Report No. 42166-AF Afghanistan Building an Effective State Priorities for Public Administration Reform January 23, 2008 Poverty Reduction and Economic Management Sector Unit South Asia Region Document of the World Bank TABLE CONTENTS OF ACKNOWLEDGEMENTS ........................................................................................................................ i EXECUTIVE SUMMARY ...................................................................................................................... i11 Buildingan EffectiveCivil Service.......................................................................................................... iv Context and Prospects for Reform.......................................................................................................... 111 ... Improving Local Governance and Service Delivery.............................................................................. ..vi Making Government Accountable to the People...................................................................................... ix A Twelve-Point Agenda for Public Administration Reform.................................................................... xi I CONTEXTANDPROSPECTSFORREFORM . ............................................................................. 1 A. Introduction....................................................................................................................................... B. Rebuildingafter the Taliban.............................................................................................................. 1 C. 1 Lessons from Research and Country Experience.............................................................................. 7 D. LoolungAhead: Challenges and Constraints .................................................................................. 12 I1. BUILDINGANEFFECTIVE CIVIL SERVICE ....................................................................... 16 A. Introduction..................................................................................................................................... 16 B. The Civil Service's Scope ............................................................................................................... 18 C. The Civil Service's Competence..................................................................................................... 25 E. D. Civil Servants' Incentives ............................................................................................................... 29 Direction o f ReformActivities........................................................................................................ 36 I11. IMPROVINGLOCAL GOVERNANCEAND SERVICE DELIVERY ................................. 39 B. The Sub-National Institutional Structure......................................................................................... A. Introduction..................................................................................................................................... 39 40 D. Strategic Framework for Addressing the Sub-National Challenge ................................................. C. Key Features o fthe Sub-National Institutional Structure ............................................................... 42 E. 45 F. Specific Actions within this Strategic Framework.......................................................................... 48 InitialProcedural Steps ................................................................................................................... 56 I V . MAKING GOVERNMENTACCOUNTABLETO THE PEOPLE ........................................ 58 A. Introduction..................................................................................................................................... 58 B. BringingPublic Accountability to the PARagenda........................................................................ 59 D. Rethinkingthe Roles o fMediaand CSOs....................................................................................... 62 C. Improving the Effectiveness o f Parliament and the Judiciary......................................................... 60 E. Increasing Information Flows and Transparency ............................................................................ 67 V . A TWELVE-POINT AGENDA FORPUBLICADMINISTRATION REFORM ..................... -70 REFERENCES .......................................................................................................................................... 80 Box 1. 1: What do we meanby Public Administration Reform?................................................................. BOXES 2 17 Box 2.2: The Civil Service's Size ............................................................................................................. Box 2. 1: Compact Goals for an Effective Civil Service ........................................................................... 19 Box 2.3: Usingcommunities instead of government mechanisms to deliver education........................... 21 Box 2.4: Alternate Delivery o f Health Services inRwanda...................................................................... Box 2. 5: Creation ofthe IARCSC............................................................................................................. 22 24 Box 2. 6: Types ofTA and their perceived strengths and weaknesses ...................................................... 26 Box 2.7: Vulnerability to Corruption inAppointments............................................................................. Box 2. 8: Old and new grade pay structures............................................................................................... 33 32 Box 2.9: World Bank's support for Afghanistan's civil service reformagenda ........................................ 36 Box 2.10: Simplification o fVehicle RegistrationinKabul ....................................................................... 38 Box 3.2: Municipal Challenges inAfghanistan ........................................................................................ 52 Box 3. 1: Sub-National Challenges inTakhar Province............................................................................. 42 Box 3. 3: The Evolution of CDCs .............................................................................................................. 54 Box 3.4: The Role o fPRTs....................................................................................................................... 56 Box 4.2: The Constitutionand Media's Freedomo f Expression .............................................................. Box 4. 1:The 2005 Parliamentary Elections inAfghanistan ...................................................................... 59 63 Box 4.4: InnovativePublic Accountability Initiatives ............................................................................... Box 4. 3: The Universe o fAfghanistan's CSOs.......................................................................................... 65 68 FIGURES Figure 1.1: State Scope and Strength ............................................................................................................ 9 Figure 1.2: Service Delivery Framework for Afghanistan.......................................................................... 10 ACKNOWLEDGEMENTS This report was prepared by a team led by Ranjana Mukhejee and Mark Baird (consultant), under the overall guidance o f Ijaz Nabi. Major contributors to the report were Anne Tully and Robert Crown (consultant) on civil service reform, and Roland White and Sarah Lister (consultant) on sub-national governance and service delivery. The report also draws on background papers by Barnett Rubin (NYU) on the political context for reform, by Jalpa Pate1 and Scherezad Latif on reforms in the Ministry o f Education, by Saeed Parto (AREU) on priority sector reforms, by Monica Das Gupta on service delivery models, by Asger Christensen and Asta Olesen on the role o f CDCs, by Soraya Goga on municipalities, and by Lorenzo Delesguez (IWA) on public accountability. Bill Byrd, leading the preparation o f the parallel report on anti-cormption, provided valuable advice. Others providing inputs include Paul Banerjee and Homayun Sediq (MoF), Mithulina Chatterjee (UNDP), Birgit Hand, Nigel Coulson and AbdulRaoufZia. Administrative support was providedbyThelma RutledgeandJuliet Teodosio. The concept note for this report was reviewed inthe World Bank in October 2006 and with government and donor counterparts in Kabul in November 2006. The report has benefited from subsequent discussions in Kabul over the past year, as well as from comments on draft chapters by Rahela Sidiq Hashim, Zabihullah Sawayz and Wali Hamidzada (all o f the IARCSC), Paul Fishstein (AREU), Yasin Osmani (Office o f President), Satyendra Prasad (DfID),Barnett Rubin, Lubomira Beardsley, Sarah Cliffe, Tony Verheijen, b c k Stapenhurst, Mohini Mallhotra and Francesca Recanatini. However, the authors are solely responsible for any remaining errors or oversights. A draft o f this report was formally reviewed inKabul on October 28, 2007 (withm the World Bank) and October 29, 2007 (with government and donor counterparts). Both reviews were chaired by Alastair Mckechnie. The peer reviewers were Ahmed Mushahed (IARCSC), Shakti Sinha (UNAMA), Paul Fishstein (AREU), Sanjay Pradhan and Chstine Wallich. i 11 EXECUTIVESUMMARY Context and Prospects for Reform 1. Buildingan effective state -that can provide security and services to the people, while protecting their rights - has always been at the heart of the reconstruction effort in Afghanistan. A lot has been achieved over the past five years. Most importantly, the economy has grown strongly and social conditions have improved inmany parts o f the country. Some progress has also beenmade inbuildingup a more professional, merit-based civil service - which can start to take on the core functions o f a "lean state". However, many functions expected o f government are still performed by the international community or not performed at all. Realistically, this will remain the case for many years to come. Yet, despite the difficult political and security situation, it i s important to persevere with the longer-term task o f building an effective state - which can gradually take on more responsibility for Afghanistan's future. This report looks at the implications for public administration reform over the next 5-10 years, and proposes a priority agenda for action by the Government o f Afghanistan (GOA)andthe donor community. 2. The review o f international experience in Chapter Iconfirms that PAR i s difficult under the best o f conditions. InAfghanistan - where informal power relationships are stronger than formal government systems, and large parts o f the country are insecure - it is a daunting challenge indeed. With the urgent need to show progress on a broad development agenda, PAR sometimes seems like a luxury that will have to wait for more normal times. The trouble with this view i s that it leaves Afghanistan with a very weak state - that cannot deliver public services without large amounts o f external expertise. This inevitably weakens local ownership and leadership o fthe development agenda. So a balanced approach i s needed: one which finds innovative ways to improve service delivery as quickly as possible, while sustaining progress on PAR over the longer term. All government and donor agencies supporting PAR need to recognize it i s a long-term task, which requires sustained support for institutional change and capacity buildingover the next decade and beyond. 3. PAR in Afghanistan i s often equated with civil service reform. Indeed, civil service reform i s essential to build a professional, merit-based bureaucracy - which can deliver public services in an effective and efficient manner. But this should not be interpreted as solely buildinga strong, centralized bureaucracy inKabul. Much o f the reformeffort will have to be directed at deconcentrating line ministry authority to provinces and districts, and improving the capacity and coordinationo f government activities at the sub-national level. This i s where most services are delivered - and where the credibility o f government will be won or lost. Similarly, civil service reforms are unlikely to work unless there are ways for the Afghan people to express their demand for public services and hold the government accountable for results. Hence the importance o f broader reforms to build trust in the effectiveness o f parliament and the judiciary, and to involve civil society and communities indecisions affecting them. 4. Against this backdrop, the rest o f the report reviews the prospects and priorities for public administration reform inthree key areas: Building an effective civil service: The GOAis about to embark on a major reform o f the civil service pay and grade structure. This i s seen as important for attracting and retaining quality human resources. However, it will be a challenge to implement a merit-based system in the face o f widespread patronage and corruption. Furthermore, to lead to improved civil service performance and service delivery, higher pay will have to be linked to other reforms in ministry roles, functions, structures and processes. Chapter IIlooks at how such a program might be implemented: the sequencing o f reform actions, the role o f the IARCSC and line ministries, and ways to monitor progress. 111 Improving local governance and service delivery: There has been growing debate over the past year, within the GOAand the donor community, about the role o f sub-national administration. While Afghanistan i s a unitary state, the Constitutionprovides flexibility to delegate functions to provincial and district government units. Chapter IIIlooks at the major challenges facing sub- national administration in Afghanistan, their impact on governance and service delivery at the local level, and the implications for public administration reform. Makinggovernment accountableto thepeople: Some o fthe checks and balances on government power come fi-om internal controls, including through better public financial management. However, most successful states recognize the importance o f external controls on executive power, formally through the judiciary and parliament, and informally through the role o f civil society and the media. Chapter IV explores what can be done to make government more accountable to the people, through better information flows and transparency, as well as institutional development o f parliament and thejudiciary. Building an Effective Civil Service 5. Civil service r e f o m is vital for sustainingthe state inAfghanistan. A civil service, that serves the public's interest rather than its own, is important for Government's credibility within the country. An effective civil service can equally reassure donors that their support i s being credibly spent. 6. The Afghan Government has controlled the civil service's size through considerable discipline in establishment control. Since the start o f reconstruction, it has reduced the number o f ministries from 30 to 25. The civil service pyramid i s characterized by a broad base and narrow-tip, respectively explained by the pressure on government to generate employment and severe scarcity o f professionals inthe civil service and society at large. H o w the government chooses to deliver services affects its size and shape now and inthe future. Consistent with the EducationMinistry'sretentionof most o f the service delivery functions, it i s the government's biggest employer. The Ministry o f Public Health, having delegated the delivery o f its Basic Package o f Health Services largely to NGOs, has much fewer staff. Drawing upon alternate service delivery models being used in other developing countries, Afghanistan could to select service delivery models most appropriate for its different sectors. 7. Functional overlap between organizations complicates the scope o f government. For example, both the Ministry o f Urban Development and Kabul Municipality claim responsibility for planning, zoning, and building regulations within the city. Equally, gaps in agency coordination exist in some critical areas o f government. Although policy making i s the most important function o f the center o f government, there exists no institutional arrangement to coordinate policy development across ministries prior to their submission to Cabinet for consideration. Greater synchronization between the Ministry o f Interior (police) and the Attorney General's office (prosecutors) will help improve rule o f law. 8. Similar to the structure o f government as a whole, individual organizations need restructuring as they had become very informal during the conflict years. The Priority Reform and Restructuring (PRR) program was the principle vehicle for restructuring and rationalizing individual government organizations. Ministries that underwent restructuring started to look like modem ministries, and merit- basedrecruitmentwas introduced. 9. IARCSC, the organization mandated to lead civil service reform and organizational restructuring, itself has conflicting roles and functions. Two distinctly separate personnel functions are merged within IARCSC: a ministry in charge o f personnel, and a traditional Public Service Commission to provide the necessary check and balance between government and employees. Having to perform all aspects o f iv Afghanistan's civil service management has created a serious capacity crisis in the Commission. Meanwhile, neither practice nor the provisions o f the Basic Civil Service Law lay down the Commission's independence. The current arrangements create conflicts even within the IARCSC's own structure. 10. Because the IARCSC is responsible for leading civil service reform, it needs to start serious monitoring o f reform's progress against the Compact's benchmarks. This will enable the Commission to point out to the Cabinet and line ministries those areas where more concerted effort i s needed. It will also enable the Commission to highlight its own achievements (such as the number o f civil servants trained, the number o f high level recruitments made by the Appointments Board etc) to partly offset leaders' unhappiness with the IARCSC's role inreducingpatronage. 11. Reforming the structure o f government and its various ministries and agencies provides the necessary framework within which the civil service can function, but the best o f structures would not result in good service delivery without competent civil servants working in these organizations. This required level o f competence among civil servants was virtually nonexistent in 2001. International technical assistance flowed into the vacuum at a rate that government was largely unable to plan for, or effectively control and coordinate. Some leaders have observed that there i s little to show for the estimated U S $1.6 billion that has been spent on technical assistance since 2002. The government has since been trying to balance initiatives on three fronts: managing technical assistance; conducting in- house training to strengthen current employees; and planning for longer term improvements in competence through schooling, training institutes and higher education. An Inter-ministerial Capacity Development Committee has been established to monitor and compile information on various capacity buildinginterventions. 12. Even competent civil servants, employed in well-structured organizations do not comprise an effective civil service unless their institutional environment motivates their accountability and high level o f performance. Informality in government administration during the conflict years distorted whatever civil servants' incentive structure might have existed prior to the conflict. As a result, at the start o f reconstruction in Afghanistan, there was very little discipline or competence within the civil service, and no highlevel consensus that it should be politically neutral. InFebruary 2005, the President and Cabinet approved a Basic Civil Service Law. Key regulations and procedures to implement the law still need to be finalized; otherwise the law will remain only a lofty ideal. 13. Merit basis in appointments is at the heart o f effective civil service management. But the volatility o f decades o f conflict eroded the limited trust there had been with merit-based recruitment as competing political factions needed loyal and known representatives within government. The lack o f transparency in the recruitment process left staff dissatisfied and demoralized, and contributed to the overall capacity depletion o f the civil service. With the establishment o f an Appointments Board in the IARCSC to process recruitment at high levels, there has been significant progress in this area. But the Appointments Board still faces considerable pressures from ministers and other leaders to appoint specific candidates, and the process and outcomes o f the merit system itself have been subject to considerable criticism across government. Some lacunae arise from the Appointments Board's insufficient technical expertise. 14. The inherited grade-pay structure in Afghanistan did not provide adequate performance incentives for civil servants. It emphasized longevity rather than competence and responsibility, and focused managers and individuals on the inputs (such and qualifications and seniority) and process rather than outputs. Pay scales for qualified staff are completely out o f line with prevailing market conditions and severely compressed. These created incentives for donors to intervene with a wide range o f ad hoc top-ups and supplements, providing further distortions inthe incentives system. The new pay and grade V structure, approved by Cabinet on June 18, 2007 i s intended to end the practice o f personal grades, make grades consistent withjob content, reduce the need for ad hoc pay arrangements, and allow government to substantially raise salaries at the top end so as to recruit and retain experienced and effective staff. The proposed salary scales, ranging from $100 to $650 per month, are significantly higher than current senior salaries, but could be still be inadequate in cases to attract the required slulls and compete with the salaries offered by donors. Schemes such as the new Management Capacity Program will be required to deal with such extra-normal appointments over the medium-term. 15. The new structure's test will lie in its implementation, and whether the higher pay will provide the right incentives for improved performance and service delivery. The introduction o f regrading and new pay scales has been postponed by a year to SY 1387. This will require compressing the implementationtimetable if its synchronization with the MTFF i s to be maintained, and covering 25% o f civil servants inone year. However, the IARCSC has not yet discussed any implementationplan with line ministries and donor partners. With elections scheduled in Afghanistan in less than two years from now, the same pressures that led to units making superficial PRR-restructuring plans, abandoning the rigorous recruitment process, and migrating all existing staff to higher pay scalesnould equally derail any logical and systematic roll-out o f the new scheme. IARCSC and line ministries have limited competence to ensure that structures, job descriptions and trained staff are in place at the right time in ministries, which can then implement the new structure ina systematic way. . ImprovingLocalGovernance and Service Delivery 16. Civil service reforms will inevitably impact on sub-national administration, as ministries devolve responsibilities and strengthen their capacity at the provincial and district levels. However, there are many other formal and informal power structures that influence the way government works at the sub- national level. Attention to these sub-national issues has risen markedly over the past year - reflecting concerns about the effectiveness o f local investment and service delivery programs, and the growing influence o f militia commanders and criminal elements in some parts o f the country. This has brought home the pointthat the key question surrounding the character o f the state in Afghanistan i s not that o f a strong center versus a strong periphery. Rather it i s whether the current de jure (and hollow) state at the periphery can become a de facto state at the periphery; whether robust and institutionalized systems o f government can be expanded and deepened within the provinces andbelow. 17. The analysis in Chapter 111identifies five important features o f the "sub-national problem" in AfghaniStan: First, the overall structure is characterized by significant systemic contradiction. On the one hand, formal functional and budget authority for the delivery of most key services in the provinces i s held by highly centralized line ministries, which work in vertically integrated silos withweak linkages between them. Onthe other hand, this highly centralized lineministry system co-exists with the Provincial Governor (PG) system, which allows the PGs to intervene in the affairs o f line ministries and other agencies (such as municipalities) through a number o f informal channels for influencing staffing and resource decisions. Second, even given the choice of a centralized inter-governmental structurefor Afghanistan, existing institutional arrangements are often highly inefficient. In particular, line ministries tend to be over-centralized, with their central offices inKabul retaining functions which could be more efficiently performed at the provincial level (teacher recruitment in the case o f education, for example). To some extent, this problem is not unrelated to the one discussed above. So long vi as line ministries perceive that deconcentration o f authority to their Provincial Departments i s likely to make them subject to greater influenceby the PGs, they are unlikely to do so. 0 Third, the current system is both asymmetric and inequitable. Different institutional models being used in different provinces (due to accidents o f history or varying donor practices) are likely to introduce significant strains and tensions as the sub-national system consolidates and attempts to cohere. An even more pressing set o f problems arises from the inequitable distribution o f resources across sub-national jurisdictions (in sectors such as education and health). Afghanistan does not as yet have strong enough mechanisms for citizen voice in setting local priorities. And large aid flows further distort resource allocations between and within provinces. 0 Fourth, the most autonomous elements of the sub-national system of governance and service delivery-the municipalities and CDCs-face issues of their own. Untilmunicipal elections are held, the potential for municipalities to deliver accountable local government intheir areas will be highly circumscribed. In addition, the fiscal and functional framework governing them i s both unclear and constraining o f effective and efficient service delivery. CDCs face questions about their future funding, and the possible extension o f their role from planning and implementing community projects to taking on broader governance functions at the local level. 0 Finally, like the rest of the Afghan state, the entire sub-national structure is afflicted by the sorts of problems characteristic of fragile states: severe human resource weaknesses, an absence o f properly functioning operational systems, shortages o f equipment, and spares supporting infrastructure (such as power and phone systems) necessary to get things functioning properly. Afghanistan i s particularly badly affected by these, and they will inevitably impact on any effort to reformand strengthenthe sub-national system. 18. The political climate in Afghanistan precludes any significant reform to the underlying inter- governmental structure inthe short to mediumterm. The Constitutionrefers to "preserving the principles o f centralism", and it i s also clear that the current political establishment is firmly committed to the current centralized structure and has no appetite for a substantial political devolution to sub-national levels. The basic intent o f any sub-national strategy, therefore, must be to improve the functioning o f the current system, rather than to restructure it. True, this structure has intrinsic weaknesses from the service delivery and governance perspectives - especially in a large, poorly-linked country with a thinly spread population. However, not withstanding these difficulties, the emerging structure does allow for improved performance with appropriately focused reforms. These reforms should address four broad challenges: Challenge 1: Establish an overall direction for sub-national reform which, while building on the current system, diminishes the existing levels of systemic contradiction. This direction should be based on the broad allocation o f roles and responsibilities outlined inFigure 3.1. Two directions o f reform stand out: (a) in order to reduce systemic contradiction, strengthen and consolidate the role o f the PG's Office inthe areas o f coordination and planning, while limiting its functional, budget and operational authority to deliver services, which falls under the Provincial Departments; and (b) in order to enhance accountability, strengthen the oversight and supervisory role o fthe Provincial Councils, to provide checks and balances on the activities o fthe delivery agencies. Challenge 2: Strengthen theperformance of the main service delivery agencies by improving their organizational structures. One key, cross-cutting activity inthis area i s for line ministries to ensure that functions are properly divided between their central offices and Provincial Departments. Ina number o f cases this i s likely to involve the deconcentration o f various powers and functions from the central to provincial level. Infact, insome cases this is already underway; vii the Ministry o f Education, for example, i s currently in the process o f deconcentrating teacher recruitment to the provincial level. The pace o f such deconcentration will be determined, among other things, by capacity constraints and appropriate service delivery models indifferent sectors. For example, deconcentration may be o f less relevance inthe health sector, which relies more on contracting out to NGOs. Challenge 3: Provide sufficient resourcesfor the key organs of sub-national governance and service delivery to play their ascribed roles. For example, CDCs face a fundamental threat as their access to funding under NSP i s due to end, and without the funds necessary for their development activities, they will wither and die. At the provincial level, organs such as the Provincial Councils need to gain access to sufficient funds to conduct their operations as envisaged under the recently amended law; while at the municipal level, an overall fiscal framework which provides a rational, equitable and affordable solution to the fiscal gap they face needs to be developed. A further dimension o f the "fiscal problem" is moving towards a more equitable distribution o f resources across the country. Arguably, this i s a second generation problem which can only really be addressed once greater stability and predictability are established in vertical and sectoral flows, including those funded by aid. But it cannot be neglected indefinitely. Challenge 4: Enhance the operational performance of all sub-national organs by providing them with the systems and capacities they need. In some areas (e.g., some PG Offices) substantial progress i s being made; in others (e.g., Provincial Councils) it has really only just ' begun. Needs obviously vary across organizations, but certain generic themes can be identified (training, IT, offices, communications etc.). Moreover, it should be possible to handle some o f these needs on a modular basis, i.e., to define a package o f "basic needs" equipment for Provincial Councils, for example, and to roll this out to PCs across the country. 19. Inorder to addressthese challenges, anumberofconcreteinitiativesareproposedinSectionEof Chapter III.Some o f these proposals are new but many leverage o f f activities which are already inplace or are planned. It should be stressed that the proposals made cover a wide area and there is no expectation that they will - or should-all be embarked upon at once. Rather, incremental movement will be required on as many fronts as is feasible at any one time. However, to bringsome order to the process, there i s an urgent need for a clear policy framework for the desired institutional structure o f the sub- national system, and a strategy to guide actions to realize it. While the Constitution and other legislation provide a legal framework for a number o f sub-national institutions, such as Provincial and Municipal Councils, this tends to be broad and ambiguous, and often begs more questions than it answers about who should be doing (and i s accountable for) what, the resource flows that are necessary for the effective exercise o fpowers and functions, and the implementing mechanisms that needto be put inplace. 20. To this end, the recent decision to establish the Independent Directorate o f Local Governance (IDLG), reporting directly to the President, is a welcome move. The IDLGi s essentially a merger o f the structures o f civil administration under the Ministry o f Interior and provincial relations under the Office o f Administrative Affairs (OM). Its basic role i s to supervise the affairs related to Provincial and District Governors, as well as municipalities (except Kabul). To be effective, the IDLG will need consistent political support from the President. It will also have to build up political credibility with the powerful and disparate interests workmg at the sub-national level, and the technical capacity to tackle complex institutional and policy issues. To support this effort, consideration should be given to forming an Inter-Ministerial Committee, chaired by the President or Vice President, with IDLG acting as the secretariat, to establish a framework for sub-national strengthening and reform. Inthe short to medium term, this should focus on providing clear direction for the evolution o f the sub-national system. In the longer term, it could oversee a more thoroughgoing policy exercise, culminating in a new Local V l l l ... Government Law (a process which began some time ago, but which now appears to be in a state o f suspended animation). 21. The IDLGwarrants strong support from the donor community. Inthe past, donor activities at the sub-national level have been poorly coordinated. The extensive funding, which flows to sub-national initiatives outside government systems, has compounded this problem. Recent efforts to establish a donor worlung group on sub-national governance are therefore welcome. This provides a useful forum for formulating common donor views on sub-national issues and approaches, which should be fed into the Sub-National Governance Donor and Government Worlung Group under the ANDs mechanism. The IDLG shouldprobably take over the chairmanship o f this group from the IARCSC. One o f its first tasks will be to set appropriate benchmarks for developing a clear policy framework for the sub-national system, and a strategy for guidingreforms, along the lines mentioned above. Making Government Accountable to the People 22. Thirty three years after the last election in Afghanistan, 6 million citizens went to the polls in September 2005 to elect the lower house o f the National Assembly (Wolesi Jirgu) and 34 Provincial Councils. Citizens have other indirect means o f holding the executive to account. Restraints on the executive are exerted formally by the National Assembly (NA), Control and Audit Office, and independent agencies at arm's length from government. However, these organizations and individuals are not generallytrustedby citizens. Afghan media and civil society organizations (CSOs) comprise a less formally organized oversight mechanism. 23. Public accountability i s the least advanced among the three elements o f Afghanistan's PAR strategy. The first Constitution o f 1923 did not introduce formal checks and balances on the King's powers. The 1964 Constitution, while giving the status o f a Constitutional body to the traditional Loyu Jirgu, also gave the Kingthe power to dissolve it. Inthe 1977 Constitution, the Party was the guarantor o f power and all branches were accountable to the party, including the President. Public accountability also has had a slow start in the reconstruction. The extreme `capacity deficit' inevitably ignored the `accountability deficit' during the early days o f reconstruction. Humanitarian and reconstruction needs were so immense and immediate, and government's capacity so sparse relative to the challenge, that the objective o f donor-assisted reconstruction was to first get the system back on its feet before fine tuning its performance. The security deficit-like the capacity deficit-also ignored the accountability deficit. Donors' reconstruction priorities were short-term, but long-term effort i s required to build institutionso f public accountability. 24. The National Assembly has started asserting its role o f oversight on executive power. It rejected some Presidential nominees for cabinet, and for the Chief and judges o f the Supreme Court. However, parliamentary scrutiny o f the government's budget and accounts has been impeded by NA members' unfamiliarity with presentation o f public financial management information. Passing o f the SY 1385 and SY 1386 budgets by the NA were delayed by several weeks. Although the NA receives the CAO's annual report o f audit o f government accounts, it has no established system o f scrutiny of public expenditure and response to these audit reports. Parliamentary effectiveness is reduced by the absence o f strong and effective political parties. Third party groups form the balance between pro and anti- government factions. Without party discipline to keep members in the house, many sessions have to be canceled for lack o f quorum. An adversarial relation sometimes develops between the NA and the executive. ix 25. Justice institutions are probably the least developed among formal oversight organizations. Stronger legislative capacity and skills inlegislative methods and techniques will be required in the NA, the Cabinet and the Ministry o f Justice; and greater coordination between the Cabinet and the NA. Customary law prevails widely, and most disputes are settled innon-state forums, but customary law and informaljustice systems are biased against women, who are not represented. At the same time, confidence inthe formaljustice institutions is low, and strengthening the sovereignty and integrity o fthe Afghanistan courts will require transparency of courts' operations and procedures, uniform application o f law with predictable and credible decision-making, and courts' capacity to deliver services and manage resources. The role o f the judiciary, especially the Supreme Court, in enforcing constitutionality o f laws and regulations and reviewing legality o f administrative decisions needs to be strengthened. 26. The Government o f Afghanistan has taken several high-level steps to fight corruption: Constitutional provision, creation o f the General Independent Administration against Corruption, adoption o f the Anticorruption Roadmap, and ratification o f the UN Convention Against Corruption. These ideals are not reflected in ground reality. Drugs and unprecedented large inflows o f international assistance are dominant sources o f corruption. Conhsion stems from lack o f clarity o f what corruption actually means in the Afghan context, blurred lines o f responsibility among agencies responsible for fightingcorruption, and the large number o funcoordinated complaintmechanisms at all levels. 27. Recent tensions between the media and government are evidence that the media has started playing its role inpublic accountability. The new media law has provisions that can be potentially used to restrict the media's freedom through classifying some material as prohibited through the constitutionally mandated supremacy o f Afghanistan's religious beliefs over other laws. As the media acquires capacity to play the public accountability role, the government i s also adjusting to having its actions being scrutinized by the media. There have been stand-offs between the executive and media over allegations o f misquoting and corruption. The concept o f investigative journalism still has not taken root in large parts o f the media and government. Accountability o f government institutions has received marginal media coverage, and more than half of published information was attributed to government sources. Considering this very low base from which media development started duringreconstruction, its growth involume is impressive. However, the printmedia is still a long way from financial independence. A majority o fthe population depends on radio for news and information. 28. There i s little contact and communication between media and civil society organizations. The media ignores thousands o f shuras, associations and NGOs as news sources; while these groups also do not use the media to voice their concerns. Afghanistan had little or no tradition o f formalized community participation in political decision-making or development planning, either at the national or local level. Shurus' top-down model and patriarchal structure impedes efforts to include female members-thus impeding the whole community's involvement indiscussion and debate on civil society. 29. The current role o f most Afghan CSOs i s more that o f project implementer than defender o f public accountability. Demands o f the donor community and their emergency aid programs led to relocation o f Afghan NGOs from Pakistan to Kabul, and mushrooming o f several new Afghan NGOs all over the country. They have carved their own niche indelivering the very much-needed humanitarianaid throughout the country while following international organizations' policies and practices. Arising partly from the general population's hstration and jealousy for NGOs' access to resources, many hold a deep mistrust for "NGOs," and perceive them as a foreign phenomenon that has gained control over the money meant for Afghans. CSOs' capacity to play the role o f monitoring government i s not fully developed. Few CSOs themselves have the capacity to analyze the reports and inform the public. CSOs rarely lobby the National Assembly or the Executive. Security conditions not only threaten the opportunities for international NGOs to assist in developing Afghan civil society, but are also diminishing the ability o f local NGOs to enhance their own capacity and legitimacy. Local perceptions about association with X internationals, particularly the Provisional Reconstruction Teams, are adversely affecting the security o f NGOs. 30. There i s no clear answer as to which amongst the current CSOs organizations has the most potential for being effective participants in local level governance-in district level planning and prioritization, as well as monitoring o f services. There i s a large gap between the mandate of a traditional community shura and a local community development association. Religious leaders and networks have always played a prominent role as civil society forces inAfghanistan, but the question is how the Taliban experience may have affected traditional Islamic leaders' potential as civil society forces inAfghanistan. CDCs, being elected bodies and having more women's representation than traditional shuras, could potentially evolve into district level governance organizations, but could potentially lose the people's voice ifthey become part o f the establishment. While the NSP has delivered actual results on the ground, the advocacy of those with political stakes in the NSP, and NGOs that functioned as implementing partners may havebeen overstated. It also remains to be tested how CDCs perform outside their currently limited remit o f planning and implementing small infrastructure projects, and with less hding than provided under NSP. 31. Timely and reliable information i s essential for public accountability. Weak communications, the steep learning curve o f parliamentarians, and relatively undeveloped CSOs make it quite impractical to expect that either parliamentarians or civil society groups can even demand the relevant information, let alone analyze it to hold government to account. Therefore, rather than waiting for demand to emerge, government and donors should take the initiative to publicize their activities. This could include, for example, information on budget preparation and execution, public financial management and procurement, and project objectives, spending and achievements. Special attention should be given to monitoring the outcomes o f government programs, so that the public can see whether they're getting value for their money and the government can make mid-course corrections when programs fail to achieve their stated objectives. Better information flows - even among government organizations - can also bolster anti-corruption efforts. A Twelve-PointAgendafor PublicAdministrationReform 32. This report has made the case that public administration reform in Afghanistan is both very important and very difficult: very important because it will provide the Government o f Afghanistan with the capacity to respond to popular demands for good governance and improved service delivery; very difficult, because o f the inherent challenges o f institutional change, especially in the current political and security situation in Afghanistan. PAR in Afghanistan i s nothing less than rebuilding an effective state after more than two decades o f conflict and administrative decay. 33. Afghanistan's experience over the past five years, as well as lessons from other fragile states, provide some clues about how to go about public administration reform. The resulting challenges, and possible responses, have been summarized above. Chapter V o f this report attempts to distil the main points into a twelve-point agenda. There i s nothing magical or immutable about this list. It i s simply intended to help shape the PAR debate within the GOA and with the donor community - with the objective o f speeding up progress on reforms and realizing tangible benefits for the Afghan people over the longer term. xi (1) Recognize that PAR is a long-term task -which will requirepersistent effort over time, as well as innovative approaches to improve service delivery in the short and medium terms. 