1 Lessons on Sequencing Financial Management and Accountability Systems in Decentralized Post-Conflict/Low Capacity Contexts: the Case of Rwanda* David Sewell January, 2004 *Draft. Not to be Cited 1 2 2 3 ACRONYMS AFRITAC-East BNR CDF CEPEX CFAA CSR DCDP DFID FAD FDP GIF HIPC HIPC AAP: LABSF MDGs MINALOC MINECOFIN MININFRA MTEF NTB OAG OBL PEM PFM PETS PSCBP PRSP RALGA RF RRA SIBET USAID VAT Africa Technical Assistance Center –East (IMF) Banque National du Rwanda (National Bank of Rwanda) Common Development Fund Central Projects and External Finance Bureau Country Financial Accountability Assessment Caisse Sociale du Rwanda (Social Security Fund) Decentralization and Community Development Project Department for International Development (U.K.) Fiscal Affairs Department (IMF) Fiscal Decentralization Project General Inspectorate of Finance Highly Indebted Poor Countries (Initiative) Assessment and Action Plan for public financial management Local Authorities Budget Support Fund Millennium Development Goals Ministry for Local Government and Social Affairs Ministry of Finance and Economic Planning Ministry of Infrastructure Medium-Term Expenditure Framework National Tender Board Office of the Auditor General Organic Budget Law Public expenditure management Public financial management Public Expenditure Tracking Surveys Public Sector Capacity Building Project Poverty Reduction Strategy Paper Rwandese Association of Local Government Authorities Rwandan Francs Rwanda Revenue Authority Système du Budget de l’État (budget execution system) United States Agency for International Development Value-Added Tax 3 4 LESSONS ON SEQUENCING FINANCIAL MANAGEMENT AND ACCOUNTABILITY SYSTEMS IN DECENTRALIZED POST-CONFLICT/LOW CAPACITY CONTEXTS: THE CASE OF RWANDA Introduction This is one of two case studies undertaken by the author for the World Bank’s PRMPS unit as a member of a team that is considering issues which arise in the sequencing of financial management and accountability measures in post-conflict or low-capacity countries where decentralization is taking place. The work was financed by a World Bank-Netherlands Partnership Program Global and Regional Initiatives grant. Terms of reference for the study are given in Annex A.1 Rwanda was one of two countries selected for case studies and visited by the mission team in October 2004, the other country being Sierra Leone. Both of these countries have emerged from vicious civil conflicts in the last decade with a common analysis that centralization had played a large part in bringing about the conflicts and decentralization was needed to address causes of the problem. In both cases, decentralization is being accompanied by reforms in financial management and accountability. Rwanda’s Recent History How did Rwanda come to be in its present situation? The civil war and genocide that accompanied it in Rwanda from 1990-1994 are without equal in recent history. While estimates vary, a frequently cited figure is that close to 1 million people perished in the genocide. The Poverty Reduction Strategy Paper (PRSP) of the GoR in 2002 noted that another three million were driven into exile and 107, 000 were imprisoned. To get an idea of the relative scale of these tragic events, the population of Rwanda was only 8.2 million in 2002. So the genocide more than decimated the population and drove more than a third into exile. 11 The other member of the team, Antonio Carvalho, is reporting separately on the importance of decentralized information and communications technology for financial management and accountability in post-conflict and low-capacity situations. For inputs into this paper, the writer would also like to thank Joseph Kizito, Mohamed Toure, Kampeta Sayinzoga, Duncan Last, Dana Weist and Kai Kaiser. 4 5 Substantial destruction of infrastructure and physical capital also took place, increasing poverty. But more than physical capital was destroyed. According to the 2002 Rwanda Country Assistance Strategy The professional and technically skilled people (doctors, teachers, nurses, etc) were singled out for killing in the genocide. While the majority of the refugees have returned, the educated and qualified among them have largely remained in exile.2 This development is of some importance to the present study. In terms of our inquiry, a post-conflict country need not necessarily be synonymous with a low-capacity country. A principal contributing factor to the “German economic miracle” after WW II, for instance, is said to have been that critical “human capital” in the form of highly trained professionals was available to spur growth, despite the fact that destruction of infrastructure and physical capital had been immense. Rwanda was obviously less fortunate in this respect in recovery from its own tragic history. The conflict in Rwanda officially ended in 1999, with the signing of the Lusaka Peace Treaty, although implementation of the treaty only began in 2001. But the security threat around Rwanda’s borders continues even at the time of writing and regional instability remains a menace to peace and development.3 There is some question, indeed, as to whether Rwanda can even now be considered to be a post-conflict country, And Rwanda has other problems that were exacerbated by the conflict. Many thousands suffer from resulting mental and physical handicaps as a result of the conflict. And Rwanda is heavily affected by the HIV/AIDS epidemic: around one in seven of those in the prime labor force years have HIV/AIDS. 4 One of the consequences of this tragic history is the extent of poverty in Rwanda. By 2002, several years after the conflict officially ended and after a period that included substantial economic growth, the GoR’s PRSP showed that 60% of the households in Rwanda population were living below the poverty line. The Government of National Unity that was formed after the Lusaka Accord embarked on an ambitious program of reconstruction and reconciliation. The National Unity and Reconciliation Commission (URC) initiated a countrywide dialogue identifying poor justice, governance and leadership, and poverty as the major obstacles to national unity. These consultations gave impetus to the efforts to reform institutions so as to foster good governance, including the decentralization of government. At the national level, a new Constitution was approved in a countrywide referendum in May 2003 and national elections followed in August 2003. But well before this the GoR 2 World Bank, Country Assistance Strategy for the Republic of Rwanda, November 21, 2002, Report No. 24501-RW, p.2, para. 7. 3 See for instance the Washington Post, “Congo, Rwanda at Brink of War as Border Fighting Surges,” Dec.16, 2004, p. A-26. 4 The 2002 Rwanda CAS reports a 13.7% incidence of HIV/AIDS among the population aged 15-49. 5 6 had taken substantial steps to decentralize government in Rwanda. In part, this emphasis on decentralization arose because the former system of centralized government, which was considered to have encouraged unquestioning obedience to orders from above, was seen as an origin of the genocide.5 Previously, local administrations had consisted of 354 communes, with their officials being centrally appointed. In 2000-2001, the GoR first adopted a National Decentralization Policy and then a Fiscal and Financial Decentralization Policy and in the same years the National Assembly passed decentralization laws, involving devolution of powers to 106 elected local governments (107 including Kigali) and deconcentration to eleven provinces.6 Devolution involves the allocation of political decisionmaking power to subnational governments. The latter are given the authority to decide what is done as well as the power to implement it and are accountable to their own electorates for their actions. Deconcentration means the dispersion of responsibilities within a central government, to regional branch offices or local administrative units. Elections were held for the new district administrations in March 2001 and in the remainder of the year the elected assemblies in the new districts were mostly occupied with forming their governments and recruiting technical staff. The first year of normal operations for the district councils was 2002. The district governments immediately assumed the functions previously carried out by the communes that preceded them. These are the core functions of local governments throughout the world: those of transportation (local roads), environmental services (local water, sewers, solid waste collection), and regulation of local land use and commerce.7 The first set of functions that are being devolved to local governments are the social services of health, education and social welfare. The next largest sector to be devolved is agriculture. More will be said subsequently about the financial management implications of devolving these substantial sectors of public spending to local governments Outline of the Paper. The rest of this paper will assess the principal areas of public financial management in Rwanda: budget preparation and planning; budget execution; and accounting and reporting, auditing and external scrutiny. As required by our terms of reference, special attention will be paid to the sequencing of financial management and accountability measures. The paper will also pay particular attention to financial management applications and decentralization in Rwanda’s local governments. “Decentralisation is central to the creation of democratic structures of governance in Rwanda. Encouraging people to work together at a local level is central to overcoming the divisions that have been so destructive in the past. We are creating a sense of local ownership of public programmes, which has been lacking. Decentralisation also allows local Governments to respond to local needs, and can increase the accountability of Government to the people.” Government of Rwanda: Poverty Reduction Strategy Paper. Ministry of Finance and Economic Planning, June 2002, p.62. 