Page 1 THE GOVERNANCE OF SERVICE DELIVERY IN ECA A REGIONAL STUDY ECAVP World Bank June 30, 2004 This study was prepared by Brian Levy, with guidance from an advisory group comprising Arup Banerji, Bill Dillinger, Alexandre Marc, Helga Muller, Gary Reid and Lee Travers. The study was sponsored by Cheryl Gray and Pradeep Mitra, both of whom have provided valuable inputs, and financed by the ECA Regional Studies Program. Peer reviewers were Shanta Devarajan, Orsalia Kalantzopoulos, Shekhar Shah and Sanjay Pradhan. Bjorn Dressel provided excellent research assistance, and was responsible for drafting the majority of the sector notes included in Volume 2. Additional research assistance was provided by Leah Cohen. Thanks also for helpful feedback (including reviews of the sector-country notes in volume 2) to Enis Baris, Lilia Buruncuiuc, Cevdet Denizer, Halil Dundar, Armin Fidler, Tim Gilbo, Ron Hood, Robin Horn, Monika Huppi, Chris Lovelace, Maureen Mclaughlin, Michael Mills, Michael Murtaugh, Andreas Rohde, and Sudipto Sarkar. 36141 Page 2 2 TABLE OF CONTENTS (Volume 1) Chapter 1 and Executive Summary: The Governance of Service Delivery: Strengthening ECAs Operational Approach Patterns of Accountability for Service Delivery 4 The long route of accountability 6 The intermediate route of accountability 9 The short route of accountability 11 Comparative empirical assessment 12 Operational implications 14 Improving governance – the service delivery lens 14 From service delivery to good governance more broadly 20 Getting the (im)balance right 23 2: The Analytical Framework 26 Issues and approach 26 The Accountability framework 28 Fitting the framework to country realities 29 Fitting the framework to sectoral realities 32 3: The Long Route of Accountability 35 The compact 35 Voice, linked to service delivery (“serdel voice”) 40 4: The Intermediate Route of Accountability – Decentralized Service Delivery 43 Sub-national electoral voice 43 The center-local compact 45 The local-level arrangements for service delivery 49 5: The Short Route of Accountability – Client Power 51 Client power through choice 51 Client power through sectoral participation 52 Community empowerment through social investment funds 54 Page 3 3 Chapter 1: The Governance of Service Delivery: Strengthening ECA’s Operational Approach This regional study has two objectives. The principal objective is to explore the extent to which the framework laid out in WDR2004, Making Services Work for Poor People might help strengthen World Bank efforts to support improvements in the delivery of services in Europe and Central Asia (ECA). The second objective is to explore the extent to which a ‘service delivery’ entry point can help provide renewed focus and momentum to efforts to improve governance and public management in ECA countries which appear locked in a dysfunctional political equilibrium. WDR04’s approach to strengthening service delivery is to focus narrowly on the institutional arrangements for assuring accountability for performance. WDR04 motivates this focus on accountability as follows: Too frequently those seeking improvement have focused only on internal organizational reforms – focusing on management of the frontline workers. If organizational failures are the result of deeper weaknesses in institutional arrangements (weak political commitment, unclear objectives, no enforceability), direct attacks on the proximate determinants (more money, better training, more internal information) will fail.” 1 While the accountability perspective is indeed a welcome corrective to a narrowly bureaucracy-oriented approach to strengthening public management, it also risks imbalance. Intensified pressure for performance without corresponding efforts to strengthen the internal capacities of the public sector – including through training, organizational system improvements, and efforts to build a service-oriented esprit-des- corps among public sector workers – can backfire, raising then dashing expectations, and reducing morale all round. The challenge, which this chapter addresses, is to find a creative balance between intensified demand-side pressures for performance, and corresponding efforts to build capacity on the supply-side. Within the confines of its accountability framework, WDR04 does not offer a ‘one size fits all’ approach but rather is explicit that “no single solution fits all services in all countries.” 2 Assessing the fit between country and sectoral conditions on the one hand, and the preferred approach to accountability on the other requires an assessment both of country conditions and of the country- and sector-specific service delivery arrangements. This implies that a comparative case study approach is the preferred methodology. The comparative analysis focuses on three sectors in each of five countries. The three sectors are ones chosen by WDR04 for intensive scrutiny – education, health, and urban drinking water. The countries selected for comparative analysis are Albania, the Kyrgyz Republic, Romania, Tajikistan and Turkey. Chapter 2 provides further detail on the analytical starting points, and on the selected countries and sectors. 1 WDR04, p. 58 2 WDR04, p.12 Page 4 4 This chapter (which serves also as an executive summary) is presented in two parts. The first part summarizes the empirical findings (presented in more detail in chapters 3-5, including examples from selected countries and sectors, and in Volume II) as to how the accountability relationships actually operate on the ground in the selected country-sector observations. The second part details the policy implications of the analysis for ECAVP. The implications are considered both through the narrow lens of service delivery and from the broader perspective of efforts to foster good governance more broadly. Patterns of Accountability for Service Delivery Figure 1 reproduces the WDR04 ‘accountability triangle’ which depicts the key principal-agent accountability relationships which link service providers, citizens/clients, and politicians/policymakers. 3 The first two accountability relationships are those between citizens/clients and politicians/policymakers ( voice ), and between politicians/policymakers and service providers ( compact ). 3 As the WDR details, each principal-agent relationship comprises five distinct facets. Principals delegate responsibilities to their agents; finance them (directly or indirectly) to carry out these functions; and enforce (including monitor and evaluate) compliance with the agreed delegation. Agents perform on the basis of agreements; and inform their principals on progress in implementation. Page 5 5 Taken together these first two relationships comprise what the WDR terms the · long route of accountability in which citizens hold politicians accountable through ‘voice’, and politicians/policymakers in turn hold providers accountable through ‘compacts’. As WDR04 clarifies, the voice-compact nexus of accountability relationships can be evident both at national and subnational levels. This regional study thus defines an · intermediate route of accountability in which some responsibilities and resources are de-centralized to subnational governments (which, in turn, can institute a variant of the long route at the subnational level). The third accountability relationship is that between citizens/clients and service providers ( client power ). The WDR refers to this as the · short route of accountability which, as discussed further below, can be in part a substitute for, and in part a complement to, the intermediate and long routes of accountability. What empirical patterns might be expected vis-à-vis how these routes are applied across countries? Recall the WDR04 proposition that “no accountability solution fits all services in all countries”. Since the five countries were selected to capture diversity in their institutional starting points, it follows that, if service delivery accountability strategies were well aligned with country’s starting points, even within the same sector the strategic emphasis would be different across countries – and that, if strategies were poorly aligned, the development impact would be mediocre. [Note the word ‘emphasis’ ; as will be developed in some detail in this chapter, the intent is not to identify a single, narrow set of actions on which all efforts to improve service delivery should focus, but rather to help provide some guidepost for prioritization, even as some efforts continue across multiple activities. Put simply, the relevant decision is not one of ‘ either/or’ , but nor is it entirely both/and. Rather, for each activity it is on the margin of ‘more or less?’] WDR04 provides a bold, heuristic starting point for distinguishing across countries in their institutional characteristics. Its key distinction is between ‘pro-poor’ and ‘clientelistic settings: “Pro-poor settings are those in which politicians face strong incentives to address the general interest. Clientelist political environments are those in which, even though the average citizen is poor, politicians have strong incentives to shift public spending to cater to special interests, to core supporters, or to ‘swing’ voters…….the payoffs in service delivery for assessing whether the environment is pro-poor or clientelistic can be high. Even if the politics are clientelist, policy choices can be made that are likely to yield better results than the misguided application of policies that work well only in pro-poor environments….” 4 4 WDR04, pp, 80, 90 Page 6 6 Of course, most countries cannot be pigeonholed neatly into one category or the other, but fall somewhere between the extremes of a clientelist/ pro-poor spectrum. Moreover, as discussed further in Chapter 2, governance is a multi-dimensional phenomenon, not one that can be captured along a single continuum. Even so, the pro-poor/clientelistic distinction offers a useful starting point for aligning accountability strategies with country characteristics: · An emphasis on client power – the short route – is identified in WDR04 as especially relevant in clientelistic settings, where governments lack the capacity and incentives to facilitate a compact, and where citizens are unable to translate their voice into effective pressure for better public sector performance (though, as discussed below, even in stronger states, the short route can be a useful complement to other mechanisms of accountability). As chapter 2 details, among the five country sample, Tajikistan best fits the characteristics of a weak state. · The long route which combines both a technically demanding compact, and politically inclusive voice is best aligned with institutionally strong countries. Romania and Turkey are the institutionally strongest of the five sample countries. · The intermediate route seems especially relevant in settings which are pro-poor and relatively de-centralized. As discussed below, and detailed in chapter 4, Albania and Romania made quite far-reaching moves in the direction of decentralization in the 1990s, the Kyrgyz Republic is in the process of doing the same, and the issue is a central part of an ongoing reform agenda in Turkey. The Long Route of Accountability As highlighted in Figure 1, the ‘long route of accountability’ comprises two interlinked principal-agent relationships – ‘ compact ’ and ‘ voice ’. The compact can be viewed as WDR04s contribution to the challenge of assuring better performance by central public bureaucracies. The service delivery role assigned to public bureaucracies in the WDR04 is a distinctive and controversial one. Highlighting the poor track record in many countries, the WDR argues strongly that the conditions are very stringent under which bureaucratic provision of services through central line ministries is likely to be effective, and focuses rather on an approach which highlights a clear separation between policymakers and front-line providers as key to effective service delivery. Such a separation, argues the WDR, enables both policymakers and providers to be more effective. It enables policymakers to define clearly what should be done, to set performance targets as a basis for monitoring the performance of frontline providers 5 and, more broadly, to govern the service activity via a regulatory system. And it enables front-line providers to be given the authority to decide how things are to be 5 The question of what kinds of performance targets are appropriate is a subject in itself. For now, it must suffice to note that a compact need not depend solely on service delivery outcome measures, but could also proceed using process and output indicators (school enrollment, or teacher absenteeism, for example). Page 7 7 done, within a framework of accountability for results. It also opens up the prospect for delivery by multiple front-line providers, potentially in a competitive market. Table 1: The Accountability ‘Compact’: 14 sector-country observations Romania Turkey Kyrgyz Albania Tajikistan EDUCATION M&E extensive extensive marginal some marginal Arms-length some marginal marginal marginal marginal Regulation Some (?) marginal marginal marginal marginal Compact average a/ 2.3 1.7 1.0 1.3 1.0 HEALTH M&E some Some* some marginal marginal Arms-length extensive Some* some marginal marginal Regulation marginal Some* some marginal marginal Compact average a/ 2.0 2.0 2.0 . 1.0 . 1.0 URBAN WATER M&E some Some* n/a marginal marginal Arms-length extensive Extensive n/a some some Regulation marginal marginal n/a marginal marginal Compact average a/ 2.0 2.0 1.3 1.3 Page 8 8 Notes : a/ average calculated on basis of: marginal=1; some=2; extensive=3 * reforms in early stage of implementation imply shift from marginal/some to some/extensive n/a no information available Page 9 9 Critics of the WDR suggest that its emphasis on the compact simply defines away a central problem of public service delivery. It is precisely because performance cannot be specified precisely that delivery finds itself in the public realm – and that all the messy instruments of public management (internal control systems, fostering an esprit des corps etc) become necessary. Further, even if the problem of specifying performance targets can be solved, putting a ‘compact-style approach in place is organizationally and politically complex, implying that compact mechanisms are a better fit for capable than for weaker states. Table 1 summarizes the role of accountability compacts among the 14 country- sector observations. The table reports the intensity of use across the five countries and three sectors of each of three facets of the compact: · an arms-length relationship which gives providers some flexibility to decide how resources are deployed; · a monitoring and evaluation system capable of assessing how well front-line providers are performing; and · a regulatory system which defines the ‘rules of the game’ within which providers are expected to operate. An average intensity-score is reported for each country observation. Consistent with expectations, the most immediately evident pattern is the differential role of compacts in one of the two institutionally strongest countries (Romania), and the weakest (Tajikistan). In the latter country, none of the elements of a compact are in place in any of the sectors. By contrast, in Romania all three sectors have in place significant elements of the compact. Chapter 3 offers further detail on how the different elements of the compact have been implemented in the sample countries and sectors. Voice, linked to service delivery (“serdel voice”) 6 comprises an ‘information strategy’ which brings into the public domain information on the performance of public agencies which can enable citizens to hold politicians, policymakers and providers accountable for the efficiency and effectiveness with which they deliver services. WDR04 argues strongly that to be useful such information must be specific: “identifying specific government decisions, specific decisionmakers, and the effect of the decision on voters individually or as a group …… Information about broad aggregates of public sector performance – whether based on surveys, budget studies or other methods – is less likely to be politically relevant. Why?….. At the end of the day, these efforts tell citizens what they already knew – that services are bad…... What citizens do not have, and what they need help in getting, is information about how bad their neighborhood’s services are relative to others’ and who is responsible for the difference.” 7 Table 2 reports on the extent to which voice of the kind described above (that is, voice linked to service delivery – referred to below as ‘ serdel voice’ ) was evident in each 6 As discussed further in chapter 3, electoral voice also is a relevant part of the long route. 