Chapter 11: Senegal

International Monetary Fund. Independent Evaluation Office
Published Date:
April 2003
  • ShareShare
Show Summary Details


1. Growing external imbalances and diminished prospects for continuing private external financing led Senegal to seek its first IMF arrangement in 1979. Since then, the country has had an almost continuous succession of IMF arrangements, except during 1992–93 (Table 11.1). A short-lived arrangement under the Extended Fund Facility (EFF) was followed by four Stand-By Arrangements (SBAs) during 1981–85. Since 1986, the bulk of IMF lending to Senegal has been through concessional facilities—a Structural Adjustment Facility (SAF) arrangement and three Enhanced Structural Adjustment Facility (ESAF) arrangements—with resort to regular facilities/resources limited to supplementing access levels (1986, 1987) or in a transition to a multiyear concessional facility arrangement (1994). Following the transformation of the ESAF into the Poverty Reduction and Growth Facility (PRGF), Senegal’s third ESAF arrangement was converted to a PRGF arrangement in 2000; it expired in April 2002.1

Table 11.1.Senegal: IMF Arrangements
Arrangement1Date of arrangementOriginal expiration dateDate of expiration or cancellationAmount agreed (SDR million)Amount agreed (In percent of quota)2Average annual access level (In percent of quota)Amount drawn (In percent of agreed amount)
SBA IMar. 1979Mar. 1980Mar. 198010.525.025.0100.0
EFF3Aug. 1980Aug. 1983Sept. 1981184.8440.0146.722.2
SBA IISept. 1981Sept. 1982Sept. 198263.0100.0100.0100.0
SBA IIINov. 1982Nov. 1983Sept. 198347.375.075.012.5
SBA IVSept. 1983Sept. 1984Sept. 198463.0100.0100.0100.0
SBA VJan. 1985Jul. 1986Jul. 198676.690.060.0100.0
SBA VINov. 1986Nov. 1987Sept. 198734.040.040.0100.0
SAF4Nov. 1986Nov. 1989Nov. 198854.063.521.278.7
SBA VII5Oct. 1987Oct. 1988Oct. 198821.325.025.079.8
ESAF I6Nov. 1988Nov. 1991Jun. 1992144.7170.056.7100.0
SBA VIIIMar. 1994Mar. 1995Aug. 199447.640.040.065.0
ESAF II7Aug. 1994Aug. 1997Jan. 1998130.8110.036.7100.0
ESAF/PRGF III8Apr. 1998Apr. 2001Apr. 2002107.
Source: IMF Treasurer’s Department.

Roman numerals are used to indicate the sequence of arrangements, by type.

The size of Senegal’s quota at the IMF increased from SDR 42 million to SDR 63 million in December 1980, to SDR 85.1 million in December 1983, to SDR 118.9 million in December 1992, and to SDR 161.8 million in February 1999.

Approved as a three-year arrangement. The first review, envisaged for completion by December 1980, was not completed.

The approved amount was increased from SDR 40 million in July 1987. The second annual arrangement was approved on October 26, 1987.

Combined with second annual SAF arrangement

The second and third annual arrangements were approved in December 1989 and June 1991, respectively.

The second and third annual arrangements were approved in December 1995 and January 1997, respectively.

The second and third annual arrangements were approved in July 1999 and February 2001, respectively.

Source: IMF Treasurer’s Department.

Roman numerals are used to indicate the sequence of arrangements, by type.

The size of Senegal’s quota at the IMF increased from SDR 42 million to SDR 63 million in December 1980, to SDR 85.1 million in December 1983, to SDR 118.9 million in December 1992, and to SDR 161.8 million in February 1999.

Approved as a three-year arrangement. The first review, envisaged for completion by December 1980, was not completed.

The approved amount was increased from SDR 40 million in July 1987. The second annual arrangement was approved on October 26, 1987.

Combined with second annual SAF arrangement

The second and third annual arrangements were approved in December 1989 and June 1991, respectively.

The second and third annual arrangements were approved in December 1995 and January 1997, respectively.

The second and third annual arrangements were approved in July 1999 and February 2001, respectively.

2. Senegal has had outstanding IMF credits and loans continuously since 1975.2 They increased from SDR 110 million (174 percent of quota) at end-1980, to SDR 221 million (260 percent of quota) at end-1990, and then fell to SDR 205 million (127 percent of quota) at end-2001, partly reflecting Senegal’s net repayments to the IMF in recent years (Figures 11.1 and 11.2).

Figure 11.1.Senegal: Outstanding Obligations to the IMF

Sources: IMF Treasurer’s Department

Figure 11.2.Senegal: Net Borrowing from the IMF

Sources: IMF Treasurer’s Department and IEO calculations

3. What factors contributed to this prolonged use of IMF resources, and what have been the effects? In particular, to what extent were the objectives of the programs supported by these arrangements achieved? To the extent that key objectives have not been achieved (or achievements have not been sustained), do the failures represent weaknesses in policy implementation or in the design of programs? What can be learned about improving the effectiveness of IMF-supported programs and avoiding permanent reliance on IMF financing? These are the main questions addressed in this evaluation.

4. The evaluation is based largely on an extensive review of (published and unpublished) IMF documents and interviews conducted in Dakar (during an IEO mission in March 2002) and in Washington with (i) current and former senior officials of the Senegalese government and of the Banque Centrale des Etats de l’Afrique de l’Ouest (BCEAO); (ii) a broad range of other Senegalese stakeholders, including leaders of political parties and representatives of trade unions, NGOs, and journalists; (iii) representatives of the donor community based in Senegal; (iv) the IMF Executive Director for Senegal; (v) current and former IMF staff; and (vi) World Bank staff.

5. Senegal’s membership of the CFA franc zone limits the authorities’ scope for independent exchange rate and monetary policy actions.3 Arrangements for pooling international reserves, limits on central bank financing of government operations, and support of the French Treasury have succeeded in maintaining the convertibility of the CFA franc. Monetary policy is set at the regional level by the central bank, BCEAO. Since 1993/94, the principal instruments of monetary policy have moved away from administratively set interest rates and country-specific credit ceilings to indirect instruments; this change further narrowed the scope for country-specific monetary policy. Thus, fiscal policy and structural reforms are the principal means available to the authorities for effecting macroeconomic adjustment. Not surprisingly, these two policy areas feature prominently in this evaluation.

Overview of Policies and Performance

Background: the 1970s and early 1980s

6. Senegal’s external current account deficit almost doubled in the 1970s from an average of about 4½ percent of GDP in the first half of the decade to over 8½ percent in the second half, and the rising deficit was financed by public sector foreign borrowing. Total external debt rose from about $130 million in 1971 to almost $1 billion in 1979 (going from 19 percent to 35 percent of GDP). Real GDP growth averaged 2 percent a year but with wide fluctuations that partly reflected the impact of exogenous shocks, especially weather conditions and fluctuations in the international terms of trade. Favorable developments in the world prices of two of Senegal’s main exports (groundnut oil and phosphates) helped temper the effect of rising international oil prices on the terms of trade; on average, the terms of trade improved by about 15 percent in the second half of the 1970s, compared with the first half. Inflation followed an upward trend in the first half of the decade, peaking at over 30 percent in 1975, and then subsided to an average of less than 7 percent during 1976–80.

7. Senegal faced a severe financial crisis in the late 1970s and early 1980s, when deteriorating terms of trade and the government’s pricing policies produced a large external current account deficit (aver-aging 13 percent per annum during 1979–83) and an expansion in the public sector deficit.4 The staff estimated that at the end of the 1980/81 fiscal year, internal arrears accumulated by the central government and public enterprises exceeded total government revenue during the year.

Objectives and policies of the IMF-supported programs

8. The principal medium-term objective of Senegal’s IMF-supported programs during 1979–85 was to reduce internal and external financial imbalances to sustainable levels. With respect to external imbalances, the objective was to reduce the current account deficit to a level that could be financed without recourse to debt rescheduling or accumulation of arrears. The objective was to be achieved mainly through policies that restrained aggregate demand. The structural weaknesses that underlay Senegal’s macroeconomic imbalances—for example, a large inefficient public sector, extensive subsidies through controls on producer and consumer prices—were recognized at an early stage, and the early programs included measures to deal with them.5 In particular, they included measures to contain financial losses associated with government intervention in sectors such as agriculture and energy, policies to limit growth in public service employment, and attempts to strengthen tax administration. The operations of agencies performing quasi-fiscal functions (e.g., the Caisse de Péréquation et de Stabilisation des Prix (CPSP)) received particular attention.6

9. Structural reforms and social policy issues received greater prominence under the arrangements supported by the concessional facilities, beginning with the 1986 SAF. The advent of the Policy Framework Paper (PFP)—associated with SAFs and ESAFs—provided a vehicle for incorporating the authorities’ sectoral and social programs into IMF-supported programs. The PFP for the first ESAF arrangement (1988) indicated a “two-pronged medium-term strategy entailing (i) a reduction in the obstacles to private sector initiative and growth; and (ii) the achievement of greater efficiency in public resource management, including a strengthening in government finances.” Policies to be implemented included: a public investment program to support directly productive sectors, abolition of virtually all price controls, reduction in labor market rigidities, strengthening of the government budget through shifting to a more stable revenue base (including re-form of taxation and pricing policies for petroleum products), and reinforcing efforts to improve the delivery of basic services (e.g., education and health care) to the population.

