Lots of talk on the potential negative effects of minimum wages, but in a new paper with Michael Reich ( @IRLEUCB) we show that minimum wages also have positive side effects, especially in terms of reducing racial inequality. A short thread. 1/n https://irle.berkeley.edu/racial-inequality-and-minimum-wages-in-frictional-labor-markets/

We find that recent minimum wages (1990-2019) increase earnings of workers of all races, ages and genders, but that earnings of black workers increase more, reducing racial wage gaps. These results are in line with Derenencourt and Montialoux (QJE, 2021) evidence for 1966. 2/n

The graph below shows that minimum wage policy reduced hourly wage gaps by 60% among workers without college degree and 12% among all workers. (Was a lot of fun to create this graph in @Stata) 3/n

The difference in earnings response cannot be explained by lower initial wages. Using entropy balancing to equalise initial wages leaves the difference intact. The graph below shows estimates from a bunching estimator (cf Cengiz @arindube Lindner @benzipperer 2019) 4/n

Strikingly, the effect on white workers is concentrated on the bins around the minimum wage, whereas most excess jobs for black workers are in the +$4 bin. 5/n

We motivate the disprepancy with Card et al. (2018) style model, with workers that differ in their outside options. Why? Because black workers tend to live in central cities and have fewer mobility options (Miller, 2018) whereas the high quality jobs are in the suburbs. 6/n

Minimum wages increase the wealth and financing options for black workers (Cooper et al, 2020; Aaronson et al., 2012), allowing them to buy cars and access distant jobs. Car commuting increases for poor black workers after minimum wage increases, but not for white workers. 7/n

Remember how Dube et al. (JOLE, 2016) found that minimum wages increase job stability? That effect is twice as large for black workers, in line with our theory that the MW allows black workers to escape the bad options-bad jobs equilibrium for a good options-good jobs one. 8/n

These results are robust across four different methods and two datasets (CPS: state variation, within individual variation, event studies, bunching estimators; QWI: food services) and we do not find any evidence for pre-existing trends. 9/n

We find no disemployment effects, in fact low wage black workers are less (!!) likely to lose their job after minimum wage changes. 10/n