Spatial persistence of Japanese unemployment rates

•This study investigates whether Japanese unemployment rates are spatially persistent.•Regional interdependence in unemployment rates is analyzed by spatial statistics.•Municipal unemployment rates show significant positive spatial autocorrelation.•Some groups of contiguous municipalities form persistent unemployment clusters.•Unemployment clusters show heterogeneities among different genders and age groups.

Regional interdependence can make a spatial structure persist, resulting in the formation of unemployment clusters. As such, this study investigates whether Japanese unemployment rates are spatially persistent using a six-period municipal panel dataset built by integrating the country's 1980-2005 population censuses. In order to capture spatially interdependent aspects of regional unemployment, this study takes a spatial statistical approach using the Moran's I and Getis-Ord Gi*(d) statistics. This study finds that municipal unemployment rates in Japan show significant positive spatial autocorrelation. Furthermore, this analysis provides evidence that some groups of spatially contiguous municipalities form unemployment clusters, with heterogeneities among different genders and age groups.