The Household Analysis of Comparative Development: An Empirical Replication

Abstract:     
Institutions are clearly exogenous to household decisions. Analysis of institutions can thus in theory proceed from the household level and avoid the problem of endogeneity. This assertion is tested through a replication of Acemoglu, Johnson and Robinson’s (2001) seminal paper. Replication at the household level is successful and with adequate controls IV and OLS household estimates are nearly identical, indicating no endogeneity problem. These results invite further analysis of institutions using household data. Estimates show a smaller effect of property rights institutions than found in the literature. Significant effects of geography, life expectancy and urban location are found.

Number of Pages in PDF File: 27

Keywords: Institutions, Growth, Endogeneity

JEL Classification: O40, P50

working papers series