Abstract
Considering that approximately two-thirds of individuals who are released from prison are re-arrested at least once within three years following release (Langan and Levin 2002), any research that seeks to identify factors associated with successful reentry is certainly warranted. This dissertation investigates the role of family support for individuals who have been deemed serious and violent offenders and recently released from state prisons. Little research has sought to quantitatively measure the extent of the relationship between levels of family support and recidivism after controlling for other known predictors of reoffending. Prior research has largely relied on fairly small sample sizes, short follow-up periods post-release, basic bivariate analyses and inconsistent conceptualizations of family support (La Vigne, Visher and Castro 2004; Nelson, Deess and Allen 1999; Sullivan, Mino, Nelson and Pope 2002; Visher, La Vigne and Travis 2004b).
As part of the evaluation of the Serious and Violent Offender Reentry Initiative (SVORI), 1,697 adult males and 357 adult females were interviewed 30 days prior to their release and then three, nine and 15 months following release. Using the data collected from these interviews, this dissertation explores the relationship between emotional family support and instrumental family support and four measures of reoffending: any self-reported criminal offending, any self-reported violent offending, any self-reported drug offending and whether any arrest occurred (using official records from the National Crime Information Center) during each of the post-release follow-up periods.
Controlling for other known predictors of reoffending, logistic regression models are used to predict the likelihood of reoffending. Considering respondent attrition over successive interview waves, all analyses are conducted using listwise deletion as well as multiple imputation to handle missing data. Results generally reveal that emotional support is associated with a significant decrease in reoffending, while instrumental support is not significantly associated with reoffending.
These findings have implications for correctional policies and programming, sentencing policies, post-release supervision policies and programming, criminological theory and future research.