Minimum wage and employment in the U.S.: an application of Bayesian quantile kink regression

We examine whether the employment effects of minimum wage depend on unknown tipping points in the labor market. We apply a continuous threshold regression model—regression kink with unknown thresholds—to U.S. state-level panel data in 1993–2016 to estimate the tipping point and quantile employment effects. Overall, we find that the marginal effect is near-zero or mildly negative below the tipping point, and it is considerably more negative above it. The tipping occurs at 50–55% of the state’s median wage among women and 40–45% among men. Simulations of minimum wage reforms reveal nonlinear and asymmetric employment effects.