Research

Social Interaction and Cultural Activities within Friends Networks

(Job Market Paper)

This paper investigates club participation among students, exploring how individual choices interact with prevailing social norms. The participation of white students in basketball and black students in baseball is found to be significantly influenced by the size of their respective racial groups, a pattern not observed in reverse race-sport combinations. Analysis with a Linear-in-Mean (LIM) model reveals that previous studies may have overestimated the peer effects on individual decisions. Employing instrumental variables (IV) does not fully resolve the reflection problem, primarily due to the endogeneity of the network structure and the presence of homophily in friendship formation. Furthermore, I introduce a benefit-driven model, where each student's participation decision involves a trade-off between the costs of participation and the benefits gained from social engagement. The empirical analysis reveals that the differing behaviors between white and black students are driven by discrepancies in their benefits obtain through social engagement. This study highlights the role of club participation in shaping the structure of social networks, and the results suggest that extracurricular clubs can foster social cohesion across different racial groups in schools. Link

Presented at: NASM 2024, CES (Rising Stars Session) 2024, AMES 2024

Interaction and Segregation:

A Networks Formation Model

This study examines how multi-dimensional characteristics, particularly race and cultural activities, shape friendship network formation. Recognizing that homophily extends across various dimensions, we explore how the layer interdependency influences the network's structure. Our model predicts several key patterns observed in real-world data: students who join clubs have more connections than those who don't; the number of connections for club members increases with the relative size of their racial group and the club's popularity; and racial segregation is minimized when individuals from different racial groups engage in cultural activities at similar rates. Our analysis demonstrates how these patterns can emerge from a biased matching process in network formation and provides evidence that participation in cultural activities reduces social segregation among racial groups, emphasizing the role of such activities in promoting inclusivity and cohesion. Link

Profit-Driven Recommendations: Navigating Consumer Preferences and Value

The development of big data offers convenience to customers, allowing them to find their preferred products more quickly and easily. Intermediary platforms, after learning a consumer's preferences, may be motivated to recommend products that are more profitable to the platform rather than the ones that are most suitable for the customer. In our model, after learning each consumer's taste, the monopolist provides an optimal recommendation strategy. By decreasing the gradient on the price-quality schedule, the monopolist offers lower prices for a given quality than those set by impersonal monopolists, while serving more customers. The monopolist could receive more profit by successfully recommending a product with a quality higher than the consumer’s optimal choice. Consumers, on the other hand, may benefit from recommendations if they prioritize efficiency and time.