Ex Post Assessment of Longer-Term Program Engagement

This Ex-Post Assessment of Longer-Term Program Engagement for Ghana explains its experience with particular attention to the IMF's role. Progress in structural reform was slow, but Ghana also avoided mistakes made by some countries that introduced reforms without enough public support. Ghana’s success in its reforms is embedded in the strengthening of democratic institutions that allow for greater public ownership of the reforms. Continual structural reform is essential to ensure that macroeconomic stabilization is durable. Recent slippages underscore the need to further improve fiscal institutions to control expenditures.


This Ex-Post Assessment of Longer-Term Program Engagement for Ghana explains its experience with particular attention to the IMF's role. Progress in structural reform was slow, but Ghana also avoided mistakes made by some countries that introduced reforms without enough public support. Ghana’s success in its reforms is embedded in the strengthening of democratic institutions that allow for greater public ownership of the reforms. Continual structural reform is essential to ensure that macroeconomic stabilization is durable. Recent slippages underscore the need to further improve fiscal institutions to control expenditures.

I. Introduction

1. Ghana is widely considered an African success story. Thus, in a continent that has repeatedly come short of expectations, it gives hope for a better future, not only for Ghana itself but for many other countries in the region. The achievements most commonly cited are its sustained growth rate which has led to a 50 percent increase in per capita GDP over some 20 years and a marked reduction in poverty, the maturity of its democratic institutions, and the lack of major armed conflict in an otherwise difficult neighborhood.

2. The purpose of this Ex-Post Assessment is to take a closer look at the Ghanaian experience, with particular attention to the Fund’s role. The assessment tries to identify lessons for Ghana itself and for the Fund’s future engagement in Ghana and similar countries. In light of these lessons, it also discusses options for future Fund engagement in Ghana and how and when Ghana might exit from its long-term program engagement with the Fund.

3. The assessment covers a period of about 11 years, from the approval of the 1995-1999 ESAF to the expiration of the 2003-2006 PRGF. Besides reviewing staff reports and internal memos, the assessment team has also drawn on external research, World Bank reports, interviews with mission chiefs, and the authorities. The team leader visited Accra twice, first in December 2006 for initial discussions, then in February 2007 to present the findings.1

4. The paper is organized in three main parts. The first part reviews Ghana’s economic performance, concentrating on: why inflation was persistently higher than expected; why debt became unsustainable; and whether the successes are solidly anchored or risk being reversed. The second part reviews the role of the Fund in Ghana’s achievements and summarizes the main lessons. The paper concludes with an assessment of future options for Ghana and vis-à-vis the Fund.

II. Ghana’s Experience

A. Overview

5. The introduction of the Economic Recovery Program in 1983 led to a dramatic change in Ghana’s economic performance. While up to then the economy was marred by spouts of hyperinflation and declining output, Ghana’s output has since exceeded that of most other sub-Saharan African countries (Figure 1). Since 1983, annual growth has averaged 4.6 percent, compared to 3.6 percent for sub-Saharan Africa. In per capita real PPP terms, the performance is even better—1.8 percent, compared to 1.0 percent in sub-Saharan Africa.

Figure 1.
Figure 1.

Ghana: Comparative Economic Performances, 1995-2006

Citation: IMF Staff Country Reports 2007, 211; 10.5089/9781451814996.002.A001

Sources: IMF, African Department database; and IMF, World Economic Outlook.

Real GDP Per Capita and Inflation Across Political Regimes

Citation: IMF Staff Country Reports 2007, 211; 10.5089/9781451814996.002.A001

Sources: IMF (International Financial Statistics); World Bank (World Development Indicators).

6. The upward trend in per capita output was, however, interrupted by several severe, though temporary, setbacks. All were the result of pre-election fiscal slippages, amplified by shortfalls in external aid. The Economic Recovery Program had brought about a sharp reduction in the inflation rate combined with stable growth and increasing current account surpluses; the first major setback came in 1992 in the run-up to the first parliamentary elections, and others followed in the two subsequent general elections. While these setbacks delayed the stabilization process and reduced average growth, the economic recovery always came back on track a few years later, often with the help of Fund-supported programs and the resumption of donor support.

7. Thanks to rapid growth and a reduction in inequality, progress in combating poverty has been encouraging. The number of Ghanaians living below the poverty line has dropped from 50 percent in 1990 to 42 percent in 1997 and further to 35 percent in 2003, lifting more than 1 million people above the poverty line. Moreover, Ghana is well poised to reach by 2010 the Millennium Development Goal (MDG) of reducing the incidence of poverty to half its 1990-level, and food security has improved markedly. Overall, Ghana’s social indicators compare well to other countries in sub-Saharan Africa (Tables 1 and 2). Despite these improvements, rural poverty remains a problem. Growth has largely benefited urban areas and export-oriented regions, leaving behind those regions relying mainly on subsistence agriculture. There are also large differences between regions in their access to education and health services.

