Job Matching and the Wage Distribution∗

We present a synthesis between the microeconomic-labor and the macroeconomicequilibrium views of matching in labor markets. We nest a job matching model à la Jovanovic (1984) into a Mortensen and Pissarides (1994)-type equilibrium search environment. The resulting framework preserves the implications of job matching theory for worker turnover and wage dynamics, while allowing for aggregation and general equilibrium analysis. We obtain two new equilibrium implications of job matching and search frictions for wage inequality. First, learning about match quality and turnover map Gaussian output noise into an ergodic wage distribution of empirically accurate shape: unimodal, skewed, with a Paretian right tail. Second, high idiosyncratic productivity risk hinders learning and sorting, and reduces wage inequality; the opposite is true in standard equilibrium search models with observable idiosyncratic productivity. We also obtain analytic solutions for the wage distribution and for aggregate worker flows − quits to unemployment and to other jobs, displacements and hires − which provide the likelihood function of the model, and allow for its parametric structural estimation.