10. With the transformation of the ESAF into the PRGF, programs continued to emphasize structural reforms to remove impediments to growth, but also made poverty reduction a more central goal. Furthermore, they have paid greater attention to the allocation of resources to priority social sectors and to rural infrastructure. In May 2002, Senegal submitted to the IMF and the World Bank, a PRSP that was produced from an extensive consultation process involving a wide range of domestic stakeholders and international development partners.7

11. The World Bank supported Senegal’s adjustment efforts in the 1980s and 1990s with, among other operations, four Structural Adjustment Loans (SALs) and several sectoral adjustment credits (including in the financial, agriculture, and energy sectors). The first SAL, which was approved in 1980, was closely aligned with the 1980 EFF.8 The second and third SALs, approved in 1986 and 1987, respectively, run in parallel with SBA and SAF arrangements. The fourth SAL, approved in 1990, aimed to build on achievements under the 1986 and 1987 SALs and to tackle some of the remaining challenges. To that end, it sought to help the authorities “restore Senegal’s competitive position and achieve growth with macroeconomic equilibrium.” It was to focus on, among other things, measures to improve production incentives (e.g., reducing tax burdens, costs of production, and labor market rigidities) and to rationalize the public sector (e.g., through civil service reform, reducing government subsidies to public enterprises, and privatization). A Financial Sector Adjustment Credit (approved in 1989) was instrumental in restructuring the banking system in Senegal and strengthening the BCEAO’s banking supervision capabilities. Adjustment credits to the agriculture and energy sectors (in 1995 and 1998, respectively) endeavored to tackle long-standing structural problems in those sectors and included some that had featured prominently in IMF arrangements.

Program implementation

12. Senegal’s record shows a stop-go pattern of program implementation as measured by compliance with performance criteria and benchmarks (hereafter referred to as “performance targets”) and timeliness of the completion of program reviews. Implementation was generally weak in the early programs during 1979–82. Two of the programs in this period (the 1980 EFF and the 1982 SBA) went off-track soon after they were approved, because of policy slippages (especially in price liberalization and tax measures) and the failure of IMF staff and the authorities to agree on policy adaptations required to attain program objectives in the face of unanticipated shocks. In the case of the EFF, part of the problem appears to have been weaknesses in the data used for establishing performance targets, which understated the magnitude of prevailing and prospective financial imbalances. By contrast, all the arrangements approved between 1983 and 1987 (two stand-alone SBAs and two combined SBA/SAFs) were characterized by high compliance with performance targets and timely completion of program reviews. This period was also marked by significant deregulation of the economy, including a reduction in the scope of price controls, partial liberalization of the agriculture sector, and the phasing out of most quantitative restrictions on imports.

13. Program implementation weakened again between 1988 and 1992. Under the first two annual arrangements of the 1988 ESAF, performance targets linked to midterm reviews were observed, but were followed by a loosening of fiscal policy after completion of the reviews. These policy slippages were judged by IMF staff to be sufficiently serious to warrant a reestablishment of a track record of good performance under a six-month staff-monitored program (July–December 1990) before the request for a third annual arrangement was submitted to the Executive Board. Under the latter arrangement, implementation was once again satisfactory prior to the completion of the midterm review (in November 1991), but this was not sustained.

14. Discussions on a successor ESAF arrangement started in 1992, before the expiration of the 1988 ESAF, but no agreement was reached over a period of nearly two years.9 In this intervening period, some measures that had been delayed under the ESAF were implemented, including extension of the coverage of VAT to the transport sector, and the starting of operations of a company hired to improve the system of valuation of imports in order to curb underinvoicing. Measures envisaged under a banking system reform program were also fully implemented. On the other hand, the authorities took several steps that worsened the public finances: an increase in the producer price of groundnuts (contrary to understandings under the ESAF) widened the deficit of the groundnut sector, and a reduction in selected customs tariffs and VAT rates lowered revenues. Furthermore, an agreed mechanism for automatic adjustment of domestic petroleum prices to reflect developments in world prices was not implemented.

15. Legislative and presidential elections in the first half of 1993 constrained policy actions needed to address the reemergence of severe financial problems. In August, the authorities announced a package of corrective “internal” (i.e., nonexchange rate) measures that included a 15 percent cut in most public sector nominal wages, increases in import duties, and increases in the retail prices of petroleum products. The reduction in wages was not implemented, following strong protests by trade unions.

16. Discussions between the staff and the authorities during 1992–93 covered the issue of devaluation as a policy option for the CFA franc zone. Similar discussions were held with the authorities in other CFA franc countries, against a background of real exchange rate appreciation and persistent adverse terms of trade developments in most of the member countries. A historic 50 percent devaluation of the CFA franc in January 1994 paved the way for new arrangements in support of a renewed adjustment effort—initially, an SBA to provide quick financial assistance, followed by a new multiyear ESAF arrangement five months later. Implementation of policies was generally good during 1994–99, al-though the pace of structural reforms was slow and there were policy slippages in the period leading up to legislative elections in 1998 and presidential elections in 2000.

17. The incumbent president, Mr. Diouf, was defeated by the veteran opposition leader, Mr. Wade, in presidential elections in early 2000 to end 40 years’ rule by the Socialist Party. President Wade’s Senegalese Democratic Party won a majority of seats in legislative elections for a new national assembly held in April 2001. Significant slippages in the timetable for structural reforms occurred during the political transition that extended into the period covered by the third annual PRGF arrangement (approved in February 2001). On the strength of some corrective measures taken by the new government, the IMF Executive Board completed the second of three envisaged reviews in April 2002, shortly before the expiration of the PRGF arrangement.

What was achieved over the extended period of programs?10

18. Over the period of Senegal’s prolonged use of IMF resources there has been a significant reduction in macroeconomic imbalances and less volatility in real GDP growth (Figures 11.3 and 11.4). Inflation has followed a downward trend, except for a spike in the rate following the large devaluation in the exchange rate in 1994 (Figure 11.5). The impact of terms of trade shocks on the overall economy has also declined, although there continue to be large fluctuations in the world prices of two commodities—crude oil and groundnut oil—which have been associated with periodic adjustment problems and linked in part to government policies (Figures 11.611.10).11

Figure 11.3.Senegal: External and Fiscal Balances

(In percent of GDP)

Sources: IMF, WEO database; and IEO calculations.

Figure 11.4.Senegal: Real GDP Growth

(In percent per annum)

Sources: IMF, WEO database; and IEO calculations.

Figure 11.5.Senegal: Inflation

(In percent per annum)

Source: IMF, WEO database.

Figure 11.6.Senegal: Impact of Terms of Trade Shocks

(In percent of GDP)

Sources: IMF, WEO database; and IEO calculations.

Figure 11.7.Senegal: Terms of Trade Indices

(1995 = 100)

Sources: IMF, WEO database; and IEO calculations.

Figure 11.8.Senegal: Export Price Index and World Price of Groundnut Oil

Sources: IMF, WEO database; and IEO calculations.

Figure 11.9.Senegal: Import Price Index and World Price of Crude Oil

Sources: IMF staff reports and IEO calculations.

Figure 11.10.Senegal: Nominal CFA Franc/U.S. Dollar Exchange Rate

(Annual averages)

Sources: IMF, WEO and WETA databases.

19. Table 11.2 presents selected indicators of macroeconomic performance during the three years before Senegal’s first IMF arrangement and five subperiods during 1979–2001. The subperiods are broadly based on “adjustment effort” as judged by consistency of program implementation (discussed above):12

  • (i) 1979–83: characterized by weak implementation.

  • (ii) 1984–88: characterized by strong implementation.

  • (iii) 1989–93: spanning the period of the 1988 ESAF arrangement (uneven implementa- tion) and a period when there was no IMF arrangement.

  • (iv) 1994–99: during which there was strong implementation of macroeconomic policies, and modest progress on structural reforms.

  • (v) 2000–01: marked by some policy reversals but continuing relatively good macroeconomic performance.

Table 11.2.Senegal: Selected Economic Indicators(In percent of GDP, unless otherwise indicated)
Period averages
Real GDP growth (percent per annum)
Inflation (percent per annum)5.310.65.0–
Terms of trade (percent change per annum in U.S. dollar price indices)7.2–5.92.8–1.80.9–2.5
External current account balance, including transfers–6.6–13.2–8.9–8.3–4.9–6.3
(balance of payments) of which: official transfers14.
Gross national saving25.8–
Gross investment12.310.711.813.517.718.1
Central government balance–1.4–6.9–2.3–1.5–0.6–2.1
Total government revenue and grants19.121.119.419.419.719.8
Of which: grants0.
Total government expenditure and net lending20.528.021.820.920.321.9
Final consumption expenditure98.6103.797.493.488.990.7
Public consumption expenditure19.219.316.214.811.513.6
Private consumption expenditure79.484.581.278.777.477.1
Gross capital formation12.310.711.813.517.718.1
Gross public capital formation4.
Gross private capital formation7.
Imports of goods and services48.748.338.931.738.138.7
Exports of goods and services37.233.829.624.731.430.0
Memorandum items

Decomposition of external adjustment (change, in percent of GDP)

Current account
Fiscal balance––1.5
Private sector saving-investment balance3–1.1–0.3–
Sources: Calculated from IMF, WEO and WETA databases.

The decline between 1994–99 and 2000–01 partly reflects a reclassification of project grants from the current account to the capital account.

Calculated as the difference between the external current account balance (balance of payment) and gross investment.

