Public institutions Institutions

Posted February 1998

Food and Agriculture Organization of the United Nations United Nations Capital Development FundInternational Fund for Agricultural DevelopmentGerman Agency for Technical CooperationSwiss Agency for Development and CooperationWorld Bank

16-18 December 1997
Technical Consultation on Decentralization

Supporting Rural Local Governments in Practice: Issues for Consideration

by Laura Kullenberg
Roger Shotton
Leonardo Romeo
United Nations Capital Development Fund


These comments are based on the practical experience of a donor involved in a number of decentralization pilot programmes in a wide range of countries. The thoughts are forwarded from the point of view of a practicioner, to share with this conference some of the challenges, questions and issues that have percolated up through our field experiences in the last several years. We don't pretend to have conclusive answers, but do offer emerging questions from the field operations which we hope that the working groups can discuss and consider throughout the week.

Over the last 5 years UNCDF has designed programmes which provide capital budgets and technical support to local governments and de-concentrated state authorities in over 15 LDC countries. These programmes are now in various stages of implementation, providing a continuous field information which we have used to modify our approach, model and assumptions. While the design of these specific UNCDF projects may be different from other donors-and there are critical differences in many cases- we do think most of the issues emerging will be common to many donors and governments engaged in supporting decentralization policy and programmes.

These projects, called Local Development Fund (LDFs) are of a pilot nature, averaging from $3m-$9m over 3-5 years. They can be characterized as practical experiments in decentralized financing and planning of rural development. They provide technical assistance and direct capital resources in the form of budgetary planning ceilings to rural authorities at district Council or Commune level, and transfer typically 1-5 US$/capita. These programs often precede implementation of national decentralization programmes or investments by larger donors such as the World Bank. Critical to our ability to fund such pilot programs is UNCDF\s mandate to provide capital assistance in the form of small grants, which also facilitates the direct channeling of resources to local levels.

The intent is to promote decentralised planning and financing of rural development AND institution building at the local level.. UNCDF's particular focus has been to enhance the capacity of local government to finance and manage capital budgets with a strong emphasis on inclusive participatory planning at the local level. Many agencies are doing similar things. A key feature of these programmes is that we work within and not parallel to the national/local planning system for public investments.

It is important to underscore the fact that given UNCDF's central mandate is poverty reduction, there is a basic assumption underlying the focus on governance; that decentralisation improves governance and improved governance has an impact on poverty reduction. We agree with the concerns that the jury is still out on the strength of this link.

Programme strategy and tensions

UNCDF shares the view of many donors who see improved local governance as the key to rural development. But more specifically we believe in direct support to decentralised "Local Government" institutions. We have found that this is a critical but often neglected entry point, and, somehow surprisingly, one which is often controversial:
  1. it often differs from the approach taken by other donors (EU, World Bank Social Action Funds)
  2. it raises at times quite strong contrary views from some of our development partners, particularly in the NGO community, who raise questions about the legitimacy of and accountability of local governments to the rural poor
For us this strategy is often problematic because of the chronic tension between our twin program objectives of expeditious delivery of goods and services for poverty reduction, on the one hand, and of local institution building, on the other .

Political dimensions

In designing our programs we have found it necessary to use a methodological framework for assessing "local state institutions", which goes beyond the "de-concentration/devolution" dichotomy. Specifically we need to look at 2 dimensions : (1) the local institutional topography of the state and (2) the political commitment to transfer power/resources/autonomy etc. to local institutions. While the latter gets a lot of attention, the significance of the former (for the scope and limits of decentralization) is often neglected. To give an example, contrast these two extremes:
  1. a Ugandan District council, governing 500,000 persons or more, with several hundred employees (many at graduate level) and a budget of several million dollars, clear statutory fiscal & legislative powers and responsibilities, etc.
  2. A Malian commune, governing 10-15,000 persons, with one or two low grade employees, a budget of several thousand dollars, with a quite limited mandate, and fiscal autonomy constrained by the "unicite de caisse" principle
Clearly , while both countries rank high on political commitment to devolution, the actual scope for decentralization of functions and fiscal responsibilities must vary dramatically, as must the expected benefits of decentralization. By illustration we could refer to variations in service delivery and political accountability.