6 As is common throughout the world for capital cities or regions, there are also separate provisions for Rwanda’s capital of Kigali. 7 Elsewhere, these core local government services might also include local public safety (police protection for property, fire protection). 5 6 7 Developments in Financial Management A brief account of the more important post-conflict developments over time in financial management and accountability in Rwanda will now be set out. The interaction between the GoR and the development partners will be emphasized. In 2002, donors provided 88% of Rwanda’s development budget and 53% of its recurrent budget, for a total of 55% of public expenditures. This substantial input of resources by the development partners has been accompanied by a joint emphasis with the GoR on increasing accountability and transparency. There were three major developments in 1998. Perhaps most important in terms of a new institutional framework for financial accountability was the establishment of the Office of the Auditor General (OAG). This is only one of two Offices of the Auditor General in Francophone African countries, and the most distinctive feature of the Office (as opposed to the Cour des Comptes which preceded it in Rwanda) is that it reports to Parliament and not the Executive. In May 2003, Rwanda’s new Constitution formally established this reporting relationship to Parliament, and the OAG completed its first report to Parliament in January 2004. The OAG conducted 30 audit missions of key ministries in 2002, and 60 in 2003, which included a number of districts. The Governments of Sweden and the Netherlands have assisted the OAG with capacity building. The National Tender Board (NTB) was also created in 1998 to oversee procurement. And in the same year, the Central Projects and External Finance Bureau (CEPEX) was established as a semi-autonomous body within MINECOFIN to improve the coordination of investment projects with donors. In 2000, the GoR adopted a medium term expenditure framework (MTEF), which is a policy-determined limit to the total expenditures that are to be undertaken during determining spending. In Rwanda’s case, the MTEF determines spending for a 3-5 year period. Such a medium or longer-term perspective on budgeting is said to have been at the heart of successful budget reform in much of the world over the past two decades. 8 The MTEF is seen as integral to execution of the PRSP. The MTEF was introduced in the 2001 Budget with the assistance of DFID and the World Bank. Its implementation has been supported by sectoral public expenditure reviews by the Bank and other development partners, including social sector expenditure reviews in 1999 and 2000, and reviews of the transport and agriculture sectors in 2002. Several diagnostic studies of Rwanda’s fiduciary framework have been undertaken. In 2001, the Bank and the Fund carried out the first of two Assessment and Action Plans for public financial management in Rwanda under the Heavily Indebted Poor Countries Initiative. HIPC AAPs provide a checklist of PEM achievements and deficiencies and give an account of the three main areas of budgeting—budget preparation, budget execution, and budget reporting. They go beyond analysis of the adequacy of PEM The Bank’s September 2004 Poverty Reduction Support Credit and Grant (PRSC) document considers adoption of the MTEF to be “a comprehensive framework to further PEM reform” (para.44). 8 7 8 systems and also create action plans to correct deficiencies and indicate what additional technical assistance should be provided to support the action plan. A second HIPC AAP was undertaken in 2004 to update the findings in the 2001 study and its findings will be reported subsequently. Following the 2001 HIPC, the World Bank and DFID sponsored the Financial Accountability Review and Action Plan (FARAP), a limited version of the more usual Country Financial Accountability Assessment (CFAA) undertaken by the Bank. The FARAP diagnostic mission was conducted in 2002, and the report issued in February 2003. The EU sponsored a follow-up to the FARAP, with recommendations on its implementation.9 Also in 2003, the IMF’s Fiscal Affairs Department issued its principal financial accountability document for Rwanda, the Report on the Observance of Standards and Codes, Fiscal Transparency Module (ROSC). Since the decentralization process was started in 2001, major steps have been taken to develop financial capacity in the elected local administrations. They include installation of the local budgetary and accounting system which was sponsored by USAID, and training accountants from all districts in use of the system.10 Similarly, training in MTEF techniques was provided to all districts. 11 RALGA, the Rwanda Association of Local Governments, has played a pivotal role in provision of this training. Another major step was the publication of the Financial Management and Accounting Procedures Manual for Local Administrations. The manual covers all phases of budgetary management including preparation and execution of budgets and internal control, accounting, auditing and reporting. A useful measure of public expenditure management of fairly recent development that has been applied in Rwanda is that of Public Expenditure Tracking Surveys (PETS). These surveys determine the extent to which funds reach the facilities that provide services to final consumers. The sole PETS that has been done in Rwanda was of the flows of resources to and from the providers of public services in the health and education sectors in Rwanda in 1998 and 1999. The 2004 HIPC AAP finds that this PETS usefully supplemented internal controls. One of the most important developments in financial management occurred in November 2003, with the signing of the Partnership Framework for Harmonization and Alignment of Budget Support. This is a memorandum of understanding between the GoR and donors providing budget support (as opposed to say, support for specific projects). The Framework will lead to better coordination of external funding of the budget. An aspect of the Framework that is of importance in its own right is that donors agree to harmonize 9 Government of Rwanda and European Commission, Development of a Public Financial Management Capacity Building and Technical Input Plan. Rapport Provisoire, by Jean François Bauer and Bernard Biche, IDC, 22 Mars 2004. 10 See Ricardo Valderama, “Budget Execution and Financial Position of Local Governments in Rwanda.” ARD Rwanda Fiscal Decentralization Project, July 2004. 11 Obidegwu says that commitment to the METF training was greater in the districts than among the leaders of many central government ministries See Chukwuma Obidegwu,. “The Medium-Term Expenditure Framework: The Challenge of Budget Integration.” p.45 8 9 their practices with respect to public financial management and government fiduciary arrangements. In collaboration with the World Bank, the GoR in 2003 completed a Country Procurement Issues Paper (CPIP), which is analogous to the Country Procurement Assessment Reviews other countries have undertaken with the World Bank. The Paper involves a number of recommendations to improve the procurement environment so as to make it conform more closely to international standards. An expenditure monitoring system (the monthly “flash” reports on actual and budgeted expenditures) was also established in 2003. 