Rapport No. 41776-ZR Democratic Republic of Congo Decentralization in the Democratic Republic of Congo Opportunities and Risks Joint Note by the World Bank and the Delegation of the European Commission in the DRC January 2008 Public Sector Reform and Capacity Building (AFTPR) Africa Region Document of the World Bank and the European Commission The views expressed in this document do not constitute the official position of the European Commission or the World Bank TABLE OF CONTENTS Acknowledgements 1 2 2. Competency Transfers and FiscalIntergovernmentalRelations 1.Introduction and Context......................................................................................................... ..................................................................................................................... ............................................... Competencies: gradual or `big bang'? ..................................................................................... 5 5 Fiscal aspects: balancing entitlements, equity and viability., 9 Constitutional interpretations. equity and viability ............................................................ .................................................... 10 The `CommissionParitaire' scenario :not enoughsolidarity ............................................ 11 Solidarity scenario: improved levels of equity. but questions remain Theagreed option. Solidarity with central guidance. ......................................................... ................................. 13 Financial managementcapacity: how to efectively manage transferred resources ............16 .......................................................................................................................... 18 3. Capacity: Creating a suitable Human Resource Base Conclusions ............................................................. 18 20 Systemic Issues..................................................................................................................... 20 23 4 Central guidance and control .Provincial level issues ........................................................................................................... ................................................................................................. 26 Budgetary and financial control ............................................................................................. 26 5. Managing public investments: the `Caisse de PCrCquation'.................................................... Legality control ..................................................................................................................... 27 6.The Education Sector ............................................................................................................ 29 31 Unfinishedcentral government reforms., ............................................................................... 31 Fiscal and Institutional Capacity inthe Provinces and ETDs ................................................. 33 33 7 Conclusions .Aligningorganizationalmodels ............................................................................................. .......................................................................................................................... 35 List of tables Table 1. Provincial expenditures on Education versus Fiscal Transfer Scenarios ........................ 7 Table 2: Provincial Fiscal Transfer Allocation Options - 8 Table 3. Transfers per province per capita : pure constitutional interpretation ........................... .............................................................. Table 4. Disparities betweenrichest andpoorestprovince per capita under the `pure 11 constitutional' scenario ............................................................................................................. 11 Table 6. Transfer level CP scenario 2008-2011,by Type o f Allocation Table 5. Transfer level CP scenario 2008-2011 ......................................................................... 12 Table 7 Disparities betweenrichest and poorestprovinces under the: CP scenario Table 8 Transfer levels per capita, solidarity scenario, By Component ..................................... 15 Table 9 Transfer levels per capita, solidarity scenario, By Component ... ..................................... 12 ...................12 ..................................... 15 Table 10. Relative and absolute level o f disparity betweenprovinces, per capita....................... Table 11.Provincial Transfers and Wage Expenditures, FY 2008 Budget ................................. 15 .Essential 17 Table 13.Wage BillEstimatesper Province (FC mln, 2008, total Education)............................ Table 12 steps to buildingup provincialpublic service systems................................ -24 31 Listof figures Figure 1. Revenue allocations under the solidarity model .......................................................... 14 Acknowledgements This note i s based on analytical work undertakenbetween January and September 2007. It forms part o f the ongoing, multi-year studies on governance inthe DRC, The program o f analytical work o f governance is joint initiative o f the World Bank and the European Commission. The team that prepared the report wishes to thank the DRC authorities for the constructive discussions undertaken during the various missions undertaken to prepare the note. Tony Verheijen i s the main author o f this report, with inputs from Teresa Polora and Thierry Vircoulon, Kai Kaiser, Jean Mabi Mulumba, Monthe Bienvenu Biyoudi, Helena Ramos, Piet van Heesewijk, Janette Uhlmann, Margo Hoftijzer and Juan Diego Alonso. The study benefited from the insightful and constructive comments o fthe peer reviewers, Deborah Wetzel, Stepan Titov and RanjanaMukhejee. This report was prepared under the general guidance and advice of Helga W. Muller, h a n d Rajaram, Pedro Alba, Judy O'Connor and Jean-Michel Happi. The report also benefited from valuable suggestions and inputs received at different stages o f preparation from Keiko Kubota, Emilie Mushobekwa, Dieudonne Randriamanampisoa, Abdou Salam Drabo, Laurent Messiaen, Alessandro Palmer0 and Raymond Muzellec. Processing and editing assistance from Mavo Ranaivoarivelo and Marie Lo10 Sow is gratefully acknowledged. Helpful logistical assistance was also received from Yvette Kandi Shungu. An earlier version o f this report was discussed during the National Decentralization Forum, held in Kinshasa from 3-5 October 2007 and benfitied from valuable comments and suggestions from participants. 1 1.IntroductionandContext 1. Over the last ten months the decentralizationprocess has emerged as the key issue on the governance agenda in the Democratic Republic o f Congo. Whereas the country faces a large number o f deep challenges in establishing an effective and democratic governance system, few issues have aroused more political debate than the decentralizationprocess, which has engaged the senior political leadership o f the country, up to the President, ina drawn out debate. The decentralizationdebate has also become a microcosm o f all other key issues inherent in the reform o f the governance system, as it touches upon the reform o f the public service, andnotably the public sector wage system, the public financial management system, the delivery o f basic public services and transparency and the fight against corruption, which represent the main cross-cutting governance issues in the Government program. The process also has implications for other areas o f the governance agenda, in particular the reform process in the extractive sector and revenue collection issues related to this. The latter issue is discussed in depth ina World Bank studyonminingsector reform(WorldBank, forthcoming). 2. The decentralization process in the DRC is an inherently political process, and part o fthe broaderpeace and reconciliation process that has taken place over the last five years. Inthis context decentralization is part o f the process o f seeking a new equilibrium and national consensus that provides a more widely accepted sharing o f resources while maintaining the national integrity o f the country. This sets the process apart from the often donor driven decentralization processes in other African states and also requires from the international community a different, less technical and more politically conscious approach. 3. The Constitution o f the 3rd Republic presents the political consensus that was achieved in2005-6. It provides the broad parameters for the decentralization process but leaves several issues with regard to the details o f the implementation arrangements to be decided. For instance, even though it was expected that the Organic Law on Decentralizationwould have been adopted before the 2006 elections, parliament failed to reach consensus, Other pieces o f legislation, such as the Law on Programming the Decentralization process and legislation on key aspects o f public financial management and public service, were also not adopted. At the same time, the Constitution is highly specific on other issues, inparticular on the division o f competencies and on the level o f fiscal transfers (40 percent of domestic revenue). 4. In the meantime, the political process has moved on, as Provincial Assemblies and Governors were elected in late 2006 and early 2007 respectively, leading to a demand for prompt implementation o f the Constitution, in particular regarding the intergovernmental fiscal arrangements. This has exacerbated the problems o f an unfinished legal framework, .i.e. Constitutional definitions o f competencies and transfer levels combine with a lack o f clarity on transitory modalities and aspects o f supervision and coordination between levels o f government. In addition, the central government effectively curtailed transfers inthe first half o f 2007, putting all provinces on a `forfait' 2 o f about 6-7 percent o f domestic revenue, instead o f the 20 percent that was common practise before the elections. Taken together, the lack o f legal clarity and unilateral actions on the side o f the central government created a crisis o f confidence between central and provincial leaders, leading to the emergence o f a highly politicized and acrimonious debate on the decentralization process in the first half o f 2007. However, following a direct intervention by President Kabila, and the National Decentralization Forum in early October 2007, some key principles underlying the process have been agreed, and the decentralization debate has become more constructive, with more attentiondevoted for the tremendous technical and operational challenges ahead. 5. This note, which i s the result o fjoint work by experts from the World Bank and the European Commission, considers technical aspects o f the various elements o f the decentralization debate. It summarizes and pulls together the conclusions o f a series of technical notes that have been produced between April and October 2007, notably the Note on Transitory Arrangements in the Decentralization process (June 2007), the Note on Fiscal Aspects o f Decentralization and Competency Transfers, the Note on Human Resource Aspects o f Decentralization, as well as Briefs on aspects o f guidance and control system (`Tutelle') (September 2007) and the creation o f the National Equalization Fund (September 2007) and an impact assessment o f the Decentralization Process on Education Sector Reform (October 2007). All these notes were developed in close consultationwith government counterparts and the donor community. 