Bidding for Talent: A Test of Conduct on a High-Wage Labor Market

About this event

This paper proposes a novel procedure for adjudicating between models of firm wage-setting conduct. Using data from a U.S. job search platform, we first flexibly estimate labor supply to differentiated firms. We then use those estimates to formulate a test of conduct based on exclusion restrictions. On average, workers are willing to pay 12.5% of their salary to enjoy a 1-S.D. improvement in non-wage amenities. Between-worker dispersion in preferences is of a similar magnitude, indicating that preferences are not well-described by a single ranking of firms. Oligopsonistic models incorporating strategic interactions between firms and tailoring of wage offers to workers' outside options are rejected in favor of simpler monopsonistic models featuring near-uniform markdowns. Misspecification has meaningful consequences: while our preferred model predicts average markdowns of 18%, others predict average markdowns of 26% (about 50% larger). Implied patterns of between- and within-firm productivity dispersion also differ markedly across models of conduct.