2007
Budget support: As good as the strategy it finances

Rebecca Carter, Stephen Lister

By signing up to the Millennium Development Goals (MDGs), the international development community has made a commitment to turn into reality the right to social security for all. To finance the MDGs aid agencies have promised to deliver more aid in a more effective way. One aid instrument that has risen in prominence is budget support. The term ‘budget support’ encompasses general and sector budget support. Budget support is an effective instrument when the government is implementing a poverty reduction strategy or a development strategy that its aid partners broadly support. The governments must be able to maintain economic discipline and control public expenditures, and there must be a high level of trust between the government and its partners. In these circumstances, budget support avoids many of the problems that accompany other forms of aid.

By signing up to the Millennium Development Goals (MDGs),[1] the international development community has made a commitment to turn into reality the right to social security for all. The International Labour Organization (ILO) definition of social security includes basic health and education services, as well as income security. There remains a strong role for state financing and provision of these services; therefore they will feature strongly in government-to-government aid for poor countries. Typically the requirement is to underpin recurrent costs, not just investment costs of basic services. This article looks at the relevance of budget support to financing the relevant MDGs, and draws some conclusions about the role of budget support, how it should be designed and the attitude civil society organizations should adopt towards it. It draws largely on the Joint Evaluation of General Budget Support[2] in which both authors were involved.


More and better aid promised

To finance the MDGs aid agencies have promised to deliver more aid: donor countries have pledged to meet the official development assistance (ODA) target of 0.7% of gross national income (GNI), and the G8 has pledged to double aid to Africa by 2010.

However, these promises are not being met, and this is putting the success of the MDGs in jeopardy. The United Nations (UN) has shown its serious concern with this state of affairs in the latest MDG progress report: “In particular, the lack of any significant increase in official development assistance since 2004 makes it impossible, even for well-governed countries, to meet the MDGs. As this report makes clear, adequate resources need to be made available to countries in a predictable way for them to be able to effectively plan the scaling up of their investments.”[3]

As well as an increased volume of aid, the aid agencies have promised, through the Millennium Declaration, the 2002 Monterrey Conference on Financing for Development, the 2005 Paris Declaration on Aid Effectiveness and the 2005 World Summit, to deliver more effective aid. Since the late 1990s there has been a growing consensus that inappropriate aid modalities had become part of the problem. Efforts to bypass weaknesses in government systems were seen to have further weakened them, to have fragmented national decision-making, and to have raised the transaction costs of aid. The aid effectiveness debate has led to a new consensus that aid needs to be delivered in a way that promotes harmonization, alignment and recipient government ownership.


Budget support: rise to prominence

In this context, one aid instrument that has risen in prominence is budget support. The proponents of budget support claim that it is an efficient mechanism for delivering scaled up aid effectively. As this view has gained ground, an increasing number of aid agencies have started disbursing more aid as budget support. The European Commission (EC) in a recent proposal declared that “a shift to more budget support will be essential to make effective use of scaled up aid.”[4]


Definition and expectations

Budget support is aid funding to government that is not earmarked to specific projects or expenditure items and is disbursed through the recipient government’s own financial management system. It is not a new phenomenon: several former colonies received general budgetary grants for some years after independence; balance of payments support – including structural adjustment lending by the World Bank and the International Monetary Fund (IMF) – often generated local currency that governments could use to support their budgets according to their own priorities. More recently, debt relief has been an important form of budget support (for example, when bilateral donors used special debt relief funds to service developing countries’ debts to the World Bank and IMF, governments were able to spend an equivalent amount on their own domestic priorities instead).

However, recent forms of budget support have focused more directly on the government budget. They have been designed to support nationally owned poverty reduction strategies in ways that would strengthen national capacity and ensure more sustainable development – hence the designations ‘poverty reduction budget support’ and ‘partnership general budget support’. (‘Partnership’ is contrasted with the imposed conditionality of the structural adjustment era.)

The term ‘budget support’ encompasses general and sector budget support. All types of budget support include a lump sum transfer of foreign exchange; differences then arise on the extent of earmarking and on the levels and focus of the policy dialogue and conditionality. Sector budget support is distinguished from general budget support by being focused on a discrete sector or sectors, with any conditionality relating to these sectors. Often the funds are not strictly earmarked to the sector. In practice the design of budget support instruments is a spectrum (see Box 1).

BOX 1. Design options along the general and sector budget support spectrum

There has been a lack of systematic knowledge on the actual design, practice and effects of budget support, due to its complex nature and the fact that in the form of ‘partnership’ budget support it has been used more widely only from the late 1990s.