6. The analytical work conducted mainly focuses on the provincial level o f government, for a number o f reasons. First o f all central-provincial relations will be a pivotal element o f the new intergovernmental architecture o f the DRC. Second, the issue o f the allocation o f finances, competencies and human resources has taken on a particular urgency with the election o f provincial assemblies and governors in late 2006 and early 2007, which basically created provincial institutions in the context o f a legal vacuum. Resolving the issues that this has generated will be crucial to safeguarding stability inthe DRC. Finally, whereas the importance o f the decentralized territorial entities (the ETD) is obvious, in particular in terms of creating democracy at local level and o f revitalizing local economies, the election o f local authorities has been postponed until Summer 2008 at least, givingthis issue slightly less immediate urgency at this point. 7. The aim o f this summary note is to highlight the technical complexity o f the decentralization process, as well as the risks and opportunities inherent in it. It also provides a review o f the in-depth work conducted on the fiscal aspects o f the decentralization process, which is crucial to obtain an understanding o f the real life impact o f the '40 percent' rule and the difficulties inherent in its practical application. It also provides an overview o f potential solutions to the dilemmas faces by DRC's policy makers. In this regard, this Note will review four key aspects o f the decentralization process (fiscal aspects and competency transfers, human resource issues and capacity development, legal supervision issues, public investment management and the equalization fund) as well as a transversal analysis o f the impact o f these four aspects o f the process on the education sector. 3 8. Over the past year there has been much emphasis on the fiscal aspects o f the decentralization process. It should be highlighted that there are several other crucial aspects o f the process that require attention from policy makers. One o f these is the issue o f sequencing. For example, this note will argue that resolving the allocation o f competencies needs to precede the transfer o f budgets. Even in states that applied a `Big Bang' approach, such as for example Indonesia, far more detailed decisions about the transfer o f competencies had been taken before the implementation o f the `big bang' fiscal decentralization. Sequencing of policy actions i s equally important in other aspects o f the decentralization process. For instance, strong improvements in the normative framework for the civil service are needed before making appointments o f civil servants to the newly created provincial ministries, whether these appointments are made to transfers o f staff from de-concentrated agencies or by recruiting new staff. 9. Another major issue is that o f accountability. This is a central part o f achieving the objectives o f the decentralization process. The transfer o f resources and o f competencies to sub-national level without equivalent strengthening o f accountability mechanisms poses risks o f increased corruption and capture o f the state. Accountability in this context has three directions: upward accountability exists when sub-national governments are accountable to higher levels o f government, for example when the provinces have to report on funds that they have received from the higher level o f government. Horizontal accountability i s the relationship between elected and appointed officials at sub national government levels. This concerns the accountability o f the provincial Division Chiefs towards the Governor or to the Provincial Assembly. The third direction o f accountability is the relation between the elected officials and the voters in their constituency. This accountability is executed by elections, but frequently the voters try to apply additional accountability, by attending assembly meetings or by contacting their assembly members. 10. All three directions o f accountability are currently under development as part o f the decentralisation process in DRC. It is essential that all three directions o f accountability are strengthened as decentralisation i s implemented. International experience shows that all decentralisation programs that ignore this aspect will eventually fail to achieve their objectives. This is, for example, the reason why all decentralisation efforts in Indonesia before 2000 failed. It was only when the new, democratic governments post the Suharto regime came to power that decentralisation made progress, because accountability was included as an integral part o f the "Big Bang" decentralisation. 4 2. CompetencyTransfers andFiscal IntergovernmentalRelations 11. As noted in the introduction, the main focus in the decentralization debate has been on when and how to increase the level o f transfers to 40 percent o f domestic fiscal revenue. Even though this issue in itself poses some serious technical questions, the picture becomes even more complex when considering the transfer o f expenditure assignments, as laid out in the constitution. As per the 2006 constitution, sub-national expenditure assignments include most o f the elements o f pre-school, primary and secondary education, the regional public service, key aspects o f the management o f delivery o f health services as well as a number o f other functions. 12. The danger inconducting the debate on transfers inisolation from a discussion on expenditure assignments is obvious, the discussion on '40 percent' provides an illusion that provinces would have significant resources at their disposal, while in reality it is unclear whether 40 percent o f revenues would be sufficient to even ensure the effective fulfilment o f transferred obligations. Furthermore, issues o f the repartition o f resources betweenprovincial and local level remain to be elaborated inthe context o f the definition o f the responsibilities o f the latter group o f entities, providing for further uncertainty on this issue. In addition, the discussion on competencies is decoupled from the debate on public investment spending, which is covered under a separate mechanism, the `Caisse Nationale de PCrCquation' (CNP), which i s conceived as a national fund for investments in the provinces and will receive 10 percent of domestic budget revenues as per the Constitution. 13. The Note on Fiscal Aspects and Competences (2007) reviews the aspects of revenue transfers and expenditure assignment in an integrated manner and the Note on Education present an impact assessment o f the potential cost o f education expenditures by province. In addition, the Note on Equalization provides options for the ongoing debate on the creation of the CNP. The below discussion o f options synthesizes the detailed analysis presented inthese notes. Competencies: gradual or `big bang'? 14. Global practise on competency transfers basically distinguish between two models, the `big bang' model of immediate transfers o f competencies or a gradual, phased approach. 15. Big bang models, such as for instance applied in Indonesia and the Philippines, are either based on political imperatives, such as the need to pre-empt centrifugal tendencies in a given countries, or on the principle o f a `clean break' with a previous system. The Indonesian case, for instance, included an immediate transfer o f responsibilities as a way to prevent the break up o f the country. Similarly, the Russian case saw rapid federalization as a response to separatist tendencies in some constituent republics. The Philippine case, even if a further example o f a rapid transfer o f 5 competencies, followed a longer preparation process o f some five years, similarly to another case o f `big bang' decentralization, that of Poland. 16. Phased approaches to competency transfers are based on a different logic, which is to transfer responsibility in line with absorption capacity, including budget management capacity. The Spanish and Slovak cases provide good examples o f this type o f approach, while inAfrica Tanzania would provide the most suitable parallel. 17. The choice o f model has significant implications for the discussion on fiscal aspects o f decentralization, while the former model requires a rapid step up in transfers (and/or revenue assignments) the latter is usually paired with a gradual, but defined increase intransfers (and/or revenue assignments). 18. Whereas the two approaches might at first glance appear to be mutually contradictory, there are some elements o f (pragmatic) mixtures between the two. For instance, in Indonesia, even though a `big bang' model was applied, central government, by agreement, retained responsibility for some key aspects o f service delivery, by taking responsibility for managing wage payments o f public officials at regional level. The resources required for wage payments were subsequently deducted from the transfers to be made. Thus, it is important to emphasize that the speed o f implementation is only one dimension, sequencing and maintaining balance are equally important. 19. The case o f the Democratic Republic o f Congo on the one hand shows features o f the `big bang' cases highlighted above, notably the case o f Indonesia with its strong political imperative for decentralization. In addition, the impending division o f 11 provinces into 26 would seem to provide a further argument for a rapid resolution o f the definition of competencies, along the lines set out by the Constitution. 20. However, initial calculations on the impact on educationwage transfers illustrated the risks inherent in decoupling these two elements o f the debate. The first cost calculations on the transfer o f education competencies, completed following the agreement on wage reform in the education sector between the Government and Trade Unions in August 2007, clearly showed the risks inherent in decoupling the debate on fiscal transfers from that on competency transfers. An initial review of potential resource sharing scenarios showed that virtually all o f the provinces would be unable to fund education sector expenditures. 6 Table 1. Provincial expenditures on Education versus Fiscal Transfer Scenarios ProjectedProvincial Transfers as Percentageof Education Wages 2008 BudgetWaaes - - Roiections CON CP SO SO-Rogr Budget IEducationSectorNote CON CP SO SO-Rogr Budget Bmdundu a &J& 121.2% 143.8% && 102.8% && 122.0% Bas-Congo 160.4% 153.9% 196.3% 127.6% 276.2% 127.7% 122.5% 156.3% 101.6% 219.9% Bquatcur J&& 275.0% 140.0% 252.8% 216.8% 110.3% 199.3% bal-Occidental 107.1% 116.6% 279.4% 149.2% 299.6% &f&& 225.0% 120.1% 241.3% Kasal-ChimtaI 255.7% 133.2% 350.1% 202.6% 105.5% 277.4% KamP 299.6% 308.3% 544.5% 358.9% 481.5% 225.6% 232.1% 410.1% 270.3% 362.6% Kinshasa 1180.6% 1116.5% 329.6% 284.9% 227.4% 667.6% 631.4% 186.4% 161.1% 128.6% Maniema J.Q& &,& 439.2% 222.6% 243.8% && 361.2% 183.1% 200.5% Nord-Kiw 100.7% 331.7% 184.5% 313.2% && 261.0% 145.2% 246.5% Chientalc 388.1% 200.7% 326.6% &,&Q $&& 310.6% 160.7% 261.4% Sud-Kin 104.3% 306.6% 167.7% 282.5% ,& , 0I &%% 251.3% 137.4% 231.6% ~ Source: Estimates for 2008 EducationWages are from 2008 draft budget. The Constitutional (CON), Commission Paritaire (CP), Solidarity (SO), Progressive Solidarity Estimates are based on 2008 revenue projections, allocated accordingto different criteria (see FiscalNote, next section). Estimates for the CON, CP, SO scenariosdo not deduct the 5 percentcharges for revenuecollectionbythe OFIDA, DGI, DGE, DGRAD. 