Bias and Careers: Evidence from the Aid Effectiveness Literature

We investigate whether estimates of the effect of aid on growth are influenced by authors' careers. We collect data on the careers of 190 authors and apply meta-regression analysis to investigate the impact of authors' age and tenure status on the reported magnitude of aid effectiveness, and on the degree of selectivity in which results are reported. On average, authors without tenure report much larger effects and they also exhibit substantial publication selection bias. These findings are consistent with differences in publication incentives between tenured and non-tenured authors. Older non-tenured researchers report the most biased findings in this literature. One explanation for this latter result is these authors' links with aid agencies.