Burundi is in great need of investment in infrastructure, but fiscal constraints leave little room for additional public spending. Despite this initial recovery, Burundi has yet to rebuild its pre-civil war level of public capital stock. Improving the business climate is one of the keys to attracting higher private investment. Since the Arusha agreement, some progress in the business climate has been made. Burundi is quickly moving away from the unsustainable debt situation and unstable exchange rate of the 1990s.

Abstract

Burundi is in great need of investment in infrastructure, but fiscal constraints leave little room for additional public spending. Despite this initial recovery, Burundi has yet to rebuild its pre-civil war level of public capital stock. Improving the business climate is one of the keys to attracting higher private investment. Since the Arusha agreement, some progress in the business climate has been made. Burundi is quickly moving away from the unsustainable debt situation and unstable exchange rate of the 1990s.

IV. Public Expenditure Management Reforms in Burundi45

A. Introduction

69. The new government elected in August 2005 after 12 years of civil conflict has made good governance and combating corruption a primary objective of its administration. The donor community has also increasingly emphasized good governance in dealings with individual countries and this is also a focus of reforms under the PRGF-supported program. One avenue to improving governance in Burundi is to bring fiscal transparency up to regional and international standards. Substantial enhancements in public expenditure management (PEM) are, as in most HIPC countries, a necessary condition for properly tracking social expenditures (including pro-poor spending) and improving fiscal transparency.46 Now, more than ever, it is necessary that a larger share of spending be allocated to, and actually spent, on pro-poor programs.

70. The main objective of this paper is to take stock of current fiscal management of public expenditures and identify weaknesses and needs for further reforms. As a basis for comparison, the paper applies public expenditure indicators formulated by the World Bank and the Fund, the HIPC Tracking Assessments and Action Plans (AAPs).47 It also compares the general fiscal transparency of Burundi against the Fund’s Code of Fiscal Transparency.48 The purpose is to identify major weaknesses and areas of further PEM reforms that could eventually be part of technical assistance programs. Section B discusses why PEM is important for macroeconomic management; why, with the approach of the HIPC completion point, Burundi needs to make further progress in controlling pro-poor expenditure; and why fiscal transparency and best practices in fiscal management matter. Section C makes recommendations for further PEM reforms so that expenditure tracking and fiscal transparency can continue to improve. Section D draws some conclusions.

B. Continuing PEM Reforms in Burundi

Macroeconomic stability and fiscal control

71. Macroeconomic stability, sustainable high growth, and poverty alleviation are positively linked with sound fiscal management.49 Private sector development depends on macroeconomic stability, which is based on sound financial policies (monetary and fiscal). Sound fiscal management is also essential to ensure that the authorities will have enough resources to enhance growth, fight poverty, and ensure financial sustainability by controlling public and external debt. Macroeconomic stability has been a main objective of the Burundian authorities’ plans and the PRGF-supported program which incorporates strong fiscal reforms. In Burundi, fiscal policy reform has been a key component of progress towards macroeconomic stability since 2004 and will continue to be the cornerstone of the program. The Fund-supported program has been directed to (i) achieving macroeconomic stability; (ii) improving transparency and governance; and (iii) supporting long-term economic growth by encouraging private sector development.

72. Fiscal policy, however, will not be effective if it is not accompanied by strong institutions, notably regarding expenditure management. Improvements in fiscal management in Burundi, especially public expenditure control, should continue to be a top policy priority and be supported by technical assistance. Through their influence on the efficiency of resource allocation, the microeconomic aspects of public sector management are critical to directing fiscal policy and control toward macroeconomic stability. Weaknesses in public expenditure management in Burundi, therefore, should be assessed against regional and international benchmarks to identify further needs of reforms in this area.

The HIPC Initiative and the need for better control of public expenditures

73. One of the HIPC completion point triggers covers the need for Burundi to demonstrate that a larger share of the budget is actually spent on pro-poor programs. When budgets are being prepared, it is easy to allocate pro-poor expenditures; it is much more difficult to ensure allocations are properly spent and consistently reported. Increased debt relief and the imminent possibility of scaled-up external aid means that donors will be more zealous in their concern that their resources be used to reduce poverty, reach intended beneficiaries, and make the intended progress toward achieving the MDGs.