Table 1.

Comparison of Poverty and Social Indicators in Sub Saharan Africa, 2004.

article image
Source: UNDP Human Development Report 2006.

Sub-Saharan Africa.

Table 2.

Economic and Social Indicators in Ghana and PRGF-Eligible Countries

(In percent per annum, unless indicated otherwise)

article image
Source: WEO, IFS, WDI, and staff estimates.


8. During the review period, Ghana was engaged in three back-to-back Fund-supported economic programs.2 The first three-year program was supported by Enhanced Structural Adjustment Facility (ESAF). It originally covered the period 1995–1998, but was later extended to 1999. The second program was also supported by an ESAF, which was later transformed into a Poverty Reduction and Growth Facility (PRGF). It covered the period 1999-2002. The third program was supported by a PRGF. It originally covered the period 2003-May 2006, but was subsequently extended to October 31, 2006.


Ghana: Fund Arrangements, 1995-2006

Citation: IMF Staff Country Reports 2007, 211; 10.5089/9781451814996.002.A001

Original duration of program Program extension

9. The period from 1995 through 2006 can be thought of in terms of two sub-periods: a not so successful stabilization (1995-2002) and a quite successful stabilization (2003-2006) (Figure 2).

  • The first two three-year Fund-supported programs, covering the period 1995-2002, were plagued by review delays and eventually went off track. Disbursements under the first annual arrangement under the 1995-1999 ESAF were delayed due to nonobservance of five performance criteria, the second annual arrangement was deferred by delays in prior actions, and implementation delays precluded a third annual arrangement. During the second three-year arrangement, the 1999-2002 PRGF, all program reviews except the first one were delayed. The fifth review was not completed due to significant policy slippages.

  • The latest PRGF arrangement (2003-2006) was successfully concluded despite some delays in completing reviews, on account of policy slippages. All reviews and disbursements under the arrangement were completed.

Figure 2.
Figure 2.

Ghana: Summary Program Performance: 1995-2006

Citation: IMF Staff Country Reports 2007, 211; 10.5089/9781451814996.002.A001

Solid line–actual values.Dashed line–staff projections.Source: IMF staff reports.

10. After the expiration of the last program, it became apparent that fiscal performance had deteriorated markedly in the second half of 2006, reflecting excessive wage increases and utility subsidies. Compared to a program projection of 3.6 percent of GDP, preliminary estimates suggest that the 2006 fiscal deficit may exceed 7½ percent of GDP. The authorities are, however, taking remedial action to ensure a lower deficit in 2007 (for details see the Staff Report on the 2007 Article IV Consultation).

B. What Changed in 2002/2003: Improved Policies, Luck, or a Bit of Both?

11. Some of the improvement in Ghana’s performance doubtlessly reflected better conditions for most low-income countries (Figure 2). While Ghana’s growth exceeded that of its most relevant comparators, growth increased throughout the developing world, as in Ghana, due to increased remittances, stronger commodity prices, and a surge in aid. But a closer review of the two periods suggests that most of the improved performance in Ghana can be attributed to better policies and institutions.

12. While other factors may have played a role it seems that better policies explain most of the different responses to the two recent terms-of-trade shocks. The recent oil price rises, which induced sharp drops in terms-of-trade, did not have the same overall economic effects as the sharp drop in cocoa prices in 1999-2000. Although the terms-of-trade fell similarly in these two episodes, the cocoa terms-of-trade shock was associated with a sharp depreciation of the exchange rate, a 2-percentage point fall in growth, a ballooning fiscal deficit, and accelerating inflation; the collapse in the cedi was thus more a reflection of the overall policy stance than the magnitude of the terms-of-trade shock. By contrast, the oil terms-of-trade shock had no noticeable impact on growth and fiscal position, and only a small direct impact on inflation. Unlike the earlier terms-of-trade shock, the authorities responded to the oil price shock with strong macroeconomic policies and contained the impact on public finances by allowing for a pass-through of the oil price increases. In the earlier cocoa terms-of-trade shock, fiscal balances worsened because of increased spending on subsidies, capital expenditures, and transfers.

13. In terms of macroeconomic stabilization, the most striking improvements were in the fiscal area. Recourse to domestic financing dropped significantly: the domestic financing need of the central government moved from 4.5 percent of GDP for 1995 through 2002 to a negative 0.2 percent of GDP since. Besides increased aid, this improvement reflected advances in both expenditure management and tax administration:

  • Improved expenditure management led to better allocation of expenditure to its intended uses. Three major pieces of legislation (the Financial Administration Act, Internal Audit Act, and the Procurement Act) resulted in better fiscal control and reporting was improved. A medium-term framework was introduced, and strict cash limits were imposed on different ministries (Box 1).

  • The tax-to-GDP ratio rose from 14.7 percent in 1995 to 19.6 percent in 2006, largely because the tax base was broadened and compliance improved. The burden of taxation was shifted toward consumption, reducing its direct impact on factors of production.

