United Nations Capital Development Fund

Corporate Policy Papers

Taking Risks: Background Papers

Decentralized Financing

Roger Shotton

Objects of financing: how discretionary are they?
Scale of funding
Funding allocations
Local revenue and resource mobilization
Conclusion

A central Capital Development Fund (CDF) strategy has been to establish local development funds (LDFs) as test mechanisms or policy models for channeling public funds to local government for rural development activities. These models are designed to be sustained and even replicated on a wider scale by the central government or by other donor agencies. Pursuing and negotiating this strategy in each country has, in every case, led the CDF to conceptualize these mechanisms as one or another variant of a discretionary block capital grant model of fiscal devolution—this was probably inevitable, given the universality of certain basic public finance principles and issues. Nonetheless, a large number of details and variations on this basic model are emerging through practice. They pose programming and policy challenges and are worth highlighting.

Objects of financing: how discretionary are they?

Defining inclusions and exclusions

Two central questions in any LDF always concern the range of activities to be funded from the LDF and the flexibility with which this range is to be defined. Initially, the CDF tended to focus on a fairly restrictive list confined to infrastructure investments in the core local public goods sectors (roads, health, education, water and the like); "recurrent" expenditures were excluded. Two key lessons to date are worth noting.

There are many legitimate candidates for local government funding that are not public goods: small irrigation schemes, NGO-run health clinics and the like. Defining exclusions too rigidly thus undercuts the discretion of local authorities (and so their accountability) as to just what is or is not in the local public interest.

It is not desirable to limit expenditure to capital infrastructure alone (despite the CDF’s own institutional bias to such investments). Early on it became obvious that we also needed to allow for funding of investment preparation costs (site surveys, technical design and the like). Later it became clear that there were other legitimate development expenditures that entail no capital investment: paying an NGO or government department to undertake an AIDS sensitization programme or a womens’ skills training course, for example. The essential feature is that the development activity should be of limited duration, with clear objectives, rather than a routine activity (for example, teachers’ salaries or hospital power costs), which poses sustainability implications.

Inherent limits to the feasible range

But the range of legitimate objects of financing is not fully elastic. There will be certain "natural limits" in any country context, which derive variously from:

The sectoral range of legally mandated or permissible functions of the local authorities being supported, which may be narrow (as with Senegalese communes) or very broad (as with Ugandan district councils).

The technical capacities for planning and implementing at the disposal of the local authorities may be either limited or quite substantial.

The range of rural socio-economic infrastructure problems to be addressed through public expenditure, which may be quite vast (as in Ethiopia) or relatively limited (as in Bangladesh).

Local choice and the limits of discretionary finance mechanisms

One important issue is emerging that links the question of local priorities with discretion. It is becoming clear that the local participatory planning and public choice procedures that are being encouraged will not always identify priorities that match wider national policy goals. Within a given budget constraint a consensual local needs identification and prioritizing process will tend to give priority to proposals with wide local benefit, with a bias towards basic social facilities and services and away from those (typically, economic activities) benefitting smaller segments of the community.

The CDF is therefore now examining the scope for trialling earmarked funds or special funding windows, which simulate specific or conditional grant-funding models, to encourage local planning procedures to give due weight to such activities. (Top)

Scale of funding

The strategy for scaling the level of LDF project funding derives from two considerations:

Local absorptive capacities. The volume of funding should be consistent with local absorptive constraints:

Local government planning, management and implementation capacities—sudden and disproportionate increases in resources will entail an obvious strain on any organization.

The supply capacities of local private operators, NGOs and the like, often quite sparse in rural areas, to satisfy the expressed demand for goods and services.

The recurrent budget resources available to local governments (through transfers and local revenues) for operation and maintenance of the additional public investments, bearing in mind LDF strategies to maximize the scope for direct cost recovery and for local community cost sharing.

Replicability levels of a block grant mechanism. This strategic goal requires that the volume of funds on a per capita annual basis, allocated through the LDF to local governments, be commensurate with the volume that would be available in the longer term for such activities, should government opt for devolution of the relevant parts of the development budget (gaining a commitment to being a key element in the CDF programme and partnership strategy in each country).

As it happens, the typical scale of LDF funds allocated to local governments is usually no more than US$1 per capita annually—the replication of which therefore entails quite modest levels of devolution of the national development budget and of donor followup. However, within individual local government areas these funds may sometimes be allocated on a more intensive per capita basis, in accordance with local planning and targetting priorities towards wards or villages with particular potential or problems.

However, it has become increasingly clear that applying funds scaled down in this way, consistent with the goal of institutionalization, does not allow LDF projects to address the wide range of development problems of a given area, and this action cannot be expected to push the local economy across any dynamic growth threshold.

