Do Superstar Employers Depress Labor Shares? Evidence from Counties in the United States

Abstract This paper investigates the relationship between the presence of ''superstar" employers--establishments with exceptionally large employment relative to sectoral peers--and labor shares at the county level across the United States. Motivated by a growing literature on the declining labor share and the rise of dominant firms, the analysis develops a novel methodology that constructs synthetic employment distributions using County Business Patterns data to identify statistical outliers within county-sector-year combinations. Fixed effects panel regressions are employed across three spatial scales (Georgia, the Southeastern region, and the nation) to assess whether either the existence or the prevalence of superstar employers is associated with changes in labor’s share of income. The findings are strikingly consistent: little evidence emerges that superstar employers reduce labor shares at the county-sector level. Where significant effects are detected, they are typically positive, particularly in service sectors. Robustness checks confirm these results across alternative definitions and specifications. These findings challenge prevailing assumptions that attracting or retaining large employers necessarily leads to declining worker income shares at the regional scale. Policy implications and considerations for the measurement of labor market concentration in local contexts are discussed. The approach demonstrates the utility of combining public establishment-size data with outlier detection to study monopsony and labor market power in settings where detailed firm-level data are unavailable. Overall, the results suggest that concerns about the effect of large-scale employers affecting the local distribution of income are unfounded. JEL Classification: J31 , R11 , R50