14. Of equal importance is the lower reliance on arrears and reduced quasi-fiscal losses. While good data are hard to come by, staff estimates that arrears added several percentage points of GDP to public sector debt between 1995 and 2002. Quasi-fiscal activities emanating from losses in Tema Oil Refinery (TOR) were estimated at 7 percent of GDP in 2000, and there were also large losses in the electricity sector. 3 These losses, which were not fully recorded in central government accounts, led to a build up of debt owed to the state-owned Ghana Commercial Bank (GCB).

15. Perhaps the biggest achievement of recent years, was the absence of major fiscal overruns in the run-up to the 2004 general elections. Unlike the 1995–1998 and 1999–2002 arrangements, which were temporarily suspended due to election-related spending in the run-up to the 1996 and 2000 elections, the 2003–2006 arrangement continued uninterrupted. While there was still a slippage of about 2 percent of GDP in 2004 (measured as the difference between the fiscal outcome relative to target as a percent of GDP), it was not large enough to interrupt the trend reduction in the fiscal deficit; and the overall deficit remained moderate.

Ghana—Public Financial Management Reform

Until 2000, progress on public financial management reforms were slow, as evidenced by the recurrence of domestic arrears and the breakdown of expenditure control in the run-up to the end-2000 elections. In contrast, under the last PRGF-supported program, public financial management efforts gained significant momentum with tangible results, reflecting strong political will and donors’ concerted efforts.

Regulatory framework for public finances

The regulatory framework for public finances has been strengthened, through the enactment of several key pieces of legislation (Financial Administration Act, Internal Audit Act, and Procurement Act).

Budget execution and reporting

  • Expenditure control and the timeliness of fiscal reporting have improved. After the 2000 expenditure control breakdown, basic systems for cash planning, commitment control, and fiscal reporting have been installed and become functional, with extensive technical assistance from FAD. Moreover, external oversight of budget execution has improved through timely submission of Auditor-General reports to Parliament and the clearance of the backlog of outstanding audit reports. Since August 2005, monthly reports on budget execution have generally been completed within six weeks, although more work is needed to reduce the discrepancy between above- and below-line data and to fully cover externally-financed projects.

  • The medium-term expenditure framework has been simplified (with the number of activities being reduced drastically) to adapt to the limited capacity of MDAs.

  • Implementation of the computerized budget and public expenditure management system (BPEMS) has advanced, with application set-up in eight pilot Ministries, Departments, and Agencies (MDAs). Similarly, the new computerized payroll management system (IPPD2) has become functional, serving all MDAs and some pilot subvented agencies.

  • A decentralized payment system is being introduced to speed up the transfer of funds to ministries, regional administrations, and district assemblies. The new payment system will help streamline treasury management and facilitate fiscal decentralization.

Legislative oversight of the budget

Procedures for legislative oversight are generally followed and adequate time provided for the legislature to scrutinize budget proposals. The rules for in-year amendments are also clear and generally followed, with expansion of the overall budgetary envelope requiring a supplementary budget. The Public Accounts Committee reviews the Auditor General’s reports, holds hearings, and submits its report to the House within 12 months of submission of the audit reports.

Tax administration and policy

The key tax administration reforms include the creation of a Central Revenue Authority, the establishment of a Large Taxpayers’ Unit, the introduction of a single Taxpayer Identification Number, the automation of customs systems. The efficiency and buoyancy of the tax system have also increased through the introduction of VAT, the reform of petroleum taxation, and a tariff reform.

16. A significant shift in monetary management helped to strengthen economic policies. Introduced in 2002, the new central bank law provided a more solid administrative and legal foundation for the Bank of Ghana, introducing, inter alia, limits on the central government’s access to central bank finance to at most 10 percent of total expenditure, an improved auditing function, and price stability as the bank’s primary objective. In addition, the Bank of Ghana introduced some elements in preparation for the introduction of inflation targeting, such as regular meetings of the monetary policy committee and the publication of economic and financial reviews. Efforts were also made to improve the Bank of Ghana’s ability to forecast inflation (with the help of the Bank of England).


Programmed versus Actual Inflation, 1995-2006

(End period, in percent)

Citation: IMF Staff Country Reports 2007, 211; 10.5089/9781451814996.002.A001

C. Why Did Inflation Stay So High for So Long?

17. Despite a significant improvement in the monetary policy framework, monetary policy was not immediately successful in achieving its stated objectives. The targeted single digit inflation rate was only reached after the end of the last program and was not achieved at any point during the review period. While the 1995-1999 program managed to bring inflation down close to 10 percent in 1999, this achievement was rapidly unwound in 2000 after the value of the cedi collapsed. Moreover, the deviation between target and outcome remained high: the one-year-ahead deviation of targets versus outcome fell from 12 percent in 1995-1999 to around 10 percent in the two subsequent arrangements.

