Inequality by Selection: The Case of Consistency

Leading positions in society typically are assigned through some type of selection contest in which applicants can strategically exploit information asymmetries. In this paper, we examine the individuals who are selected in such contests, as well as the public’s views on inequalities that could result from their selection. We focus on consistency as a selection criterion that could serve as a signal of ability in case of asymmetric information. We hypothesize that males exploit this signaling device more often than females, which could then lead to gender inequality in selection outcomes. To test this mechanism, we design a lab experiment where decision-makers with limited information select among contestants who perform a task twice. As expected, we find that using consistency as a signal increases chances of being selected. Most importantly, we observe a gender gap in selection outcomes, even though the decision-makers cannot tell the gender of the contestants and have no ability-based reason to favor men. Gender inequality appears to be due to differences in strategic behavior among individuals rather than differences in their traits or preferences. To investigate how the findings of our lab experiment are perceived and to obtain preliminary evidence on the consequences of inequality, we first ask the participants of the lab experiment about their fairness perceptions using an ex-post questionnaire. Among other things, this informs us about fairness perceptions of those who were actually engaged in signaling, thereby potentially benefiting from unequal selection outcomes, compared to those who did not benefit from employing strategic action in the selection contest. We conclude our investigation by studying fairness perceptions in a subsequent survey experiment. Using a large-scale population survey, we elicit fairness perceptions of different inequality scenarios that vary in the selection process and outcomes. The uninvolved third parties (observers) are asked to choose an appropriate distribution of the selection bonus of a future session, for which we announce to run a modified version of our lab experiment. Observers’ distribution decisions inform about fairness perceptions and, hence, possible policy implications, since changes to the unequal payoff structure can indicate less acceptance of inequality. Taken together, our findings highlight how gender inequality can emerge in competitive selection environments, even without explicit bias. By linking lab evidence on the selection mechanism to public fairness perceptions, we provide a comprehensive view of the sources and consequences of inequality, showing how unequal outcomes arise, are evaluated, and potentially contested.