The Implications of Pay Transparency in the Presence of Over- and Underconfident Agents

Many companies are under pressure to improve pay transparency; however, its impact on their agents and principals remains unclear. As a way to investigate the upside and downside of pay transparency, we conduct a study that focuses on a scenario where agents have “cognitive bias” (namely, over- or underconfident). We also capture the notion of “social comparisons” behavior (namely, behind-averse and ahead-seeking) that occurs under pay transparency. Through exploring a one-principal-two-agent model, we find that agents’ optimal effort decisions and job satisfaction are affected by the intersection among agents’ social comparison behavior, cognitive bias and the environment’s relative favoritism. Specifically, relative to the opaque policy, we find that pay transparency can (i) entice agents to improve their job performance (in general); (ii) decrease job satisfaction when the biased agent is overconfident, (iii) improve job satisfaction and job performance simultaneously under certain conditions (including when the biased agent is underconfident). Our analysis also provides a guideline for principals to implement transparent payment policies properly. Specifically, pay transparency could enable the principal to offer a lower compensation level than an opaque policy to motivate agents. Incorporating the heterogeneity of agents’ working capability into our base model, we find that the incentives derived from social comparison on biased agent’s effort decision will become stronger as the biased agent’s working capability increases. Through a two-period dynamic model with learning, we observe that the biased agent will exert the same effort as in the base model (without learning), and the sophisticated agent exerts a relatively lower effort level relative to the base model only when the biased estimation level is moderate