34. It will inevitably take time to build up the capacity and credibility o f the state in Afghanistan. That i s why it i s important to initially limit the role o f the state to core functions in line with existing capacity. This is consistent with the government's commitment to a "lean state" and the use o f non- governmental delivery mechanisms (such as in health). Innovations in service delivery should be evaluated carefully over time - to help define the appropriate role o f the state in different sectors and hence the priorities for PAR in different line ministries. This bottom-up approach should constantly challenge the top-down directions o f PAR. The real litmus test must be: how will these reforms contribute to better governance and service delivery? 35. In many ways, the opportunistic approach adopted under PRR is still very much relevant to Afghanistan today. It i s impossible to implement a broad-based, long-term agenda in a comprehensive systematic way - especially in a country like Afghanistan where capacity i s very weak and the government has limited reach outside Kabul. Rather the framework for reform must be comprehensive and coherent, and short-term actions must be consistent with that framework. At any point intime, there mustbe a clear idea ofpriority areasrequiringattention, andthose which canbe sortedout later. (2) Providestrongpolitical leadershipand effective coordinationto implement thePAR agenda. 36. Because PAR i s a broad-based effort, it must be led from the top with effective coordination across institutions. This leads to the following recommendations: 0 Strong leadership i s needed from the President -to provide a clear message about the direction o f reform and to manage conflicting political interests. The President should personally take responsibility for explaining the importance o f the reforms to Parliament and the public at large, and also reporting on the results achieved. 0 Important issues need to be discussed and decided at Cabinet level (the Cabinet's PAR Sub- committee provides a convenient vehicle) to build a consensus view even when different interests are being advocated by different agencies. Implementing agencies should also report back to Cabinet on a regular basis on progress made and issues to be addressed. Instead o f attempting to implement reforms in all ministries by itself, IARCSC needs to increasingly focus on guiding and overseeing ministries' implementation, while allowing ministries take responsibilityo f their ownrespective reforms. (3) Ensure that donor support is adequate and consistent with the PAR framework set by the Governmentof Afghanistan. 37. Although Afghanistan has receivedlarge amounts o f development assistance since the fall o f the Taliban, very little o f this has been directed specifically at the PAR agenda. Indeed, it i s hard to know exactly how much assistance has been received for PAR, given the different categorizations used by different donors, and the difficulty o f separating PAR components from projects with broader objectives. Getting a better picture o f donor support for different parts o f the PAR agenda would be useful in itself. Butevenmore important is effective coordinationo fdonor assistance bythe Government o fAfghanistan, inline with its ownPARpriorities. This should be done within the overall framework providedbythe I- ANDs and the Afghanistan Compact. Responsibility for donor coordination should then be delegated to xii the relevant government agency: for example, IARCSC for civil service reform and IDLG for sub- national reform. 38. Most o f the donor support for PAR i s in the form o f technical assistance. Initially, many had believed that the gap-filling need for civil servants' competence would last 2-3 years, and that short-term imported consultants would simultaneously train and coach their replacements while performing in-line functions. However, these assumptions proved to be overly optimistic. Furthermore, the higher salaries paid to externally-funded consultants and advisers have attracted some o f the best talent away from the government and bid up the cost o f scarce talent in the country. Therefore, in the future, more attention will have to be given to the objectives o f TA, with a clearer distinction between gap filling and capacity development. Donor funding o f salary top-ups for civil servants will also have to be carefully managed and phased out over time. (4) Match higher salariesfrom the new pay-grade scheme with other civil service reforms - to improvegovernmentperformance and service delivery. 39. The new grade structure's purpose i s to bring formality into civil service management by reducing inconsistencies and scope for arbitrary actions, while the new pay scales are intended to attract, retain and motivate well-qualified staff. Even though the new pays scales are significantly higher than the existing ones, they still cannot match what donors pay; and upon receiving higher salaries, civil servants will not automatically start performing better. The following cautions are advised inthe implementation o f the new grade-pay structure: 0 Draw upon the lessons o f PRR. The same issues that overcame the earlier exercise i.e., allowing higher pay scales to all staff could also engulf implementation of the new scheme. Therefore, IARCSC needs to prepare job descriptions carefully, and line ministries need to closely match incumbents' slulls to thesejob descriptions. Unlike the PRR exercise, line ministries shouldbe in charge o f their own implementationwhile IARCSC plays a guidance and oversight role. Recognize that even within this clearly defined role, IARCSC i s undertalung a task o f magnitude and complexity that it has never encountered before. Strengthen its management capacity, and provide it highlevel political support. 0 The "second civil service'' i s directly paid for by donors. Because o f the continuing scarcity o f qualified human resources, donors are likely to continue funding salary top-ups to civil servants from project budgets, thus rendering even the new pay scales inadequate to draw necessary slulls into the civil service. (5) Re-examine the IARCSC's role - tofocus on corefunctions and remove conflicts of interest within its structure. 40. The Commission's mandate both overloads it and creates conflicts o f interest. A thorough restructuring of the Commission may not be possible in the short term, and especially not now as it embarks upon the huge challenge o f leading the implementation o f the new grade pay structure. But some short-term steps could alleviate the problem: 0 To demonstrate the IARCSC's "independence," the Cabinet Sub-Committee's and PAR Steering Committee's supervision o f the Commission could be explicitly limited to its policy-making functions, and excluded from appointments and appeals. ... Xlll 0 The Appeals Board and the Civil Service Institute could be distanced from the Commission's own line structure. Distancing the Appeals Board will have the advantage o f upholding its credibility with civil servants whom it is meant to serve. Distancing the Civil Service Institute, currently buried three layers deep within the Commission's structure, will allow it to negotiate with donors for their support. Clarify the respective roles and responsibilities o f the Office o f Administrative Affairs (OAA) and IARCSC in civil service management, and especially in finalizing ministries' tashkeel (establishment). Donors that support civil service management could design their projects so that IARCSC's role is more o f a facilitator than project implementer. This will not only reduce strain on an already overloaded IARCSC, it will allow line ministriesto take control o f their own human and financial management functions as i s intended inthe Compact and I-ANDs. (6) Monitor progress on PAR - and be willing to rethink reforms when they're not meeting their goals. 41. The Government and IARCSC now have a systematic and well-thought out approach to PAR'S first element: civil service reform. The Compact benchmarks make clear what is to be achieved; while the Government's PAR Strategy, the ANDs-assisted sector strategy, and the costed civil service implementation plan have formulated how this can be achieved over the next three years. Thus, IARCSC needs to establish a monitoring system that regularly produces findings that are judged valuable by ministries, the PAR Steering Committee, cabinet and donors. It also needs to monitor line agencies' compliance with government-wide rules and regulations. IARCSC needs to become the powerful champion within government that generates line ministries' demand for monitoring civil service reform. Because o f ministries' lack o f awareness, their demand to understand how well or not they are performing their own restructuring and personnel management i s weak. IARCSC needs to make available the tools and their potential uses, and also create incentives for ministriesto monitor. The structural arrangements to monitor PAR should be clarified and strengthened. The current arrangements require one line unit o f the Commission to judge the performance o f the whole Commission and its Chairman's leadership. Also, it does not have the profile to influence other ministries, or to attract donor resources. The monitoring system should not be over-engineered; rather the emphasis should be on reliable data systems. What is required i s a set o f intermediate benchmarks leading up to the Compact's targets. The need for ministries to report on their compliance with new personnel guidelines and regulations could be reinforced through an appropriate government regulation andprocedures. (7) Develop a clearpolicy framework for the sub-national system - and a strategyfor guiding sub- national reforms. 42. Without a clear policy framework, initiatives to improve sub-national governance and service delivery will remainpiecemeal and often at odds with each other. Two directions of reform stand out: Inorder to reduce systemic contradiction, strengthen and consolidate the role of the Provincial Governor's Office inthe areas o fplanning and coordination, while limitingits authority to deliver services, which falls under the Provincial Departments. xiv Inordertoenhance accountability, strengthentheoversight andsupervisoryroleoftheProvincial Councils, to provide checks and balances on the activities o fthe delivery agencies. 43. The recent decision to improve the central oversight o f sub-national issues with the establishment o f the IDLGi s a welcome move. Next steps on sub-national issues should aim to: Buildupthe political credibility andtechnical capacity o fIDLG. To this end, an Inter-Ministerial Committee, chaired by the President or Vice President, with IDLG acting as the secretariat, could be formed to establish a framework for sub-national strengthening and reform. Buildon recent efforts to improve donor coordination. Inparticular, the IDLG should probably take over chairmanship o f the Sub-National Governance Donor and Government Workmg Group under the ANDs mechanism. One o f the first tasks o f this group should be to set appropriate benchmarks for developing the sub-national policy framework and reform strategy. (8) Strengthen the performance of the main service delivery institutions by improving their organizational structures-andproviding adequateresources to do theirjob. 44. Line ministries will have to ensure that their functions are properly divided between their central and provincial offices. These decisions should be built into the organizational plans being prepared by line ministries as part o fthe P W A R process, and guidedby the following considerations: 0 Any deconcentration process should be crafted flexibly enough to allow different ministries to implement service delivery models which are most conducive to success intheir sectors. 0 Ministries should consider asymmetric approaches: experimenting with different degrees o f deconcentration with different functions. The extent to which line ministries deconcentrate will need to be moderated by the degree to which this risks capture o fProvincial Departments by Provincial Governors. 45. Improvements in organizational structures need to be matched by the provision o f adequate resources for key organs o f sub-national governance and service delivery to play their roles. Initially, this means meeting basic needs for training, IT, offices, communications etc. Over the longer term, it means moving towards a more equitable distribution o f fiscal resources across the country. (9) Address theparticular issuesfacing municipalities and CDCs: the most autonomous elements of the sub-national system of governanceand service delivery. 46. Municipalities have been established under the law, as important and self-sufficient local service delivery and governance entities. To improve their effectiveness, the following measures are needed: 0 Enact new legislation which clarifies the relationship o f municipalities to the central and regional governing entities, while protecting their self-reliance. Inter alia, this should provide them with solid expenditure and revenue assignments, and enhance their capacity for providing local services sustainably. 0 Formulate a comprehensive financing framework for municipalities which clarifies and rationalizes own-source revenue assignments, and establishes a transparent rules-based system to allocate central transfers among municipalities. This should be matchedby improved budgeting, financial management and accounting systems. xv Introduce elections for mayors and municipal councils at the earliest feasible opportunity. Institutional measures should be introduced to bridge the current gap between the bottom-up system for expression o f community voice and top-down planning andbudget decision malung. 47. CDCs were established as community-level vehicles to deliver infrastructure inrural areas under NSP and have since started to take on a wider range o f local governance functions. To sustain their effectiveness, the following issues needto be addressed: CDCsneed to secure a regular and reliable funding flow to resource local development activities which provide their raison d 'etre. The funding should be set at far lower levels than under NSP. CDCs' roles should be consolidated on the basis o f the new by-law. In general, CDCs should take on only public good functions, and not compete with other local governance bodies (such as school management committees). The matter o f "ownership" o f CDCs at the central level will have to be settled within the central government by those agencies and ministries which have a core interest inthe question. Demonstrate Government'swill tofight corruption. The public has little trust in leaders and government institutions. Despite positive rhetoric, government has not initiated action in the face o f widespread and entrenched corruption. Although the Bonn accord stressed accountability o f public organizations, the political process that followed stressed representation from all quarters above accountability, and peace before justice. Afghanistan i s now responsible for 93 percent o f total global illicit opium output. Institutional disarray and in-fightingwithin government, and lack o f strong government leadership to confront corruption has resulted in the widely heldview i s that corruptionis being ignoredor tacitly allowed. This report recommends: Initial implementation of meaningful yet feasible short-run measures, with emphasis at sector and agency levels instead o f spending too much effort and resources on preparing an impressive strategy, and focus on a few agencies where the public interacts most frequently with service providers. Inthese agencies, the number o f steps, signatures and visits could be reduced. Clarifying institutional arrangements so that different agencies involved in anti-comption work harmoniously. Enforcing ministries to respond to audits and involvingMinistry o f Finance inthe response to audit reports will allow it to play a stronger oversight role. Strengthening the leadership o f institutions involved inthe fight against corruption. Much more can be done to go after key actors inthe drugindustry(larger and medium drug traders, and their sponsors inside and outside government). Even if weaknesses in the justice sector make prosecution and conviction very difficult, officials associated with the drug industry should at least be removed from their positions in government, which would send a strong signal o f government commitment to address the drugissue. (11) Buildpublic trust in parliament and thejudiciary -so they canprovide formal oversightin an effective and crediblemanner. 49. In the National Assembly, work is needed to strengthen coordination between the Cabinet and NA, and to build parliamentarians' understanding of their actual role, including how they can constructively contribute to law-making and oversight o f the executive. Particularly, Budget Committee xvi members' comprehension of how government budget and accounts are prepared and presented needs to be strengthened. 50. Inthejustice sector thisreportrecommends: Recognizing that security strengthening measures are only partially effective without accompanying strengthening o f rule o f law. Strengthening coordination between the Cabinet and the NA, referred inthe previousparagraph, will improve legislative capacity. 0 Starting small and scaling up. Bothjudicial reform and infrastructure creation should focus on the major cities because that is where the formal justice system i s most used and most needed. But during this time the role o f the judiciary, especially the Supreme Court, needs to be established in enforcing constitutionality o f laws and reviewing legality o f administrative decisions. Supporting non-state dispute resolution because it will take many years before sufficient courthouses are constructed, judges appointed and legal texts are available. (12) Provide credible information on government programs and performance - and open up governmentprocesses topublic scrutiny andparticipation. 5 1. The media and CSO sectors inAfghanistan are still finding their feet. Donor support can help to develop their capacity and effectiveness over time. But, inthe area o f public accountability, their role is very much dependent on their access to official information and their ability to provide feedback on government performance. Donors should support government initiatives to improve transparency, while being sensitive to the country's history, culture and politics. A number o f innovative approaches have already been built into donor-funded projects to provide information to affected communities, involve them in project planning and implementation, and to monitor the impact o f project interventions on economic and social conditions inthe affected area. These approaches shouldbe mainstreamed into other government programs over time. The Government should also provide the information needed to inform public debate on complex issues such as the budget and corruption, and to build public confidence in financial management and procurement processes. 52. This would be a daunting agenda inthe best o f circumstances. InAfghanistan, where the public administration has to be rebuilt from a low base in a very difficult reform environment, it can sometimes seem overwhelming. But some comfort can be taken from the progress to date, and the fact that small steps can build into a coherent program over time. Experience in Afghanistan and elsewhere shows that successful innovations in service delivery can be implemented - even in an environment o f poor overall outcomes and when systemic problems are still to be resolved. However, over the longer term, there is no substitute for a merit-based and accountable public administration that can perform its core functions well. It will simply take a long-term commitment from the Government o f Afghanistan and its international partners to stay the course. The stakes are high: only an effective public administration, that is seen to provide security and services to the people, can ensure the long-term legitimacy and credibility o fthe state. xvii I. CONTEXTANDPROSPECTSFORREFORM A. Introduction 1.1 Buildingan effective state -that canprovide security and services to the people, while protecting their rights-has always been at the heart o f the reconstruction effort inAfghanistan. At the end o f 2001, after more than two decades o f wars and civil strife, the political system had largely collapsed, with power dispersed in the hands o f regional warlords and local commanders. Although the Taliban had taken control over most o f the country, prior to being overthrown inlate 2001, they didnothing to builda modem state other than imposing, through harsh measures, a certain degree o f order. While many o f the formal structures o f public administration remained in place, appointments were based more on loyalty than merit, procedures followed the old-style soviet model, and officials lacked the financial resources and motivation to do their job. Therefore the new government faced a severe challenge: to restore security and take charge o f running the country, while also rebuilding the very limited capacity o f the public administration. 1.2 This Chapter will review progress over the past five years in building an effective public draw some lessons from successes and failures inAfghanistan -- as well as from research and experience administration in Afghanistan (for what we mean by public administration reform, see Box 1.1). It will inother countries (especially fragile states). Anditwill outline the context andprospects for accelerating public administration reforms and improving the delivery o f public services. This i s not intended to be a comprehensive review o f recent developments in Afghanistan, which are well covered in other references.' Rather it is intended to set the stage for the rest o f the report, which will develop reform proposals in three key areas: (a) building an effective civil service; (b) improving local governance and service delivery; and (c) making government accountable to the people. The final Chapter will pull this together into an agenda for public administration reform, and chart a way forward that i s both relevant and realistic inthe current Afghan environment. B. Rebuildingafter the Taliban2 1.3 Afghanistan has long been one o f the poorest countries in the world, falling near the bottom in terms o f per capita income and UNDP's Human Development Index. Nevertheless, Afghanistan was at peace between the 1930s and the late 1970s, and underwent a modest degree o f economic and social development. Modernization was concentrated in the cities and towns, however, and most rural areas retained their traditional mores, governance structures and social practices. The Afghan state remained relatively weak and had limited outreach inmost o f the country. More than two decades o f conflict after 1978, compounded by a prolonged drought, resulted in widespread human suffering and massive displacement o f population (both within Afghanistan and as refugees inneighboring countries). By 2001, an estimated 7 million people were vulnerable to famine. Afghanistan's infrastructure hadbeen destroyed or degraded, and its humanresource base severely depleted. ' See, for example, the Preliminary Needs Assessment prepared ADB, UNDP and World Bank 2002 for the situation immediatelyafter the fall of the Taliban. An excellentreview of developmentchallengesis providedinthe AfghanistanNational Development Strategy 2006. Also see World Bank 2005a on Afghanistan: State Building, Sustaining Growth, and Reducing Poverty andWorld Bank 2006dManagingPublic Financesfor Development. *This section draws heavily on World Bank 2005a and other references as noted. For an historical review of government in Afghanistan, see Evans, Manning,Osmani, Tully andWilder 2004. 1 Box 1.1: What do we meanbyPublicAdministrationReform? e.g. NationalAssembly, Auditor General n Building a capable civil service, managed according to formal rules and practices, and Civd society employed inwell-structured line organizations at both central - and sub-national levels. a I Improvedgovernance & service delivery The objective o f public adrmnistration reform (PAR) i s to deliver public services in a more efficient and effective manner. Much o f the effort towards this objective has focused, to date, on establishing a well-performing, merit- based civil service. But good service delivery requires more than competent civil servants. It requires management processes and practices that focus organizational units and their staff o n delivering the services for which they are responsible, and restructuring o f government organizations to respond to their modem-day functions. Meanwhile, there i s growing interest in sub-national administration, including deconcentration o f responsibilities within line ministries, and efforts to improve capacity and coordination at the local level. Taken together, they constitute traditional public administration `supplied' by government and are depicted by the diagram's middle box. But this core o f public administration tends to become self-serving unless stimulated by the `demand' from the Afghan people for good adrmnistration and public services (expressed through elections, civil society groups, the media etc), and the `checks and balances' provided by formal oversight institutions (such as the National Assembly) and the judiciary-institutions that hold the government accountable. Given Afghanistan's security situation, and donors' heavy presence in both security and civil administration, any discussion o f civilian public administration inevitably expands to include security issues. However tlus report does not discuss security forces per se. It references security issues only in so far as they impact on government performance. Source: Adapted from World Bank (2007) and GOAPAR Framework (discussed inChapter 11). 1.4 While many o f the formal structures o f public administration were still intact, they had become largely dysfunctional as the state broke down under previous regimes. According to the Preliminary Needs Assessment releasedinearly 2002:3 "Even at the central level in Kabul, ministries or departments are war-damaged shells, without even the most basic materials or equipment, and with few experienced staff. Before the Taliban, ADB, UNDPandWorld Bank2002, 15. 2 43 percent o f government employees were women, most o f whom were dismissed. Similarly, in other walks o f life women were largely prohibited from work. Government staff has been paid intermittently, if at all, and many o f the senior personnel have either left the country or sought alternative part-time employment." 1.5 The new Constitution, adopted in January 2004, refers to "preserving the principles o f centralism", and there i s little appetite, at least inKabul, for a substantial devolution o f political power to sub-national level^.^ As a result, on paper, Afghanistan is one o f the most centralized states inthe world: Politically, Afghanistan i s a unitary state, with all formal authority vested in the government in Kabul. Provincial governors, for example, are appointed by the center, w h c h also determines their powers and responsibilities. Municipal mayors are also approved by the center -but based on candidates proposedby local elders and the Governor. Fiscally, Afghanistan i s also highly centralized. Expenditures at provincial and district levels are made through national programs carried out by provincial arms o f central ministries. N o fiscal transfers as such go to sub-national levels. With 44 percent o f the civil service workforce in Kabul, funds are concentrated at the center. Budgets o f provincial governments are set in Kabul, and approval from Kabul i s required for even minor changes. On the revenue side, Afghan districts and provinces have no independent authority to impose taxes. The only expenditure autonomy i s at the municipal level. Although all tax and fee rates are set inKabul, revenues and fees collected by municipalities remain there and fundmunicipal spending. 0 Administratively, provincial and local governments have only a very modest formal role in decisions concerning their own structure, recruitment o f senior staff, size o f establishment, composition o f workforce etc. Such decisions are made by each ministry in Kabul, in conjunction with the Office o f Administrative Affairs (OAA), and signed o f f by the head o f the Independent Administrative Reform and Civil Service Commission (IARCSC). The ministry in Kabul or the President makes staff appointments at middle to senior grades, while the provincial governor appointsjunior staff down to the district level. 1.6 Despite this very high degree o f dejure centralization, the defacto reality i s that central control i s very weak, given the strength o f regional and local warlords, who command sub-national revenues and military power. Governors often make resource allocation decisions except on basic salaries. Staff appointments from Kabul are often rejected in favor o f those loyal to regional factions, and even Kabul- based appointments often reflect loyalties and ethnic ties, rather than merit. Factional loyalties have been strengthened by a general lack o f financial support from the center. With the exception o f offices which have been through PRR, pay levels are generally low, especially for middle-level and senior staff. Non- salary budgets for provinces and districts have also been inadequate in recent years, and allocations are very uneven across provinces. The end result i s a very inefficient and inequitable system o f sub-national administration. 1.7 This de facto reality reflects the long-standing tension in Afghanistan between traditional regional models o f administration, based on loyalty to the ruler, and tribal and family relations, and newer European models based on service to the public and merit-based appointments. In Afghanistan, the exchange o f obligations and service among different clans, tribes and families follows a set o f norms h o w n loosely as wasita, which literally means "c~nnection".~Those in government are expected to See Rubin2007 for adiscussionof the political dynamics underlyingthe preference for a strong centralstate. Understandably, there is an uneaseabout anythingwhich couldencourage "fissiparous tendencies". This situation i s not unique to Afghanistan. Accordingto North, Wallis and Weingast 2006, most societies "provide order by using the political system to limit economic entry to create rents, and then using the rents to stabilize the political system and 3 provide a valuable point o f contact and financial support for the entire extended family. As the capacity o f the state was steadily eroded over the past twenty years, there were paradoxically pressures to expand its role: "driven by the logic o f creating positions to accommodate interests, rather than the delivery o f public services."6 The next chapter discusses how wusitu plays out in all aspects o f civil service management, and how the basic tension between patronage and merit-based models lies at the heart o f the PAR agenda inAfghanistan. 1.8 Another complicating factor was the rapid expansion o f the international military and civilian presence in Afghanistan after 2001. Learning from experience in other fragile states, the UN adopted a "light footprint" approach, where the assistance mission had no direct administrative responsibilitiesand Afghan staff were used as much as possible. Even so, Afghanistan was and is highly dependent on foreign aid. InSY 1385, the international community provided US$l.5 billion o f development assistance, almost three times the level o f domestic revenues? This high dependency on aid, much o f which i s implemented outside the government budget, has created a "second civil service" of externally-paid consultants and advisers, many o f them only loosely supervised if at all by the government. Not only has this second civil service taken some o f the best talent from the government and bid up the cost o f scarce talent, but with the relatively small share o f resources at its disposal, it i s a constant challenge for the government to stay "in charge" o f the development agenda. This situation i s perhaps inevitable in a situation where local capacity has been decimated, and both the government and the international community share an interest in showing that progress i s being made. But it also creates an environment in which the long-term agenda of institution building is often sacrificed to short-term imperatives.* Indeed, some have argued that the prominent foreign presence undermines the very objective o f building a credible and legitimateAfghan state.g 1.9 Faced with these daunting conditions, the new Government o f Afghanistan, in June 2003, established the new Independent Administrative Reform and Civil Service Commission (IARCSC). It opted for an "asymmetric" approach to public administration reform, which was more opportunistic than systematic. This reflected the reality that the central government had little reach beyond Kabul and, because o f internal divisions, could not build any meaningful consensus around the key elements o f a comprehensive PARprogram. As argued inHakimi, Manning, Prasadand Prince 2004, 11: "Under such difficult circumstances, where central government has very little traction over the public sector that it seeks to reform and must, at all costs, avoid wholesale, painful reforms which would further alienate civil servants, and probably the population as a whole, the only option, other than total inaction, is to stimulate modest, targeted incremental reform o f key functions within government departments and agencies. Priority must be given to those transformations and changes that would enhance the legitimacy o f the center rather than undermine it and which both exploit and nurture the appetite for reform that does exist within certain areas o f the government administration." limit violence." They call this the "natural state". Only a handful of countries since WWII have transitioned to "open order states", where openaccess andentryinto economicandpolitical organizations sustainseconomic andpolitical competition. '*Substantialexternal Rubin2007,4. This is one of the backgroundpaperscommissionedby the World Bank for this report. assistanceto the securitysector is excluded from these numbers. It also creates resentmentamong Afghans who have livedthroughmany years of conflict, at very low pay levels, and see many ofthe best-payingjobs goingto English-speakingAfghans returningfrom overseas. See, for example, Suhrke 2007. Others argue that the foreign role has been too light to achieve its objectives(see Maley 2006 for a summary of these arguments). One such example is the initial decisionto limit the I S M military role to Kabul. This provided a window for the Taliban to re-establish in the South, and complicate the current task of providing security and development inthese conflict-affectedareas. 4 1.10 The asymmetric approach was reflected inthe Priority Reform and Restructuring(PRR) process, which allowed staff in key departments and agencies to be placed on an elevated pay scale for a fixed term in exchange for restructuring. This was intended to jump start the reform process, until more fundamental issues o f civil service pay, grading and structure etc could be sorted out. By March 2007, new job descriptions had been written for over 43,000 positions, and 7,800 of these had been filled through the new "merit-based" selection process. There i s little doubt that ministries that have gone through a serious PRR process, such as the Ministry o f Public Health (MoPH) and the Ministry of Rural Rehabilitation and development (MRRD),have benefitedinterms o f higher pay and better performance. However, there were also growing concerns about the asymmetric approach: with highly differentiated pay scales within ministries and the difficulty o f working within dysfunctionalministerial constraints. As a result, the program was expanded in late 2004 into a more ambitious ministry-wide program. In general, while many staff have been approved for higher salaries, progress on restructuring has been slower than promised in the PRR proposals. And the merit-based selection process, managed by IARCSC, has been criticized for being both slow and subject to political bias and patronage." 1.11 More fundamentally, the PAR program has been criticized for focusing too much on building technical capacity, primarily inKabul, while downplaying issues o f political legitimacy, especially at the local level. The presumption was that increased capacity would improve service delivery and thereby help to build political legitimacy throughout the country. However, Lister has argued that "powerful international and domestic actors have tended to see `state building' as creating organizations and structures and have ignored the very interventions that would have contributed to supporting the introduction and establishment o f a different set o f rules constraining powerful interests. Inthis they have contributed to the failure o f both local government reform and the wider state-building agenda in Afghanistan."" She goes on to argue that more attention to key institutional changes - such as disarmament, reformo fthe police andjudicial system, and merit-based senior appointments -would have helped to change the rules o f the game in Afghanistan from informal patronage-based systems to more formal and professional state systems. "Instead their neglect at a critical period has enabled local powerhouses to continue to use the state as a means to exercise power, resisting or co-opting attempts to create new structures and impose bureaucratic rules."12 1.12 Faced with severe capacity constraints, the government had little option but to develop a "lean state" focused on core functions. Consistent with this aim, the number o f ministries has been reduced from 30 to 25. The number o f civil servants has grown slightly, from 327,000 in 2004 to 348,000 in 2007, mainly due to the addition o f more teachers. But at just over 1percent o f the estimated population, this is still a relatively small civil service by regional and international standards. The government has also adopted some innovative mechanisms for program delivery. Two notable examples show what can be achieved under difficult conditions: 0 Delivery o f health services in rural areas was always very limited in Afghanistan, with a heavy dependence on NGOs duringthe war years. After the fall o f the Taliban, the MoPH formulated a basic package o f health services (BPHS) that i s largely delivered inrural areas by NGOs through grants or contracts. Independent assessments by Johns Hopluns University show a 21 percent improvement inthe quality o f health care from 2004 to 2006. M o P H has also experimented with a "contracting-in" management model to run publicly-financed health services. This model has also shown significant improvements in quality o f care, with as good coverage as in provinces using the NGO delivery model. While "contracting-in" model has been applied in relatively easier provinces near Kabul, with better security, the good results probably also reflect talented loSee Michailof 2007a. I'Lister2007, 15. I'Ibid, 16. 5 management and competition from NGOs. (Broader issues o f service delivery models, across sectors and over time, are discussed inChapter 11.) The National Solidarity Program (NSP) was introduced in 2003 to: (a) lay the foundations for a strengthening o f community-level governance; and (b) support community-managed sub-projects that improve access o f rural communities to basic infrastructure and services. By the end o f 2006, the program had reached more than two thirds o f Afghanistan's estimated 24,000 villages across all o f the country's 34 provinces. A key component o f NSP was the establishment o f Community Development Councils (CDCs) to develop a Community Development Plan, select sub-projects for funding under NSP, and manage their implementation, operation and maintenance. A mid-term evaluation by the University o f York in 2005-06 showed that communities reached by NSP had empowered CDCs, improved community relations and increasedpublic faith ingovernment. (Emerging issues around the role and funding o f CDCs are taken up inChapter 111.) 1.13 Similarly, a lot o f progress has been made in building up the government's public financial management systems under the direction o f the Ministry o f Finance (MOF).13 The World Bank's assessment under the Public Expenditure and Financial Accountability (PEFA) program in November 2005 "portrays a public sector where financial resources are, by and large, being used for their intended purposes as authorized by a budget which i s processedwith transparency and has contributedto aggregate fiscal di~cipline."'~These achievements are commendable compared to experience in other post-conflict countries. However, the assessment also notes that these improvements have to a large extent been dependent on external capacity brought in on an emergency basis to get things going (renting rather than building capacity). The development o f sustainable core government capacity for key public finance functions has barely begun, and there are daunting weahesses inpublic financial management in the line ministries. More fundamentally, there i s still a wide gap between public expectations and the actual delivery o fpublic services. 1.14 While the PAR agenda i s largely focused on developing an effective civil service, this objective won't be achieved without stronger institutions of public accountability. Politically, Afghanistan has come a long way since the Bonn Conference in establishing the basic institutions for a "broad-based, gender-sensitive, multi-ethnic and fully representative government." A new constitution was adopted in January 2004. Hamid Karzai, previously head o f the transitional administration, became Afghanistan's first popularly elected President in December 2004. Elections for the Wolesi Jirga (the lower house o f the National Assembly) and for Provincial Councils took place in September 2005. The inauguration o f the National Assembly in December 2005 marked the formal conclusion o f the Bonn process, a major achievement. However, Parliament's performance to date has been constrained by its fragmentation and lack o f experience. The normalization and professionalization o f political culture, including trust between individuals and groups, has a long way to go before Parliament can achieve its objectives o f representing Afghan citizens and holding the executive to account. 1.15 Similarly, despite some signs o f progress over the past two years, including the appointment o f a more reform-minded and professional Supreme Court in 2006, the justice sector in Afghanistan i s still very weak. According to a recent review by the United Nations Assistance Mission in Afghanistan (UNAMA): "The justice system still lacks sufficiently qualified officials, adequate legal education, and the necessary administrative tools and physical infrastructure to adrmnister justice properly, fairly or effectively. The absence o f strong state institutions, particularly in rural areas, low salaries for l3Issues ofpublic financialmanagementwere recentlycoveredinWorld Bank 2006d, and are not a specificfocus of this report. l4World Bank2006d, 22. 6 judges and prosecutors, rampant corruption, the ominous influence o f warlords and local commanders, and the failure to ensure a secure environment o f courts, judicial personnel, victims and witnesses also continues to severely undermine the capacity o f the legal system to act independently and impartially, and contributes to the l o w level o f public trust and confidence in these institutions.. ... Traditional dispute resolution mechanisms, especially inrural areas, remain dominant to the detriment o f women and children inparti~ular."'~ 1.16 Duringthe war, non-governmental organizations (NGOs) often substituted for a weak or non- existent state by providing basic health and education services. As noted above, this model has now been formalized to some extent inthe health sector. Civil society and the media have also been widely used for the purpose o f civic education, in order to secure the success o f the electoral process in the post-Taliban period. However, as NGO dependence on aid has increased, their local support and responsiveness to local concerns has sometimes been reduced. There are only two organizations - Integrity Watch Afghanistan and the Afghan Civil Society Forum - monitoring government performance and workmg on integrity issues. According to Integrity Watch Afghanistan, while the media has wide ownership and relative autonomy in Afghanistan, compared to other countries in the region, it lacks the investigative skills to put effective pressure on the Executive. And journalists are often threatened for raising politically sensitive issues, including cases o f corruption. 1.17 It is against this backdrop that this report looks at the priorities and prospects for public administration reform inAfghanistan over the next five years. However, before returning to that agenda, it is perhaps usehl to step back, and learn from experience inother countries about what has worked and has not worked in public administration reform, and how to increase the chances for success in Afghanistan. C. Lessonsfrom Research and Country Experience 1.18 The World Bank has done considerable research on the role o f the state and reforming public administration over the past ten years. The 1997 World Development Report on The State in a Changing Worldproposeda two-part strategy for buildingan effective state: "Matching the state's role to its capability is the first element in this strategy. Where state capability i s weak, how the state intervenes - and where - should be carefblly assessed. Many states try to do too much with few resources and little capability, and often do more harm than good. A sharper focus o n the fundamentals would improve effectiveness. But here it is a matter o fnotjust choosing what to do and what not to do -but o f how to do it well. But capability is not destiny. Therefore the second element o f the strategy is to raise state capability by reinvigorating public institutions. This means designing effective rules and restraints, to check arbitrary state actions and combat entrenched corruption. It means subjecting state institutions to greater competition, to increase their efficiency. It means increasing the performance o f state institutions, improving pay and incentives. And it means making the state more responsive to people's needs, bringing government closer to the people through broader participation and decentralization."`6 1.19 Fukuyama 2004 builds on this framework to "distinguish between the scope o f state activities, which refers to the different functions and goals taken on by governments, and the strength o f state power, or the ability o f states to plan and execute policies and to enforce laws cleanly and transparently - l5UNAMA 2007, 1. l6WorldBank 1997,3. 7 which is now commonly referred to as state or institutional ~apacity."'