Monopsony, Markdowns, and Minimum Wages

This paper presents the first direct test of two interlinked predictions at the core of the monopsony theory of the labor market: (i) that firms exploit wage-setting power by marking down wages below the marginal revenue product of labor, and (ii) that exogenous wage constraints, if binding, eliminate markdowns. Our research design revisits the 2015 introduction of a high minimum wage in Germany. Drawing on a monopsony model, we derive an empirically tractable difference-in-differences specification that provides a quantitative benchmark for the firm-level markdown response. Our main result is that empirical markdowns respond only 0–25% as much as the monopsony model would have predicted. Hence, at least for the labor market segment we study, (i) markdowns largely reflect other distortions than monopsony, (ii) markdowns are mismeasured, (iii) minimum wages induce widespread labor shortages, or (iv) the standard monopsony model does not provide a full, realistic account of the labor market.