21. Table 1highlights the financing challenges for particular provinces under various transfer allocation scenarios that have thus far formed part o f the national debate. The various options set out in Table 2 are discussed in more detail below. Even just considering education expenditures, which are mainly composed o f wages, it is clear that under the Constitutional and initial Commission Paritaire scenarios, most provinces would be unable to finance even teachers wages. For example, under a literal interpretation o f the constitution, Bandunduwould only have financed 6.6 percent o f its teacher expenses. The various alternative scenarios, with a greater emphasis on equity principles and the inability to assign particular revenues such as customs to a particular province, do allow provinces to cover basic costs, although with significant variations. The current nature o f the debate has apparently cemented the claims o f the provinces to all 40 percent o f domestic revenues in 2008. However, the simulations highlight that various phasedtransfers o f both functions andcorresponding revenues may also represent feasible options that could be managed over a 2-3 year period (e.g., as outline in the progressive Solidarity Scenario). 7 Table 2: ProvincialFiscal Transfer Allocation Options Literal Constitutional (CON) 40 % o f all non-petroleum revenues where collected lincluding customs & large tax payers) Commision Paritaire (CP) 40% of allnon-pertroleum revenues where collected, except Ofida, where partially allocated on basis on potmlation (20%) Solidarity Scenario (SO) Revenue classified as type A (easily linkedto final location o f collection) and type B (customs and large tax payers). Oilproducer revenues (type C) excluded. 40 % ofA based onlocationo f collection, 40% ofB revenues based onpopulation (50%), area (30%), and equal lump s u m for each provinced (20%). Type C n Progressive Solidarity Scenario Same as solidarity scenario, but only 20% o fB revenues (SOP) distributed in2008 (pending filltransfer o f functions). 30% in2009. and 40% in2010. 2008 Budget Scenario Proposed for 2008 budget assigns DGIand DGRAD revenues on collection basis, and all other non-oil on populationbasis 22. Considering the above, as well as the conclusions o f the Note on HumanResource Capacity and other assessments (IMF, 2005), it might be argued that it would not be prudent for DRC to follow a fully fledged `Big Bang' approach. Additional arguments against a `big bang' approach are that: a) preparatory work to establish structures and capacity at provincial level has not yet been conducted and b) that the in-depth fiscal analyses o f the fiscal viability o f individual provinces under different modalities o f resource and competency transfers still remained to be conducted. 23. At the same time, however, the clear and strict allocation o f functions, in accordance with the constitution, the planned administrative reorganization (division o f provinces) and political imperatives may make a relatively rapid decentralization o f functions all but inevitable. Considering these two contravening arguments, as well as international practise, the following options have been explored in discussions between Government and its development partners: a. a phased transfer o f competencies, according to an accelerated schedule, which would see the completion o f the transfer by the time the new administrative structure o f the state is to be created. A tentative schedule for such an accelerated transfer i s presented in the Fiscal and Competencies Note;' b. a solution whereby a formal transfer o f functions would take place, but accompanied by an agreement between central and provincial governments IThe basic premise o f the proposals inthis note is to start with systemic functions, which are essential conditions for other functions to be performed effectively. These would include public service management and financial management. This couldthen be followed by essentially local functions (such as spatial and urbanplanning, housing etc.) and ina second and thirdphase by social services, including, at a final stage, education 8 for the central govemment to temporarily continue to manage certain aspects o f these functions (inparticular salary payments), the cost o f which would be deducted from transfers. 24. Following the discussions at the National Decentralization Forum in October, it was agreed that a variant on the second option would be the most reasonable solution for the DRC at this point intime; functions and resources would be transferred inaccordance with constitutional provisions, but wage payments are to be made protected items in provincial budgets, and based on lists o f accredited officials established by central government as o f end 2007. Thus, whereas the logic o f the constitution would be respected, important safeguards to prevent a collapse o fpublic services were also agreed. Fiscal aspects: balancingentitlements,equity andviability 25. The policy debate in the DRC has focused so far primarily on the issue o f resource transfers, and the modalities (and speed) o f the implementation o f Constitutional provisions. The discussion in the above section has shown that advancing fiscal aspects o f the decentralization process in isolation from a debate on functional transfers will overlook key issues, such as the real cost o f fulfilling key functions. Nevertheless, unless the principles underlyingthe implementation o f the constitutional provisions on resource allocation are agreed, it is difficult to envisage the conduct o f a more rational debate on functional transfers and other key issues o f the decentralizationprocess. 26. How much revenue would each province receive under different fiscal allocation criteria? The bulk o f domestic revenues in DRC are collected by the central government, through collection entities. These include customs (Ofida), fees and charges (DGRAD), direct taxes (DGE), large taxpayers (DGE), and oil-related revenues (that are collected by both DGI and DGRAD). The various allocation scenario assume that different revenues can be directly attributed to a province, or whether some other criteria o f allocation (e.g., population i s more suitable). Depending on the projected growth o f each individual revenue category, this will not only impact the overall pool o f fiscal resources allocated to the provinces (e.g., a vertical share o f 40 percent o f domestic non- oil revenues), and whether particular revenue pools have more favourable prospects (e.g., large taxpayer receipts versus customs). Figure 1highlights the main revenue pools by type and how they are projected to develop over time.* O f the CFC 969 billion in domestic revenues projected for 2008, the most important revenue pool was customs and duties administered by OFIDA (40.7 percent), large taxpayer revenues administered by DGE (26.6 percent), petroleum related revenues (16.9 percent), fees and charges administered by DGRAD (8.2 percent), and direct taxes administered by DGI (7.6 percent). 'Asper IMFandgovemment estimates. 9 Figure 1:Domestic Revenue Projections, by Type (2006-201 1) National Revenues(ActuaUProjected) CFCBillions 1,200 PEXROLPROD DGRAD 0DWDGE 800 DGl 600 OFIDA 400 200 0 2006 2007 2008 2009 2010 2011 Constitutional interpretations, equity and viability 27. There are multiple ways o f interpreting the constitutional provisions that 40% o f domestic revenues should be transferred to the provinces, and that this should be done by withholding such revenues at the source. The logic inherent in the constitution is based on two premises: a. that individual provinces will be viable entities under this scenario b. that all revenues can be automatically attributed to individual provinces 28. The analytical work presented in the Fiscal and Competencies Note, has highlightedthat neither o f these premises is valid. Firstly, a literal interpretation o f the constitution, using the contribution that each province makes to domestic revenues as a basis for retrocession, would create unacceptable disparities in allocated revenue per capita by province. Secondly, it would make more than 75 percent o f all provinces financially unsustainable (see Table 1). Thirdly, several important revenue categories (off shore oil revenues, most customs revenues and revenues collected by the central Large Taxpayers Unit, which are all paid in Kinshasa) are not attributable to individual provinces, and therefore further exacerbate large income disparities, mainly between Kinshasa and other provinces with important customs posts and, on the other side the rest o f the country. 29. The purely constitutional approach to transfers does not recognize these issues o f inequality. Table 3 and Table 4 highlight the significant provincial revenue disparities evident under an application of a literal constitutional interpretation. As a first modification it i s generally agreed that oil sector revenues, or at least those that are 10 generated through off-shore operation^,^ should be set aside from the intergovernmental revenue allocations. This in view of their temporary nature and the high degree o f volatility in the size o f these revenues. This provision i s therefore applied to all o f the scenarios presented below. It should also be noted that these scenarios deal only with provincial revenue due to revenue sharing (retrocession). They do not include any own revenue generated by the provinces. Increasing these own revenues will in the longer term need to play a larger role, inorder to create a stable fiscal basis for the provinces. Table 3. Transfersper provinceper capita :pureconstitutionalinterpretation 2008 2009 2010 2011 Bandundu 215 246 285 338 Bas-Congo 4,153 4,822 5,642 6,669 Equateur 263 304 354 419 Kasal Occidental 1,566 1,784 2,050 2,439 Kasal Oriental 954 1,089 1,252 1,487 Katanga 3,450 4,159 4,905 5,700 Kinshasa 34,896 41,286 50,207 59,163 Maniema 165 198 234 273 Nord Kivu 810 964 1,138 1,332 Province Orientale 331 396 465 542 Sud Kivu 1,030 1,199 1,397 1,648 Average 5,349 6,322 7,624 8,976 Table 4. Disparitiesbetween richestandpoorestprovinceper capitaunder the 'pure constitutional' scenario 2008 2009 2010 2011 "Richest" (Kinshasa) 34,896 41,286 50,207 59,163 "Poorest" (Maniema) 165 198 234 273 Absolute Difference 34,731 41,088 49,972 58,890 Ratio 212 209 214 217 Source : Calculs du staff The 'Commission Paritaire' scenario :not enough solidarity 30. The two problems highlighted above have been recognisedby Congolese policy makers, and were discussed in the context of the 'Commission Paritaire' (CP). This commission is constituted o f representatives o f the central government, o f the provincial governors and o f the provincial assemblies. It