~He then develops a matrix (see Figure 1.1) to show how different countries rank in terms o f the scope o f state activities and the strength of capacity, as well as how these rankings change over time. According to Fukuyama, the dominant trend in recent years has been to consolidate the scope of state activities, with much less attention to (and success in) buildingstate capacity. Afghanistan would probably be close to the axis on Figure 1.1, with both limited scope and capacity. The challenge i s obviously to build the state's strength, including throughpublic administration reform, before embarking upon any major expansion of state activities. 1.20 ButFukuyama also warns about the "black hole ofpublic administration". By this he means that there are no globally-valid rules for public-sector organization, and that the field o f public administration i s "more of an art than a science.'' While most good solutions to public administrationproblemswill have certain common features o f institutional design, they will also have to incorporate a great deal o f context- specific information. Fukuyama therefore concludes: "If we really want to increase the institutional capacity of a less-developed country, we need to change the metaphor that describes what we hope to do. W e are not arriving inthe country with girders, bricks, cranes and construction blueprints, ready to hire natives to help build the factory we have designed. Instead, we should be aniving with resources to motivate the natives to design their own factory and to help them figure out how to buildand operate it themselves. Every bit o f technical assistance that replaces a comparable capability o n the part of the local society shouldbe regarded as a two-edged sword and treated with great caution. Above all, the outsiders need to avoid the temptation to speed up the process by runningthe factory themselves." 1.21 Fukuyama also draws on the analytical and historical framework developed by Pritchett and Woolcock 2002 to understand why there i s an absence of a uniformconsensus on how to improve service delivery. They distinguishbetween two aspects of public sector services: their transaction intensity (the number of decisions that need to be made by organizations) and their discretionary character (requiring judgment of imperfect or incomplete information by a shlled decision maker). A traditional, centralized approach to public administrationmakes most sense when it i s easy to monitor a few decisions based on good information, such as runninga central bank. But this approach becomes more difficult when many decisions are neededand there i s little information to monitor performance. Education i s a case inpoint, where outputs are hard to measure and it i s virtually impossible to hold individual teachers accountable. Pritchett and Woolcock develop eight different models of service delivery- including supplier autonomy, contracting out, decentralization and community-driven development - which may provide better outcomes than a centralized civil service dependingon local conditions. "Fukuyama2004,9. 8 Figure 1.1:State Scope and Strength France Japan United States USSR Turkey Brazil Sierra Leone Afghanistan Scope of state functions Source: Fukuyama2004, 18 (withAfghanistan added). 1.22 This approach was further developed in the 2004 World Development Report on Making Services Workfor Poor People. It provided a framework for thinlung about relationships between clients, providers and policymakers, and alternative models for delivering services depending on whether: (a) the political system i s pro-poor or captured by vested interests; (b) clients are homogenous or heterogeneous; and (c) services are easy or hard to monitor. This framework was applied to Afghanistan inWorld Bank2006d, as shown inFigure 1.2. Thisreportgoes onto draw some lessons for Afghanistan: One size does not fit all: different approaches can and should be used depending on the characteristics o f the service concerned and the context inwhich it i s being delivered. However, it is essential that the underlying accountability relationships support and encourage adequate service provision. Service providers need to be made more accountable directly to consumersheneficiaries: through enhanced channels for voice, choice o f providers or, in some cases, by paying directly for the service and thereby having a greater incentive to be concerned about quality. Different models o f service delivery have different implications for state building. Several dimensions are important in this regard: (a) the actual role played by government (policy, financing, management, provision etc); (b) the visibility o f the government to consumersheneficiaries; and (c) the effectiveness o f the services (access, quality, inclusiveness etc). Where possible, competition (or at least contestability) among providers can play a useful role. Mobile telecommunications i s a notable example o f competition resulting inmajor improvements inaccess, quality andefficiency o f service delivery. 9 InAfghanistan, the lack o fcapacity intraditionalcentralized service delivery systems provides an opportunity to rethink, modify or even move away from this model - although this could seriously constrain service delivery inthe short run. Alternatives include deconcentration within the government structure, more use o f commercialized modes o f provision, and private sector participation. Accountability Financing Choose, Assess Pay for service 1.23 The 2006 GlobalMonitoring Report prepared by the World Bank and IMF has a special focus on monitoring governance. In an expanded background paper, Levy 2006 concludes: "Top-down reforms o f bureaucracies and checks and balances institutions generally take a long time before they help improve the ffont-line performance o f governments. Low-income countries seeking to turnaround their performance thus confront a dilemma. Maintaining the momentum o f turnaround may depend on achieving major gains in service provision in the short-to-medium term. But reforming their top-down state accountability system to the point that it can provide the requisite combination of accountability and responsiveness i s a task which will only bear fruit over the long-run." Levy goes on to argue that top- down initiatives can be undercut insofar as they are not underpinned by a change in the incentives o f political and bureaucratic actors. "Engaging at the front-line has the potential to alter this equation - achieving development results quite quickly, while also potentially helping to accelerate the pressure and momentum for deeper top-down reforms."** 1.24 World Bank evaluation results confirm the difficulty o f implementing reform programs inpublic administration and public financial management. According to the Annual Review of Development Effectiveness 2006, while assistance has ledto improvements inthe quality o fpublic sector management processes in some countries, this has not usually translated into improvements inthe perceived quality o f governance. Yet recent progress in perceived governance quality in some countries in Eastern and Central Europe shows that it i s possible to make progress in a limited time when there i s strong country commitment: "Public sector reforms o f a technocratic nature, such as modernizing personnel practices, can succeedwhen they buildonpolitical commitment. Bulgaria's achievement inprofessionalizing its civil service, for instance, has been the product o f both donor-supported reforms in pay and recruitment andbroadpolitical interest inmeeting conditions for European Union (EU) accession. But many reform programs have been undermined by lack of political support. The extent of political opposition is often underestimated at the time o f design. In Bolivia and Yemen, Bank- supported reforms in civil service management achieved little, because there was no commitment IsLevy 2006, x-xi. 10 to ending the traditional role o f the public service as a vehicle for large-scale patronage appointments. When political conditions are not ready for wholesale reforms, it is advisable to proceed gradually, identifying opportunities for less-contentious reforms in order to build coalitions across affected interests andto gradually gain rn~mentum."'~ 1.25 State buildinghas always had a central role inthe World Bank's work on fragile states. Ina 2005 Board paper on Good Practice in Country Assistance Strategies it i s noted: "Fragile states, as the name implies, are fragile because state institutions have limited capacity to deliver services to their citizens, to control corruption, or to provide for sufficient voice and accountability. In the long-term, therefore, a successful exit from fragility will only come about through stronger and more accountable state institutions which are resilient to political and economic shocks. Building state institutions means not only focusing on state capacity, but also on the responsiveness and accountability o f state institutions to the population at large, civil society and the private sector."20 The paper goes on to stress the importance o f calibrating assistance to weaknesses in existing capacity and accountability and to the risk o f conflict or instability. 1.26 Cliffe and Manning, in a forthcoming paper for the International Peace Academy, look in more depth at the challenges o f buildingstate institutions after conflict. They argue that "effective institutions are now widely viewed as critical to address both the `capacity deficit' and the `legitimacy deficit' faced by fragile states - since only strong national institutions can ensure that the state is associated with provision o fpositive services to the population, and can be held to account by its citizens." Yet the track record on building state institutions inpost-conflict societies i s mixed at best. Four particular problems are identified: First, many reconstruction efforts have been insufficiently informed by what institutions already exist, and so have tended to reinvent the wheel, rather than build on the pre-existing institutional archtecture. Transitional arrangements should be as light as possible, and as reliant as possible on national capacity. 0 Second, efforts to support institution-building have typically been quite diffuse, spread across all sectors and all areas o f the state, without a clear sense o f priorities or sequencing over time. The initial focus should be on key functions and outcomes that are necessary for the survival o f the state and sufficient to restore credibility to the state. Third, there has been little attention to the relation between transitional oversight and delivery mechanisms and long-term national institutions. Some sectors (army, health) are more amenable to the use o f international models and capacity than others (police, education). Fourth, donors' own good intentions to support rapid recovery after a conflict have all too often unintentionally undermined long-term institution-building, by sapping the skills base available to national institutions and bypassing national decision malung structures. 1.27 Given these lessons from research and country experience, our approach to PAR inAfghanistan should follow the following principles: World Bank 2006e, xv. WorldBank2005d, 3. 11 Initially limit the role of the state to core functions in line with existing capacity and political realities.2' This i s consistent with the government's commitment to a "lean state" and the use of non-governmental delivery mechanisms insome sectors (such as health)?2 Recognize that real PAR will take a long time -- and will not generate quickreturns. Itrequires a steady hand and a lot of patience. Likewise, donor support has to be sustained to support long- term institutionalchange and capacity building. Look for innovative ways to respond to popular demands for better service delivery in the short and medium terms. Here Afghanistan's own experience in health and community development provides valuable lessons, along with experience inother countries (e.g., India). The broad nature of PAR requires commitment to a common framework, and a clear sense o f priorities and sequencing over time. PAR should build into a coherent program, even if short- term actions will often haveto be opportunistic and incremental. D. LookingAhead: ChallengesandConstraints 1.28 Before turning to the future, it's important to acknowledge what has been achieved in Afghanistan over the past five years. From a state of near collapse at the end of Taliban rule, the non- opium economy grew on average by 16 percent per annum from 2001/02 to 2005/06. Progress on social indicators i s also very significant: enrollments inprimary schools are up from one to five million children, one-third o f them girls; and the coverage of basic health services has been extended to 82 percent o f the population, contributing to a 22 percent fall inthe infant mortality rate from 2002 to 2006. Yet, nothing can conceal the fact that Afghanistan remains a very poor country, withper capita income around US$300 in2005/06. Healthindicators remain amongthe worst inthe world, just over halfofthe eligible students are inprimary school, and only about one third o f the 15-24 year old population i s literate. Progress has also failed to keep up with the expectations o f the population, which had been buoyed by new-found freedoms and promises of large amounts of aid. According to a 2006 survey of citizen perceptions sponsored by the World Bank, the biggest problem facing Afghanistan was security and law and order, followed by progress on the economy, poverty and jobs. In terms o f basic services, most complained about poor access and quality of electricity, sanitation and drinkingwater. 1.29 The Government of Afghanistan's strategy to bridge this gap i s laid out in the Interim Afghanistan National Development Strategy (I-ANDs). The primary objective of the I-ANDs i s "to consolidate peace and stability throughjust, democratic processes and institutions, and to reduce poverty and achieve prosperity through broad-based and equitable economic The strategy i s based on an economic growth rate of 9 percent per annum from 2006 to 2010: to provide people with a tangible sense of improvement, while compensating for the contraction o f the drugs economy. Together with investments in human development, this i s expected to lead to significant progress towards achieving The criteria proposed by Cliffe and Manning forthcoming to focus on key functions and outcomes that are necessary for the survival o f the state and sufficient to restore credibility to the state, are very relevant to Afghanistan. In comments on an earlier version o f this chapter, Professor Rubin questioned whether it is realistic to expect GOAto carry out even "core functions", given the low potential for mobilizing resources and the high costs o f administration, especially in insecure areas. It is certainly true that Afghanistan will be dependent on foreign resources for the foreseeable future. But, as more o f these resources are channelled through the budget, there will be more scope for their management by GOA. Therefore, the capacity o f the public administration is likely to remain the critical constraint on getting things done. 22Paul Fishstein (AREU) notes that the Government's commitment to a "lean state" i s more like acquiescence, as many middle- level officials see the use o fnon-governmental delivery mechanisms as a temporary measure only. 23Islamic Republic o fAfghanistan 2006, Executive Summary, 2. 12 Afghanistan's Millennium Development Goals (MDGs). Underpinning the strategy, i s a set o f goals to develop an effective and accountable public administration at the central, provincial and districts levels: "Through the Public Administration Reform program, the government will ensure that there is improved coordination between decision-making bodies within the central government. It will restructure ministries and simplify administrative procedures and business processes. The IARCSC will assist line ministries and decision-making bodies with reforms of sub-national adrmnistration and also monitor and evaluate the process. Government will work through line ministries and decision-making bodies to strengthen coordination and improve service delivery ministries. Government will create an effective, financially sustainable civil service and gradually phase out donor supplementation o f salaries. The IARCSC will strengthen rules and procedures for a professionally managed civil service, including promotion of high ethical standards and establishment o f disciplinary measures for corrupt or unethical practices. Government will adopt a merit-based recruitment system that promotes gender equity and ethnic diversity. It will establish a coherent, comprehensive skills development program for existing and new civil servants. The IARCSC and MoF will better manage international technical assistance to ensure faster, more sustainable skill transfer, including inthe area of project management."24 1.30 Unfortunately, while expectations are rising, the environment for implementing the I-ANDs, including proposals for reforming public administration and improving government performance, i s deteriorating. A number o f factors are at play: Politics. Like most post-conflict states, Afghanistan had to weigh up the relative merits o f co- opting or sidelining warlords and factional leaders, some o f whom have dubious pasts. Initially, a number o f technocrats, often Afghans returning from overseas, were brought into key positions. However, many o f these have now left, as more positions are allocated to balance ethic and regional interests (with the notable exception o f the Taliban).25 This is perhaps to be expected, as different factions vie for power within a maturing democracy. Even so, many observers perceive that the overall political environment for reform has deteriorated, and this has to be factored into any assessment o f the prospects for progress on PAR. Similarly, security concerns have increasingly encroachedupon the objectives and modalities o f bilateral aidprograms. Again, this is understandable, as security operations have to be quickly backed-up with demonstrable progress on reconstruction and development. However, a quick-fire approach i s unlikely to be sustainable, and i s often inconsistent with longer-term goals o f institution building and public administration reform. 0 Security. The security situation deteriorated in2006, with major challenges to the Government's authority from the M a y riots in Kabul and the growing insurgency in the South and East. By early 2007, the violence had spread, with suicide bombings in Kabul and in other parts o f the country as well. Reports from Amnesty International suggest that at least 3,000 persons have been killed since 2002, including around 1,000 civilians in 2006 alone. More generally, law and order problems - related to criminality, drugs and warlordism -- are a major concern for the population. Coordination between Afghan forces, and with international troops, is challenging, while police andjudicial reforms remain elusive. Some argue that the provision o f law and order is the primary function o f government, and until this i s done in Afghanistan, other functions will suffer. Recent attacks on schools inthe South are a case inpoint. Others argue that establishing law and order depends on a minimum level o f government presence and credibility. But, either 24Ibid, Volume 1, Chapter 7, 127. 25As noted by Suhrke 2007, 11-12: "Political criteria became increasingly evident in2006 and 2007, as PresidentHamidKarzai sought to co-opt potential rivals, rebels or critics by appointing them as special advisers to his ofice and distributing gubernatorial and police chief positions in the provinces. H e came to recognize militias run by local strongmen, a move supportedby the US.militaryfor counter-insurgencypurposes as well." 13 way, it will be very difficult to build effective public institutions in an insecure environment, without the normal checks and balances o f a working judicial system. Drugs. One o f the major causes o f insecurity inAfghanistan is the opium economy. By 2007, production had risen to an unprecedented 8,200 tons o f opium, accounting for more than one quarter o f the national economy,26 and further enhancing Afghanistan's role as the dominant supplier o f illicit opiates to the world. Opium production is increasingly associated with insecurity, with around 70 percent o f the poppy crop grown in 5 provinces along the border with Pakistan, and 50 percent from just one province: Helmand. According to a recent report by the UNOffice on Drugs and Crime and the World Bank 2007, the opium economy is a massive source of corruption and undermines public institutions especially in (but not limited to) the security andjustice sectors. There are also worrying signs o f infiltration by the drugindustryinto higher levels o f government and politics. Yet, there are no easy solutions. Counter-narcotics efforts - whether enforcement actions or development o f alternative economic opportunities - inevitably cannot be as nimble or quick as the activities they are targeted against, and they inevitably take time to be effective. Although different strategies will be needed in different geographic areas, there i s a strong argument that enforcement efforts against poppy farmers should follow rather than precede the development o f alternative income sources. Corruption. According to available statistics, Afghanistan ranks near the bottom interms o f the seriousness o f its corruption problem. Although the opium economy is undoubtedly the most important source o f corruption in Afghanistan, the unprecedented large inflows o f international assistance, and the pressures to spend money quickly, also carry associated vulnerabilities to corruption. Other, more normal sources and forms o f corruption, related to the role o f government in service delivery and regulation, are likely to increase as state activities and capacity are expanded. Similarly, there i s a growing risk o f political corruption, as the political system i s developed. Widespread corruption raises the cost o f business and deters and distorts private investment. It also undermines the effective functioning and credibility o f the state. The Government o f Afghanistan has recently produced a draft anti-corruption roadmap. While this i s an important first step, a muchmore cohesive and effective approach will be needed to reduce the costs o f corruption inthe future." 1.31 So it is in this difficult environment that the Government o f Afghanistan, supported by its international partners, has to buildan effective public administration. Inthe following three chapters, we will look at the immediateand medium-termreformpriorities inthree critical areas: Building an effective civil service. The Government o f Afghanistan is about to embark on a major reform o f the civil service pay and grade structure - to raise salaries and decompress pay scales between the top and bottom grades. This is seen as important for attracting and retaining qualified human resources. However, it will be a challenge to implement a merit-based system in the face o f widespread patronage and corruption. Furthermore, to lead to improved civil service performance and service delivery, higher pay will have to be linked to other reforms in ministry roles, functions, structures and processes. Past experience with PRR suggests that political pressures for higher salaries may well overwhelm reform efforts - unless there i s a strong commitment to implementingthe reformprogram ina systematic and phasedmanner. Chapter I1 looks at how such a program might be implemented: the sequencing of reform actions, the roles o f the IARCSC and line ministries, and ways to monitor progress. 26UNODCand the World Bank 2007 estimate that drug income accounted for 26.7 percent o f total (licit and illict) GDP in 2005/06. This share has almost certainly risen since then. 27The World Bank i s managing a separate work program on anti-corruption, o f which the roadmap is one o f the first outputs. This report refers to corruption only in so far as it's a symptom or cause o f broader weaknesses inpublic administration. 14 0 Improvinglocal governance and service delivery. There has been increasing debate within the Government o f Afghanistan and the donor community about the role o f sub-national administration. While Afghanistan is a unitary state, the Constitution provides flexibility to delegate functions to provincial and district government units. It is at this level that most services are delivered to the people - and where the credibility o f government will be won or lost. However, there are many actors at the sub-national level (including Governors, Provincial Development Committees, Provincial Councils, and Provincial Reconstruction Teams) that influence the way government works, albeit in different ways in different provinces and sectors. There are also powerful informal power structures at work, which often complement or replace government services. Chapter 111 will look at the major challenges facing sub-national administration in Afghanistan, their impact on governance and service delivery at the local level, and the implications for public administration reform. Making government accountable to the people. Some o f the checks and balances on government power come from internal controls, including through better public financial management. However, most successful states recognize the importance o f external controls on executive power: formally, through the judiciary and parliament, as well as informally through the role o f civil society and the media. Understandably, most o f the PAR focus to date in Afghanistan has been on civil service reform. But issues o f public accountability are likely to become more important as Afghanistan develops into a mature democracy, and experience in other fragile states suggest it is better to look at these issues sooner rather than later. Therefore Chapter IV explores what can be done now to ensure that government i s held accountable to the people, including through better information flows, and institutional development o f parliament and the judiciary. H o w can trust in these institutions be restored? And how i s this likely to impact on the public administration reformand anti-corruptionprograms? 15 11. BUILDINGAN EFFECTIVE CIVIL SERVICE A. Introduction 2.1 The previous chapter presented the background necessary for the rest o f the report to explore three key components o f Afghanistan's public administration reform: an effective civil service; local governance and service delivery; and Government's accountability to the people. The present chapter will examine in detail the first o f those three areas-the civil service. Using the analytical framework referenced earlier and drawing lessons from recent reform experience, this chapter's intent i s to develop proposals for buildingan effective civil service that functions at Compact-mandated standardsz8. 2.2 The civil service is being examined first because this piece o f Afghanistan's public administration is comparatively more developed than the other two. However, its success in serving the public depends on the other two components: local governance and public accountability-which are discussed in the next two chapters. The Government's PAR framework explains that a capable civil service i s not restricted to central government. Because the bulk o f public functions are delivered at the sub-national level, the capability o f civil servants working in provincial and district governments i s important along with local participation for monitoring delivery o f services. But even a civil service, strong at both the central and local levels, becomes self-serving unless stimulated by a `demand' upon government for good administration and public services. Some o f these checks and balances on government come from the National Assembly and from the judiciary, while the public and media have a role in monitoring government's performance and holding it accountable. Thus, institutions o f public accountability are very important for buildingan effective civil service.29 2.3 Civil service reform i s vital for building and sustaining an effective state in Afghanistan - one that has Afghans at the center o f determining policy and guiding utilization o f public resources. An effective civil service can reassure donors that their financial support is being credibly spent. More importantly, an effective civil service underpins government's credibility throughout the country. The civil service i s the `face' o f Government seen by citizens, whose expectation now from their Government i s that it will provide security and public services. Unless citizens perceive that civil servants are serving the larger public interest rather than their own, the Government's trustworthiness will be eroded leading to large scale dissatisfaction, and perceptions that outside actors are managing the state. 2.4 The civil service's wide extent, cutting across all sectors o f Government, makes its progress both essential and hardto achieve. Improving the civil service means dealing with individual organizations as well as the whole apparatus of government. But even more challenging than the scale o f reform i s its nature. Since the 1990s' both developed and developing governments' roles have shifted-away from being sole provider o f public services and towards policy setting and regulation-while allowing other agents to actually deliver the services. This shift in governments' priorities has meant reorienting many government organizations and their officials towards different goals, and to perform a different set o f tasks. Such transformation i s difficult anywhere: Afghanistan's circumstances enlarge the magnitude o f the challenge. At the start o freconstruction, organizational structures were still based on functions of the 28 At a conference in London on January 31, 2006, the Government of Afghanistanpresentedthe InterimAfghanistan National Development Strategy (I-ANDS) to its development partners. It set out the government's political and economic plans and priorities for the following five years up to the end of SY 1389. The framework for co-operationthat resultedfrom consultation *'Thisi on the I-ANDs at the Londonconferenceisreferredto as the AfghanistanCompact. s partofthe PAR frameworkapprovedon April 3,2007 by the Government's PAR SteeringCommittee. 16 Afghan government established duringthe 1960s and before. Their sizes had grown considerably during the Soviet period, and defunct and bloated structures were overlaid with arbitrary changes imposed during the conflict years. Civil service reform also means introducing performance incentives along with accountability mechanisms for civil servants employedwithin these organizations. This latter i s slow and painful because it involves changing officials' behaviors. In Afghanistan, the merit principle sits uncomfortably on a tradition o f administration based on wash (exchange o f favors): in spite o f discussions on this issue, the Bonn Agreement in2001 did not come out with any definite statement about a merit-based civil service.30 Progress in this area i s also compounded by a serious lack o f local competence and the continued uncoordinated and distortionary behavior o f donors in attempting to fill that void through short-term arrangements. At the sub-national level, increasing competition with anti- government forces with significant resources or influence at their disposal also creates serious impediments to improving government performance andlegitimacy. Box 2.1: Compact Goals for an Effective Civil Service The London Compact contained specific goals o f civil service reform to be achieved by SY 1389 (end 2010). The Compact stated: With respect to scope o f Government, Government machinery (including the number o f ministries) will be restructured and rationalized to ensure a fiscally sustainable public administration; the civil service commission will be strengthened; and civil service functions will be reformed to reflect core functions and responsibilities. With respect to civil servants' competence, requisite support will be provided to build the capacity o f the civil service to function effectively With respect to personnel management and incentives, a clear and transparent national appointments mechanism will be established within6 months, applied witlun 12 months, and fully implemented within 24 months for all senior level appointments to the central government and the judiciary, as well as for provincial governors, chiefs o fpolice, district administrators and provincial heads o f security. By end-2006, infurtherance o f the work of the civil service commission, merit-based appointments, vetting procedures and performance-based reviews will be undertaken for civil service positions at all levels o f government, including central government, the judiciary andpolice. Annual performance-based reviews will be undertaken for all senior staff (Grade 2 and above) starting by end-2007. Source: TheAfghanistan Compact at http://www.fco.nov.uWservletiFront?pa~ename=~enMarkeWXcelerate/ShowPa~e&c=Pa~e&cid=34650705195 11 2.5 The ingredients of an effective civil service are the same as those that make an effective state. These were outlined in Chapter I(paragraphs 1.18 to 1.22). The World Bank's experience and research has shown that effective governments have, first, deliberately selected what functions they will perform, and then, discharged those functions satisfactorily. The first step is to match the state's role to its capabilities. The next i s to raise state capability by reinvigorating public institutions. Fukuyama 2004 has described state performance interms o f the same two dimensions, which he has named-only slightly differently from World Bank 1997-as scope o f state activities, and strength o f state power. The same logic o f scope and strength also determine civil service effectiveness. A civil service can be a well- performing, effective tool o f government when its extent (scope) i s compatible with how well public organizations perform their tasks (strength). The civil service's strength i s determined by whether or not 30Rubin2007. 17 public officials have the competence to perform their tasks, and whether personnel management rules and practices induce behavior and performance consistent with the goals o f a professional politically neutral civil service. Thus the civil service's effectiveness is described by the scope o f government, and the competence and incentives o f civil servants. 2.6 The rest o f this chapter explores the Afghan civil service's current and future performance in terms o f its scope and strength. The government's scope i s evaluated in Section B interms o f its size and shape, how it i s arranged to accommodate differences between service delivery models in different sectors, and the structure o f individual organizations. The chapter then probes the civil service's strength interms o f civil servants' competence to do their work (Section C) and the incentives createdbythe laws and rules for personnel management (Section D). The last section (E) outlines conclusions from the analysis, and proposes directions for maintaining the reformagenda. 2.7 Clarifying terminology here can avoid confusion arising later inthe chapter. The word `capacity' i s used very often in Afghanistan. This chapter uses the word `competence' instead o f the wider `capacity' to mean the narrower idea o f civil servants' skills to perform their tasks. The term TA is used inAfghanistan to mean both technical assistance as well technical assistants hired from donor funds. In this chapter, TA means only the support provided by donors, while the individual contractors providing the assistance have been referred as consultants. The Priority Reform and Restructuring program may have been unpopular, but it has made the term PRR so prevalent that it i s used in Afghanistan both as a verb (describing the process that ministries undertook) as well as an adjective describing the ministry's status (whether PRR-ed or not). B. The Civil Service's Scope Whole of Government 2.8 The Afghan government has displayed considerable discipline in establishment control and civil service growth that i s rare in post-conflict countries. In Cambodia, for example, the civil service's size ballooned during reconstruction because the political coalition following the peace accord required expanding the civil service to absorb functionaries from parties joining the coalition. Close budget management has helpedcontrol the establishment inAfghanistan. Duringevery budget's preparation, the staffingdemand o f each ministryand department is examined and approved, often with reductions, during the budget formulation process. The salary bill i s released according to this approved tashkeel, and ad hoc, mid-year increases are not easy to extract from the Ministryof Finance. The persistence o f security threats and the resulting steady growth o f the security sector have acted as another check on civil service growth. The approved strength o f police and para-military forces combined is currently 62,000; but the JCMB has recently agreed a higher figure o f 82,000. Growing security sector expenditure constrains the rise o f general civilian government's wage bill. InSY 1386, the security sector's salaries and allowances accounted for nearly half o f the operating budget (48 percent) compared with the 27 percent for education and health sectors combined3'. 2.9 The Government has controlled its scope and the civil service's size by reducing the number o f ministries-from 30 at the start o f reconstruction in 2002 to the current number o f 25. Though this reduction may have been driven by political considerations rather than rationalization o f government's structure, it was a notable achievement. The Afghan delegation at the 2001 Bonn negotiations had refused to eliminate any of the existing 29 ministries, and instead added one more, a ministry for 3' Ministryof Finance's budget estimates for 1386. 18 Women's Affaid2. In addition to the 25 ministries, there are 11 independent departments, 2 Constitutional agencies and 8 other independent organizations that do not report to any ministry. The current structure o f Government offers further scope for rationalization. For example, there have been proposals within Government to merge the Ministry o f Returnees and Refugees and the Ministry o f Women's Affairs into the Ministry o f Labor and Social Affairs; and to merge both the Ministry o f Religious Affairs and Haj, and the Olympic Committee Department into the Ministryo f Information and Culture. Another proposal was to fold the Ministry o f Counter Narcotics and the National Security Department into the Ministryo f Interior33. Box 2.2: The Civil Service's Size Looking at the civil service as a whole, its size appears small by international standards. The proportion o f the population employed ingovernment (1.9 percent) is lower than the average o f other countries inthe South Asian region. The estimated figure o f 348,000 for a general civilian government includes 138,000 teachers, 10,000 employees o f municipal administrations and approximately 63,000 police but does not include employees o fpublic enterprises, the military and other uniformed personnel. Teachers comprise half o f the non-uniformed civil service; and nearly half o f the non-teacher civil servants are contract employees (agirs). However, these distinctions make little difference to civil service management or the salary bill because agirs have the same status, rights and benefits as civil servants. Equally, the 25,000 employees o f state owned enterprise (SoEs) also have civil servants' status, thus extending the size o f the civil service. Although a majority (56 out o f 65) o f the SoEs has been earmarked for divestiture, their auctions-especially o f their fixed assets-have been unsuccessful. The enterprises that were selected to be restructured and retained ingovernment employ nearly 11,000 employees, with DABM(6,660) and Millie Bus (2,146) being the two largest employers. Source: Ministry o f Finance, Verified PayrollProject. 2.10 This civil service's shape i s a very broad-based pyramid, populated at the base by employees in low-skill positions, at the tip by managers in the higher grades, and with the technical qualified staff distributed in between. The pyramid's broad base and narrow-tip are explained by the pressure on government to generate employment and severe scarcity o fprofessionals inthe civil service and society at large. The narrow tip o f the civil service pyramid may be even leaner than it appears. Following earlier civil service management rules and practices, many officials have been promoted in post, to higher grades, based on their academic qualifications and years spent in each grade. So, there are officials with low slulls occupyingpositions inhigher grades. The recently approved, new grade-pay structure attempts to restore some order to the previous inept civil service structure through classifying government positions into 8 grades based on those positions' job descriptions and responsibilities. The proposed, new distribution i s not expected to radically alter the civil service's shape. It will still generate a very broad- based pyramid, with about three-fourth o f positions at the bottom, not unlike other civil service pyramids inthe South Asian region. But eventhis structure may turn out to be hardfor the civil service to achieve as it i s extremely difficult to find skilled persons to fill technical and managerial positions in the civil 32Rubin2007. 33 Untilrecently, the Ministry of Interior was inchargeofprovincial anddistrict administration,with an unspecifiedcoordination role for Governors. A Presidential Decree on August 30, 2007 establishedthe Independent Directorate of Local Governance (IDLG) as an independent directorate under the Office of the President. The IDLG is essentiallya merger of the structures of civil administrationunder Ministry of Interior (excluding registration of births and deaths - tashkera) and provincial relations (provincial councils) under the Office ofAdministrativeAffairs (OAA). 19 service. In addition, unless donors and government can collectively address the accountability and incentive problems created by the "second civil service"34 then government will continue to be thwarted inits ability to buildin-house competence at these important levels. 2.11 How the Government chooses to deliver services should affect its scope and shape now and inthe future, as well as the level o f competence that i s needed. Apart from the general needs o f fiscal sustainability and transparency, Afghanistan's approach to service delivery would expect to be strongly determined by central government's limited reach and capacity, and even more limited resources and capacity at sub-national levels. These constraints would typically suggest that central government focus primarily on planning, support, and oversight o f service delivery, while delegating the actual delivery function outside o f central government. Since Afghanistan has extremely limited capacity at the provincial level to which central government can delegate, it i s especially pertinent to utilize alternate models such as NGOs and public-private partnershipsto deliver specific kinds o f services, and make use o f community engagement to help monitor and guide performance. 2.12 The difference between service delivery models in Afghanistan's education and health sectors i s reflected in the scope o f government in these areas, although not necessarily in the underlying organization and structure o f relevant ministries. Unlike many other post-conflict countries, Afghanistan very quickly returned to the pre-conflict centralized organization o f education after 2001. The sector's model so far has been to bring schools, including the ones run by N G O S ~into ~ , the ministry's fold, through the provision o f standardized textbooks and curriculum, and by putting all teachers on the government's payroll. Some administrative and financial controls and teacher recruitment are being examined for delegation from the national level to deconcentrated units o f the ministry at provincial, district and school levels. NGOs participation is restricted to where they have a comparative advantage relative to government, such as in-service training for some teachers, early childhood education, and community-based schooling for girls in remote areas36. Consistent with the Education Ministry's retention o f most o f the service delivery functions, it i s the government's biggest employer with 143,000 teachers and 38,000 staff inadministrative and other categories. Untilrecently, it has lagged behind other key ministries in terms o f administrative reforms; resulting in poor planning, budgetary, monitoring and technical (including teaching) capacity. Despite the high level o f donor engagement, financial sustainability will be challenging without greater prioritization or community contributions. There have been no formal evaluations undertaken to assess different service delivery models, and there i s virtually no data on education outcomes. The Ministryhas recently prepared a five-year education strategy and i s beginning implementation o f organizational and human resource management policies designed to improve performance. Along with these, the ministry could usefully examine the scope for increased partnerships with NGOs/communities invarious parts o fthe education process. 34The second civil service i s described in Chapter I(paragraph 1.8) as an army of "externally-funded consultants and advisers, many of themonly looselysupervisedifat all bythe government." 35Currentlyestimatedat between 15,000 to 25,000 schools,manyof which survivedthe conflict years. 36In 4 out of the 14 provinces where the World Bank's Emergency EducationProgramand the on going EducationQuality ImprovementProgramare beingimplemented, NGOshavebeencontractedby the ministryas facilitating agencies. 