The decision to grant conditional release from prison (aka the parole decision) has been largely neglected in the contemporary criminological literature, despite its critical implications. The current study, conducted in Pennsylvania, United States, tests for punitive themes in parole decision making by examining the impact of several measures reflective of punishment satisfaction on the decision to grant release to eligible parole candidates. The results indicate that the amount of time served in relation to the original punishment does not predict parole decisions but the nature of the original offense does. Moreover, inmates eligible for parole have to experience at least one parole denial to increase their chances of release, suggesting that parole decision makers use the parole process as a punitive means. The implications of the findings are discussed.

Parole decision—the decision to release an incarcerated individual from prison conditionally—is one of the most critical decisions across justice systems around the world. The decision carries with it significant consequences: for the freedom of the individual awaiting release (the parolee); for the safety of the community in which they will return; and for the correctional system overall, especially its organizational capacity. The current study attempts to add to the parole decision-making literature by specifically analyzing the role that mental health factors may play in explaining parole decisions. Research to date is inconclusive on whether or not mental illness is a risk factor for criminal behavior; despite this, individuals with mental health problems generally fare worse on risk assessment tools employed in justice decisions. The study relies on a 1000+ representative sample of parole-eligible individuals in Pennsylvania, United States. To increase reliability, the analyses test for several mental health factors based on information from different sources (i.e., self-reported mental health history; risk assessment tool employed by the Parole Board; and risk assessment tool employed by the Department of Corrections). To address validity concerns, the study controls for other potential correlates of parole decisions. Although the multivariate models explained a considerable amount of variance in parole decisions, the inclusion of mental health variables added relatively little to model fit. The results provide insights into an understudied area of justice decision making, suggesting that despite the stigmatization of mental illness among criminal justice populations, parole board members in Pennsylvania, United States, appear to follow official guidelines rather than to consider more subjective notions that poor mental health should negate parole release.

Research on targeted enforcement in high-crime places has focused on direct crime-reduction impacts, possible displacement of crime, and more recently, diffusion of benefits to adjacent areas. Studies have ignored other unanticipated negative effects that a place-oriented enforcement intervention may have on the justice system overall. Using the generation of fugitive defendants as one possible example of an important system side effect, this study proposes hypotheses relating to adverse, generalized, system side effects of a place-and crime-focused intervention, and it tests for target area and targeted crime-type effects, nontarget area and nontargeted crime-type effects, and overall system effects.The analysis employs a multiple interrupted time-series design [autoregressive integrated moving average (ARIMA)] to test the impact of one widely publicized, geographically targeted drug-enforcement strategy in Philadelphia (Operation Sunrise, formally launched in June 1998) on the incidence of bench warrants as a measure of fugitives (weekly aggregate bench warrants series for the period January 1994-May 2005 N = 590 observations). The findings appear to support all hypotheses as they relate to the example of the generation of fugitives, and suggest a generalized system adverse side effect from the circumscribed place-and crime-focused intervention. The implications of

No abstract

The concurrent impact of individual and neighborhood effects on defendant pretrial performance has not been studied. This study asks whether there is neighborhood-level variation in defendants’ failure to appear and pretrial crime and explores the impact of three neighborhood structural conditions (socioeconomic status, stability, and racial composition). The study was conducted in Philadelphia, Pennsylvania, on a 2005 sample of defendants (N = 800), followed for a year to record bail outcomes. Defendants’ residences were geocoded within 45 neighborhoods. Census data were used for neighborhood structural characteristics. Multilevel analyses (hierarchical linear modeling) found no neighborhood variation. However, individual-level results indicate that a defendant’s neighborhood status and stability are negatively related to rearrest. Defendants from more affluent and stable neighborhoods are less likely to be rearrested for new crime. Although reliance on individual-level prediction at bail seems warranted, the study underscores the need to further explore the linkage between neighborhood conditions and pretrial outcomes.