These case studies and the accompanying synthesis were prepared for the February 2004 Regional Seminar and Learning Event on Local Governance and Pro-Poor Service Delivery, hosted by the Asian Development Bank in Manila, Philippines. This Seminar is the occasion for practitioners and policy makers to come together to share lessons and to promote what is a critically important policy agenda in Asia.

UNCDF is delighted to be able to contribute to this agenda, and to do so in partnership with the Asian Development Bank and the Asian Development Bank Institute. We believe that these case studies on the role of local government in pro-poor infrastructure and service delivery in rural areas will help contribute to greater learning and to a more informed policy debate.

We also believe that this is an area which has been neglected. The majority of the literature on local government, and on better service delivery practice, focusses on the urban setting. Yet the challenges for improving service delivery in rural areas by rural local governments are in many ways qualitatively different, and frequently more daunting. Recipes which work for improving big city service delivery are often quite inappropriate when applied at rural Commune level. However, the fact that the great majority of the poor in Asia continue to live in the rural areas is a compelling reason for us to better understand these challenges, and also the opportunities for change.

Indeed the more positive note is that we believe there are important but little-known innovations - in policy, procedure and practice - which can greatly enhance the performance of rural local governments in Asia, and which can be replicated and mainstreamed into policy. The focus of these case studies is on a number of such innovations which UNCDF has been piloting in the countries under review. The three Asian case studies are supplemented by the case study from Uganda, in recognition of the widespread interest that Ugandan experience in decentralization is eliciting.

About this publication

This publication includes a synthesis of the case studies, prepared under the editorial direction of Roger Shotton, with substantial contributions by Richard Slater, Andrew Preston and Mike Winter. Summaries of the case studies are included in the appendix, and a CD-ROM with the full-text of the case studies is included with the publication. The case studies themselves cover the following themes:

Bangladesh: Local Governance & Service Delivery to the Poor, by Richard Slater and Andrew Preston. This study outlines lessons on the scope for piloting innovations in financing, participatory planning and implementing of basic infrastructure & service delivery by rural Union Parishad authorities in one district of Bangladesh, and documents their effectiveness, even within a national policy context which is not at the moment especially conducive for decentralization.

Nepal: Local Government Infrastructure & Service Delivery in Rural Areas, by Michael Winter. This study outlines issues and challenges in promoting innovations for more effective and equitable financing, planning and management of basic infrastructure and services by authorities at District and Village level in Nepal, and for feeding these into the national policy framework, but within the limits imposed by the Maoist insurgency and the current suspension of local elected bodies.

Cambodia: Decentralization Reforms and Commune-Level Service Delivery in Cambodia, by Leonardo Romeo and Luc Spyckerelle. This study documents the effectiveness of recent innovations in the financing, planning, and management of local infrastructure development and decentralized service delivery by commune councils. These innovations have now been mainstreamed as national policy in Cambodia, as part of the decentralized governance systems of elected commune authorities in both rural and urban areas.

Uganda: Lessons from Experience in Decentralizing Infrastructure & Service Delivery to Rural Areas, by Martin Onyach-Olaa. This study outlines the radical and comprehensive policy and institutional framework for devolution and local service delivery in Uganda and, within this, the role played by such innovations as performance-linked funding to District and sub-County governments in rural areas.

The publication of these studies, together with a set of companion case studies prepared for the Africities Conference (held in Yaounde', Cameroon, December, 2003) marks the beginning of an important new direction on the part of the UNCDF Local Governance Unit. As an organization with a unique portfolio of operations across the Least Developed Countries, all aiming to promote pro-poor service delivery innovations in partnership with rural local governments, we intend to do more to document our experience and to share the lessons which emerge from this. We therefore look forward to contributing further to policy debates in this important area and to help, in a very modest way, in achievement of the Millennium Development Goals (MDGs).