7 WDR04, pp.87, 89. Page 10 10 of the 14 sector-country observations. Overall, the pattern is uneven, with the distribution of effort skewed strongly towards the low side: there is only one instance of ‘extensive effort, five instances of ‘some effort’, and eight instances of only ‘marginal’ effort. Table 2: Voice, linked to service delivery : 14 sector-country observations Romania Turkey Kyrgyz Albania Tajikistan VOICE IN EDUCATION extensive some marginal some marginal VOICE IN HEALTH some Some* some marginal marginal VOICE IN URBAN WATER marginal marginal n/a marginal marginal * reforms in early stage of implementation imply shift from marginal/some to some/extensive The ‘Intermediate Route’ of Accountability -- de- centralized service delivery As used here, 8 the intermediate route of accountability refers to the de-centralization of responsibility for services to local governments. Three sets of accountability relationships seem key for subnational service delivery. · subnational electoral voice; · the center-local compact; and · the local level arrangements for service delivery. To begin with subnational voice, a necessary condition for de-centralization to be credible as a distinctive accountability arrangement is that it is underpinned by a tier of government (municipal or provincial) which is downwardly accountable to its citizens via electoral voice. As Chapter 4 (which comprises a more in-depth discussion of the intermediate route) details, subnational electoral voice formally is in place in all of the countries being considered here, other than Tajikistan. The second set of accountability relationships relevant for the intermediate route comprises the compact between central and local governments. The compact between central government on the one hand and local authorities and front-line providers on the other comprises at least three facets: assignment of roles; provision of fiscal resources; and regulatory, fiduciary and other forms of oversight. Table 3 summarizes the role and 8 The term is not used explicitly in the WDR. Page 11 11 resource patterns across the 14 country-sector observations. As the table suggests, there are important sector-specific differences in the extent of de-centralized service delivery, and some important outstanding challenges in clarifying the assignment of roles and responsibilities. Turnin g to central government’s regulatory, fiduciary and administrative oversight mechanisms necessary to underpin a central-local compact, the limited attention given to regulatory oversight was evident in Table 1. How fiduciary and administrative oversight of subnational authorities by central government is handled also can have important implications for the effectiveness of de-centralization: · For fiduciary oversight, in the absence of reliable mechanisms for monitoring how local governments spend resources, central governments have tended to resort to controls. · For administrative oversight, while de-centralization has the potential to enable local governments to set their own terms and conditions of employment, in practice central governments (or public sector employee associations) tend to constrain flexibility. While efforts to strengthen fiduciary and administrative systems have been part of reforms in Albania, the Krygyz Republic, Romania and Turkey, these reforms did not directly address the challenges of aligning the systems to a more decentralized setting. Table 3: Assignment of revenues and responsibilities to subnational governments ROMANIA TURKEY KYRGYZ ALBANIA EDUCATION Assignment of responsibility Extensive None/marginal Some- ambiguous Shared- ambiguous …of revenue limited None/marginal Some- ambiguous Shared- ambiguous HEALTH Assignment of responsibility None/marginal None/in transition None/marginal Shared- ambiguous …of revenue None/marginal None/marginal None/marginal Shared- ambiguous WATER Assignment of responsibility Extensive Extensive [no data] Formally extensive, de facto ambiguous …of revenue Shared Shared (de facto) [no data] Extensive Page 12 12 The local-level arrangements for service delivery comprise the third set of relationships relevant to the intermediate route of accountability. Commonly, enthusiasm for local-level service delivery stems from the proximity of citizens to – and hence better information (through direct information) about -- local leaders and providers. Often, though, local social arrangements bear little resemblance to some idealized version of grass-roots democracy. But even in settings where there is no track-record of civic participation (making the risk of elite capture higher) active interventions can help reshape relationships between citizens and local governments (and can also help build technical capacities). Unlike other regions of the Bank which are active in this area, in none of the five sample countries did Bank lending operations focus on building the participatory (and technical) capacity of local governments (though see Box 2 in chapter 4 for some examples of smaller, grant-supported initiatives). The Short Route of Accountability – Client Power In the short route of accountability – depicted in Figure 1 as ‘client power’ – citizens as principals hold front-line providers accountable for performance. WDR04 identifies two distinct variants of client power – choice and participation. Chapter 5 provides details on both variants; the focus in this summary is on participation. Within the category ‘participation’, two distinct subvariants are considered: client power within individual sectors, and multisectoral, community-driven client power. Client power through sectoral participation engages service users in decision- making -- giving parents associations oversight responsibility over how annual school budgets are spent for example, or involving farmers in operating irrigation systems via water users associations. As WDR04 highlights, participatory client power is in part a temporary substitute for weak government capability, providing a framework for getting resources to the front-line, with pressure for performance, in settings where politics is otherwise strongly clientelistic. But it also can complement other accountability mechanisms. Even where the predominant (central or subnational) service delivery arrangements are compact-based, institutional space can be made for parents associations, water users associations and the like to participate, strengthening accountability and contributing to improved performance. Table 4: ‘Client power’ through participation: 14 country-sector observations Romania Turkey Kyrgyz Albania Tajikistan EDUCATION extensive marginal Some* marginal extensive HEALTH marginal marginal marginal marginal Some? Page 13 13 URBAN WATER marginal marginal n/a some some Table 4 summarizes the extent to which participatory client power has been supported across the 14 sector-country observations; chapter 5 provides further detail. Overall, participatory mechanisms are used more intensively in education than in the other sectors; and are more in evidence in Tajikistan than in the other countries. Community empowerment through social investment funds (SIFs, with the funds generally provided by donors) comprise an increasingly common multisectoral variant of ‘participatory client power’ worldwide, including in ECA. In SIFs, communities work together with an agency which operates at arms-length from government to prioritize among a multisectoral array of local public investments, and participate in their implementation and subsequent maintenance. As chapter 5 details, in Albania, Tajikistan, Romania and the Kyrgyz Republic SIFs have become a major provider of resources for local public investments, especially in rural areas. SIFs have been an uncomfortable fit within the Bank’s country programs. Their multisectoral, community-level mechanism of prioritization and decision-making seemingly conflicts with sector-specific approaches (because, for example, communities may seek investments which are not in a sectoral master plan – or because sectorally earmarked investments may be inconsistent with community priorities). And the creation of a parallel agency responsible for local public investment seemingly is at variance with efforts to strengthen the role, resources and responsibilities of elected local governments. In practice, as chapter 5 details, experience in at least three of the country cases suggests that these seeming conflicts can be resolved in implementation – an example of complementarity between the short and intermediate routes of accountability. Indeed, the examples point to the possibility that SIFs can play a key bridging role -- as a transitional mechanism which, not only supports poverty-reducing investment in the short-term, but also helps strengthen local participation and build local capacity more broadly, exiting from the stage as local governments mature. Comparative Empirical Assessment. As noted earlier, it is neither expected, nor necessarily desirable that all interventions in any country are described as ‘extensive’ for all sectors in all of tables 1-4. Rather, the relevant question is whether, considering the tables together, any systematic approach to help strengthen accountability for service delivery is evident in country programs (with the relative emphases on the long, intermediate and short routes depending on country specific political and institutional realities). Table 5 groups the fourteen country-sector observations across two dimensions. The first dimension is whether the country approach has been reformist, reformist with constraints (including the ‘constraint’ of a reform process which currently is at an early stage) or non-reformist, with some other agenda dominating (the last might include, inter alia, maintaining the status quo; assuring discretion and related rent-seeking; or satisfying some key interest group). The second dimension is whether the Bank approach exhibits a clear application Page 14 14 of one or another of the accountability strategies highlighted in WDR04, a somewhat clear application, or whether no clear approach is evident. Viewed from the country perspective, the accountability options are clearest – and consistently have been supported as such by the Bank in country strategies – at the extremes. Romania emerges as the country which most consistently has been ready to embark on (across all three sectors, though with uneven implementation) far-reaching ‘compact-style’ reforms, and has consistently received Bank support for these efforts. At the other extreme, Tajikistan has presented the least propitious environment for systemic reforms – and the Bank’s strategy for assistance has been most consistent in support of the ‘short route’ of participatory client power in its interventions. In the remaining three countries, there is less evidence of a systematic approach to accountability. In Turkey, the pattern may perhaps be described as an ‘incomplete compact’ (with even this something of an overstatement in the education sector, where authorities have so far been unwilling to cede any flexibility at all to the school-level). In the Kyrgyz Republic, the approaches are quite different for health, with its strong ‘compact-style’ reforms, and for education, where the government has shown little appetite for reform, and the Bank is adopting a nascent participatory client power approach. Table 5: Country and Bank Approaches to sectoral accountability COUNTRY APPROACH Other agenda Reformist, with constraints Reformist No clear approach Albania health Somewhat clear approach Albania education Tajik health Tajik water Turkey education Romania health B A A P N P K R O A C H Clear approach Tajik education Kyrgyz education Romania water Turkey water Turkey health Albania water Romania education Kyrgyz health Albania emerges as an outlier. While Albania’s institutional, technical, and participatory capacities do not seem notably weaker than, say, the Kyrgyz Republic’s, the Albanian government has demonstrated at best modest commitment to credible sectoral reforms. Yet the detailed analyses in chapters 2 and 4 point to both political openness Page 15 15 (voice and accountability) and a potentially far-reaching commitment to de-centralization as assets with which an accountability-oriented country strategy might work. However there appears to have been no consistent institutional response by the Bank which, on the one hand, recognizes the limitations in commitment to reform, and, on the other, builds on the assets of openness and political decentralization in ways which can help address Albania’s accountability challenges on the ground. Turning to the sectoral perspective, the cross-country empirical patterns and the specific examples included in the later chapters reveal a series of distinctive sectoral strengths and weaknesses in engaging with the themes highlighted in the WDR: · The education sector has clear models of support both for strongly reformist countries (Romania) and for countries where systemic reform is not on the agenda (Tajikistan; Kyrgyz Republic). But it is less clear what was the vision for reform for countries that fall between the extremes (Albania, Turkey). · The health sector propagates and supports a far-reaching compact- and market- oriented model of structural change, one which has the potential to achieve major gains in accountability (Kyrgyz Republic, Turkey). But the sector seems to lack an accountability-oriented incremental agenda of reform in countries where the appetite for this model is weak (Albania, Tajikistan), or where the process of implementation is sufficiently ‘soft’ as to raise the question as to whether the systemic reforms have been worth the effort (Romania). · The urban water sector supports a compact-oriented approach to reform which seems able to be adapted readily both to institutionally weaker, slow-reforming countries (Tajikistan) to those that are stronger and more reform-oriented (Romania, Turkey) and to those in between. The dilemma, though, is that the sector’s approach seems tilted towards strengthening the arms-length flexibility of front-line providers, with relatively little attention to institutionalized mechanisms (beyond transaction-specific contracts) for strengthening the accountability side of the compact. Operational Implications Two sets of operational implications are drawn from the empirical analysis summarized above. The first set of implications focuses directly -- within the confines of the three routes of accountability -- on how ECAVP might better address the governance of service delivery. The second set of implications concerns the interactions between the governance of service delivery, and good governance more broadly. Improving Governance – the Service Delivery Lens The empirical analysis point to significant limitations in ECAVP’s approach to service delivery. As Table 5 highlighted, in six of the fourteen country-sector Page 16 16 observations there is no consistent, clear approach to sectoral accountability. But even for the eight observations where a clear approach is evident, perusal of Tables 1-4 suggests that implementation is very uneven, undercutting effectiveness. 9 Heightened, systematic attention (by both country and sector units within ECAVP) in six areas could go a long way towards addressing both levels of limitation. Each of the six is considered in turn. Better aligning sectoral accountability strategy with country settings. The WDR04 framework has the potential to enhance how the Bank’s country assistance strategies support improved service delivery in two important ways: it directs attention to the accountability underpinnings of the arrangements for service delivery; and it points to ways in which the arrangements might be aligned with the country’s political and institutional realities. The empirical assessment in this paper suggests that a more systematic approach to the governance of service delivery could yield gains in development effectiveness in all five selected countries (possibly major in some cases; more modest, but nonetheless significant in others). Three sets of questions could help guide an intensified effort in this area: · First, given the country’s political and institutional realities, should primary (not exclusive) emphasis be given to the long, short or intermediate route as the principal focus of efforts improving the governance of service delivery? · Second, in those sectors which lend themselves readily to service delivery via the route identified as primary, how systematically are the institutional arrangements needed for that route to be effective being put in place? (For example: if the long route is primary, are initiatives to strengthen ‘compacts’ adequately complemented by efforts to strengthen voice?) · Third, what would be a useful role for other routes as complements and/or secondary options alongside the preferred primary route? Addressing these questions systematically at the level of the Country Assistance Strategy could provide an important filter for prioritizing among diffferent operational approaches aimed at strengthening accountability compacts, serdel voice, the intermediate route of de-centralization, and the short route of participatory client power. Doing better on accountability compacts. As explored earlier, accountability compacts (one of two links in the ‘long route’) fit better in institutionally and technically more capable countries. In countries where the compact emerges as (part of) the primary 9 A (foot)noteworthy pattern evident in Volume 2 is a seeming inverse correlation between the level of support for physical investment, and the level of attention to institutional reform. Many of the projects reviewed for this study incorporated support for both physical investment and for institutional change. However, project after project, especially where physical investment was on a large scale, and required careful requisite attention to complex procurement issues, in the course of implementation there seemed to have been a progressive loss of focus on the intended institutional reforms. Page 17 17 route of accountability, three further sets of questions could help guide an intensified effort: · First, are front-line providers being afforded adequate flexibility to decide how organizational resources are being deployed? (Among the sectors considered in this paper, this seems to be an issue in the education sector in particular.) · Second, are efforts to improve flexibility adequately complemented by efforts to strengthen the mechanisms of M&E and regulation needed to assure accountability? (Among the sectors considered in this paper, this seems to be an issue in