Fiscal decentralization in the Democratic Republic of the Congo has been viewed as a way to hold a vast and diverse country together following a long civil war. The country—divided and governed by different factions during the 1998–2003 civil war—was officially reunited with the promulgation of the Transitional Constitution in April 2003. The creation of the Government of National Unity in July 2003 and the inauguration of the National Assembly and the Senate in August 2003 were major accomplishments of the peace process and symbolized the end of the six-year-long war. A new constitution was approved by a national referendum in 2005 and entered into force in February 2006. For the first time in half a century, parliamentary and presidential elections were held in 2006. Despite these positive steps, holding together this vast country—populated by 62 million people and 250 ethnic groups, with four times the area of France and as many as 700 local languages and dialects—has proved to be a major challenge.
The constitutionally mandated provincial institutions were created in early 2007. Direct election of the members of the provincial parliaments (Assemblées Provinciales) took place in late 2006, and the provincial parliaments elected governors and vice-governors in January 2007.2 The provincial governments, consisting of up to 10 ministers in addition to the governors and vice-governors, began operating in February 2007. However, their functioning has been impaired by the following factors: (1) delays in drafting and approving a legislative framework for the relationship with the central government, including with respect to fiscal decentralization; (2) insufficient resources for the operations of the new institutions; and (3) weak administrative capacity, partially reflecting insufficient staffing and an inappropriate skill mix in the new provincial administrations. Against this background, it remains unclear whether the constitutionally mandated creation by May 2010 of 26 provinces (25 provinces plus Kinshasa), instead of the current 11 provinces (10 provinces plus Kinshasa), will be possible.
Democratic Republic of the Congo: Indicators of Fiscal Decentralization
Democratic Republic of the Congo: Indicators of Fiscal Decentralization
| Type of government | Unitary | |
| Population (million, 2007) | 57.5 | |
| Area (million km2) | 2.3 | |
| Levels of government | 3 | |
| Provinces | 11 | |
| Municipalities | 1,005 | |
| Average municipality size (population) | 57,262 | |
| Minimum municipality size (population) | … | |
| IMF technical assistance missions on | 2005 | |
| fiscal decentralization | ||
Democratic Republic of the Congo: Indicators of Fiscal Decentralization
| Type of government | Unitary | |
| Population (million, 2007) | 57.5 | |
| Area (million km2) | 2.3 | |
| Levels of government | 3 | |
| Provinces | 11 | |
| Municipalities | 1,005 | |
| Average municipality size (population) | 57,262 | |
| Minimum municipality size (population) | … | |
| IMF technical assistance missions on | 2005 | |
| fiscal decentralization | ||
IMF staff have provided advice on the decentralization process over the last few years. A staff mission in 2004 found that central and local governments often spent their resources outside their competencies, own-revenue collection was weak, and the draft decentralization law posed large macroeconomic risks. The mission recommended a gradual and carefully sequenced decentralization strategy. Although no dedicated technical assistance on decentralization has been provided since that mission, these issues have remained on the radar screen. The World Bank has been leading the dialogue with the authorities in the area of decentralization, and has closely cooperated with IMF staff, especially regarding macrofiscal aspects of the decentralization process. Since April 2008, the World Bank has been supporting decentralization and capacity building in the provinces through the Enhancing Governance Capacity Project.
Fiscal Arrangements Prior to Enactment of Decentralization Legislation
Until recently, the Democratic Republic of the Congo had a deconcentrated system. According to the organic public finance law existing at the time, four categories of local governments, called Entités Administratives Décentralisées (EADs), maintained separate legal personalities, had their own budgets, and had the right to raise local taxes. Heads of EADs were officially nominated by the president. They were representatives both of the state (and as such, responsible for implementing central government policies) and of the local executive branch. Decisions of the local authorities were subject to central control. Their budgets had to be approved by the minister of the interior (for the provinces) or by the governor (for other EADs).
