Addressing Aid Effectiveness: A Key Challenge in Meeting the MDGs

Introduction

The amount of global resources for development cooperation has slowly begun to rise in recent years, reversing a downward trend that lasted for over two decades. Whilst aid levels are still low compared to figures in the 1970s, a growing number of donors are committing themselves to timeframes for reaching the UN target of 0.7% of gross national product (GNP) to overseas development assistance (ODA). Ireland's aid budget has doubled since 2000 and despite reneging on the commitment to reach 0.7% by 2007, the government remains committed to substantial increases in aid and to re-instating a timeframe in due course. A further four countries have recently agreed to reach the 0.7% target by 2013: Belgium, Finland, France and the UK.

This renewed focus on aid as a key component in international development efforts has cast light on the shortcomings of the existing international aid system in which there is a vast array of actors often working at cross-purposes. Improving the coherence within the system, with a view to making aid more effective, has become a major focus of the donor community. The Monterrey Consensus arising from the International Conference on Financing for Development in March 2002 made the specific commitment that donors would "harmonise their interventions behind country-owned poverty reduction strategies, to reduce the transaction costs of aid for recipients and make aid more effective". ¹ In other words, they committed themselves to working more closely with each other and in partnership with poor countries so as to ensure that poor countries had greater ownership of pro-poor development strategies. These commitments to more effective aid were futher expanded in the Tome Declaration on Harmonisation and Alignment (see box 1) ² which addressed three levels of the aid system: ownership (by developing countries of their own policies); alignment (donors align to partner countries' priorities and systems); and harmonisation (donors harmonise with one another through common arrangements, rationalising procedures and sharing information and analysis).³

Adherence to these principles by donors, international financial institutions, developing country governments and other stakeholders, including NGOs (non-governmental organisations), is essential for achieving and sustaining progress in meeting the Millennium Development Goals (MDGs).⁴

The Aid effectiveness agenda in practice

The commitments within the Monterrey Consensus and the Rome Declaration set out a clear agenda towards greater ownership and partnership. These commitments, however, have to be set against the evidence of practice on the ground. Three major surveys were carried out in 2003-2004 which shed light on the trends in fulfilment of donor commitments under the aid effectiveness agenda. The first two were commissioned by the Development Assistance Committee (DAC) of the Organisation for Economic Cooperation and Development (OECD), adn reviewed practices in 16 adn 14 developing countries respectively.⁷

Box 1: Harmonisation and aid effectiveness: the Rome Declaration

The Rome Declaration states that the key element in increasing aid effectiveness would be a ⁵ Developing countries were expected to undertake reforms necessary to enable progressive reliance by donors on their systems. Donors, for their part, committed to:

The third was by Development Finance International (DFI)⁸ based on analysis of donor policies and procedures carried out in 2004 by officials in 13 African countries. As will be seen from the discussion below, each of these reports reveals striking similarities in terms of the issues identified by developing countries as critical to aid effectiveness and the assessment of donor fulfilment of commitments.

Ownership and partnership

Country ownership of national development policies and procedures is fundamental to achieving effective development outcomes and impact. The Poverty Reduction Strategy Paper approach is based on this understanding and the Rome Declaration affirms it. However, many of the ways donors operate undermine nationally-owned development policies and procedures. The 2003 DAC Needs Assessment Survey aimed to find out which donor practices placed the highest burden on recipient countries. The main finding was that "there is a significant lack of ownership", where this is defined as "partner governments' ability to design and manage its own development plans while relying on donor assistance". The survey found that the principal burden for developing countries was "donor driven priorities and procedures", which had two dimensions: firstly, the pressure exerted by donors on partners' development policies and strategies; and secondly,donoraid managment systems which are designed to meet donors' rather than national needs and requirments. ⁹

This problem was further highlighted by the findings of the 2004 DAC surveyon aid effectiveness which examined whether developing country goverments exercised effective ownership by leading the aid coordination process. ¹⁰ In all of the 14 countries examined "it was felt that country leadership left a lot to be desired". Even those countries which have relatively advanced processes of aid coordination, at least for budget support, expressed redervations about whether a formalised government-led dialogue process was in place. ¹¹

Both surveys point to the existence of a problem at the heart of the donor-government relationship. They highlight the mismatch between donors' requirements and expectations and partner countries' capacity which leads to lack of confidence in partners' policies and systems. Donors consequently set up parallel aid implementation systems to bypass government systems. Donors also try to bring about policy change through direct (i.e. conditionality) or indirect (i.e. technical assistance) pressure. This leads to low ownership of both policies and procedures by developing country governments, which further undermines donors' confidence and trust. DAC describe this as the "low ownership trap"¹²

Development partnerships work best when they are based on the principle of mutual obligations, according to CIDSE (2005), which goes on to note that "top-down donor policies and conditionalides are fickle and are too often shaped by capricious and shifting strategic priorities". The DAC guidelines on aid effectiveness acknowledge that partnership is based on clarity of assumptions, mutual accountability for commitments and shared responsibility for outcomes, and encourage better practices. However the experience on both sides of the donoe - recipient relationship has been mixed.¹³

The problem, however, also has to be addressed from the recipient government side. African governments have often been poor at specifying their expectations of partners. That is, with the rare exception of the Government of Uganda whose Poverty Eradication Action Plan (PEAP, PRS-equivalent) contains a set of partnership principles developed by both Government and donors. The partnership principles commit both the Goverment of Uganda and donors to working within the policies and procedures of the PEAP and specify commitments on both sides of the development partnership ¹⁴ Proving that there is no development-utopia however, Uganda officals who were interviewed fo rhte ASC Needs Assesment  complained still setting up new systems of going through different channels in practice¹⁵

Alignment to country priorities

One key dimension of aid effectiveness is the alignment of donors' aid programmes to country priorities. The first DAC Guiding Principle on coordinating aid states: "Donors should support country-owned, country led poverty reduction strategies or equivalent national framework and base their programming on the needs and priorities identified in these". ¹⁶ This was endorsed in the Rome Declaration. The evidence on actual alignment of donor programmes to country priorities is dismal.