Calculated as the change in the current account balance minus the change in the fiscal balance.

Sources: Calculated from IMF, WEO and WETA databases.

The decline between 1994–99 and 2000–01 partly reflects a reclassification of project grants from the current account to the capital account.

Calculated as the difference between the external current account balance (balance of payment) and gross investment.

Calculated as the change in the current account balance minus the change in the fiscal balance.

20. The most significant adjustment took place during 1984–88, when the average annual current account deficit improved to about 9 percent of GDP from 13 percent of GDP during 1979–83 and average inflation was halved. An improvement in the fiscal balance—driven by a 6 percentage point reduction in the expenditure/GDP ratio—was the principal contributor to that outturn. The adjustment achieved during the period was aided by relatively favorable developments in the terms of trade. During 1989–93, there were slight improvements in average fiscal and current account deficit compared to 1984–88.13

21. Another significant adjustment in the external and fiscal balances occurred during 1994–99, following the devaluation of the CFA franc. This time, in contrast to 1984–88, the adjustment was accompanied by significant and sustained growth and progress on structural reforms. The period 2000–01 was marked by continued growth in spite of some backsliding on structural reforms and a reversal of the downward trend in fiscal and external imbalances. The authorities’ success in sustaining most of the real depreciation achieved in 1994, contributed to the better growth performance. At the end of 2001, the real effective exchange rate was at about the same level it was at the end of 1994, implying a real depreciation of about 30 percent compared to its predevaluation level (Figure 11.11).14

Figure 11.11.Senegal: Real and Nominal Effective Exchange Rates

Sources: IMF staff reports and IEO calculations.

22. There was a banking crisis in the late 1980s, brought about by severe liquidity problems that reflected government payments arrears and a sizable (and growing) share of nonperforming loans in banks’ portfolios. A regional (WAEMU-wide) banking system restructuring project, financed by the World Bank and other donors, succeeded in cleaning up the sector and strengthening banking supervision. A Financial Sector Stability Assessment undertaken by a joint IMF–World Bank team in 2001 concluded that the banking system in Senegal had recovered from the crisis and was in good health. However, it highlighted as a significant risk factor in the system the high exposure of banks to parastatals in the agriculture and energy sectors—reflecting the continuation, after two decades of programs, of problems with restructuring the groundnut sector and inefficiencies in the energy sector—with significant macroeconomic impact.

Why Was There Prolonged Use of IMF Resources?

23. Five main reasons were found for Senegal’s prolonged use of IMF resources.

24. First, the initial imbalances were large and deeply rooted in structural weaknesses of the economy that were likely to require a long time to address in a sustainable manner. The weaknesses included the vulnerability of the economy to weather and terms of trade shocks, and the heavy burden on public finances exerted by a large inefficient public sector, by extensive price controls over both consumer and producer prices, and by a heavy external debt-service burden.

25. The second reason was a broadening of objectives associated with programs supported under the IMF’s concessional facilities. The introduction of the SAF and its evolution to the ESAF, and the latter’s transformation to the PRGF, were accompanied by an elevation of growth, social policy issues, and poverty reduction, as explicit goals in programs. This evolution has been accompanied by a lengthening of the time frame within which users of these resources are expected to achieve goals specified under programs.

26. A third factor is the use of IMF arrangements as a seal of approval for the provision of external finance by several multilateral and bilateral creditors and donors.15 One example of this is the Paris Club of official creditors, which requires the existence of an IMF arrangement for its debt rescheduling agreements. Senegal has had 13 such agreements. The earlier approach to debt rescheduling—with its focus on restructuring of debt service falling due within the limited period covered by the IMF arrangement—provided only temporary respite, and required a succession of programs to continue to receive debt relief. Senegalese officials confirmed in interviews that this catalytic role of IMF arrangements was an important consideration in the country’s continuing requests for use of IMF resources. There is some evidence from internal documents that, on occasion, the “seal of approval” role was a factor in efforts by the staff to keep programs afloat when slippages occurred, and to try to work on corrective measures rather than interrupt the program. Senegal’s good standing among donors, based in part on its historical role as the administrative center of French West Africa and its tradition of regular democratic elections, may also have earned it the benefit of the doubt from time to time.16

27. Weaknesses in program design also contributed to prolonged use. In particular, the predevaluation programs were too optimistic about how effective the adjustment strategy being pursued would be in promoting growth and sustainable financial viability. For example, the successful stabilization during 1984–88 was accompanied by low growth and, in retrospect, programs during this period may have been too sanguine about the scope for achieving growth and external viability objectives without an exchange rate adjustment. Furthermore, the programs could have paid more attention to the consequences for growth of some of the measures employed to contain public sector deficits. For example, Rouis (1994) argues that a persistent focus on addressing short-term financial imbalances with ad hoc revenue measures and a lack of attention to needed structural reforms (e.g., to address tax administration and international competitiveness problems) produced a fiscal adjustment that hurt growth.17 Although there is a limit to how much an IMF arrangement by itself can address structural reforms, the persistence of problems in areas such as the energy and groundnut sectors, civil service re-form, labor market regulations, and public enterprise reform, raise questions of the effectiveness of IMF–World Bank collaboration in program design (including measures to enhance implementation prospects).

28. Finally, the stop-go pattern of program implementation weakened the effectiveness of programs and thus contributed to the continuing “need” for IMF arrangements. The wide variations in the degree of implementation under different arrangements reflected several factors. Successfully implemented programs tended to be characterized by strong up-front adjustment measures and adaptations of policies during program reviews when there were significant actual or prospective deviations from targets (usually fiscal targets).18 In the cases where implementation was weak, contributory factors usually included social and political concerns of the authorities which translated into (i) nonimplementation of agreed measures (e.g., the contingency mechanism of freezing lower priority expenditure in the event of a shortfall in government revenues during the second annual arrangement under the 1988 ESAF was not implemented because of concerns about social un-rest); (ii) delays in implementing measures (e.g., weakening in macroeconomic management and slippages in the timetable for structural reforms in late 1999 and early 2000, ahead of Presidential elections); or (iii) policy reversals, which were also often linked to the electoral cycle (e.g., suspension of the petroleum pricing mechanism in February 2000).

Effectiveness of the IMF-Supported Programs

Program design: the macroeconomic framework

29. Against the backdrop of very low (and sometimes negative) saving rates in the late 1970s and early 1980s, increasing the domestic saving rate has been a key objective in Senegal’s IMF-supported programs. This was not only to contribute toward narrowing the external current account deficit, but also to help boost investment and, ultimately, growth. The efficiency of investment was to be enhanced through various structural reforms.

Realism of key assumptions and projections in the macroeconomic framework

30. Between 1986 and 1992 (spanning two combined SBA/SAF arrangements and three annual arrangements under the 1988 ESAF), programs projected sharp drops in the external current account and the government budget deficit over the medium term. There were improvements in both balances, but outturns tended to fall short of projections. In particular, exports consistently fell short of projections, and budgetary revenues tended to be lower than envisaged. Projections of domestic saving and investment were consistently higher than the outturns, and the projected real GDP growth also tended to be higher than actual growth (Appendix 2).

31. Under the immediate post-devaluation ESAF (1994–97), the divergence between medium-term projections and outturns narrowed considerably for several variables, most notably real GDP growth, current account balance, government balance, domestic saving, and exports. Significant deviations between projections and outturns reappear under the 1998 ESAF/PRGF for the current account balance (i.e., outturns worse than projected) even though export performance was better than projected. In contrast to the earlier period, government revenues have tended to be higher than projected.

Progress toward external viability

32. The first six rescheduling agreements concluded between Senegal and the Paris Club (between 1981 and 1987) provided nonconcessional flow relief in successive program periods. Between 1989 and 2000, Senegal concluded another seven agreements with the Paris Club on increasingly more concessional terms, reflecting the evolution of Paris Club policies with respect to low-income countries.19 Throughout the 1980s, programs repeatedly indicated that Senegal would be able to stop reliance on “exceptional financing” (rescheduling and accumulation of arrears) within a few years. For most of the 1990s, programs suggested that once account was taken of traditional debt relief mechanisms, Senegal’s external debt would be sustainable. For example, as recently as 1998, a debt sustainability analysis (DSA) conducted by IMF and World Bank staffs and the authorities indicated that Senegal’s debt burden was sustainable when gauged against the thresholds under the initial HIPC Initiative. However, an updated DSA done in early 2000 indicated that the country’s debt burden was not sustainable when judged against the lower sustainability thresholds of the “enhanced” HIPC Initiative.20 Senegal reached the decision point under the enhanced HIPC Initiative in June 2000, and was, at that time, expected to reach completion point in 2002.

33. Several staff members interviewed acknowledged that the medium-term balance of payments projections and debt sustainability analyses prepared for Senegal had been influenced by an incentive to “overpromise” on the pace of restoration of sustainability that stemmed from internal guidelines requiring that there be significant progress toward external viability by the end of three-year arrangements. A second factor that contributed to this overoptimism was the heavy weight given to export-based indicators in HIPC thresholds; this focus on “external” burden indicators, rather than “fiscal” burden indicators, tended to downplay the extent of Senegal’s debt problems (Figures 11.12 and 11.13). Indeed, it was consistently pointed out in staff reports that the debt-service burden was much heavier when viewed in relation to government revenues rather than to export earnings.21 A number of Senegalese officials interviewed indicated that program limits on nonconcessional borrowing have been a useful device for instilling discipline in external debt management. The share of total debt owed to private (commercial) creditors fell from a peak of nearly 50 percent in 1978 to 8 percent ten years later (1988) and to less than 1 percent in 1999.22

Figure 11.12.Senegal: Debt-Service Ratios Before Rescheduling

(In percent)

Source: IMF staff reports.

