Nonviolent Recidivism Research Articles

A Meta-Analysis of the Effectiveness of Restorative Justice Programs in Reducing Recidivism: A Global Perspective

Restorative justice (RJ) is a growing alternative to traditional punitive justice, emphasizing offender accountability, victim reparation, and community reintegration. This meta-analysis aims to quantify the effectiveness of RJ programs in reducing recidivism across diverse global contexts. A systematic search of electronic databases (Scopus, Web of Science, PsycINFO) from 2018-2024 was conducted, including studies that assessed the impact of RJ on recidivism using experimental or quasi-experimental designs. The primary outcome was recidivism (any reoffense), with secondary outcomes of violent and non-violent recidivism. Random effects models were used to calculate pooled effect sizes. Thirty-two studies met the inclusion criteria (N = 18,654 participants). The overall pooled effect size for RJ on recidivism was -0.11 (95% CI -0.15, -0.07), indicating a small but statistically significant reduction in recidivism compared to traditional justice. Subgroup analyses revealed greater effects for youth offenders (-0.16) and for programs with high adherence to core RJ principles (-0.18). In conclusion, RJ programs demonstrate a modest but consistent effect in reducing recidivism. The findings support the continued implementation and expansion of RJ, particularly for young offenders and in settings with strong adherence to RJ philosophy.

Discriminative and predictive validity of risk assessment measures for women incarcerated for serious violent offences in Australia

Despite the growing population of women in Australian prisons, limited research has explored whether commonly used risk assessments – predominantly developed and tested on men – are valid for women. We investigated the discriminative and predictive validity of the Level of Service Inventory–Revised: Screening Version (LSI-R:SV), Level of Service/Risk, Need, Responsivity (LS/RNR), and the Historical, Clinical, Risk Management 20–Version 3 (HCR-20v3) for Victorian women imprisoned for serious violence (N = 79). The LS/RNR was related to any, violent, and non-violent recidivism, and both the LSI-R:SV and the H-Scale of the HCR-20v3 were related to violent recidivism, with the H-Scale demonstrating strong predictive validity for violence. Four LS/RNR needs domains demonstrated discriminative and predictive validity for any and/or violent recidivism (criminal history, family/marital, alcohol/drug problem, antisocial pattern). Findings are locally significant, showing that the LS/RNR and HCR-20v3 H-Scale are useful for the prediction and discrimination of recidivism for Australian women incarcerated for serious violence.

The Validity of Violence Risk Assessment in Young Adults: A Comparative Study of Juvenile and Adult Risk Assessment Tools

Few studies have addressed the application of violence risk assessment for individuals transitioning from youth to adulthood. For 202 young adults released from Dutch juvenile justice institutions this study investigated the predictive validity and potential disparities in impact of juvenile risk assessment tools (i.e., SAVRY [Structured Assessment of Violence Risk in Youth], and SAPROF-YV [Structured Assessment of Protective Factors for violence risk-Youth Version]), and comparable adult risk assessment tools (i.e., HCR-20V3 [Historical Clinical Risk management-20 Version 3], and SAPROF [Structured Assessment of Protective Factors for violence risk]). Assessments with juvenile and adult risk assessment tools yielded similar predictive validity for violent and non-violent recidivism. Risk and protective factors related to treatability, parents, community participation, resilience, and personality showed individual predictive validity. These findings offer flexibility when applying risk assessment in clinical practice. The choice between youth and adult assessment tools should be made considering the individual’s developmental stage.

The Predictive Validity of the Revised Screening Scale for Pedophilic Interests (SSPI-2).

The Revised Screening Scale for Pedophilic Interests (SSPI-2) is a five-item measure that assesses for pedohebephilia (sexual attraction to prepubescent and pubescent children) based on child victim characteristics. We aimed to replicate findings by Seto, Sandler et al. (2017) by examining the predictive validity of the SSPI-2 in an independent sample of 626 men referred for a sexological assessment because of sexual offending against children. SSPI-2 scores were associated with an increased likelihood of sexual recidivism but were not significantly associated with non-sexually violent or non-violent recidivism. When they were entered together, the SSPI-2 did not contribute additional variance to the Static-99R in the prediction of sexual recidivism. Results are consistent with the findings of Seto, Sandler et al. (2017) and suggest that higher scores on the SSPI-2 may be indicative of an increased risk for sexual recidivism in individuals who have sexually offended against children.