was established in June 2007. The conclusions o f the technical work done by the CP highlights both the issue o f disparities and the problems o f attributing certain revenue categories to individual provinces. In its conclusions, the CP therefore suggested to take customs revenues (about 26% of total domestic revenue) out o f the equation for direct allocation and to divide this in part according the location where the revenue was raised (80%) and in part according to a formula (20%) based on the size o f the population. The CP scenario has a certain equalising impact. However, in the CP proposals the disparities between income levels Representatives of the province o fBas-Congo have argued for the allocation o f40% o fthe on-shore revenues to be transferred to this province, as these revenues are generated there 11 per capita remain unacceptable, and they put the viability o f a large number o f provinces further into doubt, especially considering the expenditure categories that would be transferred, such as primary and secondary education. Table 5, Table 6, Table 7 below present the retrocession scenarios according to the CP, and highlights that this scenario i s still subject to significant disparities. Table 5. Transfer level CP scenario 2008-2011 2008 2009 2010 2011 Bandundu 610 694 797 949 Bas-Congo 3,984 4,630 5,423 6,407 Equateur 654 748 862 1,024 KasaY Occidental 1,703 1,940 2,229 2,652 KasaY Oriental 1,208 1,378 1,582 1,881 Katanga 3,550 4,272 5,035 5,854 Kinshasa 33,002 39,139 47,75 1 56,236 Maniema 582 670 775 918 Nord Kivu 1,156 1,355 1,586 1,865 Province Orientale 727 845 978 1,154 Sud Kivu 1,310 1,516 1,760 2,08 1 Average 5,349 6,322 7,624 8,976 Table 6. Transfer level CP scenario 2008-2011, by Type of Allocation 2008 2009 2010 2011 A B A B A B A B Bandundu 28 582 35 659 43 754 50 899 Bas-Congo 1,146 2,839 1,413 3,217 1,743 3,680 2,020 4,388 Equateur 59 596 72 675 89 772 103 921 Kasal Occidental 91 1,612 113 1,827 139 2,090 160 2,492 Kasai Oriental 65 1,144 82 1,296 99 1,483 113 1,768 Katanga 1,786 1,764 2,273 1,999 2,749 2,286 3,128 2,726 Kinshasa 23,264 9,739 28,103 11,036 35,127 12,624 41,184 15,052 Maniema 87 495 110 560 134 641 153 764 NordKivu 374 782 469 886 573 1,013 657 1,208 Province Orientale 148 579 188 656 228 751 259 895 Sud Kivu 265 1,044 333 1,183 406 1,354 466 1,614 Average 3,186 2,163 3,871 2,451 4,820 2,804 5,633 3,343 Table 7. Disparities between richest and poorest provinces under the: CP scenario 2008 2009 2010 2011 "Richest" (K 33,002 39,139 47,751 56,236 "Poorestq'(M 582 670 775 918 Absolute Dif 32,420 38,469 46,976 55,318 Ratio 57 58 62 61 Source : Calculs du staff 12 Solidarity scenario: improved levels of equity, but questions remain 31. Therefore, neither the pure constitutional scenario, nor the solution proposed by the CP provide satisfactory solutions to the essential requirements o f viability, equity and the feasibility to identify the actual origin o f revenues), a third alternative scenario was developed in the context o f the development o f the background note on fiscal and competency aspects o f decentralization. 32. This scenario builds on some o f the same elements as those developed for the CP, in particular the feasibility to attribute revenues to specific provinces. The main difference between the solidarity scenario and the CP scenario is that under the solidarity scenario both customs revenue and the large taxpayer unit's revenue are taken out o f the equation for direct allocation4. These two revenue sources are proposed to be allocated according to a formula that augments the allocations that are based on equity principles. There is limited technical feasibility o f identifying the origin o f exported and imported goods (as exemplified by the problems that have arisen on the same issue inthe Republic o f Congo). There i s an increased risk that such a process would create large opportunities for fraud and corruption. Therefore, under the solidarity approach, a different mode o f allocation o f the custom revenues across the provinces i s required. Similarly, many revenues generated by the Large Taxpayers Unit at the Tax Administration in Kinshasa actually represent revenue generated by sources in different provinces, notably in Katanga, but also by many others. Therefore, the allocation o f these two sets o f revenues through a formula would enhance the equity and viability o f a larger number of provinces andit would be more equitable with regard to the limitedfeasibility to attribute revenues to a particular province. ~~~~ ~ ~ Inadditionto off-shore oilsector revenues 13 Figure 2. Revenue allocations under the solidarity model 100.0% C:RecettesNationales (13.5%, c 90.0% FC 126billion) 80.0% 70.0% B: Formule (46.3% 60.0% 50.0% 218.9billion) Province (0-->40%) e 40.0% E 30.0% A: Derivation Nationale(60%) 20.0% Automatique 10.0% (40.3%, FC 196.6billion) 0.0% Province (40%) Recettes Budget2006 C. Hydrocarbeurs/DroitsAdministratives 1C.Revenuesdessocietes(Produitspetrolies) Hydrocarbeurs/PetroliersProducteurs B. B. Revenues des societes B. Droits Et Taxes A L'Eportation B. Droit et Taxes A L'lmportation A. Autres A. Recettes Domainele (Autres) A. Recettesdes Participations A. Recettes Domaniales (Mines) 1A.RecettesJudiciares 0A. RecettesAdministratives A. lmpots Sur Les Revenus (non-petrol/societe) A. DirectTaxes BA. Droit de Consomation 33. The allocation mechanism for the Customs and Large Tax Payer revenue sources i s proposed to be a simple and transparent formula, based on the factors population (50 percent), territory (30 percent) and lump sum (20 percent). The first two reflect specific aspects of the cost of public services (inparticular in education and health) and the third aims to guarantee a basic level of income for all provinces. The simulation results for this approach are presentedintables 6 and 7 below. I t should be noted that this approach also proposed a gradual phasing in o f transfers (starting from 28% in 2008 to 36% in 2010, which is the same level as under the other two scenarios), which makes total transfer levels lower than the other two scenarios for 2008 and 2009. This is obviously contingent on the decision on the modality and speed of functional transfers as per the above discussion. Table 8, Table 9, Table 10 highlight the projected development of transfers to individual provinces under this scenario, including a breakdown (A versus B) of revenues assignedon a locationof collectionversus formula basis. 14 Table 8. Transfer levels per capita, solidarity scenario, By Component 2008 2009 2010 2011 Bandundu 3,972 4,65 1 5,593 6,611 Bas-Congo 5,082 6,008 7,243 8,526 Equateur 4,704 5,511 6,627 7,832 Kasai' Occidental 4,083 4,780 5,738 6,782 KasaY Oriental 4,246 4,973 5,974 7,060 Katanga 6,270 7,516 9,039 10,567 Kinshasa 9,743 11,622 14,155 16,621 Maniema 7,104 8,325 10,012 11,831 NordKivu 3,807 4,486 5,400 6,364 ProvinceOrientale 4,956 5,817 6,995 8,259 SudKivu 3,851 4,527 5,444 6,423 Average 5,349 6,322 7,624 8,976 Table 9. Transfer levels per capita, solidarity scenario, By Component 2008 2009 2010 2011 A B A B A B A B Bandundu 52 3,920 61 4,590 74 4,590 87 5,519 Bas-Congo 1,525 3,557 1,843 4,165 2,234 4,165 2,605 5,009 Equateur 84 4,620 102 5,409 123 5,409 143 6,505 Kasaf Occidental 277 3,806 324 4,456 379 4,456 447 5,359 Kasai' Oriental 177 4,070 209 4,765 244 4,765 286 5,730 Katanga 1,996 4,275 2,511 5,005 3,020 5,005 3,452 6,O 18 Kinshasa 7,096 2,647 8,523 3,099 10,429 3,099 12,216 3,726 Maniema 97 7,007 121 8,204 147 8,204 168 9,865 Nord Kivu 429 3,378 531 3,955 644 3,955 742 4,756 Province Orientale 171 4,785 215 5,603 258 5,603 295 6,737 Sud Kim 362 3,490 442 4,086 531 4,086 615 4,9 13 Average 1,360 3,989 1,652 4,670 2,009 4,670 2,338 5,616 Table 10. Relative and absolute level of disparity between provinces, per capita 2008 2009 2010 2011 "Richest" (Kinshasa) 9,743 11,622 14,155 16,621 "Poorest" (Nord Kivu) 3,807 4,486 5,400 6,364 Absolute Difference 5,936 7,135 8,755 10,256 Ratio 2.6 2.6 2.6 2.6 15 34. The calculations under this scenario are overall much closer to the criteria outlined above (equity, viability, alignment with constitutional principles) than the results generated by the other two options. It brings the system closer to generating an acceptable basic level o f income for all provinces, does not adversely affect the ability of richer provinces to meet their expenditure obligations on a per capita basis and it is true to the spirit o f the constitution. Thus, from a fiscal, economic and developmental point o f view this would clearly be the more desirable option. The agreed option, Solidarity with central guidance 35. Discussions at the National Decentralization Forum and subsequent technical work by a mixed Commission o f Experts have led to the identification o f a consensus position on the fiscal aspects o f decentralization, one o f the key preconditions to move the process forward. The consensus view elaborated following the Forum has the following main features: a. It follows the logic o f the categorisation o f revenues set out under the `solidarity' scenario (attributable revenue, non attributable revenue and `off shore' revenue, as well as the principle o f applying different principles o f distribution according to this categorisation) b. A simpler formula (based on a population co-efficient only) has been devised for non-attributable revenue c. Although a re-allocation o f resources between Kinshasa and Bas-Congo (on an ad hoc basis) was introduced duringthe Forum discussions, this has not yet beenreflected inthe final 2008 budgetproposal5 36. In addition to the revenue side aspects, risk mitigation measures have been proposed on the expenditure side. Based on a calculation o f actual wage expenditure under transferred functions, wage enveloped will be protected items and provinces will be obliged to base wage payments, at least in 2008, on the staffing lists provided by the central government, thus postponing de facto some delicate problems o f human resource transfers. 