Comment on Lof, Mekasha, and Tarp (2014)

Abstract

In Nowak-Lehmann et al. (2012), we used time-series methods to investigate the impact of aid on per capita GDP. Lof, Mekasha, and Tarp (LMT, 2014) criticize our econometric approach, our interpretation, and our data-handling procedure which lead to a large share of missing observations in some specifications. Using a different time-series approach, a different aid variable, and a different sample, they claim to find a positive effect of aid on income, which contrasts with our own results. In this comment, we first explain why we disagree with LMT’s critique of our econometric method and show that our results do not depend on our way of dealing with missing data. Second, we show that the methods used by LMT are unsuitable and rely on similarly problematic data-handling procedures. Supplementing their approach with appropriate cointegration and causality tests shows that there is no robust effect of aid on income.