An Economic Analysis of Crime Costs Associated with Psychopathic Personality Disorder and Violence Risk

Given substantial national crime costs and that psychopathic personality disorder (PPD) is a robust predictor of recidivism, a research gap exists concerning the cost of crime attributable to adults with PPD. The current study employed a bottom-up cost of illness approach to estimate the association between PPD and crime costs among Canadian men incarcerated in the federal correctional system ( n = 188). Participants were rated using the Psychopathy Checklist–Revised (PCL-R) and the Historical-Clinical-Risk Management–20 (HCR-20, version 2). Group mean crime costs were highest for participants who scored highly on the PCL-R and were rated high risk on the HCR-20, and higher scores on both measures were associated with prospective costs accrued from violent and nonviolent recidivism. The findings highlight the need to improve the treatment and management of high-risk individuals with prominent psychopathic features, as it has the potential for significant financial savings for criminal justice systems.

Demonstrating the Reliability and Validity of the Self-Appraisal Questionnaire for Assessing Violent and Nonviolent Recidivism with Female Offenders in South Africa

Explanations regarding the involvement of females in crime, justice, and punishment in South Africa are primarily based on the experiences of male offenders with little knowledge about the criminogenic offender assessment of female offenders. The Self-Appraisal Questionnaire (SAQ) is an actuarial measure designed for the identification of offenders’ static and dynamic risk factors and criminogenic needs that may increase or diminish their risk upon release. This study involves a sample of 284 female offenders incarcerated in three inter-provincial correctional centers in South Africa. The goal of this study was to examine the psychometric properties of the SAQ in relation to its reliability, validity, and potential use as part of the criminogenic assessment of female offenders, within the multicultural South African context. Research results indicated that the SAQ is reliable and valid to use with South African offenders.

Dynamic violence risk, protective factors, and therapeutic change in a gender and ethnoculturally diverse sample of court-adjudicated youth.

The present study examined the predictive properties of three youth forensic measures-the Violence Risk Scale-Youth Version (VRS-YV), Structured Assessment of Violence Risk in Youth (SAVRY), and the Structured Assessment of Protective Factors-Youth Version (SAPROF-YV)-in a diverse court-adjudicated sample of 257 youth referred for assessment and intervention services at an outpatient mental health facility, and followed up an average of 9.4 years in the community. Study measures were rated from court and clinical files, along with treatment participation, and recidivism outcome data were obtained from official criminal records. The three measures had strong interrater and convergent validity, and moderate to high predictive accuracy for violent, nonviolent, and general recidivism. The measures significantly predicted outcome across male, female, Indigenous, and non-Indigenous groups; however, prediction magnitudes showed some variability with respect to specific risk/protection domains and outcome types. Cox regression survival analyses demonstrated incremental predictive validity for each violence risk measure, but not protection measures, with respect to each of the three recidivism outcomes. Moreover, pre-/posttreatment measurements of change on the VRS-YV dynamic factors were significantly associated with decreased nonviolent recidivism, controlling for baseline risk and protection. Violence risk (VRS-YV) and protection (SAPROF-YV) scores have the potential to be integrated in meaningful ways to capture the potential risk-mitigating effects of protective factors. Implications for integrating risk, protection, and treatment change information in clinical-forensic service delivery to diverse and violent youth populations are discussed. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

Anger, Violence, and Recidivism in Justice System Involved Youth

Anger is considered integral for motivating youths’ criminal activity. To assess how anger and its features (fuse and duration) are associated with youths’ patterns of violence and offending, we studied official risk and (re)offending data among a sample of justice-involved youth (Mage = 14.87, SD = 1.40) in the United States (N = 548). Short anger fuse was associated with increased likelihood of being currently violent, whereas prolonged anger was related to current violence only in mid- to late-adolescence. Youth who had exclusively nonviolent prior(s) were at greater odds of nonviolent recidivism if they reported a short fuse. Alternatively, youth who had violent prior(s) were at greater odds of reoffending if they reported prolonged anger. Overall reported anger was associated with current violence, but not recidivism. This suggests that, in youth, short fuse predicts the continuation of a nonviolent offending pattern, whereas, among violent offenders, prolonged anger predicts recidivism.