18. Until 2002, monetary policy was clearly a by-product of poor fiscal discipline. Excessive public spending combined with a shortfall in aid inflows, and inflexible revenue streams left the authorities with only two options for financing the budget gap: domestic borrowing and central bank finance. And until 2002, there was no legal limit on how much the central government could borrow from the central bank. In several cases too, quasi-fiscal operations were financed by the central bank lending to state-held banks, which also contributed to reserve money creation; monetary control was also undermined by the misreporting of reserve monetary data in 2000 and 2001. 4 Similarly, the 1999/2000 cedi collapse and the ensuing hike in inflation was largely driven by the need to provide monetary finance for fiscal slippages and shortfalls in divestiture receipts and donor inflows; 5 the sharp fall in cocoa prices seems mainly to have been the triggering factor for the collapse.

19. Why did inflation not fall more rapidly and durably after 2002, when monetary policy was freed from the need to finance large budget deficits? Even while correcting for the January 2003 petroleum price hike, which admittedly explains most of the 2003 hike, inflation remained clearly above the target.6 Moreover, a review of program documentation shows that the higher-than-targeted inflation was not the result of errors in the assumptions underlying the monetary program. Comparing program assumptions and outcomes for the 2002-2006 period shows that excess inflation was not the result of an unexpected increase in reserve money velocity, but a result of higher-than-programmed base money growth. In fact, assumptions about reserve money velocity seemed to have erred on the side of caution as base money velocity fell faster than expected, thus limiting the impact of rapid reserve money growth on inflation.

20. One explanation for the excessive growth of reserve money between 2002 and 2006 may be that capital inflows were not adequately sterilized. Until 2001, the question was how to ensure sufficient external financing to fill the large current account deficit; after 2001 excess inflows allowed for accumulation of international reserves and the repayment of domestic debt. Helped by the collapse of the cedi and increased remittances, the current account deficit fell markedly and the improved policy outlook triggered a sharp scale-up in aid inflows during 2003-2006.

21. This view is consistent with data on the sources of reserve money growth. The text table (below) shows that, after 2002, the increase in reserve money largely came from reserve accumulation, which was incompletely sterilized by the increase in central bank securities held by the private sector. By contrast, up to 2001, the increase largely came from central bank lending to central government and the financial sector.

Ghana: Contributions to Monetary Base Growth: 1995-2005


article image

Corrected for the recognition of government liabilities to recapitalize Bank of Ghana.

Source: International Financial Statistics and authors’ estimates.

22. Exchange rate policy may also have helped limit the decrease in the inflation rate. While there is not much doubt that macroeconomic stabilization contributed to increased exchange rate stability, the monetary authorities chose to let most of the inflows accumulate as reserves, thus offsetting some of the upward pressures on the exchange rate. Indeed, the US dollar-cedi exchange rate has become increasingly stable, particularly from mid-2004.

23. Supporting this view, staff estimates of the real equilibrium exchange rate suggest that the 2001 exchange rate depreciation led to a significant undervaluation of the exchange rate (Figure 3). They suggest that the exchange rate depreciation in 2000 significantly undershot its equilibrium level and that the subsequent real appreciation can be interpreted as a convergence of the actual real exchange rate to its equilibrium level (reached in early 2006). Thus, to the extent that the nominal exchange rate was kept relatively stable, the pressure to revalue the real exchange rate had to come through higher inflation: between 2002 and 2006, average yearly inflation exceeded that of Ghana’s trading partners by some 13 percent.


Ghana: Nominal Exhange Rate against the US dollar, 1995-2006

(End of month)

Citation: IMF Staff Country Reports 2007, 211; 10.5089/9781451814996.002.A001

Figure 3.
Figure 3.

Ghana: Actual and Equilibrium Real Effective Exchange Rate, 1984-2006

Real Effective Exchange Rate: Actual vs. Equilibrium Values

Citation: IMF Staff Country Reports 2007, 211; 10.5089/9781451814996.002.A001

24. More generally, overshooting the inflation target seems to be related to the lack of a clear nominal anchor to guide inflation expectations. Given the long period of double-digit inflation and the short track-record of the new institutions, the Bank of Ghana was faced with the difficult task of changing expectations without incurring too high economic costs. During most of the post-2000 period it was unclear whether monetary policy was steered by the monetary target (repeatedly overshot, whether broad or narrow money were targeted), the exchange rate (which continued to depreciate until mid-2004), or an inflation target (also missed, whether headline or core inflation was targeted). In addition, the piecemeal and stop-go implementation of energy price reform led to sudden energy price-movements, further undermining attempts to generate lower inflation expectations.

25. Nevertheless, inflation has been reduced significantly and is expected to reach single digits this year. Disinflation could possibly have been more rapid by adhering more closely to the monetary target and allowing the real exchange rate to appreciate faster, but this would—if not accompanied by a downward shift in inflation expectations—have come at the cost of lower growth. Moreover, monetary policymaking was complicated by unpredictable developments in velocity and volatile aid inflows.

