Aid, taxes, and government spending: A heterogeneous co-integrated panel analysis

Abstract

A substantial amount of aid to developing countries is given to the government, or goes through the budget, meaning it should have an impact on government fiscal behaviour (particularly on government spending). The few existing empirical studies on the effects of aid on government spending neglect variable time-series properties, cross-country (recipient) heterogeneity, and the potential for cross-country correlation. This paper examines the impact of foreign aid and taxes on government spending for a sample of 69 developing countries over 1980–2013, taking account of dynamics characterizing fiscal data, cross-country heterogeneity, and the distorting impact of cross-section dependence. Our econometric approach addresses these problems by applying the Pesaran (2006) common correlated effects mean group estimator. We show that spending, net aid (as well as variants including grants and loans), and taxes comprise an equilibrium (cointegrated) relation. Our results provide robust evidence of a positive, long-run (as well as short-run) association between aid and spending. On average, the aid coefficients are positive but smaller than the tax coefficients, indicating that, in the long run and short run, taxes have a stronger association with expenditures than aid.