Put differently, it should be recognized that there is a trade-off between the goal of direct promotion of rural development and that of developing replicable institutional models for local development. (Top)

Funding allocations

Allocations to local governments

Consistent with the block grant approach to financing local development plans, LDF allocations to individual local government units are announced at the start of the annual local planning cycle. The allocations are formulae-based, rather than being left to central discretion, with rules of calculation and access agreed and widely advertised. Negotiating this approach is not always easy, since it certainly undercuts the potential for patronage and power at the centre, but entails two key advantages:

The hard budget constraint. Setting an indicative budget figure for the local planning exercise is critical, inasmuch as it constitutes a clear funding ceiling for local prioritizing, thereby encouraging more serious weighing of alternatives, greater commitment to choices made, greater attention to choosing cost-effective solutions and a greater incentive to mobilize matching local resources. The key is that this allocation must be announced in advance of the local planning and budgeting cycle (rather than midway through, as is unfortunately common in central-local funding arrangements).

Transparency and equity. Annual allocations are derived from population-based formulae, with weighted adjustments to incorporate other development factors (poverty, "backwardness", land area and the like) that reflect national policy priorities and that are dependent on the availability of local data on these factors. This approach ensures that allocations are fair, and are seen to be so, maintaining the legitimacy of the mechanism and that they undercut the tendency of local leaders to lobby the centre for ad hoc discretionary increases, and so it counters patron-client relations and reinforces local downward accountability for proper use of funds. It must be said that these formulae are often critiqued for not providing comprehensive measures of relative welfare or fiscal capacities. The CDF’s position is that more complex criteria are rarely supported by adequate, up-to-date local data, and the criteria invite greater (real or suspected) political interference in their management, thereby undermining local faith in the mechanism.

Access to these allocations is not unconditional. The CDF has been introducing a range of minimum conditions: satisfactorily audited accounts, compliance with agreed participatory planning procedures in previous years and the like.

Allocations to lower tiers

In countries with two or more local government tiers with service provision responsibilities, simple mechanisms for downward allocations (from one tier to the next), which embody the same principles, are being tested (as from districts to subcounties in Uganda).

Allocations to communities

A further issue concerns the principles of fund allocation by local government to very local "communities" (wards, villages), whose participation in the local planning process is being encouraged. Here the choice lies somewhere between two contrasting approaches:

On the one hand, allowing these bodies to "apply" upward, with their proposals judged on their competing technical and socio-economic merits by the local government planning body, with additional scope for more or less healthy political bargaining by the respective area councillors.

On the other, announcing an ex ante funding entitlement to these bodies (for "community activities" with no wider externalities), undercutting the scope for either competition through techno-rational assessment or political bargaining.

The CDF recognizes the pros and cons of each approach and finds it impossible to be universally prescriptive on allocation principles at the local level. The preferred balance between these two extremes will depend on such issues as the nature of the local planning system, the characteristics of local political representation, the cohesiveness of local community bodies and the relative variations in poverty or development potential between areas. (Top)

Local revenue and resource mobilization

There is increasing awareness in the CDF, heightened by criticisms in the MTR, that this is a critical area that has been relatively neglected to date. Raising local government revenues is not only key for the more obvious service sustainability reasons, but also to enhance local government legitimacy in its dealings with the centre.

Unfortunately, and in contrast to, say, improving local planning systems, it is also an area where scope for change is particularly circumscribed by legislative and central policy constraints. The core problem is that the tax base and tax powers given to local government in rural areas of LDCs is usually inherently weak—the base being narrow, stagnant and both costly and politically hard to collect and local collection and enforcement powers being often undercut by central interference or apathy. It is therefore a prime area for "upstream" policy reform discussions.

Direct support to improved local revenue administration

Given this, the best that the CDF can do at the local level is to focus on such modest measures as:

Providing technical support to improving tax collection efficiency, through the updating of tax registers, assessments, billing, collecting and monitoring procedures and the like (an area with great potential for computerization).

Widening the base and improving the collection of non-tax revenues (service charges, market lease fees, and the like) and using LDF-funded investments (markets, water supply systems) as pilot cases for innovation.

Incentives in the financing mechanism

More recently, the CDF has attempted to add in incentives for increased local fiscal effort in the funding allocation mechanisms described above. These range from limited incentive adjustments to reward past improvements in local revenue mobilization performance (for example, recognizing a 10% increase in tax collection performance by a supplement to next year’s block grant allocation) to direct "matching formulae", whereby in any given year local revenue commitments are matched by block grant funds according to some agreed gearing ratio.

The CDF has learned that these approaches are certainly not without problems, and they pose questions of equity between local governments, which may face very different degrees of difficulty in increasing levels of fiscal performance, often quite unrelated to questions of "incentive". Thus different rates of poll tax collection in Tanzania are claimed to be primarily a reflection of variations in local crop harvest reliability, in the difficulty in recruiting collectors or in the supportiveness of local party leaders, all factors outside the control of even the most enthusiastic tax-levying local authority.

However, the CDF believes that the key importance of raising local revenue generation does justify carefully monitored experiments in block grant incentive mechanisms in this area.

Community self-help contributions

A further important "matching" principle, adopted in all LDFs, lies in cost-sharing arrangements for individual schemes and micro-projects, by which local community bodies are required to fund some agreed share of the investment costs (in cash or kind) and to commit to long-term operation and maintenance arrangements. However, the CDF experience is also that this principle must be applied with care to various equity considerations, by:

• Distinguishing types of schemes where self-help is simply not appropriate (for example, schemes with wider, diffuse benefit such as secondary schools or district roads) to avoid benefiting the many (free riders) at the expense of the few.

