Predicting criminal recidivism: A research note

Criminal justice researchers often develop prediction instruments as a practitioner tool for improving the allocation of resources in community corrections administration. Although best practices have emerged for developing predictions, those best practices lead to predictions that fail to distinguish risk factors from control and correctional responses to risk. The consequence is that predictions fail to predict what they purport to predict, and this limits the utility of those predictions for public policy. This note argues that when properly done, predictions pertain to a latent, unobservable population. Given that perspective, some best practices advocated for prediction should be abandoned, and new best practices should be adopted.     

Violent recidivism in criminal psychopaths

The predictive validity of the Psychopathy Checklist-Revised (PCL-R) was compared with 3 actuarial risk scales in a sample of 81 offenders followed for a maximum of 67 months (average of 30 months). The recommittal or general recidivism rate for the entire sample was 57% (40% for nonpsychopaths, 51.2% for a mixed group, and 85% for psychopaths). The violent reoffense rate was 10% for the sample (nonpsychopaths 0%, mixed 7.3%, psychopaths 25%). All instruments were significantly correlated with general recidivism; however, the PCL-R was the best predictor of violent recidivism. Compared to the actuarial scales, the PCL-R had a higher predictive efficiency (Relative Improvement Over Chance (RIOC)) and yielded fewer decision errors. Most importantly, Factor 1 was a better predictor of violent recidivism than Factor 2, suggesting that the trait construct of psychopathy makes a unique contribution to the prediction of violent recidivism.     

Early criminal recidivism among mentally disordered offenders

Criminal recidivism was studied during 2 years in a Swedish population-based cohort (N = 318) of mentally disordered male offenders who had undergone a pretrial forensic psychiatric investigation, been convicted in subsequent trials, and been sentenced to forensic psychiatric treatment (FPT; n = 152), prison (n = 116), or noncustodial sanctions (n = 50). Recidivism was analysed in relation to index sanctions, levels of supervision, diagnoses, and criminological factors. Significantly lower recidivism in the FPT group was related to lower crime rates during periods at conditional liberty in this group alone, and recidivism was significantly more common among offenders with at least one of the two diagnoses of substance abuse disorder and personality disorder than among those with psychotic or other mental disorders alone. Age at index crime and number of previous crimes emerged as significant predictors of recidivism. The results of this study suggest that the relapse rates depend as much on level of supervision as on individual characteristics.     

The impact of program developers as evaluators on criminal recidivism: Results from meta-analyses of experimental and quasi-experimental research

Using meta-analysis, we report on an investigation of the evaluator's influence in the treatment setting on criminal recidivism outcomes. Many evaluators and users of evaluation of social interventions worry that mixing of the roles of program developer and program evaluator may bias results reported in intervention studies in a positive direction. We first review the results of prior investigations of this issue across 50 prior meta-analyses, finding 12 that tested the impact of investigator influence in the treatment setting. Eleven of these reported that effect size increased positively, sometimes substantially so, when evaluators were influential or involved in the treatment setting. We followed this with a meta-analysis of 300 randomized field trials in individually focused crime reduction, also finding intervention studies in which evaluators who were greatly influential in the treatment setting report consistently and substantially larger effect sizes than other types of evaluators. We discuss two major views — the ‘cynical’ and ‘high fidelity’ theories — on why this is consistently the case, and conclude with a further agenda for research.     

Estimating the numbers of prison terms in criminal careers from one-step probabilities of recidivism

A method of using estimates of one-step probabilities of recidivism, i.e., conditional probabilities of individuals returning to prison for the jth time given release for the (j-1)st time, to estimate the numbers of prison terms expected to be accumulated by the individuals, is presented. The method is illustrated by calculating the expected numbers of prison terms separately for racial and gender groups in a large data base of Western Australian prisoners. The recidivism probabilities for these data were estimated by fitting Weibull mixture models to the (possibly censored) times to recidivate. The probabilities increase strongly asj increases from 1 to 6, then level off. Large differences between them are due to racial and gender group and these are reflected in the differing expected prison career durations for these groups. The effect of interventions which might lower recidivism is discussed in the light of the method as applied to these estimates.     

