Unemployment and Entrepreneurship in the Mid-Atlantic Region of the United States: A Spatial Panel Data Analysis

Abstract

Studies of whether unemployment leads to more or fewer firm births and whether firm births reduce unemployment have produced mixed, inconclusive, and even conflicting results. A group of studies have attempted to reconcile these findings by exploring the possibility that the assumed linear relationship may break down and become non-linear, as might be the case if the relationships were space- and time-variant. However, these studies have also produced inconclusive results. The current study extends the analysis in this small but growing literature from the usual single- and multi-country level to the sub-country level while paying particular attention to the role of spatial effects in the firm creation process. We do this by combining spatial exploratory data analysis and spatial econometrics based on panel data for 365 counties in the Mid-Atlantic region from 1999 to 2010. The results show that spatial effects play a crucial role in this process and that ignoring them substantially underestimates the long-run effects of unemployment on firm births. The results also reveal a robust and statistically significant inverted U-shaped relationship in which rising unemployment raises the firm birth rate up to a certain threshold. In addition, a U-shaped relationship where rising firm births lower unemployment for a period but subsequently increase it is observed. These non-linear results help explain some of the conflicting findings in the literature.