2004 HIPC AAP. As noted above, a second HIPC AAP was undertaken in 2004 to update the findings of the 2001 study. It is perhaps the most useful and recent assessment of Rwanda’s PEM system. Its findings are of importance to the themes of the present paper and will be briefly summarized. The 2004 HIPC assessment was undertaken in response to the request of the Executive Boards of the World Bank and the IMF for another comprehensive assessment of all HIPC countries. The 2004 assessment includes an indicator on procurement, which was not separately treated in the 2001 assessment. The 2004 HIPC AAP also focused on the central government’s capacity to track pro-poor spending, the issues arising from the program of decentralization, and public accounting practices. In 2004, Rwanda satisfied 9 of the 16 HIPC benchmarks; a performance which situates it in the middle group of HIPC countries. Rwanda’s performance was strongest in budget preparation and planning, where it satisfied 5 of the 7 benchmarks, and weakest in budget execution, where it only satisfied 1 of the 4 benchmarks. Rwanda’s performance in relation to the HIPC benchmarks is shown in Table 1 of Annex B. The deficiencies that were found in PEM in 2004 can be briefly outlined and assessed. Two principal problems were found in budget preparation. First, “own” revenues generated by extrabudgetary funds amounting to 12.5%-15% of their revenues are not reported in the budget. Some major activities are involved, including Rwanda’s pension and road funds. But the principal problem found in budget formulation was that treatment of donor funds is inadequate, and it will be recalled that donor funding amounts to over half of public outlays in Rwanda. All donor funds are included in budget reports, but the HIPC AAP finds that “it is difficult to identify exactly where and how well these funds are being spent.” There are also problems of classification in that the development budget includes many recurrent expenditures. Again, this problem arises from “incoherent reporting of donor-funded projects.” Some donor funding is recorded in the district budgets but is 9 10 missed in the national budget and some in-kind assistance is included in budget projections but not always reflected in reported spending. On budget execution, significant stocks of arrears arose in the period before and immediately after the genocide and continue to exist. There is also a smaller stock of annual arrears, and it is noteworthy that this arises principally from the uncertainties of timing of external budget support funds. Viewing Rwanda in a wider context, the writer is inclined to think lack of budget discipline is likely to be a lesser problem than elsewhere. The Rwandans have a culture of budget discipline:12 their dedication to living within their means—which is what is implied by cash budgeting—being an example. But as will be seen, cash budgeting does lead to other problems in financial management. Although line ministries, provinces and districts all have internal auditors, the HIPC AAP assesses internal control as being ineffective because the roles and responsibilities of the internal auditors remain unclear. Nor is the quality of fiscal information assured by regular reconciliation of government bank accounts with government’s accounting records. Public accounting issues were found to be of “urgent concern.” As noted above, the 2004 HIPC AAP finds that the PETS for 1998/99 usefully supplemented internal controls. The writer observes, however, that the PETS did not track salaries and wages and these amount to more than 80% of recurrent costs. Wage and salary costs will be tracked in the forthcoming 2005 PETS, however, thus rectifying this deficiency. A separate analysis of the PETS for 1998/99 by Fofack, Obidegwu and Ngong adds to the remarks in the 2004 HIPC AAP.13 Substantial delays were found in the process of transfers of public resources from the central administration to primary beneficiaries. The Survey found that large and variable discrepancies existed between the amounts transferred by the treasury to regional health and education offices for local administration of facilities, and the corresponding amounts recorded in the books of these offices. According to Fofack and his colleagues, the Survey found “rampant lack of accountability in these offices,” with poor bookkeeping and lack of internal financial controls and auditing requirements. The discrepancies could have been due, therefore, either to leakages in the system or simply to unreliable bookkeeping. But in any case, Fofack et al. find that “the lack of accountability created an atmosphere for leakages and mismanagement of funds.”14 See Chukwuma Obidegwu, “The Medium-Term Expenditure Framework: The Challenge of Budget Integration,” p.44. 13 Hippolyte Fofack , Chukwuma Obidegwu and Robert Ngong “Public Expenditure Performance in Rwanda: Evidence from a Public Expenditure Tracking Study in the Health and Education Sectors,” Africa Region Working Paper Series No. 45, March 2003 . 14 In another fair comment, the Fofack analysis asserts that “the failure of accountability was not limited to local education and health offices and was probably pervasive in the departments of ministries and other government agencies. The PETS findings have simply highlighted the problem that has been known to exist.” 12 10 11 With respect to budget reporting, the HIPC report found that there were no regular fiscal reports that track poverty-reducing spending. And the public procurement system was not considered to be efficient and effective. There is a competitive tendering process, but the AAP found that it is undermined by a weak governance environment that lacks adequate internal controls to enforce the rules. To sum up, it is significant that many of the major problems found by HIPC in both budget formulation and execution arise from donor funding, and this subject will be further examined below. Current Initiatives Legislation. At the time of writing of this paper in January 2005, further legislative developments were occurring. For instance, the provisions of the Constitution call for the Auditor General’s independence, but the FARAP had identified deficiencies in this respect in the legal framework establishing the OAG. Revisions to the Audit Act are now under consideration to rectify this problem. A second important piece of legislation still under consideration by Parliament is the Organic Budget Law. This was submitted to Parliament in June 2004, and accompanying financial instructions were issued in May 2004. Inter alia, the Organic Budget Law provides the framework for a shift to a decentralized system of financial management in which control is exercised ex-post through the OAG. Following on the CPIP, a new Procurement Law has also been drafted. Improvements in Accounting. The current consensus among donors is that the most needed public financial management improvements involve improving public accounting. The IMF’s ROSC in 2003 wanted the GoR to concentrate on “the basics” of public financial management, which it defined as developing a high quality government accounting system, and good internal and external audit. The 2004 HIPC AAP also found that public accounting issues are of “urgent concern” and highlights a number of immediate actions that can be taken to improve accounting standards and practices. Finally, the Bank’s 2004 PRSC paper concludes that “the biggest medium-term challenge [in public financial management] is to develop a decentralized, integrated, high quality accounting system” (para 112). At the time of the mission for the present paper in late 2004, a study of public accounting needs was ongoing and was led by the EU. Other Developments in Revenue Collection and Decentralization While the above changes were taking place, significant developments were also occurring in related financial fields. For instance, the Rwanda Revenue Authority (RRA) was established in 1998 as a semi-autonomous agency for revenue collection. A VAT was 11 12 introduced in 2001 as a principal revenue-raising instrument and the VAT rate was subsequently increased from15% to18% in 2003. The GOR commissioned an evaluation of its decentralization policy by the Dutch consultants VNG International, which reported in September 2003, at the same time as the five year Decentralization Implementation Program was introduced.15 An IMF study was also undertaken in 2003 “to review the allocation of fiscal responsibilities between levels of government and to prepare an action plan for the management of the fiscal decentralization process currently underway.”16 Intergovernmental Transfers Rwanda’s two main intergovernmental transfers, the Local Authorities Budget Support Fund (LABSF) and the Community Development Fund (CDF), have been instituted to finance local government recurrent and capital expenditures respectively. Both are meant to provide local governments with a steady flow of funds through a set share of national domestic revenues. Legally mandated flows in 2004 for LABSF and the CDF were 5.3 and 10 percent of national revenues (excluding donor funds) respectively. It is obvious that these funds are, of course, merely a small part of spending on health education and the social services. As Kaiser indicates, responsibilities for these functions are not clearly delineated at present.17 Teachers’ salaries, for instance, make up the bulk of educational expenditures, are paid by the central government and administered through the budgets of the deconcentrated provincial governments. Similarly, capital expenditures for schools are administered through the CDF in the central government budget. While the stated intention of the GoR is to devolve health, education and the social services to the district councils, it is worth exploring what devolution would entail. Were there to be full devolution of powers in the three areas of health, education and welfare spending, the budget of the central government would be reduced by a quarter and the collective budgets of the district councils would increase tenfold. To this writer, it is inconceivable that these resources will be raised by the central government and simply handed over to the district councils without ties. There are legitimate national interests that need to be advanced in these fields.18 Such national interests can be pursued without micromanagement of the functions concerned, however, by judicious use of transfers whose receipt is conditional on meeting defined national 15 VNG International, Appraisal of the Decentralization Process in Rwanda. Ministry of Local Government, Information and Social Affairs, September 2003. This study is, however, classified as “strictly confidential.” See Robert Gillingham, Duncan Last, Ronald Neumann and Russell Robinson, “Rwanda: Managing Fiscal Decentralization”, IMF Fiscal Affairs Department, August 2003. 