Political Institutions and Schumpeterian Growth: A New Estimation Technique.

This paper presents a statistical technique to estimate how the effects of political institutions on economic performance vary across different sets of circumstances. Recent work using a Schumpeterian growth paradigm (Aghion and Howitt, 2005) suggests that there exist different stages of economic growth and that the effect of any given set of institutions potentially varies across these stages. Catching up to the technological frontier, for example, requires that countries absorb and implement existing technologies, while continued growth at the technological frontier requires innovation of new technologies. Countries must shift from "implementation-enhancing institutions" to "innovation-enhancing institutions" in order to sustain growth. Additionally, there exists a third set of countries where institutions are so poor that neither implementation nor innovation takes place systematically. Use of this growth paradigm thus offers a method for developing context-dependent predictions concerning the relationships between political institutions and economic performance, representing a significant departure from typical approaches that apply a single economic growth model across all countries and times. Using a dataset covering 82 countries from 1960-2000, the statistical technique developed in this paper simultaneously estimates a separate set of coefficients on the key explanatory variables for each of these three stages of growth along with a weighting function that that indicates the extent to which each model applies to a particular country-year. The result is a much fuller description of the ways in which political institutions affect economic outcomes. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]