20 Box 2.3: Usingcommunitiesinstead of governmentmechanisms to deliver education Guatemala's PRONADE, which started in 1992, and now extends to 4,100 communities and 445,000 children inremote rural indigenous areas, i s an example o f government's contracting out the actual delivery ofprimary education. The Ministryo f Education is responsible for strategic planning, financial management, and monitoring and evaluation of primary education. It contracts communities to deliver the service. Each community participating in this program i s represented by a parent-run school committee, COEDUCA. Eligible contractors are communities located at least 3 kilometers from the nearest public school, having at least 25 primary school-aged chldren, and which have located a site and demonstrated ability and interest in managingthe new school. Government funding covers teacher salaries, learning materials and school snacks. Teachers and parents are trained on how to manage their community schools by NGOs, contracted by the ministry. In 2003, PRONADE schools accounted for 15.2 percent of the primary enrollment nation-wide. Fewer PRONADE schools had access to water, latrines and electricity than government schools. However, on average, PRONADE students spent more time in classroom instruction than traditional rural schools. PRONADE students were in school for 180 days during the year compared with much fewer school days o f traditional schools' years. Parental participation in school activities and contact with teachers and principals was higher inPRONADE schools than inthe control schools. Source: Rojas 2005 2.13 Incontrast, the MinistryofPublic Healthhas delegated, through grantsandcontracts, the delivery o f its Basic Package o f Health Services (BPHS) to NGOs in roughly 80 percent o f the rural areas o f A f g h a n i ~ t a n ~ ~s.reported in Chapter (see paragraph 1.12), independent assessment o f the quality o f A health care shows a significant percent improvement in NGO-provided services. The ministry i s also experimenting with contracting in management (called the ministry's Strengthening Mechanism) to run publicly-financed services. In 3.5 provinces, senior managers o f the ministry were recruited at market rates, salaries were increased through the Government's PRR process, and performance bonuses were paid according to the results o f independent assessments. Both quality o f care and coverage have also increased considerably in these provinces where the Strengthening Mechanism i s being im~lernented.~' All inall, and inrelation to other post-conflict countries at similar periods, Afghanistan's health sector is regarded as relatively functional; characterized by a strong policy framework, clear guidelines for service providers, a formal evaluation plan, and reasonable access. The Ministry itself i s regarded as the star performer o f PRR which it utilized liberally to attract well qualified candidates to the civil service, including at provincial levels. It has also benefited from a high level o f well managed national and international technical assistance which have been influential indeterminingpolicy directions and guiding donors. It i s not clear, however, that these radical changes have had a significant impact on the ministry's wider structure and organization. The Strengthening Mechanism model has in part been a result o f continuing belief within parts o f the ministry, especially at provincial levels, that the future o f Afghanistan's health system is public sector service delivery. There i s still some ambiguity about the role o f provincial health offices inthe context o f the delivery o f services, and gaps in capacity at t h s level to monitor NGO performance or oversee vertical disease control programs. Financial sustainability o f both hospital-based services and the BPHS without some consideration being given to user charges, and ministryreorganizatiodstaffing will be a challenge. 37As part of its support to Afghanistan's health sector, the World Bank finances Performance-based Partnership Agreements (PPAs) betweenthe ministryand theNGOs. 38 These and other findings are containedinIslamic Republic of Afghanistan, Ministry of Public Health2006. 21 Box 2.4: Alternate Delivery of Health Services inRwanda While re-building its health services, Rwanda at first re-adopted the district health system model in which provincial and district health management teams in the public sector were given complete responsibility for running health services. In 2001, the annual public budget for health was $3 per capita o f which it is estimated only $1 made it to the frontline facilities which generated 60-80% o f their revenues from user-fees. A series o f contracting efforts were started indifferent parts o f the country. Resources were given to health centers, not as a fixed amount but one based o n performance. An international NGO, working as intermediary, worked with funds from a bilateral agency in two health districts o f Butare province with a combined population o f 388,000 served by 19 rural health centers. These health centers had previously received a fixed budget and per diems used to be paid to health workers regardless o f their productivity. Management committees in each health center signed contracts with the NGO that provided the health centers specified amounts based on the provision o f five key services The services and the amount provided for each one were: (i) institutional delivery - almost US$5; (ii) antenatal care (as measured by tetanus toxoid at least 2 doses) - $0.50; (iii)family planning new acceptor - $2; (iv) immunization (completion o f measles and DTP3) - $1; and (v) curative care consultations $0.10 for each new one. The management committees - had considerable autonomy inhow they used the funds they received. The performance initiative cost an additional $0.24 per capita per year o f which 62% went to the health centers, 27% went inincentives to the district and provincial staff, and 11% to cover the transaction costs o f the new scheme. There were large improvements achieved in the intervention areas o f Butare province in almost all the services designated inthe contracts. The median double difference was 12.9 percentage points. Source: Loevinsohn2008. 2.14 Functional overlap of some organizations complicates understanding the true scope of government. Because of ambiguous and contradictory provisions in the legal fiamework, both the Ministry of UrbanDevelopment and Kabul Municipality claim responsibility for planning, zoning, and buildingregulations within the city. This not only leads to conflictingplanningand development inparts o f the city, but also stalemates in project implementation-both government and donor supported. Resolution occurs on a case by case basis, often only with the intervention of the Mayor and Minister o f Urban Development. Water beingan important subject in Afghanistan, 3 ministries (Energy and Water, Mines, and Agriculture) and one department attempt to control equipment and funds for hydrological studies. The current arrangement i s that when the water inquestion i s potable, the Ministryo f Energy and Water controls the hydrological studies to determine aquifers and water deposits. When the water i s meant for irrigation, the Ministry of Agriculture conducts the studies. The Ministry o f Mines, on the other hand, conducts hydrological studies when it needs soil data to identify mineral deposits and the water data to avoid flooding of mine tunnels and pits. The Department of Environmental Protection also conducts hydrologic studies to determine location of dumpsites and sanitary landfills, avoiding seepage into a water veiddeposit being used by people for drinkmgpurposes. A quick review o f organizational mandateswithin each sector can help identifyoverlaps and lacunae. 2.15 While some government functions are being performed by more organizations than i s necessary or economic, other critical areas of government needgreater coordinationbetween organizations to cover gaps. Eventhoughpolicy making i s the most important function o f the center of government, there exists no institutional arrangement to coordinate policy development across ministries prior to their submission to Cabinet for consideration. The concept o f a Cabinet Liaison Officer for eachministry, as a focal point for the ministry's submissions to the cabinet, and for implementation o f relevant Presidential decrees emanating from Cabinet meetings has been introduced only very recently. Greater coordination between the Ministryof Interior (police) and the Attorney General's office (prosecutors) will help improve rule o f law. Recent laws have generated confusions and tensions by changing the responsibility for criminal 22 investigations. Historically, the police in Afghanistan had been responsible for both discovery and investigation o f crimes, but that responsibility has now been assigned to the Attorney General's office. N o w the police do not have the authority to investigate crime. Instead, they send the file to the Prosecutor, who does not have the capacity to investigate3'. The result is that investigation suffers. Scope of Individual Organizations 2.16 The structure o f individual organizations i s especially important to Afghanistan's scope o f government because structures had become very informal during the conflict years. There was little clarity about the functions o f ministries and agencies, and even less about their constituent units. The Priority Reform and Restructuring program, introduced in July 2003, was the principle vehicle for introducing some restructuring and rationalization o f individual government organizations. As already described in Chapter I,the PRR, was an unconventional initiative for a weak environment like Afghanistan's, but it was the only way possible at that time. The asymmetrical approach allowed higher pay to staff o f selected departments o f ministries performing high priority functions in exchange for restructuring those departments. The initiative was more in line with practice in developed countries, where the sense o f discipline i s ingrained. But in 2002, this was attempted in Afghanistan because there was not much alternative: neither any assigned organization in government to undertake more comprehensive organizational restructuring, nor any organization that had the capacity to undertake the restructuring o f its entire structure4'. 2.17 The PRR program enabled ministries that underwent some restructuringin key departments start to look more like modem ministries, and provided an incentive to begin recruiting staff on a merit basis. It also providedtemporary relieffrom the very compressedwage structure for ministries and departments undertaking important functions. Indeed, the proposed regrading of the civil service and assignment o f new pay scales could not have been considered if the PRR program had not installed, in the first place, some rational organization o f ministry structures and production o f job descriptions. The program encouraged efforts to reform and restructure in at least 32 o f the 44 ministries and agencies, plus several Governor's offices, and approved 44,0004' positions for higher allowances. Some departments and ministries inevitably used this process more rigorously than others. 2.18 The program's success in achieving some degree o frestructuring inretum for higher salaries was diminished when the program was expanded in 2004 to encourage "whole o f ministry" approach rather department by department applications. Because it created inequities within ministries-between departments restructured and not-its replacement by a broad, across-the-board reform had been politically inevitable. The program was concentrated at the national level. Of all positions that have been PRR-ed, 44 percent are located in Even at national level, ministries' weak HR departments were unable to keep pace with recruitment plans; and departmental units located at sub national levels were often excluded during the program's early days. Real reform o f line ministries' provincial departments was hampered by the lack o f IARCSC's presence in the provinces and competence gaps at provincial levels to implement reforms. Later, however, provincial components were added to approved PRR applications o f the Ministries o f Public Works, Public Health, Rural Development and the mustoufiat offices o fthe Ministryo f Finance. 2.19 The restructuring process initiatedby PRR i s far from complete. After completing an initial phase o frestructuring, a number o fministries are now looking at forming units/ task forces that are dedicated to 39Wilder 2007. 40Hakimiet a12004. 41 As of 3 1 July, 2007. 42The Services Group2007. 23 making further refinements o f their organizational structures and overseeing implementation o f the reform process, including improved human resource management practices. These units replace the old PRR task forces whose ad hoc nature and temporary membership by senior civil servants with other responsibilities were not able to sufficiently monitor or guide implementation o f PRR reforms within ministries. The new task forces, popularly called Reform Implementation Management Units (RIMUS), will have their staffing supplemented by bothnational and international consultants for 18-30 months to diagnose and recommend changes. In addition, reform plans are expected to give greater attention to responsibilities at central and sub-national levels as well as to opportunities for improving service delivery or operational performance through business process improvements. IARCSC's Own Functions and Structure 2.20 In spite of being the organization mandated to lead civil service reform and organizational restructuring, the IARCSC itself has conflicting roles and functions. The Commission i s overloaded with having to perform all aspects o f Afghanistan's civil service management, creating a serious capacity crisis. Other Public Service Commissions with much longer operating histories than IARCSC (4 years) and functioning instronger institutional environments than today's Afghanistan, perform fewer functions. Box 2.5: Creationof the IARCSC Even the unusually structured IARCSC almost didn't happen. The Bonn Agreement's original draft proposed a powerful Public Service Commission to process merit-based recruitment at the national, provincial and district level. But all representatives opposed any attempt to curtail their authority to appoint whomever they wanted to government posts. In the end, the Agreement stated that the Commission would "provide the InterimAuthority and the future Transitional Authority with shortlists o f candidates for key posts inthe administrative departments (ministries), as well as those o f governors and uluswals (district administrators), in order to ensure their competence and integrity." The Agreement did not impose on the government any obligation actually to appoint officials from among the candidates on these short lists. Inpractice, the Commissionnever compiled or submitted such lists. Instead it defined its purpose as "reform" o f the system. While the core role was reduced, the IARCSC started perfonning other civil service functions generally performed by line ministries. I Source: Rubin2007 2.21 Two distinctly separate personnel functions are merged within IARCSC: a ministry in charge o f personnel and a traditional Public Service Commission providing the necessary check and balance between government and employees. Functioning as Afghanistan's Personnel Ministry, the Commission i s perceived as a "doer" organization more than a "facilitator" or "oversight" organization. It was entrusted with actual reorganization tasks o f line ministries in the PRR process instead o f being a facilitator o f an exercise ledby line ministriesthemselves. Unsurprisingly,this resulted in overstretching the Commission's capacity and limited success where the line ministry was not actively involved or sufficiently motivated. On the other hand, if the Commission functioned as a real Personnel Ministry, it would logically play a major role in establishment control-not only its own but o f the whole government. But current practice i s that the Ministry of Finance agrees each line ministry's tushkeel only with inputs from the Tushkeelut Directorate inthe Office o f Administrative Affairs but leaving the Commission out. While expected to perform both executive and check-and-balance functions, the Commission enjoys neither executive authority nor independence from the executive. Trying to perform executive functions without being backed by Cabinet-level authority slows it down. Meanwhile, although the Basic Civil Service law states that the Commission i s independent in its performance and directly responsible to the President's Office, inpractice its status i s unclear. Rather, the Commission reports to a Cabinet Sub-committee, with supervision coming from a PAR Steering Committee, the latter including a Deputy Minister and an Advisor to the President. Thus the IARCSC i s supposed to be unbiased inall its 24 actions, while being vulnerable to pressure from within the Government. Finally, there are conflicts o f interest even within IARCSC's own structure. While one part o f the Commission (Appointments Board) handles recruitment, another part (the Appeals Board) can stay those proceedings. Similarly, the Appeals Board can also rule against any directive from the Commission's Civil Service Management Department. These could pit different parts o fthe Commission against each other. 2.22 The IARCSC i s responsible for monitoring civil service reform's progress. This i s not easy for the Commission to accomplish because a majority o f the reform activities are being undertaken outside o f itself and by the line ministries. The ministries themselves, lacking awareness about the usefulness o f monitoring, have no incentives to monitor. Yet, it is in the Commission's own interest to do this systematically and generate useful and timely information, to identify areas o f progress and stagnation, and organizations that are laggmg in the reform effort. This can help IARCSC position itself as an able steward o f the reform effort. Without it, the Commission will find itself explaining and defending all achievement shortfalls at the end o f the Compact period. Currently, the Commission i s refining a set o f monitorable indicators for the civil service reform effort, in line with the Compact Benchmarks. The Human Resource Policy Department (HRPD) o f the IARCSC i s responsible for monitoring the compliance o f personnel management policies and practices across government. At this point in time, little monitoring takes place, although the HRPD sends out monitoring teams to determine the extent to which new regulations are being complied with in ministries. The Appointment Board attempts to monitor performance assessments and appointments at senior levels. The extent to which it can monitor general appointments depends upon the ability o f their staff to regularly observe ministry level recruitments. C. The CivilService's Competence 2.23 Reforming the structure o f government and its various ministries and agencies provides the necessary framework within which the civil service can function, but the best of structures would not result in good service delivery without competent civil servants working in these organizations. This required level o f competence among civil servants was quite non-existent in 2001. Afghanistan had suffered a serious exodus o f trained and experienced persons, and the education system had not replaced those who left or were victims o f violence. Shlls and practices o f those civil servants who had stayed behind had not been systematically upgraded to match evolving demands o f a modem society: rather policies and practices o f the latest pre-reform regime placed obstacles inthe way o f modernization. This left the government at the start o f reconstruction in a very weak position to establish control over its own finances and build public confidence in its ability to serve its redevelopment needs through its own efforts. As in many other post-conflict reconstructions, international technical assistance flowed into the vacuum at a rate that government was largely unable to plan for, or effectively control and coordinate. In addition, donors have attempted to compensate for the lack o f capacity in government to manage programs and projects, through an array o f salary top-ups to civil servants, replacing normal government implementation arrangements with the use o f private contractors and NGOs, and competing away whatever competent staff are available to government in order to manage their own projects/programs. This has effectively created a "second civil service" with perverse incentives for reforms within ministries and unclear accountabilities for those staff receiving salary supplements. The government has since been trying to balance initiatives on a number o f fronts: managing technical assistance; recruiting more competent staff under PRR and other schemes; conducting in-house training to strengthen current employees; and plan for longer term improvements competence through schooling, training institutes and higher education. 25 Managing TechnicalAssistance (TA) toAugment CivilService Competence 2.24 Five years into reconstruction, the challenge now before government i s how to manage TA so that there is sufficient to fill competence gaps, while replacing it with less costly and more manageable in- country-grown and national capacity on the other. OECD's Development Assistance Committee has estimated that donor spending on technical cooperation in Afghanistan reached almost $460 million in 200443 representing more than 20 percent o f total donor assistance to the country. Spealung in Washington during the World Bank Annual Meetings in September 2006, the Minister o f Finance observed that since 2002 an estimated $1.6 billion had been spent on technical assistance and capacity building projects, with little result. This estimate did not take into account the technical assistance provided through development projects and capital expenditure (investment TA) nor TA provided by many NGOs operating in Afghanistan. Therefore, annual donor spending in Afghanistan on technical assistance could well be in the range o f a half a billion U S dollars. This roughly amounts to the government's tax receipts, or about one half o f government's on-budget expenditure. Buildingin-country competence and experience, i s likely to be less costly, but it will require time and has its own risks that are discussed inthe next sub-section. Box 2.6: Types of TA and their perceived strengths and weaknesses 0 TA that is embodied in a capital investment operation, such as site engineering. This practice i s generally perceived positively because it buys international best practice for a specific task for a limited duration with no expectations beyondthe completionand functioning o fthe capital involved. 0 TA that constitutes all or a significant part o f the management o f an investment operation, such as ina Project Implementation Unit. Offering advantages in terms o f focus o n results and experience in project management, it generally creates institutional arrangements that run in parallel with regular government operations, distort incentives, and reduces the sustainability o f the project inits post-investment operations stage. 0 Operational TA delivered by contractors that directly substitutes for civil service inline positions. While justified inemergency situations, it is very costly, distorts the labor market, and discourages capacity buildingby creating dependency. It is also difficult to effectively determine when the emergency has been sufficiently calmed to beginto reduce this dependency. 0 Advisory TA, also delivered by contractors, that advises managers and policy makers but i s not involved in implementation. The distinction between the provision o f advice on performing a function and actually performing the function itself canbe quickly blurred. 0 TA that functions as coaches, trainers, and third party analysts, without involvement in operations or capital investment. This is the least intrusive form o f TA, but its success i s difficult to standardize and measure. Source: Michailof 2007b. 2.25 Dissatisfaction with the result o f TA has been wide and vocal. Among all types o f technical assistance (see Box 2.6), the government is facing the greatest challenge in managing operational technical assistance. The earlier expectation had been that donor-funded consultants would function as advisors while civil servants in line positions would manage the operations. For this reason, the Afghan Expatriates Program (AEP) and Lateral Entry Program (LEP), funded from the multi-donor Afghanistan Reconstruction Trust Fund(ARTF), were expected to recruit knowledgeable Afghans that would retum as advisors, and use their experience outside Afghanistan to introduce modem methods and practices in ministries and agencies. Progressively, however, the real demand for their competence, and that o f most TA-funded consultants, was to act as operational line officers. The persisting need to fill gaps in 43Annual figures are $19 millionfor 2001, $226 million for 2002, $407 million for 2003 and $459 million for 2004. 26 management o f key ministerial and agency functions has now been explicitly recognized in the new Management Capacity Program, which follows on the former AEP and LEP. Making explicit the demand has allowed suitable attention to accountability, transparency and management arrangements for the consultants that are being engaged, although an exit strategy for these programs has yet to be hlly conceived. Apart from this advisory versus in-line dilemma, many had believed that the gap-filling need for civil servants' competence would last 2-3 years,# and that short-term, imported consultants would simultaneously train and coach their replacements while performing in-line operations. These assumptions had been overly optimistic, considering the unprecedented length and depth o f the conflict from which Afghanistan was emergmg and its relatively low level o f development pre-conflict. 2.26 Onthe other hand, gap-filling has beenhighly successful where the objective o f the TA hadbeen clear and focused. Through a series o f grants45, the Ministry o f Finance has received direct operational assistance in managing treasury, budget, procurement and internal/external audit operations. This has allowed the government to absorb and account for an expanding operational and development budget, meeting acceptable international standard^^^. Inthiscaseitwasquiteclearattheoutsetthatthetechnical assistance was intended to maintain a set o f public financial management functions at an international standard for transparency and accountability. Training and personnel development were secondary objectives and achieved little success via this assistance: what did result was an assessment o f training needs for civil servants who would later work inthese public financial management areas. Similarly, TA intended to assist the Government in reforming the structure and organization o f ministries and agencies was clearly mandated and achieved a successful outcome. For example, the PRR program executed through the IARCSC was almost entirely executed by international contracted experts because at the program's start IARCSC was new and not equipped to undertake the mandate. It succeeded in establishing a credible base for further self-directed agency reforms and the proposed civil service pay reform. Acquiring CivilServants' Competencethrough Staff Training and Development 2.27 One o f the earliest initiatives o f the IARCSC was to propose policies for civil servants' training and a strategy to be implementedby its Training and Development Department. This strategy proposed a menu o f targeted skills-buildingactivities, to be undertaken after in-ministryneeds assessments, and to be provided intraditional classroom settings located within training centers inministries and agencies4'. A Civil Service Institute, builtwith donor assistance from the Republic o f Korea has begunto operate. The result o f this effort to remediate basic knowledge gaps has been relatively modest compared to the huge need for transfer o f shlls. The UNDP-supported Civil Service Leadership Development program, for example, i s expected to reach only about 10 percent o f officials in top and senior grades o f the civil service48. Training in basic English and computer skills are reaching a fairly modest number o f individualscompared to the size o f the civil service: current results represent about 15 percent o f staff in grades 5 and above, assuming that only staff inthese grades will use Englishand computers. The cost o f ~ 44The EmergencyPublic AdministrationProject(EPAF') (H-155)financed by IDA andDFIDprovideddirect operational support for treasury, procurement, and audit hnctions of the Ministry of Finance. Its objectives were "to put in place key public administration capacity to facilitate better use of public resources, including donor funds. " Unfortunately, the project also explicitly stated that this couldprobablybe eliminatedinabout 3 years, basedon other post-conflictrecoveryexperience. 45World Bank Emergency Public Administration Project, World Bank Second Emergency Public Administration Project, and World BankPublicAdministrationCapacity Building Projectsupportedby DFID, USAID, UNDP andADB. 46Afghanistan's public financial managementfunctioningwas rated as average and acceptableby the 16 PEFA criteria used to ratepublic financialmanagement performance. 47IARCSC2005. 48As of June, 2006, there were about 5,800 staff in the most senior level positions, graded above 3 in the current scale, while UNDP's programaims to cover 500 officials. 27 scaling up such retail training to establish a critical mass i s high, but perhaps unavoidable given the very limitedpool o fprivate training schools to which this could be outsourced. 2.28 The United Nations Development Program o P ) has introduced a new Coaches and Advisors Program (CAP) as an alternative to in-class training for civil servants. The CAP assigns a coacWmentor to an official to advise on the daily work, operating methods and resolution o f issues, and generally act as a source o f at-hand information that the Afghan official may lack. UNDP's international experience with this approach suggests its suitability insituations like Afghanistan where the government official is new to the job and inexperienced in the management process or becoming confronted with a range o f new situations (as opposed to a single task function), and where there may be uncertainty in how to proceed. Similar programs in Kosovo and Timor-Leste have been short lived, but were generally successful in transferring experience and know-how. Through making the coach / mentor available for only 6 to 12 months, CAP avoids the risk o f the coach acting as line manager as was the case under the AEP. Coaches are provided on demand, currently being sourced from among civil servants in India. Twelve are already in place and five more being processed49. Risks, however, remain because language limits communication, and differences between the culture, background and worldview o f the mentor and Afghan official reduces the sharing o f concepts. 2.29 Inreactionto the general discontent with usingTA and growing awareness that TA is not well managed, government and donors proposed to examine the prospects o f a more coordinated and planned approach to capacity building. A concept note was presented at the Afghanistan Development Forum, April 2007, to provoke thinking on a possible Capacity Development Plan for Common Functions (hereafter Plan) for the Civil Service. The concept o f capacity advanced was fairly broad, and included training, knowledge sharing, coaching, mentoring ,experience building, while recognizing the importance o f a sound enabling environment (uncluttered organization, clear job descriptions, adequate incentives) to allow knowledge and slull to be used effectively. The Plan identified 5 sets o f common functions for which capacity should be built: policy and strategy development, project management, financial management, procurement and human resource management, along with a lead agency or ministry whose legal mandate establishes it primacy in the field. A proposal to establish an Inter-ministerial Capacity Development Committee to monitor and compile information on various capacity building interventions has been made and would be helpful as knowledge sharing. While the concept paper, Plan and an Inter- ministerial Committee could be helpful inraising awareness and the focus on capacity, and serving as an information clearinghouse, there are inherent risks in formalizing this unit. Without a clear definition o f the objectives o f a capacity building policy and standards by which to monitor achievement, the Committee would readily establish bureaucratic procedures and become overly constraining and regimenting o f capacity buildingactivities when these require creativity, flexibility and responsiveness. Acquiring CompetentCivil Servantsfrom Afghan Society at Large 2.30 The civil service inmost countries hires personnel from among graduates o f universities, training schools and the general education system, directly, to fill government functions. The government's role thereafter has been to perfect job shlls required for the efficient functioning o f the service through specialist in-house training. The common belief has been that the Afghan education system has not been able to produce the graduates neededto fill the demand. 2.31 The Ministry o f Education, with international assistance i s attempting to reform the quality and educational achievement o f the basic education system. Additional projects are being launched to raise standards o f the University o f Kabul, and establish a new National Institute o f Management targeting the development o f senior levels of the civil service. The Civil Service Institute i s expected to continue 49Notes and updates from UNDP staff by e-mail. 28 playing a role inbothpre-service training, such as awarding a pre-service certificate for readiness for civil service employment, and post-employment professional straining. Unfortunately, generating candidates from the education system and hiringthem into the civil service, while vital for the long term stability o f government functions, i s at best only a long term solution. This means that current practices for recruitingcompetence (TA and MCP) and in-house training are likely to remain. 2.32 For the school and university system to be effective and efficient inproviding a source o f future civil servants, important issues will have to resolved. For example, the respective roles o f the university system's general courses inpublic administration, the role o f specialized institutes, and the role o fthe CSI will need to be clarified to minimize duplication, particularly in Afghanistan's resource scarce environment. Some general agreement between university, institute and CSI authorities and government on benchmark standards for entering the civil service at different levels would be required to assist in curriculum development and setting graduation standards. To date, little has been done in this regard. It would be incorrect to say that there are no graduates at any level, available inthe market. Over the past two years, the LEP was able to hire more than 300 individuals with adequate qualifications for relatively responsible positions. IARCSC has also been managinga graduate entry scheme. Donor-fundedprojects have been able to recruit national graduates at almost all levels to implement their activities. Most embassies and legations have had little problem in recruiting national staff. Clearly, the Government cannot pay the salaries offered by these agencies. But the Government can certainly manage its own recruitmentbetter by improvedmanagement and wide publicity o f thejob openings in Government. D. Civil Servants' Incentives 2.33 The previous section explored competence as an element o f the civil service's strength. But even competent civil servants, employed in well-structured organizations do not comprise an effective civil service unless their institutional environment motivates their high level of performance and accountability. Therefore, this present section examines whether the laws and rules to manage Afghan civil servants can induce behavior and performance consistent with the goals o f a civil service that i s professional, politically neutral, and unified across government. 2.34 A legally defined civil servants' cadre with predictable terms and conditions, protection from political micro-management and standardized application o f recruitment can motivate both employees and employers to perform according to agreed standards and for agreed purposes. Collectively these take the form o f an ingrained discipline that becomes entrenched in the organizational culture o f public sector agencies, and a political culture where actual behavior is close to the written rules. 2.35 Informality in government administration during the conflict years distorted whatever civil servants' incentive structure mighthave existed prior to the conflict5'. There were n ojob descriptions for civil servants and little clarity about the functions o f different units and agencies. A s a result, at the start o f reconstruction in Afghanistan, there was very little discipline or capacity within the civil service, and no high-level consensus that it should be politically neutral. The Government's response was to stimulate modest, targeted incremental reform o f key functions within organizations, through PRR. As patience with the asymmetric approach waned, and Compact benchmarks have laid down specific standards o f civil service management to be achieved by 2010, Government has turned its attention to accelerating the adoption o f a more comprehensive reformo f civil service management. 50Rubin2007. 29 CivilServants Legal and RegulatoryFramework ' 2.36 The detailed architecture o f civil servants' management i s usually laid down by a Civil Service Law along with secondary legislation providing regulations on how to implement the law. The purpose o f the law and regulations i s to prevent arbitrary actions by the executive and to define the civil service and the qualifications, duties andrightso fthe civil service as well as their working conditions. 2.37 Legislation for Afghan civil servants' management that existed prior to reconstruction did not contain incentives for its consistent application across government5'. There was considerable confusion over which laws were actually applicable, creating poor incentives for organizations to follow any law at all5'. No institution was given clear responsibility for managing the civil service, with human resource management functions, including recruitment, promotion and staff placement spread across different organizations. This created an incentive for ministriesto apply rules and regulations inan ad hoc manner. The legislation did not encourage Government to focus on its core tasks and employment within government was regarded as a homogenous whole. As such, there was little effective distinction between staff having different skdls, and no incentives for qualified persons to take up government employment. The elaborate system o f allowances set out in the legislation, and the nature of the grading system (as described below) meant that all staff were remunerated at effectively the same l o w level, reducing incentives for individual performance and career advancement. 2.38 InFebruary 2005, the President andCabinet approved a Basic Civil Service Lawvia a legislative decree Cfurmann taknini). It established the principle o f open competition and merit for all civil service appointments; set up the Independent Administrative Reform and Civil Service Commission to be responsible for civil service management; and specified civil servants rights and obligations in a way that provided a clear distinction and separation o f the rights from the duties o f the civil servants. While these somewhat assured civil servants by detailing their expectations and clarifying their responsibilities, the law lacked some critical accountability-strengthening provisions such as conflict of interest, civil servants' participation inpolitical affairs, and their regular performance appraisal. These provisions had been dropped from the final version o f the law that was approved. A Code o f Conduct for civil servants has been approved by Cabinet (with a Dari pamphlet version distributed to ministries) and appears to address conflict o f interest concerns butnot political impartiality. 2.39 The Civil Servants Law, separate from the Basic Civil Service Law discussed above, i s intended to provide guidance for the implementation o f pay and grading. The current draft provides scope for misinterpretation. It appears to reinforce the principle o f promotion within the civil service being based upon qualifications and longevity rather than on responsibilities and task, and allows for a considerable array o f allowances, prizes and promotions without any indication o f how their cost would be accommodated. T h i s i s in contradiction with the policy on pay and grading that the Cabinet earlier approved. After the law i s passed, key regulations and procedures will needto be finalized; otherwise the law will remain only a lofty ideal. Some regulations have been drafted but not issued (appointments, appeals, performance appraisal). Others still need to be drafted (promotions, transfers, discipline, and implementing the new grade-pay structure). There i s little capacity within government to implement new features o f the law. For example, the law requires ministries, agencies, provincial and district offices to assume the full range of civil service management functions. But, Human Resource (HR) units in organizations have been primarily concerned with personnel administration and record keeping and are inadequately staffed to guide the implementation o f new policies and procedures. Although an ideal Arrangementsfor managingthe Afghan civil service stem from the 1964Constitutionand includethe 1970 (SY1349) Law on the Status and Condition of Government Employees as amended by the 1977 (SY 1356) Decree no 143, and relevant parts of TalibanEmployment andLaborRegulationsandPersonnelLaws. 52Evanset a12004. 30 organizational structure for HR units within line ministries has been developed along with a training schedule, this has not been implemented across the civil service thus leaving many organizations with little incentive to implement the law. Merit Basis in Appointments 2.40 The merit principle in appointments i s at the heart o f effective civil service management. It provides an incentive to individuals to come forward and be considered for employment within the civil service, regardless o f any particular allegiance or affiliation. It also provides the institutional mechanism through which employers can recruit the best person for the job and make the best use o f government resources. Selection criteria for individual positions are specified, widely communicatedand contestable: failure to appoint a candidate can be appealed and reviewed against explicit specifications for the position and prescribedprocedure. 2.41 The expansion o f the state in Afghanistan inthe seventies did enable some degree o f merit based recruitment into the civil service to take place, even though subsequent assignments and promotions were largely dependent on patronage and c o ~ e c t i o n s ~But the volatility o f the subsequent three decades ~ . eroded the limited trust there had been with merit based recruitment as competing political factions needed loyal and known representativeswithin government. The lack o f transparency inthe recruitment process left staff dissatisfied and demoralized, and contributed to the overall capacity depletion of the civil service. 2.42 The merit principle's success depends on public perception about the credibility o f the recruitment process. There has been significant progress in this area, especially considering the wide acceptance o f patronage as a basis o f civil service appointments. Duringpeace discussions in Bonn in 2001, ministerial positions and the power to appoint civil servants were seen as legitimate spoils o f victory by various political factions who did not want to see limits placed on their ability to use the civil service for political gain. Subsequently the Civil Service Appointment and Appeals Boards have been established and the principle o f merit based recruitment defined through the 2005 Civil Service Law. A comprehensive set o f Appointments procedures govern the system o f appointments to the Afghanistan civil service. They have been regularly updated to reflect changes inthe Civil Service L a w and provide a firmframework to support merit-based recruitment. 2.43 The Appointments Board still faces considerable pressures from ministers and other leaders to appoint specific candidates, and the process and outcomes o f the merit system itself have been subject to considerable criticism across government. Some lacunae arise from the Appointments Board's insufficient technical expertise. Incomplete job specification by requesting ministries, insufficient advertisement, the Board's poor interview process, and the Secretariat's inadequate reference checking have enabled ministries to favor incumbents. The Appointments Board's monitoring o f appointments below the top grades, where the vast majority o f appointments are made, i s neither robust nor well understood, and enables the process to be bypassedby ministries infavor o fpatronage appointments. The Board i s not sufficiently empowered to resist these pressures. Original plans to fully establish its independence through the law have not been followed through. The only recourse at the disposal o f the Board i s the power to report to the Chairman on unlawful appointments; and members' freedom to resign from their positions iftheir advice i s not followed. Meanwhile ambitious quantitative targets threaten the Board's incentive to concentrate on quality in the face o f mounting pressure to clear the pipeline o f PRR appointments. Since September 2003, when it started its work, the Appointments Board has appointed over 1,400 candidates for grade 2 and above and monitored over 7,500 appointments at grade 3 and below. The Appointments Board's anticipation that at full capacity it can oversee some 10,000 53 Rubin2007. 