Implications for decentralized service/infrastructure delivery

  1. The range of infrastructure/service delivery functions that can be decentralized, which is more limited (both "horizontally" and "vertically") where the planning area is smaller, the mandated responsibilities fewer/weaker and the available technical capacities much more limited.
  2. The extent to which investment planning can be horizontally integrated, and integrated with the budgeting process, which is more limited where local line departments are not under the control of the local authority.

Implications for political accountability

It is common to refer to two sorts of accountability, critical for effective local government: (i) accountability of local civil servants to elected local leaders and, and (ii) of local elected representatives to their constituents. We have found that the challenges vary:
  1. the main challenge in the Malian context (small-scale commune, with most technical staff at the higher Departmental level) is to devise mechanisms by which Communes can access technical services of the de-concentrated state, and hold bureaucrats to task for service provision; by contrast
  2. a bigger challenge in Ugandan context (where bureaucrats are under the control of District authorities) is to secure the accountability of the more remote district representatives to local constituents.
We would add a third sort of accountability, that of lower level local authorities to higher levels. This has several dimensions that refer to the obligation to account for the use of resources mainly coming through fiscal transfers and to comply with higher level policy priorities. Obviously the need for this kind of accountability is the greater, the smaller is the local government unit.

Local planning and politics

A natural tension exists between local planning and local politics. To the extent that planning practices attempt to introduce better informed, more transparent and more rational decision-making, they not only specifically conflict with the practice of patronage, but also, and in more neutral/general terms, reduce the discretion of local leaders, constrain the "mode" and the "tempo" of their decisions, in one word limit the "autonomy" of politics. Paradoxical as it may seem, since decentralized planning in fact reduces local political discretion, it may have been introduced in the past as a way to avoid or limit real political decentralization.

But there is a tension also between local planning and participation. When participatory mechanisms are incorporated to planning procedure, the result is often a surge in demand for social infrastructure and welfare subsidies, a spreading of scarce resources in numerous small projects, the inability of local politicians to "control" the demand they have unleashed (often reluctantly, precisely because they knew how damaging to their status would be any successive attempt to control).

Program implementation

Our experience highlights some of the challenges of introducing participation in local planning. This is an issue of both techniques and institutions.

In encouraging local participation we face real difficulties in devising mechanisms which are sustainable and replicable, institutionally and financially. NGOs can afford to mount intensive PRA exercises in limited areas, but these can rarely be imitated on any scale. In general we find that techniques for community-level participatory projects/actions identification , have not been adequately related to the local public sector planning process.

While a lot of attention has been given to grassroots projects identification, not nearly the same has been given to increasing the capacity of local planning authorities for projects selection, appraisal and preparation. Equally deficient has been the effort to combine the process of grassroots proposals identification and selection with that of submitting local planners ideas to grassroots screening and feedback.

Less attention seems to have been given to building the institutions for participation in public affairs than to participation in project cycles. Yet the sustainability of the improved planning procedures, is highly dependent on the set up and functioning of a local planning authority which combines the legal responsibility of local authorities with the control functions of civil society organizations. The strategies to achieve this combination, again vary by context, particularly on the representativeness of the local authorities (elected versus nominated) and the relations between them and the technical agencies.

Institutional capacity

Building local capacity is obviously key to the success of any national decentralization strategy. Given the systemic nature of this capacity, the magnitude of the task may often appear overwhelming. What our experience suggests the need for an approach which is : (a) institutional, (b) focused , and (c) long-term

By institutional we mean that the main responsibility for building local capacity, must be assumed by higher levels of governments and mandated by a specific national policy. Two issues emerge in this connection. National/regional/provincial staff may not have the capacity to build capacity and this may call for a wider range of actors to be involved in this effort. NGOs and national associations of local authorities have a potentially important role to play in this respect. Also government staff leading the capacity building effort may lack the appropriate incentives if the cost and financing of the extension which may initially be covered by external financing is not taken over by national governments in the long run.

