Aid allocation within countries is often thought of as a strategic action by the incumbent leaders to further their own goals. Theoretically, however, the effects of aid may be either positive or negative and the empirical evidence is limited. By matching geo-coded data on aid projects to 101 792 respondents in five waves of the Afrobarometer, we investigate the effects of aid on incumbency support using project fixed effects. We estimate the effects for World Bank aid and Chinese aid separately and find positive effects for the former and no robust effect for the latter. For neither project donor do we find effects on turnout and that aid is not targeting areas with previously higher incumbency support. We find little support for the notion that economic voting is driving the result as individuals self-perceived economic conditions are not affected. The positive effects for the World Bank aid projects seem to be mediated by trust in the politicians, whereas we find no effects of Chinese aid on trust. (C) 2019 Elsevier Ltd. All rights reserved.