74. Well-functioning PEM systems should help reduce donor reporting requirements and assure both domestic taxpayers and development partners that funds are being used as intended. This is even more urgent now that increasing amounts of development aid are taking the form of budgetary support. The importance of PEM is clearer in the light of several instruments drafted by multilateral institutions in recent years. Box 1 summarizes the main instruments these institutions are currently using to analyze budgetary systems and PEM capabilities. All these instruments cover at least the budget execution stage of PEM, reflecting the critical role of this area in fiscal policy. The instruments also make recommendations in the form of actions plans to overcome weaknesses and bring country performance to acceptable standards.

Instruments for Assessing PEM Capabilities

  • Public expenditure review (PER), created by the World Bank, looks at “upstream” issues of budget construction, assesses a country’s public expenditure program, checks on progress in meeting accountability requirements, and gives the government an external view of its budget.

  • Country Financial Accountability Assessments (CFAA) by the World Bank look at “downstream” aspects of the budget process: accounting and internal controls, financial reporting, fiscal transparency, and auditing.

  • Country Procurement Assessment Reports (CPAR) by the World Bank comprehensively analyze a country’s public procurement system—laws, organizational responsibilities and capabilities, current operating procedures and practices—and draw up a detailed action plan to improve institutions.

  • HIPC Tracking Assessments and Actions Plans (HIPC AAPs), carried out jointly by the World Bank and Fund, survey the extent to which PEM systems can assure that debt relief is used for poverty reduction. AAPs assess the capacity of a country to plan a program of spending, execute that plan, and produce timely and reliable reports on results.

  • Public Financial Management Performance Indicators (PEFA), carried out jointly by the World Bank, the Fund, and other partners, replaced the AAPs in the second semester of 2005. PEFAs indicators are broader in scope and use a modified approach to subject coverage and scoring system.

  • Fiscal transparency reports (ROSCs), implemented by the Fund, assess fiscal transparency based on 37 good practices organized according to four main principles: clarity of roles and responsibilities; public availability of information; open budget preparation, execution, and reporting; and assurances of integrity.

  • European Commission audits, largely based on financial and technical audits, verify the legality, legitimacy, and conformity of public expenditures in relation to aid funds provided by the Commission.

75. This paper uses the AAPs as a reference point for a preliminary evaluation of the current status of Burundi PEM performance. It also identifies areas for improvement. As Burundi approaches the HIPC completion point and donors consider increasing their aid and financing, it becomes ever more important for Burundi to generate reliable reports on budget execution and pro-poor expenditures. The next subsection uses another of the instruments (Fiscal Transparency Reports, ROSCS) to broader the perspective on fiscal management in Burundi by comparing it with best international practices.

76. The AAPs include 16 benchmarks of expenditure management capacity for all World Bank/IMF Heavily Indebted Poor Countries.50 The purposes of the AAPs are threefold: (i) to assess capacity for tracking all public spending, including poverty-reducing spending; (ii) to better understand risks associated with poverty-reducing spending, especially in the context of debt relief; and (iii) to clarify what donor and technical assistance will be necessary to improve systems for managing poverty-reducing and other public spending. The benchmarks are organized in terms of three dimensions of PEM: covering budget formulation, execution, and reporting. These areas provide the basic information needed to identify and monitor functional and economic and targeted spending.

77. Of the 26 countries assessed by AAPs up to 2004, a majority require substantial upgrading to have PEM systems capable of reliably tracking public spending (Figure 1).51 Only 2 countries, Tanzania and Mali, needed little upgrading; 5 more countries met 8 to 11 of the benchmarks, meaning they need somewhat more upgrading; and 19 countries, which met from 0 to 7 benchmarks, required substantial improvement in their PEM systems. A low-income, post-conflict country like Burundi is likely to show very low compliance with the AAP indicators. The rest of this section explores the 16 benchmarks as applied to Burundi.

