Giving and Costless Retaliation in the Power-to-Take Game (with Nobuyuki Hanaki, Natsumi Shimada, Yuta Shimodaira) [Link to working paper]
Abstract: Extending the power-to-take game, we explore the impact of two forces that may shape retaliation. In our 2x2 design, i) in addition to taking, the proposers can give part of their endowment to the responders, and ii) in addition to destroying their own endowment in retaliation, the responders can destroy the proposer’s endowment. Although these added options lead the responders to retaliate more severely, they do not significantly influence the proposers’ behavior. It is only when the proposers can give, and the responders can concurrently destroy the endowment of the proposers that the proposers take significantly less from the responders.
Corporate philanthropy, social incentives and leadership by example (with Edward Cartwright and Paola Paiardini)
Abstract: Corporate philanthropy has been frequently used as a tool to motivate behaviour in modern organisations. Using a large sample of subjects from an online experiment (N = 2,376), we study the impact of social incentives on leading-by-example in two separate studies. Study 1 considers treatments where leaders and followers are exogenously assigned to donate 0%, 20% or 40% of their income from a group project to charity. We find that the introduction of social incentives leads to a non-monotonic increase in contributions to the project for both leaders and followers. Study 2 examines behaviour in endogenous treatments where the assigned leaders choose the proportion of project income donated to charity. We show that social incentives increase contribution behaviour in a monotonic way. Consonant with previous literature, we find that part of our observed social incentive effects can be explained by shifts in beliefs.
Resisting bad norms: preferences for truth-telling in Non-WEIRD Societies (with Zeyu Qiu)
Abstract: This paper studies how social norms relate to lying behavior in four non-WEIRD countries: Egypt, Kenya, Mexico and South Africa. In an online die-roll experiment with 1,091 participants, subjects privately observe a random draw and have monetary incentives to misreport it. We elicit three types of norms: beliefs about what majority of others reported (descriptive), individuals’ own appropriateness judgments (personal) and beliefs about what majority of others viewed as socially appropriate (injunctive). Across all countries, we find partial but not full dishonesty. Strikingly, participants who believe that most others lie rather report lower numbers, choose high-payoff outcomes less often, and are less likely to select the payoff-maximizing outcome than those who believe others are honest. This pattern is particularly evident among participants who regard truthful reporting as personally or socially appropriate. We interpret this as evidence of norm resistance in a simple theoretical framework: individuals may trade off the cost of deviating from prevailing norms against reputational or self-image gains from resisting a bad norm. Our results provide important implications for policy making targeted at influencing dishonest conduct.
Redistribution Over Gains and Losses: Social Preferences and Moral Rules (with Ernesto M. Gavassa-Perez)
Abstract: We investigate redistributive behavior over gains and losses. Using two preregistered experiments, we document a systematic asymmetry in behavior: people are more selfish when redistributing over losses than over equivalent gains. We use structural estimation methods and out-of-sample predictions to understand the drivers of choices made by experimental subjects, and identify that a mix of social preferences coupled with loss aversion (inequality aversion, social efficiency, and maximin) alongside moral rules (blame avoidance and praise seeking) are key to understand the individual heterogeneity of redistributive behavior.
Income inequality and group identification in public good games (with Zeyu Qiu)
Abstract: A large body of research shows that advanced economies display low levels of social capital, such as trust and cooperation, while income inequality has risen. In this paper, we use online social dilemma game experiments with over 1,600 participants to test whether group identification helps explain the negative effect of income inequality on cooperation. We provide robust evidence supporting this hypothesis using two distinct measures of group identification. Our findings offer important implications for the design of social institutions and highlight the central role that group identification plays in explaining social capital differences driven by income inequality.
The effects of time blocking and goal setting on work performance (with Joaquín Gómez-Miñambres and Praveen Kujal).
Abstract: We experimentally examine how time blocking (dividing a fixed work period into shorter intervals) and non‑binding goals influence worker performance. Measuring four outcomes, task performance, effort, accuracy, and mistake rate, we find that time blocking primarily operates through a speed–accuracy tradeoff. While it substantially increases the number of tables attempted within a given period it simultaneously reduces accuracy, resulting in only a modest net gain in output. Goal setting, meanwhile, has no significant effect overall. The effects are heterogeneous and depend strongly on worker ability. For high‑ability workers, goal setting significantly improves performance. Combining time blocking with goal setting yields the largest performance gains, indicating that the two techniques complement each other for this group. For low‑ability workers, neither technique, individually or combined, produces significant improvements: although time blocking raises effort, the associated decline in accuracy offsets any potential benefit. Our findings highlight worker ability as a key moderator of non‑monetary motivational interventions and caution against applying multiple techniques simultaneously without accounting for workforce heterogeneity.
Gender Differences in Competitiveness: The Role of Social Incentives (with Mary L. Rigdon)
Abstract: The provision of social incentives in the workplace, where performance benefits a charitable cause, has been frequently used in modern organizations. In this paper, we quantify the impact of social incentives on performance under two incentive schemes: piece rate and a winner-take-all tournament. We introduce social incentives by informing individuals that 50% of their performance earnings will be donated to a charity of their own choice. Our findings indicate that, in the presence of social incentives, women increase their performance by approximately 23% and 27% in the piece rate and tournament payment schemes, respectively. These effects are sizable and significant. Despite the fact that women also become more confident when social incentives are used, their willingness to compete is not affected due to their general lack of willingness to take financial risks.