| Summary: | Under what circumstances do citizens turn to their neighbors to confront criminals directly? I propose that social capital and distrust in authorities interact to increase the probability that citizens will attempt to enforce the law independently of the state (what I call Extralegal Collective Law Enforcement [ECLE]). That is, I argue that the link between social capital and ECLE is greater when citizens distrust law enforcement. To test this hypothesis I take a multi-method approach. In Chapter II I turn to one of the most recent and iconic cases of ECLE in Mexico, the case of Cherán. I find evidence that prior to the rise of the movement, citizens of this community had relatively strong levels of social capital, and they mobilized it in order to confront crime. In Chapter III I analyze survey data from Mexico and sixteen other comparable countries in the Americas and find that distrust in the police moderates the effect of social trust on citizens likelihood of engaging in collective anti-criminal organization. Further, I find that collective anti-criminal organization translates into citizens willingness to engage in ECLE more strongly among those who distrust the police. In Chapters IV and V I test my hypothesis experimentally. In Chapter IV I find citizens to be more supportive of others engaging in ECLE when they read that they are surrounded by trustworthy neighbors. Further, in line with my argument, I find this effect to disappear when the police are portrayed as trustworthy. Chapter V shows the results of a behavioral experiment in which subjects risk real money in a game that incorporates some of the incentives and actors involved in collective vigilantism. I find subjects to be more willing to confront a mock-thief when their neighbors have a higher likelihood of helping them (i.e., are more trustworthy) and the police are less likely to intervene (i.e., are more untrustworthy). Finally, consistent with my previous results, I find the untrustworthiness of the police to intensify the effect of the neighbors on subjects likelihood of displaying vigilante-like behavior.
|
|---|