17 Kai Kaiser, “Fiscal Equalization in Rwanda: Challenges for Decentralized Financial Management, Reporting, and Enhanced Equalization”, draft, July 12th, 2004, para. 68. 18 To state the obvious, a nation with this history must have a vital interest in the curriculum of primary and secondary education and its influence on nation-building. 16 12 13 standards and criteria. There is wide international experience that could be drawn on to fashion such grants. The desire to preserve autonomy for local governments arises for other reasons. One can have sympathy with the districts, which are expected to comply with the equivalent of many unfunded mandates coming down from the national and international levels. In addition to carrying out their purely local functions, for instance, the local governments are also expected to contribute to achievement of the Millennium Development and Poverty Reduction Strategy goals, as well as Rwanda’s Vision 2020. At present, Kigali is probably the only local government with sufficient own-source revenues to have a substantially independent voice in its own affairs. In future, however, as prosperity increases in Rwanda more district councils are likely to acquire the financial independence that sustains local autonomy. Such potentially self-sufficient districts will need to be given an incentive to implement legitimate national goals in the area of social services such as health and education. When local governments are responsible elsewhere for such services that have effects that spill over outside the local community, they are usually supported financially by national governments. The central government can encourage the provision of services with such externalities by offering to share particular costs provided that the project or program meets certain conditions or minimum standards. According to the PRSC paper (para 112), the Bank plans to undertake analytic work in 2005 to help the GoR decide questions such as the nature and size of block grants to be transferred to districts and related performance indicators and the optimal balance between central guidance and local flexibility over the near and long-term. The Availability of Financial Management Skills. Background. There is no doubt that Rwanda needs more people with financial management skills. The 2002 Rwanda Country Assistance Strategy stresses that capacity building in the decentralized administrations should be a top priority, but the Decentralization and Community Development Project PCD properly emphasizes that lack of managerial skills is a more general problem in Rwanda: “There are severe capacity shortages at all levels in the public sector as well as in the private sector and civil society.”(P.11) And we have seen that there is a tragic reason for this shortage of skilled professionals, in that the latter were singled out for killing in the genocide and many fled the country. Rwanda’s CAS in 2002 noted further that “while the majority of the refugees have returned, the educated and qualified among them have largely remained in exile.” Although Rwanda may be a more extreme example than most, its lack of skilled personnel is a problem that would be common to many post-conflict countries. Where there is civil conflict, those with marketable skills such as chartered accountants and information technology specialists are likely to be more mobile and hence able to flee from conflict. 13 14 Equally, however, many of these skilled emigres will also return when order is restored and assured—if wage levels compensate for the costs involved. Has Decentralization Exacerbated the Demand for Financial Management Skills? Claims that decentralization in Rwanda has increased the demand for financial management skills, putting stress on the supply of skills to such institutions as the National Tender Board and the Office of the Auditor General, are common.19 But does it really follow that decentralization increases the demand for financial management skills? In the case of procurement, for instance, general training is needed with respect to procurement decisions, and the NTB has assisted in procurement training and installing staff with these skills at both central government and district levels. But the NTB itself oversees the tendering of all larger procurement contracts and there is no apparent reason why the number of such decisions should increase with decentralization. In fact, there is a general point to be made here. Whatever the type of decentralization that is taking place, it is not obvious that the process itself adds to the general demand for skills. In the case of devolution, where functions are transferred from the central to local governments, officials who formerly performed the functions can be offered the choice of moving to local governments. Some senior officials in Sierra Leone who were previously appointed to administer local administrations were offered this choice when local councils were elected in that country. In the case of deconcentration, as in Rwanda, officials who previously performed their functions in the capital could be offered the choice of moving to the provinces. In either case, decentralization need not be—indeed should not be-- accompanied by any net increase in personnel. Duplication of staff should be avoided. Capacity Building Training. There is no doubt that a substantial amount of capacity building is nevertheless needed in Rwanda in financial management and information and communications technology—both at the central and district government levels. These concerns are being addressed by the GoR through various strategies such as under the Multi-Sector Capacity Building Program and the Public Sector Reform Program. The World Bank’s Public Sector Capacity Building Project and Decentralization and Community Development Projects, which were scheduled to become operational at the end of 2004, will contribute to such capacity building. Much has been accomplished already through the offices of USAID by installing an excellent local accounting system in Rwanda’s 107 districts and training officials in its For instance, according to the PRSC (para 124), “lack of proficient procurement staff [in the NTB] is also being exacerbated by decentralization.” The 2004 HIPC AAP says that “the National Tender Board and the Office of the Auditor General, two key good governance institutions, are not adequately represented at the local government level” (p.6.) While the IMF ROSC of 2003 views procurement capacity as being adequate, it also says that “additional resources may be required to ensure the smooth functioning as the activities of local governments expand.” 19 14 15 use.20 In terms of what needs to be done next in improving financial management in the districts, emphasis might be placed on consolidating existing gains: as for instance, by making sure all districts report on time and addressing the problems of those that do not and are deprived of transfers for this reason.21 Creation of a special financial management/accounting cadre. There are other means besides training to increase financial management capacity through institutional change, however. In fact, there are many complaints of officials who receive financial management training subsequently leaving for the private sector where rewards are better. For instance, the assessment of the decentralization process commissioned by the GoR in 2003 found that Staff turnover in districts and provinces is a serious challenge for human resource development in local governments. This mainly affects critical areas of financial management, audit and planning.22 One solution recommended by this assessment of the decentralization process in Rwanda is that A well elaborated cadre organique in districts might decrease the staff turnover that affects critical areas of financial management, audit and planning.23 It is worth noting that Sierra Leone intends to adopt just such a strategy in financial management capacity building, through the creation of a special financial management/accounting class in its civil service. Enclaves and Civil Service Reform. Other institutional changes undertaken to attract skilled personnel in Rwanda have met with less approval. For instance, one means of attracting skilled personnel is to create special enclaves that can pay salaries higher than civil service rules permit. Thus the GoR established CEPEX in July 1999 as a semiautonomous body in MINECOFIN, to lead the process of evaluation and selection of investment projects for the development budget. Similarly, the Rwanda Revenue Authority—the country’s tax agency—was created as a semi-autonomous body. The creation of such semi-autonomous bodies to address the problem of attracting skilled personnel has not met with general approval. In fact the FARAP Report urged the GoR to reconsider the rationale behind the establishment of such autonomous or semiThe FARAP report (para. 174, p.44 ) concluded that “the accounting framework envisaged for local government is better and more comprehensive than that which current exists for the central government.” As noted in our text, however, efforts have since been made to raise standards in the latter accounting system. 