There are a number of assumptions made about budget support. For example, when used to finance country development strategies, budget support is expected to also have a wide range of complementary effects, such as:

• Improved coordination and harmonization among donors and alignment with partner country systems (including budget systems and result systems) and policies.
• Lower transaction costs.
• Higher allocative efficiency of public expenditures.
• Greater predictability of funding.
• Increased effectiveness of the state and public administration as budget support is aligned with and uses government allocation and financial management systems.
• Improved domestic accountability through increased focus on the government’s own accountability channels.

Not all preconceptions of budget support are positive: another set of assumptions focus on the risks associated with it. A common view, for example, is that budget support is more vulnerable to corruption than other forms of aid, and sometimes it is crudely characterized as “money for governments to do what they like with.”


Joint Evaluation of General Budget Support

The debate on budget support has been moved forward with the completion of the Joint Evaluation of General Budget Support.[5] An independent study led by the International Development Department of the University of Birmingham commissioned by a group of 24 aid agencies and 7 partner governments under the aegis of the Development Assistance Committee (DAC) of the OECD, this was the first systematic attempt to assess to what extent and under what circumstances general budget support is relevant, efficient and effective for achieving sustainable impacts on poverty reduction and growth (see Box 2 for further details).

While the evaluation’s focus was on general budget support, one of its findings is that many of the lessons in design and operation of budget support are relevant to both general and sector budget support. This report draws upon the evaluation to explore the assumptions behind some commonly held views regarding the effects of budget support in general. The evaluation also highlights that the interaction between general budget support and sector budget support is an important practical consideration and we return to this in the section on budget support design.

BOX 2. The Joint Evaluation of General Budget Support 1994-2004

A broader strategy or 'package'

The evaluation found that although budget support money is not earmarked to specific expenditures, it is part of a broader understanding about how the government resources will be used. The finance is accompanied by other 'inputs'. These include: the conditions on which funding is provided and procedures for dialogue between government and donors; donor efforts to harmonize their aid and align it with national policies and procedures; and technical assistance and capacity building. Box 3 describes a typical budget support package.

BOX 3. Budget support is a package

Only as good as the strategy it finances

Partnership budget support is used to support national poverty reduction strategies, and so it reflects the strengths and weaknesses of those strategies. The first set of Poverty Reduction Strategy Papers focused strongly on expanding access to basic public services, especially primary education and health care. Budget support has been an efficient way of supporting those strategies, but they have had limited effects on growth and on raising the incomes of the poor. Second-generation poverty reduction strategies are paying more attention to growth and income poverty reduction.[6]

The general budget support evaluation concluded that budget support is an effective instrument when the government is implementing a poverty reduction strategy that its aid partners broadly support. The government must be able to maintain economic discipline and control public expenditures, and there must be a high level of trust between the government and its partners. In these circumstances, budget support avoids many of the problems that accompany other forms of aid (e.g. uncoordinated projects that undermine government systems, impose high transaction costs and lack sustainability).


The potential to strengthen government systems

A characteristic feature of budget support is a strong focus on public finance management. This stems immediately from fiduciary concerns about the resources entrusted to national public finance management systems and, more fundamentally, from the budget’s role as the key link between policy and implementation. Greater focus on the government budget (as opposed to funds separately dispensed by aid agencies) gives public agencies an incentive to compete for public funds and strengthens the budget process. This also strengthens the formulation of national policies. Budget support strengthens the demand for timely and transparent budgets and expenditure records. This complements technical assistance and capacity building efforts that focus on the supply of technical improvements. Looking at public expenditure management, allocatives efficiency is improved by making more funds available to finance poverty reduction strategy priorities and operational efficiency is improved by allowing a better balance between recurrent and capital costs, and giving governments more flexibility in the use of funds.


Not necessarily more vulnerable to corruption than other aid instruments

Budget support requires a basic level of trust between partners. Corruption – especially high-level corruption – undermines this. Corruption also corrodes public support for aid in donor countries. Corruption was perceived as a serious issue in all the study countries of the general budget support evaluation but it is inherently difficult to measure. Available data are not robust enough to indicate reliable trends in performance. Corruption can affect all modalities of aid, sometimes in subtle ways (e.g. corruption creates a bias towards capital expenditures, because investment projects offer more opportunities for illicit gain). Aid modalities themselves affect the environment for corruption (e.g. a multiplicity of donor procedures outside of government systems may complicate and undermine the role of national audit institutions; tied aid may create a non-competitive contracting environment). Box 4 highlights lessons of donor experience in combating corruption.

BOX 4. Donor experiences in combating corruption

There was no clear evidence that budget support funds were, in practice, more affected by corruption than other forms of aid. Actions against corruption were included in the performance matrices and conditions for budget support in all cases, but highly visible legal measures were rarely very effective.

