Addressing corruption in fragile states: what role for donors?

3. Cross-cutting themes

Coordinate donor responses

The international community is putting an extreme administrative burden on recipient countries under the present aid regime. In fragile states, there is no capacity for dealing with an uncoordinated group of donors.

“Managing aid flows from many different donors is a huge challenge for recipient countries, since different donors usually insist on using their own unique processes for initiating, implementing, and monitoring projects. Recipients can be overwhelmed by requirements for multiple project audits, environmental assessments, procurement reports, financial statements, and project updates. According to the World Bank, developing countries typically work with 30 or more aid agencies across a wide variety of sectors, with each sending an average of five missions a year to oversee their projects.The donors all want to meet with the same top government officials, leaving them with much less time to deal with pressing matters.”[1]

At the country level, donor to donor harmonisation and aid alignment is often in place where there the aid to GDP ratio is high, donors co-finance through general or sector budget support, and where there is a well-developed framework and at least a partial political will to fight corruption (Uganda, Tanzania, Mozambique, Vietnam, Cambodia and Nicaragua, for example). In non-aid dependent countries where large donors have great economic interests (such as Nigeria and Angola), isolationist states (such as Zimbabwe and Myanmar), weak states (such as Tajikistan) and in conflict and post-conflict countries (Sudan, Afghanistan, Nepal and Haiti, for example), frameworks for aid coordination and donor collaboration on corruption are usually poorly developed or non-functional. In these countries the government not only fails to provide a framework for donors to align to, but the donors themselves are often fragmented and sometimes even competing.[2]

This topic has recently received much needed international attention and in some cases we find the limelight of coordination to be intense. Some researchers point out that donors tend to be caught up in discussing today’s operational and security concerns rather than in coordinating and strategic planning for the medium term. Others point out that aid allocation criteria to fragile states are not developed and who gets what and when is unclear. Certainly, some states or regions receive too little or too much aid, making what is available to be distributed unevenly spread.

An emerging lesson is that an absence of government leadership on AC, or an inability to prioritise, as have been seen in countries such as Yemen, can be abated by donor community speaking with one voice. In Yemen, this has brought about promises of country leadership, which is a positive first step. There is therefore an urgent need for donors to prioritise the harmonisation agenda at the country level, with an emphasis on the coherence of donor policies.

Where donors are struggling to find a common vision and their policies cannot be fully harmonised, their objectives must at least be complementary. A starting point in fragile states would be to promote the coherence of non-aid policies and bring together the defence, diplomacy, development, aid, trade and intelligence communities in order to minimise fragmentation and duplication and maximise complementarities and common purpose[3]. The second step is for all the actors in the international community to join in establishing a high-level donor coordination group to develop a clear and coordinated message on corruption to the government; to encourage and monitor the mainstreaming of corruption in the forms of donor assistance; and to encourage greater synergy between donor corruption programmes.

Leadership of “donor coordination groups” should, if possible, be held by one strong donor which will facilitate communication between the government and the donor community at large. Bilateral donors can often serve as leaders, since they are commonly considered less interventionist than the World Bank, for example, and cultivate special relationships with certain countries. However, in most cases it would be natural for the lead to be taken by the main contributor in a certain area. There is also a case to be made for greater donor specialisation on certain sectors and for some donors not to get involved in some countries, instead allowing likeminded donors to take the lead on behalf of a larger group.

Joint efforts among donors can include common assessments, the joint funding of sectors and institutions, common procurement systems and common reporting and financial requirements. Positive steps in this regard have been facilitated by the Budget Support donors. The result has been more and better donor coordination and closer aid alignment. Even donors that do not give budget support will be drawn into coordinated sector policy discussions.

In many “established” difficult partnerships, such as those with Uganda and Mozambique, initiatives like these have been tried out. The donors have done a good job in these countries and have moved far ahead of the international harmonisation agenda. Donor coordination and aid alignment have been successful at both harmonisation and alignment. The result is a much improved dialogue with the government and better aid practices. However, they have not been able to achieve meaningful results in the fight against corruption.