37. Initial estimates o f the impact o f the above scenario shows that a large majority o f provinces would be fiscally viable, albeit with very limited room for manoeuvre interms o f expenditure, at least for the fiscal year 2008. Table 12 presents projected fiscal allocations to individual provinces as set out in the 2008 budget. DGI and DGRAD revenues are assigned on a collection bases. The bulk o f other non-oil revenues (Ofida, DGE) are assigned on a population estimate. Based on estimates of wage costs, all provinces would be able to cover education, health, agriculture, traditional authorities, public functions transferred to the provinces. However, a number o f provinces such as Bandundu and Kinshasa would have limited room to manoeuvre beyond these Based o n the so-called `osmosis' principle, which considers the interlinkage between the Kinshasa and Bas-Congo economies 16 expenditures given prevailing wage liabilities. Given recent pressures to increase wages ineducation, the realism ofthese wage estimates for the fullyear of2008 may also need to be verified. Table 11.ProvincialTransfersandWage Expenditures,FY 2008 Budget Population Ratio to PC by PCby Education Provincial (2008 Alternative PC Total Collection Population Wages/Total WagedTotal Budget) Pop Figures Bandundu 4,373 22 4,351 69.5% 86.9% 7,637,700 108% Bas-Congo 8,254 3,903 4,35 1 36.2% 48.5% 3,387,437 87% Equateur 4,396 45 4,351 39.6% 57.3% 6,364,842 98% KasalOccidental 4,415 64 4,351 33.4% 46.5% 5,326,850 99% KasalOriental 4,409 58 4,351 28.6% 41.4% 5,874,994 132% Katanga 1,775 3,424 4,351 20.8% 3 1.5% 6,230,738 71% Kinshasa 6,800 2,449 4,351 44.0% 80.0% 7,184,001 99% Maniema 4,426 75 4,351 41.1% 60.9% 1,515,381 89% Nord Kivu 4,802 451 4,351 31.9% 46.8% 3,398,078 75% Province Orientale 4,461 110 4,35 1 30.6% 44.3% 6,732,660 93% Sud Kivu 4,661 310 4,351 35.4% 49.2% 3,291,783 76% 5,347 995 4,351 37.4% 54.9% 56,944,464 93VQ 38. Whereas the compromise solution has a number o f positive features, several important issues remain to be addressed: 0 Fiscal estimates rely heavily on provincial budget estimates for own revenues. These estimates show a wide variation, posing a risk o f buildinga system o f shifting sands, especially once the protected status o f wage expenditures will be eliminated;6 0 In several instances, provinces do not appear to have budgeted adequate resources for non-salary related staff cost Cfrais defonctionnement), which, if not properly analysed could lead to the problem o f fiscal viability re-entering through the backdoor The legal basis for the above system i s still far from complete. Whereas it may be possible to operate, for the first half o f 2008, based on interim regulations, the rapid adoption o f an amended Organic Public Finance Law, based on a decentralized public expenditure management context remains an urgent priority 0 While population figures provide a simple and transparent criterion for allocating the bulk o f transfers to the provinces, the lack o f even basic data in DRC means that additional clarification may be needed as to the choice o f numbers used for the interim allocations pending a full census. For example, Table 12 highlights that there are significant differences between the population figures used in the 2008 budget, versus those projecting forward 2002 baseline figures. 0 Whereas there appears to be broad agreement on the interim solutions for wage bill management, a number o f problems (establishment o f real staff numbers, creation o f payroll management capacity in the provinces and at local level) remain as urgent as ever, and there i s a risk that a solution to these Wage expenditures will be mandatory items for the provinces at least for the year 2008. Provinces will be obliged to pay wages as per staff lists drawn up bythe central government 17 fundamental problems may be postponedindefinitely, thus leaving the process ina `half-way ho~se'.~ 39. Therefore, whereas political brinkmanship and a (reluctant) willingness to compromise have created what appears for the time being to be a viable solution to the fiscal and competency transfer dilemmas inthe DRC, a number o f crucial issues remain to be addressed over the coming months (the first three issues above) or throughout 2008 (the last issue above). Only once these points will have been addressed the process will be on solid ground. Financia1management capacity: how to effectively manage transferred resources 40. Apart from the political and technical questions surrounding the definition and phasing in o f transfer mechanisms, an important capacity building agenda remains, in particular at provincial, and later also at ETD level. Discussions with provincial authorities have revealed an important agenda o f establishing strengthened local systems for managing public resources. In particular, the development o f new rules and procedures for managing the expenditure chain is an urgent issue, as existing legislative and regulatory frameworks do not account for the existence o f a provincial government or provincial ministers. 41. The nature o f the relation between the adoption procedure o f the provincial budgets and that o f the national budget needs to be urgently clarified (see the point on the credibility o f estimates above). Furthermore, conflicts have emerged in a number of provinces due to the lack o f fulfilment o f the role o f the assemblies in the adoption o f provincial budgets. Similarly, the relation between the provincial government and the tax, customs and administrative fee collection agencies needs to be clarified. Whereas it i s clear that these agencies will continue to be subordinated to the central government, a linkage between the revenue management, budget preparation and expenditure managements systems needs to be defined and operationalized inthe relevant IT systems. In addition, a serious effort in capacity building of the newly appointed provincial governments (ministers and their cabinets) will be required, as most o f the appointees are new to public management processes. Conclusions 42. Finally, it is important to emphasize once more that in the medium term it is essential that the fiscal aspects o f decentralization are put on a firm and stable basis. There will continue to be uncertainty about provincial revenues based on the volatile nature of the political agreements that underpin the revenue sharing arrangements. Ultimately, there will need to be a re-allocation o f taxes and other revenue categories The case o fIndonesia shows the real nature o f this type o f risk 18 between the levels o f government, large improvements in own revenue generated by the provinces and a fundamental reform o f the tax administration, to enhance its accountability and to bring the tax administration in line with the new reality o f the decentralised state system. A first step in this reform process is the revision o f the nomenclature o ftaxes. 43. Finally, a similar exercise on fiscal transfers remains to be conducted at the level o f each province on the allocation o f responsibilities and o f resources between the provinces and the Territorial Decentralization Entities (ETD). 19 3. Capacity: Creatinga suitableHumanResourceBase 44. The Constitution o f the DRC foresees the creation o f a Provincial and Local Public Service, as one o f the elements o f the decentralization process. Whereas this is a key condition for other elements o f the reform process to succeed, there has been surprisinglylittle interest inthis issue, notably at the central government level. Provincial governors have frequently raised the disconnect between the principle o f decentralized governance and a centrally managed civil and public service. Equally, provincial officials have highlighted the risks that the lack o f clarity on issues o f subordination poses for institutional stability at provincial level. The risks include a disconnect and overlap between the parallel public service systems, which would be both fiscally expensive and unnecessary, and the degradationo f the quality o f the existingpublic service at provincial level, which so far had been relatively unaffected by the malaise in the public service at central level. Furthermore, based on international experience, decentralization processes pose serious organization and management issues, in particular the question o f whether and how to transfer staff currently working at de-concentrated government agencies to the provincial public service to-be-defined. 45. Several fundamental issues are to be addressed ifa working public service system at all levels o f government is to be created. Some o f these issues are elements o f overall reforms that are long overdue, such as public service wage system reform and the management and redeployment o f staff between de-concentrated and decentralized units. Others are more specific to the provincial level, such as the creation o f human resource management systems at provincial level that would be able to handle issues o f recruitment and career management and the issues referred to above, i.e. decisions on ways o f handling the transfer o f personnel that will most likely need to accompany the transfer o f competencies, and how to handle the cases o f staff that may not meet the requirements posedby the new responsibilities assigned to the provinces. Systemic Issues 46. Recent studies o f the public service system in the DRC (World Bank, 2007) highlightseveral key systemic challenges: 0 To modernise the civil service system and amend its legislation inorder to bringit up to improved standards o f efficiency in service delivery and human resources management by means o f reinforced values like merit, results, performance and professionalisation; 0 To fill the normative vacuum by setting rules and procedures for the management o f the new decentralised civil service; 0 To ensure consistency between the legal base and organization o f the national and sub-national civil services with the new constitutional order; and 20 To make the public administration a more attractive employer for young skilled professionals by reforming and improving the public sector wage system through systematic and sound reform measures. 47. The current civil and public service system, while upholding the illusion o f an integrated system, i s in fact a system fragmented along sectoral lines, based on more than 27 legal statutes that provide significantly different employment conditions for different professional groups in the public service. The harmonization o f this system, and the development o f a small number (4-5) model statutes for specific professional groups (general public service, education, health and social services, civil security) is a pre- condition for the development o f a manageable provincial public service system, as provincial statutes will invariably be based on the central model. Inthis line o f reasoning, chaos at central level will bring chaos in the provinces. The modernization o f the legal base for the civil and public service system therefore needs to be well aligned with the creation o f the provincial public service. 48. Along with the creation o f a modernized (set of) statutes for the civil and public service, the development o f secondary legislation and rules and procedures that would enhance the chances o f developing a merit based and professional public service is a further priority, as provincial legislationwould be likely to follow the central examples. 49. A second element o f systemic issues is the public service wage system. Even if there are some 27 different statutes are currently in place, they invariably create opaque and inadequate wage systems, based on a plethora o f bonuses and allowances and very low base wages. Whereas more recently the establishment o f a more transparent system has become an issue for discussion, notably inrecent discussions on education wages, the principles o f transparency and predictability are still far from being respected.* In addition, there are large differences betweenprovincial and Kinshasa wage levels, mostly brought about by differences in allowances, which are not justified by differences in responsibility or difficulty o f the job. A separate World Bank study addresses specifically the reform o f the public service wage system, and proposes a medium term strategy to address these issues. It is highly important that decisions on a wage reform strategy be taken rapidly, so that these can be taken into account when designing wage systems at provincial level. Specifically, provincial level wage systems should respect the same principles as those applicable overall (equal pay for equal work, merit, transparency and predictability). In addition, whereas regions may be given the right to vary salary levels, it is important that variations remainwithin a certain defined level (e.g. -2O>X<+20). For this to work, central level wage system issues would need to be addressed at least at the level o f strategy and policy. 