Global Policy Forum

Partnership Prospects for Donor Budget Support

View full report: Evaluation of General Budget Support (1994 -2004), published by the Development Assistance Committee (DAC) of the Organisation for Economic Co-operation and Development (OECD).

Paris in the spring may have had something to do with the prevalence of references to love and marriage at the recent presentation of the OECD - DAC Evaluation of General Budget support last 9th and 10th May. Although perhaps not concepts one might directly associate with any aid modality, nonetheless the language of love, marriage, pre-nuptial agreements and even infatuation nonetheless provided sometimes useful analogies for the dry development language of partnership, policy dialogue, conditionality and ownership!

The meeting was attended by all the major bilateral agencies, as well as the International Financial Institutions (IFIs) and UNDP. Also in attendance were government representatives from all the case study countries (see below) except Mozambique and in addition Ghana and Tanzania. A group of civil society representatives were also present including Actionaid International, Care International, INTRAC, Dóchas (Irish Platform of NGOs), CAFOD, EURODAD and ODI.

This evaluation of budget support was carried out for the OECD DAC over the last two years by the International Development Department (IDD) (University of Birmingham, UK) and it examined seven country case studies: Burkina Faso, Malawi, Mozambique, Nicaragua, Rwanda, Uganda and Vietnam. The evaluation focused on what it terms ‘partnership general budget support'. This refers to the funds which go directly through the recipient governments' own financial management system (and not earmarked for specific use) and also to the associated conditionality, dialogue, technical assistance, harmonisation and alignment that accompanies these funds.

Whilst clearly some donors are more enamoured than others with budget support, the evaluation was on balance positive – as well as realistic – about what budget support can achieve. The evaluation team adopted the term 'Partnership' general budget support (PGBS) for the purposes of the evaluation; a term they used it for a specific type of general budget support that is identified by the following criteria:

• Unearmarked funds to government budget
• Using government systems
• Complementary inputs: dialogue and conditionality, harmonisation and alignment, technical assistance and capacity building
• Focus on partnership GBS: new style of conditionality with agreed objectives rather than imposed conditions, and support to poverty reduction strategy
For the purpose of this report PGBS will simply be referred to as budget support to avoid excessive acronyms.

Main positive conclusions

The evaluation found that general budget support had a positive impact on improving pro-poor expenditure particularly on improving the amounts directed to social services, although because of the difficult of assessing causality it was not possible to say whether it has any particular effect on growth and income poverty. The evaluation concluded that there was a decrease in transaction costs for both donor and recipient governments – particularly where the level of budget support was higher – despite the considerable costs associated with negotiating and monitoring budget support itself.

Improving public financial management was seen as a key benefit, not only in terms of management of budget support itself but also of other project and sector financing as government systems are strengthened as more money is channelled through them. It was also found that budget support tends to improve policy coherence and execution.

Budget support has reinforced pre-existing macroeconomic stability and has also widened international partner involvement in macroeconomic debate.

The evaluation also concludes that budget support works best where there is significant political commitment in the government already – not just in the narrow sense of governments being willing to implement ‘good policies' but in a broader sense of recipient governments genuinely being ambitious for the country and for the people. However Vietnam was the only country that was cited as an example of this broader commitment.

Main negative conclusions

If budget support is tied to recipient governments fulfilling the IMF's macro-economic conditions it can have a perverse impact on predictability and size of resource flows. This is the case especially since other forms of project and sector support are not usually tied to such conditions and where funding for budget support is not additional money on top of money for projects and sector support. While budget support increases the sense of ownership, it often comes at the cost of greater intrusion and influence of donors on all aspects of national budgets.

Budget support creates greater sense of risk among senior aid managers and particularly donor ministries of losing money to corruption (although there is evidence that budget support is not more or less vulnerable to leakages than any other aid modality).

Technical assistance/ capacity building has been the least well integrated aspect of budget support and the political context has tended to be less well analysed and adapted to than other elements of the context. In addition, donors still have a lot to do to improve their coordination of analytic work with significant duplication still occurring.

Key recommendations of the evaluation

Key recommendations:

1. Budget support can be applied in almost any context and its application should be defined by any particular set of ‘hard entry criteria', but agreement around broad principles and performance targets.