The 2004 OECD-DAC Survey found that whilst donors had taken on board the alignment with PRSs in principle, there was little evidence that they had actually tailored their programmes in support of PRS priorities.¹⁷ This mirrors findings in the World Bank Operations Evaluation Department's (OED) review of the PRSP (Poverty Reduction Strategy Papers) initiative, which found that neither bilateral donors nor the World Bank had defined whether or how they should change the content of their programmes to align with the PRS.¹⁸ African goverments responding to DFI's survey noted that the promises by the multilateral financial institutions to align their country assistance programmes with the priorities of the country's PRS had not had a mjor impact.¹⁹

Further evidence of this gap between rhetoric and practice is seen at the sector or subsector level where programme aid is often still earmarked. Johnson et al. (2004, p.23) give the example of an EU donor who supported the PRS priority sector of health, but insisted on funding projects at sub-sector level which are not government priorities.

Most PRSP's do not currently provide an operational framework for development planning or monitoring²⁰ Nonetheless, it is also clear that donors and the international financial institutions are failing to make the changes necessary to support PRSPs, with the result that a framework with much to offer is under threat.²¹  This had a lot to do with the lack of political will to align, as identified by respondents to the DAC needs assessment.²²

Alignment to country procedures

As well as aligning to country priorities, the DAC 2003 guidelines on aid effectiveness propose a series of practical steps for alignment to country procedures and systems. The surveys found that progress in this area is very limited however, particularly for projects. DAC found that in most countries significant numbers of donors do not use country systems at all for any of their projects.²³

Donors bypass government procedures because of a perception of weak government systems, as noted above. However, this perpetuates governments' incapacity to manage aid effectively. The respondents in the DFI survey expressed a strong preference for capacity building to help address shortcomings in systems, rather than technical assistance. Indeed inappropriate technical assistance, along with procurement, was one of the two main difficulties with donor procedures identified in the DAC Needs Assessment. ²⁴ This is not surprising when experience with technical assistance proves that it is at best a costly form of tied aid and at worst, an indirect form of donor pressure for changes in policy and procedures. Meanwhile burndensome procurement procedures lead to significant delays in disbursement of agreed funding and can often also represent informal tying of aid.²⁵

Such practices represent a drain on aid resources. DAC donors agreed in 2001 to untie most aid to Least Developed Countries but tied aid continues to reduce the global vlaue of aid by almost 10% (or $5 billion) per annum. ²⁶ Oxfam calculates that nearly one third of aid from G7 countries is tied, reducing the value of aid flows by 20-30%²⁷

Another area where problems existed is around the harmonisation of monitoring and reporting procedures. The DAC calculates that a typical African country submits 2,400 quarterly reports (i.e. 9,600 per annum) to different agencies and hosts more than 1,000 annual donor missions. Each mission meets with key officals and asks the goverment to comment on its reports.²⁸

The 2004 DAC Survey found that the number of donor missions per annum in the fourteen countries studied ranged from 400 in Cambodia and Vietnam to 120 in Zambia.²⁹ Less than 10% of these missions involved more than one donor, leading to the understated conclusion that this is an area where "greater efforts could be achieved by donors". Oxfam surveyed officials in 11 developing countries in June - July 2004 and found that over half of the officials felt they spend too much time reporting to donors. They also found that the Tananis hosted 275 donor missions, conducted by a total of 516 World Bank staff³⁰

Where harmonisation effort have been made, moreover, they tend to focus on the harmonisation of procedures amongst donors rather than on aligning behind host country procedures.³¹ This is of critical concern as the costs of trying to manage a multiplicity of procedures falls disproportionately on developing country authorities. As noted above, this risk consolidating impressions that the PRS is a piece of theatre for donors who continue with thier own plans regardless of national priorities.³²

Preferred aid modalities: the case for and against budget support

Another area which has an impact upon aid effectiveness is the specific financing instrument or modality chosen. The DF1 survey found a clear preference among African governments for programme assistance, as opposed to project aid or technical assistance. The principal reason for this relates to the freedom programme assistance allows a government, through enabling it to direct aid through its own budget and allocate it according to its own expenditure priorities and procedures. In doing so, it frees governments of many of the extremely burdensome procedures associated with project aid. The Rome Declaration recommended giving budget, sector or balance of payments support, providing "appropriate policy and fiduciary arrangements are in place". There is a marked trend not only towards increased budget support by individual donors but also towards multi-donor arrangements for budget support. Ten countries have now embarked on such programmes.³³ Whilst the multi-donor system has considerable advantages, as outlined in Box 2, it also has some serious risks.