both the health and urban water sectors.) The third set of questions relevant to accountability compacts concerns the adequacy of attention to voice, serdel voice in particular. Within the long route serdel voice is a necessary complement to a compact. But its potential role is substantially broader . Doing better on serdel voice. Table 4 suggests that ‘serdel voice’ has received only modest attention in ECA. Yet transparent, readily available information on the performance of front-line service providers (‘serdel voice’) has the potential to support all three of the long, intermediate and short routes of accountability: · As noted above, it complements the ‘compact’ as one of two links in the ‘long route’ of accountability. · As discussed further in chapter 4, serdel voice is at least as important for de- centralized service delivery as for delivery via central government. · The effectiveness of the ‘short route’ also could be enhanced to the extent that explicit information enables clients to judge the more technically complex dimensions of performance, and to make comparisons as to relative performance across different agencies. Serdel voice thus has the potential to have a broad impact on the quality of service delivery. A strong recommendation follows: Each country program where the Bank is active in supporting efforts to strengthen governance and improve service delivery could usefully include a specific initiative aimed at helping produce and disseminate, on a sustainable, multi-year basis the information needed to give effect to serdel voice. In assessing whether a serdel voice strategy adequate to the task is in place, the following questions might usefully be asked: · For each service delivery sector, is a mechanism in place for making available to citizens/clients information on the performance of front-line facilities? (The mechanism could be sector-specific, though there appear to be strong economies and scope from organizing multisectorally). Page 18 18 · Is this mechanism linked to a government M&E system for monitoring results? (A strong M&E system has the potential to facilitate serdel voice – so long, that is, that performance data are made public.) · Is the mechanism institutionally credible and sustainable over time? (Is leadership in the hands of civil society organizations, perhaps in partnership with government and the private sector? Is funding for the exercise assured over multiple cycles of information collection and dissemination? Doing better on the intermediate route of de-centralization. In highlighting accountability as key to improved service delivery, WDR04 also brings de-centralization to center stage. The empirical analysis summarizede above (and detailed in chapter 5) points to both the empirical relevance of de-centralization, and the challenges which it entails: · In two countries -- Albania and Romania -- governments have given this intermediate route an important role in their approaches to service delivery since the mid-1990s. In both countries, though, the disconnects in assignments of roles and resources between central and local levels are weakening the efficacy of service delivery. · In a further two countries (Kyrgyz Republic and Turkey) de-centralization has become increasingly central to the policy agenda over the past two years. Both governments have committed to far-reaching changes in the intergovernmental system, without detailing what these might involve. The potential either for value adding change or for disillusion (whether through inaction or actions which turn out to be ill-advised) is large. · In the fifth (Tajikistan) modest de-centralization is emerging as one of the few promising feasible options in the short term for strengthening relationships of mutual accountability between the state and society. The empirical work also points to some challenges posed by de-centralization to sectoral approaches: · urban water already has moved far down the de-centralizing path almost everywhere, but major challenges remain in assuring, not only autonomy, but also accountability for front-line providers; · in the education sector, there is a disconnect between the de-centralizing commitments of countries (e..g. Albania, Romania) and the Bank’s seeming discomfort with aligning the sectoral dialogue to these realities. · in the health sector, more attention might usefully be given to how de- centralization might be leveraged to enhance accountability in setting where implementation of the insurance-driven compact model does not seem feasible. Page 19 19 Yet, for three reasons, de-centralization is an uncomfortable topic for Bank operational work. First, the process is troublingly complex. It involves a combination of changes in rules, the development of new capacities at both central and local level, and a transformation in the relationships and expectations between citizens and local leaders. Progress is measured in decades, not in years. Second, the process is troublingly political, its rhythms dictated more by the dictates of political strategies and tactics than by an orderly technocratic logic. Third, the process is troubling multisectoral, an uncomfortable fit with our sectorally stovepiped structures which encourage most units to view de- centralization as ‘some other group’s business’. For all three reasons, the inclination has been to try and keep decentralization on the margins of an already overloaded agenda. Given both its empirical salience and its potential as an accountability route, the downplaying of de-centralization is misplaced. Once the intermediate route has come strongly onto a country’s agenda, added value in dialogue and advice comes not from opposing the process (either explicitly, or implicitly through disingenuous admonitions to ‘slow down’), but rather from exploring how the challenge of change can most effectively be orchestrated and sequenced. The recommendation follows: · In countries where de-centralization is on the national agenda, engage it seriously as a cross-cutting mechanism for strengthening accountability -- at both the sectoral and cross-sectoral levels. This recommendation is not intended to imply that the Bank should become an unqualified advocate for de-centralization, nor to suggest that whatever de-centralizing agenda is proposed should be supported uncritically. Rather, it calls for a serious engagement across all sectors with the merits of the various proposals – against the backdrop of a recognition of the merits of effective de-centralization, IF it can proceed in an appropriately sequenced way. Where the intermediate route emerges as an important part of a country’s accountability system, success in service delivery will depend not only on appropriate assignments of roles and resources across central and local levels, but also on the ability of local governments to carry out their responsibilities effectively. A second recommendation follows: · Assess whether adequate support has been provided to build capacity at the local level – both the technical capacities of local governments, and the capacities of local citizens to participate effectively in local governance (including the capacity of local government to engage effectively with citizens). If not, consider how best the Bank might provide operational support for such efforts, including linking capacity development with the provision of local investment resources. Doing better on participatory client power. The data in Table 4 point to a striking disconnect between the potential role of participatory client power – as both a substitute Page 20 20 accountability mechanisms in clientelist settings, and as a possible complement in settings with more developmentally oriented states – and its limited utilization in the country-sector observations. At the level of country strategy, an illustrative example is Albania, where the lack of sustained attention to this facet of institutional reform (at least in the sectors under study) plausibly represents a missed opportunity given the limited traction of other approaches. At the level of sector strategy, the possibility arises that participatory client power could add value as an ‘incremental’ reform strategy in the health sector in settings where more systemic reforms are not yet on the agenda. Yet the empirical review poionts to the virtual absence in the health sector of what seemed prima facie to be some relatively straightforward opportunities for strengthening ‘client power’ (for example, more participation of local citizens in the governance of local health clinics; more pro-active recruitment of para-professionals from within the service areas of health centers). These and other examples suggest that the returns could be high from more systematic attention to the opportunities for intensifying the use of participatory client power in country and sectoral interventions. Key questions which could guide such an effort might include: · In country settings with weak, clientelistic governments, what additional (sector- specific and multisectoral) opportunities might there be for crowding in participatory client power more directly into the governance of service delivery? · In settings where governments are stronger and more pro-poor, what opportunities might there be for participatory client power to complement long or intermediate routes of accountability? · What opportunities might there be for using participatory client power as a mechanism for strengthening downward accountability in both the long and, especially intermediate routes? The last question points, of course, to the potential of multisectoral community-linked client power as a way of helping build both technical and participatory capacity for local self governments over the medium-term, while providing services to neglected communities in the short-term. Focusing attention on how these synergies can be realized is surely more productive than mutual sniping between protagonists of community approaches, and protagonists of an institutionally coherent role for local self governments. Doing better multisectorally. One final point warrants emphasis as to how attention to the demand-side of accountability might be intensified within ECA. Implementing the suggestions above on serdel voice, on the intermediate route, and on client power often can be done best via approaches which cut across individual sectors: Page 21 21 · A serious, sequenced approach to de-centralization cannot simply be assigned to ‘de-centralization specialists’, but requires working together with specialist staff from the affected sectors on the basis of a common approach. · While some efforts to foster client power may be sector-specific, those that aim to empower communities to prioritize and implement are multisectoral – and hence require collaboration · The returns can be very high from combining in a single institutional arrangement service delivery performance scorecards which bring into the public domain data from a variety of service sectors. The multisectoral character of the task suggests that sustained impact will be achieved Improving governance is not simply a task for a single sector. Improving the governance of service delivery is not simply a specialized sector-by-sector task. They are everybody’s business, and – this is the final recommendation of this subsection -- can be addressed most effectively with multisectoral products, supported by multisectoral teams. From Service Delivery to ‘Good Governance’ More Broadly The discussion to this point has taken the environment external to the service delivery system as given, and asked the question: “ given the context, which route of service delivery accountability would be more effective in a specific country setting? ” Yet the WDR04 accountability framework is itself part of a larger, interdependent system of governance and public management. Over the course of the past decade, broad consensus has emerged as to the characteristics of governance and public management systems capable of underpinning effective public action – in service delivery, and other public functions. Many countries initiated ‘top-down’ public sector reform programs which aimed to put in place these ‘best practice’ system. It has become increasingly evident, however, that the process of building state capacity is a dynamic one; knowing the desirable end point is only moderately helpful in discovering the path from institutional weakness to progressively stronger capability. In recognition of the dynamic character of the process, increasingly the focus of public sector reform is on initiatives which aim to alter the incentives of political leaders. The intent is to induce them to respond to a broad array of civic pressures for performance – not simply to the elites who benefited from the pre-existing status quo. The change process is understood to be cumulative, directing attention towards the identification of ‘entry points’ which, though modest in themselves, have the potential to catalyze subsequent rounds of change. Against the backdrop of this broader conception of the reform process, the question arises: “To what extent can an initiative which aims to Page 22 22 strengthen the accountability framework for service delivery serve as an entry point for changes in governance and public management more broadl y?” In ways which parallel the earlier discussion of the preferred route of accountability for service delivery, the answer to this last question depends importantly on a country’s starting point. Consider first the extremes. At one end of the spectrum are countries whose states fit the WDR04 definition of ‘pro-poor’, and whose starting points for further enhancement of state capacity are strong. Political leadership is in place with a clear developmentally-oriented vision, and with a mandate (viable over the medium- term) for acting on that vision. There exists a workable baseline of bureaucratic capabilities. In such settings, the opportunity exists for direct reform of governance and public management systems. None of the five countries which are the focus of this paper meet these criteria. At the other extreme are countries with ‘weak’ states, stark examples of the WDR04 definition of ‘clientelistic’. In such settings (among the five countries, Tajikistan is the one which fits best this ‘weak state’ category) bottom-up pressure of the kind which might be generated through efforts to strengthen accountability for service delivery is unlikely to be sufficient to effect improvements in governance more broadly without a concomitant reformist impulse from elsewhere. The more interesting opportunities for cumulative change via a service delivery entry point are in countries which fall between the two ends of the spectrum. There are (at least) two relevant sub-types. The first comprises countries with longstanding mixed economies and pluralistic polities – their decision-making processes too ‘soft’ to spontaneously engage in robust efforts to strengthen pressure for performance, but sufficiently open to be responsive to pressure from below. Turkey fits straightforwardly into this category. The second subtype is of special relevance to post-communist countries. The 2002 World Bank publication, Transition: the First Ten Years describes how some transition countries found themselves becalmed in a partial reform quasi- equilibrium.: “In many countries in the CIS and Southeastern Europe…the collapse of communism was rooted in a contest among competing elites rather than in any broad social movement. The new political arrangements in these ‘concentrated political regimes’ were designed by incumbent leaders, often as a way to consolidate their power ……. Oligarchs and insiders prefer a partial reform, and sustain the reform process through the point where their gains are maximized, and beyond which further implementation of policies of discipline and encouragement threaten to undermine gains from rent-seeking and ‘tunneling’…Although the gains from such distortions tend to be highly concentrated, the losses are dispersed…countervailing pressures from competing groups were weak and the disaffection and apathy of the ‘losers’ minimized the direct costs to politicians of poor policy choices.” 10 In countries where the dominance of insiders and oligarchs is robust, the prospects of breaking this dysfunctional equilibrium are bleak, at least in the short-term. However, in 10 World Bank (2002) pp. xxiii-xxiv, p. 112. Page 23 23 others – and both Albania and the Krygyz Republic arguably fit this latter category -- politics remains more fluid. There, the prospects seem higher of effecting change via the political strategy outlined in the Transition study, namely “….the mobilization of the losers in the existing low-level equilibrium through greater political inclusion”. 