Kadmiel Wekwete
Director, Local Governance Unit
United Nations Capital Development Fund

Introduction

Country Contexts

LDP Innovations

Outcomes Achieved

Lessons Learned and Future Challenges

INTRODUCTION

The Underlying Assumptions

This report - like the case studies it synthesizes - is founded on three assumptions: firstly, that poverty in Asia is still - and despite rapid urbanization - primarily rural (in numbers and in incidence); secondly, that rural local governments (LGs), despite their many weaknesses, have a potentially key role to play in the basic infrastructure and service delivery (ISD) which effective poverty reduction requires; but, thirdly, that the challenges for improving ISD through rural LGs are qualitatively much greater than through urban LGs, and that both these challenges - and also the opportunities - for making rural LGs more effective are poorly understood and documented.

The Local Development Programme approach

The case studies themselves focus on experiences of innovation in improving ISD through rural LGs, and their outcomes, in three Asian countries (Nepal, Bangladesh, and Cambodia) and also in Uganda (given the considerable interest, even outside Africa, that the Ugandan experience is eliciting).

These four sets of innovation all share a common framework. They were all introduced within the framework of UNCDF-supported Local Development Programmes (LDPs). The LDP is a generic approach to promoting more effective ISD through rural LGs, by twinning innovations in the funding of ISD with other capacity development innovations to support improved planning, budgeting, implementation, and overall accountability of LGs. These aim is to use the innovations as a basis for policy influence and hence for wider replication. Although the LDP model is generic, it is tailored to context.

COUNTRY CONTEXTS

Patterns of similarity & difference

The four countries are of course each very different in very many ways. But one common feature is the fact that in each country poverty is primarily a rural problem: both the relative incidence of poverty is twice as high (or higher), and the absolute number of poor is far greater in rural areas.

The policy and institutional contexts for rural local government and for ISD in rural areas exhibit patterns of similarity, but also important differences. The latter can be viewed along several dimensions:

Institutional & Functional factors:

• Size, population & population density of LG jurisdictions;
• Existence of formal institutional interface between communities and elected LGs;
• Degrees to which LGs exercise control over technical services and extent to which LGs are "organically" linked to sector line departments;
• Relationship between LGs and central government, and degree of supervision and support by the latter.

Fiscal factors:

• Local fiscal revenue raising powers;
• Relative significance and clarity of expenditure assignments;
• Importance and types of inter-governmental fiscal transfer system.

Representational factors:

• Modes of representation and election;
• Statutory representation of disadvantaged and/or minority groups.

Political drive

• The degree of political willingness and drive to empower elected LGs, entrust them with responsibilities and resources and support them.

In brief, the four country contexts can be characterized as follows:

• In Uganda, the institutional framework for sub-national government is a fully devolved on, with two main levels of elected local government: the sub-County ( a relatively small unit) and the District (a large unit, with a full complement of line department staff under full council control). There is a strong political drive for decentralization, and these LGs have been fully mandated to undertake all ISD functions; the allocation of financial resources to LGs is being gradually adjusted to match.
• In Cambodia and Bangladesh, there is a relatively small elected lower level local government unit (at Commune and Union level, respectively), with one or two general-purpose employees, and with a limited range of ISD functions. The next higher levels (the District and the Province in Cambodia, and the sub-District and the District in Bangladesh), where line department staff are deployed, are deconcentrated. While in Cambodia there is a political drive to move ahead with a decentralization agenda within this framework, in Bangladesh there is little movie and substantial opposition to reviving the role of LGs.
• In Nepal, there are legal provisions for elected LGs at both the lower Village and the higher District levels (but these are currently suspended due to the insurgency problems); both levels are entrusted with substantial ISD functions, although at District level these are "shared" with the deconcentrated line departments. The policy is to move towards integrating these departments and their funding within the elected LGs, but implementation is subject to reinstate of the elected bodies.

Opportunities & Constraints

Common opportunities to promote more effective ISD through rural LGs are largely encapsulated in a general formal commitment to diacritic decentralization, such that LGs are expected to provide public goods and services in an accountable, transparent and participatory way, with their own and other resources. They are also reflected in the very considerable organizational resource that LGs represent in rural areas for more effective ISD, albeit a largely under-used one.