50. Finally, central rules for transfer and deployment o f public servants would needto be set, along with attestation procedures, which would ensure an orderly transfer, where appropriate, o f staff between the current de-concentrated units and the new provincial public service. T h i s i s a problem inparticular inthe general public service, where base pay is between 5-10 percent o f realpay 21 51. Another set o f systemic issues is evolving around the organisational aspects o f decentralisation. Inpart these issues derive from the lack o f clarity regardingthe relations between central and provincial level and to the incomplete devolution o f functions between these levels o f government. As a result, a number o f organisational difficulties have emerged. The existing de-concentrated entities (Divisions and `Rigies FinanciBres') find themselves ina position o f uncertainty with regard to their tasks and responsibilities and the hierarchical relation with central and provincial government. Inmost provinces it i s assumed that these entities will support the provincial authorities, but the extent o f their involvement and responsibilities for provincial functions is not clear. On the other hand, the new provincial governments, who lack adequate staff o f their own, expect full support and allegiance from these de-concentrated entities. At occasions, this situation has led to conflicts and confusion, and more frequently de-motivated staff and resistance. There are cases were staff work enthusiastically for the provincial authorities, motivated by the opportunity to advance their home province, but this is not common, especially among the `Rigies FinanciBres', which are centralised entities with a national staff rotation programme. 52. Inthe meantime, new organizational structures have beenset up at the provincial level. In accordance with the Constitution, the Governors have appointed provincial ministers. These will be supported by ministries, but at present, these are not yet firmly established or staffed. Nevertheless, the organisational shape o f the provincial governments i s becoming clear, based on thejob titles o f the provincial ministers. Ideally, these job titles should reflect the competences that the Constitution transfers to the provinces to ensure that all tasks and responsibilities are adequately covered. A review o f the job titles o f the ministers inthe 11provinces shows that there i s a wide variety injob titles and in the division o f labour in the provincial governments. All provinces bar one (Bas Congo) have limited themselves to the maximum o f 10 ministers prescribed by the Constitution'. The Ministry o f Interior has made an attempt to regularize the scope and responsibilities o f these ministers but this happened after their appointments and this guidance was therefore widely ignored by the provinces. The result i s a variety that reflects the circumstances o f the provinces, but also the need to have a number o f core government responsibilities covered by the ministers. For example: all provinces have a ministers responsible for Finance (often combined with Economy), Planning and Budgeting (mostly combined, but separated from Finance), and Territorial Administration (including decentralisation and the relations with ETDs). Not well assigned among these core functions is public administration. Seven o f the provinces have not clearly allocated this function to one o f the ministers. Beyond these core government tasks, the provincial ministries show a wide variety o f combined responsibilities. In most cases these combinations make sense, in order to limit the size and cost o f the provincial administration. The only concern could be that in five provinces Health and Education are combined into a single ministry. This may create difficulties as these are large sectors with complex problems and huge numbers o f staff which from FY2008 onward will become a responsibility o f the provincial governments. 9 Bas Congohas added one CommissaireGeneraldes Mines avec rangde Ministre Provincial. 22 53. One o f the key risks inherent particularly to this element o f the reformprocess is a disconnect between the rapidly advancing debate on fiscal decentralization (and increasingly on transfers o f competencies) and the very slow movement in the public service reform process. Whereas ideally the three above issues, normative base, wage system and transfer principles, should be addressed in each province on the basis o f a central strategy, this may prove to be elusive if the speed and direction o f the public service reform process is not increased dramatically. This poses, yet again, the risk o f a legal vacuum, which could generate a disorderly process inthe development o f provincial public service systems. Therefore, if the reform process at central government level does not move forward, it could be foreseeable that provinces themselves move forward on the definition o f provisional rules and procedures for public service management. In this regard, adherence to good practise principles, both in recruitment, career management andwage policy, are crucial. Provincial levelissues 54. Considering the fact that central level reforms on the legal statute and wage and management systems for the public service may not move fast enough to address the immediate needs o f the provinces in building up their public service systems, it i s important that capacity issues at provincial level are addressed without delay. Table 8 below sets out the essential elements o f a strategy to address this immediate need. The steps highlight the complexity o f the issue, which has repercussions for three levels o f government; For central government level, which will face the need to reduce the role, functions and staffing levels o f de-concentrated agencies ina significant way; For provincial level, where legal and organizational frameworks as well as management capacity will need to be created, and processes to decide whether and how to absorb current staff at de-concentrated agencies will need to be defined and implemented For local (ETD) level, that will have to establish, eventually, its own capacity to handle public service management issues. ETDs face specific challenges in this regard, due to remoteness and the small size o f units, which will require the pooling o fresources andcapacity to handle HEWissues. 23 Table 12. Essential steps to building upprovincial public service systems Stepsfor creation HRM capacity Enabling conditions Creating provincial capacity Definition o fthe provincial authorities for establishing the ministry staff (ministries ) responsible for public ---- Resources Procedures for competitive recruitment based on service management overall legal principles and existence o f a control Establishment o fprovisional provincial mechanism rules andprocedures for public service -- Development o fplanning tools for assessing staffing management needs Create capacity to handle payroll and -- Establish interface with the public finance ministry for personnel database management management the payroll system Create processes and capacity to manage recruitment o f provincial staff, from deconcentrated entities or from the local labor market. o Staff new provincial public service systems Review of de-concentrated agencies Review o f competencies and -- Clear and transparent relations between provinces and coordination mechanisms betweenthe de-concentrated agencies, prevention o f overlap in provincial structures and de- competencies and functions concentrated entities Adoption o fthejuridical instruments -- Clear procedures and mechanisms to manage the providing for a new set o f retirement o f superfluous staff competencies and coordination mechanisms Adaptation o f the structures o f de- concentrated entities Create processes and capacity to manage attestation o f de-concentrated entity staff Adaptation o f the number o f posts, includingretirement o f staff older than 56 years inall deconcentrated entities. Incoordination withthe provincial ministries, manage the (partial) transfer o f staff from deconcentrated entities. Managing capacity at sub-provincial level (ETD) o Create capacity to manage exclusive -- Formal allocation o fpersonnelmanagement authority competencies allocated to the ETD. to the ETD. Transferring public service -- Establish clear rules for personneltransfers from de- management authority to decentralized agencies bodies. --concentrated Establishment o f pools o fpersonnel that canprovide Set up inter-municipal services to support to individual ETDbased on a rotationschedule support HRh4inlocal authorities -- Sufficient funds to cover operational cost for support functions to the ETD. 55. The essential steps outlined above are complex, and will take significant time to put inplace. Inthe current situation there is inmost provinces an uneasy co-habitation of public and civil servants (who are still accountable to the central government) and 24 provincial ministers. These ministers are appointedby the elected governors and formally they do not have authority over the de-concentrated units that operate at the provincial level. This has already ledto budget preparation processes inprovinces being handledby small groups o f politically appointed officials (cabinet staff), without being able to draw (sufficiently) on the more experienced staff available in de-concentrated units. This is a clear indication o f the need to find a solution to this anomaly, caused by the co-existence o f outdated legal statutes with new political realities. 56. Once legal and competency issues are addressed, a significant capacity building agenda will remain. Provincial governments do not currently have the capacity, the manpower or the systems to handle career and wage bill management, and they have limited, if any, capacity to manage recruitment and dismissal processes. In addition, to take responsibility for the management o f a provincial public service will include the need to negotiate with trade unions on employment conditions and wage levels. This is currently handled exclusively by the central government. 57. Like many aspects o f the decentralization process, human resource management issues both poses opportunities and risks. The creation o f provincial public service systems may help unblock stalemates on crucial issues such as statutory and wage reform, as provinces, inthe absence o f central legislation, could establish their own legal regimes, within the constraints o f constitutional principles (which outline in some detail the principles on the organization and management o f the public service i s to be based), including more rational and viable wage systems. However, at the same time this poses risk o f fragmentation and disparities incivil service employment conditions, which is far from desirable in a decentralized unitary state. It also poses risks for the very limited capacity to manage public service issues at provincial level, both in terms o f career and wage system management and interms o f negotiation capacity with the trade unions. 