• Identifying significantly poorer areas with fewer potentially matching cash or labour resources by modifying the levels of expected contributions.

• Recognizing often that there are no comparable self-help expectations in urban areas, and that this may lead to understandable resentment by rural communities.

It is also more problematic to raise local contributions for schemes implemented by contract to private operators (for example, borehole and pump installations).

The taxation-accountability link

Attention is now being paid in many quarters to the potential link between local taxation and local government accountability (no representation without taxation). This will probably be an area that will attract increasing interest and one that the CDF is keen to operationalize and monitor. It must be said, however, that the real force of this accountability linkage only applies where local governments are allowed real discretion in taxation: that is, choosing their tax sources and setting their tax rates, rather than having these both dictated by the centre, as is so often the case.

Funding channels and financial management and control

Consistent with the block grant model strategy, the CDF is now paying increasing attention to channeling funds to local authorities through existing government channels, rather than creating short cuts that seem expeditious but which bypass the government’s own accountability controls for use of public funds. The need for this was highlighted by the MTR and was strongly endorsed by the Arusha workshop.

Fund allocation, oversight and channeling arrangements

Even a formulae-based funding mechanism requires an authority to manage the process, judge whether local authority performance complies with agreed guidelines, interpret the data, authorize funds release and undertake monitoring and inspection to ensure "upward accountability".

The choice of allocation and oversight authority varies considerably between LDFs but has generally been the central Ministry of the Interior or local government (as in Uganda or Malawi) or deconcentrated regional authorities (regional administrative secretary in Tanzania, provincial governor in Cambodia or district commissioner in Bangladesh). To widen the scope for accountability the CDF ensures that this authority person acts in consultation with a national or regional committee of other stakeholders (usually key line ministries, but where possible also NGOs, local authority association representatives and the like), which the person chairs. Project support units—where they have been set up—act as technical secretariats to these authorities or committees, but they no longer have allocating or overseeing powers.

Funds are channeled from UNCDF/UNDP to bank accounts held by these authorities, who, in turn, release them to local government bank accounts in periodic tranches in accordance with the approved local development plan and the capital funding programme and on the basis of reported expenditures.

The CDF has learned that institutionalizing these procedures poses various challenges:

In countries with only embryonic local authorities and decentralized financing arrangements (as in Cambodia or Ethiopia) many arrangements have to be devised from scratch—negotiating may take considerable time.

In others, with long-standing decentralized institutions (Bangladesh or Uganda), there may be serious rigidities in existing procedures and practices and, further, there are often inconsistencies between local government enabling legislation and more restrictive existing circulars and regulations.

In any event, full compliance with government procedures for approval of local government financing is rarely practical or desirable within an LDF, since this often entails lengthy interaction between the Ministry of Finance, Parliament, and the Local Government Ministry, and very often delayed transfer to local government bank accounts.

Generally, the CDF recognizes that there is often an inherent trade-off between full mainstreaming and maintaining project credibility at the local level, and compromises must be negotiated in each case.

Financial management and control

Here too, emphasis is on using existing arrangements for local government financial management, budgeting, accounting and controlling and for procurement—while supporting their proper implementation through attention to training, preparing and disseminating manuals.

The CDF aim is to move away from the all-too-common ex ante financial controls (so that local government budgets often have to be approved and subject to arbitrary change by central officials who are ill-placed to judge), towards a system of ex post control and auditing of local government by the centre—ultimately leading to the monitoring and audit of performance rather than of budgets and expenditures. But the CDF experience is that this is often a distant reality, requiring considerable prior improvements in local management and accounting practice, much strengthened central inspection and audit capabilities and clearer procedural and legal frameworks and codes of conduct at all levels.

It is increasingly clear that there is little chance of encouraging greater fiscal devolution without much greater financial transparency and accountability of local government, to counter the pervasive mistrust of local politicians and officials. It is fair to say that these are areas that have been relatively neglected by the CDF to date, but to which we must now turn. (Top)

Conclusion

In conclusion, a number of key issues and challenges must be addressed by the UNCDF:

Policy and strategy. Greater effort, along with the UNDP, to alert governments to the inefficient constraints on local tax revenue mobilization and to convert both governments and other donors to the merits of decentralized financing arrangements, especially in countries where the bulk of the national development budget is donor-funded—this is key to any chance of replicability.

Programming and project design. Greater attention to efforts to improve local revenue administration and to strengthen local government financial management and accountability—this is key to enhancing the legitimacy of local government in the eyes of the public and central government.

Monitoring and research. Greater attention to the effectiveness of different local government revenue improvement strategies and a clearer understanding of the scope and limits of unconditional block grant mechanisms, and some trial of conditional grant or earmarked or matching funding approaches to address needs that may not be adequately expressed through local public planning procedures (such as the "green window" proposal for environmental activities in Malawi)—this is key to broadening the role of local government in rural development.