Common constraints to this include - among others - a typically very weak local fiscal base (exacerbated by low levels of rural development), inadequate or even perverse centre-local fiscal transfer arrangements, ambiguities in expenditure assignments (both between different LG tiers and between LGs and central government), varying degrees of confusion over sectoral devolution and responsibilities, and sometimes inadequate provision for transparency (exacerbated by the general weakness of the media in rural areas).

LDP INNOVATIONS

The four LDPs have piloted a range of innovations aimed at improving the provision of public goods and services by rural local government. These have aimed to take into account the overall and specific policy contexts discussed above.

Financing

Financing innovations have included the following:

• the allocation of block grants aimed at providing LGs directly with the funds to finance development expenditure, and which are allocated according to a clear and known formula, calibrated to reflect relative poverty and fiscal need;
• synchronization of these block grants with the LG budgeting timetable to provide the discipline of the "hard budget constraint" when local priorities are set;
• a "cascading" approach whereby higher level LGs, in multi-tier systems, must pass down the larger share of their grants to lower level LGs;
• providing LGs with considerable discretion over the use of such block grants so as to foster accountability and achieve allocative efficiencies, and also to encourage local participation;
• the establishment of performance-based funding mechanisms that (a) determine whether LGs access their block grants and (b) increase or decrease allocations depending on previous performance;
• earmarking proportions of block grants for specific purposes, such as investments which benefit the poor or scheduled caste groups;
• support for local revenue collection efforts.

Planning & budgeting

Here, LDP-piloted innovations have included:

• linking planning to LG budgeting, thus making both exercises more meaningful;
• testing out cost effective participatory and inclusive planning methodologies;
• devising institutional arrangements to help bridge the LG-community gap, and the LG-line department gap;
• improving the local planning process by ensuring that it goes beyond participatory needs assessment and takes into account more "technical" aspects, such as appraisal, design and costing, etc.

Implantation ("production")

These have included:

• working within the regulatory framework but adapting procreant procedures to the LG context - with flexibility to take into account rural realities;
• where private procreant is required, devising simplified procedures and formats for competitive tendering, etc.
• setting up community-based and other local committees to provide day-to-day oversight of micro-project implantation;
• where necessary, upgrading contractor and engineering skills in the private sector;
• providing a small funding window in the block grant to facilitate LG access to technical support;
• distinguishing realistic operations & maintenance roles of communities and LGs for different types of investment, and integrating operations and maintenance considerations into planning procedures.

Capacity-building

Several innovations have been introduced in these areas which underpin all the others outlined above. To help LGs and other stakeholders take full advantage of innovations and correctly use planning and other procedures, LDPs have:

• developed & provided basic training packages to a range of actors;
• developed daddies mechanisms for capacity-building, thus placing LGs firmly in the "driving seat";
• instituted local support teams, comprising local residents, local arrangements to extend and support participatory planning & monitoring of ISD;
• more generally and most importantly, fostered real-time capacity-building by "doing", by providing local stakeholders with the opportunity to plan, finance and deliver public goods and services.

Accountability

Innovations to foster greater accountability have included:

• providing local citizens with access to information about LG resources and decisions through comprehensive communications strategies;
• introducing annual self-assessments of LG performance;
• encouraging greater public insolvent in planning, budgeting and monitoring of ISD through participatory planning methods and related institutional arrangements.

OUTCOMES ACHIEVED

Parts 4 and 5 trace the outcomes documented under each of the categories of innovation highlighted above. These can be consolidated under three broad headings:

Pro-poor outcomes

It is beyond the scope of the case studies to undertake in-depth assessments of poverty impact of the ISD innovations under review. The best that can be done therefore is to examine some "proxy measures" which indicate the probable poverty impact.