D. Why Did Ghana’s Public Debt Become Unsustainable?

26. Before the recent debt relief initiatives, 7 Ghana—like most other low-income countries—was living in the shadow of an ever-increasing debt burden (Figure 4). What is striking in the case of Ghana is that the external debt burden was considered sustainable in 1996, only to be declared unsustainable some three years later, in November 1999. The assessment that Ghana’s external and public debt was unsustainable was repeated and reinforced in 2000 and 2001.8


Ghana: External Public Debt, 1995 - 2005


Citation: IMF Staff Country Reports 2007, 211; 10.5089/9781451814996.002.A001

Figure 4.
Figure 4.

Ghana: External Debt and Debt Service, 1996 - 2006

Citation: IMF Staff Country Reports 2007, 211; 10.5089/9781451814996.002.A001

Source: IMF.

27. A review of the evolution of public sector debt suggests that a number of factors contributed to the changed assessment of the Ghana government’s ability to carry its debt:

  • Fiscal slippages and quasi-fiscal losses led to increased domestic borrowing as well as some commercial external borrowing.9 Increased deficits were largely financed by foreign and domestic borrowing (from the banking system), and, in the earlier years, by slower repayment of arrears.

  • The high cost of debt also mattered. In 1996, the limit on nonconcessional borrowing was breached, and borrowed funds (amounting to almost 1 percent of GDP) were used for poverty-reducing, but low return, public projects.

  • Public enterprise liabilities were recognized. In 1999, domestic public debt was revised up by about 7 percent in GDP when debt of some public enterprises was included. External debt, on the other hand, was also revised upwards in 1996-1998 but then downwards in 2001 by a similar amount.

  • The 1996 debt sustainability assessment assumed a future real growth rate of close to 7 percent.10 Future growth in the 1999 assessment compared to (below) 6 percent 11 and 5 percent in the 2001 assessment. The reduced growth projections explain about half of the increased end-period debt projection in the 1999 assessment.

  • In 2000, the large drop in the value of the cedi against the US dollar and the euro (in 2000, nearly all external debt was denominated in either one of these or SDRs), pushed up the external debt-to-GDP ratio (Figure 4). In real effective terms, the 1999/2000 cedi collapse led to an immediate depreciation of exchange rate of about 35 percent, resulting in an about equal increase in the GDP ratio of external debt, only half of which has been reversed since.

  • The benchmark of acceptable debt levels was also changed. Under the Enhanced HIPC Initiative, the threshold for the net present value of external debt-to-exports ratio was lowered from 200–250 percent to 150 percent, and for net present value of external debt-to-revenues from 280 percent to 250 percent. With these changes both the debt-to-exports and to revenues ratios exceeded the thresholds.

Ghana: External Debt Dynamics

(Percent of GDP)

article image
Source: IMF staff.

“Other factors” include data revisions, debt relief, and other non-price or exchange-rate related valuation changes

28. Decomposing the contributions to the increase in the external debt-to-GDP ratio illustrates the relative importance of these different factors (see text table). While before 1999, new borrowing seems to have been the predominant factor in the debt increase, exchange rate changes overshadowed all other factors in 2000 and in the period after that.

29. While much of the sustainability analysis focused on external debt, the increase in domestic public debt was equally important. Ghana’s domestic debt continued to rise during the 1995–1999 program and by 1998 it was high even in comparison with other West African HIPC countries. The staff analysis of the sustainability of domestic public debt in 1998 demonstrated the importance of stabilizing domestic financing as a percentage of GDP if real interest rates were to be brought down. Based on the projected fiscal outcome for 1998, the analysis suggested that the medium-term outlook under the program would imply a leveling-off of the ratio of domestic debt to GDP at about 20 percent in 1999–2000, followed by a modest downward trend over the medium term. However, because of fiscal slippages in the run-up to the 2000 elections, domestic debt as a share of GDP increased to almost 30 percent.


Ghana: Stock of Domestic Debt, 1996-2006

(Percent of GDP)

Citation: IMF Staff Country Reports 2007, 211; 10.5089/9781451814996.002.A001

30. Finally, the deeper origins of the debt problem predates the review period (Appendix I). The most rapid build-up in the external debt-to-GDP ratio happened in the early 1980s. After the Economic Recovery Plan was introduced in 1983, multilateral institutions and donor countries scaled up their lending but there was no corresponding increase in exports. For the whole 1980–1995 period, the Fund’s role as a lender was relatively moderate. All of the debt increase after 1986 was due to other multilateral institutions and donors.

31. Reaching the HIPC completion point and then qualifying for MDRI debt relief were turning points in restoring debt sustainability. Ghana’s external debt fell from 120 percent of GDP in 2000 to 22 percent in 2006. Between 2001 and 2005, the debt-to-GDP ratio decreased to about 60 percent because of a combination of faster GDP growth, exchange rate appreciation, and the HIPC debt relief in 2004. The MDRI relief was the main factor behind the further substantial decline in the ratio in 2006.