Budget support’s contribution to the strengthening of public finance management (PFM) probably had a more significant effect on the environment for corruption. This is because “the nature and quality of a country’s PFM system to a large extent determine the ease with which public corruption can occur.”[7] Building on earlier work (notably the fiduciary analyses and assessments linked to the Highly Indebted Poor Countries [HIPC] initiative processes), budget support-related dialogue and technical assistance have continued to support improvements in transparency, procurement management and auditing; their joint involvement in budget support has tended to increase coordination among donors on such issues and added to the collective weight of donor pressure for improvements in government accountability systems. This includes specific measures such as expenditure tracking studies, which are helping to address practical issues in ensuring that resources and services reach their intended beneficiaries.

Budget support donors have also pursued anti-corruption strategies by complementary means, including specific projects and technical assistance to support accountability institutions (audit agencies, parliaments, etc.), and support to civil society organizations.


Budget support good practices

Complementarity of aid instruments

Budget support tends to enhance the country-level quality of aid as a whole. For example:

• It provides more funds for recurrent costs, so that government can operate the new facilities provided through projects.
• All forms of aid benefit from the strengthening of public finance management systems.
• It promotes better coordination among all donors, and more consistent expenditure plans across sectors.

When large amounts of off-budget project aid continue, the positive effects of budget support are weakened by: fragmentation of the planning and budget process; project management structures that undermine core government capacity; and higher transaction costs for government.

While there is often an important role for general budget support, it is not a complete substitute for other ways of providing aid. Different aid instruments can complement each other. For example, well-designed technical assistance can reinforce the capacity-building effects of budget support; projects can be useful in trying out innovations, or as a way of managing large infrastructure projects.

The general budget support evaluation therefore advocates a portfolio approach which does not assume that one modality is always superior, but rather looks explicitly at the comparative advantages and the complementarities between modalities in any given situation.

The DAC guidelines on harmonizing donor practices for effective aid delivery (see Box 5) imply a stronger discontinuity between general and sector budget support than the study found. There is a spectrum of budget support instruments (see Box 1), and many of the good practices defined for general budget support will also apply to instruments that are habitually referred to as sector budget support. Alignment and coordination among budget support instruments with different (general/sector) orientations is an important practical issue.

BOX 5. DAC (2005) guiding principles and good practices for budget support

Budget support focused on particular sectors could be a useful complement to general budget support, as long as general and sector budget support are carefully coordinated in support of consistent economic and budgetary targets. However, the general budget support instrument (with its associated dialogue and support for capacity development) plays two roles that sector budget support could not provide in isolation: (a) as the focus of support for strengthening overall public finance management, including the budget system; (b) as a force for coherence and alignment across sectors.

Design principles for budget support

The general budget support evaluation report supports the DAC advice, with some additional comments:

• General budget support needs to be conceived (and developed and managed) as part of strategy which takes explicit account of the interplay between different aid modalities and instruments, seeking to exploit complementarities and tackle dissonance between them.

• The findings from the country studies as a whole do not support the idea that there is a standard evolutionary sequence, in which project aid first gives way to sector programmes (or sector basket funds) before the eventual introduction of unearmarked budget funding. They do support the value of moving to the use of government systems as early and as completely as is practical.

• There should be an incremental approach to the use of budget support. It needs to be adapted to country circumstances, and building up effective systems and procedures is an iterative process. Where there are doubts about the quality of PFM systems, both the learning and the incentive effects of initially modest disbursements may be valuable. Aid agencies as well as governments need to learn and to adapt their capacities. Over time, and depending on performance, budget support may be scaled up in several dimensions: in volume of funds (including a contribution to the scaling up of total aid flows), as a share of aid resources, and in terms of the policy and sectoral scope of the budget support dialogue.


The need for predictable and genuinely long-term aid

The 2007 UN MDG progress report shows that adequate resources are still not being made available to countries in a predictable way. Genuinely predictable and long-term aid is not being delivered. Donors are still – by and large – unable to commit to three-year budget support cycles that would facilitate medium-term expenditure framework planning. In practice, even longer-term commitments would be necessary to assure partner governments that they have a stable source of financing for MDG-related recurrent costs of social and other public services. Social security type expenditures need to be predictable, continuous, and not subject to the ‘stop-go’ features of aid politics.

The DAC good practice guidelines[9] advise that “political conditionality should not be specifically linked to budget support or any individual aid instrument, but rather should be handled in the context of the overarching policy dialogue between a partner country and its donors.” Nevertheless, experience shows that budget support, and especially general budget support, is especially vulnerable when there is a deterioration in political relations. This undermines budget support as a long-term instrument. Apart from immediate disruptive effects, it makes partner governments less likely to treat budget support as a reliable source of financing for medium and long-term planning, and this in turn may undermine some of the distinctive benefits of budget support.