21 Kaiser notes that “The Directorate of Territorial Administration in Minaloc suggested that LABSF transfers to districts were being withheld based on non-reporting.” See “Fiscal Equalization in Rwanda: Challenges for Decentralized Financial Management, Reporting, and Enhanced Equalization”, draft, July 12th, 2004, p.9 22 VNG International, Appraisal of the Decentralization Process in Rwanda: Final Report. Ministry for Local Government, Information and Social Affairs, September, 2003, sections 3.2 23 Ibid., Section 4.2. 20 15 16 autonomous entities, on the grounds that “setting up such bodies as a means of better paying their key personnel is no substitute for strategically targeted civil service reform.” (para 17) Obtaining Financial Management Skills: Buy or Make Decisions. There is therefore a consensus that civil service pay scales need to be adapted to attract and retain skilled financial management professionals in Rwanda’s civil service, whether at the central or district level. As has also been realized, however, the market may provide an alternative to obtaining specialist financial management skills from within the civil service. Thus, to carry out her duties, the Auditor General has been building up the capacity of her office, but is not reluctant to contract out audits to accounting firms of, for example, the autonomous public institutions. Nor does the AG confine herself to hiring private sector auditors from inside the country. The AG’s use of private sector auditors might even be employed more extensively in audit of the districts, which are fourth on her list of priorities (after the central government, development projects and autonomous public institutions.) And it is to be noted (and applauded) that a specific objective of the Bank’s Public Sector Capacity Building Project is to help ministries and agencies contract out some services to NGOs. Experience in decentralization demonstrates, indeed, that contracting out can be complementary to and may even be a superior alternative to some capacity development within the civil service. A number of the more successful local authorities in North America provide an excellent example of the merits of using market solutions to provide municipal services. Thus, Southern California is made up of many small inter-connecting urban communities. Many of these communities are traditional “full-service" cities that deliver services directly through their own in-house workforce. However, many other cities - 53 in the County of Los Angeles alone – follow what is known as the contract city model and leave service delivery to the competitive marketplace. Municipal services are contracted out to other communities, to regional governments, or to the private sector. Contract city staffs are small and principally design, monitor, and manage contracts to provide services to residents. But the costs savings may be substantial compared to the price of traditional services.24 The Scope of Local Governments: Does Each District in Rwanda Need to Do it All? This leads to a more general issue—as to whether all 107 districts in Rwanda need to duplicate personnel and facilities to carry out their responsibilities. It is certainly important to establish a clear division of responsibilities for functions between the central 24 The Frontier Center for Public Policy cites the example of Dana Pointe, a contract city with 35,000 residents in Southern California. Whereas a traditional, full service city would employ hundreds of fulltime staff, Dana Pointe has a staff of only twenty-five civic managers. But “the cost of service to its residents is about a third less than that paid by people in a so-called traditional "full service" city where government employees deliver most services. “When Smaller is Better - California Contract Cities,” Policy Series, September 22, 1997. 16 17 and district governments. But it is responsibilities that need to be assigned—not provision of the services. The example of the Californian contract cities cited above illustrates that competition in provision of goods and services is, in fact, desirable: monopolies do not produce good results. It helps to have the possibility of contracting out services to the private sector, NGOs or even other levels of government. Some time before the creation of 107 local authorities in 2001, Rwanda had deconcentrated health care by creating 40 Health Districts to manage health services. An analysis in 2003 of the PETS survey of the health and education sectors was far from complimentary to the administration of health care by the central government in these Health Districts and was far more complimentary to the contribution of localities. Fofack and his co-authors considered that the most startling finding of the PETS was the very limited funding by the central government of primary education and health services, with the result that their operations and maintenance costs were paid for largely by fees for services and, in the case of primary schools, other contributions by parents. Another finding of interest was the strong local involvement in managing education and health facilities. Indeed, Fofack et al. report that “the lack of accountability in the administrative offices is in contrast with efforts at the facility level to be accountable, presumably due to local involvement.” 25 Despite this verdict, there is no reason why the functions of the 40 deconcentrated health districts and the 107 local governments should be incompatible. The decentralization theorem that provides the economic rationale for decentralization simply says that governmental functions should be carried out at the lowest level of government at which they can be executed efficiently.26 There is no reason, however, why the optimal areas for delivery of local services should coincide exactly with the jurisdictions of the 107 local councils in Rwanda and plenty of reasons to think that they do not. The creation of the 40 health districts in Rwanda may, in fact, be regarded as an example of the possible uses of the special purpose districts that are a feature of successful decentralization throughout the world. These special purpose districts recognize the fact that optimal service areas for some functions may cross the borders of local and subnational political jurisdictions. North American examples are the school districts in the provinces and states of Canada and the United States. Changes in the areas covered by these boards have accommodated a massive rural/urban shift in population over the past century. The New York Port Authority is an example of an interstate special purpose district in the United States: it crosses the boundaries of three states. See “Public Expenditure Performance in Rwanda: Evidence from a Public Expenditure Tracking Study in the Health and Education Sectors,” Africa Region Working Paper Series No. 45, March 2003. 26 The same principle is embodied in the subsidiarity principle embodied in the Maastricht Treaty forming the European Union. 25 17 18 In fact, the Decentralization Laws of Rwanda make provision for the creation of special purpose districts,27 and indeed one such has already been created to protect and manage a two lake area that is spread over four districts. Further to this point, cooperative arrangements mean that efficient sizes of local governments can be quite small. The average population of French communes (1300 inhabitants), for instance, is much smaller than the average size of local government districts in other European states.28 The communes make extensive use of cooperative arrangements (syndicats intercommunaux) to provide services. There are obvious possibilities for communities to get together to provide services in Rwanda. This is a logical role that could be organized by RALGA, the Rwandese Association of Local Governments. RALGA already lobbies on behalf of local governments and is extensively involved in training of local officials--particularly in financial management. Coordinating the supply of local services would be a logical extension of these roles. Issues raised by Donor Funding Instability in Resources As noted above, development partners supply 55% of Rwanda’s budgetary funds. While the country is obviously grateful for this level of support, it has entailed two principal problems. The first concerns stability in the resources made available by development partners and their timing. Financing flows, particularly budget support from development partners, have been uneven, leading to late expenditure release by the GoR itself. And donors’complex aid procedures have contributed to the unpredictability of disbursements. This affects decentralization in Rwanda. It has long been known, for instance, that spending at the district level has been less than budgeted: this was the finding of the PETS for 1998/1999, for example. More recently, Kaiser observes that in 2003 the Local Authorities Budget Support Fund (LABSF) that provides funding for recurrent expenditures disbursed in only 3 out of 4 quarters. And while the Common Development Fund (CDF) approved an estimated two hundred projects for RF 3.96 billion in 2003, 27 Article 19, Law No. 34/2002 amending Law No. 04-2001 on Organization and Functioning of the Districts, states that: District Councils may, upon mutual concertation, decide to work together to initiate, organize or run the activities of their common interests. That partnership is referred to as “Districts partnership.” A District Council and Town Council or several District or Town Councils may also upon concertation undertake such partnership. Instructions that are made by the members to the Partnership Council are included in a document that mentions the activities to be performed and specifies the rights and duties of each party. 