E. How Solidly Are Recent Achievements Anchored?

32. Most observers agree that Ghana has made significant strides towards achieving improved macroeconomic balance. While there have been several setbacks, fiscal management has improved markedly and inflation, though still too high, is close to the long-coveted single-digit level. Meanwhile, structural reform has moved steadily to open the economy to market forces, improving the prospects that economic growth will accelerate. In particular, the last Fund-supported economic program (2003–2006) can be considered successful in achieving its stated objective of achieving macroeconomic stability as an important precondition for accelerated efforts to combat poverty. The question is, however, whether these achievements are durable; economic history has many examples of success stories that quickly unravel, notably Ghana itself in the early 1990s.

33. One important lesson from the 1990s is that, unless combined with a system of checks and balances, democratic institutions may carry risks as well as benefits. The fiscal excesses of 1995/1996 and 1999/2000 both occurred in the run-up to general elections. The tendency to overspend was larger in the highly contested elections in 1996 and 2000 than in 2004 when the incumbent government had a more comfortable margin.

34. While the electoral spending cycle has not been eliminated, progress in structural reform and institution building may help limit future slippages.

  • The government’s powers to increase expenditures have been limited by strengthening parliamentary controls, the establishment of an independent Auditor General, and increasing press freedom.

  • Improved administrative controls have restrained wasteful spending. Systems for cash planning, commitment control, and fiscal reporting have been installed and become functional. The partial privatization and improved supervision of the financial sector has reduced the potential for the creation of large-scale fiscal liabilities through the banking system; the financial control of state-owned enterprises has also improved.12

  • The 2002 Bank of Ghana Act established price stability as the primary objective of the central bank. Legal foundations for the operational independence of the central bank, improved internal auditing mechanisms, and a ceiling on the central government’s access to central bank finance were also introduced.

  • Privatization has allowed the private sector to control more of the economy’s resources (see Appendix II for more details) and thereby make them less vulnerable to political cycles.

35. On institutional quality, Ghana compares well to its peers along most dimensions. Notably, Ghana scores better than average for sub-Saharan Africa, mature stabilizers, emerging markets, and HIPC countries. Sustained democracy has brought political stability and entrenched the rule of law (Figure 5). Moreover, despite a negative trend in the 1990s, according to Transparency International Ghana is considered to be less corrupt than most other low-income countries. Among the sub-Saharan African countries, Ghana ranks in the top eight in the World Bank’s Ease of Doing Business Index, scoring particularly well on enforcement of contracts, investor protection, and cross-border trade.


Ghana and Comparator Countries: Press Freedom Index, 1994-2006

Citation: IMF Staff Country Reports 2007, 211; 10.5089/9781451814996.002.A001

Source: Freedom House.Note: Free (F): 71-100, Partly Free (PF): 41-70, Not Free (NF): 0-40.Mature stabilizers based on Selassie et al (Occosional Paper, 2006).
Figure 5.
Figure 5.

Ghana: Indicators of Institutional Quality, 1996 and 2005 1

Citation: IMF Staff Country Reports 2007, 211; 10.5089/9781451814996.002.A001

Sources: World Bank, Transparency International1 High index values indicate high degree of voice and accountability, political stability, quality of the regulatory authorities, and low degree of perceived corruption. The Corruption Perception Index has been normalized by subtracting 3 from the index.

36. However, not enough has been achieved in reforming the electricity and the public sectors. Private sector participation in the electricity sector is negligible and tariffs are adjusted without sufficient relationship to the true costs of electricity production.

Moreover, the implicit subsidization of the aluminum smelter puts pressures on the operating costs of the electricity companies, reducing the economic incentives for investment either by the private sector or by the companies themselves, resulting in increased capacity problems. As for public sector reform, progress has been limited despite longstanding government plans for a comprehensive reform that go back to the late 1990s. Thus, without an improved wage structure and employment reform, pressures to adjust imbalances may easily lead to an overall increase in the wage bill.

III. The Fund’s Role

37. Overall, the Fund’s engagement in Ghana has helped the authorities achieve macroeconomic stabilization. While the 2003-2006 program was the most successful arrangement, it is also possible that the earlier arrangements prepared the ground for the later successes. In the area of structural reform, for example, joint efforts were helpful in reforming the petroleum sector and reducing state involvement in the economy. Similarly, long-standing efforts to improve public expenditure management—often provided as technical assistance—were helpful in improving fiscal discipline. In an attempt to glean critical lessons for the future, the rest of this assessment looks at a range of areas where there may be scope for improvement: the pace of macroeconomic stabilization, the comprehensiveness of structural reform, Fund access, and collaboration with the World Bank and other donors.

