U4 Helpdesk Query

U4 helpdesk reply

This note consists of four sections:

1. Budget support and corruption
   The first introductory section defines budget support, and briefly reviews the main reasons for the increased donor emphasis on this type of aid. The rationale for budget support is important to keep in mind, as one considers alternative types of aid in situations where the effectiveness of budget support is challenged.
   
2. How does corruption affect the effectiveness of budget support?
   The second section examines how budget support might be affected by a high level of corruption in recipient countries. This section also refers to efforts to include assessments of corruption in more general evaluations of fiduciary risk.
   
3. Does budget support increase corruption? Section three discusses whether there is a risk that budget support might increase the level of corruption in a partner country.
4. Recommendations and references
   

Many development agencies are currently increasing the share of aid that is given as budget support. DFID spends about 15% of its current budget on budget support, a proportion that is likely to increase in the coming years. The World Bank spent about 50% of their total assistance on budget support in 2002, up from about 10% in the 1980s. The question of which particular challenges budget support raises in terms of corruption, is therefore a timely and important one. This note reviews some of the main lessons that can be drawn from recent experiences with budget support, and also draws some relevant inferences from research on aid and corruption. Studies that directly address budget support and corruption are scarce. Hence, a number of issues remain unresolved, which this note also reflects. In 2005, the U4 anti-corruption resource centre will introduce a new focus area on Public Finance Management. Under this focus area, an issue paper on corruption and budget support will be developed, which will explore the issues raised here in greater detail.

[top]

Part 1: Budget support

Budget support is financial aid that is given as a direct contribution to a partner government's budget. Budget support can take the form of a general contribution to the overall budget, which is often referred to as general budget support. Or it may be channelled into a specific sector, which is commonly referred to as sector budget support. Budget support can be contrasted with project aid, through which a more closely specified and limited set of activities are funded.

Development agencies give different accounts of the advantages of using budget support rather than project support. Some of the more common reasons for budget support are:

• Aid is fungible: A partner government may redirect funds from one area to another. Attempts by donors to channel funds to specific projects are thus futile, since the final distribution of funds will reflect the preferences of recipient governments.
  
• Local ownership: By giving partner countries discretion on how funds are allocated, partner governments are more committed to the spending priorities set, and aid is therefore used more effectively.
  
• Capacity building: Partner governments expand their capacity to make allocative decisions, and conduct long term financial management. Incentives are provided for both donors and recipients to enhance the effectiveness of government systems and institutions.
  
• Reduced transaction costs: Many of the costs associated with a more piecemeal approach through project aid, are avoided.

See this ODI working paper for a structured and general discussion of the choice of aid instruments.

[top]

Part 2: How does corruption affect the effectiveness of budget support?

A main challenge to budget support is that it requires a relatively well-functioning institutional environment in partner countries, to effectively enhance development. In a weak institutional environment, there is a risk that funds are misallocated or inefficiently employed, commonly referred to as fiduciary risk.

Corruption can be seen as one specific form of misappropriation of funds, where the power of allocation of public funds is used for private gain for the officials involved. This might take the form of embezzlement or bribery, or kickbacks at the implementation stage, such as in procurement processes. Or it might be in the form of favouritism, through which certain persons, groups or constituencies in a country benefit from a disproportionate share of the funds. There are many examples of public funds being redirected this way.

When Daniel arap Moi took over as president in Kenya, for instance, government investments shifted significantly towards regions from which Moi drew his support. Studies also indicate that the opportunities for favouring certain individuals or groups are numerous.

A recent study of the budget process in Malawi, suggests that the formal budget process is merely for show ('a theatre'), whereas informal practices determine the real allocation of funds, and in practice secure the interests of the politically powerful, to the detriment of the poor. See this ODI working paper for a general discussion of development aid within the context of budget processes.

In a survey conducted by USAID of the perceptions donors and other stakeholders have of general budget support to Tanzania, it was found that budget support delivered few of the advantages attributed to it. Notably, financial management capacity was strained, dialogue between donors and partner governments has been of mixed quality, democratic accountability is unlikely to improve, and transaction costs do not seem to have been reduced, at least not in the short run. Moreover, it is argued that given the high level of leakage of funds from the central government level to local service providers, and the generally high level of corruption in Tanzania, budget support is unlikely to be effective. However, few of the donors surveyed saw corruption as a prohibitive factor in providing budget support. Some even saw budget support as offering a chance to identify leakages, and enter into a dialogue with the government to stop the leakages. Given the unclear effect of budget support on general dialogue processes identified by this study, this seems a dubious proposition.

