Does reform work? : An econometric examination of the reform-growth puzzle

Abstract

Why are socially beneficial reforms not implemented? One simple answer to this question (which has received little attention in the literature) is that this may be caused by generalised uncertainty about the effectiveness of reforms. If agents are unsure about whether a proposed reform will work, it will be less likely to be adopted. Despite the numerous benefits economists assign to structural reforms, the empirical literature has thus far failed to establish a positive and significant effect of reforms on economic performance. We collect data from 43 econometric studies (for more than 300 coefficients on the effects of reform on growth) and show that approximately one third of these coefficients is positive and significant, another third is negative and significant, and the final third is not statistically significant different from zero. In trying to understand this remarkable variation, we find that the measurement of reform and controlling for institutions and initial conditions are main factors in decreasing the probability of reporting a significant and positive effect of reform on growth.