11 Approaches to effect change highlighted by the Transition study include mobilizing the electorate, mobilizing excluded enterprises, ensuring (and using) a free media, and transforming hidden into explicit subsidies. A focus on accountability for the quality of public service delivery, though not considered in that study, is another example of a politically inclusive strategy. In this intermediate group of countries, the service delivery focus can be an entry point for a larger agenda of governance and public management reform. The reason is that a focus on service delivery also is likely to surface a series of cross-cutting constraints on performance which result from the structure of the public management system more broadly: · An effort to hold front-line providers accountable may be undercut by system- wide failures to monitor transparently how service providers and other officials use public resources, and system-wide incentives to blur lines of accountability. Non-transparency and a lack of clear lines of accountability are classic ways of maximizing the discretionary authority of senior political and bureaucratic actors. · Front-line providers may be constrained by a lack of fiscal resources, perhaps because they were never allocated in the first place (which may point to broader weaknesses in national or local budget formulation systems) or perhaps because – though formally allocated – in practice the relevant resources were redirected before they reached the front-line. · The effort to empower front-line providers may be blocked by rigidities in cross- cutting personnel and financial control systems, preventing well-performing providers from being rewarded, poorly performing providers from being sanctioned, and in general making it frustratingly difficult to get the green light to do anything worthwhile. While the intent of these controls may originally have been to assure fairness and transparency, the world over they have a tendency to ossify into obstacles to getting things done, with their rationale and benefit becoming increasingly obscure. As was noted earlier, the track record of stand-alone ‘top-down’ efforts to improve fiscal, financial management, personnel control and other cross-cutting public management systems is decidedly mixed. As was also noted, an important reason for the uneven results is that they are not underpinned by a change in the incentives of political and bureaucratic leaders. They are ‘upstream’, far removed from the service delivery front-line, and hence without well-defined beneficiaries ready to pressure for results. Heightened accountability for service delivery has the potential to focus the attention of 11 World Bank (2002) p. 112. Page 24 24 front-line providers not only on those aspects of service delivery which are under their control, but also on broader systemic constraints which block their ability to perform. Part of the increased pressure felt at the service delivery front-line may thereby be transferred ‘upstream’, spreading the impulse and pressure for reform across the public management system more broadly. Indeed, a wide variety of quite broad governance and public management reforms may be possible without confronting up-front the core levers of insider and oligarch power. To cite one example, the incentives for state capture – and hence resistance to reforms which increase accountability – are greatest in settings where rent opportunities are most concentrated. Expenditures in many local public services (notably including education and health) often are diffused among a large number of smaller transactions, suggesting that oligarchs may not feel the need to vigorously oppose reforms in these areas. To cite a second example, oligarchs may care more about political control at the national than local levels – suggesting they may be willing to countenance greater political and executive accountability at subnational levels. Incremental reform via a service delivery entry point may thus go quite far in reshaping the environment for governance and public management before it becomes confrontational, and zero-sum. Getting the (Im)balance Right The operational agenda outlined in the previous two sections lends itself easily to misinterpretation – misinterpretations which can become ‘straw men’ on the basis of which the approach is set aside. A first misinterpretation would be to draw the implication that since, in most countries, there are inadequate incentives for supply-side public management reforms to be successful, the focus should be narrowly on strengthening accountability for service delivery. Viewed this way opens up the approach to the trenchant criticism that strengthened accountability without strengthened capacity runs a high risk of backfiring. An alternative misinterpretation would posit that, since the agenda is an interdependent one, reforms should push simultaneously on all fronts. Viewed this way, the analysis also is flawed: it cannot reckon with the weaknesses in capacity that both motivate reform, and limit what can be done. What is proposed rather is a strategy of “nimble creative imbalance”. C reative imbalance is key as a way of building momentum for governance reform by embracing entry points for reform which ratchet up tension in the system. Pressure for improved service delivery seems an especially attractive example. Nimbleness is key to assure that the binding supply-side public management constraints are addressed before they short- circuit the entire process of change. For the latter, it is important that theh public management agenda not be defined in so comprehensive a way as to render it unworkable, or incapable of adapting to signals as to what are the emerging ‘choke points’ of change. All of this may seem frustratingly circuitous, perhaps even fanciful – with the broader benefits dependent on an elusive combination of policy entrepreneurs within the broader bureaucracy ready to respond to the emergent signals from service providers, Page 25 25 plus the implicit threat of electoral consequences for local and national political leaders. But consider the example of Bangalore in the Indian state of Karnataka. Frustrated by years of inaction in the face of public services which increasingly were unable to keep up with the city’s economic dynamism and population pressure, in 1994 a group of citizens introduced the idea of a user survey based ‘report card’ on Bangalore’s public services. Initially, the impact was modest. As the pioneer of the initiative put it: “It is unrealistic to expect public agencies to respond immediately and directly to the signals given by a report card. Agency leaders need the time and capacity to internalize the messages of the report card and design interventions to address the issues raised. Civil society institutions also need time and resources to get organized and plan strategies to interact with service providers.” 12 Nonetheless, the sponsors persisted, establishing a non-governmental organization, the Bangalore Public Affairs Center to institutionalize the effort, building coalitions with other NGOs, and repeating the report card survey in 1999, and again in 2003. As Figure 2 overleaf highlights, over time the turnaround has been extraordinary in perceptions of the quality of service delivery. How were these achieved? The (almost) last word goes to the Public Affairs Center: ” The first and second report cards had put the city’s public agencies under the scanner. The adverse publicity they received, according to many observers, acted as a trigger for corrective action. Inter-agency comparisons seem to have acted as a proxy for competition. Citizen activism and dialogues with the agencies also increased during this early period. These developments prepared the ground for a positive response from the Government. A good example is the political leadership and vision displayed by the Chief Minister S.M.Krishna in the past four years. He provided the framework within which a set of able administrators could set in motion a series of actions and reforms in the agencies. Many civil society groups and the media have stimulated and supported this momentum. Sustaining this movement is the challenge for the future.” The questions arise. Could similar cumulative processes unfold in ECA countries? Take the examples of Albania and the Kyrgyz Republic, in both of which the quality of governance appears to have gone into reverse in recent years. What would have been (or might still be) the impact of more systematic attention to strengthening the intermediate route of accountability by clarifying roles and responsibilities between central and local actors? of more systematic attention to an information strategy which strengthens service- delivery-linked voice? and of a more systematic effort to directly strengthen client power by more systematically supporting participatory governance arrangements at the service delivery front-line? Might a critical mass of such interventions help provide much needed momentum to more far-reaching reforms of governance and public management? A key contribution of WDR04 is its implicit invitation to countries and their development 12 Samuel Paul, Holding the State to Account: Citizen Monitoring in Action (Bangalore: Books for Change, 2002) p. 71 Page 26 26 partners not only to act to strengthen service delivery in the short- and medium-term but, more broadly to test the opportunity suggested by these questions. Page 27 27 Figure 2: Perceptions of service delivery performance in nine Bangalore agencies, 1994-2003 Legend: BMP = Bangalore Municipal Corporation; BESCOM = electricity; BWSSB = water supply; BSNL = Telecom Department; BDA=Land Development Authority; BMTC = Metropolitan Transport Corporation; RTO = Motor Vehicle Licensing Overall Satisfaction across Three Report Cards General Households 5 6 4 9 25 0 1 0 14 41 47 42 67 34 34 16 32 32 73 94 73 92 73 78 85 96 77 0 20 40 60 80 100 B M P B E S C O M B W S S B B S N L G O V H O S P I T A L S P O L I C E B D A B M T C R T O agencies % s a t i s f i e d 1994 1999 2003 n/a n/a Page 28 28 Chapter 2: The Analytical Framework Issues and Approach An analysis of the governance of service delivery in ECA countries potentially is extraordinarily broad. To narrow the terrain, four sets of decisions were taken up front. The first decision was to focus narrowly on the accountability arrangements for service delivery. WDR04 motivates a focus on accountability as follows: “ [The aim is] to improve services by changing the relationships of accountability. [The proposals] recognize that, though there are many proximate causes of service delivery failure, the deep causes lie with inadequate institutional arrangements. If frontline workers (civil servants) in the existing organizations of public production are frequently absent, have little regard for clients who are poor, and lack the technical knowledge to perform their services well, this inadequate organizational capacity is the proximate cause of poor services. Too frequently those seeking improvement have focused only on internal organizational reforms – focusing on management of the frontline workers. If organizational failures are the result of deeper weaknesses in institutional arrangements (weak political commitment, unclear objectives, no enforceability), direct attacks on the proximate determinants (more money, better training, more internal information) will fail.” 13 Note that, looking backwards, weaknesses in accountability most likely are not the dominant proximate explanation for the precipitous decline in the provision of many public services across ECA countries -- the loss of resources for public services that came with the collapse of communism mattered more. But as WDR04 shows, there is only a loose correlation between the provision of resources and the effectiveness of public service delivery. So from a forward-looking perspective, the central challenge indeed seems to be the one highlighted by WDR04: not pumping resources back into a system with is poorly aligned with the requisites of mixed economies and pluralistic polities, but rather putting in place institutional arrangements capable of using efficiently and effectively those flows of public resources that become available as a result of renewed economic expansion. The second decision was to presume as a starting point that accountability mechanisms of the kind detailed in WDR04 are important to improved service delivery. Note that the institutional perspective adopted by WDR04 is a distinctive one. WDR04 argues strongly that the conditions are very stringent under which bureaucratic provision of services through central line ministries is likely to be effective: 13 WDR04, p. 58 Page 29 29 “ For public servi ce production to be reliable and effective…the policymaker must care about outcomes, including those for poor people. That concern needs to be transmitted effectively to the public agencies that receive public resources to provide the resources. And the public agencies must hire technically qualified providers motivated to provide the service.” 14 As WDR04 details, in most developing and transition countries one or more of these conditions are likely to be absent for public production. The WDR describes a variety of accountability mechanisms which can respond effectively to weaknesses in these areas, and synthesizes the voluminous evidence which has been accumulated worldwide as to the effectiveness of these mechanisms. This study does not try to test the hypothesis that strengthening these accountability relationships will improve service delivery in the ECA context. Rather, taking the causal relationship as given, it probes the extent to which these accountability relationships are part of the institutional arrangements within specific country-sector observations. The third decision concerns the selection of countries and sectors. A central theme of WDR04 is that “no single [accountability] solution fits all services in all countries.” 15 Assessing the fit between country and sectoral conditions on the one hand, and the preferred approach to accountability on the other requires a quite in-depth assessment both of country conditions and of the country- and sector-specific service delivery arrangements. This implies that a comparative case study approach is the preferred methodology. The comparative analysis focuses on three sectors in each of five countries. The three sectors are ones chosen by WDR04 for intensive scrutiny – education, health, and urban drinking water. The countries selected for comparative analysis are Albania, the Kyrgyz Republic, Romania, Tajikistan and Turkey. 16 The following criteria guided the country selection: · That there be sufficient variation in their underlying governance characteristics to explore the comparative relevance across the country settings of each of the alternative mechanisms of accountability; · That the selected countries be ones where – as evidenced by operational demand from the relevant country management unit -- there was a desire to explore new approaches to improving governance, and an openness to linking this to sector-specific arrangements for service delivery; and · That the sample incorporate some of the more challenging governance environments within which the Bank is active in ECA, as a further way of testing the operational value added of the WDR04 framework. 14 WDRO4, p. 55 15 WDR04, p.12 16 Over the course of the fiscal year, the principal author visited and worked operationally in all of these countries other than Romania, which was added towards the latter stage of the exercise is a way of diversifying further the governance characteristics of the selected sample. Page 30 30 Within the countries and sectors selected for study, the question arises as to whether the emphasis of analysis and recommendations should be on the country or on the Bank - - on how well the countries are doing in putting in place an accountability framework adequate to govern service delivery, or on how well the Bank is doing in encouraging and supporting efforts in this direction? The answer to this question was dictated in large part by the necessity (in the absence of resources sufficient for in-depth primary analysis across all countries and success) of drawing on secondary material to learn about the details at institutional reform within sectors, and the success (or otherwise) of their implementation. The bulk of this material was drawn from within the World Bank, in the form of its Economic and Sector Work; Project Appraisal Documents; Project Status Reports; Implementation Completion Reports and Project Performance Audit Reports (the latter prepared by the Operations Evaluation Department). This material is used, to the maximum extent possible, to describe how the accountability relationships actually operate on the ground in the selected country-sector observations; but the exercise has obvious limitations. Hence the fourth decision as to how to narrow the terrain of analysis: focus the policy-oriented discussion principally on an assessment of World Bank efforts to help strengthen the governance of service delivery, given the country- specific realities on the ground. The Accountability Framework This section outlines the analytic framework used in WDR04, and explores its relevance across the five country-specific and three sector-specific governance environments in which it is to be applied. The accountability relationships. Figure 1 (repeated from Chapter 1) reproduces the WDR04 ‘accountability triangle’ which depicts the key principal-agent accountability relationships which link service providers, citizens/clients, and politicians/policymakers. 