28 Melville McMillan notes that “the mean district population in England is about 125,000, in Germany it is 20,000...but closer to 9,000 in Norway and 30,000 in Sweden.” A Local Perspective on Fiscal Federalism: Practices, Experiences and Lessons from Developed Countries. Washington, DC: The World Bank, Public Economics Division, 1994. 18 19 actual disbursements were only a quarter of this, or RF 1.0 billion.29 Valderrama also notes that “the variances between actual and budgeted activity appear to be most pronounced for development as opposed to recurrent) budgets.” The variability of donor funding has much to do with this finding: as noted above donors supply 88% of the development budget as opposed to only 53% of the recurrent budget.30 One of the consequences of the greater variability in spending in the development budget may well have been the postponement of integration of the recurrent and development budgets. This was originally planned for 2003 but has been postponed until 2006. The cash budget approach adopted by the GoR to deal with this problem of unpredictability of timing in the release of donor funds--while indicating the GoR’s admirable dedication to fiscal discipline-- compromises the budget formulation process and priority setting. Some of the more promising advances in PEM would not be possible in Rwanda without additional donor cooperation to deal with this issue. As indicated by Obidegwu, for instance, it is clear that the success of an MTEF depends substantially on the willingness of the development partners to provide resource predictability.31 An obvious way of reducing such unpredictability in aid flows is for donors to increase the amount of aid they give in general budget support and reduce the amount of aid given for project or targeted spending. Transaction Costs to the GOR . The second major issue raised by donor funding involves the transaction costs to the Government of Rwanda, and these can be substantial. As well as assuring both domestic taxpayers and development partners that funds are being used for their intended purposes, well-functioning PEM systems should contribute to reducing reporting requirements by donors. The plethora of advice the GoR receives can be overwhelming. Of course it is wellmeaning. And perhaps this is the price a nation might expect to pay when it gets half or more of its resources from development partners, as is the case for Rwanda. But even taking these factors into account, the response burden for the GoR can be imagined when it is noted that Rwanda received 240 separate missions from development partners in the past year! There is little doubt that the demands that these missions make on Rwandan officials has the side-effect of a negative impact on the routine performance of the duties of the latter. And there are two references in the Bank’s recent PRSC paper to another consequence, “the potential for ‘reform fatigue’, especially in line ministries” (paras.111 and 115). “Fiscal Equalization in Rwanda: Challenges for Decentralized Financial Management, Reporting, and Enhanced Equalization”, draft, July 12th, 2004, para.10. 30 Ricardo Valderrama, “Budget Execution and Financial Position of Local Governments in Rwanda,” ARD Fiscal Decentralization Project, July 2004, p.4 31 See Obidegwu,. “The Medium-Term Expenditure Framework: The Challenge of Budget Integration.” Both the FARAP and PRSC have also noted that cash budgeting undercuts the MTEF. 29 19 20 The problem of complexity in donor aid procedures is not one that is confined to Rwanda. With respect to financial accountability in particular, Brooke reported in 2003 that A review of existing practice found a wide range of measures required by donors/agencies. Some 60-70 different types of measure were identified. Not only are they of a widely divergent nature, but they are also applied in different ways. 32 One method of reducing this response burden is for donors to agree on a common harmonized standard in assessment of fiduciary risks. Efforts to develop such a common standard have been based on the heavily indebted poor countries (HIPC) initiative of the World Bank and the IMF, which in turn built on the IMF’s ROSC methodology. Extending these efforts to get wide international acceptability of a common standard of fiduciary risks, a working group involving staff from the World Bank, IMF and the Public Expenditure and Financial Accountability (PEFA) Secretariat of the World Bank subsequently developed a set of sixteen indicators that can be applied to measure PFM performance in countries across a wide range of stages of development.33 Efforts to standardize fiduciary risk assessments of donors and reduce instability in donor aid flows in Rwanda came together in Nov 2003, when the GoR and a number of donors providing direct budget support signed the Partnership Framework for Harmonization and Alignment of Budget Support. The donors involved are the UK DFID, the EU, the African Development Bank (AfDB), Sweden (SIDA), and the World Bank. This approach signals a major change from the way donor assistance was previously provided to the GoR. The Framework states that donors agree to harmonize their practices with respect to requirements for PFM and government fiduciary arrangements. The principal benefits to the GoR are that it will allow for improved predictability of aid flows and will also improve donor coordination and reduce transactions costs. In the latter respect, the Framework also identifies a lead sector or agency in the GoR to manage preparation of a strategy for the sector with support from a lead donor for the sector. To give examples of the latter that are particularly relevant to the present work, the lead donor for decentralization is the Netherlands, for governance the UNDP and for human resource development and capacity building the Swedish International Development agency (SIDA). The cluster groups, including government counterparts, meet regularly to coordinate the approach in the sector.34 Peter Brooke, “The Platform Approach to Public Financial Management Reform and Managing Fiduciary Risk”, Bannock Consulting for the World bank’s PEFA Secretariat, 2003, p.1. 33 See the PFM Performance Measurement Framework, revised consultative draft, PEFA Secretariat, February 12, 2004 34 For instance, a Working Group on Rwandan Fiscal Decentralization met to consider decentralization issues in a Video Conference (Kigali/Washington, DC), on July 13th 2004. Participants included GoR representatives (from MINECOFIN, MINALOC, MININFRA, CDF, RRA, DCDP), and from the development partners including the World Bank, USAID, the Government of the Netherlands and UK DFID. 32 20 21 Other Methods of Reducing Transaction Costs Availability of Documents. The writer would like too suggest another initiative to reduce transaction costs for both donors and governments receiving aid, including the GoR. Some of the analytic work done by development partners is cloaked in secrecy requirements. Access to such documents has been a persistent issue at the working level between the Bank and the IMF. 35 Governments are urged to make some of these documents public, such as the HIPC AAPs. But leaving the decision as to whether to declassify a document to government bureaucrats is likely to get a decision that errs on the side of extreme caution. What is the upside to a bureaucrat of such public release? Effects of these confidentiality restrictions may be to conceal good analytic work that might make a contribution if more widely known (and perhaps to let some questionable work go unchallenged), to duplicate analytic efforts and to increase the response burden of aid recipients, as in the documented cases where questions are repeated by “visiting firemen” who may not have seen the proscribed work. Another consequence is unevenness in application of confidentiality requirements among countries. A case of relevance in the present work relates to the HIPC AAPs. While governments are urged to make these assessments public, the Government of Sierra Leone has done so but the Government of Rwanda has not. Much of this secrecy may be simply unnecessary. Nothing this writer has seen in documents relating to Rwanda that are classified as “confidential” or even “strictly confidential” raised his eyebrows at all. Nor, one feels, are these documents likely to raise eyebrows in the post-conflict nations themselves. Given the exigencies of their recent histories, these countries are not likely to be shrinking violets. And in any event, nations that accept over half of their resources from development partners are likely to be sufficiently thick-skinned to receive some commentary on their performance with equanimity. Above all, these practices are objectionable because they offend the spirit behind the drive to get greater transparency and accountability that lies at the heart of much public financial management. The present paper suggests that the Bank not agree to label any of 35 In late 2002, the writer worked on the paper on Bank-Fund