Better Data Could Inform This Debate
Van Der Veen moves from assertion to simple empirical poking and shows that for the projects with the largest dollar amounts, donors have worked in conjunction with official authorities.
My colleague at William and Mary, Maurits Van Der Veen, recently blogged about a post by William Easterly and Laura Freschi published in the New York Review of Books. According to Easterly and Freschi, western donors are propping up dictators by providing aid in a manner that allows recipient governments to reward political supporters and punish the opposition.
Van Der Veen mostly agrees with their conclusions, but uses AidData to sort projects by size and then concludes, "Searching the database for aid projects in Ethiopia in the past few years, and then sorting by declining project size shows — not surprisingly — that all of the largest projects take place in conjunction with the official authorities." Those projects are displayed below and can be sorted by donor, year, dollar amount, etc...
For the entire Van Der Veen Post, click here.
Easterly and Freschi are probably correct when they assert that donors have not bypassed the regime in order to deliver aid to local NGOs or directly to recipients. Van Der Veen moves from assertion to simple empirical poking and shows that for the projects with the largest dollar amounts, donors have worked in conjunction with official authorities. But what all these researchers really want/need is accurate information on the implementing agency or the "channel of delivery" for these projects. That would allow them to determine what proportion of projects and what proportion of dollars are delivered through recipient government agencies. To date, the majority of projects in AidData (and the OECD's CRS database) simply do not have information about whether these projects were implemented by private contractors, western NGOs, local NGOs, local government officials, national government officials, or staff members of the donor agency. Having this information would allow researchers to definitively answer such questions about aid allocation and related questions about aid effectiveness.