17 The WDR highlights three key sets of principal-agent accountability relationships. The first two are those between citizens/clients and politicians/policymakers ( voice ), and between politicians/policymakers and service providers ( compact ). Taken together these first two relationships comprise what the WDR terms the · long route of accountability in which citizens hold politicians accountable (principally, though not only, through voting), and politicians/policymakers in turn hold providers accountable through ‘compacts’. As WDR04 clarifies, the voice-compact nexus of accountability relationships can be evident both at national and subnational levels. This regional study thus defines an 17 As the WDR details, each principal-agent relationship comprises five distinct facets. Principals delegate responsibilities to their agents; finance them (directly or indirectly) to carry out these functions; and enforce (including monitor and evaluate) compliance with the agreed delegation. Agents perform on the basis of agreements; and inform their principals on progress in implementation. Page 31 31 · intermediate route of accountability in which some responsibilities and resources are de-centralized to subnational governments (which, in turn, institute a variant of the long route at the subnational level). The third accountability relationship is that between citizens/clients and service providers ( client power ). The WDR refers to this as the · short route of accountability which, as discussed further below, can be in part a substitute for, and in part a complement to, the intermediate and long routes of accountability. Fitting the framework to country realities. How (holding sectoral characteristics constant) do a country’s political and institutional starting points influence which would be the preferred route of accountability? WDR04’s key distinction is between ‘pro-poor’ and ‘clientelistic settings: “Pro-poor settings are those in which politicians face strong incentives to address the general interest. Clientelist political environments are those in which, even though the average citizen is poor, politicians have strong incentives to shift public spending to cater to special interests, to core supporters, or to ‘swing’ · intermediate route of accountability in which some responsibilities and resources are de- centralized to subnational governments (which, in turn, institute a variant of the long route at the subnational level). Page 32 32 voters…….the payoffs in service delivery for assessing whether the environment is pro-poor or clientelistic can be high. Even if the politics are clientelist, policy choices can be made that are likely to yield better results than the misguided application of policies that work well only in pro-poor environments….” 18 An emphasis on client power – the short route – is identified in WDR04 as especially relevant in clientelistic settings, where governments lack the capacity and incentives to facilitate a compact, and where citizens are unable to translate their voice into effective pressure for better public sector performance. The intermediate route seems especially relevant in settings which are pro-poor and relatively de-centralized, and the long route is preferred in settings which are pro-poor, but not de-centralized. The distinction between pro-poor versus clientelistic settings is more an heuristic device than a depiction of a complex reality. Most countries cannot be pigeonholed neatly into one category or the other, but rather fall somewhere between the extremes of a clientelist/ pro-poor spectrum. Moreover, governance is a multi-dimensional phenomenon, not one that can be captured along a single continuum. Key influences on how a country is governed (i.e. how power is exercised in the management of a country’s economic and social resources’ 19 ) include its inherited institutional legacy, the distribution of power across social groups, and the pattern of political competition. Each of these is rooted in a country’s unique history. This cannot be captured (let alone contrasted across five countries) in a few short paragraphs. For present purposes, what must suffice as a basis for comparison (both between countries and over time) are four composite indicators, each of which is intended to capture a different dimension of governance: 20 · Government effectiveness , which combines indicators of the quality of public service provision, the quality of the bureaucracy, the competence of civil servants, the independence of the civil service from political pressure, and the credibility of the government’s commitment to policies. · Rule of law which incorporates several indicators which measure the extent to which agents have confidence in, and abide by, the rules of society, including perceptions of the incidence of crime, the effectiveness and predictability of the judiciary, and the enforceability of contracts. · Voice and accountability, which incorporates indicators which measure aspects of the political process, civil liberties and political rights, including measures of the extent to which citizens are able to participate in the selection of governments; and measures of the independence of the media. 18 WDR04, pp, 80, 90 19 World Bank, Governance and Development, 1992 20 For details of these indicators, see the website www.worldbank.org/wbi/governance and Daniel Kaufmann, Aart Kraay, and Massimo Mastruzzi, “Governance Matters III: Governance Indicators for 1996-2002”, World Bank Policy Research Working Paper No. 3106, June 2003. Page 33 33 · Control of corruption which incorporates diverse aspects of corruption, including the frequency of additional payments to get things done, the effects of corruption on the business environment, and grand corruption in the political arena or in state capture in policy formation. Table 1 reports for each indicator the country’s percentile ranking for 1996 and 2002 within a global sample of 199 countries. [Note that since Chapter 1 has been written as a self-contained piece, the tables for chapters 2-5 – which include many repeated from chapter 1 – are numbered sequentially, beginning with Table 1.] Table 1: Governance in five countries, 1996 and 1992 (0=worst; 100 =best) Government effectiveness Rule of Law Voice and accountability Control of corruption Tajikistan 2002 9 7 21 10 1996 3 4 7 3 Kyrgyz 2002 21 22 20 24 1996 35 27 35 25 Albania 2002 38 17 50 23 1996 41 42 41 60 Romania 2002 46 54 61 45 1996 31 44 55 51 Turkey 2002 51 55 36 44 1996 61 58 38 61 Source: website www.worldbank.org/wbi/governance and Daniel Kaufmann, Aart Kraay, and Massimo Mastruzzi, “Governance Matters III: Governance Indicators for 1996-2002”, World Bank Policy Research Working Paper No. 3106, June 2003. Patterns evident from the table which are especially relevant for the present paper include the following: · Tajikistan, which within three years of its independence from a collapsed Soviet Union was engulfed in a brutal civil war which lasted until 1997, has the weakest governance among the five sample countries. As Table 1 summarizes, five years Page 34 34 after the war’s cessation, the country still rated at or below the 10 th percentile globally in government effectiveness, the rule of law, and the degree of corruption. · Romania and Turkey occupy the other end of the spectrum within the five country sample. As of 2002, both countries rated close to the global median in terms of government effectiveness, the rule of law, and the extent of corruption. In Romania’s case (as evidenced by gains since 1996) this rating reflected ongoing, incremental improvements subsequent to the country’s break from communism, improvements spurred in large part by the country’s opportunity for accession to the European Union by 2008. For Turkey it reflected the country’s longstanding (though erratic, as evidenced by the country’s repeated economic and political crises) legacy of bureaucracy and law which dates back even to the pre-Ataturk Ottoman era. Note though the worsening of the corruption score in both countries. In Turkey’s case, this most likely reflects disillusion which accompanied the country’s plunge into economic crisis in 2001 (a disillusion which recent data suggests has been reversed by a combination of electoral change and economic stabilization. In Romania’s case, the data points to the worrying possibility that advances in government effectiveness have not been underpinned by strengthened accountability mechanisms for the public sector, raising the risks of reversal. · Albania’s post-communist trajectory initially seemed (at least in terms of the data) little different from Romania. Indeed, as of 1996 Albania enjoyed the reputation of being a stronger reformer. Unlike Romania, though, it experienced a sharp, systematic reverse after 1997– precipitated by a fraudulent, nationwide pyramid scheme. Interestingly, though, Albania seems to have enjoyed continuing advances in civic voice. · The Kyrgyz Republic, viewed as a promising post-Soviet reformer in the mid- 1990s (with the stated aspiration of becoming the ‘Switzerland of Central Asia’) also appears to have gone backward subsequently in the quality of governance, with no gains between 1996 and 2002 in any dimension, and notable declines in both government effectiveness and the extent of civic voice. Given these patterns, plus information on country-specific patterns of de- centralization (to be presented in more detail in Chapter 5 below), as a first approximation WDR04’s short route of accountability seems a priori to be especially salient for the Kyrgyz Republic and Tajikistan, and also relevant for Albania (though, as discussed below, the country also has underway a quite far-reaching agenda of decentralization). The normative expectations for Romania and Turkey are more mixed. Fitting the framework to sectoral realities. How (holding country characteristics constant) do the characteristics of individual sectors influence which would be the preferred route of accountability? From a normative perspective, the key distinctions highlighted in the WDR are between easy-to-monitor, and difficult-to-monitor services, and between services with homogenous demand, and those where demand is Page 35 35 heterogeneous. Viewed from this normative perspective, direct central government provision is preferred for homogenous, but difficult-to-monitor services. Conversely, compacts between local government and front-line providers are preferred for easy-to- monitor services with heterogeneous demand……with a variety of options in between. Within each sector, where are the opportunities for incorporating each of the different accountability relationships highlighted in Figure 1? Table 2 summarizes the focus of the empirical investigation for each of the three sectors. The following sector-specific features deserve highlighting: Table 2: Sector-specific Accountability Mechanisms within WDR04 Framework EDUCATION HEALTH WATER Parental participation in school governance Mechanisms for engaging communities in primary care provision Civic participation in oversight of water utilities Availability of private physicians’ services Client Power Private schools Expanding primary care network and funding mechanism in ways which expand choice to citizens Degree of decision- making autonomy delegated to school level Arms-length relationship between municipality and front-line provider (including private participation) Quality of strategic policymaking Degree of autonomy delegated to hospitals Extent of arms- length contracting for population- oriented outreach services Framework for regulating prices and service delivery standards Compact Test scores and other mechanisms for M&E of performance of front-line providers Investment in systems for M&E of arms-length providers, and of health system performance more broadly Mechanisms for M&E of performance of front-line providers Page 36 36 Voice Performance scorecards, publication of facility level performance scores, and other mechanisms for informing citizens about sector performance Governance of central-local relationships (roles; resources; regulatory, fiduciary and administrative oversight). Subnational level participatory and technical capacities De- centralization Quality of local electoral voice Page 37 37 · Education potentially lends itself especially well to client power through participation via the role of parents committees. (By contrast, for all the attention given to the mechanism in WDR04, there is no evidence in the 5 countries of any role for vouchers as a means of client power, so this option is not included in Table 2). Systematic performance assessment through student testing emerges as a potentially straightforward approach to monitoring and evaluation – with the results also potentially useful as a tool for voice. · The health sector has a product mix which is extraordinarily heterogeneous, and the sector’s demand-side and financing challenges are unusually complex, making for an unusually broad reform agenda – too broad to be considered comprehensively in the present exercise. As Table 2 summarizes, the analysis explores the role of client power principally in relation to the delivery of front- line primary care services; and explores the relevance of the compact in relation to the autonomy of primary care and hospital providers, and M&E systems, as well as investment in regulatory oversight. 21 · The urban water sector potentially offers a classic example of a compact relationship in the context of a regulated utility. Two sets of relationships are key: (i) that between central government and municipalities; and (ii) that between municipalities and the water utilities themselves. A striking feature of the sector is its ongoing, and continuing, process of experimentation with institutional alternatives to govern these two sets of relationships. With the analytical building blocks in hand, we turn now to a review for each of the long, intermediate, and short routes of accountability of the patterns prevailing across the five countries and three sectors. (Volume 2 of the study presents the more detailed, individual sector-country descriptions of the patterns of accountability on which the summary assessments in the body of the paper are based.) 21 The background analysis -- and the notes in Volume II – also examined the role of the compact in relation to the delivery of population-oriented outreach services; however, in no countries was there any evidence of competitive outsourcing of these services, so the issue is not explored further in this synthesis volume. Page 38 38 Chapter 3: The Long Route of Accountability As highlighted in Figure 1, the ‘long route of accountability’ comprises two interlinked principal-agent relationships – ‘ compact ’ and ‘ voice ’. This section considers each in turn. The Compact The compact can be viewed as WDR04s contribution to the challenge of assuring better performance by central public bureaucracies. The service delivery role assigned to public bureaucracies in the WDR04 is a distinctive and controversial one. Highlighting the poor track record in many countries, the WDR argues strongly that the conditions are very stringent under which bureaucratic provision of services through central line ministries is likely to be effective, and focuses rather on an approach which highlights a clear separation between policymakers and front-line providers as key to effective service delivery. Such a separation, argues the WDR, enables both policymakers and providers to be more effective. It enables policymakers to define clearly what should be done, to set performance targets as a basis for monitoring the performance of frontline providers 22 and, more broadly, to govern the service activity via a regulatory system. And it enables front-line providers to be given the authority to decide how things are to be done, within a framework of accountability for results. It also opens up the prospect for delivery by multiple front-line providers, potentially in a competitive market. Critics of the WDR suggest that its emphasis on the compact simply defines away a central problem of public service delivery. It is precisely because performance cannot be specified precisely that delivery finds itself in the public realm – and that all the messy instruments of public management (internal control systems, fostering an esprit des corps etc) become necessary. Further, even if the problem of specifying performance targets can be solved, putting a ‘compact-style approach in place is organizationally and politically complex, implying that compact mechanisms are a better fit for capable than for weaker states. Table 3 summarizes the role of accountability compacts among the 14 country- sector observations. The table reports the intensity of use of each of three facets of the compact: · an arms-length relationship which gives providers some flexibility to decide how resources are deployed; · a monitoring and evaluation system capable of assessing how well front-line providers are performing; and 22 The question of what kinds of performance targets are appropriate is a subject in itself. For now, it must suffice to note that a compact need not depend solely on service delivery outcome measures, but could also proceed using process and output indicators (school enrollment, or teacher absenteeism, for example). Page 39 39 Table 3: The Accountability ‘Compact’: 14 sector-country observations Romania Turkey Kyrgyz Albania Tajikistan EDUCATION M&E extensive extensive marginal some marginal Arms-length some marginal marginal marginal marginal Regulation Some (?) marginal marginal marginal marginal Compact average a/ 2.3 1.7 1.0 1.3 1.0 HEALTH M&E some Some* some marginal marginal Arms-length extensive Some* some marginal marginal Regulation marginal Some* some marginal marginal Compact average a/ 2.0 2.0 2.0 . 1.0 . 