Which Institutions Matter Most for Growth

In this paper we revisit the debate over the relative contribution of different types of institutions in economic growth. We find that the existing study (Acemoglu and Johnson 2005) linking property rights institutions and contracting institutions to long-run growth suffers from identification problems. It appears that the historical instruments used to predict the contribution of history in shaping current institutions are as good a predictor of current schooling as they are for current institutions. We use GMM and control for schooling to tackle these problems. The results show that strong 'market creating institutions' characterized by the adequate protection of private property and contract enforcement are growth enhancing. The 'market stabilizing institutions' that ensures macroeconomic stability and does not undertake distortionary policies boosts investor confidence and are also good for growth. We notice that there is a growth maximizing level of 'market regulation' beyond which it increases red tape and kills the incentive for investment. The effect of 'market legitimizing institutions' is statistically insignificant