The outcomes of the local planning process - which in turn was enabled by the financing innovations - introduced by LDPs have included the following:

• A much wider consultation of significant numbers of local citizens, thus providing significant opportunities for the poor to voice thIs elves, express their needs and influence budget allocations (a very practical form of "Improvement");
• The great preponderance of expenditure outcomes has been on those sorts of ISD-related investments which are inherently more likely to benefit the poor than the non-poor: basic primary access, health, education, and water supply facilities. Furthermore, these are facilities which it is inherently difficult - and unappealing - for "elites" to capture.

Implantation outcomes have been of benefit to the poor in a number of ways:

• an apparently greater cost effectiveness, better quality and greater timeliness in ISD implantation by LGs;
• improved access to appropriately located infrastructure, largely due to local oversight of implantation and implication in the planning process;
• sustainable benefits through adequate O&M arrangements in most (but not all) cases.

Institutional outcomes

The incentives Imbedded in the performance-linked funding innovations - allied with more direct capacity support - have led to the following sorts of outcomes:

• Greater compliance by LGs with legal and regulatory provisions regarding the conduct of LG affairs and financial manage and accountability;
• Greater effort by LGs to adopt institutional arrangements and procedures for consultation, communication and transparency to the public;
• Overall, more effective accountability to the public, and to central government.

The key role of information in fostering greater accountability has been clearly shown. Informed citizens have a greater chance of calling their elected representatives to account and of demanding better or more appropriate services.

In terms of finance modalities, whilst outcomes from LG funding pilots have been largely positive - dogtrotting the advantages of formula-driven, discretionary block grants and performance-based assessments/incentives - LDP activities aimed at improving LG own-source revenues have yielded meagre results so far.

Policy outcomes & replicability

Ultimately, the replicability of the innovations in policy and practice discussed above depends on their being adopted as national policy. Of the four cases studied, all have had some impact on policy, although this has been greatest in Uganda and Cambodia. Two sorts of policy impact are distinguished according to two types of policy environment:

• Where there is no clear policy or political drive to decentralize ISD functions and resources to LGs (as in Bangladesh now, and indeed as in Cambodia in the mid 1990s) the role of a pilot experiment such as an LDP can be to simply show what can be done, to dist the usual "weak local capacity" allegations, and generally to better arm the advocates for change and to inform national debate.
• In countries where there is a policy drive to decentralization (such as Uganda, in Cambodia now, or in Nepal until the troubles) the role of a piloting experiment can be more to test " how to …" policy options and procedures in direct partnership with national policy-making bodies.

A positive policy outcome appears to be depend on the quality of the innovations thIs elves and especially how well they are dove-tailed into the statutory institutional framework. It also requires close consultation with national and LG partners and a willingness to allow this to manage the process thIs elves. Finally, however good results and M&E systems may be, positive outcomes do not speak for thIs elves, and there is need for more pro-active communication and advocacy with policy-makers.

LESSONS LEARNED AND FUTURE CHALLENGES

Overall Role of Local Governments in Rural Infrastructure & Service Delivery

A first lesson is that the challenges in promoting more effective ISD through rural LGs are indeed much greater than in the urban LG context:

• Promoting more effective and inclusive demand through participatory planning & budgeting is harder because of much higher transaction costs (typically sparse populations, poor communications, etc), lower education levels, weaker "horizontal" associational activity, greater social stratification and dominance of patronage relations, etc.
• Promoting more effective and responsive ISD supply is harder because of typically weaker human resources, much lower fiscal resources, the typical institutional divide between elected LGs and the sector departments and their staff and budgets, and a much weaker private and NGO sector.

But a second and much more encouraging general lesson to be learned from the case studies is that - despite these inherent contextual problems - given the right circumstances and appropriate support, rural local governments can deliver useful public goods and services, contributing to poverty reduction.

Their performance in this should not be measured against ideal standards of effectiveness, inclusiveness or probity, but against the performance of alternative institutional options for ISD (central government agencies, NGOs, project teams). The evidence of the case studies is that rural LGs perform quite as well as these other arrangements- indeed considerably better in some cases - given a conducive framework of policies and procedures, and some minimum level of funding.