1.0 URBAN WATER M&E some Some* n/a marginal marginal Arms-length extensive Extensive n/a some some Regulation marginal marginal n/a marginal marginal Compact average a/ 2.0 2.0 1.3 1.3 Notes : a/ average calculated on basis of: marginal=1; some=2; extensive=3 * reforms in early stage of implementation imply shift from marginal/some to some/extensive n/a no information available Page 40 40 · a regulatory system which defines the ‘rules of the game’ within which providers are expected to operate. An average intensity-score is reported for each country observation. Consistent with expectations, the most immediately evident pattern is the differential role of compacts in one of the two institutionally strongest countries (Romania), and the weakest (Tajikistan). In the latter country, none of the elements of a compact are in place in any of the sectors. By contrast, in Romania all three sectors have in place significant elements of the compact. Indeed, Romania’s education sector provides the strongest example 23 across the entire sample of how a compact might function: · Schools enjoy quite substantial operational flexibility. They are managed by principals who are elected by teachers, and assisted both by pedagogical councils (comprised of teachers) and the School Management Councils described earlier. Schools are responsible for their own budgets, have the right to raise supplementary funding (including, if they so choose, from user fees), and can select themselves from a menu of approved textbooks. Deconcentrated inspectorates are expected to function as active partners in the process of school improvement, rather than simply instruments of central control. · There has been substantial investment in M&E, including: the establishment of a National Assessment and Examination Service which has drafted standardized assessment instruments for all levels of education; the introduction of standardized national examinations at the conclusion of 8 th and 12 th grade; and participation in standardized international tests. A variety of central-level institutions are in place to oversee the system, under the overall umbrella of the Romanian Ministry of Education and Research, including (in addition to the NAES), a National Curriculum Board a, National Teacher Training Board, a Textbook Evaluation and Distribution Board, a Council for Occupational Standards and Assessment, a National Board for the Development of Rural Education, and the Institute for Education Science. The Romanian and Tajikistan ‘extremes’ aside, the remaining three countries show a strikingly uneven picture as to the extent of compact-oriented reforms. The unevenness is evident across both sectors and countries. Consider first the sectoral patterns: · The education sector appears to have given quite significant attention to monitoring and evaluation in Albania and Turkey. In both countries, the Bank has supported efforts to enhance the use of standardized testing. In Turkey, for 23 Note, though, that -- as discussed further in the context of the intermediate route -- on the fiscal side the re have been some serious shortfalls in the provision of resources to Romania’s education sector. Note also that Romania’s education compact operates within a framework of delegated de-centralization – a variant of the intermediate route of accountability, discussed in the next section. Page 41 41 example, a 1990 World Bank project supported the establishment of an Education Research and Development Department within the Ministry of Education; within a decade, the department had developed and conducted student achievement assessment tests in Science, Mathematics, Social Sciences and Turkish, and had introduced the use of standardized international tests. But neither in Turkey nor Albania were enhancements in M&E accompanied by measures to give more day-to-day operational flexibility to schools. [In the Kyrgyz Republic where, until very recently the World Bank stood aside for the Asian Development Bank, there have been no significant efforts either in M&E or school-level flexibility.] · In the health sector, the Bank has supported a strong, compact-oriented reform program in the Kyrgyz Republic (see Box 1). There have also been repeated events to promote compact-style health reform in Turkey; though earlier efforts were blocked by political opposition, there appears to be renewed momentum, which the Bank is supporting via a 2004 Health Transition Project. But in Albania, neither of the two IDA credits provided by the Bank have supported systemic compact-oriented reform in any significant way, with the focus rather on upgrading existing skills and infrastructure in a system which remains institutionally dysfunctional. · Urban water sector reforms over the past two decades have been pre-occupied with enhancing the operational autonomy of front-line utilities from their political principals – initially via efforts to build institutional buffers to shield parastatal utilities from pressure from municipal or central government political leaders, subsequently via efforts to bring private participation into the sector. Private providers currently operate major urban water systems in all the countries where the Bank has been active – on the basis of management contracts in Albania, Tajikistan and Turkey 24 , and a 25 year concession contract in Romania. But outside of the specific terms of individual contracts, there have been no systematic efforts in any of the countries to invest in M&E systems in the sector. [But see Box 2 in Chapter 4 for a promising initiative in Turkey – not water specific, but with potentially major impact in the sector -- supported by the Bank’s Institutional Development Fund.] The regulatory environment for water utilities remains similarly confused, with no regulators in some countries, regulators with ambiguous mandates in others, and seemingly no systematic underlying framework for water regulatory systems on the basis of which to provide advice. 24 Especially in Turkey, where the process has been underway for longer, these contracts have been fraught with difficulty. In one case it took five years from project effectiveness before a management contract finally was concluded. In another case, after six difficult years, the private contractor withdrew, and legal proceedings currently are underway. Page 42 42 Box 1: The Kyrgyz Republic’s Ambitious ‘compact-style’ Health Sector Reforms The Kyrgyz authorities Manas Plan of fundamental health reform, adopted in 1994, was a bold response to the urgent need to restructure the country’s health care system in the wake of the collapse of the Soviet Union. Sustained efforts at implementation over the subsequent decade offer useful lessons as to the potential of compact-style reforms – and the challenges of assuring their effectiveness in institutionally and politically challenging environments. Consistent with the compact-style approach, the Kyrgyz authorities moved boldly to separate health care providers from purchasers, and to provide patients with some choice among providers: · For primary care, a nationwide network of Family Group Practices (FGPs) has been established, initially via an early phase of pilots supported by USAID and a first World Bank Health Sector Reform Project, and thereafter scaled-up with the support of the Bank’s second Health Sector Reform Project. By 2002, 798 FGPs had been formed, and about 90 percent of the population in targeted regions are enrolled with FGPs. · Patients are free to choose their FGPs, which are paid on a capitation basis by the Mandatory Health Insurance Fund (MHIF), established in the mid-1990s as an independent purchaser of health services. The MHIF is funded through a combination of earmarked direct health taxes, general earmarked social security taxes and the general budget. · As of 2002, 65 pilot hospitals were being paid by the MHIF for in-patient care on a case- based system, rather than receiving off-the-top budget allocations. · A Licensing and Accreditation Center has been established. As of 2000, 81 inpatient facilities, and 401 FGPs had been licensed and accredited. While much has been achieved, it is still too soon to conclude that the Kyrgyz compact-style health sector reform is a success. The major hospitals in the system (in the two largest cities, Bishkek and Osh) continue to be funded off-the-top on the basis of historical costs, locking in a disproportionate share of health budget resources, and leaving the arms-length payment system chronically under-resourced. Nor has the impact on system effectiveness of the purchaser- provider split yet been evaluated. Experience elsewhere (Romania, for example) points to the risk that, the formal structure notwithstanding, the system could become a de facto mechanism for funding existing providers, with little attention to service quality, and little real responsiveness to patients’ desires to move to preferred providers. Turning to country-specific cross-sectoral patterns, while not as unambiguous as for Romania and Tajikistan, Table 3 and the above discussions point to a fairly clear trajectory vis- à-vis Turkey – namely an effort to effect reform in a compact-oriented direction, though one which is halting and uneven in the face of resistance from status- quo oriented interests. In the Kyrgyz Republic, the disconnect between a strongly reform- oriented health sector, and a laggard education sector might be interpreted as signaling differences among sectoral political leaders and donors in their commitment to change in a setting which is not rigidly control-oriented. The outlier seems to be Albania, where there has been almost no systematic progress in a compact-style direction in any sector Page 43 43 even though, prima facie, the data in Table 1 suggest that the environment is no less favorable than, say, the Kyrgyz Republic for such efforts. Voice, linked to ser vice del ivery (“serdel voice”) The second link of the ‘long route’ accountability chain, voice – with citizens as principals and politicians/policymakers as agents – enables citizens to hold politicians/policymakers accountable, thereby providing the incentives for the latter to follow through on their ‘compact’ agreements with service providers. As WDR04 details, ‘voice’ can be manifested through two distinct, but complementary mechanisms. The first mechanism comprises elections, through which citizens pass judgment on the performance of policymakers via their votes. Because the impact of electoral voice at the national level on service delivery is highly circuitous,and is diluted by the myriad of other influences on how people vote nationally, (and also because reforms of the national electoral system lie very far from the World Bank’s mandate) its role will not be considered further in this study. [By contrast, because the responsibilities of local politicians and policymakers are narrower, and more directly linked to service delivery, subnational electoral voice can have a more direct impact – and so will be examined in Chapter 4.] The remainder of this section will focus on the second voice mechanism, namely ‘voice, linked to service delivery’. The second ‘voice’ mechanism comprises an ‘information strategy’ which brings into the public domain information on the performance of public agencies which can enable citizens to hold politicians, policymakers and providers accountable for the efficiency and effectiveness with which they deliver services. WDR04 argues strongly that to be useful such information must be specific: “identifying specific government decisions, specific decisionmakers, and the effect of the decision on voters individually or as a group. [In an education example], the information was not about the general quality of education, or general budget support, or even about leakage from national education budgets. It was about one type of transfer, disaggregated to the school level, responsibility for which was easy for parents and voters to assign to the school principal. In addition, the information make clear to voters what the school should have received…… Information about broad aggregates of public sector performance – whether based on surveys, budget studies or other methods – is less likely to be politically relevant. Why?….. At the end of the day, these efforts tell citizens what they already knew – that services are bad. They might tell them exactly how bad and which services are worse than others – roughly the information citizens already had, but more quantified. What citizens do not have, and what they need help in getting, is information about how bad their neighborhood’s services are relative to others’ and who is responsible for the difference.” 25 25 WDR04, pp.87, 89. Page 44 44 Table 4 reports on the extent to which voice of the kind described above (that is, voice linked to service delivery – referred to below as ‘ serdel voice’ ) was evident in each of the 14 sector-country observations. Overall, the pattern is uneven, with the distribution of effort skewed strongly towards the low side: there is only one instance of ‘extensive effort, five instances of ‘some effort’, and eight instances of only ‘marginal’ effort. In relative terms, the effort has been greatest in the education sector, which has incorporated serdel voice quite systematically into sector reforms in Romania, Turkey and Albania. · Education reforms in Romania included (i) a commitment to public access to the results of national and international education assessments; (ii) a redesign of the National Education Database to make it accessible to a large number of people; (iii) establishment of a Romania Education Network to make the operation of the system accessible to the public; and (iv) (to begin in 2006) publication of an annual synthesis report on the education sector, using the new indicators and database. · Turkey committed to make M&E results -- including the results of performance on international standardized tests -- available on the Ministry of National Education’s web page. · Albania’s far-reaching formal commitments to voice (incorporated both in Bank-supported education sector projects, and in the country’s Poverty Reduction Support Credits) included: (i) publication of learning assessment results and other education performance indicators; (ii) publication of a first Statistical Scorecard on Albanian Education, and an Annual Education Performance Report; and (iii) publication in newspapers of a full list of schools rehabilitated the previous year, including information by school on the levels and source of funding. It remains uncertain, though, to what extent the Albanian authorities will follow through on these commitments. Table 4: Voice, linked to service delivery : 14 sector-country observations Romania Turkey Kyrgyz Albania Tajikistan VOICE IN EDUCATION extensive some marginal some marginal VOICE IN HEALTH some Some* some marginal marginal VOICE IN URBAN WATER marginal marginal n/a marginal marginal * reforms in early stage of implementation imply shift from marginal/some to some/extensive Page 45 45 For health, the Romania, Kyrgyz and (upcoming) Turkey sector reforms all support the preparation of regular household surveys, and surveys on consumer satisfaction, health-related knowledge, attitudes and behavior. In addition, in all three countries the process of reform included active efforts to solicit participation from multiple stakeholders, both professionals and users. In none of the countries, though, do the reforms seem to have incorporated efforts to report publicly on a systematic basis as to how well the sector is performing (though admittedly, certainly at facility level, the health sector less readily lends itself to metrics which could provide the basis for performance data than does the education sector). In urban water no systematic efforts to strengthen serdel voice were identified in any of the countries – although (as Box 2 in Chapter 4 details) a broader local-government performance monitoring initiative in Turkey could have some positive spin-offs vis-à-vis water utilities as well. Page 46 46 Chapter 4: The ‘Intermediate Route’ of Accountability -- De-centralized Service Delivery As used here, 26 the intermediate route of accountability refers to the de-centralization of responsibility for services to local governments. [This paper hyphenates the term ‘de- centralization’ to underscore that, especially in ECA, the process need not be one of undermining sensible central government in favor of some utopian notions of community and local control, but rather one of re-aligning states which had been excessively centralized to achieve a more politically and economically pluralistic balance between central and local levels.] Both the political and economic logics for engaging de- centralization are potentially compelling (though, as will emerge from the discussion, realizing the gains is harder than might appear at first sight): · Viewed through a political lens, the intent of de-centralization is to transform state-society relationships which had become dysfunctional into ones with the prospect of providing a new platform of legitimacy for governance more broadly. If successful, de-centralization can help harness civic energies for the provision of collective public goods, can provide a seedbed for the emergence of new generations of political leaders and, most broadly, can strengthen checks and balances in society in an inclusive way, by providing a multiplicity of new outlets for political action. · Viewed through an economic lens, aligning responsibility for services according to the principle of ‘subsidiarity’ 27 has the potential to improve both technical efficiency (by activating local information to help hold service providers accountable for performance), and allocative efficiency (by facilitating trade-offs among competing priorities at the levels where the costs and benefits are actually realized). Figure 2 depicts how the intermediate route of de-centralization translates into the accountability framework of the WDR. Three sets of accountability relationships seem key for subnational service delivery. · subnational electoral voice; · the center-local compact; and · the local level arrangements for service delivery. Each will be considered in turn. [As will become evident, these three overlap with – but are not identical to – the traditional political, fiscal and administrative trio, which ordinarily guides analysis of de-centralization.] 