Figure 1.
Figure 1.

Relative Need for Upgrading PEM Systems in HIPC Countries

Citation: IMF Staff Country Reports 2006, 307; 10.5089/9781451802900.002.A004

(Number in parentheses = total number of benchmarks met)Source: Fund-Bank AAP database

78. A preliminary analysis suggests that it meets only 3 of the 16 AAP benchmarks (Table 1). This preliminary analysis is not based on a full HIPC-AAP assessment process, and therefore the ratings should be regarded with some caution. At present, Burundi appears to meet the benchmarks on the reliability of the budget as a guide to the future, timely accounting for transaction and budget classification. Although Burundi has not yet implemented programmatic classification, it is not sufficient for arguing that the indicator cannot be met since the country is already reporting expenditure by administrative economic and functional classification. Recent improvements on budget execution and reporting will allow the country meet the internal fiscal reporting as soon as the quarterly budget execution report is consolidated and starts including comparisons with previous years.

Table 1.

Burundi: Comparison Between HTPC AAPs Indicators and Current PEM Practices

article image
article image

79. The poor performance of Burundi against the 16 AAP benchmarks should not be a surprise considering that the country has only recently emerged from a prolonged civil conflict. Deterioration of macroeconomic management capacity and collapse of the expenditure control system and revenue collection have been a common outcome of civil conflict in Africa. In several cases, the implications for macroeconomic control have resulted in monetization of an uncontrolled budgetary deficit and have fueled inflation. Lessons from other post-conflict countries highlight the importance of an early strengthening of budget control and tracking of expenditure—including budget formulation and execution and reform of civil service.52 PEM reforms currently being put in place, especially through the implementation in 2004-05 of an interim information financial management system (interim IFMIS),53 with technical assistance from donors suggest, however, that Burundi is likely to meet more of the indicators soon. Consolidation and deepening of PEM reforms is critical at this juncture and further steps and measures are discussed in section C.

Fiscal transparency

80. The empirical literature suggests that there is a strong positive relationship between fiscal transparency and economic outcomes.54 In Burundi, there was increased attention to good governance in the late 1990s—ultimately, the literature has shown, good institutions contribute to economic growth. Because of this link, fiscal transparency55 has received considerable attention from both policy makers and researchers, who have been particularly concerned with economic governance, particularly public sector management.

81. Why is it important for Burundi? Greater transparency will generate better information for fiscal policy design and decisions, allow stronger public and parliamentary oversight, make government more accountable for its actions (governance), and improve the environment for investment, both domestic and foreign. Fiscal transparency also lowers government borrowing costs and helps keep budget deficits low, which has positive implications for macroeconomic stability and stronger economic growth. Fiscal transparency also helps to identify weaknesses in fiscal policy, risks, distributional implications, and poverty effects. Finally, fiscal transparency may enhance public understanding of reforms and thus public support for responsible policies. Building political consensus is fundamental for countries like Burundi.

82. It has been shown that better-performing countries generally follow more transparent fiscal practices. Hameed (2005) reviewed the main literature, developed indices of fiscal transparency for a broad range of countries based on the Fiscal Transparency Code, and carried out a cross-section analysis using the transparency indices to examine how fiscal transparency relates to market credibility, fiscal discipline, and corruption. Some of the main conclusions:

  • Transparency makes it easier to achieve fiscal discipline, which, in turn, helps explain macroeconomic stability. For example, a medium-term budget outlook and careful analysis of fiscal risks are likely to result in more-disciplined budgets and better fiscal outcomes. Econometric analysis (Hammed, 2005) shows a significant relationship between primary balances and the main indices of fiscal transparency (including fiscal risks).

