Initial Diagnostics of a Regional Adjustment Model

Abstract

Adjustment models are used increasingly to analyze population and employment changes in regional economies. However, questions remain about the most appropriate geographic scales and time lags for these models. In this paper we estimate a well-known adjustment model for a recent 25-year period in the USA. Regional population and employment changes (levels and densities) are examined at three scales (states, Bureau of Economic Analysis regions, counties) using various time lags (one to ten years). Two-stage least squares regression estimates, based on Regional Economic Information System data running between 1969 and 1994, are generated and discussed. Analysis is restricted to the core relationships between population and employment; the roles of other exogenous variables, normally included in adjustment models, are not considered. Instead, concern is focused on issues such as stability and directional causality of the interacting population—employment systems. Some brief suggestions regarding future research conclude the paper.