2. Disbursement should not be tied to achievement of specific performance targets but based on an overall assessment of progress which should be jointly done.

3. The role of the IMF in monitoring, reporting and advising on macroeconomic performance should be maintained but PGBS funds should not be mechanistically tied to IMF conditions.

4. There should be deeper application of principle of mutual accountability and process and space to develop ideas and define what mutual accountabilities mean and how they will be managed.

5. There should be joint approaches to working through the following tensions: a. Autonomy of government vs. poverty focus of donors
b. Government vs. country ownership
c. Role of Ministries of finance/planning vs. role of government more generally across all ministries
d. Ownership vs. conditionality

NGO assessment

NGO representatives present welcomed the evaluation which they thought provided many useful proposals for moving forwards. They agreed with the evaluation's recognition of the role that donors play as day-to-day actors within the political and policy environment, not merely as external influences, as well as the appreciation that poverty reduction is a highly political and not a technical issue. This inevitably requires that donors improve their political knowledge and awareness of both formal and informal structures and processes in each country.

One element that was missing from the report was the question of domestic accountability. Although general budget support is generally thought to improve domestic accountability and shift the direction of accountability towards domestic constituencies, this issue was barely analysed in the report. Development of effective states requires greater domestic accountability, which should not be restricted to policy implementation only but must include holding governments to account for the entire policy process. Donors should not therefore view domestic constituencies as instruments of their agenda but as strategically vital for building genuine political commitment to development.

NGOs welcomed the recognition that a rigid link between the IMF's PRGF and general budget support disbursement conditions can lead to unnecessary volatility and endorsed the recommendation that donors refrain from tying general budget support funds to the IMF's macroeconomic conditions. However, they expressed their concern that the IMF nonetheless dominates in an environment where there is little openness in terms of its advice to governments. Macroeconomic discussions are still largely held between the IMF, Central Bank and Ministry of Finance.

Evolving consensus from divergent debate

The conference provided the opportunity for a frank debate on the conclusions of the evaluation and the positions of donor, recipient and civil society organisations. Though the discussion focused on budget support specifically, clearly many of the issues raised are relevant for the quality of aid delivery generally.

Courtship and conditionality

There was considerable discussion on when donors should start budget support in individual developing countries. The importance of trust between donors and recipient governments was emphasised many times as well as the partnership – or marriage between donors and recipients – which like any marriage has repercussions if the relationship breaks down. One African representative however reminded us that this marriage is of course not one of between two equals with donors tending to impose their views and lack humility in their actions towards the recipient government.

Conditions attached to development loans and grants continue to be controversial. However the discourse at least seems to be shifting; donor representatives were talking about identifying ‘common targets', ‘performance indicators' and ‘standards' and not attaching disbursement or minimum conditions to budget support grants. According to the World Bank representative "we have no minimum standards (for giving budget support) as long as there are commitments for improvements form the government". Nonetheless the unequal power relations make it more of a challenge for such common targets and agreed indicators to mean something substantially different from conditionality. Furthermore the evaluation stressed the importance of not proliferating indicators; one participant noted the tendency for donors to want to insert their own indicators into performance frameworks so as to monitor where ‘they had made a difference'. The World Bank said that they believed "the notion of criticality was very important' and that they should ‘choose just those areas that are important for the operation".

According to the evaluation team, donors should be quite flexible on ‘entry conditions' and willing to take risks as well as incremental approaches. The evaluation pointed to the positive experiences of budget support in both Mozambique and Rwanda, where budget support was started at not so stable periods. However some donors are clearly more cautious than others. Luxembourg admitted it was very ‘reticent' about budget support; USAID said that a "courtship" might be necessary before it could tie the knot! They thought that "whilst it is fair to say no rigid standards should be in place… we feel one needs to be selective in the countries one engages in PGBS (…. and) we shouldn't expect to see partnership in the same way in fragile states as in performing states". Germany also stressed the ‘basic criteria' that needed to be in place before providing PGBS, namely "good governance, good performance and strong political will". However based on these criteria, positive examples such as Mozambique and Rwanda would have been unlikely to have ever received budget support in the first place.