All that said, improving ISD through LGs can only be one component of a much broader strategy for rural poverty reduction: poverty reduction requires many other things than improved ISD, important as that is (appropriate macro-economic, private sector & environmental policies, microfinance, etc); and many key pro-poor ISD functions thIs elves must be retained by central or higher levels of government (public health, vaccination, crop protection, agricultural research, etc.).

Sector-specific issues

Looking at the role of LGs in specific sectoral ISD activities, several lessons Eire.

Firstly, there are two lessons regarding those types of ISD which it is just simpler for rural LGs - especially the lower level LGs - to manage:

• "User Community". Rural LGs are best able both to plan delivery and to ensure O&M arrangements where investments serve a small, definable community of users. This is simply because in such cases it is much easier to organize "user communities" for implantation and long term manage. Examples are tubewells, small irrigation shIes, village tracks. (These are in contrast to investments such as inter-village or inter-Commune roads, or hostels serving pupils from a wide area, which are inherently more problematic).
• "Vertical Linkages". The sectors which it is simplest for rural LGs to manage are those where issues of "vertical integration" are minimal - i.e. where there is little reliance on higher levels of local government or on deconcentrated line departments for "co-production" (recruitment and deployment of staff, or other recurrent inputs, or technical backstopping). Examples of "stand-alone" investments are rural water supplies, tracks & trails, repairs to existing schools or clinics, etc. By contrast, new health facilities pose considerable challenges of vertical linkage.

The broader lesson, however, is that devising a more comprehensive and appropriate policy, institutional and financing framework for decentralization of each of the key ISD sectors - and tailored to the institutional topography of each of the countries in question - raisins a major challenge and one where piloting, research and policy guidance is badly needed.

Financing

In the area of financing, key lessons include:

• pre-determined and predictable allocations make the planning process meaningful;
• providing LGs with discretionary powers over the use of capital expenditure can lead to positive outcomes and is not at all a recipe for disaster as is often suggested (provided it is accompanied by clear rules and incentives);
• performance-based incentives can lead to significant improvement in the quality of LG processes;
• block grant funding approach also encourages LGs and communities to focus efforts on mobilizing their own resources to complement this amount; conversely, it tends to switch the focus of LG politicians away from the unhealthy lobbying of central politicians and officials for extra funds.
• LG own-source revenues are vital (not only as a more discretionary funding source but to enhance the legitimacy of the LG as local polity), but rI ain a major prolix area in rural areas, given the inherent weakness of the rural fiscal base.

Planning & Budgeting

Lessons for planning and budgeting include:

• Without tying planning to LG budgeting, and ensuring access to some degree of certain and discretionary funding, there is little chance of encouraging serious participatory planning;
• Participatory approaches to LG planning and budgeting are both helpful and feasible; and LG politicians may see their own interest in promoting this. But care must be taken to introduce approaches which are cost-effective and sustainable;
• Participatory processes should not simply be viewed against ideal standards. It must be recognized that all organizational activity tends to be dominated by a few (the "elite", which at times may also be a progressive one). The key is to ensure transparency and checks.
• Promoting participatory planning is about much more than "PRA tools", useful as some of these can be. Crucially, attention must also be paid to developing the LG institutional framework itself, to ensure more effective community-council interface at village or ward level.
• Planning must also go beyond the expression of needs and should include simple "technical tools" to help appraise, cost and rank very different proposal in a transparent and consistent manner. There is also scope for "tools" to promote gender and poverty targeting, but they need to be devised very clearly and applied carefully.
• Planning procedures based solely on "community proposals" tend to be biased towards very small, local initiatives, and to neglect more strategic and public forms of ISD. Procedures must therefore also allow for inclusion of proposals from higher institutional levels, to capture the broader public interest.
• The politics of LG resource allocation also tends to be biased: towards "equal" shares for all councilors or villages, at least initially; towards social facilities yielding broad-based benefits, neglecting economic investments typically favouring a few; and towards initiatives with short term benefits. Such biases can be partly corrected through performance incentives - but this may all too easily lead into unreasonable "second guessing" of local choices.