31 appointments per year at senior andjunior levels seems not only wildly ambitious, it is also significantly short o f those expectedto be appointed or reconfirmed duringthe pay and gradingprocess. Box 2.7: Vulnerabilityto CorruptioninAppointments As the Appointments Board's volume o f work i s expected to increase rapidly when the new grade pay structure's implementation starts, its vulnerability to political pressure and corruption opportunities was independently assessed in January 2007. The report c o n f i i e d an earlier evaluation's finding that significant progress had been made. But it also highlighted that there was a deep dissatisfaction and sometimes strong skepticism about the way in which the overall merit-based recruitment process is handled; and especially, perceptions of: Inadequate and poorly respectedprocedures; Inadequate experience, competence and training o f Commissioners resulting in variable application o fprocedures and inappropriate behavior; Narrow bureaucratic decision making where common sense and experience should drive the process; Ethnic and politicalbias; 0 Patronage by IARCSC itselfand "men of influence;" and Corruptioni s o fprimarily, but not exclusively, an individual (petty) nature. To remediate issues raised inthe two reviews, the IARCSC has adopted a comprehensive actionplan. It includes future de-concentration o f senior appointments and a monitoring mechanism to allow evidence- based evaluation; and strengthening the Appointments Board's t e c h c a l competence. Source: Michailof 2007a. Gradingstructure andpay scales 2.44 When civil service grade and pay structure i s rational and predictable, it assures officials within the system, and observers outside, that assignment to positions i s according to the predictive ability o f the official to do the job, and it reduces arbitrary decisions and patronage in civil servants' pay. Like compensation for any other group o f employees, civil service pay i s an important determinant o f employees' motivation. InAfghanistan, the civil service's pay structure is particularly relevant because it determines whether or not the civil service i s able to attract and retain skills from the limited local pool. 2.45 The inherited grade-pay structure in Afghanistan did not provide adequate performance incentives for civil servants. It emphasized longevity rather than competence and responsibility and focused managers and individuals on the inputs (qualifications and seniority) and process o f jobs rather than outputs. Grades were applied randomly and inequitably across ministries for similar functions. For example, under the current system, a Director o f Finance in one ministry can have a different personal grade than a Director o f Finance in another ministry, based upon longevity o f service, rather than the job's content. An employee can remain inthe same post and receive a grade increase every 3 years until he/she reaches "beyond grade" . This "rank inperson" grade and pay system, as distinct fiom internationally more common "rank in post" systems, has resulted in severe distortions in the grading structure, with many staff in lower level posts earning higher salaries than their supervisors. The distortion i s partially masked by the acute wage compression that has prevailed, but it i s a serious impediment to the development o f a more effective and responsible management ethos and to the ability '`Except for promotion from grade one to "above grade" and from "above" to "beyond" grade, each of which requires four years service. 32 o f government to attract, motivate and retain qualified staff. Pay scales for qualified staff are completely out o f line with prevailing market conditions and severely compressed. Because base pay had essentially remained unchanged during the conflict years, a complex set o f allowances, with food allowances being the predominant, had served to lower the compression ratio to around 1.01-compared to the regional average o f nearly 10. The seriously compressed pay scale provides no incentive for performance or career management. This in turn has created incentives for donors to intervene with a wide range o f ad hoc top-ups and supplements, providing further distortions inthe incentives system. 2.46 The new grade and pay structure, approved by Cabinet on June 18, 2007 i s intended to end the practice o f personal grades, make grades consistent with job content, reduce the need for ad hoc pay arrangements, and allow government to substantially raise salaries at the top end so as to recruit and retain experienced and effective staff. The proposed salary scales range from $100 to $650 per month. Although significantly higher than current senior salaries, there will continue to be cases where even the highest salary will not be adequate to attract the required skills and compete with the salaries offered by donors. Schemes such as the new Management Capacity Program will be requiredto deal with such extra- normal appointments over the medium-term. Unusuallythe top end o fthe civil service scale remains well below that o f the police and the militaryscales. This decision was taken on grounds o f the extra security risks affecting those personnel, however poses risks overall for the government, in terms o f relative strengths o f these arms o f government. II L Box 2.8: Old and new grade paystructures Old II o w pay, hlghly compressed pay scale- I Higher ~~~ salary andNew ~ proportionate with little difference between hghesdlowest the workload 12 grades based on qualifications and Less cumbersome 8 grades based o n length o f service job responsibilities and duties Promotionbasedmore on length o f Promotion based o njob performance service, and less on quality ofperformance and achieving results Larger number of employees, low output Reasonable number o f employees withimprovedoutput Inadequately qualified and skilled staff Mechanisms to appoint better aualified and comuetent staff 2.47 The policy has allowed for a limited number o f special cases for expanded pay scales in individual grades where there i s a market related premium on particular slulls, such as an expanded pay scale for grade-four doctors. There have been demands in Cabinet for such higher premiums to many other groups o f employees, including education professionals. In addition, there have been pressures from within traditional segments o f the civil service for retention o f old practices - such as promotion basedupon longevity. Inthe current fluid political environment these pressures are inevitably difficult for the IARCSC to resist but their acceptance will seriously weaken the technical integnty o f the reform. 2.48 The new structure's test will lie in its implementation, and whether the higher pay will provide the right incentives for improved performance and service delivery. As reported in paragraph 2.35, the current version o f the Civil Servants Law could issue implementation guidance that contradicts the provisions o f the new grade pay structure. Introduction o f regrading and new pay scales has been postponed by a year to 1387. This will require compressing the implementation timetable if its synchronization with the MTFF i s to be maintained, and covering 25% o f civil servants in one year. However, the IARCSC has not yet discussed any implementation plan with line ministries and donor partners. Capacity within IARCSC and line ministries i s very limited in terms o f ensuring that ministry structures, job descriptions and capacity are in place at the right time in order to implement the new 33 structure, including appointments, in a systematic way. With elections scheduled inAfghanistan in less than two years from now, the same pressures that led to some units under PRR making superficial restructuring plans, abandoning the rigorous recruitment process, and migrating all existing staff to higher pay scales4ould equally derail any logical and systematic roll-out of the new scheme. IARCSC will need to ensure frequent consistency checking as to the application o f the new grading across ministries. Without a centralized HR information system, this will be challenging. Ministry o f Finance could make more explicit its role as a gatekeeper to various ministries receiving new pay for re-graded positions so as to ensure that the overall application o f the new pay i s fiscally sustainable. Finally, proposed salary scales will still be below those prevailing in the "second civil service" and paid directly for by donors, and unconstrained donor actions could undermine the intent o fthe policy. . Performance Evaluation 2.49 Performance evaluation can be a key aid to good management in the civil service. International good practice requires that performance management should be task-oriented, based on results, and measuring results against pre-defined goals and targets, rather than an assessment o f traits. Traditionally, performance evaluation, if conducted at all inthe civil service, used a trait-based methodology - loolung at behavioral factors such as neatness, punctuality etc. The employee typically did not participate and was often not told what rating they received. 2.50 A Performance Appraisal procedure and related forms for managers and employees have been drafted and piloted by the HR Policy Department o f the IARCSC based on the principles o f international best practice (based upon the individual's job description, allowing for employer and employee feedback etc). Ifthe individual civil servant is unhappy with the appraisal he or she has the right o f appeal to the Appeals Board. The updated appraisal forms and process i s only for merit-based appointments (PRR-ed positions). For non-PRR-ed positions, the traditional method o f assessment will be used. Roll out o f the new staff appraisal system across the Civil Service was originally scheduled to take place within two months o fthe enactment of the new Civil Service Law, buthas been delayed. 2.51 Incentives do not appear to be strong for either ministries or individuals to implement the current performance appraisal mechanism across government. Both the pilots and early implementation have been difficult, technically and culturally. Managers and staff have no experience o f meeting together to discuss and assess performance, and there are considerable cultural barriers to conducting an objective and honest face to face assessment o f a subordinate's performance. Many ministries report the system as beingcomplicated to manage with staff laclung understanding as to the purpose and process, despite the workshops that have been held and are planned (IARCSC 2005b). Disciplinary and reward mechanisms associated with the performance appraisal process are not clear, and the process itself i s technically difficult in cases where there are no clear job descriptions with associated performance standards. The lack o f fimctioning HR departments in ministries is a critical constraint to understanding and implementationo fperformance evaluation. Exit Management 2.52 Job security and pension have been inherent attractions o f the civil service and have partially compensated for pay levels lower than the private sector's. Pension prospects encourage long careers in the civil service. A workable and affordable severance scheme (that might also encompass retraining, re- sktlling and redeployment) can provide a useful incentive to organizations to restructure and reform their functions and enable the recruitmento f more appropriate staff to new functions. 34 2.53 Afghan civil servants' pension, along with that o f the military and police, is administered by the Pension Department under the Ministryo f Labor and Social Affairs. The legal framework o f this scheme is defined by PensionRules and various decrees issued from time to time. The scheme operates on a Pay As You Go basis and is funded by individual contributions and, significantly, by transfers from the state budget. Inorder to receive the regular old age benefit, a civil servant must have accumulated at least 10 years o f service by the time o fretirement. The normal retirement age i s 65 for men and 55 for women. 2.54 The rate o f pension, being dependent on the base pay, and excluding various allowances, i s very low providing no incentive for individuals to consider a career inthe civil service, or for those within the service to retire on time. In addition, the ad hoc adjustments in benefit administration and method o f paying arrears have resulted invery different absolute pension levels for individuals retiring at nearly the same time--creating further disincentives and dissatisfaction within the civil service. For example, civil servants that retired before and after 2003 in the same rank and with the same length o f service receive different pensions, on account o f what length o f service i s considered as pensionable. Wealmesses in administration, including record keeping, and a disconnect between the process o f collecting employees contributions (managed by the Ministryo f Finance) and the process o f assessing entitlement (managed by the Ministry o f Labor and Social Affairs) have further undermined the usefulness o f pensions as an effective career incentive. 2.55 Staff being considered for redundancy i s the responsibility o f the Ministry of Labor and Social Affairs. The legislationregarding severance of staff i s unclear. The February 2007 Labor Lawprovides a basis for severance policy but it also refers to external provisions that are yet to be derived. There i s no determination o f who will pay employees "on suspension" and no referral to centralized budgets for the purpose o f payments. In principle, and traditionally, such staff join a surplus list maintained by the ministry for six to twelve months while attempts are made to find them employment inother ministries. They receive "in-waiting'' payments for this period, which amount to a flat rate six month salary payment for retrenched staff. Under PRR, unqualified staff are assumed to have been made surplus and suspended. A Severance and Pensions Proposal estimates the number o f civil servants "suspended" under PRR at 5,000. It i s not clear whether any o f these civil servants have received any payments at all, and those ministries who have suspended staff have no allocatedbudgets to pay any ongoing salary. 2.56 Inadditionto providinganenhancedcareer incentive, there isa strong fiscal incentiveto urgently reforming the pension scheme. Direct budget allocations have constituted between 75 and 81 percent o f the funding source for the pension scheme in the last two years. Assuming a conservative 3 percent discount rate, the World Bank estimates that the pension debt for retirees i s 2.4 percent o f GDP; and outstanding pension debt for active employees+urrently 3.8 percent o f GDP-will rise to 27.7 percent o f GDP following the new pay structure. As the pay reformprogresses and individuals retire with a new reformed base pay, pension expenses will start growing and may increase 4 to 5 times within the next ten years. Ifat the same time pensions of the current retirees get adjusted to be at par with the benefits o f the new retirees, the immediate effect may be an 8-fold increase in the current pension spending for civil servants. Inorder to make any new scheme fiscally sustainable there will need to be changes inboth the distribution o fthe Employee/Employer contribution shares and the formula for benefit cal~ulation~~. 2.57 Both Ministry o f Finance and IARCSC have incentives to put in place severance arrangements that are fiscally affordable, encourage the right people to leave the civil service, are fair and reasonably attractive, and can be utilized inthe longer term for subsequent early departures from the civil service. A draft regulation and proposed benefits package for newly redundant staffing was prepared in 2005 but n o action appears to have been taken on this. The Ministry o f Finance has estimated (GOAMinistry o f Finance 2006) that some 40,000 existing civil servants would not be qualified for their own, or any other 55Palacios and Sluchynskyy 2007. 35 position within the administration and would need to be retrenched and receive severance. It had earlier estimatedthe cost o f severance payments would amount to $1.8 million, $3.24 million, $3.64 million and $3.69 million over the four-year period o f implementation. A more recent analysis o f civil servants length-of-service statistics by the IARCSC indicates that the cost of the initial `in waiting' payments for these staff would be closer to $34 million based upon existing legal provisions. 2.58 Pension and severance-policy and administration-need to be finalized alongside the new grade pay structure. Indeed, the new grade and pay structure cannot be implemented without reforming pension and deciding severance policy at the same time. Various proposals for retraining, re-skilling and redeployment have been discussed. Beyond that this matter has not received attention from Government. Government i s now preparing an issues paper, that would look at the various options including early retirement, severance payments (lump-sum, flow, mix), retraining and re-skilling (for in and out o f service employment), inaddition to changes inthe pensions policy and administration. Any final package of options considered by the government, particularly inthe severance area, needs to be administratively simple. E. Direction of ReformActivities 2.59 The analysis in sections B, C and Dshows that some progress has been made towards meeting the Compact's goals. But progress in both areas-scope and strength o f government-has been fragile and requires sustained follow through. Through the Compact benchmarks, the Government has determined that the reform it will adopt will be comprehensive and cover all ministries. This is understandable but ambitious inthe context o f disparate capacity and leadership across ministries. Box 2.9: World Bank's support for Afghanistan's civil service reformagenda Starting with the establishment o f IARCSC, the World Bank has been supporting Afghanistan's civil service reform since the beginning o f reconstruction. It has helped the Government and IARCSC develop an action agenda for the rest o f the Compact period, estimate investment needs to meet the Compact targets, and indicators to monitor reform's progress. Donors are now aligning their support for elements of the commonly agreed action plan. Organizational restructuring and rearrangement o f civil servants' incentives are being supported by the World Bank's Civil Service Reform Project. This includes support to a number o f key ministries as they further their restructuring efforts, support for implementing pay and grading across government, and further development o f and training in new human resource management practices within government, including merit based appointment. Meanwhile, the Bank also helped shape the Management Capacity Project which provides ARTF resources for approximately 240 highlevel managers within government duringthe next 3 years. The World Bank's support for civil service reform is not limited to assisting the IARCSC. Civil service strengthening measures are part o f the Bank's ongoing interventions inthe Ministries o f Finance, Education, and Reconstruction and RuralDevelopment. 2.60 The overall conclusion about government's scope i s that it offers the opportunity for less government. Scope, interms o f number ministries, i s unlikely to undergo serious reduction by the end of the Compact period. While there are technical reasons for recommending fewer ministries and agencies, social and political realities o f governing the country are unlikely to allow for a significant compression o f ministerial ranked agencies. Nevertheless, the issue could continue to have relevance indiscussions o f reducing the cost o f government. The scope o f government, in terms o f the functions o f individual ministries and how they organize themselves to deliver those, does provide greater room for modification. The two largely different models o f service delivery shown by the Ministry o f Education (sole service provider, employing all personnel as civil servants, some de-concentrated management) along with 36 contrasting model inMinistryo f Public Health (ministrystaff mostly performing core functions, multiple service providers, worlung under a central strategic andprogram leadership) offer alternative concepts for other ministries to consider when deciding how to meet their client's needs more effectively. The present reconstruction phase allows Afghanistan the luxury o f a clean slate and comparatively few unwanted practices to undo when compared with other countries o f the regon. However, those choices need to be tempered by Afghanistan's severe resource and capacity crisis. Modifying the scope o f individual organizations may be a more promising avenue for reducing the scope o f the government as a whole than compressing ministry numbers. But the risk i s that IARCSC, currently responsible for supporting and overseeing ministryreorganization, may interpret its mandate to mean controlling ministryreorganization byoverly conditioningits assistance, andproscribe methodologies and operational models that infact will discourage ministries from considering alternate options for achieving operational effectiveness and cost efficiencies. 2.61 The conclusion about civil servants' competence is that an ideal solution to the competence problem-currently framed as training a larger number of regularly employed nationals to perform an increasing number o f key functions-is unlikely to be established before the end o f the Compact period. In-house efforts for developing the competence o f existing staff, and developing a broader pool o f individuals employable as civil servants are important initiatives to be pursued, but will not deliver the critical mass o f competent civil servants until well beyond the Compact period's end. Therefore, despite the disappointment with the cost, co-ordination and control being exercised over the use o f much of TA, it i s likely to be needed until proven alternatives start providing the same level o f competence. The deployment o f consultants can be made less problematic through making their terms o f reference and accountabilities more precise, the sources more competitive (for example, using more regional and national contractors), but it will not remove the need for their continued use. 2.62 The conclusion about the civil service's incentive framework i s that vigorous attention to making improvements inthis area offers the best short-term opportunities for strengthening the quality o f the civil service within the Compact period. The introduction o f formality via new laws has driven reform in the right direction, butitneeds strong follow upwithregulations andguidelines, and strong monitoring o fthe compliance o f ministries with these regulations. The application o f the new pay and grading policy as intended would be paramount in this effort, but needs to be supplemented by an effort on the part o f donors and government to address the disincentives created by the continued practice o f salary top ups and the existence o f the "second civil service". Some clear guidance on the extent o f donor top-ups in government and remuneration o f donor project staff worlung within ministries would go some way to limitingthe most objectionable practices. The new laws andregulations should be clear that the IARCSC has authority to hold ministries accountable for applying the regulations and specify sanctions for non- compliance. 37 Box 2.10: Simplification of Vehicle Registration inKabul Pre-existing conditions. The process for vehicle registration in Kabul typically involved: 30 administrative steps; more than one month to complete; visits to multiple offices at various locations; andpayment ofbribes at most steps inthe process. Reform eforts. The "Simplification o f Administrative Processes Commission" established by the President undertook a four phase cycle o f diagnosis, process redesign, implementation and evaluation. This methodology could be applied to the simplification of other key processes. Results. A one stop shop has been established; the number o f steps has been reduced; processing time has declined from around one month to two days; volumes o f applications processed and revenues have increased; the unofficial cost associated with bribes has been dramatically reduced; andstaffmotivationhas improved. 2.63 The approach to building an effective civil service can benefit from two changes from the present direction. First, government's credibility will have to be demonstrated strategically and deliberately. Without some quick wins, status quo will remain the norm past the end o f the Compact period. The government can create "islands o f excellence" even within a weak civil service, as it did with the simplification o f the vehicle registration process in Kabul (see Box 2.10). IARCSC's quick response to the Appointments Board's vulnerability to corruption (described earlier in Box 2.6) bolstered its credibility. Further quick wins could come from improvements in reporting and accountability, and simplifying selected administrative procedures. At the sub-national level, quick wins could also come from rapidly improving administrative capacity in Governors' offices and municipalities. Increasing women's access to training could increase their numbers in Government and demonstrate Government's commitment to reform. Second, the concept that any one organization e.g. IARCSC can build an effective civil service largely through technocratic means should be rejected. Instead, visible, high-level political commitment for civil service reform i s needed, along with patience for a sustained effort to overcome resistance from those who lose from the changes. This i s even more relevant in Afghanistan than inmany other places because previous administrative traditions were not based on merit or formality. Technical solutions and best practice models are helpful. But even when timely and well-coordinated, they alone cannot transform the public service. 2.64 In addition to these more strategic requirements, the action that can most effectively strengthen the civil service i s sound implementationo f the new grade pay structure. This needs a pragmatic plan that takes available resources and competences into account, a strong leadership to manage the risks and coordinate all strands (e.g. new pension and severance schemes, civil servants' database) into the task, and focusing resources as a priority on those ministries with the greatest likelihood o f success and where the outcome i s likely to be most significant interms o fpotential improvements inservice delivery. 38 111. IMPROVINGLOCALGOVERNANCEAND SERVICEDELIVERYs6 A. Introduction 3.1 Civil service reforms, as described in Chapter 11, will inevitably impact on sub-national administration, as ministries devolve responsibilities and strengthen their capacity at the provincial and district levels. However, there are many other actors that influence the way government works at the sub- national level, albeit in different ways in different provinces and sectors. There are also powerful informal power structures at work, which often complement, undermine or even replace government services. Attention to these sub-national issues has risen markedly over the past year - indeed, as this report has been prepared. Questions and concerns raised by the Government o f Afghanistan, the donor community and policy analysts reflect three main developments: A number o f sub-national institutions mandatedby the Constitution (such as Provincial Councils) have been established, lifting expectations but also creating confusion about their roles and responsibilities. Insome cases - such as District and Municipal Councils - their introductionhas been deferred indefinitely, creating a constitutional and institutional vacuum, and raising questions about how this is to be filled. While numerous local investment and service delivery programs have been rolled out, problems with coordination and alignment have impactedon their effectiveness. Moreover, security-dnven investment approaches (often emanating from PRTs), which focus mainly on the rapid creation o f physical assets using non-government systems, have sometimes conflicted with government institutions and processes. And, where broadly successful programs have generated institutional capacity at the sub-national level, key choices concerning the future of this capacity now have to be made. What, for example, should become o f the thousands o f CDCs established under NSP when their current fundingruns out? The continued wealmess o f sub-national structures in certain parts of the country has allowed militia commanders and criminal elements to assert themselves, undermining local governance and service delivery, and aggravating security problems. This has brought home the point that the key question surrounding the character o f the state inAfghanistan i s not that of a strong center versus a strong periphery. Rather it i s whether the current de jure (and hollow) state at the periphery can become a de facto state at the periphery; whether robust and institutionalized systems o f government can be expanded and deepened within the provinces and below. 3.2 One o f the key constraints to strengthening the sub-national system in Afghanistan has been the absence o f a clear policy framework regarding its desired institutional structure, and a strategy to guide actions to realize it." While the Constitution and other legislation provide a legal framework for a number o f sub-national institutions, such as Provincial and Municipal Councils, this tends to be broad and ambiguous, and often begs more questions than it answers about who should be doing (and i s accountable for) what, the resource flows that are necessary for the effective exercise o f powers and functions, and the implementingmechanisms that needto be put inplace. The costs o f these lacunae have mounted steadily 56 This chapter draws heavily on the background paper prepared by Roland White and Sarah Lister on Service Delivery and Governance at the Sub-NationalLevel inAfghanistan, July 2007. A draft ofthis paperwas presentedinKabul inJuly 2007, and the feedback from that discussion, as well as additional inputs on municipalities and CDCs, have been incorporated into the analysispresentedinthis report. 57 The Government of Afghanistan recently pointed this out, identifying a need for "a strategic institutional framework to improve the operating capacity o f sub-national governing units..... [with] a need for higher clarity on the roles, functions, relationships and resource allocation for the various entities within the provincial administration. See Islamic Republic of Afghanistan 2007a, 5. 39 and have become increasingly visible to all. Both government and donors need greater clarity on the basic structure o f the sub-national system, how its different elements fit together, how best to allocate roles and responsibilities between them, where to focus fundingand institutional efforts, and which sorts o fplanning, coordination and implementingmechanisms to support. 3.3 This chapter is an initial effort to address this question. It builds on the considerable analytical work that has already been done on sub-national issues and seeks to add value infour areas: (a) to provide additional information and analytical insight into the nature o f sub-national service delivery and governance in Afghanistan; (b) to develop a simple and coherent fi-amework regarding the sub-national system for discussion with the government and donor community; (c) to propose a number o f concrete initiatives that could be taken in order to strengthen the sub-national framework as proposed; and (d) to make some procedural recommendations to move things forward. B. The Sub-NationalInstitutionalStructure5* 3.4 Territorially, Afghanistan i s divided into 34 provinces, 2 o f which were created by the Karzai administration immediately before the Presidential election o f October 2004. There is disagreement about the number o f districts, with different ministries and government agencies disputingboth overall numbers and boundaries. As o f April 2007, the Central Statistics Office considered there were 364 rural districts, and 34 provincial urban centers, some o f which are sub-divided into nahia or urban districts. The Ministry of Interior and Afghanistan Geodesy and Cartography Head Office (AGCHO) recognize different numbers and boundaries. There i s also disagreement about the number o f municipalities. However, a commonly agreed figure i s that there are a total o f 217 provincial and rural municipalities with a population o f more than 5,000 people. According to the household listing exercise (2003-2005) there are an estimated 40,020 rural villages, but some ministries have very different estimates (some as low as 25,000 to 30,000). 3.5 As administrative units were mostly formed for political, rather than technical, administrative reasons, the population and land area o f each o f these units vary enormously. They also vary significantly interms o f terrain. All ofthis has implications for the development of the sub-national systems and the delivery o f services. The largest province, Kabul, has an estimated population o f around 4 million, and the provinces o f Ghani, Herat, Helmant and Nangahar all have estimated populations over 1 million. Nimroz and Panjshir, on the other hand, have populations o f less than 120,000. The situation i s the same for districts - with districts in the same province ranging from populations o f less than 6,000 to nearly 100,000. Administrative units also cover vastly different geographical areas - with estimates suggesting Helmand covers more than 57,000 sq. km, whereas Panjshir covers less than 4,000 sq. km. Some rural districts cover less than 50 sq. km, whereas others are nearly 600 sq. km. 3.6 The institutional structure o f sub-national governance and service delivery inAfghanistan can be divided into four main levels: two o f which (provinces and districts) are hierarchical sub-components o f the central government; and two o f which (municipalities and village-level structures) have a certain degree of autonomy, at least informal terms. a) The provinciallevelcomprises: 0 Provincial Line Departments (PDs), with basic service delivery responsibility in key sectors (health, education etc.); 58For brevity, this section does not pursue a detailed description of the main elements for the existing institutional structure and the key issues that each of these confronts. This may be found inWhite and Lister 2007, Part 2. 40 Provincial Governors (PGs), appointed by the President and until recently supervised by the Ministry o f Interior, with considerable formal roles in areas o f MOI's responsibility (e.g., security, police) as well as informal powers, including sign-off on expenditures, participation inprocurement andcertainpowersofappointment; Elected Provincial Councils (PCs), with advisory, monitoring and conflict resolution roles; and Provincial Development Committees (PDCs), with limited coordinating and planning functions. b) The district levelcomprises: 0 District Offices (DOs) o f some central ministries with limited functional responsibilities; and District Governors (DGs), selected by the Appointments Board (for most central provinces and Bamiyan) or MoI, with a relatively limited formal (but sometimes important informal) role. c) The municipal level comprises municipal administrations led by mayors, currently appointed by the President, with functional and service delivery responsibility mainly for urbanservices. d) The village level comprising a range o f traditional governance entities (such as shuras) and the newer CDCs. CDCs (currently in about two-thirds o f the country) are elected according to different systems in different areas, and are responsible for local infrastructure development and some local planning and administrative functions. 3.7 Provinces and districts are graded, determined largely by population, but with the influence o f additionalpolitical factors. Grade 1i s the largest, and Grade 3 i s the smallest. The grade o f a province or district affects the size o f the governor's office, as well as the position grades o f the staff. This formal governmental structure i s complemented by a large additional service delivery and development infrastructure, includingNGOs, PRTsetc. 3.8 The 1382 Constitution provides the over-arching legal framework for sub-national government, establishing Afghanistan as a unitary state, but stating (Article 137): "The government, while preserving the principle o f centralism, shall delegate certain authorities to local administration units for the purpose o f expediting and promoting economic, social and cultural affairs, and increasing the participation o f people in the development o f the nation." It mandates the establishment o f elected provincial, district, municipal and village councils, and the election o f mayors. 3.9 The legal framework below the constitution i s rather unclear, with a variety o f old and new laws and decrees which impinge in different ways on the institutions o f sub-national government. The 1379 Law for Local Administration (issued under the Taliban) made minor revisions to various earlier local government laws. Revision of this law has been underway for several years but is currently stalled as key issues are sorted out. Relevant new laws include the 1384 Provincial Council L a w and its recent revision, the 1384 Civil Service Law, and the 1385 Procurement Law. A recent by-law has also started the process o f institutionalizing the CDCs. InJuly 2007, the GOAmoved to strengthen the central oversight o f sub- national issues with the establishment o f the Independent Directorate o f Local Governance (IDLG), reporting directly to the President. This i s potentially an important step - which i s discussed further in Section Fbelow. 41 C. Key Features of the Sub-National Institutional Structure 3.10 The institutional structure described above exhibits five important features. Collectively, these constitute the core o f the "sub-national problem" in Afghanistan, and generate the basic challenges that efforts to strengthen the system must address. The impact o f these challenges inone province -Takhar - i s illustrated inBox 3.1. Box 3.1: Sub-National Challengesin Takhar Province Takhar Province i s located inNorth Eastern Afghanistan. It has a population o f 886,000 and ranks 24 out o f 32 provinces interms o f human development indicators. It ranks inthe best fifteen provinces for overall primary school attendance and girls primary school attendance. However, it i s among the worst fifteen provinces for under-5 mortality rates, and among the worst five for its maternal mortality ratio. The World Bank's PAR team visited Takhar inJune 2007 for a series o f consultations with the Provincial and District Governors, Provincial Directors o f Line Departments, the Provincial Council and local CDCs. These consultations confirmed a number o fthe issues raised inthis report: The Provincial Governor does not have a good overview o f government spending inTakhar. The Provincial Development Plan calls for spending o f around US$lOO million per annum, while actual spendingfromthe core budget (excluding external assistance) was lessthanUS$1million last year. Members o f the Provincial Council were concerned that the lack o f good budget information made it difficult for them to carry out their oversight responsibilities. At the same time, the Governor was concerned that the Councilwas pushing for executive powers to implementdevelopment programs. There were also concerns about the impact o f uncoordinatedaid. The Provincial Education Director complained about schools being built by the PRT and then handed to the government to run - without any consideration for recurrent budgets. Mechanisms for feeding local priorities into planning andbudgeting processes inKabul are still very weak. There was good understanding o f the Priority Reform and Restructuring Program (PRR). However, the Provincial Health Director complained about the impact o f multiple pay scales on staff morale. Andmembers o fthe Provincial Council wanted to ensurethat appointments really were meritbased. The CDC met by the team was newly elected, and was just starting on the process of selecting priority projects for fknding under NSP. They were concerned that the facilitating partner had not been back since the elections to help with the project selection process. The CDC does interact regularly with the District Governor. Source: World Bank. 3.11 First, the overall structure is characterized by significant systemic c~ntradiction.~~ On the one hand, direct formal functional and budget authority for the delivery o f most key services in the provinces - such as education, health, water and roads - i s held by highly centralized line ministries, which work in vertically integrated silos with relatively weak linkages between them. Thus, in an organizational sense, provincial and district "administrations" - defined as horizontally integrated entities in which different agencies responsible for specific sectors fall under a unified point o f budget and executive authority - do not actually exist in Afghanistan. What is often referred to as a "provincial 59 The focus here i s on the contradictionsbetween systems internalto the government structure at the sub-national level. There are also contradictionsbetweendonor allocationandimplementationsystemsandthose o f the government, which are well known and not necessarily specific to the sub-national level. Those which are most germane to sub-national service delivery and governance (e.g., PRTs) will bediscussedat various other points inthis chapter. 42 administration" i s really just the aggregate o f the provincial offices o f the line ministries (i.e., the Provincial Departments), the Provincial Governor's Office, the Provincial Development Committee and the like, with no "hard" institutional linkbetween them. Equally, there is no such thing as a "provincial" budget, in the sense o f identifiable fiscal resources dedicated to expenditures in a specific province, over which a single authority has allocation and execution responsibility. There are only aggregates o f expenditure which take place for services delivered at the provincial level, which the central line ministries largely control, and which cannot even be accurately captured by the existing accounting systems. 3.12 On the other hand, this highly centralized line ministry system co-exists with - and is even cut across by - the Provincial Governor system, which allows the PGs to intervene in the affairs o f the line ministries and other agencies (such as municipalities) through six "softer" institutional channels, some legislated, some not: 0 As noted in Chapter I,Provincial Governors hold extensive powers over the appointments o f lower-grade civil servants to departmental positions intheir provinces. 0 PGs are able to influence the day-to-day activities o fthese civil servants simplyby virtue o fbeing close to them. In fact, this power is being extended in a number o f provinces as PG Offices undergo PRR reform, gain increasing capacity in technical areas, and consolidate their influence through mechanisms such as Provincial Administrative Assemblies where they formally bring heads o f provincial departments together in order to make joint decisions on operational matters such as financing priorities and procurement. 0 In some cases, PGs have been able to gain effective influence over funds which allow them to build infrastructure in areas for which line ministries are responsible (e.g., directly through the Governors Discretionary Fund, now narrowed in scope, and indirectly through their influence over PRTs intheir province). 0 PGs hold separate approval powers for all local recurrent expenditures o f Provincial Departments funded from the core budget. This power also underlies the other powers they exercise, for example on PDprocurements through the Administrative Assemblies. PGs are able to exercise (varying degrees of) control over municipal budgets (as indicated in recent researchby the Asia Foundation). 0 These formal and informal avenues o f influence all exist within, and are reinforced by, an overall institutional culture which gives PGs a central position in the political structure o f the country, and which tends to defer to the inheritance o f existing practices even when they are no longer consonant with institutional developments. For example, when actual allotments to provincial departments were significantly less than budgeted, there may have been a legitimate role o f PGs inthe area ofexpenditure authorization -after all, someone hadto rationthe cash. This problem was solved some two years ago, but the practice continues. 3.13 Second, even given the choice of a centralized inter-governmental structure for Afghanistan, existing institutional arrangements are often highly inefficient. Inparticular, line ministries tend to be over-centralized, with their central offices in Kabul retaining functions which could be much more efficiently performed at the provincial level; teacher recruitment in the case o f education, for example. To some extent, this problem i s not unrelated to the one discussed above. So long as line ministries perceive that any significant deconcentration o f budget or functional authority to their Provincial Departments i s likely only to make these authorities subject to influence and capture by the Provincial Governors, they are unlikely to have much incentive to do so. 43 3.14 Third, the current system is both asymmetric and inequitable. Institutionalasymmetries in inter-governmental arrangements are not uncommon internationally, o f course, and are not intrinsically problematic - it all depends on why they exist and how they affect the inter-governmental system as it evolves. InAfghanistan's case, it i s fair to say that while some o f these appear to make sense, others are the result o f arbitrary processes connected with history, varying donor practices and the like, rather than considered design choices. The models o f the PG Offices that are being implemented in the Policy Action Group (PAG) provinces, for example, are rather different from those in B a l k and Herat - resulting in higher degrees o f governor influence in these provinces - and it i s not at all clear why. Ultimately, these are likely to introduce significant strains and tensions at the sub-national level as the system consolidates and attempts to cohere. 3.15 The inequitable distribution o f resources across sub-national jurisdictions presents a more obvious and pressing set o f difficulties. Some preliminary work conducted by the Ministry o f Finance indicates that in sectors such as education and health, per capita expenditures differ widely between provinces and much more than would be expected as a result o f "neutral" factors such as variations in input-cost. Inother environments, a limited degree o f variation inper capita expenditure on health and education would not necessarily be regarded as problematic. Part o f the rationale for increasing the autonomy o f sub-national entities i s that efficiencies arise when they are able to allocate their expenditures in line with (varying) local priorities. Afghanistan, however, does not possess strong local governance mechanisms for citizen voice. Thus expenditure variations are unlikely to reflect local preferences and so do not contribute to allocative efficiency at the sub-national level. It should also be mentioned that these problems are aggravated by the fact that the size, composition and regional distribution o f the external budget, which finances about two thirds o f public spending in Afghanistan, i s largely unknown. It i s therefore impossible to assess, let alone address, the equity o f total public expenditure. 