  • Transparency makes it easier to combat corruption because the government is more accountable and there is less tax discretion and more effective auditing. After controlling for certain geographical economic and demographic factors, transparency has been positively correlated with control of corruption.56

  • Transparency gives countries more credibility in financial markets and lowers the premium. For a country with access to capital markets, controlling for economic fundamentals, international financial markets are likely to demand lower premiums from countries that are forthcoming about their fiscal position and risks. Hammed (2005) analyses the relationship between a credit ratings variable57 and fiscal transparency indices for countries that have published fiscal ROSCs. The analysis showed that fiscal transparency matters for credit ratings. Although this result may not be of immediate interest to Burundi—participation in the international financial markets is still incipient—it may become important for later economic development because Burundi, a land-locked country, will depend strongly on trade, foreign direct investment, aid, and external financing.

83. Cross-country comparisons of fiscal transparency show differences in performance across regions and level of development. Observations for the 57 countries that have published fiscal ROSCs, 20 specific good transparency practices were assessed as observed, largely observed, or not observed (Table 2).58 Main results in Table 2 indicate that (i) advanced economies, as expected, generally comply best with good fiscal transparency practices, and (ii) sub-Saharan Africa countries generally have low compliance with good practices but they perform relatively well on some.

Table 2.

Observance of Selected Fiscal Transparency Practices by Region1

article image

Regional divisions reflect IMF area departments (except for the Western Hemisphere Department, since Canada and the U.S. are omitted from this table).

Data in the table show the number of countries where the practice is Observed or Largely Observed as a ratio of the total number in each group for which there was sufficient information to assess observance/nonobservance.

Indicates practices for which more than 10 percent of the ROSCs in the study did not have sufficient information from which to assess observance/nonobservance.

Preliminary assessment included in this report, which needs to be confirmed by a fiscal ROSC assessment.

84. It appears from a preliminary analysis that best practices of fiscal transparency in Burundi, like other sub-Saharan countries, are not being observed (Table 2), except for those on budget classification. Some of the benchmarks in fiscal transparency overlap with the AAPs indicators already analyzed. It is worth noting, however, that the PEM reforms put in place in recent years are improving the capacity of the country to become more transparent in terms of fiscal management. Consolidation of the current reforms, especially the interim IFMIS, additional capacity building, and further expansion of specific measures on improving universality of the budget and reporting, will substantially improve Burundian fiscal transparency.

85. Burundi is now poised to make quick progress on at least two areas of fiscal transparency practices—an open budget process and medium-term budgeting which correspond to best practices 1 to 6 and 17 to 20 in Table 2. The interim IFMIS system is being consolidated and with the first quarter of 2006, quarterly budget execution reports are now being published. Some improvements in the government’s capacity to prepare annual government accounts and the Audit Court’s capacity to analyze them will allow Burundi to comply with most practices encompassed by the title of “open budget processes.” Likewise, by including more information (forward estimates) in the budget and carefully identifying poverty-reducing activities, and by linking the annual budget with an MTEF, Burundi can be expected to improve the practices for numbers 17 to 20 in Table 1. Improvements in most of the fiscal transparency practices related to data quality, assurances of integrity, and reporting on fiscal risks, will take more effort but should be tackled by the time reforms on the open budget process and medium-term budgeting are in place.

C. Further PEM reforms

86. Although some progress was made in 2004 and 2005, weaknesses and inefficiency in PEM in Burundi are still hampering fiscal policy implementation. Section B indicated that Burundi is considered far behind in complying with indicators to track expenditures; especially pro-poor spending and fiscal transparency could be considered below the regional average, which suggests that Burundi is not using best practices.

87. Further PEM reforms should aim at consolidating the current implementation of an interim IFMIS and expand in areas that will help to improve compliance with the HIPC AAPs indicators and fiscal transparency. Public expenditure management, for example, is being reinforced from a very low base by an interim IFMIS, a crucial step toward improved budget execution and capacity to generate more reliable and detailed reports. The system is still restricted to the Ministry of Finance and there are several capabilities still to be developed, such as expansion to line ministries and completion of interfaces with systems like payroll and public debt management.

