Comparative regional income dynamics: Clustering, scale, and geocomputation

The relationship between regional economic growth and inequality is difficult to ascertain. In order to truly uncover the mechanisms involved in regional income dynamics researchers must move past notions of correlation to address the more intricate issues of directionality and simultaneity. The spatial mechanisms involved in income convergence and inequality have only recently begun to take focus in the literature. A portion of this trend can be attributed to advances in spatial statistics and econometrics, which have helped to diffuse the knowledge of several inferential pitfalls associated with spatial effects in empirical data. The focus on space in economic growth and inequality is further apparent on an applied front, as evidence points to interesting socioeconomic constructs such as poverty traps and convergence clubs. This research examines these multifaceted phenomena using three distinct approaches. The first is an exploratory geovisualization framework which is used to generate new insights into the functional relationship between regional growth and inequality in the presence of spatial dependence. A formal spatial econometric approach is then provided that addresses the role of intraregional spatial clustering and simultaneity in the evolution of regional incomes. This project also examines regional economic change in a relative distribution framework in order to relax the assumptions often underlying the analysis of income data. Monte Carlo techniques are used to evaluate the properties of the relative distribution subject to varying levels of spatial association. Several techniques are then employed to control for spatial effects thereby allowing for a decomposed and robust outlook on regional income dynamics.