Evan D. Sadler

Preferences, Homophily, and Social Learning

We study a sequential model of Bayesian social learning in networks in which agents have heterogeneous preferences, and neighbors tend to have similar preferences—a phenomenon known as homophily. We find that the density of network connections determines the impact of preference diversity and homophily on learning. When connections are sparse, diverse preferences are harmful to learning, and homophily may lead to substantial improvements. In contrast, in a dense network, preference diversity is beneficial. Intuitively, diverse ties introduce more independence between observations while providing less information individually. Homophilous connections individually carry more useful information, but multiple observations become redundant.

Learning in Social Networks

This survey covers models of how agents update behaviors and beliefs using information conveyed through social connections. We begin with sequential social learning models, in which each agent makes a decision once and for all after observing a subset of prior decisions; the discussion is organized around the concepts of diffusion and aggregation of information. Next, we present the DeGroot framework of average-based repeated updating, whose long- and medium-run dynamics can be completely characterized in terms of measures of network centrality and segregation. Finally, we turn to various models of repeated updating that feature richer optimizing behavior, and conclude by urging the development of network learning theories that can deal adequately with the observed phenomenon of persistent disagreement.

Customer Referral Incentives and Social Media

We study how to optimally attract new customers using a referral program. Whenever a consumer makes a purchase, the firm gives her a link to share with friends, and every purchase coming through that link generates a referral payment. The firm chooses the referral payment function and consumers play an equilibrium in response. The optimal payment function is nonlinear and not necessarily monotonic in the number of successful referrals. If we approximate the optimal policy using a linear payment function, the approximation loss scales with the square root of the average consumer degree. Using a threshold payment, the approximation loss scales proportionally to the average consumer degree. Combining the two, using a linear payment function with a threshold bonus, we can achieve a constant bound on the approximation loss.

Innovation Adoption and Collective Experimentation

I study a game in which individuals gather costly information about an innovation and share their knowledge through social ties. A persons incentive to experiment varies with her position in the network, and strategic interactions lead to counterintuitive behavior among the most connected players. The structure of the social network and the distribution of initial beliefs jointly determine long-run adoption behavior in the population. Networks that share information efficiently converge on a decision more quickly but are more prone to errors. Consequently, dense or centralized networks can have more volatile outcomes in the long run, and efforts to seed adoption in the network should focus on individuals who are isolated from one another. I argue that anti-seeding, preventing key individuals from experimenting early in the learning process, can be an effective intervention to encourage adoption because the population as a whole may gather more information.

Resource Allocation with Positive Externalities

We study a budget allocation problem between two players where budget allocations entail positive externalities. We characterize an optimal mechanism when the designer is unable to commit ex ante to the allocation rule. Without commitment, every incentive compatible mechanism the designer can implement is a hierarchical mechanism --- the allocation rule partitions the type space into intervals and allocates the budget to the player in the highest interval, dividing it evenly if both are in the same interval. The optimal mechanism uses a partition with infinitely many intervals. With full commitment power, this hierarchical mechanism remains optimal for a family of distributions that includes the uniform.