Collaboration on Public Expenditure Issues that went to the Boards of both the Bank and the IMF. The paper involved a survey of those working on public expenditure issues in the Ministries of Finance of developing countries, the staff of development agencies and staff of the Bank and Fund themselves. A third of respondents thought problems of collaboration reduce the effectiveness of the PEM advice of the Bank and the Fund. Bank staff complained about the "unwillingness of IMF to share relevant reports and other documentation." One comment from World Bank staff in the Africa region was that: We in the Bank's AFR region have excellent working relationships with the Fund's area department but people in FAD (especially those working on fiscal transparency issues) seem to operate in an excessively secretive environment. They refuse to share documents with us (and even with their own area departments) 21 22 the PFM documents on which it works as confidential and to impress on client governments that it expects documents such as the HIPC AAPs to be released to the public. The Need for a Plan. Although Rwanda started its reconstruction earlier than Sierra Leone, it might learn something from the latter country with respect to the sequencing of PFM in post-conflict situations. In both countries there is more than enough to do in the field of PFM and absorption capacity is limited. Hence there is a need to assign priorities. As noted in a previous paper, the Government of Sierra Leone does have an Action Plan that assigns such priorities. There has been quite a lot of effort by the development partners to establish a PFM action plan in Rwanda, with priorities being assigned. This started with the FARAP initiative and the EU follow-up to FARAP and, although in principle all donors are involved, is supported principally by the EU, DFID, the IMF (AFRITAC-East), as well as a number of European development partners who give bilateral aid. The FARAP report says that its “action plan will provide the appropriate sequence of activities and proposals for institutional capacity building on a sustainable basis.” The principal--and critical-- difference from the approach being followed in Sierra Leone is that the GoR does not appear to have assigned priorities, although it has a mechanism to do so: a national Steering Committee is supposed to coordinate and monitor reforms in public financial management. The absence of prioritization in Rwanda may be a victim of the abundance of outside advice and actors, with the outcome that the emphasis is on doing everything at once. Development and issue of a prioritized PFM plan is an essential step that the GoR should take. Adoption of Advanced PEM Techniques This writer is skeptical about the adoption of some advanced PEM techniques whose application is being urged in Rwanda and thinks use of such techniques should be deemphasized. Rwanda may yet be ahead of some developed countries—even perhaps of some of its development partners—in adoption of some PEM techniques. While it has been said that a medium or longer-term perspective on budgeting is at the heart of successful budget reform in much of the world over the past two decades, less than two out of three OECD countries were recently found to employ medium term expenditure frameworks.36 It is therefore impressive that the HIPC AAP concluded that Rwanda met the benchmark of Sixty three per cent of OECD countries answered affirmatively to the question, “Is there a consistent medium-term fiscal framework stating targets or ceilings for expenditures, deficits and revenues for the medium-term…” OECD/World Bank Budget Practices and Procedures Database, 2003, available on the World Bank’s website. 36 22 23 multiyear planning, in that multiyear forecasts are formulated at a program level and are then used as the starting point for annual budgets.37 Another advanced budgeting technique whose use has been suggested in Rwanda is that of results-oriented or performance budgeting. The recent PRSC Paper, for instance, wants results-oriented budgeting developed in education, health, water and energy (para 5). There is no doubt that the goal of budgeting should eventually shift towards getting results from the spending of public resources rather than just allocating inputs. While Rwanda is a long way now from results-oriented budgeting, this goal should always be kept in mind as being that to which this country and all others should be striving. Nor is it too early to start toward this goal in Rwanda, for instance, by beginning to incorporate some performance indicators as part of the budget. But the writer is entirely in accord with the judgment of the IMF’s ROSC in 2003 that adoption of results-oriented budgeting should be delayed in Rwanda.38 The record shows that implementation of results oriented budgeting is challenging enough in OECD countries, where implementation started in the 1960’s. After continuous work for 40 years, most OECD countries still face difficulties in establishing links between budgets and results achieved. Conclusion: the Sequencing of Financial Management Measures in Rwanda The Government of Rwanda, with the aid of its development partners, has made impressive progress since its civil conflict in installing two hallmarks of democratic societies: responsible local government and accountability for financial management. It helps that this is a nation that has developed a formidable sense of discipline.39 And the considerable efforts that it is making to help itself should be taken into account. An 18% VAT rate speaks volumes in the latter respect. Despite its formidable accomplishments in the post-conflict period, what the Government of Rwanda needs in an area such as financial management where much needs to be done, is a prioritized plan. In this respect, it could learn from the example of the other postconflict country visited by the mission: Sierra Leone. Perfectionism also needs to be avoided. The GoR should proceed with caution in considering introduction of some advanced techniques of financial management. This 37 A possible reason for the greater success of medium term expenditure frameworks in Rwanda is because this is the vehicle whereby the PRSP is being put into effect. 38 “Only at a much later stage should one move towards adopting accrual accounting or output-oriented budgeting...” International Monetary Fund, Report on the Observance of Standards and Codes, Fiscal Transparency Module, Rwanda. IMF Fiscal Affairs Department, 2003. 39 A noneconomic example occurred during the mission for this paper, when the government decided to ban the use of plastic bags on environmental grounds. Compliance was instantaneous. In the Maghreb countries where the writer also works, one can tell when one is approaching a major town because the plastic bags on the fence lines by the roads grow thicker. 23 24 writer was forced to wonder how many of these advanced financial management techniques apply in the countries of the donors who advocate them. Performance or results-oriented budgeting, for instance, is beyond Rwanda at present—as indeed it is beyond most OECD countries. The present paper agrees with in-depth assessments such as that of the IMF ROSC and the HIPC AAP that a high quality government accounting system is a priority in Rwanda. Agreement with the IMF ROSC extends to the arguments that other basics of good PFM are good internal and external audit. Rwanda is well on the way to achieving these goals. If anything needs to be added to these recommendations, more emphasis might be placed in Rwanda on using Public Expenditure Tracking Surveys to supplement internal audit. More effective use has been made of this PEM tool in other African economies, including the other country visited by the mission, Sierra Leone. In terms of what needs to be done next in improving financial management in the 107 district governments, emphasis might be placed on the straightforward goal of consolidating existing gains: as for instance, by making sure all districts issue financial reports on time and addressing the problems of those that do not and are deprived of transfers for this reason. In terms of intergovernmental transfers, this paper has noted that the proposed devolution of functions and accompanying transfer of resources to local governments is substantial. But here national interests should be served at the same time as local autonomy is maintained. Attention needs to be devoted to transfer arrangements that permit national goals to be attained in areas such as health, education and welfare while avoiding micromanagement of local governments. In this respect, the central government might give some thought to the constraints it operates under itself as a result of the substantial proportion of its budget that is provided by donors. The fact that 88% of the development (or capital) budget is provided by donors, for instance, is perhaps the biggest obstacle to the necessary step of integrating the development and recurrent budgets. But coordination with donors has led to some sensible loosening of constraints on the GoR. The Partnership Framework for Harmonization and Alignment of Budget Support has been a major step forward, both in making more aid available as general budget support and in agreement on a common standard of fiduciary requirements for donors. Although the major donors are in the Partnership, more