The association between sentencing and recidivism in men who have committed a sexual offence

ABSTRACT There have been few studies examining the impact of sentencing decisions on recidivism in individuals who commit sexual offences, which is important as this population may face more stigma, and different conditions in the community, than individuals who have committed non-sexual offences. The present study examined how sentence type and custodial sentence length impacted the incidence of sexual, violent, and non-violent recidivism in 659 Canadian men who were convicted of a sexual offence. Findings suggested that custodial sentences were associated with a greater likelihood of reoffending for all types of recidivism, except sexual recidivism. Custodial sentences of less than two years were consistently associated with an increased likelihood of non-sexual recidivism compared with those who served no time in custody. There was more mixed evidence for whether custodial sentences and custodial sentence length added to the prediction of recidivism once risk and sexual offence type were accounted for. Practice Impact StatementCustodial sentences, including shorter custodial sentences, was associated with increased violent and non-violent reoffending. Accounting for risk and type of sexual offence introduced greater uncertainty regarding the relationship between sentencing factors and recidivism. These findings contribute to the larger body of literature that questions the rehabilitative impact of lengthy custodial sentences.

Evaluating Bias, Shrinkage and the Home-Field Advantage: Results from a Revalidation of the MnSTARR 2.0

ABSTRACT In 2016, the Minnesota Department of Corrections implemented the Minnesota Screening Tool Assessing Recidivism Risk 2.0 (MnSTARR 2.0), a fully-automated, gender-specific instrument that predicts multiple types of recidivism. Using multiple performance metrics, this study externally validates the MnSTARR 2.0 on a sample of 8,997 people released from Minnesota prisons. The results show that while the MnSTARR 2.0 overestimated risk for felony and nonviolent recidivism, it achieved adequate predictive discrimination, with an average area under the curve (AUC) of 0.73 for males and 0.74 for females. The MnSTARR 2.0 produced predictions that were generally consistent with the observed recidivism outcomes by race/ethnicity, and the amount of “shrinkage” from the internal validation sample to the external validation sample was minimal. The MnSTARR 2.0 significantly outperformed the LS/CMI on a sub-sample of individuals who received both assessments, which provides additional evidence that customized instruments tend to perform better than off-the-shelf tools that are designed to be used with a variety of correctional populations.

The Ontario Domestic Assault Risk Assessment: Predicting Violence Among Men With a Police Record of Intimate Partner Violence in the United States

The Ontario Domestic Assault Risk Assessment (ODARA) is an actuarial risk assessment tool for intimate partner violence (IPV) recidivism. Despite its international use, there is no published validation of the ODARA’s predictive accuracy in a U.S. sample. We studied 356 men in New York police records of IPV against a female partner to examine the ODARA’s predictive accuracy for IPV recidivism (base rate 35%), non-IPV violent recidivism (against a nonpartner; 16%), any violent recidivism (49%), and nonviolent recidivism (50%), in a fixed 2-year follow-up. Using 11 scorable ODARA items, area under the curve values were significant and ranged from .590 to .630, indicating small to medium effects. Expected/Observed indices revealed poor calibration with 2-year IPV recidivism rates in ODARA construction and cross-validation samples. Findings support the generalization of the ODARA’s predictive accuracy in different populations and outcomes, but a need for new norm development for higher risk populations.