Box 2 illustrates that although budget support throws up difficulties in ensuring that aid reaches its intended beneficiaries, this has to be weighed against the need to ensure that the institutions of state are strengthened, and not weakened, by development assistance. This is based on the premise that effective public financial management systems which are accountable to domestic constituencies are key to the effective use of aid and to overall development.

The multiplicity of administration and accounting processes required by thousands of separately funded projects offers a clearer picture of impact for specific groups of beneficiaries, but on a narrower canvas and with the effect of undermining efficient public expenditure managment.³⁴ The Mozambican Finance Ministry recently found that it had 1200 bank accounts, instead of a single consolidated account. This is largely as a result of donors insisting on separate accounting procedures for each project.

Box 2: Advantages and disadvantages of budget support

Advantages include:

 

Box 3: The advantages and disadvantages of multi-donor budget support

The advantages include:

 

For donors, moving to budget support would also mean a shift in emphasis at country office level which are not necessarily positive or negative, but would imply a certain amount of organisational change, More decision-making authority and analytic capacity would have to be transferred to the country office. Donors may find that their engagement would become less focused on project portfolio management and more on broader policy dialogue and coordination. ³⁶

The major risk facing recipient governments, however, would be increased aid volatility if performance targets and conditions are not met, or if donors fail to disburse aid in a timely manner and in accordance with national budget cycles. Moreover, whilst reporting requirements may be reduced through the use of one set of tools, recipient governments may actually experience increased transaction costs as a result of the elaboration of complex monitoring, reporting and evaluation systems.

Recipient governments will also be faced with policy conditionality from bilateral donors, whereas previously this had been the preserve of the international financial institutions, who have been engaged in budget / balance of payment support for decades. Killick (2004, p.21) notes that the partnership approach associated with budget support is supposed to do away with the need for extensive policy conditionality. However the IMF and World Bank have not introduced any substantial reduction in reliance on conditionality. Recipient governments are therefore liable to find themselves hemmed in by a proliferation of intrusive conditionality, with the prospect of donors converging increasingly around IFI (international financial institution) conditions.

The common architecture of multi-donor arrangements outlined in Box 3 offers advantages over individual arrangements, however it also has complicating factors beyond those experienced in single-donor budget support.

There is a need for careful assessment of the benefits and drawbacks of such a system as the results (positive or negative) are not necessarily automatic and may hinge on local political Economy factors.³⁸ A recurrent issue in multi-donor budget support arrangements is the need for absolute clarity on commitments undertaken and procedures to be followed. In particular, the identification and definition of conditionality and the monitoring and evaluation of compliance need to be carefully planned and managed. This is an added burden on all parties, but it is apparent that the time spent in preparation is crucial if the system is to function.

Conditionality

Conditionality is possibly the most contested of all areas in the debates over aid effectiveness. The legitimacy and effectiveness of conditionality have been heavily contested for two decades but a number of conclusions are by now axiomatic. Domestic political economy factors determine policy reform and attemps to buy reform typically fail.³⁹ This was a conclusion of the influential World Bank Assessing Aid report in 1998 and has been repeatedly affirmed by analysis of IMF conditionality, notably in the work of Graham Bird (University of Surrey), the IMF's Independent Evaluation Office (IEO) and Tony Killick (Overseas Development Institute), amongst others.

1. This central conclusion has informed an ongoing debate, which seeks to address the tension between national ownership, donors' policy preferences and donors' concerns around the management and impact of the aid they provide. A number of conclusions on conditionality are now widely accepted, at least in principle. For example, there is broad agreement that all conditionality should be based on benchmarks drawn from governments' PRSs or equivalent national development plans;
2. Conditionality should be clear, consistent and explicit, in order to improve the predictability of aid flows;
3. Conditionality should be streamlined across all donors, notably the international financial institutions;
4. Conditionality should take the local political economy and potential shocks into account;
5. Targets (particularly macro-economic targets) should be rigorously stress-tested for realism and achievability;
6. Whilst there is agreement on these issues, many points on conditionality are still the subject of intense debate, including: conditionality around economic governance, economic policies and civil and political governance and the role of the IMF in conditionality and signalling.

Economic governance: it is generally accepted that fiduciary conditionality which assures donors that their money is being used for the purposes for which it was intended, and which also enables accountability to citizens, is a legitimate form of conditionality. However, broader conditionality covering areas such as legal framework building, corporate governance, tax reforms, public procurement procedures and anti-corruption initiatives are contested. It has been asserted that these can be used by donors to impose thier own policy priorities, including frameworks for liberal market ecinomic policies, with unproven poverty reduction impacts⁴⁰

Economic policy, including fiscal, monetary and structural policy (trade liberalisation, privatisation etc.): a sovereign government should have the right to determine its own economic growth and development model and should be accountable to its own parliament and people for the policy choices it makes. It can be argued therefore, not only is imposing economic policy ineffective, but that donor involvement in this area is contrary to national ownership and the development of participatory accountability.

Civil and political governance, including human rights, free and fair elections, participatory decision-making and parliamentary oversight. The key question in this area is whether external pressure can engender greater respect for human rights, democracy and government accountability, and whether conditionality is the right vehicle for achieving these aims. One benefit of the PRSP approach, which is a form of macro-level conditionality, has been the opening of space for civil society to participate in policymaking and monitoring.⁴¹  However, the human rights question is potentially more difficult, particularly where this involves government-sanctioned aggression against Its own citizens or those of another country.

Conditionality, signalling and the role of the IMF: despite the evidence that IMF programme conditionality is ineffectual⁴² and results in severe destabilisation for poverty reduction  programmes,⁴³ it remains the cornerstone of IMF operations in low income countries. IMF conditionality is critically important because almost all budget support donors require a recipient government to have an on-track IMF programme. Disruption of such programmes results in suspension of budget support. The IMF is obliged to assess countries' commitment and capacity to repay short term balance of payments loans. Where it feels this commitment is lacking, as illustrated by failure to meet its benchmarks and performance targets, it will suspend its programmes. However, this assessment generally bears little relation to the information which donors need. That is, whether the government in question has the commitment and capacity to use aid resources effectively for the purpose for which they intended.

There is therefore a mismatch in bilateral donors' and the IMF's information needs. However, donor suspension of aid programmes when IMF programme's go off-track is a frequent occurence. The IMF currently remains formally outside of all multi-donor partnership frameworks which have developed around the PRSP and budget support systems. The Fund has participated in the negotiations around joint donor-government frameworks (for example in Mozambique and Rwanda) but claims that for legal reasons it cannot be bound by such agreements. Hence parallel negotiation frameworks are in place, where the IMF negotiates a macro-economic framework with a limited number of officials in the Ministry of Finance / Central Bank and donors negotiate a broad development plan and implementation strategy with government. The parameters of the latter will depend on the targets in the former - notably inflation and fiscal deficit targets, which can curb spending of available aid.

The macro-economic policy content of PRGF's predetermined the policy orientation of almost all countries' first PRS, even though this is sipposed to be the other way around, this is continuing into the second round of PRS.⁴⁴ Significant effort needs to be made to ensure that the IMF's policy prescriptions are verifiably based on the PRS or equivalent national framework, are realistic and achievable given economic and social conditions, will contribute to poverty reduction and are only adopted following an open debate in public and in a multi-stakeholder dialogue.

Existing mechanisms, such as PRS-related macro-economic working groups and dialogue processes under multi-donor budgetsupport arrangments, offer oppertunities to bring the IMF a broader process of policy dialogue.⁴⁵ Over the medium term, there is a need to assess whether and how the IMF should be formally involved in donor-government agreements on budget support.

In the meantime, there is a need for absolute clarity around the integration of macro-economic conditionality into a multi-donor framework. The response of each donor to failure to meet macro- economic conditionality agreed between the IMF and government should be clearly specified. Indeed, African governments have expressed a strong preference for donors retaining the freedom to respond individually to performance criteria. They propose that donors state explicitly what share of their disbursements is to be linked to specific preformance criteria, enabling partial disbursements if there are performance shortfalls.⁴⁶

Predictability and timeliness of aid: The DAC guidelines on aid effectiveness state that "donors, whereever possible, should programme their aid over a multi-year timeframe that is consistent with the palnning framework of the partner goverment and [be] transparent about the circumstances under which aid flows may vary.⁴⁷ However, the predictability and timeline of aid disbursal remain a major issue and aid volatility is a major source of economic instability in low-income countries.⁴⁸

A seminal study by IMF staff Bulir and Hamann (2001) showed that aid is far more volatile than domestic revenue. The gap between pledges and actual disbursements can be particularly significant for some donors (notably the EU which has $20 billion of approved but undisbursed funds). Both Tanzania and Mozambique experienced macro-economic and budget implementation problems when donor budget support decisions were not finalised until the government was already into the year of budgeted support.⁴⁹ Aid flows are so unpredictable that some governments (including the Government of Ugnada) arereportedly discounting projected donor and creditor inflows to project more realistic trends.⁵⁰

Donors are notoriously poor at notifying recipient governments of aid disbursements. Bulir and Hamann (2001) showed that the information content of commitments made by donors is either very small or statistically insignificant. In nine of of fourteen countries surveyed by DAC,⁵¹ clear procedures for notifying disbursements were not in place. Oxfam found that in Tanzania, 20 out of 39 donors did not return information about aid spending to the Tanzanian government when asked to do so.⁵²

The 2004 DAC survey⁵³ showed that few donors make commitments on a three-year basis. Some donors, such as the UK for example, are in a position to guarantee multi-annual support because of domestic multi-annual spending reviews. Donors such as Ireland, however, are not unequivocally able to do the same as domestic budgets and hence aid budgets are allocated on an annual basis. Nonetheless Ireland is ranked by African governments as having one of the best track records on predictability.⁵⁴

Flexibility: Finally, African goverments have identified lack of donor flexibility as the most problematic policy area in aid managment⁵⁵ African governments argue in favour of readily accessible contingency financing to address budget or balance of payments difficulties due to shocks. The IMF is working on a revised approach to shocks within its PRGF programme, but these programmes are invariably expensive, come with additional onerous conditionality and can undermine rather than regenerate economic growth. ⁵⁶

Other issues around flexibility include the continued earmarking of aid for projects and programmes at sector or subsectoral level, which increases the management and reporting burden for donors and undermines ownership. DAC guidelines recomment that donors make their procedures simpler and more fexible in order to allow them to work collaboratively and in support of country-owned approaches.⁵⁷ This implies a number of changes in work practice and in instruments, including the decentralisation of decision-making power to country offices and the establishment of substantial contingency lines in finance plans.

Proposals for partnership and mutual accountability

The dynamic of continued donor leadership and control over aid management is apparent from the three survey reports reviewed here, but the solutions proposed are subtly different.

The DAC guidelines emphasise donors setting transparent performance standards for themselves in consultation with national partners and being "willing to participate in assessmentsof performacein aid management".⁵⁸ While welcome, these commitments are relatively weak and seem to anticipate a rather passive response from developing country ("partner") governments.

It is clear from the DFI report that an explicit mutual accountability framework is a prerequisite for a well-functioning and effective partnership. Furthermore, the starting point for arriving at such a framework of agreement is not what donors are willing to offer, but what African governments determine to be their needs and preferred financing options.

Most African countries don not have a comprehensive external finance policy. This enables donors and creditors to maintain their leading role in driving a country's aid policy.⁵⁹ The DFI proposal would therefore involve developing country governments: identifying their finance needs and optimal types of finance; assessing donors on objective criteria and targeting those whose policies and procedures best match their needs; and designing a matrix of mutual accountability for a quality aid programme with the chosen partners. The matrix would include priority areas for improvement of both donors' and goverments' polices and proceedures.⁶⁰

Such an initiative holds considerable promise for breaking the low ownership trap. Its implementation requires many of the steps outlined in the DAC guidelines for aid effectiveness, but it is more progressive and more specifc than that set of guidelines. In fact, the unique characteristics of the DFI model are the exercise of control by national governments over development finance choices through explicit statements on aid policy and systems. This will require developing countries to be more assertive, as noted in the DAC survey,⁶¹ but it will also require donors to give their partners the space to carry out such analysis and then refrain from undermining it, either by bypassing government or by attempting to impose policy change through conditionality.

The reality is that, for most governments, constraints in access to finance will still leave them in an unequal position vis-a-vis donors. However, having prepared an external finance policy, developing countries would at least be able to attempt to discriminate in favour of preferred donors. Some donors may of course be preferred for political reasons, rather than because of the quality of their aid programme. This is inevitable. Nonetheless, building a system of mutual accountability at national level, based on developing country priorities and preferences, is a necessary step for effective aid management. Basing it on a rational and explicit statement of needs and preferences is logical and worthy of support.

Ireland and aid effectiveness

The DFI survey of African countries asked them not only to specify major obstacles, preferred solutions but also to rank donors according to the quality of their aid policies and practices. This survey sheds light on how Ireland is perceived as a donor, most of which affirms findings from the 2003 DAC Peer Review of Ireland.

The DAC Peer Review found that Ireland had a strong commitment to government ownership, illustrated through its support for pooled-funding arrangements, including budget support and SWAPs. The DFI survey showed that Ireland, along with the Netherlands and the UK, was perceived as one of the best bilateral supporters of government priorities.

Ireland also makes significant efforts to align to government procedures, according to the DAC Peer Review. While not highlighting Ireland as an above or below average donor in terms of procedures, the DFI survey nonetheless implies that as a member of the group of countries committed to various forms of (harmonised) budget support, Ireland scores well.

On aid modalities- the African respondents to the DFI survey stated a clear preference for programme (particularly budget) support, as noted above. Ireland, although a small donor, is favourably evaluated. The 2003 DAC Peer Review highlights the political difficulties for the Irish Government in engaging in budget support (see below) but gives approval to the triple lock of budget support, SWAPs and area-based programmes. The logic is that continued engagement at a local and sectoral level allows verification of policy implementation and enriches the quality of dialogue which Ireland can therefore hold with authorities at central and regional level.

There is the danger of an inherent contradiction between a continued focus on sectoral support (through SWAPS) and budget support. The host government can find itself involved in SWAPs negotiations in parallel with negotiations in budget support, rather than dealing with those sectors within the overall budget management system. In Mozambique, for example, the continued lack of articulation between donor-driven sectoral policies and budgets is a serious issue for public financial management and a drain on government capacity.

Ireland scores well in the DFI survey with respect to conditionally, possibly because its formal conditionality is largely limited to that included in multi-donor sector and budget support. Ireland tends to argue for streamlining of conditionalities and felxible of conditionalities to avoid these leading to aid disbursal disruption.⁶²

However, Ireland's independence of IFI conditionality is open to challenge. Ireland requires an on-track IMF programme for disbursal of budget support. Given the widely recognised problems with this signal, it would appear that the requirement of an IMF programme is more for reasons of domestic politics than development policy. Ireland's acceptance of the World Bank's role in managing donor coordination through the framework of its Poverty Reduction Support Credit (i.e. in Uganda), with attendant conditionalities, is also a cause for concern. Ireland does not have a formal policy on conditionality, which is a critical weakness given its engagement in budget support.

Ireland is viewed as a reliable and flexible donor. The predictability of Ireland's aid is affirmed by African governments, placing Ireland in the top four along with the Netherlands, Sweden and the UK. However, this is based on experiences with Ireland before the Irish Government reneged on its 2000 commitment to reach the target of allocating 0.7% of GNP on aid by 2007. Between 2000 and 2004 Ireland doubled its expenditure. The pressure on Ireland to disburse would have been significant therefore but Ireland has no official policy on predictability beyond its country strategy plans. Ireland's aid budget is still decided on an annual basis, leaving a very high chance of unpredictability. Ireland urgently needs to adopt a formal policy of predictability and tackle this through a political commitment to an increasing aid budget, guaranteed by legislation.

African governments ranked flexibility of donor funding, particularly to enable reallocation or increase of funding at times of shocks, highest in terms of their needs. They found that the most flexible sources of funding are Denmark, Ireland, the Netherlands, Sweden and the UK. Once again however, Ireland has no explicit policy on flexibility. In tandem with a policy on predictability, Ireland should make clearer to partner governments what flexibility they can expect.

Ireland is a member of the like-minded group of donors⁶³ who are seeking to progress the harmonisation and alignment agenda fast. It has advantages which should allow it to make significant progress in implementing a strong ownership, alignment and harmonisation agenda. Its concentration on a small number of countries is a key characteristic of the Irish aid programme. This should allow it to build quality relationships with partner governments and invest in a deep understanding of the local political, social and economic environment.

However, as a small donor, Ireland faces significant challenges. As the DAC Peer Review Report (2003) points out, Ireland suffers from a barely adequate level of staffing and uncertainty in terms of its aid increases. Ireland simply cannot be a driver of change without sufficient qualified staff at country level and headquarter levels. Indeed, it can be asked whether Ireland's highly-rated performance in harmonisation is driven as much by the need to rely on other donors as by policy choice.

As noted throughout this article, an increased commitment to country ownership, alignment and partnership often require a change in the capacity and focus of donor country offices, making them more autonomous and increasing capacity for policy dialogue. It is questionable whether there is a sustainable level of adequately qualified staff to carry the burden of monitoring and policy dialogue.

The 2003 DAC Peer Review Report points out that "sustaining the Irish commitment to ownership and co-ordination will be increasingly restricted by the centralization of major decisions in Dublin".⁶⁴ Furthermore the DAC peer review finds that, "giving further autonomy to the field mission is constrained by the need to deepen analytical capacity in key areas". Recently, Ireland has increased its technical capacity at headquarters, but this has involved relocating experienced specialist staff such as regional economists from country offices to headquarters. Posts have been filled by nationals at country level, which is commendable and a solid long-term strategy. However, it appears that there is still a lack of capacity in key policy areas, particularly where budget support is moving ahead fast.

There is a sound logic behind delegating cooperation and advisory roles to technical staff in other donors' offices. However, this risks compromising the extent and quality of dialogue Ireland can hold with all relevant development partners, which may prove essential in building constituencies for country- driven alignment and harmonisation.

Overall, it can be argued that Ireland needs a stronger policy presence and voice in the limited number of priority theatres where it has chosen to be present - i.e. programme countries and key international institutions. Experience with the domination of policy space by the World Bank and IMF in particular suggests that this could be a valuable priority for increased policy attention by Ireland. Similarly, the welcome inclusion of policy coherence in the new national trade policy statement requires a government resource commitment. Ireland should develop an increased capacity to assess and maintain ongoing dialogue with relevant actors on the implications of bilateral, regional and international trade negotiations for its programme countries in particular. This would help to break down a shallow and often polarised debate in Ireland with respect to agricultural trade in particular.

One of the critical problems for Ireland will be to resolve its policy dilemmas around budget support in politically-sensitive aid environments. In Uganda, for example, concerns around democratisation, human rights violations, corruption and above all defence spending and the role of the Government of Uganda in the conflict in the Democratic Republic of Congo sparked intense media and public interest in 2003. This led to transferal of $12.7 million in programmed budget support to the ring- fenced Poverty Action Fund.

Ireland does not have a formal budget support policy. This is a key weakness, as illustrated by the fact that a media row - rather than a policy position and clear criteria and conditions - led to a significant shift in how aid was disbursed to a core programme country. A deeper and more considered public debate on budget support, leading to a formal policy position, is necessary. A process to develop a White Paper may offer the opportunity to fulfill this requirement, but caution also needs to be exercised to ensure that the budget support debate does not eclipse other issues of development policy to be outlined in a White Paper.

Conclusions

Donors have long recognised that ownership, partnership and participation are central to development effectiveness. The failure of structural adjustment illustrated the implausibility of expecting developing countries to placidly accept donor-designed policies and programmes which were politically and socially untenable and were not necessarily proven to raise growth.

In recent years, the emergence of the PRSP approach based on ownership, partnership and participation and the MDG imperative could have been expected to lead to rapid progress in an agenda around aid effectiveness which would deliver measurable results, fast. It is apparent that the results are not coming fast and that the development community is still learning about how principles can translate into practice.

This article points to several key lessons: multi-donor budget support is the preferred aid modality of developing countries, but it needs to be implemented from a base of trust, with the highest degree of clarity in assumptions and expectations. Mutual accountability is the keystone for successful partnership agreements, but these have to be born out of explicit statements on the developing country's part as to its preferences with respect to aid relationships.

The proposal that each developing country draw up a comprehensive external finance policy, assess donors on their performance against objective aid management criteria and target those donors where there is a natural basis for partnership, is a potentially powerful model which deserves greater exploration.

Ireland is a well-regarded donor among DAC peers and developing countries, but it faces significant challenges nonetheless. In particular, Ireland's policy base and voice need to be strengthened. It needs to eliminate the uncertainty surrounding its policy on budget support and on future aid levels. Ireland's commitment to partnership is credible, but these critical issues must be addressed before it can justifiably claim a role as a first class donor.

Footnotes

1. Report of the International Conference on Financing for Development March 2002, para 43 (back to text)
2. The Rome Declaration on Harmonisation was issued in Italy on 25 Febrnary 2003 by heads of multilateral and bilateral development institutions, international financial institutions and developing countries. It was intended to build on the International Conference on Financing for Development held from 18 - 22 March in Monterrey, Mexico.(back to text)
3. ODI (2004) (back to text)
4. For a critique of the MDGs and their implementation from the perspective of social and economic justice, see Gold, Lorna, "Are the Millennium Development Goals addressing the underlying causes of injustice? Understanding the risks of the MDGs" in this issue. (back to text)
5. Rome Declaration, para. 3. The Rome declaration endorsed the DAC's Good Practice Guidelines on Harmonisation and Aid Effectiveness (OECD-DAC 2003a). (back to text)
6. The Rome Declaration refers only to streamlining conditionality while the DAC's Good Practice Guidelines on Harmonisation and Aid Effectiveness include the commitment to draw conditionality from the PRS, or equivalent national framework. (back to text)
7. The OECD-DAC carried out a survey of 16 developing countries as part of a needs assessment in preparation for drawing up guidelines for donors on good practice in aid delivery, which were published in early 2003: OECD - DAC (2003a). It subsequently surveyed 14 countries in Africa, Asia, Latin America and Eastern Europe in preparation for a bi- annual review in March 2005 of the implementation of the Rome Declaration: OECD-DAC (2004). (back to text)
8. Development Finance International is a UK-based non-governmental organisation that works with finance ministers and officials in heavily indebted poor countries (HIPCs) in building capacity for management of debt and financial flows: www.dri.org.uk . The report used a methodology which it developed to help governments assess the quality of aid they receive (writren by Alison Johnson Marhew Martin and Hannah Bargawi). (back to text)
9. OECD-DAC (2003a), pp.13-14 (back to text)
10. OECD-DAC (2004), p.2 (back to text)
11. Including Ethiopia, Tanzania and Mozambique: OECD-DAC (2004), pp.3-4 (back to text)
12. OECD-DAC (2003a), p.113 (back to text)
13. See OECD-DAC (2003b), pp. 67-75 for a description of Ireland's experience of partnership. (back to text)
14. OECD-DAC (2003a), p.121 (back to text)
15. OECD-DAC (2003a), p,104. For more analysis on how donors, international financial institutions and developing country governments undermine the principles and practice espoused under the PRSP approach, see CIDSE (2004). (back to text)
16. OECD-DAC (2004), p.19 (back to text)
17. ee pp.7-8 (back to text)
18. World Bank (2004), p. 18 (back to text)
19. Johnson et al. (2004), p. 25 (back to text)
20. IEO (2004) (back to text)
21. See CIDSE (2004a); IEO (2004); ActionAid International Uganda and US (2004). (back to text)
22. OECD-DAC (2003a), p. 118 (back to text)
23. OECD-DAC (2003a), p. 10 (back to text)
24. OECD-DAC (2003a), p.14 (back to text)
25. Johnson et al. (2004), p.29 (back to text)
26. Ibid. Technical assistance and food aid are not subject to the 2001 DAC Guidelines on untying aid in spite of a clear rationale for including them.(back to text)
27. Oxfam (2004 a), p.50 (back to text)
28. OECD-DAC (2003a), p.56, based on calculations by Van de Walk and Johnston. (back to text)
29. OECD-DAC (2004), p. 13 (back to text)
30. Oxfam (2004a), p.53 (back to text)
31. Johnson ct al. (2004), p.30 (back to text)
32. CIDSE (2004) (back to text)
33. Benin, Burkina Faso, Ethiopia, Ghana, Madagascar, Malawi, Mozambique, Rwanda, Tanzania and Uganda (Johnson et al. (2004), p. 20) (back to text)
34. In Mozambique the benefits to the government of budget support are more likely to be stronger government systems than reduced costs of coordinated aid: OECD-DAC (2003a), p.122). (back to text)
35. The average shortfall in budget support is 14% compared to 26% for projects: Johnson et al. (2004), p. 11. Note that this contradicts findings by IMF staff members Bulir and Hamann (2001) who found that there are much larger prediction errors in programme aid than with project assistance, in patt because of the application of conditionality (Killick 2004, p.24). (back to text)
36. Personal communication with die author (back to text)
37. Trocaire (2004b); Martin et al. (2004); Wood, A. (2004); ActionAid International UK (2004)(back to text)
38. A synopsis of recent research carried out by ODI / Oxford Policy Management concluded that the linkages between budget support and positive development outcomes, while plausible, are not automatic and require supporting changes that are only likely to arise from national political processes (ODI Opinion, February 2004).(back to text)
39. Assessing Aid found that conditionality had not been successful in inducing policy reform in a climate where commitment to that reform was weak.(back to text)
40. Reality of Aid Report (2004) (back to text)
41. Trocaire (2004a); CIDSE (2004)(back to text)
42. Bird (2004); IEO (2002); Killick (1998); Killick (2004); Martin and Bargawi (2004)(back to text)
43. Trocaire (2004b); Oxfam (2004b); Oxfam (2004c)(back to text)
44. Trocaire (2004a); Martin and Bargawi (2004) (back to text)
45. Such proposals are derailed in Trocaire (2004b), CIDSE (2004) and Martin and Bargawi (2004). (back to text)
46. Johnson et al. (2004), p.25 (back to text)
47. OECD-DAC (2003a), p.20 (back to text)
48. Johnson et al. (2004), p. 11 (back to text)
49. OECD-DAC (2003a), p.I23 (back to text)
50. Johnson et al. (2004), p.24 (back to text)
51. OECD-DAC (2004), p.I4 (back to text)
52. Oxfam (2004a), p.53 (back to text)
53. OECD-DAC (2004), pp.13-4 (back to text)
54. Johnson et al. (2004) p.24 (back to text)
55. Followed by conditionality, poorer types of aid mechanisms, non- alignment to government priorities and low predictability: Johnson et al. (2004).(back to text)
56. Bird (2004) (back to text)
57. OECD-DAC (2003a), p.25 (back to text)
58. OECD-DAC (2003a), p.26 (back to text)
59. Johnson et aJ. (2004), p.18 (back to text)
60. Ibid., pp.16, 39 (back to text)
61. OECD-DAC Survey (2004), p. 1 (back to text)
62. Personal communication, DCI staff (back to text)
63. The like-minded group includes the Nordic donors, Ireland, the UK and the Netherlands. (back to text)
64. OECD-DAC (2003b), p.68 (back to text)

References

• ActionAid International Uganda and UK (2004), "Rethinking participation - questions for civil society about the limits of participation in PRSPS", Washington DC
• ActionAid International UK (2004), "Money talks - How aid conditions continue to drive utility privatization in poor countries", London
• Bird, Graham (2004), "The IMF and poor countries: towards a more fulfilling relationship", University of Surrey, paper presented to conference on developing countries, global finance and the role of the IMF: Towards a New Relationship?, The Hague: Fondad
• Bulir, A. and Hamann, A.J. (2001), "How volatile and unpredictable are aid flows and what are the policy implications?", Washington DC: IMF Working Paper WP/01/167,
• CIDSE (2004), "PRSP as theatre - backstage policy making and the future of the PRSP approach", Brussels, September
• CIDSE (2005), "Justice not charity: policy recommendations to donors ahead of the G8 Summit 2005", Brussels
• Independent Evaluation Unit-IEO (2004), "Evaluation of the IMF's Role in the Poverty Reduction Strategy Papers and the Poverty Reduction and Growth Facility", Washington DC: IMF
• Independent Evaluation Unit-IEO (2002), "Evaluation of prolonged use of Fund resources", Washington DC: IMF
• Johnson, A., Marrin, M. and Bargawi, H. (2004), "The effectiveness of aid to Africa since the H1PC initiative: issues, evidence and possible areas for action", background paper for Commission for Africa (Discussion Draft), Development Finance International UK, August
• Killick, Tony (1998), Aid and the Political Economy of Policy Change, London: Routledge and OD1
• Killick, Tony (2004), '"Politics, evidence and the new aid agenda", Development Policy Review, 2004, 22 (I), pp.5-29
• Martin, M. and Bargawi, H. (2004), "The role of the IMF in low income countries - study for Swedish Ministries of Finance and Foreign Affairs", London
• Overseas Development Institute (2004), "Harmonisation and alignment in fragile states", 17 December 2004, draft report by ODI for DAC Senior Level Forum on Development Effectiveness in Fragile States, London 13- 14 January 2005
• OECD-DAC (2004), "Results of the OECD-DAC Survey on Harmonisation and Alignment (first draft)", Paris
• OECD-DAC (2003a), "Harmonising donor practices for effective aid delivery - good practice papers", DAC Guidelines and Reference Series, Paris
• OECD-DAC (2003b), "DAC Peer Review - Ireland", Paris
• Oxfam (2004a), Paying the Price - Why Rich Countries must invest now in a War on Poverty, Oxford
• Oxfam (2004b), Undervaluing Teachers - IMF Policies squeeze Zambia's Education System, Oxford
• Oxfam (2004c), The IMF and the Millennium Goals - Failing to deliver for Low Income Countries, Oxford
• Trócaire (2004a), "The other side of the coin...an alternative perspective on the role of the International Monetary Fund in low-income countries", Trócaire Policy Briefing, Maynooth, Ireland
• Trócaire (2004b)"PRSP: lessons learnt - recommendations to the World Bank, IMF and donors for the second generation of PRSPs" Trócaire Policy Briefing, Maynooth Ireland
• Reality of Aid (2004), Reality of Aid 2004 - Governance and Human Rights, IBON, Manila
• Wilks A. and Lefrancois, F. (2003), "Blinding with science or encouraging debate? How World Bank analysis determines PRSP polices", Bretton Woods Project / World Vision UK
• Wood, A, (2004), "One step forward, two steps back - ownership, PRSPs and conditionality", Milton Keynes: World Vision International
• World Bank (1998), Assessing Aid - What works, what doesn't and why, Washington DC
• World Bank Operations Evaluation Department (2004), The Poverty Reduction Strategy Initiative: An Independent Evaluation of the World Bank's Support Through 2003, Washington DC: World Bank