26 The term is not used explicitly in the WDR. 27 The principle of subsidiarity posits that responsibility for service delivery should be assigned to the lowest level of government capable of internalizing the relevant externalities. Page 47 47 Sub-national electoral voice. A necessary condition for de-centralization to be credible as a distinctive accountability arrangement is that it is underpinned by a tier of government (municipal or provincial) which is downwardly accountable to its citizens via electoral voice. WDR04 highlights both the promise and limitations of electoral voice in de-centralized settings. The promise arises from the close proximity between voters and local politicians/policymakers: “voters may make more use of information about local public goods in their voting decisions because such information is easier to come by and outcomes are more directly affected by local government actions. And political agents have greater credibility because of proximity to the community and reputations developed through social interaction over an extended period” (p. 90). As the WDR makes clear, the presence of these attractive attributes cannot be taken for granted. In some settings, the results of local elections may simply reflect longstanding power imbalances and patterns of social polarization. In other settings, these attractive attributes may be latent – but may need proactive initiatives to be activated. “ [Where] de-centralization is driven by a desire to move services administratively closer to the people……the assumption is that [it] works by enhancing citizens’ voice in a way that leads to improved services. But on both theoretical and empirical grounds this could go either way. The crucial question is always whether de-centralization increases accountability relative to its alternatives. If local governments are no more vulnerable to capture than the center is, de- centralization is likely to improve both efficiency and equity.” Sub-national electoral voice is formally in place in four of the five countries being considered here: Page 48 48 · Turkey has a well-developed system of elected municipal government, within the framework of the centralized state structure, established during the post-1923 Ataturk era. The system has as its foundation the 1930 Municipalities Law (which in turn had, as an international precedent, the French system of the time and, as a domestic precedent, a ‘prehistory’ of local governments dating back to the establishment of the Istanbul municipality in 1854). Currently, there are 3,200 municipalities, the largest 16 of which have been designated ‘metropolitan municipalities’. There also are 81 Special Provincial Administrations, which have functioned largely as deconcentrated entities, though supported by elected councils. Turkey’s draft Public Administration Framework Law, and related laws also currently under discussion, propose potentially far-reaching changes in the country’s intergovernmental system. · Romania and Albania both effected far-reaching reforms in the 1990s to (re)create, in the wake of the collapse of their centralized, communist regimes, a de-centralized structure consistent with the requisites for accession to the European Union. Albania started on a path of political de-centralization with local democratic elections in 1992. In 1998, the country incorporated the key principles of the European Charter of Local Self-Government into its new Constitution. Albania’s territory is formally divided into 374 first level local self- government units (65 municipalities and 309 communes). Romania’s 1991 Constitution and subsequent Law on Local Public Administration re-established a two-tier system of government which had existed before the advent of the communist regime. The subnational system comprises 41 counties (judets), 79 municipalities, 182 towns, and 2,682 communes; the last three all have the same legal authority and administrative structure. A 1998 Law on Public Finance expanded and clarified the role of local governments (along lines discussed further below) · The starting point for both the Kyrgyz Republic and Tajikistan is the four tier structure (center, oblast, rayon and commune), left behind by the Soviet Union. The Kyrgyz Republic has made substantial progress in clarifying the ambiguity between deconcentration and de-centralization inherent in this structure, with mayoral elections for its 470 communes (ayl okmotu/kenesh) in 2002, and 2003 legislation which explicitly defines only the commune tier as ‘local self government; a dialogue currently is underway within the country as whether both of the remaining two deconcentrated tiers are indeed necessary. By contrast, in Tajikistan the ambiguity remains. There is no credible tier of elected local self government; the lowest tier (known in the country as the jamoat) remains an administrative appendage of the rayon tier immediately above. The center-local compact . The second set of accountability relationships relevant for the intermediate route comprises the compact between central and local governments. (This encompasses, but is broader than, the traditional categories of fiscal and administrative de-centralization.) As the three arrows which emanate from the ‘center’ in Page 49 49 Figure 2 highlight, a central-local compact could refer to a variety of quite different institutional arrangements: · Devolution implies the handing over of power, resources and responsibility in key areas to subnational governments (that is, locally elected politicians, and the policymakers who work for them). For those responsibilities which are transferred, central governments have no residual authority over their local counterparts. · Delegation implies a more conditional handing over of responsibilities to elected subnational governments. Here the notion of a center-local accountability compact is quite precise, with subnational authorities and service providers accountable to the center for performance, and hence a regulatory oversight role for the center. Delegation is the predominant form of de-centralization in all four countries under study which have in place subnational electoral voice. · Deconcentration is not a form of de-centralization at all (though it is sometimes misleadingly treated as such). As figure 2 implies, it changes not at all the compact relationship between central governments and front-line providers (though the former may now be represented by regional offices, rather than those in the capital cities). Subnational governments, if indeed there even exists any politically representative body at that level, remain an irrelevancy to the service delivery accountability chain. Table 5: Assignment of revenues and responsibilities to subnational governments ROMANIA TURKEY KYRGYZ ALBANIA EDUCATION Assignment of responsibility Extensive None/marginal Some- ambiguous Shared- ambiguous …of revenue limited None/marginal Some- ambiguous Shared- ambiguous HEALTH Assignment of responsibility None/marginal None/in transition None/marginal Shared- ambiguous …of revenue None/marginal None/marginal None/marginal Shared- ambiguous WATER Assignment of responsibility Extensive Extensive [no data] Formally extensive, de facto ambiguous Page 50 50 …of revenue Shared Shared (de facto) [no data] Extensive The (delegated) compact between central government on the one hand and local authorities and front-line providers on the other comprises at least three facets: assignment of roles; provision of fiscal resources; and regulatory, fiduciary and other forms of oversight. To begin with the role and resource dimensions of the central-local compact (this comprises, of course, the central preoccupation of work on fiscal de- centralization) Table 5 summarizes the patterns across the 14 country-sector observations. As the table suggests, there are important sector-specific differences in the extent of de-centralized service delivery: · Consistent with a familiar pattern the world over, in all sample countries other than (non-de-centralized) Tajikistan, locally elected municipalities have extensive responsibility for water systems. 28 · Though, as noted earlier, far-reaching health sector reforms have been undertaken or are on the agenda in at least three of the countries (Kyrgyz Republic, Romania, Turkey), in all cases the principal shift in responsibility is towards a newly-created insurance agency, not to subnational authorities; 29 for this reason, health sector reforms will not be considered further in this section. · Education sector reforms in Albania, Kyrgyz Republic and Romania all mandated a substantially greater role for local authorities in basic education, with responsibilities shared to some degree with central government. As is also suggested by Table 5, in all of the countries under study some important challenges remain in clarifying the assignment of roles and responsibilities. · Turkey confronts both longstanding and new challenges. A longstanding challenge concerns the financing of investment in the water sector where, despite the formal assignment of responsibilities (including investment responsibility) to municipal authorities, there has been a repeated pattern of de facto centralized financing of investment via the nonpayment by municipalities of loans either guaranteed by central government, or made directly by the parastatal Iller Bank. The new challenge concerns the ambitious program laid out in the draft Public Administration Framework Law to transfer a wide variety of responsibilities in at least seven sectors 30 from central to provincial or municipal authorities. The 28 The bank has not been active in the Kyrgyz Republic, so this country-sector observation is not considered in the present paper. 29 Although note that Turkey’s proposed administrative reforms vest formal ownership of hospitals and other health sector infrastructure – which are intended to operate with significant autonomy, funded via insurers – in the hands of the Special Provincial Authorities. 30 These include significant subsets of activities currently within the Ministries and Directorates of Health, Culture and Tourism, Environment and Forestry, Agriculture and Rural Affairs, Social Services, Youth and Page 51 51 details of what will be entailed – both as to which specific functions within sectors, and as to the sequencing within and between sectors – have not yet been worked out. · In both Albania and Romania, rapid de-centralization led to imbalances in the assignment of responsibilities revenues. Romania’s assignment of revenues both minimized the role of equalization transfer grants, and transferred a variety of responsibilities (notably for education) without transferring commensurate revenues; the result – which urgently needs addressing – is underexpenditure in education, especially in poorer areas. Albania designated a variety of functions – including education and health – as ‘shared’ between central and municipal authorities, but did not clarify in a disaggregated way how these shared responsibilities were to be allocated across tiers of government. The result has been confusion, finger-pointing, and an evasion of accountability – precisely the opposite of the sharpened accountability which de-centralization was intended to achieve. Delineating these shared responsibilities more precisely is thus an urgent priority. · In Central Asia, Kyrgyz Republic has made the political commitment to de- centralization; the country held its first mayoral elections in 2002. But the elected mayors were provided with neither clear roles nor any resources to follow through; the result is a growing disillusionment – including the reported resignation of up to 30 percent of the elected mayors. It is urgent that the country makes an initial round of assignment of responsibilities (and associated resources). As for Tajikistan, the country has not yet decided whether or not to begin the process of building a genuinely downwardly accountable tier of local self-government. Meanwhile, donor-supported activities at the community-level are accelerating. As discussed further in the section on ‘client power’, if government chooses to take a significant step in the direction of de-centralization, these community-level activities could provide a useful platform for further progress. Turning to central government’s regulatory, fiduciary and administrative oversight mechanisms necessary to underpin a central-local compact, the limited attention given to regulatory oversight was evident in Table 3. The gap seemed especially striking in the urban water sector: as a classic utility, the relevance of a regulator seems prima facie most evident for this sector; yet this role has not been realized in practice in any of the countries. How fiduciary and administrative oversight of subnational authorities by central government is handled also can have important implications for the effectiveness of de- centralization: Sports, Industry and Trade, and Public Works and Settlement. As of June 2004, discussions of the law in Parliament had largely been completed, and it was working its way through the constitutional process. Page 52 52 · For fiduciary oversight, in the absence of reliable mechanisms for monitoring how local governments spend resources, central governments have tended to resort to controls. · For administrative oversight, while de-centralization has the potential to enable local governments to set their own terms and conditions of employment, in practice central governments (or public sector employee associations) tend to constrain flexibility. In Turkey, for example, the quite far-reaching formal authority enjoyed by municipalities is undercut by a myriad of ex-ante controls on inputs and processes imposed by central government, including requirements for submission of local acts, decisions, local and regional plans, budgets and financial statements. To cite just one instance, all appointments of staff by Turkish municipalities need approval from the Prime Minister’s office; and all appointments are under quite precise terms and conditions mandated by the national civil service law. The point is not that subnational governments should be exempt from scrutiny, but rather that the processes of monitoring should follow the principle of ‘watch don’t control -- proceed in ways which strengthen transparency without inhibiting the ability of local authorities to adapt flexibly to their local circumstances. The local-level arrangements for service delivery. The third set of relationships to be considered as part of the intermediate route of accountability comprises the local-level arrangements for service delivery. As noted earlier, the special appeal of local-level service delivery is the proximity of citizens to – and hence better information (through direct information) about -- local leaders and providers. Two weaknesses at the local level often are identified as potentially undercutting these potential advantages. The first is technical, namely a thin local pool of skills. The second is more directly relevant to accountability -- namely that local social arrangements bear little resemblance to some idealized version of grass-roots democracy, and that the risks are high of elite capture. Survey results in the five sample countries (and elsewhere) do not support the view that elite capture and corruption are higher at the local than the central level. Surveys conducted in Romania and Turkey (both in 2000) showed that citizens perceived local governments more positively than their central counterparts across a variety of indicators (e.g. service quality, extent of corruption), though the differences are quite small. A survey conducted in Albania in 1996 showed little significant difference in how central and local authorities were perceived, but revealed a strong preference on the part of citizens for shifting more responsibilities to the local level. 