Predictive Properties of a General Risk-Need Measure in Diverse Justice Involved Youth: A Prospective Field Validity Study

The current investigation was a prospective field validity study examining the discrimination and calibration properties of a general risk-need tool (Level of Service Inventory–Saskatchewan Youth Edition [LSI-Sk]) in a diverse sample of 284 court adjudicated youths, rated by their youth workers on the measure and followed up an average of 9.3 years. The overall risk level and need total demonstrated moderate predictive accuracy for general, violent, and nonviolent recidivism in the aggregate sample, although area under the curve (AUC) magnitudes fluctuated among gender and Indigenous ethnocultural subgroups. Variability in AUC values for the measure’s eight criminogenic need domains further reflected greater salience of certain needs among subgroups. Finally, clinician rated level of gang involvement incrementally predicted recidivism to varying degrees after controlling for overall risk and need. Implications for responsible use of risk assessment tools as part of individualized and gender/ethnoculturally responsive risk assessment practices with youth are discussed.

Predictive validity of the Structured Assessment of Violence Risk in Youth (SAVRY) among a sample of Asian Canadian youth on probation.

Although past studies suggest that the Structured Assessment of Violence Risk in Youth (SAVRY; Borum et al., 2006) has moderate predictive validity, its predictive validity with Asian youth in Western countries is unknown. We therefore compared the SAVRY's predictive validity in a sample of Asian Canadian versus White Canadian youth. Given that the SAVRY is normed on samples comprising mostly youth who are White, we expected its predictive validity for recidivism would be lower for Asian Canadians than White Canadians. We examined youth probation officers' SAVRY assessments for 573 youth (445 White Canadians, 56 East/Southeast Asian Canadians, and 72 South Asian Canadians) on community supervision (i.e. probation) in a Canadian province. Youth were prospectively followed for an average of 1.97 years (SD = 0.56 years) to determine if they were subsequently charged with violent or nonviolent offenses. Asian Canadians scored significantly lower on risk total scores compared to White Canadians. Predictive validity for violent and nonviolent recidivism fell in the medium to large range for East/Southeast Asian Canadians (AUCs = .69 to .89) and South Asian Canadians (AUCs = .64 to .83). In comparison, predictive validity for White Canadians was generally lower (AUCs = .63 to .77; small to large range). Risk total scores and nonviolent risk ratings significantly predicted nonviolent recidivism better for East/Southeast Asian Canadians (AUCs = .89 and .87, respectively) than White Canadians (AUCs = .77 and .71, respectively). Despite few significant differences between Asian subgroups, predictive validity for nonviolent risk ratings was significantly higher in East/Southeast Asian Canadians (AUC = .87) than South Asian Canadians (AUC = .64). The SAVRY may be a useful tool for predicting recidivism with Asian Canadians. However, future research should examine the SAVRY's predictive validity for youth of Asian descent in different countries and contexts. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

The utility of the Psychopathy Checklist-Revised (PCL-R) facet and item scores in predicting violent recidivism.

There is relatively limited research on psychopathy in non-Caucasian ethnic groups and even less on the utility of the Psychopathy Checklist-Revised (PCL-R) that focuses on PCL-R facet and item scores in predicting violent recidivism. In this study, we assessed the utility of the PCL-R in prospectively predicting violent versus nonviolent recidivism during an 11-year follow-up window. A high-risk sample of 451 incarcerated Korean male offenders was assessed on the PCL-R at baseline. A total of 445 were reconvicted after release (353 violent and 92 nonviolent recidivists). Psychopathy facet scores were higher in violent compared to nonviolent recidivists. Facet 2 (affective) showed the strongest effect size (Cohen's d = 0.53; Percentage change in odds = 22.6%) in predicting violent recidivism. Analyses of the four items constituting the affective facet indicated that callous/lack of empathy (Percentage change in odds = 134.4%) and failure to accept responsibility (Percentage change in odds = 94.5%) were the strongest predictors of violent recidivism. Findings are to our knowledge the first to document the utility of the PCL-R in distinguishing violent from nonviolent recidivism and highlight the role of affective impairment (particularly lack of empathy) in violent recidivism.

