Examining parolees in their communities: Poverty, rurality, and criminal justice resources
Open Access
- Author:
- Burden, Frances Frick
- Graduate Program:
- Crime, Law and Justice
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- March 02, 2009
- Committee Members:
- Richard Barry Ruback, Dissertation Advisor/Co-Advisor
Barry Ruback, Committee Chair/Co-Chair
D Wayne Osgood, Committee Member
Barrett Alan Lee, Committee Member
Alexander Klippel, Committee Member - Keywords:
- recidivism
neighborhood effects
prisoner reentry
parolees
concentrated disadvantage
social disorganization - Abstract:
- Over the past thirty years, the number of people incarcerated in the United States has increased dramatically and since 93 percent of offenders are eventually released, recent years have witnessed a corresponding dramatic increase in the number of ex-prisoners returning to communities. Among these released offenders, two out of three parolees are returned to prison within three years of release. This study examines community-level risk factors for parolees in order to learn how different communities affect parolee recidivism. Specifically, the parolee population in Georgia was used to investigate how parolee recidivism is affected by three community characteristics – poverty, rurality, and the presence of criminal justice resources. Using Geographical Information Systems, a multilevel dataset was created that included individual-level parolee data, socio-structural measures and crime data of communities, and the location of parole offices in Georgia. Using Hierarchical Linear Modeling, parolees were followed over a two-year period in order to learn which community characteristics contributed to their return to prison. Each community characteristic was examined independently in order to avoid issues of multicollinearity between the socio-structural variables measuring poverty, rurality, and criminal justice resources. From these analyses, six important findings emerged. First, parolees who lived in urban communities with high levels of concentrated disadvantage, high population density, and high crime rates had significantly lower odds of being returned to prison. Because parolees increasingly reside in these impoverished urban communities, these areas are precisely the type of communities where parolees are safest from being returned to prison. This finding might be explained in part by the presence of informal social control and criminal justice resources, which could affect the level of parolee recidivism across communities. Although this study attempted to evaluate whether this finding could be explained by local criminal justice resources, these effects were not statistically significant. Additional analyses should be conducted using finer measures of criminal justice resources. The second and third findings showed that community effects were not uniform across parolee subopulations. Specifically, parolee race and risk scores each significantly interacted with community poverty. While all parolees had similar odds of recidivism in communities marked by extreme poverty, this recidivism gap widened in wealthier communities with white parolees recidivating significantly more than minority parolees. Further, a significant interaction was found between risk scores and racial inequality, with minority parolees more likely to be returned to prison in high racial inequality communities that favored African Americans. These findings suggest that some populations can benefit from residing in certain types of communities, while other populations reintegrate successfully regardless of where they live. This study also sought to better understand how five different theoretical and empirical poverty measures affected recidivism. The fourth finding established that only concentrated disadvantage significantly affected individual parolees’ odds of being returned to prison. The fifth finding showed that only extreme poverty significantly affected community rates of recidivism. Poverty measures of relative deprivation, racial inequality, and spatial proximity to poverty did not significantly impact parolees’ odds of being retuned to prison. These latter two findings suggest that poverty measures examining the magnitude of poverty are more important predictors of recidivism than relativistic measures of poverty (i.e., relative deprivation, racial inequality) or spatial measures of poverty (i.e., neighboring poverty levels). It is important to note that the geographic level at which this data was analyzed may affect the significance level of certain poverty measures, meaning that smaller community measures may result in different poverty outcomes. Finally, individual characteristics of parolees were found to be signficant across every model in this study. The magnitude of the effects of these parolee characteristics, both static (e.g., race, gender) and dynamic (e.g., drug and alcohol use, risk scores), suggest that while understanding community effects is important, one should still consider the impact of individual parolee characteristics on recidivism. Several limitations to this study are noted. First, this study provided a thorough examination of socio-structural variables, yet was unable to measure important mediating variables (e.g., neighborhood attachment). Second, the lack of randomness of parolees’ reintegration into residential communities is an inherent limitation in all community-level research, including the present study. Although there are a few examples of random assignment of communities, these examples are usually extremely expensive as well as politically unpopular. Finally, given that the poverty findings from this study are inconsistent with previous parolee research, the results from this study showing that the community effects affect parolees differently across geographical areas may not be generalizable to the entire parolee population of the United States. The community effects findings from this study indicate that communities can be important to the reintegration efforts of parolees. This study, in conjunction with other studies, found that various geographic areas can affect parolees differently. Additionally, community effects are not constant across different racial and ethnic groups. Given the differential effects communities have upon parolees across states and among different racial and ethnic groups, implications from this study suggest that a “one size fits all” approach to parole policy will be ineffective. Indeed, these results indicate that parole policy should be crafted locally with an understanding as to how communities operate in different jurisdictions in order to be successful.