88. The extension of the interim IFMIS to all public expenditure and revenue will allow the Ministry Finance to prepare its annual government accounts a few months after the end of the fiscal year. In turn, this will permit the Audit Court to expedite its analysis and technical opinion on government accounts. It is feasible that by 2008, the Audit Court analysis can be considered by the Parliament within a year after the end of the fiscal year, allowing Burundi to comply with one more best practice in fiscal transparency. Additional training and resources for the Audit Court are still needed and the calendar of the government accounts must also be reviewed to reflect (i) time for preparation and approval of government accounts by the Ministry of Finance and Council of the Ministries (three to four months); (ii) time for Audit Court analysis and preparation of a formal opinion (six months); and (iii) submission of the conclusion to the Parliament.

89. Full functioning of the Audit Court and finalization of the first audit of the government accounts in 2005 will help to improve external controls and give independent assurance of the integrity of government activities helping Burundi to comply with another best practice of the Fiscal Transparency Code. Much more remains to be done related to internal controls and public availability of information. The new Audit Court needs to be reinforced. Recently, the government announced a reorganization of the Ministry of Finance and created a new department of internal control. This decision has obfuscated internal control roles because the Ministry of Good Governance has been in charge of internal control matters. The lack of clarity undermines the efficiency of the system and compromises the use of limited financial and human resources. The government should consider placing the government-wide internal control system back under the Ministry of Finance, which is more familiar with financial operations, budgeting, and cost-benefit analysis.

90. The interim IFMIS should be upgraded into a fully integrated budget financial management system when it is extended beyond the Ministry of Finance to line ministries and government agencies. This step should create additional capabilities in the system that will allow the line ministries to perform part of the budget execution process, for example by managing their own expenditure commitments. Currently, authority to spend budget allocations has been concentrated exclusively in the Ministry of Finance. Controllers reporting to the Ministry of Finance perform ex-ante controls on almost every phase of expenditure process (commitment, verification, service rendered). These cumbersome procedures generate delays and inefficiency in government spending. Furthermore, they dilute responsibility and accountability and encourage irregular practices to overcome them. Streamlining ex-ante controls in budget execution is urgently needed. It is also important to delegate to line ministries the responsibility to enter commitments into the interim IFMIS. This would be a critical step in PEM decentralization to line Ministries and government agencies, though the Ministry of Finance as the primary fiscal authority would retain firm control on global limits for commitment and payment.

91. Investment/project spending and recurrent expenditures can be integrated shortly using the capabilities of the current interim IFMIS. The potential scaling up of external aid in Burundi underscores that integration is a necessary condition to improve the quality of budgeting. Integration would increase efficiency because forecasting of future outlays (future recurrent expenditures linked to current capital expenditures) would be more effective and realistic.

92. Incorporating into the budgetary procedures donor investment and public projects and other extrabudgetary activities will help Burundi to comply with indicators 2 and 4 of the AAPs (Table 1) and fiscal transparency benchmark 8 (Table 2). This is related to a core universal principle in budgeting: comprehensiveness. In their technical assistance reports by World Bank and the Fund have made several recommendations to overcome this weakness: (i) inventory all extrabudgetary resources; (ii) standardize the reporting of expenditures and revenues by semiautonomous agencies; and (iii) identify off-budget expenditures and their effect on recorded revenue and expenditures with a view to their later inclusion in the budget.

93. Government bank accounts can be better controlled by setting up a single Treasury account (or a consolidated government bank account) in the central bank. By controlling the use of financial resources in government accounts, the Treasury could substantially reduce the need for short-term bank financing. This would also improve central bank liquidity control and implementation of monetary policy because use of government bank resources becomes much more predictable. This measure will help Burundi to comply with HIPC AAPs indicator 11 (Table 1).