of the bilateral development partners could also get on board. This type of agreement is something that should be pursued with vigor in other post-conflict and low-capacity countries. And even more could be done by donors to lower the response burden on countries such as Rwanda. The number of missions this country receives is staggering and anything that would reduce potential duplication of effort by development partners would help. A reassessment of the confidentiality classification attached to documents by particular development partners would assist in this respect and is long overdue. Such secrecy does 24 25 not sit well with the drive to get greater transparency and accountability that lies at the heart of much public financial management. In terms of using financial management, Rwanda obviously needs to augment capacity by training, such means as creating a financial management cadre in the local and central bureaucracies and pushing ahead with civil service reform that rewards financial management skills appropriately. By the latter means Rwanda may be able to attract back more of the skilled workers who fled in the civil conflict. And in particular, contracting out should be pursued more wholeheartedly. The ideal would be to have a marketplace where the private sector, NGOs and even other levels of government compete in providing such skills as an alternative to the local government districts doing everything for themselves. Indeed, use might be made of the market in extending financial management reforms in more institutional ways. The 107 districts may be the appropriate political jurisdictions for responsible local government but they do not need to duplicate provision of every service and their jurisdictions may differ from optimal areas for providing some services. Special purpose districts, cooperative arrangements and contracting out are all ways of getting economical provision of services while maintaining responsible local government. 25 26 Bibliography Brooke, Peter. “The Platform Approach to Public Financial Management Reform and Managing Fiduciary Risk”, Bannock Consulting, 2003. Fofack, Hippolyte, Chukwuma Obidegwu and Robert Ngong. “Public Expenditure Performance in Rwanda: Evidence from a Public Expenditure Tracking Study in the Health and Education Sectors.” Africa Region Working Paper Series No. 45, March 2003 Frontier Centre for Public Policy Series. “When Smaller is Better-California Contract Cities.” Winnipeg, Manitoba, September, 1997. Gillingham Robert, Duncan Last, Ronald Neumann and Russell Robinson. “Rwanda: Managing Fiscal Decentralization.” IMF FAD, August 2003. Government of Rwanda. Poverty Reduction Strategy Paper. Ministry of Finance and Economic Planning, June 2002. Government of Rwanda and European Commission. Development of a Public Financial Management Capacity Building and Technical Input Plan. Rapport Provisoire, by Jean François Bauer and Bernard Biche, IDC, 22 Mars 2004. International Monetary Fund. Report on the Observance of Standards and Codes, Fiscal Transparency Module, Rwanda. IMF Fiscal Affairs Department, 2003. Kaiser, Kai. “Fiscal Equalization in Rwanda: Challenges for Decentralized Financial Management, Reporting, and Enhanced Equalization.” Draft, July 12th, 2004. McMillan, Melville, L. “A Local Perspective on Fiscal Federalism: Practices, Experiences and Lessons from Developed Countries.” Washington, DC: The World Bank, Public Economics Division, 1994. Obidegwu, Chukwuma. “The Medium-Term Expenditure Framework: The Challenge of Budget Integration.” World Bank, Washington DC, March 2004. Valderrama, Ricardo. “Budget Execution and Financial Position of Local Governments in Rwanda.” ARD Rwanda Fiscal Decentralization Project, July 2004. VNG International. Appraisal of the Decentralization Process in Rwanda: Final Report. Ministry for Local Government, Information and Social Affairs, September, 2003 World Bank. Country Assistance Strategy for the Republic of Rwanda. November 21, 2002, Report No. 24501-RW __________ and IMF staff. Rwanda: Tracking Poverty-Reducing Spending: Second Assessment and Action Plan, July 2004. 26 27 Annex A TERMS OF REFERENCE Lessons on Sequencing Financial Management and Accountability Systems in Decentralized Post-Conflict/Low Capacity Contexts Background Effective financial management and accountability arrangements (FMAAs) are a key ingredient to improve public service delivery at various levels of government. FMAA measures include procedures and systems for budgeting (planning, formulation and execution) and forecasting (including annual budgets, MTEFs, and revenue forecasting), accounting standards, treasury management, audit and accountability functions, and financial reporting (to both local and central users). Post-conflict/low capacity environments pose particular challenges given the need to enhance access and quality of public services to avert renewed conflict, and the generally weak institutional capacity. The PREM Public Sector Group, with regional colleagues, is keen to provide a better understanding of what financial management reforms are “needed” (intergovernmental, central and local) and how to prioritize and sequence reforms by developing the capacity to choose among them in post-conflict/low-capacity contexts. To inform this broader agenda, work being conducted with the support of a BankNetherlands Partnership Program (BNPP) Global and Regional Initiatives grant will: (i) provide a framework/stocktaking of local FMAA arrangements for public service delivery, and (ii) analyze the cases of Rwanda and Sierra Leone, in terms of lessons from past sequencing to current achievements, and potential lessons for prioritization and future sequencing. The overall activity will examine three components: (i) prioritization and sequencing of decentralized FMAA, (ii) decentralized Information and Communications Technology (ICT) for FMAA, and (iii) local revenue mobilization options. Each of these components will be synthesized with regard to the overall decentralization framework, including prioritizing and sequencing expenditure assignments, intergovernmental transfers, and other critical elements. Level of Effort and Timing For this assignment, 30-35 days of effort are expected over the period 15 September – 15 November 2004. Effort would include travel and field work in Rwanda and Sierra Leone. Requirements Extensive background experience in FMAA and decentralization issues Strong analytical skills Regional experience in low-capacity/post-conflict context, Rwanda/Sierra Leone (preferred) 27 28 Good verbal and written communication skills in English (required) and verbal skills in French (preferred) Specific Tasks Working closely with World Bank PRMPS staff (Dana Weist, Kai Kaiser and others); Rwanda country team staff (Kene Ezemanari and Joseph Kizito); and Sierra Leone country team staff (Yongmei Zhou) the consultant will: Write a 10-15 page report of the sequencing and achievements of FMAA and decentralization/deconcentration in various countries based on a desk review of project documents, donor assessments, staff interviews (documents to be provided/specified), and the case studies listed below. Travel to Rwanda and Sierra Leone to participate in two, 1-week missions (plus travel time to/from base station), ideally in late September/early October 2004. These trips will be scheduled in conjunction with country-team missions (e.g., to assess decentralized financial management issues and priorities in Rwanda, as part of the PRSC and overall country dialogue). Prepare two, 10-15 page case studies of FMAA in Rwanda and Sierra Leone to include (i) current achievements in country-level FMAA, (ii) gaps and priorities, (iii) generic lessons about effective FMAA interventions and sequencing/prioritization priorities and sequencing The consultant may also be asked to present their findings to World Bank staff members working in Washington, DC. Reporting The consultant will report to Dana Weist (Lead Public Sector Specialist, PRMPS). Outputs The outputs of the activity include: [i] Stocktaking and Synthesis of FMAA Lessons for Low Capacity/Post-Conflict Contexts (10-15 pages) [ii] Rwanda Case Study (10-15 pages) [iii] Sierra Leone Case Study (10-15 pages) [iv] BBL presentation in Washington DC 28 29 Annex B. Table 1. Extent to Which HIPC Benchmarks were Met in 2004 Rwanda I. Budget Preparation 1: Coverage of the budget or fiscal reporting entity. 2: Degree of spending funded by inadequately reported extrabudgetary sources. 3: Reliability of budget as guide to the future. 4: Inclusion of donor funds in budget and fiscal reports 5: Budget expenditures adequately classified 6: Poverty reducing expenditures identified as priority programs in the budget 7: Medium-term forecasts integrated into budget formulation. II. Budget Execution 8: Arrears evidence of problems 9: Internal control system is effective 10: Internal control supplemented by public expenditure tracking surveys. 11: Quality of fiscal information assured by regular reconciliation of bank accounts with accounting records. III. Budget Reporting 12: Internal fiscal reporting regular and timely. 13: Regular fiscal reports track poverty-reducing spending 14: Transactions recorded in a timely fashion 15: Audited financial information is timely IV Public Procurement System 16. Public procurement system is efficient and effective 29