INSTITUTION AND DEVELOPMENT REVISITED: A NONPARAMETRIC APPROACH

The paper uses nonparametric methodology to examine the role of institutions in understanding differential levels of development across countries. By using the Li-Racine (2004) generalized kernel estimation methodology, our paper allows a deeper look into the impact of institutions on development. The analysis is carried out for a set of 102 countries over 1980 to 2004. Similar to parametric results established in the literature, the nonparametric analysis lends further support to the view that institutions matter in the development of countries in the context of economic policies and geographic factors. There is minimal evidence to suggest that institutions have a negative impact on development. Our results further indicate (a) parametric estimates suffer from misspecification bias and (b) the impact of institutional quality on development quality is heterogeneous across countries and time periods.