The transfer system was based on largely arbitrary revenue sharing. Revenue transfers from the central government to provinces were supposed to be based on the amount of central government taxes and duties collected in the provinces by the three revenue collection agencies, and to be paid into the provinces’ accounts at the central bank.3 Since 1998, transfers had been officially set at 20 percent of revenues collected in the provinces; three-quarters of that 20 percent was supposed to be transferred automatically by the central bank to the provinces. However, actual revenue transfers were 6.4 percent and 6.5 percent of total domestic revenue in 2006 and 2007, respectively. Moreover, important differences surfaced in the distribution of transfers between provinces. The three richest provinces (Kinshasa, Katanga, and Bas-Congo), which together account for 76 percent of total domestic revenues, benefited from the highest absolute transfers, but were also affected by the largest shortfall with respect to the 15/20 percent rule.4
Deficiencies of the Constitutionally Mandated Fiscal Decentralization
The new constitution leaves fiscal arrangements largely undefined. Although provinces enjoy far-reaching fiscal autonomy, the directly elected provincial governors perform a dual role by simultaneously representing their provinces and exercising central government tasks. The constitution establishes three government levels: central, provincial, and subprovincial (Entités Territoriales Décentralisées, ETDs).5
The constitutionally mandated revenue-sharing arrangements are vaguely defined. The constitution stipulates that 40 percent of “national revenues” collected in each province will be devolved to them (Art. 175), while another 10 percent will be assigned to an investment fund with equalization purposes (Caisse Nationale de Péréquation) in charge of reducing the development differential between provinces (Art. 181). However, the constitution neither assigns responsibilities for revenue collection to the different government levels nor clarifies the methodology, including the definition of the tax base, for calculating the 40 percent of revenues to be transferred to the provinces and the 10 percent for the Caisse Nationale de Péréquation. The constitution also lacks clarity about expenditure assignments across government levels, thereby giving rise to the risk of a duplication of responsibilities and unfunded mandates in sectors with competing legislative competence.
The approach to fiscal decentralization enshrined in the constitution results in a number of significant fiscal risks:
The new revenue-transfer system would lead to a sizable resource transfer to largely unprepared provincial and local authorities, possibly resulting in poor-quality spending and large-scale misappropriation of funds. It would also create a major vertical imbalance in favor of the provinces because the public services that are supposed to be transferred to the provinces (health, primary and secondary education, and agricultural services) account for less than 20 percent of total expenditures, leaving the central government with sizable unfunded spending responsibilities.
The new transfer system would also create serious inequity among provinces. Three provinces would get the lion’s share of revenue transfers (about 77 percent to Kinshasa, about 9 percent to Katanga, and about 5 percent to Bas-Congo). The remaining eight provinces together would receive about 10 percent (compared with their share in total population of 70 percent), with four provinces (Bandundu, Equateur, Maniema, and Province Orientale) each receiving less than 1 percent, which would prevent them from providing even the most basic public services. The Caisse Nationale de Péréquation would not help mitigate these horizontal imbalances because it is limited to investment financing in its foreseen format.
Mechanisms for fiscal policy coordination and control remain undefined, which is a critical concern in a system with a highly fragmented economic policy management framework. Key responsibilities are divided between the Ministries of Finance, Budget, and Planning, with none of them having sufficient oversight over subnational governments. Moreover, the absence of clearly defined taxing powers for subnational governments also weakens an important accountability mechanism in the conduct of fiscal policy at the subnational level.
Against this background, World Bank and IMF staff have emphasized the merits of a gradual approach to fiscal decentralization. Revenue transfers to the provinces should be increased only gradually and in line with the devolution of expenditure responsibilities. Expenditure responsibilities, in turn, should go hand in hand with progress in building provincial capacity.
Most important, the pace of revenue and expenditure devolution should depend on progress in public financial management (PFM) reforms. World Bank and IMF staff called for transitional arrangements that would make the establishment of certain institutions and the transfer of spending responsibilities contingent on administrative capacity. Such arrangements could help avoid a situation in which newly created institutions at the level of provinces and ETDs would be required to fulfill their legally established duties without a clearly defined resource envelope and while lacking the capacity to perform the functions transferred to them. IMF staff suggested that each line ministry develop medium-term sectoral transfer plans, to be accompanied by performance indicators that would trigger moving to the next stage.
The National Forum on Decentralization
Cognizant of the deficiencies of the constitutionally mandated approach to fiscal decentralization, the authorities, with broad donor support, organized in October 2007 a National Forum on Decentralization to develop recommendations with respect to several aspects of the decentralization process, including fiscal and human resources management. The forum concluded that the base for calculating the revenue transfer to provinces should be narrowly defined by excluding oil revenues, which accounted for 20 percent of total revenues (excluding grants) in 2007. This narrow definition would help contain the risk of vertical imbalances by reducing the gap between the supplementary revenue transferred to provinces and expenditure devolution.
