Spatial Dependence in the Evolution of Regional Income Distributions

This paper introduces three measures of spatial dependence for use in the analysis of regional income distributions and their evolution. The first builds upon the notion of regional conditioning (Quah 1993), and is derived as a trace statistic from a modified Markov transition matrix. The remaining two statistics are intended for use in a dynamic context and measure the degree of spatial clustering and regional cohesion in income rank mobility. All three measures are applied in an empirical analysis of per capita income patterns in the lower 48 United States over the 1929-99 period.