94. Short-term measures can be taken to help harmonize budget execution reports and other fiscal reports on such topics as banking financing and public debt. The interim IFMIS allows the Ministry of Finance to put together the main information related to budget expenditures and revenues. By incorporating other sources of revenue, such as revenues generated independently by government agencies, public debt operations, and bank financing, the Ministry of Finance can generate consistent reports on government financial operations. The Ministry is currently producing a quarterly report on budget expenditure; the objective is that it covers all budgetary transactions. The budget execution report should at least cover functional and economic classifications and comparisons with the total allocation (as modified during the fiscal year), and quarterly projections should be made for the main items. Comparisons with actual figures in the same period in previous years would also help make fiscal performance clearer. A section on poverty expenditure should detail pro-poor spending in terms of evolution and main priority areas. These measures, together with inclusion in the budget of contingent liabilities and quasifiscal activities and regular publication of public debt in the budget execution reports, will improve Burundi observance on fiscal transparency best practices No. 11, 12, 13, and 14 (Table 2).

95. The law governing public finance needs to be completely overhauled. Budgetary and financial resources need to be consolidated because not all revenue is deposited in the Treasury account. There is no formal obligation to produce end-of-year government accounts, which may complicate the functioning of the Audit Court. Furthermore, there is currently a serious potential conflict of interest built into budget execution: the person or area in charge of issuing payment orders is also the government accountant, which can obviously lead to undesirable management practices. Best practices recommend that these roles be carried out by different entities. In general, the Organic Budget Law (Loi Organique des Finances) needs to be updated to provide: (i) a clear definition of the main principles and procedures for managing public resources; (ii) a requirement that government produce governments accounts on schedule and submit them for audit by the Audit Court, whose reports will be analyzed by Parliament; (iii) integration of current and investment expenditure budget; and (iv) a requirement that annual budget global ceilings of expenditure be linked with MTEF which has to be produced regularly.

96. Now that the first complete PRSP is being finalized, it is feasible to harmonize medium- and long-term planning and monitoring instruments. Therefore, short- and medium-term fiscal documents like the annual budget and its reporting system should be totally consistent with the PRSP. It is recommended that an MTEF be progressively introduced starting with the complete PRSP. One option would be to phase in basic elements of program budgeting, for example, linking budget request to outcomes in specific areas, such as primary education and immunization programs. The challenge for program budgeting is that activities that are part of a single program are implemented by different agencies and even line ministries—another reason for consolidation of current and capital expenditure programs.

97. Accumulation of expenditure arrears have been substantial—about 3 percent of GDP in 2004 and has certainly undermined the quality of fiscal policy implementation and has impeded the country to comply with indicator 8 (Table 1). The efficiency of the government may decline when budgetary arrears are high and persistent; certain areas, such as health services, can experience severe deteriorations in service delivery from late or no payment of invoices. In 2005 and 2006, the government made progress on clearing expenditure arrears (Box 2) and it should attempt to meet obligations as they fall due avoiding any form of nontransparent financing in order to comply with indicator 8 of Table 1.

Budgetary Arrears Audit and Strategy for Clearance

Arrears: a recurrent problem

The weak budget preparation and execution associated with civil conflict in Burundi over the past decade have led to a significant accumulation of budgetary arrears to about 3 percent of GDP. Besides the negative implications for fiscal policy, the accumulated arrears have also generated lawsuits against the government; impaired the mechanisms for public procurement; and created a public perception that the government has not managed its budget and finances properly.

Government arrears inventory and independent audit

The government decided to face this problem by inventorying the precise amount of arrears and the creditors. The arrears inventory was finalized in 2005 and audited by an external firm in the first quarter of 2006. The amount still outstanding in 2006 is FBu 25.7 billion or 3 percent of GDP. The creditors were varied: private banks, companies, and suppliers; public enterprises and autonomous entities; and even some individuals.

Strategy for clearance: transparent and consistent with fiscal policy

The next step was to design a strategy to clear the arrears without impairing budget balance, undermining fiscal policy consistency with macro variables, and compromising priority expenditures. The strategy being adopted comprises the following principles and measures:

  • The strategy should be transparent. The government decided to make it explicit in the 2006 revised budget. Regular payments and issuing of additional government liabilities are now being presented in the budget law and incorporated into the budget execution process.

  • An official government decision will state the total amount of arrears for each creditor and the modalities for their settlement.