In Uganda, multilateral and bilateral donors have established a collective position on AC, which has been effective in negotiating with the authorities. Acting with “one voice” has not only reduced the transaction costs for the government of dealing separately with each donor, but has also enabled donors to increase their leverage vis-à-vis officials in order to get some reforms going. Nothing indicates, however, that this increased political capital has been translated into a stronger emphasis on fighting corruption.

In Kenya, a much less donor-dependent country with less donor cooperation and unity, more significant change has nevertheless taken place. This indicates that donor coordination per se can do little more than reduce transaction costs between donors and vis-à-vis the government – at least if no aid conditionality is in play. Other factors embedded in national politics are clearly more important.

Use conditionality with care

There is a clear and inescapable tension between country ownership on the one hand and donor priorities and conditionality on the other. The impact of aid conditionality on persuading governments to address corruption has been questioned, mainly because aid flows often continue even if conditions are broken. In the front line against the war on terror, donors continue to tolerate corrupt practices (warlords and nepotistic ministerial appointments in Afghanistan, for example), and where countries accept the policies of international financial institutions, such as trade liberalisation and macroeconomic reform, pressure is effectively taken off[4]. At the country level, conditions have also often been too complex, shifting and numerous to be practical.

Positive outcomes from aid conditionality would be expected in heavily donor-dependent countries where donors are well organised. The evidence from Uganda and Ethiopia tells a different story, as these regimes are willing to sacrifice aid in order to hold on to power. The only progress made has been in cosmetic changes, such as the prosecution of one or two individuals of minor importance. In recognising this, the donor community has been asking these countries for governance reforms linking aid to the expansion of civil liberties and administrative reform. This governance conditionality is soft by nature, and with large-scale budget support in place, these countries know that aid money will remain on the table. It remains to be seen whether governance-focused conditionality will be effective.

The same concern is raised over policy conditionality. It does not appear to be any more successful for anti-corruption than for other purposes, unless it reflects government's real priorities. A broad consensus has emerged from studies of policy conditionality that policy changes cannot be "bought" in this way. Domestic political considerations emerge as the prime factor in determining economic and political reform, and donor pressures do not easily influence these. Kenya under Moi in the 1990s is an example. Insofar as the new Kenyan government is now making a more serious AC effort (a commitment which has brought back most donors), genuine electoral competition and an active civil society appear to have been the crucial elements in this change, not conditionality.

There must, however, be a line in the sand for what the international community can accept. Regimes that operate as a de facto mafia, for example, should not be supported.  Le Billon states that the international agencies should be ready to suspend non-humanitarian aid to avoid consolidating the power of corrupt politicians.[5] However, such direct measures should only be undertaken when there is a clear consensus in the international community as to the status of the regime, and only when there is strong donor coordination on the ground.

In these worst cases the donor community could also consider restrictions on travel for regime dignitaries and their families, expulsion from international fora or even more drastic means. But this kind of pressure should only be applied with the greatest care as it may backfire into isolationism and repression.

Another lesson, however, is that external pressure can be successfully applied in countries which aspire to join international fora such as the WTO or EU. A good example is Albania, where a persistent campaign from the international actors in Tirana informed everyone, including the electorate, that the lack of progress in talks was due to the government’s unwillingness to address these issues in a genuine way. A combination of diplomatic methods and public shaming did not induce the former government to change its position. This was, however, picked up by the electorate, which became aware that their low living standards were linked to 1) corruption 2) the sitting government. In July 2005 the governing party lost power in a campaign that was fought by the opposition on an anti-corruption ticket.

The way out of the conditionality conundrum is to provide partner countries with credible medium-term commitments, and to minimise the variability of aid over the partner government’s fiscal cycle. Donors should programme aid through a multi-year framework, aligned with national budgeting and programming cycles. They should fully disclose both the expected flows and any triggers for their reduction or suspension if the terms of the agreement are not respected. The aid agreement should be a public document.