58. Considering the above, it i s essential that rapid initiatives are put inplace to assist provinces in putting in place management capacity for the provincial public service. In addition, in the absence o f progress in reforming employment systems at central government level, provinces could need to be assisted indrawing upprovisional rules and regulations for the provincial public service, until key issues at central level might be unblocked. 25 4. Centralguidanceandcontrol 59. The Constitution provides, at a general level, guidance to the management o f relations between the central level and the provinces and, ina more limited way, between the Provinces and the ETD. The Constitution clearly distinguishes between exclusive powers (at provincial and central government level) and concurrent powers. 60. The Constitution focuses on the legal aspects o f these relations, Le. the powers o f the national and provincial assemblies to adopt laws that affect the lower levels o f government. The background Note on issues o f control and supervision raises a number o f options to operationalize these legal provisions, in particular with regard to public expenditure management and budgets. The note provides examples o f the different options for ex-ante and ex-post control. 61. It should be noted that the Constitution does not provide clarity on the issue o f authorities the competencies which are to be shared between the national and the provincial authorities, apart fi-omreaffirming the acknowledged legal hierarchy according to which a national law shall prevail on a provincial law. The principle o f subsidiary should help filling the gap, providing for that an action should inprinciple be taken at the level which i s the closest to the citizen for as long as it appears as the level where the action can be carried out more effectively. 62. A first and crucial task is to provide detailed guidance to central and provincial authorities on the nature and implications o f the changes in the guidance and control function o f the central government. This implies adopting appropriate legislation and appears particularly urgent as the last months have already shown an increasingly uneasy cohabitation between provincial ministers and their cabinets, whose authority i s derived fromthe Governor, andthe de-concentrated entities, which are subordinated to the central ministries." 63. The delay in the creation o f a Constitutional Court, foreseen in the Constitution, makes taking decisions on such issues highly complicated in the interim assured by the Supreme Court alone. . Budgetary and financial control 64. As far as the control o f budget and finances is concerned, the principles set out by the constitution (providing for a decentralization unitary state) leave open four options to define control systems inthis area: 10Traditionally these units also reportedto the Administrative Directors at Provincial level. However, since these are centrally appointedpositions this does not resolve the current stalemate on issues of administrative subordination 26 C. Ex-ante control o f budgets by the central government, based on a review o f budgets adopted by the provincial assemblies. Central authorities would in this case submit their observations to the provincial authorities and ask the provincial executive and assemblies to reflect these in an amended version. Under this model, incase of disagreement the view o f the central authorities could prevail and be reflected in a veto, but only in case of three specific circumstances, being; a) a lack o fbalance between revenues and expenditures; b) the lack o f commitments on obligatory expenditure items, and; c) the lack o frespect for budget adoption deadlines d. A review modality (`droit de regard'), under which the central government would review provincial budgets following their adoption by the assemblies, and provides its comments to the provincial authorities without, however, giving a veto right.Review wouldbe limitedto obligatory expenditure items e. Third, and [in our view] the most literal interpretation o f article 171 o f the constitution, the provinces would only send their budgets to the central government for notification and consolidation within the central government budget. An accounting control carried out by the regional audit institutions could be If. added to the third option above, to provide an opinion on the economic viability o f the provincial budgets, notably on the balance between revenues and expenditures. The creation o f regional audit institutions would be a precondition inthis regard. 65. Inour view, a combination of the third and fourth control systems would be the closest to the spirit o f the constitution, though inview o f the serious capacity constraints that remain at provincial level the second option might also pass scrutiny. It is important, however, to emphasize that whatever the option selected, budgets will continue to be subject to ex post legality control. Finally, ex-post control by the Supreme Audit Institutions would also apply to provincial budgets. Legality control 66. On the control o f legality o f provincial and local government acts, there are equally two potential options inthe context o f the DRC constitution: a. Simple ex-post review, which would apply inany case b. Ex-post review combined with ex-ante review. The latter would be applicable only to the acts passed by ETDs, not by the provincial authorities, and would be limited to certain categories o f acts, as for instance: a) budget acts; b) organizational structures and; c) decisions related to the status of personnel 67. In terms of the options outlined above, control functions over the legal acts of ETDscould be assumed either bythe provinces or by the central government. inthe case o f the DRC, considering the size of the state and the potential cost o f establishing a review system managed by the central government, it would be preferable to move 27 towards a system where the provinces would take on this role. However, in view of the still weak capacities at this level, it would be desirable to postpone a transfer of these responsibilities for a certain period of time, to prevent an overload of the provincial institutions. 68. The scope and depth o f control and guidance systems, both in terms of financial and legal issues, is curtailed by the definition, in the constitution of the DRC as a decentralized unitary state. In line with international practice and norms, this limits the options to be explored in this regard. Inview of the still recent establishment o f elected provincial authorities, one option to be considered is the introduction of light control mechanisms, along the lines set out under the 3rd and 4' options under the financial control section, along with the definition of a transitory period duringwhich the 'droit de regard' would continue to apply. 28 5. Managingpublicinvestments:the `Caisse de PCrCquation' 69. The Constitution prescribes the creation o f an Equalization Fund (EF), the Caisse de Pkrkquation, which would absorb 10 percent o f national revenues. This Fund would only provide resources for investments in the provinces. Within the constitutional definition there i s a range o f options for the role o f the Equalization Fund and for its institutional model. 70. The EF provides the basis for a stronger element o f solidarity in the financial relations between the provinces and the central government. It should be noted that this, inpart, may compensate for the lack ofsolidarity that has so far emerged inthe proposals on retrocession. At the same time, however, it is difficult to imagine that the investments fundedby the EFcouldbe managedeffectively ifprovinces lack the basic means to build up their internalcapacity to manage these investments. 71. There are several options for the institutional model for the EF. However the Note argues in favour o f an institutional model that i s runby representatives o f boththe central State and the provinces, i s a reliable funding agency for useful infrastructure development and is able to buildconsensus between the provinces andthe State. 72. There are several models that can be considered to determine the eligibility criteria for accessing the EF's funds by the provinces. It is essential that these criteria are transparent, updated and open for scrutiny and that they are based on data that easily accessible given the scarcity o f quantitative data in DRC. The Note indicates that for the effective functioning o f the EF there is a need to develop a detailed planning and appraisal mechanism that will enable the provinces to prepare plans and investment proposals that produce the most viable investment initiatives and that allows the EF to select the most appropriate proposals. Moreover, the Note recommends several selective principles for the investments in the provinces such as: respecting the decentralization legal framework for competences, co-financing, programming, compatibility with the national investments programme, auditing. 73. As a funding agency, the EFwill have specific capacity demands, which will need to be met by a relatively lean organization that minimizes the overhead cost burdeno f the EF. The EF will be regularly audited and placed under democratic control in order to make sure it remains accountable. The Agency will have to develop legal and financial selection criteria. 74. The EF will also impose capacity demands on the provinces. It is expected that preparation o f public investment proposals that are submitted to the EF will need to meet these access criteria. At a minimumthese proposals will need to spell out the economic and financial case for the proposal. In addition, the proposals would need to look at the impact o f the proposal, for example on income levels o f the population or revenue generated for the province. To prepare such proposals will require certain capabilities in provincial government which at the moment do not yet exist. 29 75. The EF is a key element in the financial solidarity system between the State and the provinces and between the provinces. Therefore, it is important to put in place this institution inthe first phase o f the decentralization process. Inthis regard, the experience o f the EUinmanaging regional development funds could be particularly instructive. 30 6. The EducationSector 76. A final background note pays specific attention to the education sector. The education sector constitutes the main part o f the public sector wage bill (about 45 percent), as well as almost 50 percent o f public sector employment. At provincial level these levels are even higher; the proportion o f education wages in total wage bill expenditures in the provinces is likely to range between 57 and 85 percent (see table 9 below).At the same time, the Constitution foresees a significant decentralization o f the sector, with most functions in pre-school, primary and secondary education (apart from setting standards and curricula and inspection) to be transferred to provincial and ETD level. Table 13. Wage BillEstimatesper Province(FC mln, 2008, totalEducation) Province Total wage bill EPSP wage bill EPSP share of total Bandundu 27,368 23,224 85 Bas-Congo 12,708 10,119 80 kquateur 14,048 11,072 79 Kasa'i-Occidental 9,75 1 7,852 81 Kasa'i-Oriental 9,343 7,402 79 Katanga 13,352 10,055 75 Kinshasa 37,985 21,479 57 Maniema 3,355 2,759 82 Nord-Kivu 6,622 5,212 79 Orientale 11,498 9,203 80 6,618 5,425 82 Source: Ministbrede Budgetet MEPSP,Novembre2007. 