Growth and Convergence: An Alternative Empirical Framework

Abstract

The variance of the logarithm, often used to test for s-convergence, does not respect the properties expected for an inequality measure and it has problematic implications for world growth. Empirical tests for b-convergence are so weak that b-convergence can be observed simultaneously when one moves forward or backward in time. Taking into account these weaknesses in past empirical work, this paper proposes an alternative framework in which to analyze the relationships between growth, inequality, mobility, and social welfare. An application of the framework to worldwide and regional data on per capita GDP suggests a lack of convergence at the world level, opposite trends in convergence in various regions of the world, and a fairly low level of mobility or reranking between countries over time.

Keywords: Growth, Inequality, Mobility, Convergence

JEL Classification: O11, O40, D63