Does foreign aid help to promote democracy in developing countries? This question has gained critical importance after years of donors’ failed attempts to improve institutional capacity and the quality of governance in aid-recipient countries. A wave of recent research has illuminated potential detrimental effects of foreign aid on economic development and governance. One common critique of such development projects lies in the lack of donor coordination. Scholars and practitioners alike argue that donor fragmentation—characterized by the presence of multiple donors or funding agencies with divergent policy objectives—deprives both donors and recipient countries of necessary resources and commitment to institutional building, which is deemed critical for injecting democratic accountability into political life.
Evidence for the potential impact of donor fragmentation on democracy and the quality of governance, however, has been largely anecdotal. To my knowledge, there is no study in the scholarly literature to expose the hypothesized causal linkage between donor fragmentation and democracy to rigorous statistical testing. In this paper, I argue that foreign aid and political conditionalities will not be effective in promoting democracy when aid flows are fragmented across too many donor countries and funding agencies. To test this hypothesis, I analyze panel data from up to 91 countries for the period 1972-2007. I construct indices of donor fragmentation based on the PLAID/AidData database, which provides comprehensive data on aggregate amounts of aid commitments made by both bilateral and multilateral donors. To explicitly address the issues of endogeneity, which blurs the causal relationship between foreign aid and the level of democracy, my statistical analysis employs Arellano and Bond’s (1991) method of moments (GMM) estimator to instrument these endogenous regressors in my regression models.
