114P
Exploring Correlates of Prison Violence in Chilean Facilities: Creating Pre-Conditions for Rehabilitation
Despite that prison violence is a common, widespread problem in Latin America (including Chile), up to date no systematic research has been conducted to study this phenomenon or its correlates nationwide beyond anecdotal evidence. As a result, prison administrators and social workers have scarce systematic knowledge to reduce prison violence and to promote a supportive environment for rehabilitation. Currently, the Chilean penitentiary system holds about 49,000 prisoners with an incarceration rate of 295 individuals behind bars per 100,000 inhabitants, one of the highest in the Americas, and an estimated recidivism rate of 50%. The Chilean Bureau of Prisons (Gendarmeria) is the public entity in charge of providing security, services and rehabilitation programs for the sentenced or held-for-trial individuals under its control. The overall objective of this research is to enhance our understanding about prison violence in Chilean prisons, to explore its main correlates and to offer comparison points with what happen in developed countries like the US.
Methods
This research analyzed secondary, administrative data from official records from Gendarmeria de Chile in 2012. Data corresponded to facility-level data from two sources: i) administrative data on prison violence [ADM] and ii) aggregate results from a national-level inmate survey conducted in 2013 [SURVEY]. In this study “prison violence” is understood as the total number of violent events at the aggregate, prison level during 2012. Despite the fact that prison violence implies more than what is officially recorded as such, prison officers in Chile are mandated by law to register this type of events so the use of this variable represents at least a baseline for the estimation of violence. Descriptive statistics on the dependent and independent variables are provided, along with correlation matrix and cross tabulations. Five negative binomial regression models were employed to analyze the correlates of violent events as the dependent variable was over dispersed, positively skewed, and many facilities reporting zero incidents during the observation period (2012).
Results
Results from the full model show that whereas total prison population and proportion of inmates in solitary confinement were positively associated to violence, the proportion of inmates participating in prison programs was negatively associated to violent events. No significant interaction effects between high deprivation conditions (proportion of inmates reporting mistreat by guards) and poor management practices (proportion of inmates in solitary confinement) were found. Contrary to expectations, occupancy rate and mistreat by guards were not significantly associated to prison violence in the full model, perhaps as a result of a their widespread (and unfortunate) presence within the Chilean prison system, characterized by overcrowding and human rights violations.
Implications
Despite two important limitations of this research; namely, its cross-sectional nature and the possibility of selection bias, these findings lend preliminary support for the administrative-control hypothesis of prison violence that assigns a critical responsibility to prison officers, administrators and social workers in the development and implementation of initiatives that can decrease the occurrence of violent events inside prison walls.