31 But even in settings where there is no track-record of civic participation (making the risk of elite capture higher) active interventions can help reshape relationships between citizens and local governments (and can also help build technical capacities). Unlike 31 See Fikret Adaman, Ali Carkuglu and Burhan Senatalar, Household view on the causes of corruption in Turkey and suggested preventive measures , Istanbul, TESEV, 2002; and World Bank, Diagnostic Surveys of Corruption in Romania. June 2001. Page 53 53 other regions of the Bank, ECAVP appears to have no operations in its loan portfolio which focus on building the participatory (and technical) capacity of local governments. But three grant-based initiatives were identified, and are described in Box 2. Also, though not explicitly described in these terms, community-driven local initiatives in Albania and the Kyrgyz Republic (described in the next section) also make some contribution to local governance capacity building. Page 54 54 Box 2: Strengthening local voice, linked to service delivery, in three ECA countries Recent Bank-supported initiatives in Turkey, Albania and Ukraine illustrate three different approaches to using information to strengthen the interactions between local governments and their citizens. A US$350,000 Institutional Development Fund grant, awarded in 2001 to Turkey’s Ministry of Interior (which is responsible for oversight of the country’s municipalities) supported the development of a local government performance monitoring system. Data has been collected on the demographic and economic characteristics (e.g. population, per capita and income) and on dozens of performance-related indicators (including information on the efficacy of local-level revenue raising, the composition of local expenditures, and the efficiency and effectiveness of the provision of water, public transport and other local services) of over 100 municipalities. Customized computer software enables users to make direct comparisons across municipalities as to both their characteristics (so like can be compared with like) and performance. The system is expected to become publicly available on the internet by the end of 2004. Albania’s Tirana Transparency Project (funded in 2001 via a US$615,000 grant from the Dutch Trust Fund) worked to improve the flow of citizen-to-government information via the development of a Citizen Report Card, and to improve the flow of government-to-citizen information via the establishment of a Citizen Information Center and the development of a municipal public information strategy. The results of the Report Card (which was based on a combination of user surveys and technical service parameters) were presented at a conference in late 2003, and made available publicly. The Information Center operates principally as a one- stop-shop for citizens seeking information (including clarification of rules and regulations) from local government. As of early 2004, approximately 3500 people visited the center monthly. The People’s Voice Project in Ukraine (which was featured at the Shanghai Conference on scaling-up poverty reduction) aimed to strengthen the capacities of citizens and local government officials to interact with one another via a combination of service delivery surveys, policy development training for public officials, support for public hearings, and media participation. A 2002 final evaluation of the initial, four city pilot (funded via a US$1.1 million grant from the Canadian International Development Agency) found “continuing active relationships between NGO coalition members, initiative working groups and municipal officials; municipal officials still seek citizen feedback and input through the use of public hearings, consultations, and through plans for further surveys; and an increase in transparency of local government decision-making as evidenced by budget hearings and public hearings in education, transportation and housing issues.” 32 CIDA has provided a second grant of US$2.4 million to enable the project to be expanded into six additional cities. 32 Kazuko Ogawa and Tom Monastyrski, “Case Study: People’s Voice (Ukraine): Better Governance through Public Participation” paper submitted to 2004 Shanghai Poverty Conferenc, p.16. Page 55 55 Chapter 5: The Short Route of Accountability – Client Power In the short route of accountability – depicted in Figure 1 as ‘client power’ – citizens as principals hold front-line providers accountable for performance. WDR04 identifies two distinct variants of client power – choice and participation. While both empower clients, they differ starkly in their implications for civic engagement in the political process: the choice mechanism distances citizens from politics; participation deepens civic engagement. This section will consider both variants of client power, though given the paper’s governance focus, more attention is given to participation. Within the category ‘participation’, two distinct subvariants are considered: client power within individual sectors, and multisectoral, community-driven client power. Client power through choice. ‘Consumer sovereignty’ – expressed in the ability of purchasers to select among multiple competing providers – is the classic example of client power through choice. The ubiquity of private schools and private doctors in many countries make these straightforward market transactions part of the service delivery landscapes in both the education and health sectors (though, of course, unsubsidized and unregulated private delivery can address neither the equity nor the service-standard challenges that are central to both sectors). All five sample countries make allowance for private schools, though in practice in none of the countries do they account for more than 2-3 percent of total primary or secondary school enrollments (hence the consistent ‘marginal’ rating in Table 6). 33 Privately provided and funded health care is somewhat more common, notably in Turkey where about 15 percent of doctors work on a fee-for-service basis only, and about 60 percent of public sector physicians also work part-time in private practice. (The third service considered in this study, the provision of piped water into households, is a classic example of a natural monopoly, so does not lend itself to ‘client power through choice’ – though the failure of such monopolies to provide potable water has led to the emergence in many countries of a competitive marketplace for bottled water). Table 6: ‘Client power’ through choice: 14 country-sector observations Romania Turkey Kyrgyz Albania Tajikistan EDUCATION marginal marginal marginal marginal marginal HEALTH extensive Some* Some* marginal marginal URBAN WATER N/a N/a N/a N/a N/a 33 Note, though, that the situation is different for tertiary education. In Romania, for example, 120,000 students are enrolled in private universities, as are 15 percent of students in the Kyrgyz Republic. Page 56 56 Even with public financing, client power through choice potentially can be a feasible option for the governance of service delivery. Key here is that providers are not directly funded publicly, but rather that public financing follows the user, who is free to choose among competing providers. In education, this vision underlies the passionate quest of some policy advocates for a voucher-based system (though there was no evidence of any such system in any of the five sample countries – reinforcing the ‘marginal’ ratings in Table 6). In health, as was summarized in Chapter 3 (Table 3 and Box 1), the equivalent result can be achieved via a combination of health insurance and capitation payments to providers – part of the actual or proposed institutional landscapes in Romania, Kyrgyz and Turkey (hence the ratings in Table 6). Client power through sectoral participation. ‘Client power’ through participation at the sectoral level engages service users in decision-making -- giving parents associations oversight responsibility over how annual school budgets are spent for example, or involving farmers in operating irrigation systems via water users associations. WDR04 lauds the increased attention to participatory client power as a “welcome tonic for the top-down, technocratic orientation that has characterized much development thinking until now”. It points also to a number of key limitations: · The delivery of some services requires specialist skills and decision-making of a kind which are best left in the hands of technical agencies; · The risks of elite capture sometimes can be even greater at the local than national levels, implying that a shift of authority from central government to communities could leave poor people worse off than before; · Participation takes time, effort, and social commitment, so may be more plausible for some services (e.g. schools and local irrigation, where participants have a direct incentive to engage) and communities than others; · It may be relatively easy to mobilize participation temporarily (e.g. for a new local investment) than on an ongoing basis (e.g. for maintenance), raising the risk of unsustainability; · Parallel, client-power approaches to service delivery can drain the best talent from government (and ‘cherry-pick’ also the most socially gratifying roles), thereby reinforcing a vicious circle of weak government capability. Participatory client power is thus in part a temporary substitute for weak government capability, providing a framework for getting resources to the front-line, with pressure for performance, in settings where politics is otherwise strongly clientelistic. But it also can complement other accountability mechanisms. Even where the predominant (central or subnational) service delivery arrangements are compact-based, institutional space can be made for parents associations, water users associations and the like to participate, strengthening accountability and contributing to improved performance. Page 57 57 Table 7 summarizes the extent to which participatory client power has been supported across the 14 sector-country observations. Overall, participatory mechanisms are used more intensively in education than in the other sectors; and are more in evidence in Tajikistan than in the other countries. A comparison of Tajikistan and Romania reveals how client power can play a role in education both in institutionally weaker and stronger settings: · In Tajikistan, where systemic reform of education has not been on the agenda (at least until very recently) the World Bank facilitated the creation of Parents- Teachers Associations through an Education Reform Learning and Innovation Loan, and subsequently facilitated the involvement of these PTAs in school rehabilitation (as contributors, and as informal supervisors of the work of contractors), as well as in the preparation of proposals to access Bank-funded thematic grants for schools (e.g. to ease transport or financial obstacles to school attendance); · In Romania, by contrast, there has been far-reaching reform of the education system. Even so, as noted earlier, the Romanian authorities promulgated in 1995 a Law on Education mandating the establishment of School Management Councils of five to eleven members, including the principal plus representatives of teachers, local public administration, local firms and parents, and gave them quite substantial operational autonomy. Table 7: ‘Client power’ through participation: 14 country-sector observations Romania Turkey Kyrgyz Albania Tajikistan EDUCATION extensive marginal Some* marginal extensive HEALTH marginal marginal marginal marginal Some? URBAN WATER marginal marginal n/a some some More broadly, what is most striking about the empirical pattern in Table 7 is the unevenness of the application of ‘participatory client power’ at both the sectoral and country levels. · Almost no evidence was found of the use of participation as a means of client power in the health sector 34 , 34 The health sector-country reviews scanned systematically for evidence of (i) recruitment of paraprofessionals from within the community; (ii) inclusive governance structures for primary health clinics; and (iii) mechanisms for information exchange between front-line service providers and users. Page 58 58 · In water, active efforts to involve users emerge more as the idiosyncratic consequence of specific project circumstances than as part of the standard toolkit of doing business. In Albania, the Bank has facilitated the establishment of Consumer Panels to work with water utilities in four pilot cities (channeling service suggestions, interacting with management, and conducting public awareness campaigns). And in Tajikistan, the Bank is piloting an effort to replace the now almost entirely defunct municipal housing organizations, by fostering community participation in the management of the end-stage (e.g. apartment building) elements of the Dushanbe water network. · Turkey stands out for the consistent absence in any sectors of mechanisms for actively engaging users – a pattern which is consistent with the country’s longstanding legacy (though perhaps one which currently in the early stages of profound change) of a bureaucratic, top-down, control-oriented central government. · Participatory mechanisms also appear to be underutilized in Albania, even though client power might seem to be a logical institutional response to the combination of political openness plus poor governance prevailing in that country, Community empowerment through social investment funds. As noted, ‘participatory client power’ need not only be sector-specific. Social investment funds (SIFs, with the funds generally provided by donors) comprise an increasingly common multisectoral variant worldwide, including in ECA. In SIFs, communities work together with an agency which operates at arms-length from government to prioritize among a multisectoral array of local public investments, and participate in their implementation and subsequent maintenance. In four of the five sample countries, SIFs have become a major provider of resources for local public investments, especially in rural areas: · The Albania Development Fund was established with the support of the World Bank in 1993, initially as a short-term poverty alleviation program in the wake of the collapse of Albanian communism. Since its inception, the ADF has received commitments of US$73 million (US$26 million from the Bank, and US$47 million from other donors). As of 2003, the ADF had implemented over 1,200 microprojects. About half comprised the rehabilitation of local roads and about a quarter improvements in local water supply; with the remainder spread among rehabilitation of schools, health posts, local irrigation systems, plus myriad other local activities. · Tajikistan’s National Social Investment Fund was established in 1997, in the immediate aftermath of that country’s civil war, with the World Bank providing an initial US$4 million. By 2002, NSIFT had implemented 184 microprojects, Other than some modest evidence for the last of these (which is closely akin to the ‘serdel voice’ mechanism considered in Section III) no examples were uncovered. Page 59 59 covering approximately 600,000 people. In 2003, a process of scaling-up began, with an additional US$10 million from the Bank, plus increasing interest among bilateral donors in using NSIFT as a vehicle for channeling additional resources to local public investment. · Romania’s Social Development Fund was launched in 1999, with an initial round of support of US$10 million from the World Bank, plus US$17 million from other donors. Within two years the RSDF has approved 264 subprojects, including 126 for upgrading rural roads, 27 for rural water supply systems, 25 for infant and child care centers, and 56 for income generating activities. Based on this rapid implementation, in late 2001, the Bank provided a further US$20 million in a second phase of support. · The Kyrgyz Republic’s Village Investment Program was launched in 2004 on the basis of a US$18 million (check) World Bank project. A pilot phase successfully initiated over 150 sub-projects, spread across multiple sectors. SIFs have been an uncomfortable fit within country programs. Their multisectoral, community-level mechanism of prioritization and decision-making seemingly conflicts with sector-specific approaches (because, for example, communities may seek investments which are not in a sectoral master plan – or because sectorally earmarked investments may be inconsistent with community priorities). And the creation of a parallel agency responsible for local public investment seemingly is at variance with efforts to strengthen the role, resources and responsibilities of elected local governments. In practice, experience in at least three of the country cases 35 suggests that these seeming conflicts can be resolved in implementation: · Following a 1999 restructuring of the ADF, Albania’s local (commune) governments have played an increasingly central role in ADF-funded public investments. Communes prioritize and select which investments are to be funded from a list of submissions made by communities. They also are taking on an increasingly central role in both procurement and implementation (with the ADF serving as guarantor of the competitiveness and transparency of the process). The ADF also has built strong working relationships with sectoral line ministries, including mechanisms which ensure that supported investments are consistent with sectoral master plans. · The Kyrgyz Republic’s Village Investment Program has been designed in a way which gives a central role to the emergent local self-governments (the local ayl kenesh councils, and their ayl okmotu administrative arms). Ayl kenesh and ayl okmotu officials participate (‘a voice, but not a veto’) throughout the community- level decision process; the ayl kenesh formally adopts the community investment plans (indeed, in project where the VIP is active, these plans have replaced earlier, perfunctory exercises as the ayl okmotus submission to the rayon tier of 35 For the reason outlined in footnote 4, the Romania case was not examined in sufficient depth to address the issue of the relation between the social fund and local government. Page 60 60 government); and representatives of local self-governments comprise one-third of the members of the national governing board of the VIP implementing agency. · In post-civil war Tajikistan, many donors (including the Bank) directed their support to communities, bypassing government structures entirely. Though this facilitated investment in the short-term, an increasing number of donors recognize that it is not institutionally sustainable. Consequently, an intensive dialogue is underway, both among donors, and between donors and the government, as to how these initial efforts might evolve into a ‘community-linked development’ program, which progressively engages and builds the capacity of a nascent jamoat tier of local self government. In sum, these experiences point to a potentially important contribution of SIFs, going beyond their role as a ‘client power’ alternative to dysfunctional national governments. As was highlighted in Chapter 4 and Box 2, the intermediate route of accountability only is effective if participation at the local level is inclusive, ensuring that local governments are not captured by elites. If designed along lines evident in Albania and the Kyrgyz Republic, SIFs can play a key bridging role -- as a transitional mechanism which, not only supports poverty-reducing investment in the short-term, but also helps strengthen local participation and build local capacity more broadly, exiting from the stage as local governments mature.