3 2.0 2.3 1.9 1.7 2.2 7.7 Industry and Handicrafts Tourism 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.1 Other (including Studies and Research) 1.5 1.6 1.3 1.4 1.7 1.2 1.5 5.1 Educational, cultural, and social 6.6 6.8 7.8 9.0 8.4 9.1 8.0 27.9 action Education and Training 3.8 3.6 4.2 4.7 4.9 5.4 4.4 15.5 Health, Hygiene, Prevention, and 1.6 2.0 2.2 2.2 2.1 2.4 2.1 7.3 Nutrition Environment, Culture and 0.9 0.8 0.9 0.9 0.9 0.9 0.9 3.0 Communication, Sports, Scientific and Technological Research Family, Social Development, and 0.3 0.4 0.6 1.2 0.5 0.5 0.6 2.1 National Solidarity Public debt 4.2 2.5 4.2 3.1 3.0 2.6 3.3 11 Other unclassified 3.4 6.8 2.9 2.9 2.7 2.5 3.5 12.3 TOTAL 28.3 30.2 27.9 28.7 27.0 29.0 28.5 100.0 Source: Government of Senegal 72 9. The 2005 draft budget also provides for the use o f HIPC resources inthe total amount o f CFAF 59.6 billion, o f which 16.8 percent i s for operations and 83.2 percent for investment. The allocation o f resources for operations favors the social sectors such as education, health, and decentralization, which receive CFAF 4.5 billion, CFAF 4.2 billion, and CFAF 1.2 billion respectively. As for investment, the sum o f CFAF 39 billion, i.e. close to 80 percent, is earmarked for rural areas. Revenue analysis 10. Total revenues are projected to reach CFAF 922 billion for 2005, i.e. an increase o f CFAF 92 billion in absolute value and 10 percent in relative value over 2004. This increase i s mainly due to the volume o f fiscal revenue, which increases by CFAF 66 billion, largely as a result o f an increase in revenue from the import value-added tax, the special tax on petroleum products, alcohol, and tobacco, and the corporate tax. The import value-added tax should record a substantial increase o f CFAF 20.9 billion, boosted by the favorable trend o f economic activity and the efforts made by officials to ensure that it is collected. c. PROGRESSINIMPLEMENTINGTHE ACTION PLANS DERIVEDFROMTHE CFAA AND THE CPAR 11. InJune 2003, the Senegalese Government adopted two actions plan aimed at improving the government's budget procedures: the CFAA action plan and the CPAR action plan. Implementation o f these two plans falls under the responsibility o f a Secretariat, located within the Ministry o f Economy and Finance and supported by development partners. Their implementation i s one o f the central pillars o f the budgetary support currently provided by donors, including the World Bank, the European Union, and the French, Dutch, and Canadian foreign aid agencies, and it plays a central role in the program to which agreement was reached with the IMF. 12. Although detailed information on implementation o f these two action plans can be found in the periodic progress reports of the Secretariat, we propose to summarize here' the main issues.53 First o f all, it should be re-emphasized that Senegal has made strides, as indicated by the recent Assessment and Action Plan for HIPCs (AAP)conductedjointly by the World Bank and the IMF. Indeed, o f the 16 criteria, Senegal met seven in 2003 versus four in2001. Despite this progress, Senegal remains well behind the best performers, especially Mali. 13. CFAA action plan. Over the last two years, Senegal has initiated implementation o f the CFAA action plan through the following measures: (i) the beginning of decentralization o f payment order issue in four ministries; (ii) adoption o f the decree establishing new rules for finance operations oversight in July 2004; (iii)transfer o f the government's accounting statements to the Audit Office for the years from 1998 to 2002; (iv) implementation o f the government worker recruitment policy; and (v) completion o f the study on civil service pay, which led to a series o f actions inOctober 2004, including an increase ingovernment wages. 53Monthlyprogress reports as ofDecember 31, 2004 are available onrequest. 73 14. With respect to budgetpreparation, the government instituted a newbudget nomenclature inthe 2005 budget law which permitted unifiedpresentation of the budget. Efforts should be made to accelerate budget preparation on a programmatic basis, which will make it possible to better align public spending with the priority objectives o f the PRSP. The strategy followed by the government i s gradual inasmuch as it begins by developing the Medium-Term Expenditure Framework in four sectoral ministries (health, education, justice, and the environment), the results o f which will be incorporated into the 2006 budget law, with this approach then expanded to other ministries over time. The current system o f budget preparation should be further improved as regards projections o f investment expenditures, recurrent costs, and sources o f financing, particularly development assistance anticipated from donors. 15. With respect to budget execution, the government should try to accelerate the implementation o f a number o f actions to address problems that continue to undermine the transparency o f government finance in Senegal. For example, the posting and monitoring o f accounts show worrisome gaps. While the recent introduction o f the SIGFIP software program should improve the monitoring o f public expenditures, several technical problems remain and should be resolved as a matter o f great urgency. There are still considerable delays in the preparation o f the Treasury's monthly accounts, sometimes exceeding two months, which i s unacceptable even by sub-regional standards. The Table o f Government Finance Operations (TOFE) also i s not prepared on a monthly basis, or else after some delay. In this regard, it i s essential that Senegalese officials complete on time the technical and financial audit o f pay, as this will permit an initial readjustment o f the civil service wage bill and greater transparency. These gaps undermine the monitoring and management o f government finance and prevent officials from detecting potential problems in timely fashion. Senegalese officials are aware o f these problems and plan to correct them by the end o f 2005. 16. It is important that officials attempt to streamline their internal oversight operations. As such, they have launched an evaluatiodstudy that should help define the main problems and a priority action plan. As regards external oversight, while accounting statements through 2002 have been transferred from the Treasury to the Audit Office, the limited capacities o f the Audit Office are the principal reason for delays in inspecting these statements. Senegalese officials have received donor support to help them build the human and financial capacities o f the Audit Office. 17. The strategy for decentralizing management o f the investment budget has been finalized by an ad hoc committee composed o frelevant officials o fthe Ministryo f Economy and Finance, the sectoral ministries involved, and the ministry responsible for local government. It has been submitted for approval by the Interministerial Council, which is late inmeeting and starting the process. 18. CPAR action plan. The overall assessment o f implementation o f the CPAR action plan is less positive than inthe case o f the CFAA. First o f all, it should be pointed out that, although the National Steering Committee (CNP) responsible for supervising the plan was established in November 2004, it only includes representatives o f the Ministry o f Economy and Finance and the Office o f the President. There i s an urgent need to expand the composition o f this committee to other sectoral ministries and representatives o f the private sector and civil society. Secondly, the government has undertaken a number o f studies, including studies that should lead to a 74 revision of the procurement code before September 2005. From a legal standpoint, and as a temporary measure, the government adopted inthe 2005 budget law a measure that enables it to limit financial compensation for the private sector if a contract was not approved in advance by the Ministryo f Economy and Finance. The government also agreed with the IMF that: (i) under 20 percent o f the contracts that go through CNCA shall be approved through direct contracting; and (ii)no contract shall be signed in the absence o f a budget consideration approved by Parliament. 19. The interpretative circular on the government procurement code (CMP) and the circular on the compulsory preparation o f procurement plans by government entities subject to the CMP have been reviewed and approved by the Ministry o f Economy and Finance and sent on to the Prime Minister for the necessary directives to be given. It i s urgent that these directives be signedas soon as possible. 20. From an institutional perspective, the studies now in progress should lead to: (i)the creation o f an autonomous appeals body, which would rely on the participation o f the private and public sectors and civil society; (ii)establishment o f the Central Procurement Directorate within the Ministry o f Economy and Finance; and (iii) creation o f units responsible for coordinating the and supervisingprocurement procedures inall agencies that handle contracting. 75 ANNEX B: PUBLICINVESTMENT AND ECONOMICGROWTH: A SIMPLEMACROECONOMICMODEL 1. As the point o f departure we postulate an output function o f Cobb-Douglas form that incorporates the government's physical capital, G, as an input:54 Y = AK"GPLX a , / ? , ~0 + (1) where A stands for the economy's productivity index, K i s private capital, G i s public capital, L i s the level o f employment, and Y is output. 2. Taking the logarithmic hnction from both sides o f the equation above, and indicating the logarithms o f the variables inlower case, we find: The coefficients a,p, and x represent the elasticities o f each o fthe inputs. 3. Both equations are deduced from a condition of long-term balance, which is certainly not appropriate for an emerging country such as Senegal.55 The usual response to this type o f criticism i s to relax the hypothesis by considering the fact that the model can serve to project the behavior o f the variables during the transition. Our method was to approximate the speed o f convergence toward long-term balance inthe following manner: Ay =Aa PAk + +xAG (3) 4. Equation (3) stipulates that the economic growth rate (Ay) i s positively correlated to technological change (A a ) and to changes in stock levels o f physical capital, both public (AG) and private (Ak). For purposes o f simplification, we postulate that the level o f employment i s constant during the transition (AL=O). 5. We follow a majority o f the literature by adopting the hypothesis that change inthe stock o f public capital i s determined in a relatively exogenous manner; i.e. it i s not influenced by growth o f the economy during the year in progress even if such a connection exists in the long term.56 On the other hand, it would be unrealistic to hypothesize that change in the stock o f private capital is totally exogenous. This i s why we assume that this last variable i s influenced by contemporary economic conditions, following the approach proposed by the flexible accelerator investment model based on the methodology proposed by Blejer and Khan (1984).57 j4 For a similar approach, see J. Lighthart, Public Capital and Output Growth in Portugal: An empirical analysis, IMF, working paper, January 2000. "Alimitednumberofcountrieshavetestedacoherenteconomicgrowthtrajectorywithahypothesisoflong-term balance. See Barro and Sala-I-Martin for detailed explanations. 56 For a discussion, see M.Ramirez and N.Nazmi, Public Investment and Economic Growth in Latin America: An empirical test, Review of Development Economics, 7 (l), 2003. j7 M.Blejer and M.Khan, Government Policy and Private Investment inDeveloping Countries, IMF Staff Papers, 31.3 1984. 76 To sum up, the behavior o f private investment, measured by coefficient p ,is assumed to vary on the basis o f the private sector's capacity to finance its desired level o f investment, which is itself determined by: (1) the trend o f economic activity; and (2) the availability o f financing. These two factors can be expressed through the following equation: p =Ao + 4Y+ A2Ab AK 4 *O J , *0 (4) where y i s the trend o f economic activity and Ab i s the extent o f private financing on domestic markets.58 It is assumed that an increase in public financing leads to a reduction in the self- financing capacity o f the private sector and thus restricts private investment. This relationship, called the crowding out effect, influences the speed o f adjustment rather than the desired level o f private investment. 6. SubstitutingEquation(4) inEquation(3), we find: Ay =Aa +AoAk+4 y+ A2Ab XAg (5) + 7. We incorporate a special factor inthe case o f a country such as Senegal by assuming that technological progress i s relatively stationary in a poor emerging country, which allows us to write the variable Aa as a constant. Finally, we still need to perform one additional transformation due to the absence o f statistical data on private and public physical capital stock inSenegal. We assumethat changes inthese stock levels canbewritten as follows: where Ip and Ig stand for private and public investment respectively; 6p and 6g are the private and public capital stock depreciationrates. We adopt the hypothesis that physical capital stock i s determined inthe following manner: K, =SPY (8) K, =SgY (9) where 8, and 8, are fixed coefficients. Lastly, we can write the following equation which will be tested inthe case o f Senegal. Ay =a1+Af0I,I Y-, At2Ab + +x'I, +E Y-] (10) with a'= (a-h16,-h$g), h' l= hllep, hY2=h2ie, ~~~~ ~~ 58 We postulate that external financing i s available to the government. 77 ANNEX C:OVERVIEW OFSENEGAL Sub- POVERTY and SOCIAL Saharan Low- Senegal Africa income Developmentdlamond' 2003 Population,mid-year (millions) 10.0 703 2,310 GNI percapita (Atlas method, US$) Lifeexpectancy 550 490 450 GNi (Atlas method, US$ billions) T 5.6 347 1,038 Average annual growth, 1997-03 Population(%) 2.3 2.3 1.9 Labor force (%) 2.4 2.4 2.3 GNI Gross per primary Most recent estimate (latest year available, 1997-03) capita enrollment Poverty (% of populationbelownationalpoverfy line) 57 Urban population (% of totalpopulation) 50 36 30 Lifeexpectancy at birth (years) 52 46 58 1 Infant mortality(per 1,000live birfhs) 79 103 82 Chiid malnutrition(% of childrenunder 5) 23 44 Access to improvedwater source Access to an improvedwater source (% of population) 78 58 75 Illiteracy(% ofpopulation age 15+) 61 35 39 - Gross primaryenrollment (% of school-age population) 75 87 92 Senegal Male 79 94 99 Low-incomegroup Female 72 80 85 KEY ECONOMIC RATIOS and LONG-TERM TRENDS 1983 1993 2002 2003 Economlc ratios. GDP (US$ billions) 2.5 5.4 5.0 6.5 Gross domesticinvestmenffGDP 12.8 13.7 18.5 20.1 Exports of goods and services/GDP 31.7 22.2 30.1 28.4 Trade Gross domestic savingdGDP -1.9 7.7 8.8 8.0 Gross national savings/GDP -6.3 4.4 12.6 13.8 Currentaccount baiance/GDP -18.8 -10.6 -5.9 -6.3 Interestpayments/GDP 1.7 0.4 1.3 1.o Total debffGDP 83.8 69.2 77.5 64.1 Total debt service/exports 11.3 9.1 11.4 9.9 Presentvalue of debtlGDP 47.9 L Presentvalue of debffexports 125.4 Indebtedness 1983-93 1993-03 2002 2003 2003-07 (average annualgrowth) - GDP 2.1 4.9 1.1 6.5 5.9 Senegal GDP per capita -0.7 2.3 -1.3 6.0 3.7 Low-incomeWOUD Exports of goods and services 1.5 7.2 -1.7 0.5 4.9 STRUCTURE of the ECONOMY (% of GDP) Agriculture 21.5 19.0 Industry 15.5 19.1 Manufacturing 10.6 13.0 Services 63.0 61.9 63.4 62.0 Privateconsumption 83.9 77.5 Generalgovernment consumption 18.0 14.8 importsof goods and services 46.4 28.2 39.8 40.5 1983-93 1993-03 2002 2003 (average annualgrowth) Growthof exports and imports(Oh) 1 Agriculture 1.2 2.4 -19.9 19.2 30 Industry 3.3 6.7 9.5 4.6 20 Manufacturing 3.3 5.9 10.1 0.3 Services 2.1 5.1 4.8 4.1 10 Privateconsumption 1.6 1.2 -0.2 2.7 0 Generalgovernment consumption 1.7 7.3 1.2 8.3 -10 Gross domestic investment 4.7 10.0 5.4 16.4 I -EXDO~~S *Imports I Importsof goods and services 1.o 1.8 -2.6 -1.6 Note: 2003 data are preliminalyestimates. The diamonds show four key indicatorsin the country (in bold) comparedwith its income-groupaverage. If data are missing,the diamondwill be incomplete. 78 ANNEXC: OVERVIEWOFSENEGAL(CONTINUED) PRICESand GOVERNMENTFINANCE 1983 1993 2002 2003 Domestlc prlces lnflatlon (Oh) (% change) 5 Consumer prices 11.4 -1.o 2.3 0.0 0 ImplicitGDP deflator 8.9 -1.4 2.7 0.8 -5 Government finance -10 (% of GDP,includescurent grants) -15 Current revenue 18.8 16.8 19.0 19.6 -20 Current budget balance -4.0 0.2 5.4 5.5 Overall sumlus/deficit 4 . 1 -4.0 -3.0 -3.0 -GDP deflator &CPI TRADE 1983 1993 2002 2003 (US$ millions) Exportand import levels (US$ mlll.) Total exports (fob) 606 719 1,066 1,332 I 2.5w T Groundnut products 170 47 69 61 Phosphates 48 43 174 168 Manufactures 160 164 250 304 Total imports (cif) 1,042 1,235 1,775 2,247 Food 265 348 335 410 Fueland energy 239 124 284 387 Capitalgoods 164 159 317 325 Export priceindex (1995=100) 62 56 87 87 97 98 99 w 01 02 Import price index (1995=lOOj 53 61 83 83 Exports mImports Terms of trade (1995=100) 117 90 104 105 BALANCE of PAYMENTS 1983 1993 2002 2003 (US$ millions) Currentaccount balance to GDP (YO) Exports of goods and services 955 1,272 1,518 1,847 Importsof goods and services 1,312 1,674 2,004 2,628 Resource balance -358 -401 -486 -760 Net income -103 -191 -184 -92 Net current transfers -6 15 373 464 Current account balance -467 -578 -298 -408 Financingitems(net) 414 466 368 342 Changes in net reserves 53 112 -70 67 Memo: Reserves includinggold (US$ millions) 23 15 579 689 Conversionrate (DEC,/ocal/US$) 381.1 283.2 697.0 580.1 EXTERNALDEBTand RESOURCEFLOWS 1983 1993 2002 2003 (US$ millions) Composltlonof 2003 debt (US0 mill.) Total debt outstandingand disbursed 2,078 3,760 3,904 4,167 iBRD 86 52 0 0 G 156 IDA 183 918 1,578 1,806 I F 51 Total debt service 117 127 220 237 iBRD 10 16 0 0 IDA 3 10 16 24 E 1.208 B 1,606 Compositionof net resourceflows Officialgrants 108 299 190 190 Officialcreditors 274 101 93 83 Privatecreditors 58 -6 1 1 Foreigndirect investment -35 -1 93 Pottfoiioequity 0 6 0 0 I C:240 World Bank program Commitments 59 40 45 46 A. IBRD E. Bilateral Disbursements 32 46 114 107 B IDA D Other multilateral . F Private - ~ Principalrepayments 6 15 6 13 C - IMF G .Short-term Net flows 26 31 108 95 Interest payments 7 11 10 12 Net transfers 19 20 98 83 Development Economics 9/29/04 79 STATISTICAL APPENDIX 80 Table 1: Gross domestic product, investment,and bank credit, 1966-2000 (CFAFbillion) Private Iomestic bank GDP-constant GDP-current ublic investment investment- :redit-current Year value value constantvalue onstantvalue value 1966 869.0 206.3 23.0 72.5 30.8 1967 858.4 206.5 24.8 79.7 27.1 1968 912.4 218.3 25.9 88.7 32.6 1969 852.6 218.0 26.4 96.1 34.7 1970 925.6 241.5 32.7 114.0 37.6 1971 924.3 248.4 33.7 123.3 39.8 1972 983.3 275.1 38.4 149.4 47.3 1973 928.4 279.7 37.5 133.4 63.5 1974 967.4 340.3 40.6 127.2 89.7 1975 1040.3 408.5 42.1 135.3 106.9 1976 1133.1 461.8 40.0 124.5 121.9 1977 1102.7 486.1 44.4 113.8 144.6 1978 1059.1 498.5 36.6 98.8 196.1 585.2 38.7 83.4 232.4 I1979 1133.3 1980 1095.7 631.O 56.4 53.8 269.7 1981 1082.8 673.6 40.3 64.9 324.1 1982 1248.9 848.9 41.4 96.8 354.9 1983 1276.0 944.9 42.4 104.5 368.6 1984 1225.0 1021.2 43.6 97.3 373.7 1985 1271.5 1158.5 46.1 70.3 396.8 1986 1329.1 1303.3 53.9 77.8 391.7 1987 1382.3 1382.3 57.0 115.4 403.7 1988 1452.4 1483.3 55.7 123.9 432.1 1989 1432.0 1475.8 57.2 101.6 442.4 1990 1487.7 1551.5 55.9 142.3 411.5 1991 1481.8 1551.5 57.5 119.1 396.6 1992 1514.6 1595.4 65.9 146.8 419.9 1993 1481.0 1537.8 51.7 142.0 426.6 1994 1523.5 2022.3 44.9 124.1 275.6 1995 1602.2 2234.0 58.2 171.4 283.8 1996 1684.5 2371.8 86.2 203.0 344.7 1997 1769.5 2553.2 93.8 159.8 392.7 1998 1871.1 2746.0 112.8 172.1 437.1 1999 1964.7 2925.0 143.3 232.4 483.0 2000 2074.4 3114.0 118.1 251.5 622.2 81 Table2: Publicinvestmentpattern,1997-2001 (%oftotalpublicinvestmnt) - 1997- 1W 1998 1999 2000 - 2001 2001 T N l - R % G o w R " T E N " L 97% 97% %YO 95% 95% %% 3OvERNMENT 96% 96% 94% 94% 93% 95% ConsolidatedInvesttllentBudget 92% 92% 89% 88% 85% 89% operatingbudget 4% 4% 6% 5% 8% 5% Investmntsexecutedbydonors 0% 0% G?? 0% 0% 0% Miscellaneouscentral gov't entities(ODAC) 1% 1% 1Yo 2% 1% 1% L o c A L G 0 W R " T E " 3% 3% 2% 3'/o 4% 3% LOCALGOVERNMENTS 3% 2% 2% 3% 3% 3% R@OnS o?? 0% 0% 0% 0% 0% Municipalities 2% 2% 2% 2Yo 3% 2% Rlxalc"ities 0% 0% 0% 0% 0% G% Miscellaneoulocalgov't entities (ODAL) 0% 0% 0% 0% O??O G% chambaofcorn= 0% 0% G?? 0% O??O 0% chambersoftrade 0% 0% ooh 0% G% 0% [SBL(NGO) 0% 0% 0% 0% 0% 0% SOCIALSECUlUTYADMINLSTRATION 0% 0% 2% 2% 2% 1Yo Social security find 0% OYO 1Yo 0% 0% 0% PRES 0% 0% 2% 2% 2Y0 1% TOTAL 100% 100% 100% 100% 100% 100% 82 Table 3: Public investment inAfrica, YOof GDP, 1990-2002 Annual average I990 1995 2000 2001 2002 75-84 85-94 I Angola 6.3 6.0 6.9 Benin 7.4 10.4 7.6 7.8 6.4 8.1 Botswana 12.7 11.0 12.3 10.7 11.7 13.0 Burkina Faso 3.9 9.8 12.1 8.3 1.4 6.8 Burundi 12.5 8.3 5.6 3.6 4.0 14.1 12.6 Cameroon 5.5 1.2 1.4 2.1 1.6 4.7 6.0 Cape Verde 10.3 27.1 12.0 10.0 11.2 20.7 Central African Republic 4.7 9.8 7.1 7.4 7.6 4.7 6.4 Chad 10.0 6.3 10.5 9.1 10.0 2.2 5.6 Comoros 5.0 6.4 3.8 4.5 4.8 23.3 10.8 Congo, Dem. Rep. o f 4.0 4.4 0.2 0.1 0.1 3.9 3.5 Congo, Rep.o f 5.6 9.3 7.0 10.1 8.6 35.5 9.0 CBte d'Ivoire 3.8 5.1 2.8 1.8 3.2 9.9 4.2 Djibouti 3.6 Equatorial Guinea 10.5 1.9 0.8 0.4 0.4 4.7 Eritrea 12.3 11.7 20.3 17.3 9.5 Ethiopia 5.7 7.5 5.2 8.5 11.5 5.8 Gabon 3.9 5.1 4.8 5.6 5.4 15.8 5.5 Gambia, The 7.4 10.0 3.8 5.8 7.8 13.8 1.5 Ghana 7.5 14.0 10.4 10.4 9.6 2.5 8.6 Guinea 9.2 6.3 7.1 7.1 7.8 7.2 Guinea-Bissau 21.5 15.2 10.0 13.7 8.7 25.9 Kenya 9.4 7.4 3.1 2.7 2.7 8.9 8.1 Lesotho 23.O 15.6 8.5 11.3 10.2 20.0 19.0 Liberia Madagascar 7.9 5.8 6.7 7.3 4.8 6.4 7.0 Malawi 7.7 9.4 10.0 8.4 7.9 13.6 9.0 Mali 10.5 9.4 9.1 7.8 8.2 10.6 Mauritania 6.2 3.8 7.5 8.3 10.7 6.0 Mauritius 4.6 3.5 3.7 6.7 3.9 7.8 4.5 Mozambique 12.0 12.0 10.6 16.6 14.3 10.4 10.5 Namibia 8.2 7.1 6.0 9.3 9.5 13.9 1.7 Niger 7.4 5.2 5.9 6.5 7.4 13.8 7.1 Nigeria 10.8 5.3 9.3 14.1 11.7 10.0 Rwanda 5.9 8.1 6.0 6.6 6.8 13.7 8.4 Sgo Tome and Principe 33.1 26.0 19.3 12.4 24.4 Senegal 4.1 4.4 6.2 6.4 9.1 4.1 4.2 Seychelles 8.2 18.7 13.7 5.8 9.4 16.3 9.2 Sierra Leone 3.9 2.4 5.2 4.8 5.3 5.1 3.7 Somalia South Africa 1.4 2.0 2.2 i Sudan 2.6 2.5 3.3 Swaziland 7.0 5.9 6.2 6.7 5.9 10.9 9.3 Tanzania 10.5 3.4 3.4 3.5 3.3 7.0 Togo 7.3 3.5 3.3 2.5 2.6 12.2 7.1 Uganda 6.2 5.4 6.4 6.4 5.8 5.5 Zambia 6.2 5.1 10.0 11.9 11.8 4.1 5.3 Zimbabwe 3.4 2.9 0.0 1 .o 1.7 2.9 3.0 Source: Indicators o f development in Africa Table 4: Public investment in Africa, % of total budget, 1990-2002 Annual average 1990 1995 2000 2001 2002 75-84 85-94 95-MR Angola 18.0 10.3 10.5 12.9 14.3 13.8 13.1 Benin 26.9 36.6 38.1 38.4 31.8 27.1 36.0 Botswana 48.8 31.6 30.3 26.6 26.8 43.3 44.7 30.7 Burkina Faso 27.8 45.9 54.3 52.2 47.3 38.2 52.8 Burundi 41.9 30.3 22.5 29.9 17.3 119.7 51.9 25.7 Cameroon 26.2 7.1 15.4 18.0 14.6 40.3 31.1 13.5 Cape Verde 37.2 52.4 26.1 33.2 35.3 67.8 52.6 39.8 Central African Republic 41.4 50.2 52.4 55.3 56.4 42.1 47.3 Chad 54.9 50.4 50.8 50.2 50.4 50.9 50.8 Comoros 16.4 20.5 21.6 20.1 17.7 0.0 41.1 23.6 Congo, Dem. Rep. of 4.1 1.5 9.3 5.O Congo, Rep. of 7.9 9.1 27.3 31.8 24.6 61.7 16.1 20.7 C6te d'Ivoire 7.6 20.9 15.3 11.0 13.0 11.4 19.6 Djibouti 9.4 9.9 11.2 12.3 15.9 11.9 Equatorial Guinea 62.2 18.0 31.3 41.9 43.5 43.7 30.1 Eritrea 21.4 12.0 23.4 24.1 21.3 28.8 Ethiopia 27.3 37.7 21.0 32.8 34.7 24.5 31.2 34.5 Gabon 16.1 20.9 24.9 27.0 26.4 58.1 27.1 24.C Gambia, The 34.0 35.4 17.3 22.2 29.5 36.5 31.6 26.7 Ghana 41.0 46.1 33.1 39.1 23.4 14.5 39.5 38.3 Guinea 34.9 48.6 55.3 37.5 38.6 43.0 44.5 Guinea-Bissau 63.4 49.6 25.3 37.8 31.4 32.0 62.6 39.7 Kenya 19.4 21.1 10.7 14.1 10.8 23.5 16.3 15.8 Lesotho 44.8 32.8 20.8 20.2 20.6 37.9 30.2 28.3 Liberia Madagascar 48.9 36.2 49.4 43.8 33.9 39.9 39.8 41.0 Malawi 22.4 21.4 35.8 26.4 22.4 38.0 24.1 25.6 M a l i 61.7 53.5 52.0 48.7 49.2 47.9 56.9 52.7 Mauritania 27.2 27.6 42.1 35.1 37.9 28.2 31.8 Mauritius 18.2 15.9 15.9 10.3 16.6 21.8 19.0 15.6 Mozambique 51.9 57.3 50.7 57.7 54.0 40.1 43.3 53.5 Namibia 15.7 13.2 14.4 15.3 17.2 21.2 16.2 14.2 Niger 39.0 27.1 29.5 33.9 37.7 46.9 34.6 31.4 Nigeria 41.3 16.6 20.6 18.8 41.1 38.8 Rwanda 30.1 39.4 31.9 31.6 29.4 27.9 35.5 SZo Tom6 and Principe 56.0 69.1 63.4 55.4 55.0 50.5 59.6 Senegal 12.2 27.8 34.0 29.5 38.1 16.4 16.4 35.9 Seychelles 23.3 10.7 24.6 14.1 18.5 19.9 19.6 Sierra Leone 17.8 13.4 21.1 10.4 10.5 22.8 15.3 Somalia South Africa 10.9 9.2 5.6 7.6 8.5 23.6 10.3 7.5 Sudan 11.1 15.1 21.3 19.6 23.6 30.6 14.8 14.7 Swaziland 23.8 20.5 25.1 24.7 26.1 21.8 Tanzania 17.7 17.6 30.8 21.4 18.7 26.2 19.8 22.4 Togo 