Implementation

Lessons here are principally:

• In general, this is an aspect of ISD which is very frequently underplayed and neglected. Policy initiatives and projects all too often assume that good planning alone will lead to better ISD - but this is obviously not the case;
• There is need for flexible procurement arrangements, allowing for a range of options, in order to take into account the specificities of sometimes rIte, rural areas (as in Nepal);
• While it is often claimed that ISD production can be fully outsourced to the private sector this is not always so easy in rural areas. Frequently, rural LGs face the constraint of very weak and uncompetitive local private sectors; and, politically, private outsourcing by LGs can be problematic, where this is associated with corruption: i.e. outsourcing may simply be risky, objectively, or LGs themselves may be reluctant for fear of being suspected of corruption.
• That said, simple skills upgrading and introduction of simple tools and procedures for more transparent and competitive procreant can greatly facilitate use of private capacities.
• O&M raisins, in many cases, a bugbear. This points to the need to ensure a clearer distinction up front on the sorts of facility which can reasonably be managed and maintained at use community level, and which must be looked after by LGs or line departments; and, especially for the latter, a better articulation between recurrent and capital expenditure.

Accountability

Some lessons Eire on the scope for promoting LG accountability for ISD:

• The main focus of the innovations has been on enhancing downward accountability. Indeed the experience suggests that more open and consultative planning and budgeting procedures, allied with more systematic provision of information to citizens, can lead to improvement at relatively low cost. But challenges rI ain in: (a) extending this to the very poor and other marginal and more geographically isolated communities, and extending accountability to include other areas of LG business, beyond ISD planning; (b) strengthening the feedback between individual councilors and constituents.
• Strengthening the horizontal accountability of local civil servants to elected LGs is problematic, except in situations such as the Ugandan districts where the former are under full control of the latter. Indeed the term "horizontal" is often a misnomer since civil servants are typically deployed to a level above that of the elected rural LGs in many cases (Cambodia illustrates one interesting approach to this).
• Upward accountability of local government to the centre has been encouraged through the establishment of upward reporting systems, and compliance with central policy and procedures has been encouraged through the performance-linked funding incentives. This too is an area where much more needs to be done, but where there are constraints inherent in the capacities of Ministries of Local Government or of the Interior - or equivalent central departments - as regards supervision of local government. One particularly problematic issue is the typically very weak capacity to ensure regular external audit - and audit follow up - of hundreds or even thousands of rural LGs.

Capacity-Building

Finally some lessons on this cross-cutting theme:

• Generally, the Case Studies have shown that the "weak local capacity" mantra is often exaggerated; or is based on the false presumption that poor LG performance is due simply to inadequate or poorly trained staff. The case studies show that LG performance can be greatly improved - without changes in staffing or educational background - by a combination of learning-by-doing within a more enabling framework of policies, procedures and systems.
• Demand-driven capacity-building (as under the Uganda LDP) can play an important role for optional or location-specific needs, but it cannot replace a supply-driven component, which provides LGs with a menu of essential CB activities.
• In devising sustainable "supply-driven" training and support mechanisms Bangladesh and Uganda have indicated the value of 'cascade' and 'peer' capacity-building through training (based on centrally produced materials), mentoring and/or exchange of experience.

A more general lesson Verging is that LGs are not and can not be "autonomous". If they are to play their proper role in more effective ISD they require a whole tissue of inter-relations with higher levels of sub-national government, and with central government in order to ensure access to technical support and fiscal transfers, coordination of planning and budgeting, the upward transmission of "demand" for services, and the downward transmission of policy and guidelines.

Above all, LGs can only prosper with active support and effective monitoring by central government, and by higher levels of sub-national government. Devising these arrangements, however, is a major organizational challenge - given the large numbers of rural LGs typically involved and the turnover of elected representatives, and given typical civil service staffing constraints.