Using a new instrumental variable strategy, we examine whether bilateral development aid increases military expenditure in recipient countries. The instrument is the interaction of donor government fractionalization and the probability of receiving aid. The dataset includes new data on military expenditure for 124 recipient countries over the 1975–2012 period. When accounting for outliers, our results do not suggest that development aid affects military expenditure in the full sample. However, the effect of aid on military expenditure varies across characteristics of recipient and donor countries, even after excluding outliers. First, aid increases military expenditure in countries that depend on aid and are prone to conflicts. Second, aid provided by coordinated market economies increases military expenditure.
Aid fungibility military expenditure instrumental variables causality