A similar USAID evaluation study of budget support to Mozambique, finds that the high level of corruption threatens development gains. The study also notes that general budget support, by nature of the way it is disbursed, can easily be spent on non-development related activities. For instance, 70 per cent of Mozambique's budget is spent on the relatively prosperous Maputo region, which indicates that allocation decisions might not be made primarily with a view to reducing poverty. Finally, the study argues that budget leakages might actually force aid ministries to cut back on budget support, i.e. budget support is not sustainable in the current environment. For budget support to be a viable strategy, a significant improvement in financial management is required.

The views of development agencies differ, however, as to whether budget support is inappropriate in corrupt environments. Some, such as USAID, suggest that project aid makes more sense where corruption is rampant. Others do not see corruption as a prohibitive factor. It is important to note, however, that the effectiveness of budget support depends on the characteristics of partner countries, and the proper response varies from one country to another. Moreover, budget support can be provided in a variety of ways and with different conditions attached, thus adapting budget support to the particular partner country context.

For instance, the pattern of corruption in a country may suggest that sector budget support should be given rather than general budget support. Or it may be appropriate to attach conditions that permit donors to monitor and respond to misappropriation of funds. In many cases, the effectiveness of budget support can also be enhanced through complementary investment, in the form of technical assistance or project support, to expand the financial management capacity of a partner country.

In an effort to evaluate whether partner countries have the institutional characteristics required for budget support to be effective, donor agencies have developed various tools to diagnose the level of fiduciary risk. For instance, DFID have developed a set of good practice principles and related benchmarks, by which to assess fiduciary risk. Similarly, the World Bank has developed a Country Financial Accountability Assessment procedure (CFAA).

These tools to some extent also cover the question of corruption. Moreover, efforts to harmonize fiduciary risk assessments are made under the Public Expenditure and Financial Accountability (PEFA) programme. Further details on existing tools for fiduciary risk assessment and the harmonization process can be found here.

[top]

Part 3: Does budget support increase corruption?

A basic characteristic of budget support is that it gives developing countries a greater say in the allocation of development aid. In other words, budget support increases the discretion of partner country governments in allocation decisions. From a theoretical point of view, increased discretion for public officials may be problematic in a corrupt environment, as it expands the opportunities for giving or demanding bribes, or for favouritism. A relevant question to ask is therefore whether budget support promotes corruption, relative to other types of aid.

At present, there do not seem to be any empirical studies on how budget support affects the level of corruption. However, certain lessons can be derived from existing studies of aid in general and corruption. The available literature provides mixed results on the effect of aid on corruption. One study by a World Bank economist suggests that discretionary aid may lead to an increase in rent-seeking activities in an economy. The study finds that foreign aid is associated with higher corruption in countries with more strongly competing social groups. By implication, donors should be careful in using budget support in highly fractionalized societies.

A more recent study finds that foreign aid decreases corruption. Two reasons are given for this conclusion:

• Foreign aid is associated with conditions which limit the discretion of the recipient country. It suggests that the result that aid decreases corruption, cannot be extended to unconditional budget support.
• Foreign aid provides added liquidity which alleviates public revenue shortages, and may thus be conducive to increased salaries for public officials. It suggests that corruption may decrease if budget support is used to boost the salaries of public officials, which, however, remains an open question.

In sum, the effect of budget support on corruption seems to depend on whether the characteristics of the partner country induce rent-seeking activities, what conditions are attached to budget support, and whether funds can be expected to be channelled in ways that reduce incentives for corrupt activities. Donors should be wary of using budget support in highly fractionalized societies, or societies that in other ways are vulnerable to rent-seeking activities. And in cases where increased corruption is a risk, donors should consider attaching conditions to budget support that counteract this effect. Where possible, donors could implement complementary programmes to specifically reduce incentives for corruption. As these are inferences drawn from studies of aid in general, more knowledge is needed to establish the impact of budget support on the level of corruption in partner countries.

[top]

Part 4: Recommendations and resources

Main lessons/recommendations:

• Corruption makes budget support less effective, and budget support may increase corruption.
  
• In partner countries where corruption may substantially decrease the effectiveness of budget support, donors should consider whether effectiveness can be enhanced through the mode in which budget support is given, by attaching conditions, or by means of complementary investments/technical assistance.
  
• Donors should be careful in using budget support in highly fractionalized societies, as rent-seeking activities are likely to increase. In cases where there is a risk that discretionary budget support may increase the level of corruption, donors should consider ways in which to counteract this effect.
  
• In choosing between budget support and project support, donors should consider whether budget support actually delivers the advantages attributed to it, and whether these advantages outweigh the potentially negative effects of corruption.