The formula would also go a long way toward mitigating horizontal imbalances. The forum advocated dividing domestic non-oil revenues into two categories:
Category A revenues: 40 percent of all nontax revenue and revenue collected by the tax collection agency (DGI) minus revenues collected by the large-taxpayer unit would be automatically transferred to the provinces in which these revenues were collected.
Category B revenues: 40 percent of all customs duties, excise taxes, and revenues collected by the large-taxpayer unit would be distributed to the provinces based on population.
The distribution mechanism for category B revenues denotes a significant horizontal redistribution, with the share of the three richest provinces in total revenue transfers declining to 41 percent from the 92 percent it would have been under the constitutional scenario. Still, this mechanism may not be capable under all circumstances of providing sufficient resources to the poorest provinces to cover wages and goods and services for the responsibilities transferred to them (primary and secondary education, health, and agricultural services). Simulations prepared by World Bank staff suggest that the coverage of current spending in the poorest provinces crucially hinges on wage policy assumptions and the provinces’ capacity to generate own-revenues.
The 2008 Decentralization Legislation
Three laws promulgated in October 2008 contribute little to clarifying the fiscal relationship between the central and the provincial governments.6 These laws specify the legislative powers and the organization and functioning of public administration at the level of the provinces and the ETDs. However, they leave the methodology for implementing the constitutionally mandated revenue-sharing mechanism undefined; and instead of providing for a more-precise delineation of expenditure assignments in areas with competing competencies, these laws repeat the definitions in the constitution. Fiscal aspects of the decentralization process are now supposed to be regulated in a new organic public finance law (discussed in the last section of this chapter).
The decentralization legislation does, however, deviate from the consensus reached at the National Forum on Decentralization with respect to a key element. Art. 55.3 of the Loi sur la libre administration des provinces stipulates that all revenues, including oil revenues, are part of the base for calculating the revenue transfer to the provinces, thereby compounding vertical fiscal imbalances. It is unclear whether there is scope for the organic public finance law to reverse this stipulation, at least on a transitory basis.7
Current Transitional Arrangements
In the absence of a new organic public finance law, the 2009 budget reffects an ad hoc approach to fiscal decentralization. In compliance with Art. 55.3 of the Loi sur la libre administration des provinces, the 2009 budget provides for the transfer of 40 percent of all domestic revenue—including oil revenue—to the provinces. The transfers to the provinces (rétrocession) have been divided into three components:
Wage bill of public services transferred to the provinces. Payments for this component (36 percent of the total rétrocession in the 2009 budget) continue to be executed, as under the 2008 budget, by the central government on behalf of the provinces. This reflects a consensus that provincial capacity is still insufficient for carrying out timely wage payments on a large scale.
Transfer for provincial spending on goods and services. These are the only actual non-earmarked transfers to provinces in 2009, amounting to 28 percent of the total rétrocession to provinces.
Public investment in the provinces. Resources corresponding to this component (36 percent of the total rétrocession) are not being transferred to the provinces. The government stated that it envisages managing these resources jointly with the provinces, but discussions on the definition of comanagement are ongoing. Although some share of the contracting and implementation is supposed to be carried out at the provincial level, contracts that would have economies of scale would be handled by the central government. This approach raises a number of issues.
Although the continued management of the wage bill of transferred public services reflects a consensus with the provincial governments, the withholding of the investment component constitutes a new earmarked portion of the rétrocession that is not provided for in the legislative framework. In fact, this approach has been criticized for being contradictory to the constitution and the Loi sur la libre administration des provinces, which provide for a clear separation of the public finances of the provinces from those of the central government. Earmarking 36 percent of the total rétrocession for investment is also problematic because it prevents the provinces from freely choosing the ratio of investment to goods and services that they consider most beneficial. Moreover, this approach raises the question of whether the central government might avail itself of a part of this resource envelope to finance its own expenditures—a risk that may increase with weakening revenue performance in the context of a deteriorating global environment.
Ensuring the timeliness and predictability of transfers will be important. Although an accord had been reached between the provinces and the central government in July 2008 on the modalities for transferring the rétrocession, the center has not adhered to it and actual monthly transfers fell short of the agreed amounts.8 To instill confidence in the decentralization framework and allow for predictability of the resource envelope available to the provinces, adhering to the rules will be crucial.
