Work in progress

We propose and test a novel mechanism through which inequality sustains itself: when inequality rises, the greater rewards at the top motivate people to become more optimistic about their economic prospects. Based on a conceptual framework, we conduct an online experiment where participants experience different levels of initial inequality, form beliefs about reaching the top position, and make income allocation decisions. In line with theoretical predictions, the results show that: (1) high inequality exposure significantly increases confidence in reaching the top; (2) this overconfidence increases inequality acceptance, leading to more unequal allocations; and (3) these effects persist even with accurate relative performance feedback. Our analysis of cross-country survey data complements these experimental results: citizens in more unequal societies display greater optimism about upward mobility despite lower actual mobility, and this overconfidence correlates with reduced redistribution support. 

We study the impact of confidence on belief updating and recall of feedback. Previous research has focused on how (asymmetric) belief updating affects individuals’ confidence. We instead ask how exogenously induced high or low confidence impacts immediate and delayed feedback processing. In an experiment where we manipulate participants' confidence levels, underconfidence leads individuals to underupdate to positive feedback than overconfidence. Yet, for negative feedback we find no treatment difference for immediate belief updating. Further, we find individuals have generally correct memory of the feedback they receive, apart from some suggestive evidence of false memory of negative feedback among underconfident individuals.

Firms increasingly use algorithmic tools in hiring, but these systems can reproduce or amplify biases in the data they are trained on. In algorithm design, the exploration–exploitation tradeoff may matter for discrimination: exploratory algorithms deliberately sample less-observed options and may better reveal the productivity of underrepresented groups. Yet hiring often remains a human–AI collaboration, where managers can accept or override algorithmic recommendations.

We examine whether exploratory hiring algorithms can reduce discrimination and how managers respond to their recommendations in an online experiment. Participants act as managers making repeated hiring decisions between male and female candidates from fixed applicant pools for a task that is typically male-stereotyped. After each round, they receive performance feedback and report beliefs about the average performance of the hired group. Participants are randomly assigned to one of three conditions: no algorithmic recommendation, an exploitative recommendation that favors the stereotypically better-performing group, or an exploratory recommendation that more often suggests the less-sampled group. Managers can accept or override the recommendation in every round.

This design allows us to test whether exploratory algorithms increase hiring from the female candidate pool, whether managers follow exploratory recommendations, and how algorithmic advice shapes beliefs about group productivity. We also examine whether early exposure to recommendations affects later hiring decisions and whether gender stereotypes moderate responses to algorithmic advice.

Publications

This study explores the role of trait self-control in individuals’ changes in performance and well-being when working from home (WFH). In a three-wave longitudinal study with UK workers in the midst of the COVID-19 pandemic, we find that low self-control workers experienced a significant positive adjustment to WFH over time: The number of reported work distractions decreased, and self-assessed performance increased over the period of four months. In contrast, high self-control individuals did not show a similar upward trajectory. Despite the positive adjustment of low self-control individuals over time, on average, self-control was still positively associated with performance and negatively associated with work distractions. However, trait self-control was not consistently associated with changes in well-being. These findings provide a more nuanced view on trait self-control, suggesting that low self-control individuals can improve initial performance over time when working from home.