Impact of defelonizing drug possession on recidivism

Research SummaryCalifornia's Proposition 47 (Prop 47), passed in November 2014, sought to scale back punishment for selected drug and property offenses, making them misdemeanors rather than felonies. Although others have examined the impacts of Proposition 47 on crime rates, here we examine the impacts on a range of recidivism outcomes specifically for individuals convicted for drug possession offenses. We focus on the defelonization of drug possession because nationally evidence suggests that public and policy maker sympathy for reducing incarceration use is greatest for nonviolent drug offenders. Thus, the greatest relevance of Proposition 47 to the nation lies in its strategy of defelonizing drug possession. We draw on an ongoing data collection effort on criminal justice populations in 12 California counties and link these data to rearrest and reconviction data from county and state‐level sources. We find people who received drug possession convictions after Prop 47 had lower overall rearrest and reconviction rates than people with comparable convictions and criminal histories released prior to the proposition. Overall reductions in recidivism rates for Prop 47 drug offenders were driven by reductions in rearrest and reconviction for drug possession offenses. These findings are robust to a variety of specifications and sensitivity analyses. When we complicate the narrative of declines in nonviolent recidivism, however, we find evidence of a 2.8‐percentage‐point increase in rearrest for crimes against persons, mainly assaults and domestic violence. Reconviction rates for crimes against persons for drug possession offenders also increased by 1.1 percentage points.Policy ImplicationsThese findings add to the growing body of evidence that a targeted reform like Prop 47 does not result in increases in most measures of reoffending for people convicted of drug possession. Although we cannot rule out that the declines we observe in rearrest and reconviction are the result of behavioral change among offenders, it seems likely that those declines are partly, perhaps mostly, the result of changing decision‐making among law enforcement and prosecutors. Converting drug possession to be chargeable only as a misdemeanor has significantly reduced the likelihood individuals convicted of drug possession will return to jail or prison once they have been released and therefore stemmed some of the flow of people with such offenses through the system. Evidence from California, however, shows that there can be small increases in crimes against persons, which provides a cautionary note and points to the need to consider interventions that do not involve criminal justice responses in response to the negative aspects of drug use. Our conclusion discusses considerations in applying our findings to other states that have defelonized drug possession.

Psychopathy, self-identified race/ethnicity, and nonviolent recidivism: A longitudinal study.

Psychopathy has long been noted to play an important role in the prediction of criminal behavior and offending. Although many studies have demonstrated that psychopathic traits are predictive of violent recidivism among offenders, relatively few studies have examined the predictive validity of psychopathic traits for nonviolent recidivism and very few have examined this issue in a sample of offenders in the United States. To address this issue, we examined the predictive validity of psychopathy for both nonviolent and general recidivism using the Psychopathy Checklist-Revised (PCL-R) in a sample of 422 county jail inmates. We also examined whether ratings on Factors 1 and 2 differentially predicted nonviolent and general recidivism and whether predictive validity varied among European American, African American, and Latino American male offenders. Psychopathic traits were modestly predictive of nonviolent and general (total) recidivism. Factor 2 ratings were not more predictive of nonviolent recidivism than Factor 1 ratings in this sample, but the two factor scores also predicted nonviolent recidivism interactively. Psychopathic traits were also predictive of both outcomes in subsamples of European American and African American offenders, but not among Latino American offenders. Findings are consistent in magnitude and pattern with prior studies addressing the prediction of violence, and they show that the relationship between psychopathy and criminal conduct generalizes to the prediction of nonviolent crime in a United States offender sample. Results suggest potential differences between the predictive validity of psychopathy among Latino American offenders and other racial/ethnic groups, which suggest the need for additional research. (PsycINFO Database Record (c) 2018 APA, all rights reserved).

Reducing Violence Risk? Some Positive Recidivism Outcomes for Canadian Treated High-Risk Offenders

In pursuit of “what works” in violent offending behavior programs, there remain insufficient evaluations of program outcomes. Three hundred forty-five offenders from the Canadian Violence Prevention Program (VPP) were compared after an average 3-year follow-up with 338 non-VPP participants. Outcomes measured were new convictions for violent, sexual, or general offenses. Intent-to-treat design was used. Subsequently, participants who completed or did not complete the program were compared with the non-VPP group. Further analyses considered Indigenous and non-Indigenous subgroups. Overall, lower recidivism rates were associated with VPP completion, both in the complete sample and ethnic subgroups. However, the main finding of significantly lower likelihood of violent recidivism was found only for the Indigenous offenders, while significantly lower likelihood of general (nonviolent) recidivism was specific to non-Indigenous offenders. Results are interpreted cautiously in relation to program effectiveness given the quasi-experimental design and the important implications of outcome studies for correctional services.

