最低工资对青年就业的影响：基于美国州级差异的证据

The minimum wage remains one of the most contentious policy instruments in labor economics. Since the seminal work of Card and Krueger (1994), a vast empirical literature has sought to estimate the causal effect of minimum wage hikes on employment, with results ranging from large negative effects to zero or even positive estimates. This lack of consensus reflects genuine methodological challenges: minimum wage changes are endogenous to local economic conditions, the treatment intensity varies across workers and firms, and general equilibrium effects may offset partial equilibrium disemployment. In this paper, we leverage three decades of state-level variation in U.S. minimum wages to provide new estimates that address each of these concerns.（节选）

This paper makes three main contributions. First, we construct a novel panel dataset that tracks minimum wage changes at the state-quarter level linked to individual-level employment outcomes from the CPS, allowing us to estimate both extensive and intensive margin responses. Second, we show that the canonical two-way fixed effects estimator substantially overstates disemployment effects when treatment effects are heterogeneous across cohorts; our event-study and stacked DID specifications yield smaller and more precisely estimated effects. Third, we provide the first systematic evidence that sectoral reallocation—workers moving from affected to unaffected industries—accounts for approximately 40% of the gross job loss in exposed sectors, a margin absent from most prior work.

Taken together, our results paint a nuanced picture: minimum wage increases do reduce employment among teenagers and low-skill workers, but the effects are modest (around 1.2% per 10% wage increase) and concentrated in specific sectors. Importantly, we find evidence of substantial worker reallocation across industries, suggesting that aggregate employment effects overstate the welfare consequences for affected workers. Policymakers should weigh these modest disemployment effects against the documented earnings gains for workers who retain employment.