This paper analyses econometrically the determinants of the violence that some governments inflict on some minorities within their own countries. It first discusses the domestic-level diversionary theory of war of Tir and Jasinski (2008). It argues that their econometric results do not allow them to draw the conclusions that they present, and offers a more satisfactory approach. Using the same data set, but a more satisfactory approach, we show that the domestic-level diversionary theory of war is strongly rejected by the data. The paper then presents a more satisfactory approach, suggesting that foreign aid has some influence on this type of violence, which it reduces marginally.