Ideally, donors and partner governments should move from binary “yes/no” to graded “more/less” types of trigger, and increasingly base these on monitorable results set in the PRS. Donors should use partner governments’ performance assessment frameworks to monitor progress against the PRS. Indicators must, however, go beyond creating strategies and passing laws to implementing real reforms that increase transparency and establish real accountability in key government functions and sectors. In Afghanistan, for example, the single “anti-corruption” indicator in the National Development Plan is “ratification of UNCAC by the end of 2006”. This overly general, much less unrealistic, prescription is highly unhelpful to the case for reform.

Governments bent on improving their image and increasing aid flows will be keen to help develop some “evidence-based” indicators. Successful use of outcome-based conditionality may help build a consensus for broad-based sector-wide and general budget support among key donor constituencies.[6]

Donors should also look for “sideways” entry-points, relying more on what actually matters for government. Government may not care that much about the poor (aid money), but it might care about economic growth, about FDI, and about its international reputation. Threats to trade and the voice of the private sector are potentially much more powerful levers than donor funding, especially in countries that are not aid-dependent.

Focus on results-based management

In difficult partnerships generally, but more specifically in emergency and post-conflict situations, donors are faced with the combined challenge of the need for rapid disbursement while having to deal with poor oversight systems and inadequate monitoring institutions. In learning from past experience it is crucial to budget for, and establish, oversight institutions properly from the outset. Steven Radelet emphasises the need to demonstrate the effectiveness of aid through improved monitoring, evaluation and results-based management.

“Aid programs should aim to achieve very specific quantitative targets, and decisions about renewing or re-allocating aid going forward should be based on those results. There are three basic objectives: (1) helping donors allocate funds towards programs that are working; (2) detecting problems at an early stage to help modify and strengthen existing programs; and (3) improving the design of future programs. Stronger monitoring and evaluation would help improve principal-agent relationships so that aid agencies have clearer incentives and taxpayers have better information about the impact of aid on its intended beneficiaries”.[7]

Building reliable monitoring and evaluation processes should thus be an integral component of implementation. The literature on performance management is clear: what is measured is what gets done. Emil Bolongaita writes in this study of the Philippines:

“In my study of the Office of the Ombudsman, I found that the principal performance measure used by the organization to measure itself and its staff was case disposal rate – that is, the rate at which they are able to act upon complaints relative to the number of complaints it receives. Acting upon a complaint could mean that it is dismissed, forwarded to another office, or filed as a case in court. Nowhere in the organization’s plans or programs entailed measuring itself based on outcome-oriented standards, such as conviction rate, rate of winning appeals, citizen satisfaction of Ombudsman performance, etc.”[8]

The emphasis on monitoring should not sideline the state but instead provides an incentive for it to improve its performance while diagnosing key problems. But if there is too much pressure to show short-term results, the risk is that donors will get back into the driving seat and that the longer-term objectives of strengthening local institutions and policy-making processes will be undermined.

Orienting technical and financial assistance towards checking on the sensible use of funds should not, however, become overly focused on the Ministry of Finance and a few selected ministries in the capital. There is only so much an accountant can find out. The monitoring of public sector expenditures and projects should include views from users of public services and investments. This may take the form, for example, of establishing local citizen oversight boards within existing health programmes to keep officials accountable for budget expenditure in hospitals and clinics, and of creating internal audit units at the municipal level to monitor the procurement process within local governance programmes.

Publication of all monitoring efforts may serve to increase democratic accountability. These same monitoring efforts could also be directed towards international aid agencies, increasing the accountability of both national and international actors.

Donors should encourage peer review mechanisms like NEPAD at an early stage and be active in establishing a monitoring mechanism under UNCAC that will be meaningful in the sense that countries would feel compelled to take the necessary steps to be in compliance.