Figure 11.13.Senegal: Debt-Service Ratios After Rescheduling

(In percent)

Sources: IMF, GFS database; and IEO calculations.

Dealing with uncertainty

34. For the major shocks that Senegal is susceptible to—droughts and terms of trade—programs have tended to deal with their effects in the context of program reviews, rather than through prespecified contingencies. The 1994 SBA and 1994 ESAF arrangement went a limited way toward prespecifying contingency measures for some terms of trade shocks (discussed below). In principle, program reviews provide greater flexibility than prespecified contingencies, but if there is not a broad understanding of how policies will respond to the major risks, they increase the chances of disagreements over, and hence delay in, the appropriate policy response that may push programs off-track.

35. Virtually all the arrangements have had some mechanism for automatic adjustment of selected quantitative financial performance criteria in the event of deviations from underlying program assumptions. Typically, ceilings on net domestic assets of the banking system and on the banking system’s net claims on the government were automatically adjusted for deviations in (i) the amount of crop credit extended by the banks; and (ii) external financing of the budget (excluding grants). There tended to be full accommodation for crop credit deviations. For external financing, there tended to be a requirement to either use “excess” amounts to reduce domestic debt (especially expenditure arrears) or to save such amounts and discuss their use during subsequent missions. Shortfalls tended to be partially adjusted for, up to prespecified amounts; shortfalls beyond these amounts were to be offset by additional measures.

36. Since 1994, programs have contained quarterly benchmarks for government revenue and the government wage bill, deviations from which were to be corrected by additional tax measures or reduction in nonpriority expenditures, in order to achieve the fiscal objectives of the programs. This feature was deemed by IMF staff to be an important device to ensure that the fiscal policy stance was appropriately supportive of the devaluation. The 1994 SBA and the 1994 ESAF (but not the 1998 ESAF) contained prespecified contingencies in the event that world prices for groundnut products and cotton turned out to be lower than projected under the programs; they required that the fiscal implications of such shortfalls be offset fully by additional revenue-raising or expenditure-reducing measures. This was too rigid an approach, since it implied that there would be no situation in which a higher fiscal deficit would be allowed to accommodate a temporary adverse terms of trade shock. Moreover, the lack of a clear indication of the types of revenue and expenditure actions increased the risks of ad hoc adjustments that would be unsustainable and inconsistent with the medium-term growth-oriented strategy.

Structural reforms

37. Senegal has made major strides in some structural reforms (notably in the areas of price liberalization, trade liberalization, and simplification of the tax system). But reforms aimed at restructuring the groundnut sector and at liberalizing petroleum product prices have proved to be intractable, resurfacing periodically with significant adverse effects on the government budget and on production incentives in the economy. Moreover, tax reforms have not yielded significant increases in revenue (measured in relation to GDP), reflecting, in part, the short-term revenue-reducing effects of some of the reforms (e.g., tariff reduction).

Restructuring of the groundnut sector

38. A lack of clarity on the aims of restructuring (expansion of production, diversification of agricultural output, or stabilization of farmers’ incomes) and sociopolitical sensitivities explain the lack of progress on restructuring the groundnut sector. The recurring issues about the groundnut sector in programs have revolved around government interventions that contribute to financial losses of the sector—namely subsidization of inputs, and the setting of producer prices above world market prices. However, there have been periods when the sector has been in surplus (reflecting favorable developments in world prices that were not passed on to producers). During 1984/85 and 1985/86, producer prices were increased in order to improve production incentives. A subsequent sharp drop in the world price of groundnut oil led to the reemergence of a financial deficit for the sector in 1986/87, just as output was responding positively to the increased incentives (and good weather). The government was reluctant to reduce producer prices so soon after they were raised, and in any case, expected that financial assistance from STABEX would compensate for the revenue loss (which it did for 1986/87).

39. Under the 1988 ESAF, the government reduced producer prices and undertook to adopt a flexible system for the determination of producer prices that would take account of developments in the world market. Under the 1994 ESAF, attention shifted toward privatization of SONACOS (the groundnut milling and marketing parastatal). After a delay of about a year from the initial target date, bids were issued in December 1995, but the authorities did not consider the bids received to be satisfactory. A second call for bids in 1997 also proved unsuccessful.23

40. After adhering to a producer pricing policy based on world prices for about four years, the government returned to a more interventionist policy in 2000. In response, the third annual arrangement under the PRGF stipulated as performance criteria a return to the pricing mechanism based on world market prices (by end-September 2001), and the withdrawal of SONAGRAINES (a wholly owned subsidiary of SONACOS) from the collection and transportation of groundnuts (by December 2001). Both measures were implemented but the dissolution of SONAGRAINES did not lead to the liberalization envisaged in the program, as the authorities continued to set indicative margins rather than allow the market to determine transport and collection costs.

41. Although the sector’s share in the economy has dwindled since the 1960s, it remains a source of income for the majority of the population in the rural areas, and the authorities regard it as a key sector in the fight against poverty. The failure to deal permanently with the problems in the sector contributed to periodic fiscal pressures, as in 2001 when the government took over obligations of SONACOS to the tune of about 2 percent of GDP. Over the years, programs have applied various types of conditionality with little lasting effect, reflecting wavering political commitment of the authorities to the reforms. The move away from the earlier ad hoc discretionary adjustments worked for a while but did not prove a permanent solution because the underlying institutional approach remained unchanged.

Petroleum pricing policy

42. As was the case with groundnuts, the early IMF-supported programs were concerned with establishing a pricing mechanism for petroleum products that would reflect world market prices and obviate the need for a subsidy from the budget; the 1983 and 1984 SBAs each had a benchmark to that effect. However, in 1986 (under the SBA/SAF), the policy changed to “mobilization of prospective surpluses of the oil sector in support of the budget,” by maintaining retail prices while import costs were falling. While this change was prompted by revenue difficulties, it damaged the credibility of the system of automatically linking retail prices to developments in the international market. Eventually, this system was reinstated in 1998, but it was suspended once again in 2000, this time in order to avoid passing on rising costs to consumers in a period leading up to presidential elections—at a cumulative cost to the 2000 and 2001 budget of about 1 percent of GDP.24 While the focus of programs on the issue was appropriate (given its macroeconomic implications), use of ad hoc discretionary deviations for revenue purposes, rather than a consistent approach to establishing the pricing mechanism, probably undermined the strategy.

Tax reform

43. Since the mid-1980s (following the recommendations of an IMF technical assistance mission in 1985), the authorities have endeavored to modernize the tax system, as well as broaden the base and increase yields. Measures undertaken through 1991 included introduction of a new tax code (with most specific duties converted to ad valorem rates), simplification of the structure of the tax system (by reducing the number of rates for import taxes and VAT), and increased coverage of VAT (to include the service sectors). However, the reforms did not yield much improvement in revenues for reasons that included pervasive exemptions and weak tax and customs administration.

44. There have been further significant reforms since the 1994 devaluation, mostly in the context of a harmonized tariff regime and domestic taxes within the West African Economic and Monetary Union (WAEMU). The 1994 and 1998 ESAF/PRGF arrangements contained several measures aimed at strengthening tax administration and expanding the tax base (e.g., computerization and expansion of coverage of VAT).

45. Conditionality (in terms of performance criteria and benchmarks) on tax reform measures was virtually absent in programs until 1997. Since then there have been several, including on the elimination of hundreds of tariff lines, the establishment of a large taxpayer unit, the implementation of a single taxpayer identification number in all revenue collecting agencies, and a single rate VAT.25

46. In the event, the revenue effort did eventually show some improvements—rising from an average of about 16.5 percent of GDP in 1994–99 to 18.1 percent in 2000–01 in spite of a lowering of tariffs in the context of a common external tariff in WAEMU (Figure 11.14).26 However, greater emphasis in the early programs on improving tax administration and reducing exemptions could have yielded significant additional benefits and would have reduced the reliance on various ad hoc revenue and expenditure measures at times of fiscal pressure.

Figure 11.14.Senegal: Composition of Government Revenue

(In percent of GDP)

Source: IMF staff reports.

Social policies

47. While several programs contained social safety net measures to cushion the impact on the poorest groups of some price increases, problems with targeting of the measures limited their effectiveness. Two examples are:

  • (i) Under the 1988 ESAF, a reduction in the producer price of groundnuts was a key measure for eliminating the deficit of the sector. In order to provide some relief to farmers, administered prices of selected key consumer goods (rice, sugar, groundnut oil) were subsidized.

  • (ii) Both the post-devaluation SBA (1994) and the subsequent ESAF arrangement that replaced it contained budgetary provisions for subsidizing the price of a number of products deemed “sensitive” for low-income households (e.g., bread, rice, medicines).

48. Neither of the two programs was well targeted. An internal World Bank evaluation of the Bank’s Economic Recovery Credit (approved in March 1994 to provide emergency support to the post-devaluation reform program) concluded that the subsidies designed to limit the impact of price increases of basic consumer goods had been poorly targeted and had not delivered the expected benefits to the most vulnerable groups.

Collaboration with the World Bank

49. Program documents presented to the IMF’s Executive Board convey the impression of close col laboration between the staffs of the IMF and the World Bank, especially on sectoral issues (e.g., the banking system, agriculture, energy, and industry). Discussions with staff members of the two institutions confirm that there have always been regular contacts and exchange of information, including in areas not highlighted in staff reports (e.g., the preparatory work that informed the staffs’ advice to the authorities of the CFA franc zone in the period leading up to the 1994 devaluation). A recent staff report clearly described the division of responsibilities between the two institutions.27

Box 11.1.Selected Lessons from World Bank Evaluations

The World Bank employs a range of evaluation reports to take stock of achievements and failures, and to draw lessons from experiences under its lending operations. At the completion or termination of such operations, a “project completion report” or “implementation completion report” is prepared. The Operations Evaluation Department (OED) may comment on these reports, and/or undertake a “performance audit” of its own. This box draws on a range of such evaluation reports on the four SALs and the Agricultural Sector Adjustment Credit (AGSAC) extended to Senegal, since a number of the lessons highlighted are also relevant for the IMF’s operations.

  • Under the first three SALs, incomplete and tenuous links between proposed measures and objectives gave rise to overestimation of likely performance and underestimation of constraints (e.g., in agriculture and industry).

  • Notwithstanding significant achievements, including transformation of the economy away from excessive state intervention, about one-third of measures envisaged under SAL II and SAL III were not implemented as scheduled. These measures were concentrated in areas—labor regulations, parastatal reform—where there was strong political opposition from vested interests. The lesson drawn was that politically sensitive reforms are unlikely to be implemented unless a prior internal debate has taken place and consensus reached on key issues. It is important to encourage internal agreement on key issues, and based on that formulate upfront measures to be implemented under programs.

  • A basic design flaw in SAL IV was expecting a positive supply response to improvements in regulations and incentives at a time when demand compression was being relied on to achieve external equilibrium. By the time of SAL IV (1990) work undertaken in the World Bank suggested that devaluation would be essential for restoring competitiveness and boosting growth.

  • Lack of consensus among donors and the government on the strategy for restructuring the groundnut sector has contributed to the slow pace of reform. In retrospect, too much emphasis may have been placed on privatization of SONACOS, at the expense of other reforms to improve the efficiency of the sector.

50. Nevertheless, there have been some problems in matching the time frame and priorities of the two institutions, which affected the timeliness of World Bank inputs on some sectoral issues in IMF-supported programs. Some of the differences in the timetable may be related to the World Bank’s attempts to encourage internal consensus on structural reforms before going ahead with lending programs (Box 11.1).

51. Several Senegalese officials interviewed emphasized the need for better collaboration between the IMF and the World Bank, especially with respect to the structural reform and social policy content of programs. The PRSP and the recent formalization of the “lead agency” approach are the major instruments for improving such collaboration, although in the case of Senegal it is too early to judge their impact.

Ownership, Conditionality, and the PRSP Approach

52. The evaluation sought views on what impact, if any, prolonged engagement in IMF arrangements has had on (i) “ownership” of programs and on the economic policymaking process; (ii) capacity building in institutions responsible for formulating and implementing macroeconomic policies; and (iii) the effectiveness of IMF conditionality.

Ownership and capacity issues

53. Interviews with many Senegalese officials and with IMF staff indicate a striking difference in perceptions about the degree to which policies had been “imposed” on the authorities without sufficient consultation. However, there was broad agreement that the successful programs had been those to which there was strong domestic political commitment, regardless of the precise nature of IMF conditionality. The main lesson drawn by the World Bank from a review of its structural adjustment loans to Senegal was that politically sensitive reforms are unlikely to be implemented unless there has been prior internal debate and consensus building on key issues (see Box 11.1).

54. A number of Senegalese officials who have participated in negotiations with the IMF over time thought that the level of “country ownership” of programs has generally been low. In their view, the “seal of approval” role of IMF arrangements in unlocking other sources of financing gave the IMF the upper hand in negotiations with the authorities, and that sometimes the authorities went along with these proposals—even though they had doubts about their ability to deliver on implementation—in order to secure urgently needed resources. Several indicated that, more often than not, they were in agreement with the IMF’s diagnoses of the country’s financial problems, but sometimes differed on the pace of implementation of measures. In their view, the IMF tended to underestimate implementation constraints and overestimate the speed with which some structural reforms (e.g., privatizations) could be brought to conclusion. However, a few officials put the onus of responsibility for the content of programs and for implementing them on the authorities (including themselves), noting that while it was convenient to blame the IMF for implementation failures, the primary responsibility for difficulties in addressing some well-recognized problems belonged at home.

55. The IMF staff appears to share the characterization of weak ownership of programs, at least with respect to structural reforms: the staff report for the 2001 Article IV consultation noted that the new government that took office in April 2000 “inherited a legacy of weak implementation and ownership of the structural reforms program, particularly the longstanding structural problems of the groundnut and energy sector.”

56. A number of staff interviewed thought that officials tended to underestimate the scale of quasi-fiscal activities and the risks such activities pose to macroeconomic stability. However, they noted that there had been periods when senior Senegalese officials clearly were convinced of the merits of addressing these issues in IMF-supported programs and had been able to persuade the highest political authorities to implement difficult reforms. They accepted that there was a need to bring greater transparency to discussions with officials and other stake-holders about financial constraints in the broadly defined public sector, highlighting the risks of disorderly adjustment in the absence of measures to tackle the financial problems. It appears that the active involvement of senior political leaders in program negotiations can help increase political commitment to programs.

57. With regard to the effects of “prolonged use” on the policymaking process, there was broad recognition by those interviewed in Senegal that the IMF had contributed importantly to an “internalization” of the need for prudent fiscal management. However, many pointed to problems in the way programs had been negotiated, notably that they (i) tended to undermine the policymaking processes of the government, especially parliament’s role in economic decision making; and (ii) provided an excuse for government to stifle domestic policy debate.

58. On capacity building, many officials interviewed thought there had been significant knowledge transfer from IMF staff at the technical level. At the same time, they felt that a lack of flexibility in the formulation of IMF-supported programs had discouraged local initiative, and that the system had become too dependent on the IMF for diagnosing and proposing remedies to the country’s macroeconomic problems. Several also expressed regret that an excessive focus on short-term solutions to financial problems had led government officials to abandon medium-term “planning” and long-term strategic thinking about Senegal’s development challenges. The response of staff to these criticisms was that programs are usually constructed in a medium-term framework with growth as one of the key planks, and that the PRSP process had been designed to help improve the integration of programs with the longer-term strategy into PRGF arrangements (see below).


59. The evolution of the IMF’s concessional facilities has been marked by significant changes in conditionality: from the low-access, low-conditionality SAF; to the relatively high-access, high-conditionality ESAF; to the PRGF, which has been accompanied by efforts to streamline structural conditionality but at the same time has introduced some new conditionality (e.g., preparation of PRSPs through a participatory process). The average number of structural conditions (prior actions, performance criteria, and benchmarks) per program year increased steadily in successive multiyear ESAF/PRGF arrangements, from 4 in the 1988 ESAF to 10 in the 1998 ESAF/PRGF (Figure 11.15).28

Figure 11.15.Senegal: Structural Conditionality Under ESAF/PRGF Arrangements

(In number of conditions per program year)

Sources: IMF Policy Development and Review Department and IEO calculations.

60. With regard to the impact of “prolonged use” on the effectiveness of conditionality, views ranged from “conditionality had lost its bite over time” to “effectiveness of conditionality had been strengthened, as officials have learned from experience the importance of compliance.” On balance, the evaluation found no clear evidence of weakening; there was some evidence of “elevating” targets from benchmarks to performance criteria and prior actions in response to implementation slippages. However, as noted earlier, in the discussion of several “intractable” structural problems, a “hardening” of conditionality does not appear to have been successful in fostering permanent changes if political commitment was weak. For example, prior actions were effective in the early programs in bringing about once-off discretionary changes (several increases in administered prices were effected before programs were presented to the Board), but they did not permanently change the price-setting systems. Prior actions have also been extensively used recently in conjunction with completion of reviews for measures whose implementation had been delayed.29 But they cannot ensure that there will be no policy reversals, as demonstrated by the abandonment of the mechanism for setting petroleum product prices in the lead-up to elections in 2000 when world prices were on the rise.

The PRSP approach

61. In many respects, changes under way at the IMF—most notably, the transformation of the ESAF to the PRGF (and related to that, the PRSP approach), the increased emphasis on country “ownership” of programs, and attempts to streamline structural conditionality—already have begun to respond to many of the criticisms of the IMF that the IEO mission heard in Senegal (and reported above).

62. Most of those interviewed in Senegal for this evaluation welcomed the PRSP process as an appropriate way of promoting country ownership and hence increasing the likelihood of consistent imple-mentation of IMF-supported programs. A common view was that it goes some way to address criticisms of the process of formulating previous programs. However, most argued that the jury was still out on whether the PRSP (which has only recently been completed) will significantly influence the policy content of the next PRGF arrangement. A number of people interviewed stressed the key role IMF resident representatives can play in the implementation of the PRSP approach by widening the range of stakeholders with whom the IMF interacts on a regular basis.

IMF Internal Governance Issues

Were IMF policies to contain “prolonged use” applied to Senegal?

63. The IMF has developed policies aimed at containing the phenomenon of prolonged use, the main elements of which are (i) reduction in access levels (in relation to size of IMF quota); (ii) front-loading of adjustment effort and close monitoring of program implementation; (iii) comprehensive ex post assessment of programs; and (iv) the formulation of “exit” strategies.30

64. Average annual access levels in Senegal’s IMF arrangements have followed a declining trend, although the country has enjoyed above-average access levels in its three ESAF/PRGF arrangements (Figure 11.16).31 Over the course of the 1998 ESAF/PRGF, there was a move toward closer monitoring (more reviews in annual arrangements), and conversion of measures that had not been implemented at the time of review missions into “prior actions” for the completion of reviews.

Figure 11.16.Senegal: Average Annual Access Levels Under IMF Arrangements

(In percent of quota)

Source: IMF staff reports.

65. With regard to ex post assessments, a country strategy paper prepared in early 1998, before the negotiations for a new three-year ESAF, took stock of performance under previous arrangements and set out an agenda of remaining challenges to be tackled under the new arrangement. However, it had little discussion of lessons from the nonimplementation of earlier reforms. Nor did it have a forward-looking medium-term macroeconomic framework and a discussion of alternative scenarios and policy options. Another ex post assessment, this time limited to Senegal’s pre-devaluation IMF-supported programs, was contained in the background paper to the 1995 Article IV consultation. It contained a comprehensive analysis of the relative roles of shocks and policies in Senegal’s economic performance from the late 1970s through the early 1990s. However, it is not clear what impact the analysis had on the design of subsequent programs.

66. One set of “stepping-back” exercises that definitely influenced the design of a subsequent program was work done during 1992–93, some of it in close collaboration with the World Bank that provided advice given to CFA franc countries prior to the 1994 devaluation. The key issues tackled related to the limits of a purely “internal” adjustment strategy for dealing with loss of competitiveness and low growth. The analysis presented in internal documents was significantly more comprehensive than that contained in subsequent staff papers to the Board. However, we were told that the Board was briefed at greater length in an informal session.32

67. We found no evidence that the issue of an “exit strategy” from the use of ESAF/PRGF resources had been discussed in any systematic manner within the staff (e.g., in the internal review process) or with the authorities.33 On the occasion of Board consideration of the request for the 1998 ESAF arrangements, one Executive Director expressed the hope that the third set of ESAF arrangements would represent an “exit” arrangement for Senegal. The absence of systematic discussion of the issue of exit from the use of concessional facilities may have reflected an implicit acceptance that the deep-seated structural problems of Senegal would require its continued use of IMF resources for some time—but it also probably reflected the fact that overall, IMF-wide criteria and strategies for exit are not well defined in such cases.

Surveillance and program activities

68. To assess whether program activities have “crowded out” surveillance, the evaluation examined how staff reports on Senegal measured up against a checklist of “best practices” inferred from various internal IMF guidelines on reporting on Article IV consultations.34 Virtually all such consultations with Senegal have tended to be combined with either requests for arrangements or program reviews during 1980–2001; there were only two stand-alone Article IV consultation discussions at the Board (out of a total of 16) during this period.

69. For a persistently active program country like Senegal, the line between program and surveillance activities becomes blurred; indeed, an effective program design process should cover all the issues on the surveillance “best-practice” checklist. But regardless of how and in what context, there should be opportunities to step back periodically from the details of designing and monitoring compliance with program targets, so as to review the overall adjustment strategy underpinning programs. A natural point to do this for “prolonged users” is between arrangements and in the context of surveillance. In this context, the main points from the review are as follows:

  • Staff reports generally did a reasonable job on the presentation of medium-term scenarios, but sensitivity analyses tended to consider relatively small shocks—in other words, they did not “stress test” the consequences of markedly different, but still possible, outcomes. Nonetheless, the reports from the mid-1980s through the early 1990s demonstrated how even small adverse changes in the external environment could significantly delay the attainment of external viability. The sensitivity analyses in recent reports have tended not to include any stress testing of major vulnerabilities and downside risks. Consequently, they have been somewhat perfunctory.

  • In general, there was relatively little discussion of risks of deviations from assumptions and projections, and how programs might be adapted. A handful of reports contained candid discussions of the risks of policy slippages and policy reversals.35

  • Reports usually provide an indication of the authorities’ views; in most cases, however, there were no reports of significant differences and so no sense of any compromises that might have been reached between the staff and the authorities. Two exceptions to this were the 1992 Article IV staff report (which covered a period when there was no IMF arrangement in place), and the 2001 Article IV consultation report (which indicated differences on the timing of an increase in electricity tariffs).

  • Presentation of the reform strategy and reviews of performance under previous or ongoing programs were broadly satisfactory. However, discussion of alternative options and trade-offs was almost completely absent from reports, except in the 1992 report, which presented an account of the authorities’ internal adjustment efforts.

  • Little detail was provided in reports about the substance of the collaboration with the World Bank; often there were statements affirming close collaboration and—as is general practice—an appendix listing World Bank operations in Senegal, but with no sense of how the strategies of the two institutions were integrated and what the mutual priorities were. In a few cases, where there had been a recent World Bank operation with objectives that were similar to those in IMF arrangements—for example Structural Adjustment Loans—a summary of the content of the authorities’ Letter of Development Policy to the World Bank was provided.36

70. In terms of topics covered, it is striking how little explicit discussion of exchange rate misalignment there was in the reports before the 1994 devaluation, with most of the focus on traditional measures of the real effective exchange rate.37 This was explained by the staff as the result of a deliberate decision by management to keep discussion of a possible devaluation of the CFA franc confidential. As discussed above, the issue was analyzed extensively in internal documents during 1992–93.

Conclusions and Recommendations

71. Five main reasons were found for Senegal’s prolonged use of IMF resources.

  • First, the initial imbalances were large and deeply rooted in structural weaknesses of the economy, including vulnerability to terms of trade and weather-related shocks, which were likely to require a long time to address in a sustainable manner.

  • Second, Senegal’s prolonged use of IMF resources can also be attributed to the broadening of objectives associated with programs supported under the IMF’s concessional facilities.

  • A third factor is the use of IMF arrangements as a seal of approval for the provision of external finance by several multilateral and bilateral creditors and donors.

  • Fourth, overoptimism about the effectiveness of the pre-devaluation adjustment strategy on promoting growth may have contributed to prolonged use.

  • Weaknesses in program design that were compounded by a stop-go pattern of program implementation, especially with regard to structural reforms, undermined program effectiveness.

Ex post assessments and exit strategy

  • We recommend that guidelines on ex post assessments be applied more consistently and forcefully for countries that have had a succession of multiyear arrangements.

  • A more explicit strategy for eventual exit from use of IMF resources is needed. Attainment of the completion point under the enhanced HIPC Initiative would be one possible opportunity for discussion of such a strategy for Senegal, drawing on the assessments of previous programs.

Program design

  • Successful programs have been those to which there was strong political commitment, regardless of the precise nature of conditionality.

  • Key lessons for improving the effectiveness of Senegal’s IMF-supported programs include (i) fostering greater ownership of reform programs; (ii) reflecting capacity constraints more realistically in implementation schedules; and (iii) articulating measures for improving implementation—including through a greater flow of technical assistance on implementation aspects.

  • Programs need to be set in a realistic time frame that avoids overpromising on the pace of adjustment toward external viability; if the adjustment is realistically expected to extend well beyond the period of a three-year arrangement, then the consequences for IMF involvement should be discussed explicitly as part of a medium-term strategic framework.

  • To help programs adapt better to uncertainty, the principal risks to the program, and how policies would adapt to those risks, should be set out explicitly.

  • There is room for improving Bank-Fund collaboration to address fundamental institutional/structural issues at the heart of prolonged use. The PRSP provides an instrument to that end, but will need to be matched by operational changes to en-sure that the work priorities and time frames of the two institutions mesh effectively.

  • In countries like Senegal where quasi-fiscal activities are significant, the macroeconomic framework needs to focus on a broad coverage of the public sector accounts (instead of focusing mainly on central government balances), in order to provide a more comprehensive and more transparent basis for considering public sector financial problems, and for discussing risks of disorderly adjustment in the absence of corrective measures.


  • Surveillance discussions and reports should be used as an occasion to “step back” and reconsider the overall strategy. While such reconsiderations should be a normal part of any effective program process, a natural time to reconsider the overall strategy is between arrangements and in the context of surveillance. Such a strategic review should include an assessment of performance under previous programs, including a frank appraisal of previous assumptions about implementation; a stress-testing of major vulnerabilities and downside risks; and a discussion of alternative policy options and trade-offs. The results of such reassessments should be reported candidly to the Executive Board.


  • Persistence of criticisms of the IMF, in spite of its embrace of the PRSP approach, suggests that many people do not believe that the IMF has changed the way it does business. The IMF has increased its outreach activities in Senegal in recent years, including through the role of the resident representative. These efforts need to be sustained and broadened, including more public discussion of the rationale for policy advice, especially on sensitive issues.

Appendix 1 Senegal: Implementation of IMF Arrangements
Arrangements1Financial and structural PCs and benchmarks2Program reviewsSlippages and proximate causesShocksComments
SBA I (1979)No formal PCs, but significant deviations from indicative targets.None.
EFF (1980/81-1982/83)Observed at first test date only.First review was not completedNonimplementation of price liberalization measures and lack of agreement on policy adaptations; political changes took focus away from economic policy.Drought.President Senghor (President since independence in 1960) retired in December 1980; replaced by Mr. Diouf (Prime Minister at the time).
SBA II (1981/82)Original program set PCs for first half of program year only. These were observed.Midterm review completed four months behind schedule.Delay in completion of midterm review. The original program stipulated that additional measures to reduce projected fiscal deficit be agreed during midterm review mission and implemented in January 1982.Good weather.

Decline in world price of groundnut oil.
Original program envisaged only one (midterm) review. Delay in completing thatreview led to modification of arrangement to add a second review and rephase disbursement.

Decision to modify rather than cancel arrangement partly influenced by link to Paris Club agreement.
SBA III (1982/83)Most not observed.Midterm review was not completed.Tax measures envisaged for November 1982 not implemented; election pressures.Sharp downturn in world price of groundnut oil.Presidential and legislative elections held in February 1983.
SBA IV (1983/84)Observed.Midterm review was completed on time.Severe drought.

Higher-than-expected export price for groundnut products (partly offset large fall in output and export volumes).
Measures implemented before Board approval included significant increases in prices of several consumer goods.

Expenditure cuts more than compensate for revenue shortfall.

Increase in some petroleum product prices to avoid deficit in National Energy Fund
SBA V (1984/85-1985/86)Almost all observed.Envisaged reviews (three) were completed on time.Improved weather conditions.Significant policy adaptation in response to shortfalls in export earnings and government revenue (e.g., extension of service tax to telecom services, tightening of credit ceiling).
SBA VI and SAF_1 (1986/87)All PCs under SBA observed; nearly all benchmarks under SAF observed.Midterm review under SBA completed on time.Delay in formulation of action plan for determination of producer prices of cereals (benchmark under SAF).Sharp decline in export price of groundnuts; decline in import price of petroleum products.Windfall profits from petroleum sector and from rice imports mobilized for the government budget.
SBA VII and SAF_2 (1987/88)All PCs under SBA observed. Benchmarks under SAF observed, with slight delay in one.Midterm review under SBA completed on time.Slight delay in removal of remaining quantitative restrictions on imports (benchmark under SAF).Good weather conditions.

Sharp deterioration in the terms of trade.
Presidential and parliamentary elections held in February 1988.
ESAF I_1 (1988/89)Observed.Midterm review completed on time.Slippages in implementation of tax reform measures in the second half of the year (after midterm review).Locust infestation of crops.

Uneven distribution of rainfall
Social unrest in second half of program period.
ESAF I_2 (1989/90)Observed.Midterm review completed on time.Slippages after midterm review:

—delay in implementing tax reform measures (expanding VAT base).

—contingency mechanism for freezing lower priority expenditures, in the event of a revenue shortfall, not implemented.
Measures to increase tax revenue implemented prior to approval of arrangement.

Tighter monitoring: introduction of monthly revenue targets, and a PC on minimum revenue collection.
ESAF I_3 (1991)Observed.Slight delay (one month) in completion of midterm review.Weakening of performance after midterm review.

Delays in implementation of structural measures.
Sharper drop in export.

Poor weather conditions.
Request for third annual arrangement was delayed until successful implementation of six-month staff-monitored program
SBA VIII (1994)All PCs for March 1994 were met, except accumulation of new arrears.Only the purchase associated with approval of the arrangement was made before the SBA was replaced by an ESAF.
ESAF II_1 (1994/95)Observed.Midterm review completed slightly behind schedule (two months).
ESAF II_2 (1995/96)Most observed.Midterm review completed slightly behind schedule (two months).Delays in reform of energy sector.

Slow pace of privatization program.
ESAF II_3 (1997)Most observed.Midterm review completed slightly behind schedule (two months).Delays in public enterprise and energy sector reforms; resistance from interest groupsUnfavorable weather conditions.
ESAF III_1 (1998)Most observed.Midterm review completed slightly behind schedule (one month).Drought.

Declining oil prices.
ESAF III_2 (1999/2000)Most of targets prior to December 1999 test date were met. Thereafter, most structural and some financial benchmarks were not observed.Completion of first review delayed by six months.

Envisaged second review was not completed.
Slippages in the implementation of fiscal program and structural reforms in the run-up to presidential elections.Presidential elections held in February and March 2000; won by opposition candidate (Mr. Wade). New dates for meeting structural benchmarks agreed with new government.
ESAF/PRGF III_3 (2001)Several PCs related to second and third reviews were not observed.Delays of two months and four months for first and second reviews, respectively.

Envisaged third review was not completed.
Eight months elapsed between completion of review under second annual arrangement and approval of year (first, second, or third) arrangement.

Legislative elections held in April 2001.

Implementation of some corrective measures allowed completion of second review, but not the third review, just before the expiration of the arrangement.

Roman numerals are used to indicate the sequence of arrangements, by type. For multiyear arrangements, number suffix refers to arrangement year (first, second, or third).

PCs = Performance criteria.

Roman numerals are used to indicate the sequence of arrangements, by type. For multiyear arrangements, number suffix refers to arrangement year (first, second, or third).

PCs = Performance criteria.

Source: IMF staff reports.
Appendix 2 Senegal: Projections and Outturns of Selected Variables
(In percent of GDP, unless otherwise indicated)
Current account

deficit (excluding

official transfers)
Export growth

(in percent

per annum)


Real GDP growthGross domestic

Gross domestic

Medium-term projections2
ESAF I1988/899.110.0–0.911.813.4–1.60.2–2.12.317.716.80.94.2–0.54.710.
ESAF II19949.–3.0–0.7––0.315.116.2–1.1
ESAF/PRGF III19987.–1.1–1.0–0.3–0.715.616.8––0.413.810.43.419.217.51.7
Annual projections3
Source: IMF staff reports.

Projection minus outturn.

Projections in the document for the initial request for the multiyear arrangement.

Updated projections for the year.

Source: IMF staff reports.

Projection minus outturn.

Projections in the document for the initial request for the multiyear arrangement.

Updated projections for the year.

Appendix 3 Senegal: List of People Interviewed in Connection with the Evaluation of the Prolonged Use of IMF Resources

Current Senior Officials

Mr. A. Boucar, Deputy Governor, BCEAO

Mr. Ababacar Diop, Cabinet Director

Mr. Serigue Babacar Diop, Directeur de Cabinet, Ministère des Mines

Mr. Pape Diouf, Minister of Agriculture

Mr. A. Fall, President, Power Sector Regulatory Commission

Mr. Alioune Gassama, Director of Analyses, Forecasting and Statistics

Mr. Ambroise Koné, Director of Research, BCEAO

Mr. Seyni N’Diaye, National Director, BCEAO

Mr. Souleymane Saib, Director of Administrative and Financial Affairs

Mr. Macky Sall, Minister of Energy

Mr. Abdoulaye Sene, Technical Advisor

Mr. Chiekh Soumare, Minister of Budget

Former Senior Officials

Mr. Serigne Lamine Diop, Minister of Finance

Mr. Mamadou Lamine Loum, Prime Minister, Minister of Finance

Mr. Famara Ibrahim Sagna, Minister of Finance

Mr. Djibril Sakho, National Director, BCEAO

Mr. Magatte Pathé Sené, National Director, BCEAO

Mr. Mamoudou Touré, Minister of Finance

Representatives of Political Parties

Mr. Amath Dansokho, Parti pour l’Indépendance et le Travail (PIT)

Mr. Sakhma Diouf Faye, PIT

Mr. Sémun Pathé Gnèye, PIT

Mr. Makhtar Mbay, PIT

Mr. Ibrahima Sène, PIT

Mr. Dibo Ka, Union pour le Renouveau Démocratique

Mr. Amath Wade, PIT

Mr. Landing Savane, Jef/PADS


Mr. Abdoulaye Diagne, Director of Research, Economics Faculty, Cheik Anta Diop University

Business Community

Mr. Amath Benoit Gaye, Secretary General, National Union of Chambers of Commerce, Industry & Agriculture

Mr. Nor Talla Kane, Executive Secretary, Conseil National des Employeurs du Senegal

Mr. Mamadou Lamine Niang, President, National Union of Chambers of Commerce, Industry & Agriculture


Mr. Abdou Diarra, Le Matin

Mr. Malik Diaw, Le Soleil

Ms. Aissatou Fall, Walf-FM

Mr. Mounirou Fall, Sud Quotidien

Mr. Amadou Gueye, Le Journal de l’Economie

Mr. Josias Toba Tanama, Le Journal de l’Economie

Nongovernmental Organizations

Mr. Mame Gueye, President, Civil Forum, National Chapter of Transparency International

Mr. Jacques Habib Sy, Founder, Aid Transparency

Trade Union Representatives

Mr. Pape Diallo, UNSAS

Mr. Madia Diop, President, CNTS

Mr. Abdoulaye Gueye, President, Union Nationale des Syndicats Autonomes du Sénégal (UNSAS)

Mr. Mody Guiro, Secretary General, Confédération Nationale des Travailleurs du Sénégal (CNTS)

Mr. Jhohume Kout, Secretary General, AGHTS

Mr. Waly Ndiaye, Confédération des Syndicat Autonomes

Mr. Mademba Sock, Secretary General, UNSAS

Representatives of Donors

Mr. Jean de Gliniasty, Ambassador of France

Mr. Matar Fall, Senior Sanitary Engineer, World Bank Acting Country Director, and water sector lead specialist

Mr. Alain Frossard, Conseiller économique et commercial régional, French Embassy

Mr. Xavier Rose, Counselor, French Embassy

Mr. Jean-Luc Supera, Country Director, Agence Française de Développement

Mr. Daniel Voizot, Chef de la Mission de Coopération, French Embassy

Mr. Richard Young, Economic Adviser, European Union Delegation

The mission also met with a large number of current and former IMF and World Bank staff involved with these institutions’work on Senegal.

