Governance Indicators and the Level of Analysis Problem: Empirical Findings from South America

Abstract

Studies of the link between state capacity and development often utilize national-level governance indicators to explain fine-grained development outcomes. As capacity in some bureaucratic agencies matters more for these outcomes than capacity in others, this work proxies for capacity within the set of relevant agencies by using a measure of ‘mean’ capacity across all agencies in a polity. This practice is problematic for two reasons: (1) within-country, cross-agency diversity in capacity often overwhelms the variation encountered across public sectors considered in their entireties; (2) national-level reputations for capacity are not particularly informative about differences in capacity in functionally equivalent agencies in different countries. The article draws on the author's survey of public employees in Bolivia, Brazil and Chile to establish these points.