Work on necessary interim solutions for provincial public financial management and personnel management is still ongoing. To bridge the gap until the approval of the new organic public finance law and the law on civil service in the provinces, five decrees were drafted in August 2008. These decrees provide for the establishment of minimum requirements with respect to an orderly budget execution, including through the introduction of an expenditure chain in provincial governments. Moreover, they aim at clarifying the reassignment of civil servants working for transferred public services to provincial governments.9 However, these decrees had not yet been implemented as of mid-2009. The reassignment of civil servants to the provinces is also complicated by the fact that the census of teachers and health sector employees has not been finalized. On the positive side, the responsible line ministries and the Ministries of Budget and Decentralization recently agreed on a road map for decentralization in health, education, and agriculture, which takes into account stages of capacity building.
The Key Role of the New Organic Public Finance Law
The new organic public finance law will be crucial for containing vertical and horizontal fiscal imbalances. Following several months of preparation involving technical assistance from World Bank—financed consultants, the draft organic public finance law was scheduled to be discussed by the government in April 2009, together with the accompanying draft law on the public finances of the provinces and the ETDs. It is essential that the organic public finance law establish clear regulations for the revenue-sharing mechanism, preferably excluding oil revenues, and the allocation of expenditure responsibilities.
The new law will need to ensure appropriate budgetary control. The central government must firmly manage provincial spending by enforcing budget monitoring and reporting mechanisms, including expenditure procedures to be used at the subnational levels, particularly with respect to internal and external financial control. Successful execution of this task will require giving the Ministry of Finance an explicit role in regulating, coordinating, and supervising local government finances. Mechanisms for sanctioning provinces that do not comply must be included, preferably in the form of withholding future rétrocession.
The new law should also regulate provincial borrowing. IMF staff have suggested direct controls on borrowing, including annual ceilings on debt flows and the overall debt stock of provinces as well as central government review and authorization of individual borrowing operations. Limiting provincial borrowing to the level of public investment will not be sufficient to prevent a rapid and unsustainable accumulation of debt. The law should also specify that provinces and local authorities are not allowed to borrow from the central bank, and should prohibit the use of shared revenue to collateralize subnational borrowing.
In a similar vein, tax assignments and responsibilities for revenue collection will require further clarification. The central government should remain in control of major taxes and customs. At the same time, the administration of all national domestic taxes (both direct and indirect taxes), as well as customs duties, should remain the responsibility of the central tax administration. In the medium term, building local-government ability to collect own-taxes and to reinforce budget preparation and execution will be critical, as will be strengthening treasury management and the accounting framework.
This case study is based on Seade, Brosio, and others (2005).
Art. 198 of the constitution stipulates that the elected governors and vice-governors have to be installed by the country’s president.
Revenue collection functions are split between OFIDA (customs revenue and excise taxes), DGI (tax revenue), and DGRAD (nontax revenues).
The high concentration of large companies in the capital Kinshasa, which also hosts the large-taxpayer unit (the agency responsible for collecting taxes from large taxpayers), results in a high share of total income tax collection. Moreover, Kinshasa and Bas-Congo (with the main port Matadi) are the main entry points of the country. Katanga has the highest concentration of mining companies.
The ETDs consist of “villes [cities], communes urbaines, communes rurales [rural and urban communities], secteurs, and chefferies [the latter are rural units].”
The decentralization legislation consists of the following three laws: (1) Loi sur la libre administration des provinces [Law on the Autonomous Administration of Provinces]; (2) Loi sur les Entités Territoriales Décentralisées [Law on Decentralized Territorial Entities]; and (3) Loi sur la Conférence des Gouverneurs de province [Law on the Conference of Provincial Governors].
A transitory clause on the exclusion of oil revenues could be justified on the basis that it will take a few years to establish the necessary administrative capacity in the provinces to manage the new resources efficiently.
The 2008 budget law reflected the consensus reached during the National Forum on Decentralization with respect to the determination of the revenue base and the revenue distribution formula. However, in the actual budget execution, particularly in the first half of the year, rétrocession payments fell well short of the amounts that would have resulted from a strict application of the distribution formula.
This covers a portion of employees of other deconcentrated services, in addition to civil servants working for transferred public services (primary and secondary education, health sector, and agricultural services).