Best practices in foreign aid: Do donors measure up?

By Meredy Throop

Have international aid donors delivered on their rhetorical promises to improve aid practices? For the most part, no, according to a new piece about aid effectiveness—or perhaps more aptly, aid ineffectiveness—by William Easterly and Claudia Williamson of the Development Research Institute.

Their paper, Rhetoric versus Reality: The Best and Worst of Aid Agency Practices, measures donors’ adherence to five “best practices” that have been prominently featured in the Paris Declaration, aid agency documents and academic literature alike. These include: (1) maximum transparency; (2) minimal overhead expenses; (3) specialization of programs by sector or country; (4) delivery through more effective channels; and (5) selective disbursement of aid based on poverty and good government. The findings for each of the five indicators are summarized below and explained in great detail in the paper.

Although the paper finds poor overall performance and little to no evidence of improvements in specialization, fragmentation and selectivity, the Global Fund to Fight AIDS, TB and Malaria was a real outlier, scoring highly across all indicators and donor categories. Not only is the Global Fund leading the aid community on transparency, specialization, and selectivity, but it has the lowest overhead costs for any multilateral or UN agency. For every employee, the Global Fund disperses over $5 million of highly specialized aid dollars to low income, less corrupt recipients.

The UK performs best among bilateral agencies, followed by Japan, New Zealand and Germany. The US falls below average, and the Scandinavian agencies rank surprisingly low. The UK disperses $4.4 million aid dollars per employee, publishes budget and staffing information on the website, fully reports aid flows to the Organisation for Economic Co-operation and Development (OECD), and does not tie any of its aid. (In 2001, the UK set an important precedent for the donor community by untying all development assistance.)

The worst performing agencies are all bilateral donors: United States, Portugal, Sweden, Switzerland, Spain, Belgium, Finland, and finally Greece. The US ranks highly on transparency, but, despite decades of criticism, still ties a quarter of all foreign aid, is a leading exporter of food aid, and does not select recipient countries based on poverty or good governance. Given this poor performance on ineffective channels on selectivity, Easterly and Williamson hypothesize that domestic lobbying and national security interests “dominate the politics of American aid.”

Transparency

Easterly and Williamson use two indices to measure transparency. The first transparency index is based on the extent of data reported directly to the OECD in 2008. Bilateral agencies reported more data to OECD than both multilateral and UN agencies.

The second transparency index, overhead costs, measures four categories of overhead costs including: international staff, administrative expenses, salaries and benefits, and total development assistance disbursed. The study found that despite rhetoric surrounding the international aid transparency process, most bilateral agencies fail to report budget and operations data, and few responded to the researchers’ direct request for information. Two agencies (MOFA Japan and France’s DgCiD) failed to report any data whatsoever and three other agencies, all Scandinavian, fall below the half-way benchmark. Though multilateral agencies made the most non-OECD development assistance data publicly available, the UN agencies performed the worst, with zero reporting for the UNDP.

In total, five bilateral agencies—including the US and UK--got a perfect overall score on transparency, as did four multilaterals and one UN agency. The worst performers were all UN agencies: the FAO, UNDP, UNTA and UNIFEM. One of the few positive trends emerging from the study was the indication that both bilateral and multilateral agencies are becoming more transparent over the long-run. However, the authors still claim that transparency is “shockingly poor” in most aid agencies.

Overhead

The second best practice, low overhead costs, was measured according to: (1) ratio of administrative costs to ODA (official development assistance) or ODF (official development financing); (2) ratio of salaries and benefits to ODA; (3) and the level of ODA or ODF disbursed divided by numbers of employees. Bilateral agencies were found to have lower salaries and overhead costs relative to lending than multilaterals, which in turn have lower cost rations than UN agencies. The worst performers were the UNDP and UNFPA, which were found to spend more on administrative costs than on aid disbursements.

Specialization

Specialization, the converse of fragmentation, has been a key area of reform to emerge from the Paris Declaration and Accra Agenda for Action. According to Easterly and Williamson, “Most agree that there have been too many donors in too many countries, stretched across too many sectors or projects.”

Given its mission (Palestinian refugees), UNRWA ranked at the top of specialization by country, as did all regional development banks, potentially distorting their overall scores. Similarly, some agencies that specialize by sector, like the UNFPA (population and reproductive health) or the WFP (food aid) received high ratings. However, fragmentation was rampant among bilateral agencies that lack a specialized mission. The political economy incentives for fragmenting aid, such as appeasing sector and country lobbies, are likely responsible for the stark lack of progress in this area.

Selectivity

The fourth measure is selectivity. To measure selectivity, the researchers positively scored aid to low-income countries, and negatively scored aid to corrupt or “unfree” countries. Countries were classified as free if they received a Polity IV democracy score equal to 8, 9, or10, where 0 represents autocracy and 10 is fully democratic.

The World Bank received the top score on selectivity, closely followed by the Asian Development Bank and the Global Fund. These scores reflect a commitment to mobilizing aid primarily for “non-corrupt” poor countries. Thought the jury is still out on how to support impoverished citizens in corrupt societies, UN agencies such as the UNDP and UNICEF were penalized for prioritizing poverty selection criteria above good governance. At the other extreme, Japan and the Caribbean DB channel aid exclusively to non-corrupt democracies, regardless of economic need. The United States is one of the biggest losers on the selectivity measurement, doing badly on both poverty and governance scores. This likely reflects the primacy of narrowly-defined foreign policy objectives when determining aid recipients.

Findings confirm that donors did not substantially change aid disbursement in response to democratization. Instead, the significant decline in the share of aid to autocratic countries after 1990 is driven by governance transformations within recipient countries. Easterly and Williamson find that the negligible changes in aid shares to corrupt countries reflect the most egregious contradiction between rhetoric and actual outcomes.

Ineffective channels

The final measure used is ineffective channels, which are defined as the share of aid that is tied, food aid, or allocated as technical assistance. Tied aid--when a certain percentage of aid must be spent on the donor country’s goods or services--and technical assistance are widely seen as a way for donor countries to promote their own commercial interests above the development needs of recipient countries. Food aid is also recognized as an inefficient way to provide assistance, as it frequently undermines local and regional agricultural markets.

Perhaps most disquieting, new data reveals that the US still ties a quarter of all foreign aid.  The US is also one of the three largest donors of food aid, along with the EC and Australia. Though there is a positive long-run shift away from food aid and instead towards cash for local and regional purchase, the US agricultural lobby continues to resist this change. Globally, food aid has gone from 9 to 1 percent of aid from 1979 to 2008. Thus, the combined reduction in food aid and tied aid constitute the most positive reform trend overall.  

In conclusion authors find that not only does donor rhetoric fail, by and large, to match reality, but there is little to no evidence of improvement in doing so.

Meredy is the Policy and Advocacy Coordinator for Partners In Health's Institute for Health and Social Justice.