3.16 Fourth, the most autonomous elements of the sub-national system of governance and service delivery -the municipalities and CDCs-face issues of their own. Municipalities have been established, under law, as important and self-sufficient local service delivery and governance entities. CDCs were established as community-level vehicles to deliver infrastructure in rural areas under NSP and have subsequently started to take on a wider range o f local governance functions. A common feature o f these entities, which distinguishes them from provincial-level entities, is that, inprinciple, the elected authority inthe municipal or community area is itself directly responsible for infrastructure andor service delivery inthat area. As aresult, they do not suffer from the accountability and autonomy problems characteristic o f other organs o f sub-national government. 3.17 On the other hand, they do confront significant issues o f their own. Municipal council elections have yet to occur and, until they are held, the potential o f municipalities to deliver accountable local government intheir areas will be highly circumscribed. Inaddition, the fiscal and functional framework governing them i s both unclear and constraining o f effective and efficient service delivery. CDCs face existential questions derived from the possible discontinuation o f the funding which provides them with the basic wherewithal to play a meaningful role. Moreover, the level and scale at which they operate means that the roles that they play and the issues they face are very different from those o f higher-level bodies. These challenges are discussed in greater detail later - for the moment the important point i s that the main issues CDCs and municipalities face are different in complexion to those at the provincial level and needto be addressed intheir own right. 3.18 Finally, like the rest of the Afghan state, the entire sub-national structure is afflicted by the sorts of problems characteristic of fragile states: severe human resource weaknesses, an absence o f properly functioning operational systems, shortages o f equipment, and sparse supporting infrastructure (such as power and phone systems) necessary to get things functioning properly. Afghanistan i s 44 particularly badly affected by these, and they will inevitably impact on any effort to reform and strengthen the sub-national system. D. StrategicFrameworkfor Addressingthe Sub-NationalChallenge 3.19 Any effort to address these sub-national problems has to recognize that there is a constitutional dispensation and political climate which precludes any significant reform to the underlying inter- governmental structure inthe short to medium term. The Constitution refers to "preserving the principles o f centralism", and it i s also clear that the current political establishment is firmly committed to the current centralized structure and has no appetite for a substantial political devolution to sub-national levels. This is a fundamental point o f departure for the discussion o f reformpaths that follows. 3.20 Given this reform environment, the basic intent o f any sub-national strategy must be to improve the functioning o f the current system, rather than to restructure it. True, this structure has intrinsic weaknesses from both the service delivery and governance perspectives. A strongly centralized state, in a large, poorly linked country with a relatively thinly spread population, inevitably confronts problems with regard to both efficiency and accountability. It also imposes strong constraints on coordination and integration between different agencies. However, not withstanding these difficulties, the emerging structure does allow for improved performance with appropriately focused reforms. Such intra- structural reforms need to be designed with two sorts o f impacts in mind: (a) to strengthen specific elements o f the system (organizational impacts); and (b) to ensure that the functionality o f the system as a whole improves (systemic impacts). This approach leads to four broad challenges that need to be addressed if service delivery and governance are to be improved. 3.21 Challenge I: Establish an overall directionfor sub-national reform which, while building on the current system, diminishes the existing levels of systemic contradiction and begins to reduce the unproductively "mked" character of current arrangements. This direction should be based on the broad allocation o f roles and responsibilities outlined in Figure 3.1. Among other things, this suggests two significant orientations for reform. First, in order to reduce current levels of systemic contradiction, it is important to strengthen and consolidate the role o f the Provincial Governor's Office inthe areas o f coordination and planning, while limiting its functional, budget and operational authority to deliver services, which falls under the Provincial Departments. Second, in order to enhance accountability, it i s important to strengthen the oversight and supervisory role o f the Provincial Councils to provide checks and balances over the activities o f the delivery agencies. 45 Figure 3.1: Roles and Responsibilitiesof Sub-National Institutions in Governance and ServiceDelivery Province level District level , :: + Village 8 Community level 1 1 Beneficiaries Responsibility for direct service delivery in sectors such as health, education, water and roads should fall under the operational jurisdiction o f the line ministries (and, in their areas, CDCs and municipalities) using a variety o f differeni service delivery models. These organizations need to improve their internal organizational efficiencies by deconcentrating functions and expenditure responsibilities from the central offices to provincial departments. Their internal operations should become less permeable to the interventions o f external agencies, particularly the PG offices. PRTs will also play 5 key investment role for the foreseeable future. Their activities need to be increasingly aligned with those of the organs 01 government and/or directed ina way which does not negatively impact institutional development processes. In the absence o f district and municipal elections (and except in respect o f CDCs) the basic "bottom-~p~'accountabilitJ function inthe provinces should be exercised through the Provincial Councils. Inessence this means that their monitoring and oversight functions inrespect o f both the PDs and the PG offices need to be expanded andconsolidated. Going forward their capacity to play this role needs to be strengthened through increasing their resources, improving their systems anc making them increasingly self-sufficient inperformingthis function. A further dimension o f the oversight role o f the PCs i s the approval power they hold in respect o f provincial planning and budgeting, which appears now to be appropriatelj defined. At the community level, CDCs are structured in a downwardly accountable manner, hence exercise this function ir respect o f the activities which they undertake. They may also support the PC accountability role through providing information etc. "Top-down" accountability for delivery agencies (e.g., the PDs) is exercised through existing institutiona processes and fiduciary mechanisms within the line departments and the mustofiat (at the provincial level), municipalitie: and CDCs. Coordination of PD activities and provincial planning is the core responsibility of the Provincial Governors in anc through the institutions which fall under their jurisdiction (e.g., PAAs) or over which they preside (e.g., PDCs). Given tha budget, functional and operational authority lie with the line ministries the lateral co-ordination links will be institutionallj "soft" rather than "hard" in character. Nonetheless they are important, and the province-level planning and coordinatior function needs to be strengthened by properly resourcing the key organizations which are responsible for it, particularly thc PG offices and the PDCs. At lower levels, CDCs and municipalities have planning responsibility for their areas Federationsor alliances o f CDCs may be able to play a planning role at the district level. Conflict resolutionis dealt with by a combination of bodies, mainly PCs, CDCs and the District Governor - depending or the type and scale o f the conflict. These bodies needto be resourced to be able to play this role more effectively. Source: White and Lister 2007,38. 3.22 It is importantto stress, however, that these are suggested as broad orientations to guide a process of ongoing activity. There is no expectation that, in a country characterized by a weak central state, ethnic complexity, dispersed settlement and a history of political decentralization, specific steps in these 46 directions will be easily agreed or implemented. Moreover, as pointedout elsewhere, it appears that some o f the current reform efforts cut against rather than in favor o f this overall orientation. However, without some movement in this direction the service delivery apparatus that i s currently being built, and which has achieved some apparent success, i s likely to run into increasing operational difficulties and the progress that has been made may bejeopardized. 3.23 Challenge 2: Strengthen the performance of the main service delivery agencies by improving their organizational structures. One key, cross-cutting activity inthis area is for line ministries to ensure that functions are properly divided between their central offices and Provincial Departments. Ina number o f cases this i s likely to involve the deconcentration o f various powers and functions from the central to provincial level. Infact, insome cases this i s already underway; the MoE, for example, is currently inthe process o f deconcentrating teacher recruitment to the provincial level. The pace o f such deconcentration will be determined, among other things, by capacity constraints -although it shouldbe recognizedthat, in the absence o f functional deconcentration, capacity at the sub-national level will not be built. Ultimately, functional deconcentration drives capacity building, not the other way around. 3.24 Four important observations can be made here: The current PRRL'AR process is directly relevant. As discussed in Chapter 11, a new phase o f PAR i s currentlybeingplanned, and affected ministrieswill have to develop a new organizational structure to support their staffingplan. For those ministries active at the provincial level, one key decision will be the division o f functions between central and provincial offices. As proposed below, this process should be used as an opportunity to ensure that more efficient structures emerge. Notwithstanding this point, any deconcentration process should be crafted flexibly enough to allow different ministries to implement service delivery models which are most conducive to success in their sectors. For example, whereas deconcentration may be of significant benefit to MoE, which relies on a state-based delivery model, the same may not be true o f MoPH, which relies more on contracting out to NGOs and where centralized authority could remain relevant. Ministries could consider asymmetric approaches. MRRD, for example, has deconcentrated certain functions, including procurement, to its provincial office in Kandahar, an experiment which appears to be going well. Other ministries could similarly pilot deconcentration in this manner. Challenge 2 must be addressed inthe context o f Challenge 1. Inother words, the extent to which line ministries deconcentrate will need to be moderated by the degree to which this risks "capture" o f Provincial Departments by Provincial Governors. There i s little point in trying to improve organizational efficiency through vertical deconcentration if this makes the delivery agencies more vulnerable to horizontal power which inhibitsoperational effectiveness. 3.25 Challenge 3: Provide sufJicient resourcesfor the key organs of sub-national governance and service delivery toplay their ascribed roles. For example, as pointed out later, CDCs face a fundamental threat as their access to funding under NSP i s due to end, and without the finds necessary for their development activities, they will wither and die. Similar issues arise with other elements o f the sub- national structure. At the provincial level, organs such as the Provincial Councils need to gain access to sufficient funds to conduct their operations as envisaged under the recently amended law; while at the municipal level, an overall fiscal framework which provides a rational, equitable and affordable solution to the fiscal gap they face needs to be developed. 3.26 A further dimension o fthe fiscal resource issues is movingtowards a more equitable distribution o f fiscal resources across the country. The specific conditions o f Afghanistan make this uniquely 47 challenging: there i s an understandable tendency to concentrate expenditures in those areas which are most vulnerable to security threats, and the large off-budget flows (through PRTs etc.) distort distributional outcomes, whatever efforts are made inrespect o f the core budget. Arguably, too, this is a second generation issue which can only really be addressed once greater stability and predictability are established in vertical and sectoral flows. But it i s ultimately an important question which should not be neglected indefinitely. 3.27 Challenge 4: Enhance the operational performance of all sub-national organs by providing them with the systems and capacities they need. In some areas (e.g, some PG Offices) substantial progress i s being made; in others (e.g., Provincial Councils) it has really only just begun. Needs obviously vary across organizations, but certain generic themes can be identified (training, IT, offices, communications etc.). Moreover, it should be possible to handle some o f these needs on a modular basis, i.e., to define a package o f "basic needs" equipment for Provincial Councils, for example, and to roll this out to PCs across the country fairly efficiently. E. SpecificActionswithin this StrategicFramework6' 3.28 In order to address these challenges, a wide range o f concrete initiatives will need to be undertaken. This section makes proposals for these with reference to each o f the main organs o f sub- national service delivery and governance. Some o f these proposals are new but many leverage off activities which are already in place or are planned. It should be stressed that the proposals made here cover a wide area and there i s no expectation that they will - or should - all be embarked on at once. Rather, incremental movement will be requiredon as many fronts as i s feasible at any one time. 3.29 Provincial Departments. Broadly speakmg, PDs are over-centralized, suffer too much interference from Provincial Governors in their day-to-day operations, and are under-capacitated. Their performance needs to be improved through a combination o f capacity development and reorganization, with specific reference to the division o f powers and functions between central and provincial offices to improve efficiency. This needs to be designed and implemented as appropriate to the service delivery model prevailing in the sector. In general, however, it i s likely to involve both vertical intra-ministry deconcentration and increasing operational autonomy from PG Offices at the horizontal level. Ministries which demonstrate a willingness to pursue reform in this overall direction should receive concentrated support. More specifically: Cross-cutting PAR reforms should be accelerated at the sub-national level, with appropriate coordinationwith the capacity development plan for common functions.61 Ministries formulating restructuring plans under PAR should explicitly address the division o f powers and functions between central and lower levels in areas such as budgeting and expenditure authority. Inorder to ensure cross-ministry learning and smooth implementation, IARCSC and the line ministries mustwork closelytogether. 0 Reform Implementation and Management Units (RIMUS)should continue to be established to manage and oversee ministerial reforms, although care should be taken that RIMUSdo not become a large structural parallel to the civil service and a substitute for sustainable capacity development o f staff inline positions. 6oAnnex 1 o f White and Lister 2007 categorizes these actions into each o f the four general challenges identified above. It also provides some (incomplete) reference to activities that are already underway and some summary comment on what is needed, thus outlining a rough "gap analysis" for discussion with the government and donor community. The list o f existing activities is partial, based on information available, andneeds to be expanded and improved. See Islamic Republic o f Afghanistan 200%. 48 0 The provincial budgetingprocess should be deepened and expanded, with the proposed expansion innumbers of ministriesand provinces, as well as management at the local level of "goods and services" and "asset acquisition" allocations, as planned. 0 Other PFM reforms at the sub-national level should be continued and expanded to ensure a strengthened oversight and monitoring role for the MoF. Ifthe ADB does not continue its current fundingfor mustofiatreformafter May 2008, thenan alternative donor shouldbe found. 3.30 Provincial Governors' Offices. PG Offices are defined by a simultaneous ability to become deeply involved inthe affairs o f the service delivery agencies inthe provinces and a lack o f any structured accountability for the performance o f those agencies. Therefore, the core reform agenda here needs to focus on consolidating the role and capacity of PG Offices for coordination and planning at the inter- departmental level, while reducing their incentive to become involved in the operational functions of the line agencies. Given the important role that PGs currently play inprovincial affairs, change inthis area i s likely to be slow and incremental. However, the proposed initiatives are closely in line with GOA'Sown identified areas for reform:62 0 Regulatory reform i s needed to ensure consistency between the powers given to different institutions at the provincial level so that PDs are able to operate effectively. For example, the sign-off powers o f the Governor on line department expenditures should be eliminated or at least reduced. Intime, a revised Local Government Law should clarify the coordinationrole o f the PG Offices at the inter-departmental level while reducing the scope for their involvement on intra- departmental matters. 0 In order to avoid conflicting with the mandate of line ministries, donors should not make substantial sums available to Provincial Governors for infrastructure projects and service delivery activities. The capacity o f Governors to play a coordination and development planning role inline with the overall institutional framework should be strengthened through the continued roll out o f reforms o f PG Offices through the PAR process. It i s important that the specifics o f the models being implemented and the capacity provided i s reviewed to ensure that the PG Offices are not being developed in a manner which encourages their deep involvement in intra-departmental activity, especially at the operational 3.3 1 Provincial coordination and planning. In the current institutional structure, Provincial Development Committees (PDCs) and Provincial Administrative Assemblies (PAAs) are the only loci where cross-sectoral coordination and province-wide development planning can take place. PDCs are also developing an important role in the M o F provincial budgeting initiative, participating in the formulation o f budgets. However, practices vary widely across the country and capacity is generally weak. While ProvincialDevelopmentPlans (PDPs) are emerging as central to the planning process at the provincial level, there remains a need to clarify exactly what PDPs are - and can achieve - and to manage expectations around the process. Specifically: 0 The PDCs need to be given sufficient support to play their mandated role in development planning and budgeting, according to the overall institutional framework. This mainly requires support in basic systems and organizational infrastructure, as well as institutionalization o f internal practices through the development o f standard practices and manuals. Donors over time should withdraw from providing the secretariat functions to the PDCs. 62See Islamic Republic ofAfghanistan 2007a, 21. 63Inthis context, an in-depth review ofthe Balkhexperience should becarried out, and the actual impact on institutional issues at the provincial level assessed. 49 The relative roles o f the PAA and PDCs should be reviewed, as there i s currently the potential for overlap and duplication o f effort. The PAA, under the leadership o f the Governor, could have a stronger reportingrole to central government on provincial affairs. The active and direct participation o f the Governors in the PDCs needs to be strengthened through provision o f appropriate support and encouragement from the Office o f the President. Over time, the secretariat and support function for the PDCs could be moved away from a central ministryinto the PGOffices to cement thisparticipation, althoughthis is not apriority reform. Given the problems with linking PDPs directly to resources or institutional action within the existing framework, PDPs should be seen less as costed plans and more as strategic documents giving some sense o f priority issues to guide line ministry and donor decision making in the provinces. It i s important that information about what PDPs can and cannot achieve is widely disseminated to lower expectations. 3.32 Provincial Councils. The key role o fPCs is to exercise an accountability function -infact, aside from CDCs, they are currently the only source o f "bottom up" accountability in the sub-national system. The recent revision o f the legal framework has strengthened the scope o f the PCs' mandate considerably, and this mandate will increase in importance as and when line ministries deconcentrate. However, there are significant constraints, both financial and capacity-related, to their ability to exercise this mandate effectively. Two types o f initiative are therefore needed: to clarify and inform all actors about their roles; and to provide the resources and capacity neededto play these roles properly. Specifically: Ensure widespread dissemination o f and training to councils, government bodies and other sub- national actors on the roles of the councils as laid out inthe revisedlaw. Clarify the mechanisms by which the PCs can exercise their oversight role and act on the problems they identify. There is, for example, currently confusion about whether they should report issues to the Governor or (through the PAA) to the President. To strengthen the accountability function, clarify with all actors the role o f PCs in providing qualified sign-off on budgets formulated by PDCs. Provide substantially more resources for internal systems and institutional development, also IT infrastructure, transport etc. To ensure improved coordination in this, re-establish the joint government-donor technical worlung group on PCs, while ensuring some means o f PC input into it. Enable councils to link up with government and other (for example, PRT) data sources to improve monitoring and oversight. In the medium to long term, assess whether the current institutional location under the OAA allows the proper development of the councils as independent bodies. Districts. There i s a gap at the district level in terms o f both an agreed overall institutional framework and the activities to strengthen existing institutions. The reform o f District Governor Offices has not proceeded as planned under the PRR initiative. It i s also unlikely that there will be any directly electedbodies at this level inthe mediumterm. The appropriate structures and roles at the district level i s a topic for decision by the proposed inter-ministerial commission (see para. 3.42 below). Inthe absence o f an agreed framework, limitedreform at this level i s possible: District offices o f line ministries should be included inintra-ministry deconcentration initiatives. 50 0 Itisnot clear that DGOffices shouldbe extensively strengthenedinterms o f staffnumbers, given their limitedrole, the large number o f districts and resource constraints. However, it i s important to ensure that new procedures for selecting DGs on merit are followed in a transparent manner and that DGs are held accountable for their actions. 0 Donor programs operating at the district level, and attemptingto build institutions, should ensure that they are establishing consistent structures across the country. Impromptu donor initiatives at the district level, inthe absence o f a wider institutional framework, should be avoided. 0 Promote district-level planning and service delivery processes as they begin to develop (for example, through NABDP), but ensure that planning at this level i s kept relatively simple, with clear links both upwards and downwards, and that it i s tied to allocated funding. 3.34 Municipalities. The municipal system faces numerous challenges as outlined inBox 3.2. These couldbegin to be addressed through a number o f broadmeasures: 0 Enact new legislation which clarifies the relationship o f municipalities to the central and regional governing entities, while protecting their self-reliance. This should provide them with solid expenditure and revenue assignments, enhance capacity for providing local services sustainably, provide a clearer categorization o f different local administrative bodies that should be created to manage different settlements, and establish real self-governance and downward accountability in line with the text and spirit o f the Constitution. Clarify the functional roles and expenditure assignments governing service delivery for municipalities. 0 Develop and implement service delivery frameworks for key municipal services including sanitation, solid waste management and maintenance o f roads and drains to incorporate existing community, private sector, NGO and public sector operations as well as lessons from international experience. Formulate a comprehensive financing framework for municipalities which clarifies and rationalizes own-source revenue assignments; introduces measures to strengthen municipal revenue administration; and establishes a transparent rules-based system to allocate central transfers equitably and rationally among municipaljurisdictions. 0 Develop improved budgeting, financial management and accounting systems, and institute measures to build capacity to administer these systems within municipal administrations. Introduye elections for mayors and municipal councils at the earliest feasible opportunity. 0 Develop institutional measures to deal with the CDC-mayor interface, and to bridge the current gap between the bottom-up system for expression o f community voice and planning and budget decision malung. 51 Box 3.2: MunicipalChallenges inAfghanistan Background: Municipalities are remarkably autonomous, inspite o f their reporting relationship to the Ministry o f Interior and Ministry o f Finance. They play a key development role, delivering services (municipal solid waste, sanitation, roads) and managing municipaYurban development. * Intergovernmentalrelations: Municipalities are governed primarily by the 2004 Constitution and the Municipal Law (2000), although the Public Finance and Expenditure Management Law (2005) also governs financial operations. However, the present legal framework is inconsistent and ambiguous. The Constitution visualizes a multi-layer intergovernmental system where the sub-national level includes provincial administration, district councils, municipalities, and village councils; but their interrelation is not stipulated. In contrast, the Municipal Law (2000) depicts a two-tier system, recognizing only municipalities which are obliged to report solely and directly to the Ministries of Finance and Interior (now IDLG) and the Council o f Ministers (apart from Kabul Municipality); while ignoring villages or other small settlements. T o mitigate this vacuum, many rural municipalities and villages have established informal "client" relationships with neighboring provincial municipalities. Financial status: Municipalities are remarkably self-reliant, receivingno financial support from the central or sub- national government tiers (except Kabul more recently). Unfortunately, the revenue assignment framework i s governed by several laws including the Municipal Law, the City Services Fees L a w and the Safayi Tax L a w and is thus vaguely defined. Under these circumstances, survival strategies o f municipalities have included the introduction of a multitude o f taxes, fees, and charges - many o f them miniscule. Finally, where transfers have occurred (to Kabul Municipality) they have not been governed by transparent rules. Municipal functions: The municipal finctions and expenditure assignment frameworks are equally vague, governed again by several laws including the Municipal Law, the City Services Fees L a w and the Safayi Tax Law. The ML describes functions most comprehensively but these are poorly defined, often confusing substantive and adrmnistrative functions, excluding some key functions (such as environmental management) and including some obsolete functions such as setting market prices - a throwback to the soviet era. Service delivery: Most municipalities do have the organizational structure to deliver services. But local service delivery is extremely poor, a result o f inadequate financing, poor capacity and the lack o f service delivery frameworks. For example, the Kabul solid waste management system provides no public primary collection inthe informal settlements, which cover 70 percent o f the city and house 80 percent o f its population. Instead, primary collection is private or community based. The municipality manages the secondary collection system, but waste collection rates are still lower than they should be due to capacity constraints o f staff, routing, and vehicles. A framework is required to integrate public, private and community based systems. Finance and accounting: The internal management of finance and accounting requires extensive attention. Financial statements are unavailable, a situation compounded by an inadequate Chart o f Accounts. Records are inaccurate and there i s little financial management and poor or no audit reports. The budgeting framework i s incremental and bears no relationship to city needs. Accountability relationships:Finally, the accountability relationships inthe municipality are vague, resulting ina centralised, non-transparent organization. Currently, there is no clear hierarchy between the mayor and the council, so de-facto, the mayor runs the council. Mechanisms for community inclusion are in place through the wakil-i- gozar system. But this voice is not represented at the center and there i s no forum for the development needs to be taken from the gozar and district level to the central level o f the municipality for incorporation into the municipality's broader development planning agenda. Finally, and most fundamentally, the fact that mayors and municipal councils are not yet elected critically constraints their potential as organs o f accountable and efficient service delivery. Source: Goga, Kopanyi, Chakravarty and Ibrahimi2007. 52 3.35 Community Development Councils. The recent evolution o f CDCs i s summarized in Box 3.3. They now face three basic challenges: an existential threat deriving from the potential drying up o f their fundingas NSP comes to an end; a more robust definition o f their role; and the question o f which central agency takes ownership and responsibility for them. In order to address these challenges, the following steps should be taken. 3.36 First, CDCs need to secure a regular and reliable funding flow to resource local development activities which provide their raison d 'etre. The primary initiative here should be to establish a reliable, ongoing fundingmechanism which gives them a source of dedicated fundingwhich they can then allocate to local projects at their discretion. This should be distributed horizontally between CDCs according to a simple and transparent formula. Given international standards, and to ensure sustainability, the funding should be set at far lower levels on a per capita (or per CDC) basis than under NSP. A donor operation, underpinned by a multi-year commitment from more than one donor source, will probably be needed to put this system in place. A secondary initiative, equally important in the short term, should seek opportunities for steering funding which is meant to support sector-specific investments at the local or community level via CDCs or federations o f CDCs. 3.37 Second, CDCs' roles should be consolidated on the basis o f the new by-law, i.e.: (a) allocating their "own" funding, and planning and implementinglocal projects usingthis funding, with a similar role being played with respect to sector-earmarked funds; (b) a limited administrationrole, particularly inthe areas o f civil registration (e.g., birth and death registers) and land verification; (c) planning and allocation at the district or sub-district level where CDCs are able to federate or cluster; and (d) providing information on service delivery and governance to PCs to assist them with their accountability function. Ingeneral, CDCs shouldtake onpublic goodfunctions, andshouldnotbecome involved inactivitiessuch as the selection o f individual micro-finance beneficiaries, as this would entail risks o f conflict o f interest, nepotism and corruption. It i s also important that CDCs do not become responsible for the governance and management o f local service-delivery units (such as schools) in competition with bodies (such as school management committees) which have been specifically set up for this purpose.64 3.38 Third, the matter o f "ownership" o f CDCs at the central level will have to be settled within the central government by those agencies and ministries which have a core interest inthis question. While it would be premature to make a substantive proposal, as a matter o f procedure the best way forward would probably be to have this issue looked into and decided by the Inter-Ministerial Committee proposed below. The allocation of general CDC resources to specific schools within their areas would, of course, be another matter -here CDCs @ have a legitimate role. 53 Box 3.3: The Evolution of CDCs Field implementation of the National Solidarity Program (NSP) began inAugust 2003 and, by the end o f April 2007, the program hadreached more than two thirds o f Afghanistan's estimated 24,000 villages across all of the country's 34 provinces. The two stated objectives o f NSP are: (a) to lay the foundations for strengthening o f community-level governance; and (b) to support community-managed sub-projects for reconstruction and development. CDCs constitute the basic buildingblock to achieve the governance objective o f NSP. They are supposed to be established as representative bodies through secret ballot election bybothmen and women. The election process that is called for i s structured in such a way that promotes inclusiveness across spatial, occupational, gender, factional and ethnic sub-divisions withm the village. Under NSP, the specific roles o f the CDCs are to develop, in consultation with the wider community, a medium-term prioritized Community Development Plan (CDP), specific sub-project proposals that address one or more o f the CDP priorities and are funded by the NSP block grant, management o f sub-project implementation, and operation and maintenance as required. The governance impacts o f NSP were assessed during the mid-term evaluation by the University o f York in 2005-06. The evaluation concluded that, incommunities reached by NSP, there was significant evidence of: (a) empowerment of CDCs; (b) improved community relations; and (c) increased public faith in the system o f government. (Ongoing analysis i s likely to be less clear cut than this, which i s not surprising inthe Afghanistan context.) Case studies conducted as part o f the evaluation found that, while their composition and functions varied considerably between communities, the CDCs are emerging as legitimate local governance institutions. Similarly, a recent study by CSIS (p.47) found that "the CDCs and tribal shurus are seen as more responsive to Afghan needs than provincial governments and provincial councils, and in many cases are the only sign o f improvement villagers have seen inthe past five years." There are also indications that the presence o f NSP correlates with the stabilization o f a locality. MoE reports that there were 72 school attacks in7 provinces in 1384 followed by 73 attacks in26 provinces inthe first halfo f 1385. Although NSP to date has constructed 196 schools in 18 provinces, only 5 cases o f attacks on NSP-built schools have occurred - even though 52 o f the NSP schools were constructed in the five provinces (Kandahar, Helmand, Ghazni, Khost and Balkh) which account for half o f the school attacks. Even so, the deteriorating security situation has affected the ability o fNSP to work inhighrisk areas. A CDC by-law issuedbypresidential decree inJanuary 2007 represents an important step towards consolidating CDCs as local governance entities mandated to be the entry point for village-level development activities. Key provisions o f the by-law are that: (a) CDCs are neither an arm o f the state nor part o f the government administration; (b) their primary role is to plan, organize, implement and monitor development activities at the community level; (c) CDCs can form associations with other CDCs and can cooperate with government organizations at the provincial and district levels on governance and development issues; (d) CDCs have the administrative function o f maintaining and providing data o n births, marriages, deaths and ID cards to the district administration; and (e) CDCs will have a Community Development Fund- financed by the government, donors NGOs etc. -- to implement its Community Development Plan. Source: Christensenand Olesen2007. 3.39 Provincial Reconstruction Teams. As noted inBox 3.4, PRTs face a critical dilemma: trying to create the space for the Afghan state to develop while runningthe risk o f undermining it. A number o f steps may be identified inorder to address this risk: Scale down PRT activities in relatively secure areas and begin to route the associated resources through government systems and budgets. PRTs should really only exist where security conditions make them absolutely necessary. The ESC should undertake a review o f PRTs in terms o fthis criterion and begin to phase them out, or right-size them, accordingly. 54 EachPRT commander should develop a condition-based plan for phasing out hisher PRT. These plans should be subject to scrutiny and oversight by the PRT ESC, which should stipulate a standardformat and timetable for their production and monitor progress against it. 0 PRTs should begin to move away from small-scale community-based projects, where the conflicts between local prioritization and governance and top-down PRT processes are likely to be most pronounced. Instead, their resources should be focused increasinglyon large-scale, bulk and linkinfrastructure, where the need i s obvious and process conflicts are likely to be fewer. Insofar as PRTs continue to provide resources outside the government budget, they need to ensure maximum alignment with it. This involves producing accurate and timely information on project costs etc., and ensuring that government inputs are accurately costed, particularly salary and operating costs once the project i s completed. More generally, ESC should beginto provide regular aggregatebudget and expenditure information on PRTactivities country-wide. The ESC should issue a policy discouraging the poaching of government staff and provision of salary top-ups. An overall orientation supporting local institutions and people to learn by doing, even at the cost of time and quality, shouldbe encouraged. 55 Box 3.4: The Role of PRTs There are currently 25 PRTs in Afghanistan, led by 13 different countries. They are dispersed widely throughout the country, covering both secure and insecure areas, but tend to be less concentrated inthe central and northern areas than elsewhere. The role o f PRTs has attracted much comment and debate. According to the PRT mission statement: "PRTs will assist the Islamic Republic o f Afghanistan to extend its authority, inorder to facilitate the development o f a stable and secure environment in the identified area o f operations, and enable Security Sector Reform and reconstruction efforts." And the most recent edition o f the PRT handbook emphasizes that "PRTs play a vital role inoccupying the vacuum caused by a weak government presence and hence deterring agents of instability." In other words, PRTs have a broad mandate to pursue a wide range o f military and development activities withintheir operational areas. It's therefore not surprising that different PRT models have developed in different provinces, driven by variations inthe conditions they face and different national leadership styles. Inessence, however, PRTs do two basic development-related things: (a) they provide a protective home for bilateral donors and other government agencies which undertake development programs in the province; and (b) under the overall leadership o f their commanders, PRTs identify and implement development projects funded from military budgets (e.g., the Commanders' Emergency Response Program for the United States) or occasionally development funds. While it has not been possible to get comprehensive data on PRT budgets and activities, an examination o f CERP expenditures over the past two and a half years indicates that: (a) the funding flows are substantial, with total expenditures exceeding US$200 million over this period; and (b) the expenditures are distributed over a wide range o f activities and sectors (including transport 25 percent, energy 17 percent and governance 15 percent). In addition, the hding approval process is not necessarily conducive to effective or rational programming o f development initiatives. Typically, a U S commander has discretion over sums o f up to US$50,000 per month, while larger projects must be sent up the line o f command for approval. This process, along with short staffrotations, tends to favor a multiplicity o f small-scale short-term projects over larger, bulky investments, irrespective o f the development needs o f the area. PRTs execute their development activities through institutional modalities which operate outside and parallel to those of government. While they may consult with local officials, PRTs ultimately decide on development priorities and projects themselves, contract directly with companies to implement these projects, and provide supervision and oversight. This inevitably makes it difficult to align PRT-determined priorities with those o f local communities, and to ensure that there is adequate recurrent funding to keep projects running. It i s also difficult to strengthen the institutional systems o f government if they remain unused and under-resourced. In fact, "government weakness" can become a self-fulfilling prophecy as skills are attracted away by PRTs and other agencies. PRT leaders are aware o f these risks and have taken a number o f steps to manage them. Even so, the core PRT dilemma remains - and will continue as long as they retain their current institutional character. This does not mean there is no need for PRTs or that they should be disbanded overnight. But it does suggest that definite measures are needed both to manage their risks and to truly contribute to their stated missiodmandate - while moving steadily onto an institutional track which makes them increasingly redundant. Source: White and Lister 2007. F. InitialProceduralSteps 3.40 The strategic framework and activities outlined above provide an agenda for strengthening the sub-national system in Afghanistan. This i s one o f the first attempts to focus comprehensively on the overall system and its institutional arrangements. As such, it i s best seen as a basis for stimulating 56 dialogue within the government and with the donor community. This process has already started, with the dissemination and discussion o fthe backgroundpaper inKabul inJuly 2007. 3.41 Since then, the Government o f Afghanistan has moved to strengthen the central oversight o f sub- national issues with the establishment o f the Independent Directorate o f Local Governance (IDLG), reporting directly to the President. The IDLG i s essentially a merger o f the structures o f civil administration under the Ministry o f Interior (excluding registration o f births and deaths) and provincial relations (Provincial Councils) under the Office o f Administrative Affairs (OM). Its basic role i s to supervise the affairs related to Provincial and District Governors, as well as municipalities (except Kabul). The Directorate i s proposing to adopt a two-prong approach, focusing on: (a) short-term initiatives to address symptoms o f poor governance in specific provinces and districts; and (b) longer- term solutions to deal withthe root causes o fpoor governance inthe sub-national system. 3.42 The establishment o fthe IDLGhas been warmly welcomed inthe donor community. Inthe past, a number o f ministries had overlapping responsibilities for the various sub-national organs, and capacity to monitor and manage any given aspect o f the sub-national system has been poor. Inaddition, ministries which are active at the sub-national level exhibit high degrees o f bureaucratic and political rivalry, making it difficult to forge common perspectives on important reform initiatives. To be effective, the IDLG will need consistent political support from the President. It will also have to build up political credibility with the powerful and disparate interests working at the sub-national level, and the technical capacity to tackle complex institutional and policy issues. To support this effort, consideration should be given to forming an Inter-Ministerial Committee, chaired by the President or Vice President, with IDLG acting as the secretariat, to establish a framework for sub-national strengthening and reform. Inthe short to medium term, this should focus on providing clear direction for the evolution o f the sub-national system. In the longer term, this sort o f body could oversee a more thoroughgoing policy exercise, culminating in a new Local Government Law (a process which began some time ago, but which now appears to be ina state o f suspended animation). 3.43 The IDLGwarrants strong support from the donor community. Inthe past, donor activities at the sub-national level have been poorly coordinated. The extensive funding which flows to sub-national initiatives outside government systems has compounded this problem. Recent efforts to establish a donor workmg group on sub-national governance are therefore welcome. This provides a useful forum for formulating common donor views on sub-national issues and approaches, which should be fed into the Sub-National Governance Donor and Government Workmg Group under the ANDs mechanism. The IDLG should probably take over the chairmanship o fthis group from the IARCSC. One o f its first tasks will be to set appropriate benchmarks for developing a clear policy framework for the sub-national system, and a strategy for guidingreforms, along the lines mentioned above. 