  • FBu 100 million. Total cash payments in 2006 are estimated in FBu 7.4 billion (0.9 percent of GDP).

  • For the remaining validated arrears, a Treasury certificate (with no interest) recognizing these debts will be issued for the nominal value, which is estimated to total in FBu 18.3 billion (2.1 percent of GDP).

  • The Treasury certificates could either be freely exchangeable for Treasury bonds of three to five years maturity auctioned by BRB on behalf of the Treasury, or used in the privatization program to settle winning offers of government assets.

98. Debt management and reporting (which is part of the fiscal transparency benchmarks) is also a HIPC trigger according to the decision point document. Supported by technical assistance from donors, the government has installed some small capacity in the Ministry of Finance, in coordination with the central bank, which has maintained a public database on domestic and external debt. Monthly reports on debt are now being produced although data reliability needs considerable work. Accounts are not regularly reconciled with creditors’ statements and the information in the system is not being updated on a timely basis.

99. Wage bill control in Burundi is unsatisfactory, while its share of budgetary allocations has been increasing (see Box 3). Though the government wage bill is the most important single item in the budget, amounting to some 10 percent of GDP, the Ministry of Finance does not control payroll and normal budget procedures are not followed. Urgent reform and clear definition of long-term policy are needed. Like expenditure on investments, some items of recurrent expenditures such as wage bill need to be factored into the MTEF and linked to annual budgets.

The Wage Bill in Burundi

Because payroll has become the largest recurrent expenditure item in the Burundi budget, improving its management has become a priority for public expenditure reform. Reports prepared by the government and donors have identified specific weaknesses in current payroll procedures and controls.

The magnitude of the problem

  • Wages constitute 24 percent of total expenditure, which in 2006 is expected to achieve about 10 percent of GDP;

  • Their weight in the budget has been increasing steadily, from 7 percent of GDP in 2002 and 9 percent in 2004;

  • The government has around 100,000 people on the payroll, a high proportion of the total population compared to neighboring countries;

  • About 50 percent of these are in the security forces (the army and the new national police).

The civil servant average salary is still low (about US$60 dollars per month) and there is a clear need for a firm strategy to increase average salaries while bringing the total wage bill outlay below 10 percent of GDP. Demobilization and integration of rebel groups plus eliminating ghost workers from the payroll will certainly help to increase the average salary without jeopardizing the budget.

Lack of control

A recent government census of civil servants identified more than 4,000 ghost workers of a total of 46,000 employees. Currently, computerization of the payroll is limited and subject to errors, and the civil servant roster is not timely updated.

A reform strategy might be to transfer management of the civil payroll to the Ministry of Finance, which is better equipped and more familiar with the financial aspects of payroll payments. That would leave the Ministry of Good Governance in charge of designing civil servant wage policy (subject to broad budgetary limits defined in the budget and in the MTEF).

D. Conclusion

100. This paper has stressed the importance of fiscal policy for macroeconomic stability and the critical role of PEM in fiscal policy. By using benchmarks provided by the HIPC AAPs and fiscal transparency ROSCs, the paper concludes that it is urgent that Burundi consolidate its current PEM reforms and move forward purposefully to bring fiscal management to regional and international standards.

101. Although Burundi has been making progress in PEM since 2004, as this paper shown, much remains to be consolidated, extended, and advanced. Some of the PEM instruments analyzed in Section II might be helpful at this juncture. For example, a complete fiscal transparency ROSC and HIPC PEFA (the replacement for the AAP) for Burundi would provide more specific information on areas that need attention and would give detailed actions plans for continuing the reforms. A World Bank PER exercise would also benefit PEM reform, both current and future. Two priorities areas for a PER would be drafting medium-term plans and an MTEF and integrating them into the annual budget, and identifying pro-poor spending priorities for follow-up in budget preparation and execution.

References

  • Abed, George, and Sanjeev Gupta (editors), 2002, Governance, Corruption and Economic Performance (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Alt, James E., and David D. Lassen, 2003, “Fiscal Transparency and Fiscal Policy Outcomes in OECD Countries,” Economic Policy Research Unit Working Paper, No. 2003-2 (Paris: OECD).

