The Wrong Crowd Effect: Is Police Targeting Contagious?

• In Event: Gangs, Co-Offending, and Social Networks Across Space and Time

Abstract

In a given population, the majority of individuals have very few friends, while a handful have many friends. Many social networks tend to exhibit a highly skewed distribution of the frequency of ties. However, such distributions are not unique to social networks; they also characterize the structure of the Internet and the web of actors who co-starred in the same movies. The present study examines the structure of a large co-arrest network of 90,759 individuals arrested over 6 years in Long Beach, CA and finds that it exhibits a similar property. I examine how principles of that have been hypothesized to lead to such a structure in other networks—preferential attachment and homophily—operate in this specific context. I hypothesize that being arrested with highly connected individuals subsequently increases one's odds of being targeted by police in the future. I examine the role of age-graded subgroups and brokers between them in the "contagion" of police attention over time, which eventually leads to the emergence of large component connecting a non-trivial proportion of the total population of arrestees. I discuss the implication of this theory for targeted policing, gang injunctions and other types of police practices.

Author

• Jason Gravel, University of California, Irvine