Social Science Division Working Papers at NYUAD http://hdl.handle.net/2451/74526 2026-09-10T05:43:07Z 2026-09-10T05:43:07Z Baranski, Andrzej Chauhan, Prinsa Dariel, Aurélie http://hdl.handle.net/2451/76049 2026-09-03T06:56:50Z 2026-09-01T00:00:00Z

Title: Who is watching? Gender composition of observers and risk-taking Authors: Baranski, Andrzej; Chauhan, Prinsa; Dariel, Aurélie Abstract: We experimentally investigate whether the gender composition of an observing peer group affects individual risk-taking. Participants are randomly assigned to three-person groups differing only in gender composition. They complete an incentivized lottery task knowing that their choices will be observed by the other group members. We find that men’s willingness to choose the riskiest lottery increases monotonically with the number of female observers, whereas women exhibit no comparable response to the number of male observers. Men in all-male groups display levels of risk-taking statistically indistinguishable from those of women in all-female groups, indicating that the observed gender gap emerges primarily in mixed-gender groups. A mating-motivated signaling mechanism is consistent with our results.

2026-09-01T00:00:00Z Baranski, Andrzej Cooper, David J. Lee, Jeong Kyu http://hdl.handle.net/2451/76048 2026-09-03T06:50:03Z 2026-08-14T00:00:00Z

Title: CAT: An LLM-based tool for content analysis in experimental economics Authors: Baranski, Andrzej; Cooper, David J.; Lee, Jeong Kyu Abstract: We introduce the Communication Annotation Tool (CAT), an LLM-based tool for content analysis. CAT is intended to analyze free-form communication from economic experiments, accommodating a wide variety of communication structures. We describe CAT’s features including a broad array of customization options and a template for developing a coding manual. A detailed guide for using CAT is provided, including a step-by-step demonstration showing how CAT is used to quantify the content of communication from an experimental dataset. CAT is designed for easy monitoring of the coding process which ensures replicability of the communication coding method and simplifies identification and resolution of problems. We conclude by discussing CAT’s limitations and issues that users should keep in mind.

2026-08-14T00:00:00Z Van Essen, Matt Wooders, John http://hdl.handle.net/2451/76046 2026-08-24T06:27:40Z 2026-08-10T00:00:00Z

Title: Divide-and-choose games and the Shapley value Authors: Van Essen, Matt; Wooders, John Abstract: We establish an equivalence between two classic approaches to fairness: the Shapley value and divide-and-choose. We study the allocation of a single indivisible item among N players with heterogeneous values, where the player receiving the item compensates the others. On the cooperative side, we introduce sharing games, in which players are partitioned into syndicates and a coalition's value is determined by the highest-valued player accessible through its members' syndicates. On the procedural side, we introduce an N-player divide-and-choose game inspired by Kuhn's classic cake-cutting procedure. Our main result links the two approaches: for every ordering of proposers in the divide-and-choose game, there is a unique assortative syndicate structure for which the Shapley value of the corresponding sharing game coincides with the payoffs under maxmin-perfect play. Conversely, the Shapley value of every assortative sharing game can be implemented in a divide-and-choose game with an appropriate ordering of proposers. The equivalence provides a procedural foundation for the Shapley value and a cooperative interpretation of fairness in divide-and-choose.

2026-08-10T00:00:00Z Dariel, Aurélie Ham, John Nikiforakis, Nikos Stüber, Robert http://hdl.handle.net/2451/76045 2026-08-24T05:00:05Z 2026-08-01T00:00:00Z

Title: The family origins of the gender gap in competitiveness Authors: Dariel, Aurélie; Ham, John; Nikiforakis, Nikos; Stüber, Robert Abstract: The gender gap in competitiveness predicts disparities in life outcomes, yet its origins remain unclear. We investigate whether the gap originates within the family by exploiting quasi-experimental variation in sibling presence and sibling sex composition. We combine this variation with an incentivized measure of competitiveness in two culturally distinct populations to identify their causal effects on the gender gap. In a representative U.S. sample (N = 2,471), first-born women with younger brothers are far less likely to compete than those with younger sisters—who compete as often as men—driving the gender gap. In China (N = 421), the gap appears only among cohorts born before the One-Child Policy, who grew up with siblings. In both samples, the gap is concentrated in households with traditional gender norms. Our findings point to gendered upbringing within the family as one origin of the gender gap in competitiveness.

2026-08-01T00:00:00Z