By focused we mean that, despite the systemic nature of the local capacity to be built, a realistic strategy should identify specific entry points and limit itself , initially, to building the local capacity required in connection with these entry points. In our programs we focus on capacity for local capital programming and therefore develop the local planning, project and financial management capacities.

By long term we mean that the current 3-5 years frame of the programs we support, is clearly inadequate. Capacities for local planning and project implementation are dependent on changes in attitudes and social interaction and negotiation abilities, as much as on acquired skills. This makes it impossible to substitute intensive training for actual on-the-job experience.

Capacity building strategies are also obviously very dependent on specific contexts. One critical issue is the nature and stability of the local civil service. If municipal employees are hired and fired at any changes in local political representatives, (as in Nicaragua) the efforts to build local capacity are obviously greatly frustrated. Equally problematic is the situation where all local civil servants are centrally deployed, giving rise to higher turnover and lower degrees of local loyalty.

Also, somehow paradoxically, capacity building and introduction of innovative practices, in countries with long traditions of local government (e.g. Bangladesh) may face the extra challenge of fighting entrenched procedures and outmoded practices.

Reform of institutions and the role of the State

We recognise the provision/production distinction and try to design/implement accordingly. But we also have to deal with the limitations:

In many countries, despite apparently favorable central government policy on privatisation, at local government level there is often resistance: fear by local civil servants that the politicians use this for cronies, and fear by politicians at implications for laying off government staff In many rural areas (Ethiopia, N. Uganda) the only source of supply capacity (for engineering, technical, construction, management, etc. functions) lies in NGOs or in government departments (the private sector often finding higher returns in urban areas). This poses challenges:

  1. Devising ways of accessing government agency services (by local governments or communities) on as near as possible a "contractual" basis - though government regulations often prohibit this and allow only payment of staff travel costs
  2. Devising arrangements and incentives to bring in NGOs as service providers - when these NGOs are often quite skeptical about collaboration with local governments

Fiscal decentralization, regional equity, rural growth

The extent & pattern of feasible fiscal decentralisation also depends on the local institutional topography of the state. Thus, where there are only development committees, rather than corporate local government bodies, (e.g. as in Bangladesh at Thana level, or in Malawi) - there is only scope to decentralize development budget funds for local discretionary inter-sectoral allocation; recurrent budgets are inevitably channeled down through the line agencies, so undercutting the scope for local inter-sectoral management .

Since the development budget in many countries where we work is funded 80-90% by donors, this raises major issues of donor coordination and commitment to supporting decentralization objectives, and of reconciling government and the various donor financial monitoring procedures and expectations. In this respect we believe "working models" should be developed to provide both local authorities and other donors interested in decentralized financing, with the tools for integration of different flows of funds, in support of a unified local planning process, whilst satisfying donor accountability requirements. It is worth noting in this connection that As a donor channeling funds to local authorities (and trying to pilot better IGFT systems) we are often caught between two objectives: (1) trying to work through official central-local fiscal transfer mechanisms but (2) aiming for expeditious local delivery.

For fiscal decentralization to promote rural growth requires that local fiscal allocations entail a bias to "efficient" rather than to "fair/equitable" allocation of resources, and so a shift from the natural tendency to allow equal shares for all in the local fiscal resource pie. Our experience is that the feasibility for such a shift again depends on a number of factors:

  1. obviously, on local planning capacities.
  2. but also on the legitimacy of local authorities and their ability to make and defend decisions that may be perceived as "unfair" by some constituents;
  3. and on faith that resources will also be coming down in the future as well, and so that everybody will get their turn sooner or later (this of course poses a problem for a short duration donor funded project, if there is no prospect of longer term funding of the same magnitude)
These, in turn, also to some extent depend on the institutional topography context (eg in some countries local government has only mandate for basic social infrastructure, or is less equipped with the sort of technical planning capacity to allow or to defend efficiency-focused planning decisions - eg contrast Mali and Uganda)

