Experimental Economics, Behavioral Economics, Information Economics, Labor Economics, Decision Theory
Information gathered throughout hiring processes has a great degree of subjectivity. Most job interviews include a cluster of questions that could be evaluated in very heterogeneous ways by different decision maker, even though its effects various aspects of the labor market are unknown. Using a simulated labor market experiment, we explore the effect of subjective information in hiring decisions. We find that subjective information changes the overall valuation managers assign to worker profiles, but it doesn’t improve the hiring results. Furthermore we find that subjective information strongly reduces the well-known bias in favor of male workers when it comes to hiring for a stereotypically male tasks. This implies that subjective information is a potentially useful tool in combating discrimination. Finally, we find that experienced managers obtain worse results when subjective information is part of the workers’ profile.
Under regret theory, decision-makers derive utility both from the outcome of their chosen action and the counterfactual. Evidence for anticipatory regret aversion has been found in one-shot settings, with "regret lotteries" that provide counterfactual information being valued higher than more-standard lotteries. These one-shot findings have motivated a literature that advocates the use of regret as a policy tool to boost incentives for behaviors such as exercise and drug adherence, often as recurrent decisions. However, differences in learning opportunities and the interaction between anticipated and realized regret make the consequences of regret in repeated settings far from clear. Through a series of controlled experiments, we replicate the one-shot result that regret lotteries have higher valuations than standard lotteries. In contrast, for sequential repeated decisions, the pattern reverses. For repeated decisions, regret lotteries are valued significantly less than standard lotteries, from the very first decision and onwards. Our results serve to highlight the issues that can arise when extrapolating behavioral effects from one-shot to repeated settings.