57 IV. MAKINGGOVERNMENTACCOUNTABLETO THE PEOPLE A. Introduction 4.1 Afghan citizens have two channels through which they hold the government to account: the direct channel o f elections and the indirect channel of checks and balances on the executive's abuse o f power. These restraints on the executive are exerted formally by the National Assembly (NA), Control and Audit Office (CAO), and independent agencies at arm's length from government; while the media and civil society organizations (CSOs) comprise a less formally organized oversight mechanism. This chapter will mainly explore the latter indirect means o f citizens' control over the executive i.e. public accountability, which requires the executive to explain and justify its decisions and actions to citizens, who can pose questions and passjudgments. Although very far from a reality inAfghanistan, the judiciary is meant to be protecting the rights o f citizens (and the executive as well). As Afghanistan's political systems are still evolving, the media and CSOs have the potential to become important means o f self determination and self expression, providing citizens the opportunity to engage more fully in political and economic decision making. 4.2 Public accountability i s a critical element o f PAR, whose objective is to bring security and services to the people. Through sharing o f information and feedback between the government and citizens, reform's progress against its objectives can be checked by both on a continuous basis without waiting for the next election. Public accountability i s closely linked to the other two elements o f PAR that were discussed inthe two preceding chapters. The executive's accountability i s closely linked to the first PAR element (Building an Effective Civil Service, discussed in Chapter 11) because it i s the means by which the public puts pressure on elected and appointed public officials to serve the public's interest instead o f their own. Government's accountability to its people i s linked to the second PAR element (Local Governance and Service Delivery, discussed in Chapter 111) because it i s at local levels and through delivery o fpublic services that citizens interact most often with Government. 4.3 Butpublic accountability is the least advanced among the three elements o f Afghanistan's PAR strategy. The legal framework i s uncertain and unenforceable, while the judiciary is virtually absent. Insecurity and intimidation prevent even oversight agencies from performing their functions in some provinces. In these places, law enforcement agencies cannot make arrests without agreement from the local strongman. Political interference-from members o f the Cabinet, parliamentarians, and provincial counselors-hampers all stages o f the Attorney General's investigation and prosecution process65. Members o f the Afghanistan Independent Human Rights Commission were threatened when the Commission issued reports on abuse o f power by those in the Government66. Large sums o f donor assistance, almost three times what the Government itself earns in revenue, i s spent on reconstruction activities; and about half o f it flows outside the national budget67. 65UNDP200%. 66 See, for instance, statements from Human Rights Watch at htt~://www.rferl.or~/featuresarticle/2005/1 O/ac3308eb-df83-4738- 9f6a-efOb86e9a8fa.html at Esfandiari 2005. 67 InSY 1385, actual domestic revenueswere $580 million. Donor assistanceto the core operating budget was 5380 million, to the core development budget $320 million, and to the external budget $743 million. 58 B. BringingPublicAccountabilityto the PARagenda 4.4 The direct election channel o f public accountability was exercised in September 2005 when more than 6 million Afghans went to polls to elect the lower house o f the National Assembly (Wolesi Jirga) and 34 Provincial Councils. The 249-member directly elected lower house, and the 102-member upper house (Meshrano Jirga) comprise Afghanistan's National Assembly (NA). Within four years o f the Bonn Agreement, Afghanistan had had a Presidential election (2004), an elected National Assembly, and officially established bodies for oversight o fthe executive6'. 4.5 While the direct channel o f public accountability was developed quite quickly, the indirect channel had a necessarily slow start in Afghanistan's PAR agenda. In 2002, at the start o f rebuilding Afghanistan, humanitarian and reconstruction needs were so immense and immediate, and government's capacity so sparserelative to the challenge, that the objective o f donor-assisted reconstructionwas to first get the system back on its feet before fine tuning its performance. Eventhe World Bank, which supported development o f Afghanistan's public financial management systems, began funding treasury functions (2002) before turning its attention to a procurement law (2004). Thus, the extreme "capacity deficit"69 inevitablyignoredthe "accountability deficit." Box 4.1: The 2005 Parliamentary ElectionsinAfghanistan The Joint Electoral Management Body (JEMB), consisting o f 9 Afghan members, 4 international members plus the (non-voting, UN-appointed) Chef Electoral Officer had overall authority for the 2005 election. Actual implementation o f the election was the responsibility o f the JEMB secretariat, headedby the Chief Electoral Officer. The techcal and logistical challenges o f conducting the elections were formidable. The Taliban had threatened to disrupt the election. A combination o f international military forces, the Afghan National Army, the Afghan National Police, andthe National Security Directorate provided election security. The Electoral Complaints Commission, an official, independent body established by electoral law to handle electoral grievances, examined 5,397 registered complaints. Forty five candidates were rejected because scrutiny o f their nomination found evidence of connection with armed groups, or for not giving up their government jobs. The Media Commission, set up under electoral law, had the mandate to ensure that voters had an opportunity to make an informedchoice when they voted. The strong performance o f female candidates characterized these elections. The Afghan Constitution guarantees 27 percent lower house seats, and election law guarantees 25 percent Provincial Council seats for women. However 19 women inthe WolsesiJirga, and 29 women inProvincial Councils won their seats in their own right, and ended up not needing the quota provisions o f the Constitution and election law. Source: Wilder 2005 4.6 Security and stability so overwhelmed reconstruction needs during the early days o f reconstruction that it detracted government's and donors' attention from public- accountability issues. Military assistance vastly outweighed developed assistance, at an estimated ratio o f 10:1"; and this asymmetry was reflected inGovernment's strategies. Good governance was one o f the three pillars of the 68The upper house i s meant to have one-thirdof its members elected from within eachof the Provincial Councils, one third from the District councils, and one third are to be appointed by the President. As district elections have not been held, and are unlikely to be held anytime soon, a Presidentialdecree was signed authorizing each provincial council to elect a transitional upperhousemember inadditionto theirregularmember untildistrict elections can beheld. 69CliffeandManningforthcoming. 'OLockhart 2006. 59 National Development Strategy. But, the government's 2004 economic development strategy" concentrated only on the core of public administration, (the center o f Box 1.1) by choosing security, an effective civil service, and good budget management among focus areas of governance, and omitting institutions of public accountability and the fight against corruption. Investment in media was clubbed with culture and sports, with the hope that these would help `create an open society andrevive the cultural diversity o f the country.' The security sector received a high level o f donor assistance ($3.9 billion during 2003-06), with its major part (79 percent) remaining outside the core national budget. Developing security forces was essential to good security on the ground. But security forces' effective management and oversight by civil authorities, which was lost during the long period of conflict, was slow in returning. The Ministry o f Defense's organizational restructuring and institution-building remained outside Government's reform arrangements. During these early years, the justice sector received very few resources72. Thus, the "security deficit"-like the capacity deficit-ignored the accountability deficit. 4.7 Donors' short-term reconstruction priorities were not compatible with the long-term effort required to build institutions o f public accountability. For exam le, in the education sector, donors helped to reconstruct / rehabilitate 3,516 schools during2002-20057', but it was only in2006 that school management committees for community oversight of schools became part of the sector's strategy, whose preparation was supported by donors. Many donors channeled their support through the external budget using NGOs and private contractors. The Control and Audit Office estimates the amount o f donor assistance spent through NGOs and private contractors duringthe past five years to be about $11billion. Most of these funds have never been audited, while many audit reports have remainedwith donors only74. 4.8 Past social and cultural traditions partly explain the absence of formal accountability institutions during the conflict years and even before it. While Afghan society's traditions have included informal checks and balances, the first Constitution o f 1923 did not introduce formal checks and balances on the King's powers. This was reinforced by the 1964 Constitution, which while giving the status o f a Constitutionalbody to the traditional Loya Jirga, also gave the Kingthe power to dissolve it. Inthe 1977 Constitution, the Party was the guarantor of power and all branches were accountable to the party, including the President. Throughout the turbulenttimes, the Shari 'a continued to be the source o f Afghan law. N o elections were heldduringthe Taliban period because (they explained) the Shari'a didnot allow politics or political parties. Instead o f an election, their leader's legitimacy came from Bay'ah or oath o f allegiance inimitation o fthe Prophet and early Muslims. C. Improvingthe EffectivenessofParliamentandthe Judiciary 4.9 Although formal oversight organizations have been set up, a broad understanding o f their roles and functions i s still evolving. These institutions are as much affected by institutional weaknesses as executive organizations: absence of clear legal framework, lack o f understanding o f legal provisions and organizations' mandates, and little coordination among government organizations. Insecurity deters the voice of dissent. Inaddition, a huge obstacle to these oversight organizations i s citizens' general lack o f "IslamicRepublicofAfghanistanandInternationalAgencies. 2007.TheADB,UNAMA,UNDPandtheWorldBank helpedpreparethe reportSecuring Afghanistan s Future. It canbe seen at httu://www.adb.or~/Docunients/Reuo~s/Af~hanistan/securing-afghanistan-future-fnal.~df 72 In World Bank 2005e, the Bank reportedthat during 2002-05, Government's expenditure on the justice sector (the ministry, Supreme Court and Attorney General's office) was less than 3 percent of what was spent on nationaldefense, internal security, the National Directorateof Security in charge o f gatheringintelligence,Borders and Tribal Affairs, counter narcoticsand mine clearance. 73Islamic Republico f Afghanistan, Ministry ofEducation2007. 74UNDP 200%. 60 trust in individuals and institutions. Respondents o f a 2006 World Bank-sponsored survey in 32 provinces reported much higher trust in extended family and village elders than in central government, provincial governors and provincial councils. Given the years o f conflict and the on-going resource- scarce environment which produces extreme competition for resources, it is almost impossible to get people to believe that decisions are made for objective reasons, rather than for personal, tribal, or ethnic ones. The perception i s exacerbated when these organizations' members display their unwillingness to follow laid-down mandates and procedures. 4.10 Duringits first year, the National Assembly started asserting its role o f oversight on executive power. It interviewed the nominees for Cabinet and rejected 5 out o f the President's 25 nominations requiring him to make fresh nominations. It also rejected the President's nominee for Chief o f the Supreme Court and three other nominations for the 9-judge bench. However, parliamentary scrutiny of the Government's budget and accounts has been impeded by NA members' unfamiliarity with presentation o f public financial management information, and passing o f the SY 1385 and SY 1386 budgets by the NA were delayed by several weeks. In SY 1385 parliamentarians refused to pass the budget until Government committed to raise civil servants' salaries and martyrs' pensions. The NA receives the annual report o f audit o f Government accounts from the Control and Audit Office, but it has no established system o f scrutiny and response for audit reports. Although the Constitution provides for legislative scrutiny bodies, there i s no parliamentary committee with a charter similar to that of a Public Accounts Committee to review the effectiveness o f public expenditure. The media reported that when parliamentarians summoned a minister or head o f independent agency, members spent most o f their time resolvingpersonal disagreements and ignoring issues o fpublic interest. 4.11 The NA plays a role in controlling corruption. It oversees governance via its Complaints and Petitions Commissions that exist inboth houses. The Commission inthe upper house has the mandate to receive citizens' complaints on all matters, including those related to the executive's corruption and mal- administration. The Commission has the right to summon ministers for questioning or follow-up on issues raised by citizens. In general the Commission acts as a parliamentary ombudsman trying to solve disputes peacefully. For example, the Commission intervened when the issue o f 1,200 prisoners serving prison time beyond their terms. However, its interventions carry the risk o f violating the constitutional principle o f separating state powers. 4.12 Parliamentary effectiveness is reduced by the absence o f strong and effective political parties. Political parties suffer a major image problem inAfghanistan because people continue to associate them with communist or jihad-era political parties that played such a negative role in Afghanistan's tragic history. Inthe lower house, less than one-third o f members are generally believed pro-government, and they are aligned to 13 political parties, factions and independents. Thus third party groups form the balance between pro and anti-government factions making legislation a challenging, time-consuming task, not to mention a potentially expensive one in an environment o f high levels o f graft and patronage expectations. Duringthe first year o f the National Assembly, many sessions had to be canceled for lack o f a quorumbecause there was noparty discipline to keep members inthe house. This was inspite o fNA members having to record their attendance in order to receive their salaries. More seriously, absence o f political parties to aggregate interests resulted in the NA and Government becoming de facto political parties, an adversarial relation between the two, with parliamentarians attacking the Government instead o f legi~lating'~.Inhis January 2007 speech to the NA as it started its second year, the President had to reiterate several times that the NA could usefully consult with government officials, and give them healthy advice for improvingtheir performance. 75Wilder 2005. 61 4.13 Justice institutions are probably the least developed among formal oversight organizations. These include the Supreme Court, the Ministry o f Justice, and the Attorney General's office. The legal framework's consistency needs to be urgently developed. Not only is applicable law in Afghanistan difficult to determine due to numerous regime changes, it i s further complicated by the power, that the Bonn Agreement gave to the interim head o f state, to issue decrees until a National Assembly was in place. Several hundreddecrees were passed starting from early 2002 and untilNovember 2005, when the NA first met. Stronger legislative capacity and skills in legislative methods and techniques will be required in the NA, the Cabinet and the Ministry o f Justice. The legislative process, while being participatory, requires greater coordination between the Cabinet and the NA. Delays arise because o f limited drafting capacity o f Ministry o f Justice staff, language differences with international experts, and lack o f explanatory notes when presentedto the NA. 4.14 A majority of Afghan citizens know little about the formal legal framework. Customary law prevails widely, and most disputes are settled in non-state forums76. But, customary law and informal justice systems are biased against women, who are not represented. At the same time, confidence in the formal justice institutions i s low, and strengthening the sovereignty and integrity o f the Afghanistan courts will require transparency o f courts' operations and procedures, uniform application o f law with predictable and credible decision-making, and courts' capacity to deliver services and manage resources. Some 330-360 functioning courthouses are required, each with a minimum o f 2-3 judges and several prosecutors. The Supreme Court has approximately 1,350 official judge positions. Of these, 50 percent are currently occupied, and o f that 50 percent one-third are estimated to have been educated to university standard. According to the Prosecutor general, 2,2 12 legal professionals are needed nationwide. Legal information, education and service are critical for the courts productivity and accountability. Manyjudges do not have access to legal texts, and simply apply their own version o f sharia law to many disputes. The role o fjudiciary, especially the Supreme Court, in enforcing constitutionality o f laws and regulations and reviewing legality o f administrative decisions needs to be strengthened. The Supreme Court's interpretation o f the Constitution i s rarely available, and ignored even when it is. Especially with regard to Parliamentary vetting of ministers, it i s unclear whether the NA also has the authority to dismiss a minister. The Supreme Court has ruled that the Constitution has not granted such power to the NA, but this advice has generally been ignored. D. Rethinkingthe RolesofMediaandCSOs 4.15 Historically, Afghan press had always been the domain o f the state, andunder Taliban rule, media became severely restricted. Radio Afghanistan was renamed Radio Shuriut to reflect the Islamic fundamentalist values o f the Taliban; and Television Afghanistan, believed to be a source o f moral corruption, was closed down. The Ministry o f Culture was mainly concerned with suppressing music for entertainment, and banning photographs and depiction o f living things-including destruction o f the Buddha statues at Bamiyan. 4.16 Considering this very low base from which media development started duringreconstruction, its growth in volume i s impressive. Today, nearly 350 publications are registered with the Ministry o f Culture and Youth Affairs, many appearing inboth Dari and Pashto, and with a majority operating from Kabul. Afghan print media i s still a long way from financial independence, so newspapers are either closely associated with political and military factions, or depend on international donors for funding. L o w literacy rates, lack o f efficient distribution networks, almost non-existent culture o f advertising in newspapers and difficulty inaccessing many rural areas keep Afghanistan steeped ina culture o fradio, on which the majority o f the population depend for news and information. Radio inAfghanistan comprises 76Barfieldet a12006. 62 the state-run Radio Television Afghanistan (RTA), which broadcasts from Kabul and 18 provincial stations; more than 50 local FM stations; a national radio show funded by Internews; international broadcasters BBC, Voice o f America and Radio Free Europe; and stations and programming funded by NATO and US Coalition forces. Television has once again become the country's main source o f entertainment, though even in Kabul only 25 per cent o f families have a television and power blackouts are 4.17 Recent tensions between the media and government are evidence that the media has started playing its role inpublic accountability. There was considerable media consternation at and opposition to the provisions o f the new media law that can be potentially used to restrict the media's freedom78. It draws upon the constitutionallymandated supremacy o f Afghanistan's religious beliefs over other laws to prevent dissemination o f prohibited material. The new media law provides for a HighCouncil of Media to keep track of income and expenditure o f mass media, ensuring that they are overt and transparent. The Religious and Cultural Affairs Commission's recommendation on the HighCouncil's composition i s still influx, but initialrecommendations o fthe Commissionomittedrepresentation ofboththe media industry and civil society. The new law reiterates an independent commission for management and oversight o f RTA, but the commission existingunder the earlier law is inactive. So, currently the Ministry o f Culture and Youth Affairs takes all decisions about RTA's management. The number o f registered cases o f journalists' intimidation rose from 45 in 2005 to 50 in 2006. Media leaders claim that these are gross under-estimates as most cases go unreported, and journalists' insecurity has resulted in their self- censorship. The Government in Kabul i s concerned about who controls radio programming in the provinces and how safe each station is from the possibility o f dissident elements gaining control. But the Government's determination o f what are facts could result in one-sided reports. Sections of the media point out this law's contravention o f Article 19 o f the Universal Declaration o f Human Rights, to which Afghanistan is a signatory. The Government had sought to introduce a clause in the newly passed amnesty bill immunizing alljihadis from prosecution that would force the media to honor thejihadis in any reports, butthis was withdrawn. Box 4.2: The ConstitutionandMedia's Freedomof Expression Article 34 states that "freedom of expression is inviolable... [and] every Afghan has the right to express h i s thought through speech, writing, or illustration or other means, by observing the provisions of the constitution. The same article further gives every Afghan the "right to print or publish topics without prior submission to the state authorities inaccordance with the law." Freedom o f expression is further strengthened by Article 7, which obliges the state to "abide" by international conventions to which Afghanistan are a signatory, including the Universal Declaration o f Human Rights But the freedoms enshrined in Afghanistan`s Islamic constitution are also guided by Article 3, which stipulates that "inAfghanistan, no law canbe contrary to the beliefs and provisions of the sacred religion o f Islam." 4.18 As the media acquires capacity to play the public accountability role, the Government is also adjusting to having its actions being scrutinized by the media. Recently, the country's most popular TV channel has been locked in a standoff with the Attorney General after police raided its office following ''Thebill "TheKillidGroupandInterPressServices2007. hasbeenpassedbybothhousesoftheNationalAssembly andneedsto beendorsedbythe Presidentbeforeit becomes law. 63 allegations o f misquotingby the channel. The management o f a weekly is inconflict with the Speaker of the NA after having alleged his misappropriationo f a large sum of money. The concept o f investigative journalism still has not taken root in large parts of the media and Government. Thousands o fjournalists and would-be journalists have been trained in donor-sponsored workshops, there are two journalists' unions and two media watchdog groups, but the current belief and training in universities still i s that a journalist should reproduce the handed out communiqd. The situation of both state-run and private media i s more sparse inthe provinces than inKabul. At least 9 provinces have no local media. A recent review of cormption articles that appeared during the last two years in five of Afghanistan's largest newspapers found that 90 articles had been published, but most of them had reproduced official news handed out by Government organization^'^. Among newspaper articles about corruption, accountability of government institutions received marginal (0.82 percent) coverage. A 2006 media coverage survey showed that more than half (53 percent) o f publishedinformation was attributed to Government sources, domestic and foreign; and civil society was a very small (8 percent) source of information. The media ignores thousands of shurus, associations and NGOs as news sources; nor do these groups use the media to voice their concerns". 4.19 The communication gapbetweenmedia and CSOs is partly due to the current role o fCSOs: more that of project implementer than defender o f public accountability. International humanitarian aid programs had continued during the years of conflict and drought. These were implemented by international organizations that lacked lmowledge and expertise and needed a local partner to implement their projects. Many international organizations eagerly sought "implementing partners" and found Afghan CSOs, most of which were located in Peshawar in Palustan, well equipped for this task. Comprising returning refugees that had been educated in their host countries, these CSOs had the `capacity' that donors sought. The CSOs, on their part, found it comparatively easy to work intheir own and familiar regions of Afghanistan, and many were motivated to contribute to their country's reconstruction. Demands of the donor community and their emergency aid programs led to relocation o f Afghan NGOs from Pakistan to Kabul and mushrooming o f several new Afghan NGOs all over the country, who carved their own niche indelivering much-needed humanitarianaid throughout the country while following international organizations' polices and practices. The private sector, being hostage to the vagaries of insecurity, provides few opportunities, and donors' compensation is much higher than civil service salaries. So, a majority o f Afghan CSOs continued to work as contractors or "implementing partners" of the donor communityg1. 4.20 There i s deep mistrustfor "NGOs" as a group, malung it hard for civil society groups in general to transition from their current role of donors' sub-contractors to honest brokers of public accountability. NGOs are largely perceived by the population as a foreign phenomenon that has gained control over the money meant for Afghans82. This mistrust extends to Afghan NGOs as well. Arising partly from misplacedfiustration and out o fjealousy for NGOs' access to resources, blame for wasting development aid sometimes gets directed as NGOs. As others in similar post-conflict circumstances, the Government resents the level o f international resources that flow to NGOs and private contractors in Afghanistan. Owingto a 2-year freeze on NGOs' registration, and to avoid the stigma attached to the termNGO, many civil society groups registered themselves as Social and Cultural Organizations taking advantage o f the 2002 law that providedfor such organizations to register with the Ministryo f Justice. ''DelesguesandTorabi 2007. *'TheKillid Group and InterPressServices2007. Padamsee2004. 82Moore 2005. According to Ministry of Financeinformation, betweenJanuary2002 and September 2004,45.5 percent of donor funding went directly to the UN; 28.5 percent went directly to the Afghan Government; 16.4 percent went directly to private contractors; and 9.6 percent went directly to NGOs. These statistics do not reveal the total percentageof donor assistance that ultimatelywent to NGOs, as theUNor AfghanGovernment often re-programfundingthroughNGOs. 64 4.21 Having worked for long as "doers" within a rigid structure, CSOs may themselves find it hard to transcend from their current role to becoming custodians o f the voice function. Afghanistan had little or no tradition o f formalized community participation inpolitical decision-making or development planning, either at the national or local level. "New democracy" starting with the Constitution o f 1964 was mostly a "democracy from above" without involving local communitiesg3. Civil society groups were not included in the Bonn political process: individuals' initiatives led to the creation o f the Afghan Civil Society Forum that took place in parallel to the Bonn process. Duringthe 2005 parliamentary elections, civil society groups played a valuable role incivic educationby going out into communities to explain the electoral process but these tasks were performed on behalf o f the Chief Electoral Officer and the messages were not something that the CSOs themselves had developed. Women's involvement in discussion and debate i s impeded in traditional arrangements as shuras (see Box 4.3) have a patriarchal structure. Because youth groups lack professional contacts, they are rarely counted among CSOs. Box 4.3: The Universeof Afghanistan's CSOs Civil society is the sphere o f institutions, organizations, and individuals-located between the family, the state, and the market-in which people associate voluntarily to advance common interests. The four main types o f CSOs currently observable in Afghanistan are: Community Development Councils (CDCs), Shurus, Non Governmental Organizations (NGOs) and Social and Cultural Organizations (SCOs). In general, NGOs are perceived to make profits while SCOs do not. Where they are registered sometimes turns out to be the only differentiation between CDCs (registered with MRRD), NGOs (registered with the Ministryo f Economy), and SCOs (registered with the Ministryof Justice). However, some organizations areregistered withmorethanone ministry! Community Development Councils (CDCs), also known as Village Organizations (VOs), were created to implement the National Solidarity Program (NSP). Project funds up to $200 per family could be allocated to communities for infrastructure related community projects, but applications for those funds had to come through CDCs, representing 25 to 30 families. Lacking the capacity to build CDCs, the Ministry o f Reconstruction and Rural Development (MRRDFthe line ministry for the NSP-contracted 20 facilitating partners to take responsibility for creating CDCs in one or more province. As o f 30 April, 2007, 16,753 CDCs had been in 279 districts, and registered with MRRD. Shurus are seen, inmost parts of the country, as functional local decision-making mechanisms. Their membership i s comprised almost wholly o fmale village elders, who are not elected. Shurus do not represent the younger cohorts in the population, "alternative thinkers", and the less-well-to-do. Because donors typically implement their projects through CSOs, any shuru that wants to become eligible for a grant becomes registeredas anNGO or SCO. NonGovernmental Organizations (NGOs), o f whom 2000 are registered inthe Ministryo f Economy's Registration Department. A large proportion o f them are construction organizations not fitting any conventional definition o f a CSO, or "inactive storefronts" or "briefcase NGOs". The number of "real" civil society organization among NGOs may be any figure between 150 and 400. A smaller number o f internationalNGOs (333) are also registered with the Ministry of Economy. There was strong internal pressure on Government to adopt a more controlling stance towards NGOs. With representation from the CSO community, a working group drafted a law for regulatingNGOs' activities, but since 2003 the draft law remained untouched for more than two years until NGOs' registration responsibility was transferred to the Ministry o f Economy and the new minister had a less adversarial approach towards NGOs. Registrationo fNGOs also remained frozen during this two year hiatus. Social and Cultural Organizations (SCOs) comprise political parties as well as associations, unions, and tribal shurus-in addition to a smaller number o f project-based CSOs, o f which 242 are registered with the Ministry o f Justice. As NGO registrationhadbeen frozen in2005 and to avoid the stigma which got attached to the term NGO, many CSOs have beenregistering themselves as SCOs. Source: Ayrapentas 2005. 83Boesen, 2004. 65 4.22 CSOs' capacity to play the role o fmonitoring government isnot fully developed. Although some government reports and external evaluations are available online, few CSOs themselves have the capacity to analyze the reports and inform the public, and fewer are willing to take the risk. InAfghanistan, the international community continues to monitor the Government, and only a few umbrella CSOs monitor the development community. CSOs rarely lobby the National Assembly or the executive. The first exception to this happenedwhen Afghanistan's two major CSOs together exerted pressure on government for Afghan SCOs to be allowed to receive funds from non-Afghan donors. The second instance was when CSOs lobbied the President and Supreme Court to release ajournalist who had been condemned to life imprisonment for criticizing Islam84. Ina recent survey o f 678 CSOs in 22 provinces, a majority o f respondent organizations reported fund-raising as their main need. The also cited advocacy as a main need along with communication o f all types - from travel to electronic . Security conditions not only L threaten the opportunities for international NGOsto assist indeveloping Afghan civil society, but are also diminishingthe ability o f local NGOs to enhance their own capacity and legitimacy. Last year 24 NGO personnel were killed. This year, althoughthe number o f fatalities is currently less, the number o f attacks on NGO staff has dramatically increased. Local perceptions about association with internationals, particularly the ProvisionalReconstruction Teams, are adversely affecting the security o fNGOs. 4.23 Given their different geneses and orientations, there i s no clear answer as to which amongst the current civil society organizations have the highest potential for being effective participants inlocal level governance-in district level planning and prioritization, as well as monitoring o f services. There is a large gap between the mandate o f a traditional community shura and a local community development association. Shurus ' traditional roles have been more reactive than proactive: their main focus i s on problem solving, including the resolution o f local conflicts. Religious leaders and networks have always played a prominent role as civil society forces in Afghanistan, but the question i s how the Taliban experience may have affected traditional Islamic leaders' potential as forces in civil society.86. CDCs, being elected bodies and having more women's representation than traditional shuras, could potentially evolve into local governance organizations. 4.24 The CDCs themselves have asked to be considered as village councils, and to be involved in the planning process. A CDC conference took place in Kabul during November 10-15, 2007, bringing together 450 delegates from all provinces, over a third o f whom were women. Some evaluations have found evidence o f CDCs' future potential because NSP communities have far more tools than non-NSP communities for envisaging problems, linkingpriorities, and planning87. Some argue that while the NSP has delivered actual results on the ground, the advocacy o f those with political stakes in the NSP, and NGOs that functioned as implementing partners may have been overstated. Others point out that more than half o f CDC members are local elites, and dominance by traditional leaders and local elites may deter community participation. And once CDCs become `part o f the establishment', they will be much less the `voice o f the people' than they are now and they will potentially compete with existing authority structures at the village level. It also remains to be tested how CDCs perform outside their currently limited remit o f planning and implementing small infrastructure projects, and with less funding than provided under NSP. 84Delesgues andTorabi 2007. 85Ayrapentas 2005. 86Harpikvenet al, 2002. "SeeChristiansenandOlesen2007. Thiswasoneofthebackgroundpaperscommissionedforthereport. 66 E. IncreasingInformationFlows andTransparency 4.25 Timely and reliable information is the basis o f public accountably, and Government and donors can demonstrate their commitment to transparency by effecting greater information flows within government and from the government to the outside. Weak communications, the steep learning curve o f parliamentarians, and relatively undeveloped CSOs make it quite impractical to expect that either parliamentarians or civil society groups can even demand the required information, let alone analyze it to hold government to account. Therefore, rather than waiting for CSOs to demand information, government can open up spaces into which civil society can come in later. Indeed, in many successful experiences o f developing social accountability, it has been the government or international agencies that were the ones who took the first initiative@. 4.26 The overall objective o f these information flows initiated by Government will be to create conditions that encourage public scrutiny o f government actions. The interventions will have to be in one or two selected areas, and in topics o f significant resonance in society. Worldwide, independent budget analysis i s a common task o f many CSOs. InAfghanistan, budget information, including execution rates, exists from the impressive achievements in public financial management, but that information i s not available in real time on the Finance Ministry's website. Even if it were, citizens, lacking access to the internet, would not find it useful. However, citizens would be able to access the information if Government made printed copies freely available at provincial governors' offices, district level offices and in mosques. The technical language o f budgets could be a deterrent to the average citizen. The government could draw upon the experience of the Malawi Economic Justice Network, which in a young democracy with weak checks and balances inpublic financial management systems, simplifies technical materials, such as the national budget and PRSP, and is widely distributed. Government could also telecast hearings o f parliamentary committees probing specific issues, at the same time issuing instructions that government officials are bound to answer parliamentarians' questions. Opening up the Pakistan parliament's Public Accounts Committee hearings to the media greatly increased public interest and participation in government performance. Afghan ministers have presented their programs during Accountability Week and then answered questions. They could occasionally be interviewed in radio shows, in a format similar to the e-GAMES forum o f Hungary but adapted to the status o f communications in Afghanistan. Public procurement i s a substantial fraction o f the GDP, so the government could adapt from Mexico's Social Witness o f Public Procurement program, which has significantly reduced the costs o f pubic contracts and increased the number o f bidders participating in procurement processes. 4.27 Better information flows-even among government organizations-can bolster anti-comption efforts. For example the police and the Office o f Bribery and Anti-Corruption inthe Ministry o f Interior have no obligation to inform General Independent Administration against Corruption (GIACC) or the Attorney General's Office that a corruption case in being investigated. The Auditor General has also drawn the President's attention to the lack o f communication between organizations and overlapping responsibilities for fraud and corruption, which contribute to the current ineffectiveness. Inthe absence o f a central traclung system, overall statistics on corruption complaints and cases under investigation are not available and figures provided by one agency are contested by others. The current confusion i s fuelled by the large number o f complaint mechanisms at all levels without there being any guidelines or laid-down procedures o f how these should be addressed or managed. See Ackerman 2005,34-35. 67 Box 4.4: Innovative Public Accountability Initiatives Hungary's e-Games (Government Assessment, Measuring and Evaluation System) could be modified to the Afghan context. This i s a sophisticated forum integrated into the Hungarian Government's internet portal, allowing citizens to interact with each other and ministers inreal time. Users cannot be anonymous; they are identified by their real names. They can assess each other's comments with positive and negative points, providing a value judgment on every user's participation. Aggregated points show a picture o f public opinion. Inaddition to value judgments, the number o f contributions to the topics forum leads to a popularity index. Public officials are users but cannot comment on the opinions expressed by citizens. Ministers are regularly invited to chat with citizens at a predefined time. The responses during these online "office hours" as well as their contributions, are measuredby points from the users. InMexico, a "social witness" participates ingovernment procurement as an observer. The social witness is a representative o f the civil society, which i s recognized for its professional and ethical approach. (S)he works with the individuals and organizations participating in the bidding process. The social witness can make recommendations to enhance transparency and increase efficiency o f the procurement process. After the contracting process i s completed, the social witness releases a public testimony describing what (s)he observed during the process and makingrecommendations for future improvement. I Source: Caddy, Peixoto and McNeil2007 4.28 Donors can support the transparency agenda by intervening opportunistically while being sensitive to the country's history, culture andpolitics. They could: Encourage the Government to move beyond the legal basis of public accountability and corruption, and to initiate a few practical steps. GIACC has been established in June 2004 with the mandate to coordinate all anti-corruption efforts inthe country. The government presented an Anticorruption Roadmap Paper (which had been drafted with several internationalpartners) at the AfghanistanDevelopment Forum inApril 2007. However, there was little follow-up or action on the recommendation o f this report. InAugust 2007, the NA ratified the UNConvention Against Corruption. The current anti-conuption law, which pre-dates the formation o f the NA, is considered .suspended. N o w that Transparency International has included Afghanistan in its world-wide rating and ranking o f countries' Corruption Perceptions Index, publicizing this information by the Government can start a debate on the subject. Help gather baseline information in critical areas such as poverty reduction and service delivery. This has begun inthe health sector. As reported in Chapters Iand 11, health outcomes are being tracked according to the models being used to deliver health services. This can enable an informed decision about what models are most suited to different sectors. The World Bank, EC and ADB have also helped implement the National Risk and Vulnerability Assessment. In addition to supporting critical surveys in a scattered way, donors could help strengthen the Central Statistics Organization which could then start implementing these and service delivery surveys. 0 Stress a communications component in each o f their projects, and provide information on their own projects: objectives, expenditures and results achieved. On its external website, the World Bank publishes details o f all contract awards as well as the rules governing them. It also provides detailed implementationupdates o f all its projects. 0 On a selective basis, encourage participatory monitoring of their own projects. The National Solidarity Program requires that CDCs, assisted by facilitating partners, monitor progress o f 68 project implementation based on a few selected key indicators; and the information be made available to communities by CDCs through regular briefings and information sharing. A community participatory monitoring system, being piloted, will monitor procurement, financial management, implementationand maintenance o f the subprojects as well as facilitating partners' performance. Allow better information flows among themselves and in critical areas such as narcotics control. For example, Afghanistan i s now responsible for 93 percent of the total global illicit opium production, and drug-related activities are the dominant source o f corruption. Corruption in counter-narcotics efforts has inadvertentlycontributedto makingthe drugindustrystronger (more consolidated, with fewer, powerhl players with strong political connections) and parts o f the Government severely compromised. Greater information sharing and consensus between the USAandUK(lead donor for counter-narcotics) can helpstrengthen the fight against narcotics. 4.29 Systematic and comprehensive implementation o f a broad-based, long-term agenda i s unrealistic in Afghanistan: because of the unstable security situation, the Government's lack of capacity, and its limitedreach outside Kabul. For government organizations and donor agencies this translates into seizing entry-points and opportunities as they arise. But this asymmetric (or opportunistic) approach can also be very inefficient, ifeveryone pullsindifferentdirections. That's why it i s so important for the government to have a clear idea of priorities at any point in time: what must be done now, and what can wait until later. Donors must be willing to work within this framework, and support the government's priorities. PARpriorities and a 12-point agenda to implement themare discussedinthe nextchapter. 