Using career history data from employees' online profiles, we provide empirical evidence regarding how human resources systems can affect the gender gap in job tenure. Our results indicate that female workers experience shorter average job tenure than male workers. Our results further show that abandoning forced ranking systems helps to narrow this gender gap. We present experimental evidence indicating that in forced ranking environments, women experience shorter tenure than men and are more likely to exit their roles. This is true even when women are on equal footing with men. Finally, we present survey evidence that under forced ranking systems, more women resign as compared with men. In sum, our empirical, experimental, and survey results indicate that forced ranking systems may contribute to a gender gap in job tenure. Accordingly, abandoning such systems may help promote workplace gender equality.

We study dynamic mechanism design in non-clairvoyant environments, where sellers cannot forecast future demand and buyers hold private information. Using two laboratory experiments, we evaluate two non-clairvoyant mechanisms: the dynamic maximin-optimal non-clairvoyant mechanism (DM, Mirrokni et al., 2020) and the repeated static optimal mechanism (RS, Myerson, 1981). Experiment 1 compares DM and RS between demand-distribution scenarios and shows that non-clairvoyant mechanisms perform as intended: DM either outperforms or underperforms RS depending on the scenario, consistent with theory. Building on these results, Experiment 2 examines the optimal clairvoyant mechanism (OC) and a different implementation of DM. In surprising contrast to theory, OC using full information performs no better than the non-clairvoyant mechanisms, while DM performs equally well across implementation forms. Our results highlight the practical importance of non-clairvoyant mechanisms as implementable approaches to dynamic auction design in uncertain markets.

We study the impact of honesty on electoral outcomes and effective leadership. We predict that dishonest candidates are more likely to win elections, and yet be less likely to lead effectively. Consequently, groups who elect honest leaders are more likely to achieve socially efficient outcomes. We use laboratory experiments to test these predictions. We find that cheating rates are high among leaders when cheating is possible. Further, we obtain causal evidence that, in relation to dishonest leaders, honest followers are more likely to obey honest leaders, resulting in improved cooperation, higher social efficiency and greater equity among group members.

Alcohol consumption is an important component of business negotiations across many cultures, yet this behavior remains unmodeled and its potential explanations untested. Here, we develop a theory that combines guilt-aversion with the canonical alcohol myopia framework. Our GAAM (guilt aversion and alcohol myopia) model predicts that intoxication increases promise-making, and will not decrease rate of promise-breaking. We test these predictions using a Prisoner's Dilemma game with pre-play communication in a lab-in-the-field experiment. Among males, we find behavior consistent with predictions: intoxication promotes promise-making but does not impact the rate at which promises are broken. Importantly, this implies intoxication increases the efficiency of communication. We do not observe intoxication to impact female promise-making or promise-breaking behaviors. This is consistent with previous empirical findings that females can display less sensitivity than males to alcohol-induced myopia. Our results provide an explanation for the widespread phenomenon of business drinking.

We extend the war of attrition by studying a three-period dynamic contest game. In our game, players can fight against their opponents at certain period of the contest and can flee at any time. Waiting is costly. We focus on the role of waiting costs and show that the value of waiting costs is a key factor in determining the type of equilibrium in such dynamic contests. Specifically, as waiting costs increase, contests end earlier, battles are less likely to occur, and the weaker player in a pair is more likely to flee. A lab experiment verifies most key features of our model. However, unlike theoretical predictions, we find that as waiting costs increase, the duration of contests and the frequency of battles fail to drop as significantly as theory predicted. Moreover, we find that in each treatment, individual players exit the contest significantly earlier than predicted.

We develop a method for random assignment of language to participants in a controlled laboratory experiment, and use this to test the learning efficiency hypothesis (LEH henceforth), that languages are learned more quickly when they can be determined with fewer number of observations according to theory. This hypothesis plays a crucial role in the development of the theory of learning efficiency of compositional languages. Despite this, evidence on the empirical validity of the LEH has been elusive. Here we develop a novel extension of a coordination game within which languages emerge endogenously. We show that one can control features of an emergent language by varying the game's incentives. This enables us to compare speed of learning across participants randomly assigned to different languages. Our data provide cogent evidence supporting the LEH: languages with structures that theory implies can be determined with fewer observations are learned more quickly in the lab, and are in this sense more efficient.