    • Search Google Scholar
    • Export Citation
  • Blejer, Mario, and Ke-young Chu (editors), 1989, Fiscal Policy, Stabilization and Growth in Developing Countries (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Calcoen, Nicolas (2004), “Rebuilding Fiscal Institutions”, in Postconflict Economies in Sub-Saharan Africa: Lessons from the Democratic Republic of the Congo, ed. by Jean A. P. Clément (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Gupta, Sanjeev, Benedit Clements, and Gabriela Inchauste, 2005, Helping Countries Develop: The Role of Fiscal Policy (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Hameed, Farhan, 2005, “Fiscal Transparency and Economic Outcomes,” IMF Working Paper, No. 05/225 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • International Development Association and International Monetary Fund, 2005, “Update on the Assessments and Implementation of Action Plans to Strengthen Capacity of HIPCs to Track Poverty-Reducing Public Spending,” FAD of the International Monetary Fund and PREM Sector Group of the World Bank, April 2005 (Washington: International Development Association and International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, 2001, “Code of Good Practices on Fiscal Transparency,” (Washington: International Monetary Fund).

  • International Monetary Fund and the World Bank, 2003, “Public Expenditure Management—Country Assessment and Action Plan (AAP) for HIPCs,” PREM Sector Group of the World Bank and FAD of the International Monetary Fund, October 2003 (Washington: International Monetary Fund and the World Bank).

    • Search Google Scholar
    • Export Citation
  • Kopits, George, and Jon Craig, 1998, “Transparency in Government Operations,” IMF Occasional Paper No. 158 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

APPENDIX: Summary of the Burundi Tax System

(as of December 31, 2005)

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

STATISTICAL APPENDIX

article image
article image
Sources: Burundi authorities; and Fund staff estimates.

World Bank development indicators.

Including net short-term capital and errors and omissions.

Table 1.

Burundi: Gross Domestic Product at Current Prices, 2001–05 1/

(In billions of Burundi francs)

article image
Sources: Burundi authorities; and Fund staff estimates.

No official National Accounts have been produced since 1998. However the authorities do provide estimates, based on expert judgment and the few statistics available, at the Ministry of Planning.

Table 2.

Burundi: Gross Domestic Product at Constant 1996 Prices, 2001–05 1/

(In billions of Burundi francs, unless otherwise indicated)

article image
Sources: Burundi authorities; and Fund staff estimates.

No official National Accounts have been produced since 1998. However, the authorities do provide estimates, based on expert judgment and the few statistics available, at the Ministry of Planning.

Table 3.

Burundi: Supply and Use of Resources at Current Market Prices, 2001–05 1/

article image
Sources: Burundi authorities; and Fund staff estimates.

No official National Accounts have been produced since 1998. However, the authorities do provide estimates, based on expert judgment and the few statistics available, at the Ministry of Planning.

As recorded in the balance of payments.

Derived as residual.

Table 4.

Burundi: Savings and Investment, 2001–05

article image
Sources: Burundi authorities; and Fund staff estimates.

No official National Accounts have been produced since 1998. However the authorities do provide estimates, based on expert judgment and the few statistics available, at the Ministry of Planning.

Excluding official transfers.

Including official transfers.

Table 5.

Burundi: Key Coffee Sector Indicators, 2000/01–2004/05 1/

article image
Sources: Burundi authorities; and Fund staff estimates.

The coffee crop year extends from May 1 to April 30.

Coffee export taxes were abolished in 2000.

Table 6.

Burundi: Structure of Arabica Coffee Prices at Different Stages of Production and Export Results for the Sector, 2000/01–2004/05 1/

article image
Sources: Burundi authorities; and Fund staff estimates.

The coffee crop year extends from May 1 to April 30.

Determined as average rate, on the basis of actual payments in Dar es Salaam in relation to exportable coffee output.

Deflated by the consumer price index (1991=100).