The authorities are expected to request a new PRGF arrangement as part of the process

toward reaching completion point under the enhanced HIPC Initiative.

Including loans from special facilities—Oil Facility and the Compensatory Financing Facility—which did not require a formal arrangement to be in place.

Senegal is a member of the eight-country West African Economic and Monetary Union (WAEMU), which together with the six-member Central African Economic and Monetary Union and the Comoros form the CFA franc zone. Each of the three parts of the zone has its own central bank. The other members of WAEMU are Benin, Burkina Faso, Côte d’Ivoire, Mali, Niger, Togo, and Guinea-Bissau (which joined in 1997).

The pricing policies that led to problems were (i) not passing on to consumers increases in import costs of several consumer goods and (ii) setting the producer price of groundnuts above world market prices.

The 1980 EFF was an attempt at a comprehensive approach to tackling Senegal’s financial imbalances using the IMF’s only available “structural” facility at the time; the authorities had expressed a strong preference for an EFF over an SBA. In the event, the EFF was short-lived and was replaced by a succession of SBAs.

The CPSP was responsible for administering producer prices for agricultural products, notably groundnuts and cotton. Its financial position went from substantial surpluses in the 1970s to large deficits in the early 1980s.

Preparation of the PRSP is one of the conditions for reaching completion point under the enhanced HIPC Initiative.

A program performance audit report of the SAL prepared by the World Bank’s Operations Evaluation Department (OED) reported that “full cooperation between the staffs of the Bank and the Fund was achieved” during the preparation of the SAF and the EFF and that some of the performance targets in the EFF and the SAL were identical or very similar. The authors did not think this overlap in conditionality was appropriate, arguing that failure to meet “short-term IMF performance criteria” should not automatically lead to disruption of a SAL.

The staff report on the 1992 Article IV consultation noted that “it had not been possible to reach understandings with the authorities on the required set of strong measures that would permit a resumption of credible adjustment, and thus establish a firm basis for a new Fund-supported program.”

The data on which the tables and graphs in this section are based are mostly from the IMF’s World Economic Outlook (WEO) database supplemented by data from the following databases: the African Department’s WETA database, International Financial Statistics (IFS), and the Information Notice System (INS).

The “impact of terms of trade shocks” measures the effect (in percent of GDP) of annual changes in export and import prices, holding trade volume constant. See McCarthy, Neary, and Zanalda (1994).

A summary of the factors considered in coming to a judgment on each of Senegal’s IMF arrangements, including annual arrangements within multiyear arrangements, is presented in Appendix 1.

This comparison of averages obscures trend improvements that were partially reversed, especially in 1993.

NEER and REER are indices of the nominal effective exchange rate and the real effective exchange rate, respectively.

During the period under review (1979–2001), presidential and legislative elections were held in 1983, 1988, and 1993. In line with changes in electoral laws, presidential elections were also held in 2000, and legislative elections in 1998 and in 2001. Governments of national unity (which included members of opposition parties as cabinet ministers) were formed in 1991 and 1994 in efforts to diffuse rising social and political tensions in the country.

Rouis (1994) and Tahari and others (1996) provide comprehensive analyses of economic performance in Senegal during this period.

Examples include the 1983 and 1985 SBAs and the 1994 ESAF (see Appendix 1). The up-front measures in the earlier programs included increases in the administered prices of consumer goods and petroleum products.

Three were on Toronto terms (one-third reduction in the net present value of eligible debt), one was on London terms (50 percent reduction), two were on Naples terms (two-thirds reduction), and the last was on Cologne terms (up to 90 percent reduction).

Under the original HIPC Initiative, the sustainable threshold was the range of 200–250 percent for the ratio of NPV-of-debt to exports. Countries with very open economies (export/GDP ratios of 40 percent or higher) and that had government revenue/GDP ratios of at least 20 percent were eligible for assistance if the ratio of their NPV-of-debt to government revenue exceeded 280 percent. Under the enhanced HIPC Initiative, the threshold for the NPV-of-debt to export ratio was lowered to 150 percent (no longer a range), and the requirements for eligibility for assistance through the fiscal window were changed to an export/GDP ratio of 30 percent, government revenue/GDP ratio of 15 percent, and NPV-of-debt/revenue ratio exceeding 250 percent. For further details, see, for example, Andrews and others (1999).

Foreign exchange restrictions are sometimes cited as a reason for favoring export-based debt burden indicators over GDP- and government revenue–based indicators (see, for example, “Staff Response to the External Evaluation of the ESAF” (IMF, 1998). For Senegal, the convertibility of the CFA franc removes this concern, and allows other factors to come to the fore (e.g., dominance of public debt in total debt, a “reasonable” revenue/GDP ratio, and considerations of debt service crowding out “productive” government spending).

The ratios were calculated from the World Bank’s Global Development Finance database.

23A World Bank evaluation of performance under the 1995 agriculture sector credit noted that SONACOS was not privatized because the conditions imposed by the government (e.g., requirement to provide seeds and fertilizers to farmers on credit and to maintain the integration of the industry) were not attractive to potential investors.

The pass-through mechanism was reinstated in June 2001.

Most of the measures were specified as performance criteria, except the establishment of the large taxpayer unit, which was a benchmark. Most were observed on time or with slight delays. One measure that was delayed and that became a prior action for completion of a review was the introduction of the single VAT rate, which was implemented in September 2001 (instead of May).

Figure 11.14 breaks down government revenue into: (i) income and property taxes; (ii) taxes on domestic goods and services; (iii) taxes on nonpetroleum imports; (iv) petroleum revenues; and (v) other revenues. For the period 1981/82–1982/83, petroleum revenues are distributed between import taxes and “other” revenues.

Box 4 in EBS/02/50, March 21, 2002.

The increase in the 1998 ESAF/PRGF partly reflects the effect of carrying over missed performance targets from one annual arrangement to a successor annual arrangement.

The reintroduction of the pass-through mechanism for petroleum product prices and the submission to parliament of a draft bill introducing VAT at a single rate were performance criteria for May 2001 that were missed. They became prior actions for the completion of the first review under the third annual PRGF which was completed in September 2001.

These policies are discussed at greater length in Chapter 3 of Part I

Access under the 1988, 1994, and 1998 ESAFs was 90 percent, 81 percent, and 66 percent, respectively. This compares with IMF-wide averages for first, second, and third three-year ESAF/PRGF arrangements of 85 percent, 72 percent, and 65 percent. Current access norms under the PRGF are 90 percent of quota over three years for first-time users and 65 percent of quota over three years for second-time users.

The staff report for the 1992 Article IV consultation contained some discussion of the issue, in rather general terms, leading to some discussion at the Board meeting of the appropriateness of using a purely “internal” adjustment strategy to address external competitiveness issues in Senegal and other CFA franc countries.

The issue of IMF long-term involvement in Senegal arose during the process of clearing briefing papers in 1987. The remedy for the “problem” of the country’s prolonged use of “regular” IMF resources was to shift to arrangements under a concessional facility (at the time, the SAF).

The factors considered were (i) presentation of alternative medium-term scenarios; (ii) quality of sensitivity analyses; (iii) discussion of risks of deviations from assumptions and projections; (iv) discussion of risks and impact of policy slippages; (v) reporting on the views of the authorities; (vi) cogent presentation of proposed/recommended policy course; (vii) discussion of pol-icy alternatives and trade-offs; (viii) critical and frank review of performance under (previous and ongoing) IMF-supported program(s); and (ix) account of collaboration with the World Bank.

Examples include the report on the midterm review under the third annual arrangement under the 1988 ESAF and the 2001 Article IV consultation report.

For example, in the reports for 1983, 1988, 1989, and 2001.

The staff report for the 1989 Article IV consultation indicated that the authorities held the view that “the benefits of full convertibility and stability of the CFA franc afforded by membership of WAMU outweighed the potential advantages of a more flexible exchange rate policy.” The 1991 Article IV staff report elaborated on the benefits (promotion of confidence in the currency/country and contribution to domestic price stability—essential factors for attracting foreign investment and mobilizing domestic saving), and on how the authorities intended to support the system (appropriate financial and structural policies, notably a restrained incomes policy, a restructuring of energy pricing, and reforms designed to increase productivity). In neither report were the staff’s own views spelled out. However, the latter report noted that there had been a 5.7 percent depreciation in the real effective exchange rate between 1985 and 1990, at a time when the terms of trade had improved, thus betraying no “overvaluation” concerns. Even the 1992 Article IV report, which went beyond the standard presentation of real effective exchange rate indices to consider other indicators (e.g., relatively high cost of labor), stopped short of advocating an exchange rate adjustment, though it did provoke a discussion of devaluation of the CFA franc at the Board discussion. The inadequacies of the real effective exchange rate index as an indicator of international competitiveness are well known. Using a method that incorporates terms of trade shocks, Devarajan (1997) estimates that on average the real exchange rate was overvalued by about 30 percent for 12 countries in the CFA franc zone prior to the January 1994 devaluation. He estimates the real exchange rate in Senegal was overvalued by 22 percent.

    Other Resources Citing This Publication