The importance of forensic evidence for decisions on criminal guilt

Recent studies have found that the general public perceives forensic evidence to be relatively inaccurate and to involve high levels of human judgement. This study examines how important the general public finds forensic evidence by comparing decisions on guilt and punishment in criminal cases that involve forensic versus eyewitness testimony evidence and examining whether a CSI effect exists. Specifically, this experimental survey study utilized a 2 (crime type: murder or rape) × 4 (evidence type: DNA, fingerprint, victim eyewitness testimony, or bystander eyewitness testimony) ? 1 (no victim testimony for murder scenario) design, yielding seven vignettes scenarios to which participants were randomly assigned. Results indicate that forensic evidence was associated with more guilty verdicts and higher confidence in a guilty verdict. Forensic evidence did not change the expected sentence length and did not generally affect the ideal sentence length. However, for rape, respondents believed that the defendant should receive a longer sentence when forensic evidence was presented but forensic evidence did not alter likely sentence that respondents expected the defendant to receive. The results of this study did not support a CSI effect. Overall, this study suggests that forensic evidence – particularly DNA – has a stronger influence during the verdict stage than the sentencing stage.     

Reducing criminal recidivism: evaluation of Citizenship, an evidence-based probation supervision process

‘Citizenship’ is a structured probation supervision program based on ‘what works’ principles, designed for offenders on community orders or licenses supervised within the UK National Probation Service. The program was evaluated using survival analysis comparing the reconvictions of a cohort of all offenders in one probation area eligible for Citizenship over a 2-year period (n = 3,819) with those of a retrospective cohort of all eligible offenders in the same probation area receiving ‘traditional’ probation supervision (n = 2,110), controlling for risk related factors. At the 2-year stage, 50% of offenders in the comparison group had reoffended compared to 41% in the experimental group, and the difference between the survival curves was statistically significant. The hazard ratio was 0.69, which represents a 31% reduction in reconvictions in the experimental group over the proportion in the comparison group at any given time. Time to violation of a supervision order or post custody license was also statistically significantly longer in the experimental group. A key element of the program, promoting contact with community support agencies, was statistically significantly related to reduced reoffending in the Citizenship group. The overall effects remained after controlling for differences in risk scores although effectiveness varied by risk level. Contrary to other ‘what works’ research findings, the program was found to be most effective across the low–medium and medium–high risk thresholds, and was not effective with the highest risk group. This difference can be explained and is discussed in terms of risk, need, and responsivity principles. The Citizenship program was found to be cost-beneficial.     

Diurnal relations of criminal behavior

In order to study the correlation between criminal acts and time of day about 900 police reports which covered a period of eight weeks were analyzed. The analysis of 561 reports which have been usuable for evaluation showed that offences against property (theft, burglary) predominantly occur during the day. The activity of burglars more often then the thieves' one goes over into the night. Aggressive offences (including robbery) are mainly observed in the evening and at night. The activity of robbers more often than the ordinary aggressive offenders' one goes over into the next morning. Alcoholic intoxication hardly occurs with offences against property but is significant as far as aggressive offences are concerned. However the day-time variation of occurrence of aggressive offences without intoxication is similar to that of intoxicated offenders.     

Parenthood and adult criminal offending: The importance of relationship quality

Drawing on social control, social learning, and general strain theories, it was hypothesized that the quality of a parent-child relationship would be negatively related to adult crime. In addition, it was predicted that relationship quality would more strongly impact the criminal behavior of single parents, teenage parents, and those with prior criminal records. Data from the National Youth Survey were analyzed using a combination of probit and truncated regression. Results provided partial support for hypotheses.     

Using a natural experiment to measure the impact of swifter punishment on criminal recidivism

Journal of Experimental Criminology - Measure the impact of swifter punishment on the timing of first imprisonment and on criminal recidivism among young violent offenders. A policy reform in 1994...     

