Wage Transparency, Negotiation, and Reference-dependent Utility

Abstract

Wage transparency has become increasingly prevalent across industries as a result of government mandates, companies' voluntary disclosure, and digitization. A commonly cited benefit of wage transparency is that it can prompt underpaid workers to negotiate and thus reduce wage inequality. To empirically address how peer wage information influences wage negotiation and the subsequent implications for wage inequality, we conducted two field experiments on online labor markets where we truthfully varied information about historical peer wages and assessed workers' responses to job offers. Contrary to predictions of the prior literature, workers become more likely to negotiate not only when they learn that they are paid lower than others (by 34%-132%), but also when they learn that they are paid the same wage as others (by 38%-88%). We conjecture that the latter finding occurs because at least some workers believe they should be paid more than the average due to (warranted or inflated) positive assessments about themselves. This conjecture is supported by a laboratory experiment. Based on our empirical findings, we build a reference-dependent model to describe worker behavior and use it to derive the long-term implications of wage transparency for wage inequality. Our model and the subsequent numerical study reveal that wage transparency may amplify wage inequality by prompting workers who are already highly paid to ask for more. Our work highlights that wage transparency is not necessarily a panacea for wage inequality, and discusses additional measures that governments and organizations may consider along with wage transparency to combat wage inequality.

Keywords: Field Experiment, Reference Effects, Service Operations, Social Responsibility, Empirical Operations