My own research is designed to shed light on strategic decision-making in social and political networks. Below are a few of my papers.

The effect of holding office
on the behavior of politicians
Does holding office change individual’s fundamental behavioral traits? We recruit politicians in Zambia to participate in experiments measuring reciprocity—a behavioral trait that underpins a wide range of political interactions, from legislative bargaining to lobbying and even bribery. We use a regression-discontinuity design to identify the effect of office-holding, by focusing on the behavior of politicians who barely won or lost elections. We find that attaining office causes an increase in reciprocity. This finding implies political institutions can be designed to shape the behavior of society’s leaders.

Knowledge and networks: how a wider 
view of the network facilitates coordination
One goal of network science is to identify patterns of connectivity that facilitate collective action among distributed agents. Many scholars believe that dense networks are ideal, but adding connectivity to a network can be costly, and experiments show that some connections inhibit coordination. We show that giving agents more information about the network itself facilitates collective action more reliably than does adding connections. 

Does more connectivity help groups solve social problems?
Behavioral and computer scientists often claim that a denser network of connections among individuals facilitates collective action. We use computational complexity theory to identify conditions in which adding connections can hinder coordination, and verify our predictions with an experiment. The results help us understand how to design networks that allow for efficient problem-solving in the real world. 

Can Legal Actors Play Equilibrium Strategies?
The economic analysis of law models legal actors as equilibrium players of strategic games. This approach depends on two assumptions: (1) actors have the information and computational resources to calculate equilibria, and (2) the modeler has a precise understanding of actors’ incentives. These are unrealistic assumptions, especially when they apply to the behavior of lay people such as jurors, litigants, or criminals.

Compromise Between the Echo Chambers:
Online Partisan Sorting as an Impediment to Political Consensus

The clustering of ideological preferences in social networks, both in person and online, may exacerbate partisan polarization and inhibit consensus-building. Unfortunately, the causal effect of ideological-preference clustering has been difficult to identify because social networks cannot easily be manipulated. I present an experiment in which subjects with conflicting preferences are paid to solve coordination problems over exogenously varied networks. When subjects are clustered by preference, they are less likely to achieve consensus and take longer to do so. These experimental results suggest the clustering of ideological preferences may make compromise and consensus more difficult in democratic politics.

Does Social Capital Habituate Cooperation?
Social capital theory claims associational institutions encourage mass cooperation, but scholars disagree on how this happens. Some argue that socially engaged individuals develop habits of cooperation; others argue social engagement creates incentives to cooperate through the shadow of the future. I show, through a laboratory experiment and survey, that subjects’ behavior is consistent with the incentive-based view of social capital but conflicts with the habit-based view.