Source: Authors' analysis of annual, state-level aggregations of Current Population Survey data, 1979-2016. Each coefficient is from a separate regression of the race-specific log of average weekly hours worked, or the black-white log difference, on the overall labor market tightness outcome measured in percentage points (e.g., an unemployment rate of 6.0 is 6.0% percent). Regressions include state and year fixed effects and standard errors are clustered at the state level.