Changes in the relative importance of dynamic risk factors for recidivism during adolescence

This study examined which dynamic risk factors for recidivism play an important role during adolescence. The sample consisted of 13,613 American juveniles who had committed a criminal offense. The results showed that the importance of almost all dynamic risk factors, both in the social environment domain (school, family, relationships) and in the individual domain (attitude, skills, aggressiveness), decreased as juveniles grew older. Therefore, the potential effect of an intervention aimed at these factors will also decrease as juveniles grow older. The relative importance of the risk factors also changed: In early adolescence, risk factors in the family domain showed the strongest association with recidivism, whereas in late adolescence risk factors in the attitude, relationships, and school domain were more strongly related to recidivism. These results suggest that the focus of an intervention needs to be attuned to the age of the juvenile to achieve the maximum potential effect on recidivism.     

解读刑法第18条——兼谈对精神病人违法行为处理的不足和完善

司法实务中，人们对《刑法》第18条的理解有些偏差，如①对条文中提到的“精神病”只停留在精神医学的层面来加以解释；②错误地认为辨认能力和控制能力是法学条件；③错误地认为发病期的精神病人作案皆无罪或减罪；④错误地认为鉴定医生是评定刑事责任能力的主体。此外，该条规定需要完善，如①是否可考虑去掉第2款；②对有病无罪的那部分病人处理措施规定得太单一；③宜增加对吸毒者刑事责任能力的规定。     

解读刑法第18条——兼谈对精神病人违法行为处理的不足和完善

司法实务中,人们对《刑法》第18条的理解有些偏差,如①对条文中提到的"精神病"只停留在精神医学的层面来加以解释;②错误地认为辨认能力和控制能力是法学条件;③错误地认为发病期的精神病人作案皆无罪或减罪;④错误地认为鉴定医生是评定刑事责任能力的主体。此外,该条规定需要完善,如①是否可考虑去掉第2款;②对有病无罪的那部分病人处理措施规定得太单一;③宜增加对吸毒者刑事责任能力的规定。     

The generality of criminal behavior: A confirmatory factor analysis of the criminal activity of sex offenders in adulthood

The presence of a general construct of deviance had not been investigated in sex offenders in spite of the presumption of specialization that characterized them. Consequently, the aim of this study was to investigate whether sex offenders' criminal behavior could be explained by a general construct of deviance. A semi-structured interview was conducted with each subject in order to assess deviant behavior during the life course. The official criminal behavior of 388 convicted sex offenders was analyzed using confirmatory factor analysis. The pattern of covariance observed showed that the onset, frequency, and variety of criminal activity in adulthood could be explained by a general construct of deviance in sexual aggressors. The results, however, also suggested that child molesters' frequency of sexual crime was not part of a general construct of deviance. In other words, the offending behavior of sexual aggressors of women was predominantly versatile, whereas that of sexual aggressors of children tended to be more specific, at least for one dimension of their criminal activity.     

A cross-sectional study on expression of anger and factors associated with criminal recidivism in prisoners with prior offences

Objective: The purpose of this study is to determine demographical characteristics leading to crime recidivism and define anger levels and anger expression manners for those who re-commit crime.Method: All the literate inmates in zmit Closed Penitentiary were included in this cross-sectional study. The prisoners were asked to respond to State-Trait Anger Expression Inventory. Their socio-demographic data were collected and a questionnaire was given to them to determine their state of imprisonment, sentence, nature of the crime in which they were involved, their criminal history, their relationship with inmates and prison staff and substance and alcohol use.Results: Of the 438 prisoners, 302 (68.9%) responded to the questionnaires. Crime recidivism among the study cohort was observed to be 37.4%.Mean trait anger, anger out and anger in scores were significantly higher in prisoners with criminal recidivism in comparison with those who did not have prior criminal records. However, mean anger control scores for prisoners with or without criminal recidivism were similar. Unemployment, education level completed at secondary school or below, having committed a crime under the influence of alcohol or narcotics, having been involved in prison fights, having resisted police officers, caused damage in their vicinity when angry and violent crimes were all found to be possible causes of criminal recidivism. Educational level completed at secondary school or below, getting into fights with other prisoners, unemployment and resisting police officers were determined to be the strongest indicators to predict criminal recidivism when all variables were considered according to a logistic regression model.Conclusion: It can be proposed that those who have problems with officials or hostile towards others constitute a risk group for criminal recidivism. If prisoners with criminal recidivism can be helped to identify and control their anger, their risk of committing a new crime can be minimised.