Be politically sensitive

Donors tend to see elites only as part of the problem (vested interests getting in the way of reform), rather than as part of the solution. Steps should be taken to maximise the chances of finding common ground with local power holders and opinion formers. A starting point would be for donors to avoid jargon and to look for language that removes unnecessary obstacles to dialogue. Particularly in countries with an unstable political environment (such as Zimbabwe), AC activities may be interpreted as undermining those in power.

At the same time, working with corrupt governments can be used by the regime to gain legitimacy and entrench its position. In-country donor AC collaboration on priorities and initiatives therefore needs to be undertaken with care, with actions directed towards upholding what the state is formally committed to: for example, statements by political leaders, international AC conventions and laws relating to the transparent and accountable use of public funds.

Post-conflict deal negotiators and dignitaries are more comfortable speaking about accountability than corruption, although the former is part of the solution to the latter. Packaging is therefore important and needs to be taken seriously. Two identical anti-corruption initiatives can be presented in different ways depending on the context. However, some agency staff feel that caution has overcome the international actors. Two examples:

“I would like to see donors being more outspoken on the issues of corruption. If donor ministers and senior staff spoke out about corrupt issues and practices, it would give the general population of all countries a greater sense of the importance of why systems and processes free from corruption is the ideal”.

“A more direct involvement in the political sphere has to be accepted. The pure a-political technocratic approach can no longer be maintained in any segment of international cooperation: all policies have political dimensions that would be better be brought into the open and be discussed”.[9]

Protect aid money from diversion

Donors must be willing to accept risk and engage even in the most fragile states. But operating in corrupt environments does not exclude taking responsibility for the use of taxpayers' funds. Essentially, donors could support each other in not funding activities vulnerable to corruption if a government is unwilling to address the problem. Another proposition is to focus more on knowledge than on finance, and on using grants rather than loans where feasible.[10]

The mechanisms for aid delivery are clearly crucial in fragile states. Donors must select aid modalities according to country circumstances, in a way that builds government capacity to provide services but at the same time provides adequate safeguards in high-risk environments.

Direct Budget Support (DBS) should remain the preferred modality, but only in those cases where its long-term viability cannot be put into question[11]. This will rule out most fragile states. The drivers-of-change framework has the following sensible advice for the budget support donors if a situation deteriorates:

1. Slower increases in the share of general budget support as a proportion of total aid commitments
2. A reduction in the overall amount of budget support
3. A switch to earmarked budget support for the social sectors
4. Ultimately, delay or cancellation of tranches of scheduled budget support

In the face of blatant abuses of donor funds some hard and uncomfortable choices have to be made. That means retaining some control and not automatically accepting everything a government decides. For example, donors at one point rightly objected when the government of Tajikistan wanted to prioritise infrastructure construction [highly vulnerable to corruption] over the social sectors.

In many countries the third option of sector support, preferably linked to specific service delivery targets, could be the next best option. Aid fungibility would still be an issue. Earmarked support may free up resources that the government can spend as it wishes, but it would be harder to question the legitimacy of the aid being given.

In some contexts DBS is just not feasible and donors must go with third best options of providing project support in order to provide better safeguards against the diversion of funds. Project support should, however, only be considered in specific cases where state fragility or political concerns prevent the use of other modalities, or for specific support to capacity development efforts. The reason is that ad hoc donor funding of projects may reinforce relations of patronage. If projects are the preferred modality, they must still be in the budget. A complete by-pass of government systems should be actively avoided unless there is a serious breakdown of trust.

The assessment and management of fiduciary risk is becoming increasingly important to the donor agencies. Donors must be willing to accept high levels of risk in a transitional period and use pull factors associated with budget support in a positive way. The move towards DBS should be forcing donors to become hands off in order to spur national ownership. Most governments would like to see higher shares of aid given as DBS, which then can be used to propel the anti-corruption drive to the top of the agenda. Focusing on which absolute standards government needs to meet before proceeding to higher levels of DBS might incentivise domestic leaders. At a minimum, donors must require that a credible reform programme is in place. In effect, however, we have to face the fact that DBS can represent another form of conditionality.

[top]

Footnotes

[top]