Structural Risk and Cognitive Protection: Social Capital and Criminal Victimization in Latin America and the Caribbean

Abstract

Latin America and the Caribbean has one of the highest rates of violence and crime in the world, and the region faces a crisis of insecurity in which quality data and evidence-based policy are missing. At the same time, studies have shown that increased social capital is significantly associated with reduced levels of violence and crime, and both government agents and development agencies in the region have begun to place a major emphasis on programs that promote social capital. Nevertheless, no detailed analysis of the complex ways in which the dimensions of social capital affect violence and crime in the region has been conducted. Those studies that have analyzed elements of this relationship have all been subject to two-way causality bias.

To fill this significant gap in the literature, this thesis examines the effect of structural social capital (social organization characteristics) and cognitive social capital (social trust characteristics) on the risk of exposure to crime in the region. The included empirical analysis uses cross-sectional survey data from AmeriasBarometer. The sample includes more than 135,000 survey responses from 24 countries between 2006 and 2014, and measures exposure to crime as respondents’ self-reporting on whether they had been victims of crime in the last 12 months. The quantitative analysis uses generalized linear models (GLMs) to investigate the effects of increased structural and cognitive social capital on the likelihood of criminal victimization. Due to the cross-sectional nature of the data and the eroding effect of violence on social capital, the analysis also instrumental variable estimators to account for potential joint endogeneity.

The results of the empirical analysis offer strong evidence in support of previous studies suggesting that cognitive and structural social capital are inversely related to the risk of being a victim of crime. The results show that respondents with high levels of cognitive social capital have a significantly lower risk of being exposed to crime than those with low levels, while respondents with high levels of structural social capital have a significantly higher risk of being exposed to crime than those with low levels (p < 0.01). As such, this thesis provides strong empirical support for the view that cognitive social capital acts as a protective factor for violence while structural social capital acts as a risk factor for violence in Latin America and the Caribbean, while indicating that social capital should not be viewed as a unified concept. This finding has significant implications for violence prevention and intervention policies in the region, as well as for the theory of collective efficacy and the on-going debate over how best to measure social capital.