A. What Could Have Been Done to Achieve More Rapid Macroeconomic Stabilization?

38. While in the end relatively successful, a number of years were lost before macroeconomic stabilization was achieved. Fiscal deficits were higher than targeted—in particular in the first two three-year programs—and inflation remained too high and has only recently come down to an acceptable level. On average therefore, growth and poverty reduction have consequently been slower than they otherwise could have been.

39. As for fiscal policy, more intense efforts in improving fiscal institutions at an earlier stage could potentially have prevented some of the worst slippages. Fiscal slippages predominantly reflected expenditure overruns rather than revenue shortfalls and occurred mostly and most severely in the run-up to elections. Improved expenditure management seems to have been key to bringing fiscal management under better control. Earlier offers of extensive fiscal assistance in this area could thus potentially have helped.

40. Until 2002, the lack of progress in achieving rapid reduction in inflation was largely linked to poor fiscal discipline. Given the limited access to large-scale borrowing, monetary policy was thus largely dominated by fiscal developments. After external and domestic sources of finance were exhausted, inflationary finance provided the only available means of finance. It is thus doubtful whether an earlier introduction of a limit on central bank borrowing or a stricter commitment to nominal targets would have done much to reduce inflation faster, as long as the underlying fiscal discipline was lacking. During 1998-2000, the introduction of performance criteria on reserve money was unsuccessful in limiting the resurgence of inflation in 2000, largely because the reserve money data was misreported (Box 2 and Table 3).

Ghana: Misreporting and Non-complying Purchases

During the 1999-2002 arrangement, the Executive Board discussed four cases of Ghana’s misreporting, which resulted in four noncomplying disbursements (Table 3).

  • The second disbursement (November 1999) of the 1999-2002 arrangement was found to be noncomplying in August 2000 because a performance criterion prohibiting the introduction of a multiple currency practice was not met. However, the Executive Board granted a waiver for the nonobservance of the performance criteria because: (i) the deviation was considered temporary; and (ii) Ghana adopted additional policy measures to achieve the objective of the program.

  • The third disbursement (September 2000) was found noncomplying because of misinformation on the amount of external arrears and nonconcessional borrowing. In this case, no waiver was granted as the Executive Board found that the two cases of misreporting pertaining to the third disbursement were neither temporary nor minor. The Board determined that the external arrears incurred had had a significant impact on program performance and were large enough that their timely reporting might have led to different policy recommendations. Regarding the misreporting on external arrears, both problems were due to Ghana’s limited capacity to manage external debt. To remedy the situation, Ghana has since instituted better procedures for calculating grant elements and limited authority to contract and guarantee debt to the Aid and Debt Management Unit of the Minister of Finance.

  • The third and the fourth disbursements (September 2000 and July 2001) were found noncomplying due to the misreporting on reserve money data. A waiver was granted because Ghana had adopted additional measures to achieve the objectives of the program. It was also mentioned that the new governor of the Bank of Ghana had promptly reported the misreporting and taken corrective action to improve procedures in the Bank of Ghana. The 2002 Bank of Ghana Act established the legal basis for a system of improved internal and external controls.

One case of misreporting was identified during the 2003-2006 arrangement. The first disbursement (December 2003) was found noncomplying since the prior action on water tariffs had not been implemented. A waiver was granted in view of the prompt action taken by the authorities to raise water tariffs by the programmed amount and their commitment to implement future adjustments.

Table 3.

Ghana: Cases of Misreporting and Non-complying Disbursements1

article image
Source: IMF staff

PC stands for performance criterion.

41. In the 2003-2006 program, however, monetary policy was less constrained by the need to finance fiscal overruns. Tighter monetary policy—e.g., by limiting monetary growth to the program targets—could have led to a more rapid reduction in inflation, but was avoided, potentially out of a concern about competitiveness of the exchange rate and the inability, or unwillingness, to fully sterilize the ensuing reserve inflows. Here, the Fund could have insisted on monetary growth targets being observed, perhaps by setting performance criteria on monetary base growth: unlike the 2000-2001 annual programs, the 2003-2006 annual programs did not include a ceiling on monetary growth, only a limit on net domestic assets and a floor on net international reserves (Table 4). This was, however, not done due to perceived instability in the demand for money and the uncertain monetary impact of volatile aid inflows (see Section II. C).

Table 4

Observance of Quantitative Performance Criteria, 1995-2006

article image

Defined as net domestic financing of the budget by the banking system and net sales of government debt to the nonbank public.

Defined as the difference between total revenue (excluding grants and divesture proceeds) and noninterest domestic expenditure (excluding foreign-financed capital expenditure)

Net foreign assets in 1995-2001.

42. Overall, the relatively poor observance of performance criteria prompts the question of whether there was adequate country ownership of policies. During the first two three-year arrangements (1995-2002), a large number of waivers and extensions were granted; this did not, however, prevent either of these arrangements from being extended. Although hard to gauge, it is possible that an earlier interruption of the arrangements could have led to a more rapid turnaround in policy implementation.