77. This poses three main questions: first, the viability o f decentralization of functions in the light o f the large unfinished education reform agenda at central government level, second, the ability, both fiscal and capacity wise, o f provincial governments and ETDs to take on, inthe short term, the responsibilities assigned by the Constitution and, third, the need to align the organization o f the education sector with a decentralized service delivery modality. Unfinishedcentralgovernmentreforms 78. There are several ongoing initiatives to address the irregularities in the education sector. These include, in particular, the question o f the large number o f non-accredited teachers working inthe system, which is a result o f two separate developments. First, the inability (financially and organizationally) o f the central government to treat accreditation processes, leading to a backlog o f an estimated 50 000 cases. These teachers are currently 31 directly paid by parents through a system o f `primes' which the government is trying to eliminate. Second, a large number o f teachers in the post-conflict provinces are `nouvelles unit&', which means that not only are they not accredited, but also that their positions are not authorized by the central government (though they were authorized by the provincial authorities during the conflict period). It is important that the exact number and accreditatiodintegration processes are clearly defined before a transfer o f education competencies, or at least before expenditure responsibilities are transferred. If not, provinces will be faced with unknown levels o f expenditure assignment. Whereas this is currently also a problem at the level o f the central government, the problem would be seriously exacerbated if spread over 11provinces. 79. A second key issue is the management o f salary payments, which has been identified as a major issue inthe Public Expenditure Review (World Bank, forthcoming). The complex system o f salary payments is known to lead to significant `leakages' o f funds and less than full payment o f teachers. Whereas transferring responsibility to the provinces and ETDs may shorten the `delivery line' for salary payments, it should be noted that part o f the problems inthe salary payment system are due to weaknesses inthe control environment at provincial level, that would need to be addressed before salary payments are transferred. 80. Thirdly, the wage system itselfposes problems. Some o f the features o f this issue were discussed in section three above. Salaries are not harmonized across the country (in the current system there are three salary zones, i.e. Kinshasa, Lubumbashi andthe rest o f the country, but the differences between these are based on payment or non payment o f `primes' not o f differences in base pay levels), and consist in a large proportion o f a) allowances and indemnities and b) direct payments by parents. Discussions (August 2007) between the government and the trade unions have sought to address part o f these issues (integration o f part o f indemnities and allowances inbase pay, move towards two salary zones insteado f three, official elimination o fparent contributions. 81. Finally, reforms in the sector have barely started to have an impact on the issues with regard to accountability and performance o f the education system. Extensive systems o f additional payment by parents remain in place without adequate accountability. Oversight over schools and teacher performance remains limited without parent involvement. These are all issues that in due course will need to be addressed in the context o f decentralization, in particular with regard to the roles o f ETDs in the education sectors. 82. Whereas it might be impossible to address all the above issues before functions are transferred, an effort should at least be made to calculate the impact o f wage system reform and integration and accreditation o f teachers. This would imply that even if competencies be partially transferred, the central government will need, through an interimarrangement, to guarantee wage paymentsfor a periodo ftime. 32 Fiscal andInstitutionalCapacityinthe Provincesand ETDs 83. The organizational and fiscal capacity o f the provinces to take on the responsibilities assigned by the constitution is currently not in place. In terms o f organizational capacity, the issues in the education sector are a microcosm o f those discussed in the section on human resources above. There is no clarity as yet on the process for separating functions and resources between the centrally subordinated education departments are provincial level and the emerging education management structures at provincial level. Whereas for some aspects o f the system the situation is clear (Le. inspections well remain centrally subordinated) for others this is not. This makes provincial governments ill-equipped, at this point in time, to take on large scale responsibilities in the sector. It is unlikely that this issue will be solved quickly in the absence o f the required legislation. Therefore, unless satisfactory interim arrangements are put inplace, this i s likely to create serious organization and management problems in the education system. Aligningorganizationalmodels 84. The organizational model o f the education sector will require a major adaptation in order to comply with the new realities o f decentralization as envisaged by the Constitution. One important factor in this regard is that the state i s not the only, and not even the mainprovider o f education inDRC. The non-state education sector (mainly faith based schools) is the main provider o f these services. This will provide its own dynamic inthe decentralizationofthe education sector. 85. Another aspect to be considered is that the geographical dimension o f the education system does not necessarily coincide with the administrative territorial division. This will need to be adapted inorder to enable ETDs to play a major role inthe decentralized model for the education sector. 86. The organizational structure o f the education sector is very elaborate. It will need major changes in order to adapt to the new reality o f decentralization. Most effected are the departments for supervision and inspection o f the schools at lower levels. The role and nature o f administrative supervisionwill need to undergo fundamental changes inthe future. 87. A summary review o f key issues to be addressed to ensure a successful decentralization process in the education sector highlights the challenges faces by the DRC authorities. This confirms again the need for the process to be properly designed from a technical perspective, to prevent institutional and political turmoil that could affect social stability and put the viability o f the whole process indanger. The case study also shows the interlinkages (and mutual impact) between what remains a largely unfinished central level reform agenda and the challenges faces at provincial level. Whereas the decentralization process in some instances may help unblocking frozen reform issues at central level, in other aspects there is a clear risk o f currently 33 unresolvable issues at central level being transferred to provincial level and putting an undue burden on their currently limited fiscal and administrative capacity. This is an important point to be considered in making the final decisions on the phasing and programming o fthe decentralizationprocess. 34 7. Conclusions 88. The previous sections include a brief review o f the key issues outlined in a set o f five policy notes and briefs developed jointly by World Bank, EU and government experts. The key issues and recommendations that follow from the analysis are the following: i. Fiscalsolidarityshouldremainanimportantelementofthedebateon fiscal transfers, even if solutions have been identified that would assure the fiscal viability o f most provinces, a number o f questions *. remain on the modelthat need to be urgently answered; 11. A detailed assessment o f the cost o f competency transfers should be done urgently, as a rudimentary assessment for the education sector has already raised several important issues to be addressed before ... educationcompetencies are transferred; 111. The political and administrative infrastructure o f the provinces is starting to develop, with the political side o f provincial governments now in place and functioning in all provinces. This now needs to be followed by a rapid resolution o f issues related to the definition o f authority o f provincial authorities over current public and civil servants working at provincial level. There are a number o f modalities that could be used to separate the remaining central and provincial public servants, based on international practice, in particular in areas where the provinces have exclusive competencies. The definition o f an orderly process o f staff transfers, including a recruitment and attestation processes, i s a matter o f urgency, as is the design of, at the very least, interim arrangements at provincial level to define the mode o f cooperation between ministers and public servants subordinated to the central government; iv. The elaboration o f permanent consultation mechanisms between central and provincial governments is urgent. This should help prevent situations such as the one highlighted for the education system, where centrally managed negotiations between trade unions and the government have generated agreements that could impose serious fiscal burdens on provincial authorities (and effectively make most o f them unviable), if competencies would be transferred along with expenditure assignments in 2008. Such mechanisms should be used to define common positions between central and provincial authorities before such negotiations are initiated; v. It is essential that clarity is provided on the schedule o f the introduction o f some key reform measures at central level that are crucial to the ability o f provincial authorities to function effectively, including public service reforms (legal, wage system, etc.), crucial elements o f sector reforms (in particular in education) as well as the introduction o f key framework laws on public financial management. 35 89. Provided the above measures issues are carefully considered and appropriate pragmatic and practical solutions are addressed, decentralization would still hold significant potential benefits to the DRC. The dynamic environment at provincial level andthe rapid formation o fprovincial structures provides an indication o f the interest and commitment that exists at sub-national level to address the key issues facing the population. It is important that this momentum i s built upon to make decentralization work for DRC's citizens and to view decentralization as a potential `win-win' process, rather thanto think interms o f short term political objectives. 36