25.7 15.5 15.0 15.6 14.7 17.2 15.4 Uganda 43.8 45.1 53.0 36.3 40.2 39.7 42.1 42.6 Zambia 18.9 17.5 46.9 38.8 38.9 11.6 8.1 36.1 Zimbabwe 18.2 14.5 5.2 2.8 4.7 42.7 16.9 9.1 Source: Indicators of development inAfrica 84 Table5: ConsolidatedInvestmentBudgetas voted, 2000-2005 2000 2001 2002 2003 2004 2005 PRIMARY 107839 96535 97139 116488 101396 120260 4griculture 48346 38124 39792 39258 37747 46403 Livestock 2624 6180 5741 5373 5396 7994 Water andforests 16187 12039 13133 10928 10821 15081 Fisheries 11572 9792 7600 10968 9238 5596 Ruralandagriculturalwater supply 12207 11528 10244 25882 19670 19434 PRIMARYstudies andresearch 219 720 720 439 0 1398 PRIMARYsupport, reinforcement,TA 16684 18152 19909 23640 18524 24354 SECONDARY 52473 44029 32317 28821 34888 21051 Mining 750 550 750 999 899 600 [ndustry 3951 1433 1350 3073 3724 1000 Handicrafts 1470 1707 874 3580 202 492 Traditionalandrenewableenergy 34965 30070 23031 15918 7569 10861 SECONDARY studies andresearch 2075 430 400 909 1840 1800 SECONDARY support, reinf.,TA 9262 9839 5912 4342 20654 6298 TERTIARY 62970 52290 63091 83101 84043 162163 Commerce 428 120 150 250 225 877 Tourism 1326 1135 1000 1350 370 202 Roadtransport 47827 44400 57316 73037 70173 135540 Rail transport 5063 2160 900 800 900 19940 Sea transport 1810 100 1600 1160 3323 1500 Air transport 1330 250 200 236 1000 1320 Postal andTelecommunications Service 0 0 800 1023 0 1053 TERTIARY studies andresearch 821 673 0 0 3750 425 TERTIARY support,reinf.,TA 4365 3452 1125 5245 4302 1306 QUATERNARY 142533 189445 216001 234903 241243 287467 Urbanwater supply andsanitation 32293 30766 32269 34658 42213 41520 Culture, youth, andsports 2160 5255 4951 7615 6931 4603 Housingandurbanplanning 17858 22091 24561 20151 26796 40398 Healthandnutrition 25312 42503 46486 44665 47477 55777 Educationandtraining 23846 35365 31571 37840 36698 38700 Social development 7582 13330 16566 28140 16521 18102 Administrativeequipment 18685 21144 42302 43536 44938 71436 QUATERNARYstudies andresearch 1615 1290 700 733 850 1000 QUATERNARYsupport, reinf., TA 13182 17701 16595 17565 18819 15931 TOTAL 365815 382299 408548 463313 461570 590941 Source: Governmentof Senegal 85 Table 6: ConsolidatedInvestmentBudgetas executed, 2000-2003 (CFAF' million) 2000 2001 2002 2003 PRIMARY 53419 59618.575 71533 47246 4gricultwe 21855 21582 27610 18614 Livestock 1078 1722 2391 3148 GVater andforests 5661 >086.5749 9100 3936 hheries 9036 7732 10521 3057 h a 1andagricultural water supply 5905 3880 4511 6464 PRIMARYstudiesandresearch 48 34 720 33 PRIMARYsupport, reinforcement,TA 9836 15582 16680 11995 3ECONDARY 39241 10748 28753 19886 ming 704 550 704 995 [ndustry 1173 575 984 3028 Handicrafts 440 201 767 1182 Traditional andrenewableenergy 32426 2256 19752 8242 [ndustrialstudies/research 431 451 390 383 SECONDARY institutional support 4067 6715 6156 6056 E R W Y 43311 28542 46546 42664 Commerce 300 120 150 250 Tourism 0 430 865 950 Roadtransport 39308 26809 41181 34516 Radtransport 0 0 0 0 Sea transport 0 0 0 35 Air transport 1130 0 200 0 PostalandTelecommunications Service 0 0 800 1023 TERTIARYstudies andresearch 0 29 0 0 TERTIARYsupport, reinf:,TA 2573 1154 3350 5889 QUATERNARY 79134 133252.5 189439 171466 Urbanwater supplyandsanitation 9378 12904 16902 13551 Culture, youth, andsports 795 3514 4543 5275 Housing andurbanplanning 7967 17979 17953 15189 Healthandnutrition 27582 42329 43118 29622 Education andtraining 12952 19226 23231 31666 Social development 2184 5965 6470 16444 Administrative equipment 11805 16625.5 57768 40043 QUATERNARY studies andresearch 1331 366 872 661 QUATERNARY support, reinf,, TA 5140 14344 18582 19015 TOTAL 215105 232161.O; 336271 281262 Source: Governmentof Senegal 86 Table7 : BCIbeakdownbetweencapitalexpendituresandrecurrentcosts, 1999 CFAF million 1 Recurrent Capital Recurrentcosts Typeofoperation costs Expenditures mAL (%oftotal) PRrmARY 9993 19303 29296 34% &culture 4139 7301 11441 36% Livestock 1092 139 1231 89% Water andforests 2197 4475 6672 33% Fisheries 0 758 758 0% Ruralandagricultural wter supply 862 6432 7295 12% PRLMARYsupport, remforcenmt,TA 1703 198 1901 90% SECONDARY 5533 1229 6763 82% Mining 1006 18 1024 98% Industry 16 224 240 7% Handicrafts 942 35 976 96% Energy 1382 400 1782 78% SECONDAYstudies andresearch 24 149 173 14% SECONDAYinstitutionalsupport 2163 403 2567 84% TERTIARY 2914 32191 35105 8% chnmrce 990 453 1443 69% Tourism 250 0 250 100% Roadtransport 1675 30979 32653 5% Airtransport 0 760 760 OYO QUARTEXNARY 60302 61222 121525 50% Urbanuatersupplyandsanitation 18571 0 18571 100% Culture,youthandsports 0 288 288 0% Housingandurbanplanning 1753 1668 3421 51% Healthandnutrition 35656 19388 55044 65% Educationandtraining 1598 37388 38985 4yo Social developmnt 2063 583 2646 78% Administrative equipment 216 1629 1845 12% QUARTERNARYs a e s andresearch 446 279 725 62% 78743 113946 192689 41% 87 Table 8: PTIPforecast,2002-2007 (CFAF bill n) PTIP(2002- PTIP (2003- PTIP (2004- PTIP (2005- 2004) 2005) 2006) 2007) 2002 408.1 PRIMARY 97.2 Agliculture 39.8 Livestock 5.8 Water and forests 13.1 Fisheries 7.6 Rural and aglicultural water sumlv 10.2 PRIMARY studies and research 0.7 PRIMARY support. reinf.. TA 20 SECONDARY 30.5 TERTIARY 67.9 OUATERNARY 212.5 Urban water supplv and sanitation 32.3 Culture. vouth. and sports 4.9 Housing and urbanplanning 24.6 Health and nutrition 46.5 Education and training 28.4 Social development 16.6 Administrative procurement 42 OUATERNARY studies and research 0.7 OUATERNARYs u ~ ~ o rreinf..TA t . 16.5 2003 396.4 449.9 PRIMARY 87.2 111.3 Aaiculture 35.9 41.3 Livestock 4 5.7 Water and forests 13.5 13.4 Fisheries 12.4 11 Rural and amicultural water sunolv 12 17.7 PRIMARY studies and research 0 0.4 PRIMARY suu~ort.reinf.. TA 9.4 21.8 SECONDARY 21.4 28.7 TERTIARY 80.5 79.8 OUATERNARY 207.3 230.1 Urban water sutmlv and sanitation 64.8 36.6 Culture. vouth. and sports 3.3 5.6 Housing and urban planning 22.3 21.3 Health and nutrition 38.7 50.4 Education and training 37.7 36.3 Social development 13.6 19.1 Administrative eauipment 13.5 42.5 OUATERNARY studies and research 0 0.7 OUATERNARY support. reinf.. TA 13.4 17.6 88 377.1408.2 468.6 101.9 93.5 110.6 37.8 39.8 45.3 11.9 5.4 5.7 10.6 14.3 12 20 9.7 9.2 10.9 13.9 19.9 0.5 0 0 10.2 10.4 18.5 10.5 23.5 34.9 100 92.3 80.9 164.7 198.9 242.2 51.6 61.8 41.9 1.2 6.9 21.8 23 28 21.5 43.4 47.5 37.5 41.5 36.7 10 12.9 16.5 9 13.1 44.9 0 0 0.9 13.3 2 18.9 2005 334.5 415.0 591.7 PRIMARY 93.1 102.1 121.0 Aaiculture 37.6 37.5 46.4 Livestock 8.0 6.7 8 Waterandforests 12.2 15.6 15.8 Fisheries 7.5 9.1 5.6 Ruralandaaiculturalwatersuuulv 17.5 20.1 19.4 PRIMARYstudies andresearch 0 0 1.4 PRIMARYsuuuort.reinf.. TA 10.3 13.1 24.4 SECONDARY 18.8 20.2 21 TERTIARY 80.3 95.5 162.2 OUATERNARY 142.3 197.2 287.5 Urbanwater sumlvandsanitation 46.4 46.4 41.5 Culture,voutk andsmrts 0.2 0.2 4.6 Housingandurbanolannine 21.5 20 40.4 Healthandnutrition 24.8 39.8 55.8 Educationandtraining 27 38.5 38.7 Socialdeveloument 8.3 15.7 18.1 Adrmnistrativeeauiument 11.6 19.5 71.4 OUATERNARYstudiesandresearch 0 0 1 OUATERNARYsuumrt, reinf...TA 2.5 17.1 16 89 310.1483.2 87.7 96 36 45 7.5 6 Water andforests 10.3 12.1 6.6 2.4 Ruralandamiculturalwater suuulv 18 18.1 PRIMARY studies andresearch 0 0.8 PRIMARY suuuort. reinf.. TA 9.3 11.6 17.8 25.1 52.9 135.1 OUATERNARY 151.7 227 Urbanwater supulvandsanitation 37.4 53.3 Culture.vouth. andsports 0.4 0.8 Housingandurbanplanning 15 37.2 Healthandnutrition 27.2 44.8 Educationandtraining 40.4 39.4 0.5 8.6 14.5 23.I 0 0 16.3 19.2 373.9 94.3 miculturalwater suuulv studies andresearch 22.9 87.7 169 Housingandurbanulanning Healthandnutrition Educationandtraining Social development Administrativeeauiument OUATERNARYstudies andresearch ~OUATERNARY suuport. reinf.. TA Source:Governmentof Senegal 90 t 8v R 8 8 i r . s 2 t I 8 0 Jp P ==i= g N P O , rlf I 10 p' N N s m O 2 "l R m 8 2 m ". 10 - N 10 N h 10. 0. N I C c 8 5 c C C C IC c c -5 h c C v c C C C 5 c v P P 0 0 x E. 0 f E '0. 81 0 s 8 8 d Y 0 - 0 0 n I I 0 I( 0 m - m 81 4 j m c e 6e .- 4.-.-t .L h c) E; k E 5 c .-2x e i;l s z " r- o o gic L rrt L ; o o o c $lg > o o z N Cll > o o z I N C i g o 0 0 z 0 c C1 T e 5 PI eF t w i i I I tr---Plf-lsi I I-- - o o o o o o o m 0 0 C e m 0 0 0 0 0 0 0 ~ 0 C m 3 P z 0 c E e c ir