43. In 2001-2003, the sharp increase in donor support and unpredictability of disbursements complicated macroeconomic policy implementation (Appendix III and Figure 6). Aid inflows were used both to build up reserves and to strengthen the fiscal position: Ghana has been referred to as an example of the “don’t spend, don’t absorb” approach.13 The monetary impact of the inflows was significant since the government used the improved fiscal position to repay domestic debt instead of reducing its net debt to the central bank obligating the central bank to engage in increasingly large sterilization operations.

Figure 6.
Figure 6.

Ghana: Managing Aid Inflows

Citation: IMF Staff Country Reports 2007, 211; 10.5089/9781451814996.002.A001

Source: IMF Database.

44. Uncertain aid inflows contributed to the macroeconomic instability during 2000-2002. In particular, the nonobservance of donor conditionality on divestiture led to a large drop in program aid in 2000, contributing to external financing problems and the collapse of the cedi. Thus, in these years, better donor coordination could potentially have softened the macroeconomic impact of the policy slippages. Better prediction of the impact of policy slippages on aid disbursement could also have helped (see Appendix III).

B. Could the Fund have Done More to Ensure Progress in Structural Reform?

45. Despite relatively successful macroeconomic stabilization, risks remain due to repeated delays in structural reform, particularly in the energy and public sectors. Although conditionality required the introduction of a market based energy-pricing mechanism and increased private participation in the sector, progress was slow. Public sector reform was not a focus of Fund conditionality until towards the end of the 2003-06 program, when the macro-relevance of progress in this area became increasingly clear: ceilings on the aggregate wage bill of the public sector were introduced in the last three reviews together with structural performance criteria on the need to establish a computerized personnel database and a report on public sector salary structure.

46. On energy sector reform, a review of the structural performance criteria on the energy sector shows a strikingly low degree of compliance. Performance criteria on the introduction of an automatic pricing mechanism were in several instances undone soon after being introduced. The performance criterion on the introduction of the automatic price adjustment mechanism on petroleum was applied four times: first met in the 2003 program, it had to be reintroduced in the 2004 and 2005 programs after the mechanism was not allowed to work. Similarly, the automatic pricing mechanism on electricity was introduced three times and—while the performance criterion was met twice—such a mechanism has still not been fully implemented. Thus, while about half of the performance criteria on the energy sector (including divestitures of energy companies) were not met or waived, most of the performance criteria were only temporarily met as the same or similar performance criteria had to be reintroduced at a later stage. Similarly, a performance criterion on the divestiture of the state’s ownership of energy companies was waived at several reprises.

47. Significant progress in the reform of the energy sector was not achieved before well into the 2003-2006 PRGF, largely reflecting the increased focus on the sector in work done by the World Bank and the introduction of the Poverty and Social Impact Analyses (PSIA). Moreover, technical assistance on the petroleum price adjustment provided an important input for policymakers on the design of mitigating measures, 14 which helped soften the public resistance to the removal of petroleum product price subsidies.

48. This experience suggests that structural performance criteria are poor instruments to use to achieve lasting price-setting reforms. With poor ownership or unchanged regulatory structures such price setting schemes can be rapidly undone. Thus, unless the Fund is willing to interrupt the arrangement due to lack of observance of the performance criteria after the test date, the commitment is unlikely to be binding. While the introduction of a continuous performance criterion is possible, it would seem more fruitful to work closely with the authorities to ensure that a more profound change in economic structures takes place. It also suggests that to ensure improved ownership in this area, energy pricing reforms would need to be accompanied by adequate social safety nets to mitigate the impact on the poor. And it is probably crucial to undertake a public information campaign in support of sensitive economic reforms, as illustrated also by a successful reintroduction of the VAT, after a failed attempt in 1995.

49. The set-back in structural conditionality indicates that most programs, in particular the earlier ones, suffered from a lack of ownership. On the surface, the programs seem to be broadly in line with the Fund’s efforts of streamlining the use of structural conditionality (Box 3), but the extensive use of prior actions at program reviews—in particular in the second three-year program—suggest that Fund staff considered it necessary to impose additional conditions to prevent a derailment of the reform process (Figure 7 and Table 5). At 36, the number of prior actions in the 1999-2002 ESAF/PRGF arrangement was also exceptionally high; in the 2003-2006 PRGF the average number was in line with countries considered to have weak-track records.15 Moreover, as with the performance criteria on energy pricing, the prior actions were often easily reversible, making it questionable whether these conditions effectively reached their intended goals. In broad terms, however, ownership seems to have improved over time: compliance with program conditionality increased and the number of structural performance criteria and prior actions fell toward the end of the last program.16

Figure 7.
Figure 7.

Ghana: Structural Conditionality, 1995 - 2006

Citation: IMF Staff Country Reports 2007, 211; 10.5089/9781451814996.002.A001

Source: IMF.
Table 5.

Ghana: Structural Conditionality

article image
article image
article image
article image
article image
article image
article image
article image
article image