The purpose of this study was to gain a detailed understanding of the correlates of variation in sentencing outcomes (i.e., whether to incarcerate and the length of sentence) among those convicted of sexual offenses. In particular, this research sought to examine the extent to which contextual factors influenced variation in sex offender sentencing decisions.
Specifically, the following research questions were addressed:
- Research Question 1a: Which individual-level factors account for variation in sentencing outcomes for those convicted of sexual offenses?
- Research Question 1b: To what extent do community-level factors predict variation in sentencing outcomes among sex offenders?
- Research Question 2: Do the effects of these individual and community-level predictors vary significantly across the conditional distribution for sentence length?
- Research Question 3: Are conclusions drawn for Research Questions 1a, 1b, and 2 similar for those convicted of sexual offenses and violent, non-sexual offenses?
The researchers utilized seven years of data (2004-2010) obtained from the Pennsylvania Commission on sentencing (PCS), reflecting trial court convictions under Pennsylvania Sentencing Guidelines. The data, made publicly available by the PCS on an annual basis, provided information on conviction offenses, the defendant's criminal history, demographic characteristics, and the sentences imposed. The researchers restricted the units of analysis to the most serious offense per judicial transaction. The raw annual data files provided by the PCS, when merged, had an overall sample size of 318,906, including 16,860 sex offenders (5.3%), as defined by the Pennsylvania Consolidated Statutes.
The PCS data were supplemented with data from additional sources to capture information on court and community-level influences. Consistent with previous analyses of the PCS data, the researchers utilized the judicial district as the court/community-level unit of analysis. Data derivatives include:
- Data from the Administrative Office of Pennsylvania Courts' annual Caseload Statistics of the Unified Judicial System of Pennsylvania were used to create court size and caseload variables.
- Data from the Association of Religion Data Archives were used to create composite measures of the religious composition of the judicial districts, and from Congressional Quarterly to measure the political landscape.
Defendants convicted in Pennsylvania county criminal courts between January 1st, 2004 and end of year 2010.
Cleaned-Data-2015-R2-CX-0039.sav (cases=318048; 31 variables): this file includes data derived from administrative records for individuals convicted in Pennsylvania criminal courts between January 1st, 2004 and end of year 2010. Variables include information about individual's sentencing, type of offense, disposition, and past criminal history. Judicial district level variables include number of judges, size of court, court caseload, local jail capacity, and density of religious adherents.
Demographic variables include gender, race, and defendant's age at sentencing.