In a new World Bank working paper, Steve Knack and Lodewijk Smets examine whether untying aid and defragmenting aid are conflicting or complementary objectives. Conventional wisdom holds that tying aid, or providing aid on the condition that recipients purchase goods and services from the donor country, inflates costs by 5% to 30% and diminishes aid effectiveness. Policymakers and development practitioners have also coalesced around the notion that donor fragmentation -- the number of donors distributing aid in an individual country -- should be reduced. Donor fragmentation is thought to increase transaction costs for recipient governments, thereby reducing the overall impact of aid.
The Paris Declaration urges donors to untie their aid and to establish a more effective division of labor. However, researchers have questioned whether these practices may be counterproductive if pursued simultaneously. One hypothesis is that as the number of donors in a given developing country shrinks, the remaining donors will face less competition and use their increased influence to pursue non-developmental interests, such as tying aid to the purchase of goods and services in developed countries.
Knack and Smets pose an alternative hypothesis. They suggest that possessing a higher share of the "aid market" may help donors deliver aid more effectively. In a developing country where many donors have a local presence, no single donor can claim credit for improved development outcomes. This, Knack and Smets argue, creates a collective action problem. As the old adage goes, "if everyone is responsible, no one is responsible." Conversely, if one donor (or a few donors) dominate the aid market in a given recipient country, it has (they have) a greater stake in providing predictable aid flows and improving development outcomes.
Knack and Smets test these competing hypotheses with project-level data from AidData on aid tying, the number of donors with a presence in a recipient country, and each donor’s share of total aid to a recipient country. They find no empirical support for the claim that higher concentrations of aid allow donors to exploit their influence by tying more of their aid. Instead, they find that as donor fragmentation declines, individual donors are less likely to tie their aid. These results suggest that efforts to untie aid will not conflict with efforts to encourage donors to establish a clear division of labor.