What Determines the Quality of Institutions?

Abstract

Various papers trying to explain institutional quality have come to conflicting conclusions. This paper shows three things: First, openness is positively and pretty robustly associated with institutional quality. To minimize selection bias we use data sets with the greatest cross country coverage, though we also test the significance of the variables for smaller sample sizes. The paper confirms that both natural and policy measures of openness are important. Concentration of trade in natural resource exports continues to be associated with poor institutional quality after accounting for openness in trade. Second, it shows that social” variables such as income inequality or ethnic diversity are not associated with institutional quality. The significance of th e inequality variable disappears when continent dummies for Latin America and Africa are included. Third, it shows that features of specific institutions such as a variable accounting for checks and balances in the political system, and freedom of the press are positively associated with overall perceptions of institutional quality and that these finding hold strongly across different data sets and samples even after we control for those variables commonly used in the literature.