My Research

I study social networks in the economic, political, and criminal arenas. Most broadly, my research focuses on how interdependent actors collectively shape social structure. I work with data from historical sources, surveys, and the web using a combination of network analysis, statistical modeling, and computational methods. While addressing diverse empirical and theoretical puzzles, my research features a core focus on the mechanisms giving rise to cooperation and conflict, economic organization, and network structure.

Dissertation Project

Closure and Collaboration in the American Mafia

How do organizations obtain access to valued resources without diluting the loyalties and identities of their members? Network analysts suggest focusing on the boundary-spanning activities of “brokers” who bridge gaps in social structure. In many contexts, however, brokers are viewed with suspicion and distrust rather than rewarded for their diversity of interests. My dissertation examines organizations in which the theoretical deck is seemingly stacked against brokerage and toward parochialism: American-Italian mafia families. Using a historical network data set, I document a division of network labor in which a small number of brokers—often, surprisingly, ethnic outsiders excluded from formal membership—bridged otherwise disconnected islands of criminal activity to gain power within exclusive mafia circles. While social closure in solidary groups ensures a heavy premium on insider status, it can also paradoxically increase the returns to outsider brokerage, albeit only when taken up in a way that does not violate group norms.  


ABSTRACT - Criminal networks are thought to be biased toward decentralization and security rather than integration and efficiency. This article examines this tradeoff in a large-scale national criminal network spanning more than 700 members of 24 distinct American mafia families operating in the mid-20th century. Producing a novel network image of the American mafia as a set of highly differentiated yet intertwined islands of criminal activity, the analysis uncovers a small-world structure that allowed both for strong intragroup closure and high intergroup connectivity. This balance reflected a division of network labor in which integrative bridging connections were disproportionately concentrated among a small number of criminals. Furthermore, the criminals who held such bridging ties tended to be either low- or high-status - but not of middling status - within their respective organizations.

Endogenous Dynamics of Institutional Change

Co-authored with Victor Nee and Sonja Opper

Rationality and Society (Forthcoming)

ABSTRACT - A parsimonious set of mechanisms explains how and under which conditions behavioral deviations build into cascades that reshape institutional frameworks from the bottom up, even if institutional innovations initially conflict with the legally codified rules of the game. Specifically, we argue that this type of endogenous institutional change emerges from an interplay between three factors: the utility gain agents associate with decoupling from institutional equilibria, positive externalities derived from similar decoupling among one's neighbors, and accommodation by state actors. Where endogenous institutional change driven by societal action is sufficiently robust, it can induce political actors to accommodate and eventually to legitimize institutional innovations from below. We provide empirical illustrations of our theory in two disparate institutional contexts - the rise of private manufacturing in the Yangzi delta region of China since 1978, focusing on two municipalities in that region, and the diffusion of gay bars in San Francisco in the 1960s and 1970s. We validate our theory with an agent-based simulation.

Why Do Liberals Drink Lattes?

Co-authored with Yongren Shi and Michael Macy

American Journal of Sociology (2015) 120: 1473-1511

ABSTRACT - Popular accounts of "lifestyle politics" and "culture wars" suggest that political and ideological divisions extend also to leisure activities, consumption, aesthetic taste, and personal morality. Drawing on a total of 22,572 pairwise correlations from the General Social Survey (1972-2010), the authors provide comprehensive empirical support for the anecdotal accounts. Moreover, most ideological differences in lifestyle cannot be explained by demographic covariates alone. The authors propose a surprisingly simple solution to the puzzle of lifestyle politics. Computational experiments show how the self-reinforcing dynamics of homophily and influence dramatically amplify even very small elective affinities between lifestyle and ideology, producing a stereotypical world of "latte liberals" and "bird-hunting conservatives" much like the one in which we live. 

ABSTRACT - This article documents heterogeneous economic returns to military service that vary with the individual propensity to serve, even within a relatively privileged sample of mostly white high school graduates. Using a rich set of covariates from the Wisconsin Longitudinal Study, I estimate propensity scores for male respondents' likelihood of voluntary military enlistment or involuntary draft conscription. Then, I use recently developed HLM-based methods for causal inference to analyze systematic variation in veteran status' effect on later earnings as a function of these propensity scores. Among individuals with low propensities for military service, but not among others, veterans suffer large wage penalties. While this pattern applies to both voluntary enlisters and draftees, the timing of the wage penalty differs by mode of military entry. These effects are shown to correlate strongly with differences in educational attainment between veterans and nonveterans with low propensities for military service, suggesting the greater value of opportunities for human capital accumulation in the civilian sphere. 

ABSTRACT - Research on contemporary European politics has shown that immigrant population size is strongly associated with vote totals for anti-immigrant political parties. Competitive threat theories suggest that this association should be positive, whereas intergroup contact theories imply that it should be negative. A two-level analysis of vote totals for the French Front National (FRN) suggests that the direction of this association depends critically on the level of analysis. At the department (i.e., state or regional) level, large immigrant populations are associated with higher FRN vote totals. At the commune (i.e., town or city) level, however, large immigrant populations are instead associated with lower FRN vote totals. These findings challenge the conclusions of previous analyses of populist-right voting and provide further evidence that contact and threat dynamics often operate simultaneously, albeit at different levels.