Design of Inter-Governmental Fiscal Transfers (IGFT)

A key feature that distinguishes UNCDF supported programs, from Micro-projects or Social Funds programs, is the combination of decentralized planning with access to financing in the form of an Indicative Planning Figure (IPF): an annual or multi-annual allocation known and agreed ex-ante. The virtue of the IPF is:
  1. that it constitutes a hard budget constraints essential for a meaningful local planning exercise and priority setting,
  2. that it encourages mobilization of local matching resources and can itself be used to leverage additional external funding,
  3. that it undercuts perverse local patron-client relationships which often characterize relations between different levels of government, and consequently
  4. shifts the focus of local leaders' accountability from seeking higher officials favors to obtaining local constituents consensus.

Decentralization and provision of rural infrastructure and services

The range and complexity of infrastructure provision which can be decentralised depends, once again, on local government context. The mandate, the planning area, the relationship between local authority and line departments, all make a difference in this respect. For example, Ugandan DCs can be entrusted with a gamut including hospitals, markets, technical schools, etc., while Bangladesh Union councils or Vietnam Communes, can be entrusted with little more than the lowest tier of rural roads, tubewells and drainage, small, irrigation schemes, primary schools and healthposts.

While , in general , the distribution of infrastructure provision responsibilities should be governed by the subsidiarity principle, the application of this principle often faces some constraints. Even the most local type of social infrastructure may require some form of joint provision in which higher levels of government typically assume responsibility for financing of operating costs. The rationale for this is often strengthened by central authorities concerns with the maintenance of minimum service standards, as happened in Vietnam where provincial authorities re-centralized the management of health and education, previously assigned to districts, to secure uniform levels of service.

Decentralization and natural resources management

While the principles of subsidiarity generally requires community management of local natural resources, this does not mean that there is no role for local governments.

First, basic local transport and market infrastructure, and improved extension and support services are critical. These of course are all preeminently activities undertaken by the state, and may often be the mandate of local authorities

Second, supporting community NRM institutions also requires (decentralised) state financing of community activities, such as community irrigation schemes, watershed protection measures, etc.. In preparing projects of this sort (Malawi, Ethiopia, ..) two key issues which emerge are (a) how to adapt local participatory planning procedures to ensure that such proposals are adequately fed into the process and (b) what are the appropriate terms of financing of such support, so as not to discourage local initiative, nor to undercut emerging micro-finance institutions.

Third , and more problematic, improved NRM often requires greater access to credit/micro-finance. Clearly local authorities have no comparative advantage in this area and should not be involved in direct management of micro-finance facilities. But, in many rural areas there are simply no micro-finance institutions and there is often great pressure to relax this principle. An intermediate solution adopted in some countries (Cambodia, Ethiopia ..) is for local authorities to provide grant injections of loan capital as revolving funds for community savings/credit groups (coupled with an extension service), to help develop embryonic village banks. More generally, a major challenge lies in devising arrangements by which local authorities can provide a more direct support to farm and non-farm enterprises.

Finally, because of the frequent association of natural resources degradation with difficulties of local communities in enforcing common property management rules, local governments can also play an important supporting role, through the enacting and enforcing of local regulations on forest and pasture uses, etc.


This is not meant to be only a list of worries about limits of decentralization and these types of support programmes. Yes, the jury is still out on the impact these new decentralization programmes and policies can have on poverty, rural development, governance. It is important to track these experiences carefully, to establish a strong link to policy dialogue, and to bring findings and concerns to such fora as this one.

But we are also quite excited about the potential of these programmes. Already many possibilities have presented themselves. For example, there is still not very much being done in the area of performance based allocation systems-moving beyond simple grant financing to deliver services to looking at how transfers from the outside can improve mobilisation and use of local resources. Another area of interest is how and under what conditions local governments should be assisted to move beyond provision of social infrastructure only into other important areas such as natural resource development, promotion of local economic development. In these cases donors can provide a useful service through pilot programmes that can help underwrite the risk of partner governments.

SD Homepage Back to Top FAO Homepage