Can donors pull out when things go wrong?

There was a broad consensus that pulling out in an ad-hoc or sudden manner undermines the very predictability and macro-economic stability that budget support aims to improve. Nonetheless donors are clearly apprehensive about the political risks and perceptions of the public in the north regarding political instabilities/ corruption allegations etc. The Irish representative referred to a ‘pre-nuptial agreement' and argued that "from the outset we should make it clear that there are issues that need to be discussed and on the table (…) (and) as we have genuine dialogue with our partners we have to make it clear what we are going to do if things go wrong". Donors agreed on the need for transparent Memorandums of Understandings (MoUs) of agreement and also graduated responses if the terms of the agreement are not respected. Simply publicising information and signalling to a government and the people is a graduated response in itself.

The Uganda representative challenged the CPIA scorecard against which Uganda at the moment would score badly on governance concerns despite the fact that it has ‘performed well in most areas' in particular in reducing poverty. He suggested that perhaps what is needed is to "list inter-alia the basic tenets that should be observed (which) could maybe help to minimize tension and improve dialogue". Overall there was an agreement that donors need much better political analysis and understanding in each country. Ethiopia was citied as a good example of the need for donors to understand much more clearly how political and administrative systems work with a suggestion that donors' reactions in the country had in fact been counterproductive.

On the question of predictability more generally, there was agreement that donors really need to give longer term multi-year commitments. The evaluation team argued that the debate on predictability has focused on predictable flow of funds within a given year, but that even thinking from one year to the next is a very hand to mouth way of living. Within a given year, as one participant said "surely it is possibly (for donors) to phase disbursements across the year to ensure there is no shortfall at the beginning of the year?"

Coordination or specialisation?

In all of the case-study countries, budget support was seen to have improved donor coordination and harmonization. This has even meant in some cases that other donors who are not giving budget support are brought into the coordinated sectoral policy discussions. That said there is still a strong argument for greater donor specialization and for some donors not to get involved in certain sectors or even in certain countries but to let other donors take the lead. In terms of pulling out of particular sectors, there is some evidence of progress on this front (e.g. Mozambique and Uganda) but there is little evidence of donors not proliferating across more countries. It does appear to be a paradox that donors are still playing such an active role in discussing sectoral policies during budget support.

In fact, as aid levels increase, there is a danger that donors proliferate even more and extend their limited capacity to even more countries. One Malawi representative suggested that it would be preferable were donors to channel their funds in a more substantial way through other bilateral donors and thus limit the numbers of donors present in the country. This would reduce the ‘deluge of evaluations, missions and new assessments of everything".

` Who is setting the agenda?

Although on balance general budget support is viewed more positively by southern governments than other forms of project/ programme aid, the increased influence and penetration that multiple donors have into the government means that there is a real risk that that donors crowd out national policy makers and in a way ‘become the government'. This is because donors suddenly have access to an increased amount of national government decision making, have considerable expertise and analytic resources and can in many cases write and recommend the policies that they think should get implemented. Donors are not very clear what to do about this particularly in situations where they feel that there is weak capacity to design ‘good policies' and where they think they can help influence a more pro-poor agenda through ‘policy dialogue'. However as Geske Dijkstra, a member of the evaluation reference panel. suggested that there needs to be more "dialogue" and less "monologue".

Samuel Wangwe, the Tanzanian Reference panel member said donors currently had too much influence and asked ‘will donors give up part of their influence? Will donors be modest enough to let part of the power go to domestic constituencies?" Frans Ronsholt, director of Public Expenditure Financial Accountability (PEFA) proposed that donors should at least "hold back a bit on providing ready-made recommendations once analysis is done" rather than the government to say what it wants or is capable of doing. In that way he argued "governments can do more to set the agenda rather than reacting to the particular agenda (of the donors)".

The Ghanaian representative said that one way to improve the capacity of developing country governments was to have both technical and political delegates in the negotiations. He said that ‘in Ghana we believe we are capable of determining where to go… we understand that donors must report back to their citizens". Ultimately donors should not lose sight of the longer term vision where they in fact have worked themselves out of a job, and take concrete steps to make sure that they are not undermining state building and democratic decision-making in the medium to long term.