A randomised controlled trial of a cognitive skills programme for offenders with mental illness.

Interventions for offenders with mental illness have tended to be confined to treatment of illness, with the expectation that symptom reduction will be accompanied by reduced criminal recidivism, but recent evidence suggests that other treatment targets may be more effective against recidivism. The aim of this study was to examine the effect of a cognitive skills programme (Reasoning and Rehabilitation 2: Short Version for Adults [R&R2]) among offenders with mental illness. Our first hypothesis was that participation in this programme would result in significantly greater improvement in antisocial attitudes than among similar prisoners who did not participate; both groups received "treatment as usual" (TAU). Our second hypothesis was that those receiving R&R2 would show less post-treatment violent or general recidivism than those receiving TAU alone. Incarcerated offenders with serious mental illness (N=101) were randomly assigned to R&R2 or TAU alone. Criminal attitudes and mental state were examined before and after treatment. Violent and non-violent recidivism was measured, on average, 18months after release. In both intervention and TAU alone trial arms, there were significant pre- to post-treatment changes in criminal attitudes and symptoms or signs of mental disorder. There was no difference between groups in these respects. These pre/post changes were not associated with reductions in recidivism over time, whether or not controlling for baseline risk. Although there was no demonstrable advantage of R&R2 over TAU alone, non-significant trends towards lower violent recidivism in the R&R2 group and general recidivism in the TAU group suggest that it may be worth repeating the trial in a larger sample with more differentiated control groups.

Antisocial Process Screening Device Subscales Predict Recidivism in an Australian Juvenile Offender Sample

The present study is the first to examine the psychometric properties of the self-report Antisocial Process Screening Device (APSD-SR), and the predictive utility of its subscales for reoffending, among Australian juvenile offenders (N = 308, M age = 17.00, SD = 1.49). Exploratory factor analysis supported a modified three-factor structure in which four items loaded differently to prior studies. Total APSD-SR and modified subscale scores were positively associated with criminal history and mental health problems (e.g., internalizing and externalizing problems, alcohol and substance abuse/dependence). Survival analyses indicated that youth scoring high on the APSD-SR total score were faster to reoffend nonviolently (Hazard Ratio [HR] = 1.31, p = .0003) and violently (HR = 1.42, p = .0003) than those scoring low. Whereas the modified grandiose-manipulative subscale predicted faster time to nonviolent recidivism (HR = 1.18, p = .026) as a single predictor, when all subscales were simultaneously entered into the model only callous-unemotional (CU) traits and impulsivity predicted nonviolent recidivism (HR = 1.19, p = .026 and 1.22, p = .015, respectively), and only impulsivity predicted violent recidivism (HR = 1.26, p = .014). Findings inform current understanding of the relative contribution of adolescent psychopathy dimensions to designating a particularly high-risk group of Australian youth in custody.

Therapy drop-out as apredictor of re-delinquency : Legal probation of substance addicted patients according to §64 of the German Legal Code

Recidivism rates in substance-addicted patients placed in institutions according to §64 of the German legal code are approximately 50%, 3 years after discharge from inpatient treatment. The recidivism rates of patients with premature termination of inpatient treatment who had then been referred back to prison and were finally discharged into the community are unknown. Is premature termination of treatment arisk factor for recidivism? Patients released from forensic treatment according to §64 of the German legal code were followed up for violent and non-violent recidivism. Full data were acquired for Baden-Württemberg patients released in 2010 and 2011 with regular vs. premature termination of treatment. All measures revealed highly significant group differences: 48% of the patients discharged after subsequent prison sentences recidivated within the first year and 73% within 3 years after discharge. Among recidivists, the severity of offences was much higher (odds ratio> 3.8 each). Regularly discharged patients also re-offended to aremarkable extent (50%). Patients serving prison sentences after unsuccessful forensic treatment are ahigh-risk group for recidivism. Alternative concepts of clinical and legal treatment of this group should be developed.