The typical identification strategy in aid effectiveness studies assumes donor motives do not influence the impact of aid on growth. We call this homogeneity assumption into question, first constructing a model in which donor motives matter and then testing the assumption empirically.

The defense-growth nexus is investigated empirically using longitudinal data for Guatemala and allowing the effect of defense spending on growth to be nonlinear. Using recently developed econometric methods involving threshold regressions, evidence of a level-dependent effect of military...