City size, labor productivity and wages in Korea

This study attempts to derive policy implications for spatially-balanced growth through empirical analysis on determinants of regional wage gaps in Korea. Combes et al. [(2008) Spatial wage disparities: Sorting Matters. Journal of Urban Economics, 63, 723–742] suggest that regional wage gaps result from the regional differenc es in workers’ human capital, nonhuman endowments, and agglomeration economies. The current study applies a similar two-stage estimation model to the 2006 cross-sectional data for 4009 workers from the Korean Labor Panel Survey (KLPS) performed by the Korea Labor Institute. Localization economies are positive external effects from the geographic concentration of firms in the same industry. We find evidence of localization economies in our research. The second stage estimation results for the determinants of regional wage gaps show that the average wage is higher in areas that concentrate in a small number of industries rather than in areas that are diversified with many industries. This result is in direct contrast to the findings from Combes et al. [(2008) Spatial wage disparities: Sorting Matters. Journal of Urban Economics, 63, 723–742] who analyze the French labor market data. This difference can be explained by the possibility that in Korea the improved quality of life (e.g., better education services) and/or the increase in job mobility in large diversified metropolitan areas induce workers to accept relatively lower wages in those areas. In order to resolve the bias in the estimation of the agglomeration effects caused by the heterogeneity of workers, we also performed panel regressions of the 2nd 2000 and the 7th 2005 KLPS panel data of 7431 observations. The panel regression results also support our original findings from regressions of the 8th 2006 KLPS data.