The challenge is to find ways of reliably delivering aid through government systems to poor people even when there are political issues with the government. A step in the right direction is the recent EC proposal to provide more long-term and predictable general budget support, which is to be called ‘MDG contract’ to highlight the contractual nature of its long-term financial commitments and its focus on MDG-related results.[10] However, the EC proposal does not answer all the questions. The EC MDG contract concept is commendable, but it is worrying that it is seen as for ‘good performers’. Designs are needed that provide social security for people who live under all sorts of governments.


Practical designs: basic education in Ethiopia

A relevant and interesting example is Ethiopia. Ethiopia presents a direct challenge to Gleneagles and other international commitments concerning the MDGs. The country is exceptionally poor and receives less aid per capita than most of sub-Saharan Africa. Yet the government has demonstrated commitment to poverty reduction, backed by the mobilization of domestic resources and an effective administration. It has an exceptional track record in expanding basic education, and there could hardly be a case more deserving of international support. Yet donors have so far failed to deliver predictable financing on an appropriate scale. Successive Education Sector Development Programmes over nearly a decade have drawn declarations of donor support, but actual financial aid to the education sector has been disappointing. Goodwill has not been converted into long-term predictable funding.

Aid has been disrupted by political concerns. Direct budget support was suspended in the wake of the civil conflict that followed the disputed elections of 2005. In its place, the so-called Protection of Basic Services (PBS) project was developed and became effective in mid-2006. The PBS programme explicitly recognizes that support to help poor people towards the MDGs should not be jeopardized by the ups and downs of political relationships. (Withdrawing such aid is not an effective way to put pressure on governments, and would not be an ethical approach even if it were effective.) However, the political context required it to be delivered with additional transparent safeguards.

PBS as developed builds on the fact that decentralized governments are responsible for the bulk of primary service provision, largely financed by the federal block grant. PBS therefore augments the federal block grant. PBS arrangements include the monitoring of intergovernmental fiscal transfers as a whole, and include tests of additionality and fairness as well as other fiduciary monitoring. As the largest expenditure commitment of local governments, education is the main beneficiary of PBS funds. Box 6 describes the key features of PBS.

BOX 6. The Protection of Basic Services (PBS) project in Ethiopia

Although introduced as an emergency substitute for more conventional general budget support, the basic services approach is a superior one because it offers credible safeguards against political disruption. However, the PBS instrument is still essentially hand-to-mouth (a one or two-year time horizon for commitments is far too short compared with the eight-year primary education cycle, for example). The principal challenge is to develop the instrument into one that remains politically robust but provides genuinely long-term commitments on a larger scale.[11]


Conclusion

Budget support is not a panacea, but it should play an important role in meeting MDG commitments. Donors need to demonstrate political will and a willingness to innovate, so as to develop forms of budget support that ensure continuous support to poor people, even when the political context is difficult. For civil society organizations there is also an important role: in advocacy, to hold donors to their funding commitments, and in terms of strengthening social accountability for public expenditures (including aid) in recipient countries.

Notes:

[1] For details of the MDGs, see the article by Joyce Haarbrink in this Report.
[2] IDD & Associates (2006). Joint Evaluation of General Budget Support: Synthesis Report. May. Glasgow: DFID.
[3] United Nations (2007). The Millennium Development Goals Report 2007. New York: UN. Available from: <www.un.org/millenniumgoals/pdf/mdg2007.pdf>.
[4] EC (2007). Technical Discussion Paper on a “MDG Contract”. A Proposal for Longer Term and More Predictable General Budget Support. 19 June. EC DG Development.
[5] IDD & Associates (2006), op. cit.
[6] Driscoll, R. et al (2006). Trade and Growth in Second Generation Poverty Reduction Strategies. Report for Department for International Development. London: DFID.
[7] Dorotinsky, W. and Pradhan, S. (2007). “Exploring Corruption in Public Financial Management”. In Campos and Pradhan. Eds. (2007). The Many Faces of Corruption. Washington DC: World Bank.
[8] IDD & Associates (2007). Joint Evaluation of General Budget Support 1994-2004 – Briefing Paper: Policy Questions and Answers. March. Glasgow: DFID. Adapted from OECD DAC (2005). Harmonising Donor Practices for Effective Aid Delivery: Volume 2 – Budget Support, Sector Wide Approaches and Capacity Development in Public Finance Management.
[9] OECD DAC (2005), op. cit.
[10] EC (2007), op. cit.
[11] Lister, S. (2007). Scaling Up Aid for Education in Ethiopia. April. Oxford: Mokoro Ltd.

The authors are members of the staff of Mokoro Ltd which was formed as a not-for-profit company in 1982 to provide technical assistance and support for economic and social development and resource management.