This study explores how emotion expression (EE) can help promote positive indirect reciprocity while also breaking chains of negative indirect reciprocity. We report data from a two-stage dictator game and compare the decisions subjects make towards a third party both with and without the chance to express emotion. Standard economic models suggest opportunities to express emotion should have no impact on behavior. In sharp contrast to these predictions, we find that subjects who are treated unkindly behave more generously towards others in subsequent interactions if they have the opportunity to convey their emotions to a third party. Our findings have important welfare implications. In particular, they suggest that opportunities to express emotion can break negative reciprocity chains and promote generosity and social well-being even among those who have not themselves been previously well-treated. We argue that reference-dependent valuations can help to explain our results.

Effective leaders promote cooperation among their followers. Although many studies have focused on approaches to effective leadership, few have studied how social identity impacts a leader’s effectiveness within the context of a controlled laboratory environment. Using a novel laboratory experiment and a guiding theoretical framework enables us to shed new light on how social identity impacts leader-follower interactions in a decontextualized environment. In our experiment, the leader, if present, may or may not share a social identity with the rest of the group. Our main finding is that ingroup leaders are more effective than outgroup leaders, and outgroup leaders are no more effective than groups without a leader. In particular, ingroup leaders are more likely to suggest greater cooperation among followers, and their suggestions are more likely to be followed. Further, we find that outgroup leaders are less credible leaders. These findings suggest that diversity alone does not drive inclusion and provide evidence of an important challenge to achieving the well-documented organizational benefits that stem from a diverse and inclusive workplace.

We experimentally test convergence to the core in two-sided markets for heterogeneous indivisible goods under different trading institutions. We use bargaining and strategic games as predictors that naturally generalize the core, accommodating non-equilibrium behavior. The performance of the competing theories reflects the differences in trading procedures—market outcomes are close to Nash equilibrium predictions under auction-like institutions and close to bargaining for institutions that feature decentralized negotiations. This difference may be driving the documented effect of fewer no-trade outcomes at the expense of a higher chance of suboptimal match under free-form bargaining.

The effect of endowment equality on trust may stem from outcome inequality aversion or changes in expected trustworthiness. Here, we measure trust as the expectation of honesty in a sender-receiver game, where participants must make trust decisions without knowing the outcome. Our design enables us to isolate the effect of initial endowment inequality on trust. Our results show that endowment inequality reduces trust regardless of whether it favors the sender or the receiver. We further find that the frequency of lies is insensitive to endowment inequality. Our results amplify the importance of equal starting positions in promoting trust.

Dynamic mechanisms provide a powerful means for optimizing repeated auctions. Imple- menting them is complicated, however, due to a number of conditions that are difficult to satisfy in practice. These include the fact that the auction designer must be clairvoyant, in the sense that they must have reliable forecasts of participants’ valuation distributions in all future periods. Recently, Mirrokni et al. (2020) introduced a non-clairvoyant dynamic mech- anism (NC) and showed it is optimal within the class of dynamic mechanisms that do not rely on strong assumptions about the future. We showed, however, that an optimal repeated static mechanism (RS) (a Myerson auction) can sometimes outperform NC. Here, we report data from an experiment to test NC in relation to RS. Our results support the theory: NC outperforms or underperforms RS according to theory predictions. Our results highlight the practical value of non-clairvoyant mechanisms as implementable approaches to dynamic auction design.

Under optimal tournament design, we would expect agents to exert identical effort regardless of the shape of the contest function’s error component. We report data from laboratory experiments that provide a first test of this prediction. We find that efforts do not significantly differ when the shock distribution exhibits negative skewness versus a uniform distribution; however, subjects react substantially differently to random shock realizations under different treatments. Specifically, tournament winners demonstrate stronger reactions, economically and statistically, to negatively-skewed shocks than to uniform shocks. Meanwhile, tournament losers are less likely to be affected by negatively-skewed shocks. Our results highlight the importance of accounting for the influence of the shape of the shock distribution on a contest participant’s effort.

We study how parental absence impacts children's norm obedience. Due to the prevalence of migrant labor in China, 61 million Chinese children have been "left behind," with one or both parents absent from the home. Using a nationally representative sample of secondary school students, we find that rural "left-behind" girls exhibit norm disobedience at a higher rate than other girls or boys. They are more likely to cheat in school, use offensive language, and display anti-social behavior. We further find that being cared for by grandparents exacerbates the "leftbehind" effect in rural girls, but not in other groups. This indicates that the son preference in rural China's older generation may leave rural "left behind" girls in disadvantaged situations. This finding is relevant for policy choices in developing countries, particularly to advancing and empowering women.