69 V. A TWELVE-POINT AGENDA FORPUBLICADMINISTRATIONREFORM 5.1 This report has made the case that public administration reform in Afghanistan i s both very important and very difficult: very important because it will provide the Government o f Afghanistan with the capacity to respond to popular demands for good governance and improved service delivery; very difficult because of the inherent challenges o f institutional change, especially in the current political and security situation in Afghanistan. PAR in Afghanistan i s nothing less than rebuilding an effective state after more than two decades o f conflict and administrative decay. 5.2 Afghanistan's experience over the past five years, as well as lessons from other fragile states, provide some clues about how to go about public administration reform. The resulting challenges, and possible responses, have been discussed in some detail inprevious chapters o f this report. This chapter attempts to distil the main points into a twelve-point agenda. There i s nothing magical or immutable about this list. It i s simply intended to help shape the PAR debate within the Government o f Afghanistan and with the donor community - with the objective o f speeding up progress on reforms and realizing tangible benefits over the next 5-10 years. (1) Recognize that PAR is a long-term task -which will requirepersistent effort over time, as well as innovative approaches to improve service delivery in the short and medium terms. 5.3 It will inevitably take time to build up the capacity and credibility o f the state in Afghanistan. That i s why it i s important to initially limit the role o f the state to core functions in line with existing capacity. This is consistent with the Government's commitment to a "lean state" and the use o f non- governmental delivery mechanisms in some sectors (such as health). Experience in Afghanistan and elsewhere shows that successful innovations in service delivery can be implemented - even in an environment o f poor overall outcomes and when systemic problems are still to be res~lved.~'These innovations should be evaluated carefully over time - to help define the appropriate role o f the state in different sectors and hence the priorities for PAR in different line ministries. This bottom-up approach should constantly challenge the top-down directions o f PAR. The real litmus test mustbe: how will these reforms contribute to better governance and service delivery? 5.4 Inmany ways, the opportunistic approach adopted under PRR is still very much relevant to Afghanistan today. It i s impossible to implement a broad-based, long-term agenda in a comprehensive systematic way - especially in a country like Afghanistan where capacity i s very weak and the government has limited reach outside Kabul. Rather the framework for reform must be comprehensive and coherent, and short-term actions must be consistent with that framework. At any point intime, there mustbe a clear idea o fpriority areasrequiring attention, andthose which canbe sorted out later. 5.5 PAR in Afghanistan i s often equated with civil service reform. Indeed, civil service reform i s essential to build a professional, merit-based bureaucracy - which can deliver public services in an effective and efficient manner. But this should not be equated with building a strong, centralized bureaucracy inKabul. Mucho f the reform effort will have to be directed at deconcentrating line ministry authority to provinces and districts, and improving the capacity and coordinationo f government activities at the sub-national level. Similarly, civil service reforms are unlikely to work unless there are ways for the Afghan people to express their demand for public services and hold the government accountable for *'For a review of relevantexperience from India, see World Bank 2006a. A similar example from Afghanistan, simplifying the processofvehicleregistrationinKabul, is providedinChapter 11, Box 2.10. 70 results. Hence the importance o f broader reforms to strengthen the role o f parliament and the effectiveness o f thejudiciary, and to involve civil society and communities indecisions affecting them. (2) Provide strongpolitical leadership and effective coordination to implement the PAR agenda. 5.6 Because PAR i s a broad-based effort, it must be led from the top with effective coordination across institutions. This leads to the following recommendations: Strong leadership i s needed from the President -to provide a clear message about the direction o f reform and to manage conflicting political interests. The President should personally take responsibility for explaining the importance o f the reforms to Parliament and the public at large, and also reporting on the results achieved. Important issues need to be discussed and decided at Cabinet level (the Cabinet's PAR Sub- committee provides a convenient vehicle) to build a consensus view even when different interests are being advocated by different agencies. Implementing agencies should also report back to Cabinet on a regular basis on progress made and issues to be addressed. Instead o f attempting to implement reforms in all ministries by itself, IARCSC needs to increasingly focus on guiding and overseeing ministries' implementation, while allowing ministries take responsibility o f their ownrespective reforms (see Message 5 below). (3) Ensure that donor support is adequate and consistent with the PAR framework set by the Governmentof Afghanistan. 5.7 Although Afghanistan has received large amounts o f development assistance since the fall o f the Taliban, very little o f this has been directed specifically at the PAR agenda. Indeed, it is hard to know exactly how much assistance has been received for PAR, given the different categorizations used by different donors, and the difficulty o f separating PAR components from projects with broader objectives. Getting a better picture o f donor support for different parts of the PAR agenda would be useful initself. Butevenmore important is effective coordinationo f donor assistance bythe Government o fAfghanistan, inlinewith its ownPARpriorities. This should be done within the overall framework providedbythe I- ANDs and the Afghanistan Compact. Responsibility for donor coordination should then be delegated to the relevant government agency: for example, IARCSC for civil service reform and IDLG for sub- nationalreform. 5.8 Most o f the donor support for PAR i s in the form o f technical assistance. The main challenge now i s how to manage TA so that there i s sufficient to fill competence gaps in the short term, while replacing it with less costly and more manageable in-country-grown and national capacity over time. Initially, many had believed that the gap-filling need for civil servants' competence would last 2-3 years, and that short-term imported consultants would simultaneously train and coach their replacements while performing in-line functions. However, these assumptions proved to be overly optimistic. Furthermore, the higher salaries paid to externally-funded consultants and advisers have attracted some o f the best talent away from the government and bid up the cost of scarce talent in the country. Therefore, in the future, more attention will have to be given to the objectives o f TA, with a clearer distinction between gap filling and capacity development. Donor fundingo f salary top-ups for civil servants will also have to be carefully managed and phased out over time (see Message 4 below). 71 -(4)toimprovegovernmentperformance andservicedelivery. Match higher salariesfrom the new pay-grade scheme with other civil service reforms 5.9 The new grade structure's purpose i s to bring formality into civil service management by reducing inconsistencies and scope for arbitrary actions. It will do so by linktng grades to jobs, qualifications and responsibilities; and de-linlung them from longevity in position. The new pay scales are intended to attract, retain and motivate well-qualified staff. The new scales, ranging from $100 to $650 per month, although significantly higher than the present scales, are still less than what donors pay and may be inadequate to draw slulls into the top levels o f the civil service. Upon receiving higher salaries, civil servants will not automatically start performing better. (If this were the case, then each across-the-board pay increase for civil servants in different countries would result in citizens' noticing an increase o f public servants' performance.) The following cautions are advised in implementing the new grade-pay structure: Draw upon the lessons o f PRR. The same issues that overcame the earlier exercise could also engulf implementation o f the new scheme. With elections scheduled in Afghanistan in less than two years from now, the same pressures that led to units making superficial PRR-restructuring plans, abandoning the rigorous recruitment process, and migrating all existing staff to higher pay scales-could equally derail any logical and systematic roll-out o f the new scheme. Recognize that IARCSC i s undertaking a task o f magnitude and complexity that it has never encountered before. Although line ministries will be incharge o f re-grading the majority o f civil servants and assigning them to new grades (because the bulk o f the civil servants are inthe lower grades), it i s the IARCSC's responsibility to sequence the new structure's roll-out among ministries, providing guidance to ministries on technical issues and solving problems as they arise, enforce merit-based recruitment in the lower grades, while managing the workflow o f its own Appointments Board for recruitmentinthe higher grades. The Commission i s yet to prepare an implementation plan in discussion with the line ministries where this scheme will be implemented. To add to its duties, in advance o f implementation, IARCSC will have to draft laws and regulations, help formulate the process o f what to do with those who don't qualify, and communicate throughout the process with the public and civil servants on the objectives and progress o freform. Donors' response to the new grade-pay structure has the potential to distort it. Proposed salary levels for the upper grades will be well below levels inthe "second civil service" directly paid for by donors. Because o f the continuing scarcity o f qualified humanresources, donors are likely to continue funding salary top-ups to civil servants from project budgets. (5) Re-examine the IARCSC's role - to focus on core functions and reduce conflicts of interest within its structure. 5.10 The Commission's mandate both overloads it and creates conflicts o f interest, as described in paragraphs 2.21 and 2.22. The Commission is overloaded by having responsibility for making civil service management policy, guiding ministries' implementation, monitoring reform's progress and conducting civil servants' training-in addition to typical service commission functions o f high-level recruitment and hearing appeals against administrative actions. The potential conflicts o f interest arise because it reports to the executive while being expected to play a check and balance role. Both o f these come inthe way o f IARCSC's effectiveness. 72 5.11 A thoroughrestructuring o fthe Commission may not bepossible inthe short term, and especially not now as it embarks upon the huge challenge o f leading the implementation o f the new grade pay structure. But some short-term steps could alleviate the problem: To demonstrate the IARCSC's "independence," the Cabinet Sub-Committee's and PAR Steering Committee's supervision o f the Commission could be explicitly limited to its policy-making functions, and excluded from appointments and appeals. 0 Boththe Appeals Board and the Civil Service Institute could be distanced fkom the Commission's own line structure. It is not proposed that they be immediately detached from the Commission as many Administrative Tribunals and many civil servants' training institutes are. Rather, they could have `staff reporting arrangements to the Commission's Chairman instead o f the current line reporting arrangement. In such a lay-out, the Commission would continue to provide administrative support to these two units, but they would report directly to the Chairman. Distancing the Appeals Board will have the advantage o f upholding its credibility with civil servants whom it is meant to serve. Distancingthe Civil Service Institute, currently buried three layers deep within the Commission's structure, will allow it to negotiate with donors for their support. This i s the area o f Afghanistan's civil service management that needs the most immediate assistance, but those needs tend to get subsumed within the more general needs o f the Commission. 0 Clarifying the respective roles andresponsibilities o f Office o f Administrative Affairs (OAA) and IARCSC in civil service management, and especially in finalizing ministries' tashkeel (establishment). Donors that support civil service management could design their projects so that IARCSC's role i s more o f a facilitator than project implementer. This will not only reduce strain on an already overloaded IARCSC, it will allow line ministries to take control of their own human and financial management functions as i s intended inthe Compact and LANDS. (6) Monitorprogress on PAR -and be willing to rethink reforms when they're not meeting their goals. 5.12 The Government and IARCSC now have a systematic and well-thought out approach to PAR'S first element: civil service reform9'. The Compact benchmarks make clear what i s to be achieved; while the Government's PAR Strategy, the ANDS-assisted sector strategy, and the costed civil service implementation plan have formulated how this can be achieved over the next three years. Continuous monitoring o f progress in civil service reform will be important so that the Government can revise direction as necessary, report on progress to various stakeholders and also reassess the attainability o f different goals and benchmarks. The IARCSC i s leadingthe Government's effort on civil service reform. It is inthe Commission's interest, therefore, to not wait until the end o f the Compact period to discover that civil service targets had not been met, and instead regularly check: What progress has already been achieved? What else needs to be done to meet the Compact's targets? 5.13 Thus, IARCSC needs to establish a monitoring system that produces, on a regular basis, findings that are judged valuable by ministries, the PAR Steering Committee, cabinet and donors. It also needs to monitor line agencies' compliance with government wide rules and regulations. Towards this, the World Bank has assisted IARCSC take stock o f what part o f reform goals have already been achieved, and Benchmarksandindicators for the secondandthird element havenotyet beendeveloped. 73 suggested tools for IARCSC's monitoring activities. The following tactics could be helpful for IARCSC to perform its monitoring task: IARCSC needs to become the powerful champion within Government that generates line ministries' demand for monitoring civil service reform. Because of ministries' lack o f awareness, their demand to understand how well or not they are performing their own restructuring and personnel management i s weak. IARCSC needs to make available the tools and their potential uses, and also create incentives for ministries to monitor-perhaps by highlighting successes in Government forums and actively reaching out to ministries through training, workshops and guidance. The structural arrangements to monitor PAR should be clarified and strengthened. Currently the Administrative Reform Secretariat, headed by a Director i s responsible for monitoring PAR and also to serve the secretariat for the PAR Steering Committee. This potentially asks one line unit o f the Commission to judge the performance o f the whole Commission and its Chairman's leadership. Also, it does not have the profile to influence other ministries, or to attract donor resources. The monitoring system should not be over-engineered; rather the emphasis should be on reliable data systems. What is required i s a set o f intermediate benchmarks leading up to the Compact's targets. Locally developed and well-understood indicators are more likely to be utilized rather than borrowed or expert-developed indicators that do not reflect the program's characteristics. The need for ministries to report on their compliance with new personnel guidelines and regulations could be reinforced through an appropriate government regulation and procedures. (7) Develop a clear policy framework for the sub-national system --and a strategy for guiding sub-national reforms. 5.14 Without a clear policy framework, initiatives to improve sub-national governance and service delivery will remain piecemeal and often at odds with each other. While the Constitution and other legislationprovide a legal framework for a number o f sub-national institutions, this tends to be broad and ambiguous, and often begs more questions than it answers about who should be doing (and i s accountable for) what, the resource flows that are necessary for the effective exercise o f powers and functions, and the implementing mechanisms that need to be put inplace. Two directions o freform stand out: Inorder to reduce systemic contradiction, strengthen and consolidate the role of the Provincial Governor's Office inthe areas o f planning and coordination, while limitingits functional, budget and operational authority to deliver services, which falls under the Provincial Departments. Inorder to enhance accountability, strengthen the oversight and supervisory roleofthe Provincial Councils, to provide checks and balances on the activities o f the delivery agencies. 5.15 The recent decision to improve the central oversight o f sub-national issues with the establishment o f the Independent Directorate o f Local Governance (IDLG) i s a welcome move. Inthe past, a number o f ministries had overlapping responsibilities for the various sub-national organs, and capacity to monitor and manage any given aspect o f the sub-national system has been poor. Inaddition, ministries which are active at the sub-national level exhbit high degrees o f bureaucratic and political rivalry, making it difficult to forge common perspectives on important reform efforts. Donor coordination has also been weak, with extensive fundingflowing to sub-national initiatives outside government systems. 74 5.16 Next steps on sub-national issues should aim to: Buildupthe political credibility o fILDGwiththe powerful and disparate interests working at the sub-national level, and the technical capacity o f IDLG to tackle complex institutional and policy issues. To this end, consider forming an Inter-Ministerial Committee, chaired by the President or Vice President, with IDLG acting as the secretariat, to establish a framework for sub-national strengthening andreform. Buildonrecent effortsto improve donor coordinationthrough the Donor Working Group on Sub- National Governance and the Sub-National Governance Donor and Government Working Group under the ANDs mechanism. The IDLG should probably take over chairmanship o f this group from the IARCSC. One o f its first tasks should be to set appropriate benchmarks for developing the sub-national policy framework and reform strategy. (8) Strengthen theperformance of the main service delivery institutions by improving their organizational structures -andproviding adequate resources to do theirjob. 5.17 Line ministries will have to ensure that their functions are properly divided between their central offices and Provincial Departments. Ina number o f cases, this i s likely to involve the deconcentration o f various powers and functions from the central to the provincial level. In fact, in some cases this i s already underway (e.g., teacher recruitment). The pace o f such deconcentration will be determined, among other things, by capacity constraints - although it should be recognized that, in the absence o f functional deconcentration, capacity at the sub-national level will not be built. Ultimately, functional deconcentration drives capacity building, not the other way around. This leads to the following recommendations: e Decisions on the division o f fimctions between central and provincial offices should be built into the organization plans being prepared by line ministries as part o f PRR and integrated into the new pay and grade reformprocess. e Any deconcentration process should be crafted flexibly and opportunistically enough to allow different ministries to implement service delivery models which are most conducive to success in their sectors. Centralized authority may be more relevant inthe health sector, for example, which relies more on contracting out to NGOs for service delivery. e Ministries should not reject asymmetric approaches, and rather seize opportunities as they arise. MRRD, for example, has deconcentrated certain functions, including procurement, to its provincial office inKandahar, and this experiment seems to be working well. e The extent to which line ministries deconcentrate will need to be moderated by the degree to which this risks capture o f Provincial Departments by Provincial Governors. There i s little point in trying to improve organizational efficiency through vertical deconcentration if this makes delivery agencies more vulnerable to horizontal power which inhibitsoperational effectiveness. 5.18 Improvements in organizational structures need to be matched by the provision o f adequate resources for key organs of sub-national governance and service delivery to play their roles. Initially, this means meeting basic needs for training, IT, offices, communications etc. It should be possible to meet some o f these needs on a modular basis, i.e., to define a package o f "basic needs" equipment for Provincial Councils, for example, and to roll this out to PCs across the country. Over the longer term, it means moving towards a more equitable distribution o f fiscal resources across the country. The specific conditions o f Afghanistan make this uniquely challenging: there i s an understandable tendency to 75 concentrate expenditures inthose areas most vulnerable to security threats, and the large off-budget flows (through PRTs etc.) distort distributional outcomes, whatever efforts are made in respect o f the core budget. Arguably, this i s a second generation issue which can only really be addressed once greater stability and predictability are established invertical and sectoral flows. But it is ultimately an important question which cannot neglected indefinitely. (9) Address the particular issues facing municipalities and CDCs: the most autonomous elements of the sub-national system of governance and service delivery. 5.19 Municipalities have been established under law, as important and self-sufficient local service delivery and governance entities. CDCs were established as community-level vehicles to deliver infrastructure inrural areas under NSP and have since started to take on a wider range o f local governance functions. A common feature o f these entities, which distinguishes them from provincial-level entities, i s that, inprinciple, the elected authority inthe municipal or community area i s itself directly responsible for service delivery in that area. Inother words, they do not suffer from the accountability and autonomy problems characteristic o f other sub-national institutions. 5.20 However, they do confront significant issues o f their own. In the case o f municipalities, the following measures are needed: 0 Enact new legislation which clarifies the relationshipo f municipalities to the central and regional governing entities, while protecting their self-reliance. This should provide them with solid expenditure and revenue assignments, enhance capacity for providing local services sustainably, provide a clearer categorization o f different local administrative bodies, and establish real self- governance and downward accountability inline withthe text and spirit o fthe Constitution. Formulate a comprehensive financing framework for municipalities which clarifies and rationalizes own-source revenue assignments, introduces measures to strengthen municipal administration, and establishes a transparent rules-based system to allocate central transfers equitably and rationally among municipalities. This should be matchedby improved budgeting, financial management and accounting systems, and capacity to administer these systems within municipal administrations. 0 Introduce elections for mayors and municipal councils at the earliest feasible opportunity. Institutional measures should be developed to deal with the mayor-CDC interface, and to bridge the current gap between the bottom-up system for expression o f community voice and top-down planning and budget decision malung. 5.21 For CDCs, the following three issues needto be addressed: 0 CDCs needto secure a regular and reliable funding flow to resource local development activities which provide their ruison d'etre. The primary initiative here should be to establish a reliable, ongoing funding mechanism which gives CDCs a source o f dedicated funding which they can then allocate to local projects at their discretion. This should be distributed between CDCs according to a simple and transparent formula. Given international standards, and to ensure sustainability, the fundingshould be set at far lower levels than under NSP. 0 CDCs' roles should be consolidated on the basis o f the new by-law, i.e.: (a) allocating their own funding, and planning and implementing local projects using this funding; (b) a limited administrative role, particularly for civil registration and land verification; (c) planning and allocation at the district or sub-district level where CDCs are able to federate or cluster; and (d) providing information on service delivery and governance to PCs to assist them with their 76 accountability function. In general, CDCs should take on only public good functions, and not compete with other local governance bodies (such as school management committees). The matter o f "ownership" o f CDCs at the central level will have to be settled within the central government by those agencies and ministries which have a core interest inthe question. The best way forward would probably be to have this issue looked into and decided by the Inter- Ministerial Committee proposed above. (10) Demonstrate Government's will tofight corruption. 5.22 The public has little trust in leaders and government institutions. Despite positive rhetoric, Government has not initiated action in the face o f widespread and entrenched corruption. Although the Bonn accord stressed accountability o f public organizations, the political process that followed stressed representation from all quarters above accountability, and peace before justice. Opium production rose substantially in2006 and 2007 from previous levels, and Afghanistan is now responsible for 93 percent o f total global illicit opium output. Institutional disarray and in-fighting within government, and lack o f strong government leadership to confront corruption has resulted in the widely held view is that corruption i s being ignoredor tacitly allowed. This report recommends: 0 Initial implementation o f meaningful yet feasible short-run measures. This would mean emphasizing improvements at sector and agency levels instead o f spending too much effort and resources on preparing an impressive strategy. This could focus on a few agencies where the public interacts most frequently with service providers. Inthese agencies, the number o f steps, signatures and visits could be reduced. Computerization, communication with the public, and replacingthe heado fthe organization worked well inthe motor vehicles licensing office. 0 Clarifying institutional arrangements so that different agencies involved in anti-corruption work harmoniously. The current state o f institutional coordination and weak criminal justice system provide few prospects of investigation and prosecution. Predictability o f Government's actions, and pro-actively sharing information with the media and civil society organizations will demonstrate government's will. Enforcing ministries to respond to audits and involvingMinistry o fFinance inthe response to audit reports will allow itto play a stronger oversight role. Strengthening the leadership o f institutions involved in the fight against corruption, which must meet highethical and professional standards. Muchmore can be done to go after key actors inthe drug industry (larger and medium drug traders, and their sponsors inside and outside government). Even if weaknesses in the justice sector make prosecution and conviction very difficult, officials associated with the drug industry should at least be removed from their positions ingovernment, which would send a strong signal o f government commitment to address the drugissue. (11) Build public trust in parliament and the judiciary - so they can provide formal oversight in an effective and credible manner. 5.23 Public trust in government organizations is currently low; there is need to quickly gain public confidence. These organizations, having high visibility, have the best chance to demonstrate that they themselves are abiding by the rules o fthe game, and win people's trust. 5.24 IntheNationalAssembly, workisneededtobuildparliamentarians' understandingoftheir actual role, including how they can constructively contribute to law-making and oversight o f the executive. 77 Particularly, Budget Committee members' comprehension of how Government budget and accounts are reported needs to be strengthened. 5.25 Inthejustice sector thisreportrecommends: Recognizing that security strengthening measures are only partially effective without accompanying strengthening o f rule o f law. Even a well-trained police force cannot enforce law without a functioning criminal justice system. Rather, those responsible for security could become the sources of injustice. Strengthening coordination between the NA and Cabinet can improve legislative capacity. Starting small and scaling up. Infrastructure projects can have limited impact without institutions, butjudges need courthouses and legal texts. Bothjudicial reform and infrastructure creation should focus on the major cities because that i s where the formal justice system is most used and most needed. Focus judicial reform on provincial rather than district courts. An improved provincial court could provide a check on as-yet unreformed district courts throughout the province. However, constructing courts, and appointing sufficient trained judges will take time. What needs to be established much sooner is the role o f the judiciary, and especially the Supreme Court, in enforcing constitutionality of laws and regulations, and reviewing legality o f administrative decisions. Supporting non-state dispute resolution. Traditional conflict resolution mechanisms such as shuras can reduce conflict because they are more focused on negotiating and mediating disputes rather than the adversarial win-lose outcomes o f the formal court system. Moreover, this may be all that i s available for many years. (12) Provide credible information on governmentprograms andperformance -and open up governmentprocesses topublic scrutiny andparticipation. 5.26 The media and CSO sectors in Afghanistan are still finding their feet. Donor support can help to develop their capacity and effectiveness over time. But, inthe area o f public accountability, their role i s very much dependent on their access to official information and their ability to provide feedback on government performance. Donors should support government initiatives to improve transparency, buildingon innovations in donor-funded projects, while being sensitive to the country's history, culture and politics. Inparticular, donors could: Encourage a public debate about corruption and practical steps that can be taken to fight it. GOA should take the lead, possibly using the new ranlungs o f Afghanistan in the TI Corruption Perceptions Index to inform the debate. Support efforts to gather baseline information incritical areas such as poverty reduction (such as through the National RiskandVulnerability Survey) and service delivery (as i s done inthe health sector). Inparticular, donors could help strengthen the Central Statistics Organizations capacity to do this survey work ina more systematic manner. 0 Build communication components into their own projects, and provide public information on project objectives, expenditures and results achieved. Encourage public participation in project planning and implementation, by providing project information to affected communities and involving them in project decision malung. The National Solidarity Program already requires CDCs to monitor project implementation based on a 78 few selectedkey indicators; and also to make that information available to communities through regular briefings and information sharing. Improveinformation flows among donors incritical areas such as narcotics control. 79 REFERENCES CHAPTERI ADB, UNDP andWorld Bank. 2002. Afghanistan: PreliminaryNeeds Assessment for Recovery and Reconstruction. Carnahan, Michael, Nick Manning, RichardBontjer, StephaneGuimbert, eds. 2004. Reforming Fiscal and Economic Management inAfghanistan. World Bank and Islamic Republic o fAfghanistan, Directions inDevelopment Series. Cliffe, Sarahand Nick Manning. Forthcoming.BuildingState Institutionsafter Conflict. International PeaceAcademy. Evans, Anne, NickManning, Yasin Osmani, Anne Tully and Andrew Wilder. 2004a. A Guide to Government inAfghanistan. AREU and World Bank. Evans, Anne, Nick Manning and Anne Tully, with Yasin Osmani and Andrew Wilder. 2004b. Subnational Administration inAfghanistan: Assessment and Recommendations for Action. AREUandWorldBank. Evans, Anne, and Yasin Osmani. 2005. Assessing Progress: Update Report on Subnational Administration inAfghanistan. AREU. Fukuyama, Francis. 2004. StateBuilding: Governance and World Order in the Twenty-First Century. Profile Books. Hakimi, Eklil, NickManning, SatyendraPrasad, and Keire Prince. 2004. Assymetric Reforms: Agency- LevelReforms inthe Afghan Civil Service. South Asia Region, PREM Worlung Paper Series, World Bank, Washington, DC. IntegrityWatch Afghanistan. 2007. Reconstruction National Integnty System Survey: Afghanistan 2007. Islamic Republicof Afghanistan and InternationalAgencies. 2004. Securing Afghanistan's Future: Accomplishments and the Strategic Path Forward. Islamic Republic of Afghanistan. 2006. AfghanistanNational Development Strategy: An InterimStrategy for Security, Governance, Economic Growth and PovertyReduction. Islamic Republic of Afghanistan. 2007. Draft Anti-corruption Roadmap for the Islamic Republic of Afghanistan. Levy, Brian. 2006. Governance Assessment: Towards an Actionable Approach. Draft Paper. Lister, Sarah, and HamishNixon. 2006. Provincial Governance Structures inAfghanistan: From Confusionto Vision. AREUBriefingPaper. Lister, Sarah. 2006. Moving Forward? Assessing Public Administration Reform inAfghanistan. AREU BriefingPaper. 80 Lister, Sarah. 2007. Understanding State-Buildingand Local Government inAfghanistan. Crisis States Research Centre, Working PaperNo.14. London Conference on Afghanistan. 2006. The Afghanistan Compact. Maley, William. 2006. Rescuing Afghanistan. UNSW Press. Michailof, Serge. 2007a. Assessment o f ProgressMade inthe Implementation ofMerit-Based Appointments inthe Afghan Civil Service. Consultant Report for IARCSC. Kabul. Michailof, Serge. 2007b. Review of Technical Assistance and Capacity BuildinginAfghanistan, Discussion Paper prepared for the ADF. North, Douglas, John Wallis and Barry Weingast. 2006. A Conceptual Framework for Interpreting Recorded HumanHistory. National Bureauo f Economic Research, WorkingPaper. Pritchett, Lant and MichaelWoolcock. 2002. Solutions when the Solution i s the Problem: Arraying the Disarray inDevelopment. Center for Global Development, Worlung PaperNo. 10. Rashid, Ahmed. 2001. Taliban. Yale UniversityPress. Rubin, Barnett. 2007. The Political Context ofPublic Administration ReforminAfghanistan. Revised draft background paper commissioned for this report. Suhrke, Astri. 2007. Reconstruction as Modernisation:The `Post-Conflict' Project inAfghanistan. To be publishedinThirdWorldQuarterly, October 2007. UNAMA(UnitedNationsAssistanceMission inAfghanistan). 2007. Afghanistan Justice Sector Overview. UNOffice onDrugs andCrime, andthe WorldBank. 2007. Afghanistan's DrugIndustry:Structure, Functions, Dynamics, and Implications for Counter-Narcotics Policy. UNOffice onDrugsandCrime. 2007. AfghanistanOpiumSurvey 2007. ExecutiveSummary. World Bank. 1997. WorldDevelopment Report 1997: The State in a Changing World.Washington, DC. World Bank. 2003. WorldDevelopment Report 2004: Making Services Workfor Poor People. Washington, DC. World Bank. 2005a. Afghanistan: StateBuilding, Sustaining Growth, and Reducing Poverty. A World Bank Country Study, Washington, DC. World Bank. 2005b. Afghanistan: Poverty, Vulnerability and Social Protection: An InitialAssessment. Washington, DC. World Bank. 200%. TheInvestment Climate in Afghanistan: Exploiting Opportunities in an Uncertain Environment. Washington, DC. 81 World Bank. 2005d. Fragile States: Good Practice inCountry Assistance Strategies. Operations Policy and Country Services. Washington, DC. World Bank. 2006a.India: Reforming Public Services in India, Drawing Lessonsfrom Success.Sage Publications. World Bank. 2006b. InterimStrategy Note for Islamic Republic of Afghanistan for the PeriodFY07- FY08. Washington, DC. World Bank. 2006c. Islamic Republic o f Afghanistan: InterimPoverty ReductionStrategy and Joint IDA- IMFAdvisory Note. Washington, DC. World Bank. 2006d. Afghanistan: Managing Public Financesfor Development. Washington, DC. World Bank. 2006e.Annual Review of Development Efectiveness 2006: GettingResults. Independent EvaluationGroup. Washington, DC. World Bank. 2007. Strengthening World Bank Group Engagement on Governance and Anti-Comption. Washington, DC. World Bank and MF. 2006. Global MonitoringReport 2006: Millennium Development Goals, Strengthening Mutual Accountability: Aid, Tradeand Governance. Washington, DC. CHAPTER I1 BritishCouncil. 2005. Quarterlyreports under the IDA SecondEmergency Public Administration Project. Consultant Report for IARCSC. Kabul. Caddy, Joanne, Tiago Peixoto, and Mary McNeil. 2007. Beyond Public Scrutiny: Stocktaking of Social Accountability in OECD countries. OECD (Organization for Economic Cooperationand Development) and World Bank: WashingtonDC. Creighton, Paul. 2006. Report from the review of outcomes and progress o fmeritbased appointments system. Consultant report for IARCSC. Kabul. Evans, Anne, NickManning,Yasin Osmani, Anne Tully and Andrew Wilder. 2004. A Guide to Government inAfghanistan. AREU and World Bank. Gopalakrishnan, Krishnamurthy, NdolaPrata, Dominic Montagu, BrianMitchell, and Julia Walsh. 2002. Case study Janani-India, inNdolaPrata and FionaGreig(eds.) NGOsProviding Low Cost, HighQuality Family PlanningandReproductiveHealthServices, University ofCalifornia, Berkeley: Bay Area International Group Monographseries Vol 1N o 2. Hakimi, Eklil, Nick Manning, SatyendraPrasad, and Keire Prince. 2004. Asymmetric Reforms: Agency- Level Reforms inthe Afghan Civil Service. South Asia Region, PREM Workmg Paper Series, World Bank, Washington, DC. IARCSC (Independent Administrative Reform& Civil Service Commission). 2005. Training & Development (T&D) Policy & Structure for Government of Afghanistan. 82 Islamic Republic of Afghanistan and InternationalAgencies. 2004. 2004. Securing Afghanistan's Future: Accomplishments and the Strategic PathForward. Islamic Republic ofAfghanistan, Ministryof Finance. 2006. Briefon the proposed Grade Pay Structure. Islamic Republic of Afghanistan, Ministryof Finance. 2007. Budget estimates for 1386. Islamic Republic of Afghanistan, Ministry of Public Health. 2006. Afghanistan Health Sector Balanced Scorecard: National and Provincial Results Round Three.By Ministryo f Public Health, General Directorate o f Policy and Planning, Monitoring and EvaluationDepartment with Johns Hopkins University Bloomberg School o f Public Health andIndian Institute of HealthManagement Research. Kabul. Loevinshon, Benjamin 2008. Performance-Based ContractingFor Health Services inDeveloping Countries - A Toolkit. UnpublishedWorld Bank document. Michailof, Serge. 2007a. Assessment o f ProgressMade inthe Implementation of Merit-Based Appointments inthe Afghan Civil Service. Consultant Report for IARCSC. Kabul. Michailof, Serge. 2007b. Review of TechnicalAssistance and Capacity BuildinginAfghanistan, Discussion Paperprepared for the ADF. Palacios, Robert and Oleksyy Sluchynskyy. 2007. National Pension Scheme: FromCrisis Management to Comprehensive Reform Strategy. Draft internal note by the South Asia Human Development Unit,WorldBank, Washington, DC. Parto, Saeed. 2007. Sector Reform inPublic Health, Education, and Urban Services-Evidence from Kabul andHerat. Rojas, Carlos, Alexandria Valerio, and Angela Demas. 2005. Decentralizing EducationinGuatemala: school management by local communities, WashingtonDC: The World Bank, EducationNotes February 2005. Rubin,Barnett. 2007. The PoliticalContext ofPublic Administration Reform inAfghanistan. Revised draft background paper commissioned for this report. The Services Group, 2007. RapidReview ofPRR. Consultant Report for IARCSC. Kabul. UNDP 2007a. Quarterly ProgressReport, SecondQuarter 2007: Support to Center o f Government Project. http://www.undp.or~.af/WhoWeAre~PinAf~hanistan/Projects/Reports/SCoG/4. percent20SCoGpercent20Q2 percent202007 percent20Progresspercent20Report.pdf Wilder, Andrew. 2007. Cops or Robbers?IssuesPaper Series, Kabul. AREU (AfghanistanResearchand Evaluation Unit). 83 CHAPTER I11 Center for Strategic andInternationalStudies. 2007. Breaking Point -MeasuringProgress in Afghanistan. Chstensen, Asger and Asta Olesen. 2007. Afghanistan: Assessment of the Scope for Strengtheningthe Role ofNSP Community DevelopmentCouncils inLocal Governance. Draft backgroundpaper commissioned for this report, World Bank, Washington DC. Goga, Soraya, Mihaly Kopanyi, Bappa Chakravarty and Wali Ibrahimi. 2007. Municipalities in Afghanistan: A BriefReview of their Framework for Operations. Draftbackgroundpaper commissioned for thisreport, World Bank, Washington DC. Islamic Republic of Afghanistan. 2007a. Provincial Development, Provincial Budgetingand Integration of the ProvincialDevelopment Plans into the Afghan National Development Strategy. Draft Discussion Paper for the ADF. Islamic Republic of Afghanistan. 2007b. Capacity Development Plan for the Civil Service in Afghanistan". Draft Discussion Paper for the ADF. Rubin, Barnett. 2007. The Political Context ofPublic Administration ReforminAfghanistan. Revised draft background paper commissioned for this report. White, Roland, and SarahLister. 2007. Service Delivery and Governance at the Sub-National Level in Afghanistan. Backgroundpaper commissioned for this report, World Bank, Washington DC. CHAPTER IV Ackerman, JohnM.2005. Social Accountability inthe Public Sector: A Conceptual Discussion. Social Development Papers 35733. WashingtonDC: World Bank. Ayrapetyants, Anika. 2005. Afghanistan Civil Society Assessment: Initiative to Promote Afghan Civil Society. Counterpart. http://www.reliefweb.int/librarv/documents/2005/cp-afg-3iun.pdf Barfield, Thomas, Neamat Nojumi, and J. Alexander Thier. 2006. The Clashof Two Goods: State and Non-State DisputeResolution inAfghanistan. UnitedStates Institute o fPeace. http://www.usip.org/ruleoflaw/proiects/clash two goods.pdf Boesen, Inger, 2004, From Subjects to Citizens: Localparticipationinthe National Solidarity Program. AREU Workmg Paper Series. Kabul: AfghanistanResearchand EvaluationUnit. BritishCouncil Afghanistan. 2004. Afghanistan Cultural Profile. BritishCouncil. h~://WWW.culturalprofiles.net/afphanistan/Directories/AfPhanistan Cultural Profile/:G.html Cliffe, SarahandNick Manning. Forthcoming.Cross-Cutting Challenges. InBuilding States to Build Peace, ed. Charles T. Call and Vanessa Hawluns Wyeth. Boulder: Lynne Rienner Publishers. Delesgues, Lorenzo, and Yama Torabi. 2007. Reconstruction National Integnty System Survey: Afghanistan 2007. Integnty Watch Afghanistan, London: TIRI. http://www.tiri.org/dmdocuments/RNISSpercent2OAfghanistan.pdf. 84 Esfandiari, Golnaz.2005. Afghanistan: HumanRights Commission Strives for Justice. Radio Free Europe http://www.rferl.org;lfeaturesarticle/2OO5/lO/ac3308eb-df83-473 8-9f6a-efOb86e9a8fa.html Harpikven, Kristian, Arne Strand and KarinAsk. 2002. Afghanistan and Civil Society. Consultant report commissioned by the Norwegian Ministry o f ForeignAffairs, PeshawarBergen. Islamic Republic of Afghanistan. 2007. Anticomption Roadmap for the Islamic Republic of Afghanistan. Paperpresented at the Afghanistan Development Forum, Kabul, April 30, 2007. http://www.adf.g;ov.af/src/stratee;y uauerslhti-Corruptionuercent20Roadmau uercent20GoA percent20- uercent2OEng;lish.udf Islamic Republic of Afghanistan, Ministryo f Education. 2007. National Education Strategy 1385-1389: Towards Equal Access to Quality Educationfor all, Implementation Plan 1386. Paper presented at Afghanistan Development Forum, Kabul, 30 April, 2007. http://www.moe.g;ov.af/Misc/Nationaluercent20Educationuercent20Strateg;icuercent20Plan.pdf Lockhart, Clare. 2006. The Aid RelationshipinAfghanistan: Struggling for Government Leadership. Global Economic Governance Program, ManagingAid Dependency Project, GEGWorkmg Paper2007127, Oxford University. httD://www.globaleconomicg;overnance.ore;/docs/Lockhart Afghanistan 2007-27.udf Moore, D.2005. Civil Society and Law Reform inAfghanistan. InternationalJournalfor Non-Pro$t Law 8( 1). h ~ : / ~ w ~ ~ . i c n ~ . o r g ; ~ o ~ ~ l e d ~ e / i i1n.hm#oedn2i s s l ~ a ~ ~ ~ v l 8 Padamsee, S. 2004. Towards a Legitimate Civil Society inAfghanistan. Review o f International Social Questions. http://www.risq.oralarticle39 1.html Sherman, Zoe. 2006. Afghanistan's Constitutions: A Comparative Study and their Implications for Afghan Democratic Development. Naval Post-Graduate School, Monterrey. The KillidGroup and InterPressServices. 2007. Mediais Development: ReportofProceedings. Paper preparedfor Afghanistan Media and Civil Society Forum, Kabul, March 28-29, Kabul. http://www.iusnews.net'afghanistan/MedaiReuort.PDF UNDP (UnitedNations Development Program). 200%. MidYear Project Report 1'' and 2"dQuarter 2007: Accountability and Transparency Project. http://www.undp.ore;.af/WhoWeAre/UNDPinAfghanistan/Projects/Reports/ACT/ACT percent20Combined percent2001percent20& percent2002 uercent20Pronress percent20Reuort 17Seut07.pdf Wilder, Andrew. 2005. A HouseDivided? Analyzing the 2005 Afghan Elections. Afghanistan Research and EvaluationUnit. http://www.areu.org;.af/publicationslAuercent20House percent20Divided.udf World Bank. 2005e. Afghanistan: ManagingPublic Finances for Development. ImprovingPublic Financial Management inthe Security Sector. Report No. 34582-AF, Washington DC: World Bank. 85