Report: Predicting New Criminal Convictions while on Parole—The Role of Offense Type 2019

Summary:

Understanding factors behind sustained criminal justice involvement is a primary focus of criminal justice research and professionals. Here we examine the relationship between offense type and new criminal convictions for offenders on parole in Utah. By analyzing 3,173 parolees released from Utah’s state prison in 2013 and 2016, we find that offenders convicted of a sex-offense have a reduced likelihood of being convicted of a new crime when compared to other offense types using an average follow-up time of 382.5 days. We further find that new criminal convictions while on parole are predominantly made up of non-violent offenses, with less than 2 percent of sex-offenders being re-convicted of a sex-related crime. The costs associated with incarceration and lengthy supervision terms merits a careful evaluation of the risk, severity, and cost of a particular offense to re-occur.

Data & Method

As informed by the literature, demographics, including age at parole start, gender, and race/ethnicity are included as important control variables. Additionally, explanatory variables, including the parolee’s offense severity (felony 1, felony 2, or felony 3), an indicator variable denoting if the current parole start is the first start associated with the current incarceration, and risk to re-offend score are included in the model. The outcome variable of interest “nc_2” denotes a return to prison within 2-years of parole start while “parole_los” denotes the number of days spent on parole before being revoked to prison or discharged. A full definition of variables is presented in Table 1.

A total of 3,173 unique parole starts were obtained from the Utah Department of Corrections’ (UDOC) database O-Track and then followed for a maximum of two years. Summary statistics are presented in Table 2. The average length of stay on parole before a revocation or being discharged was 382.5 days, or 1.05 years. Overall, 84 percent of the parolees were males and had an average age of 35. Seventy-one percent were high or intensive risk to re-offend, which differed by group (see figure A.1 in Appendix A). Specifically, the percent of parolees that were high or intensive risk to reoffend had a range between 44 (those whose primary offense was a sex offense) and 82 percent (those whose primary offense was a drug possession only offense).

In terms of the parolee’s primary offense type,8 property offenses (32%) made up the largest percent of offenses followed by person offenses (19%), sex offenses (15%),9 alcohol/drug and drug (13%), and drug possession only offenses (9%). A smaller percent of offenders pertained to driving, weapon, murder,10 and a category denoted as “other” offenses. Summary statistics (mean values) by offense type are further available in Table A.1 in Appendix A.

Severity of New Convictions

In this section, we explore the severity of the new criminal convictions that occurred during the two-year follow-up time. Specifically, here we examine the nature of these offenses. Figure 2 illustrates the distribution of offenses that constituted these new criminal convictions. As seen in the figure, close to 30 percent of new convictions while on parole were drug possession only offenses. This was followed by person offenses (24%), other (13%), driving (12%), person (9%), alcohol/drug (8%), weapons (4%), and sex offenses (1%).

Read the full report: