Exposure to the criminal justice system itself increases some of the risk factors used to predict recidivism. This is the bottom line of a recently published article in Law and Human Behavior. Below is a summary of the research and findings as well as a translation of this research into practice.
Featured Article | Law and Human Behavior | 2019, Vol. 43, No. 5, 477-490
Criminogenic or Criminalized? Testing an Assumption for Expanding Criminogenic Risk Assessment
Seth J. Prins, Columbia University
Objectives: Proponents of criminogenic risk assessment have called for its widespread expansion throughout the criminal justice system. Its success in predicting recidivism is taken as evidence that criminogenic risks tap into the causes of criminal behavior, and that targeting these factors can reduce correctional supervision rates and even prevent crime. This study challenges these assertions, by testing the implicit assumption that populations in which recidivism risk factors were identified are interchangeable with populations experiencing the onset/duration of exposure to the criminal justice system. Hypotheses: Exposure to the criminal justice system increases some of the risk factors used to predict recidivism; therefore, risk factors for recidivism and onset/duration of exposure to the criminal justice system are not interchangeable. Method: Secondary analysis of data from 503 boys followed prior to first arrest through early adulthood. Inverse-probability-of-exposure-weighted marginal structural models and fixed effects models were employed to test whether arrests and convictions increase antisocial attitudes, behaviors, and peers. Results: Being arrested or convicted resulted in
subsequently higher levels of antisocial attitudes, behaviors, and peers. Risks for recidivism, which include the effect of exposure to the criminal justice system, are not identical to the risks of exposure to the criminal justice system. Conclusions: Results caution against the uncritical expansion of criminogenic risk assessment from community corrections to policing, pretrial, and sentencing. Researchers and policymakers should engage with the social conditions that put people at risk of criminogenic risks and more cautiously communicate the scope of reform that criminogenic risk assessment can deliver.
criminogenic risk assessment, criminal justice, methodology, theory, algorithmic accountability
Summary of the Research
“Over the past several decades, criminogenic risk assessment has become an integral, “evidence-based” component of the criminal justice system. Currently, in the ‘fourth generation’ of criminogenic risk assessment, there has been a shift in focus to not only assess risk, but also to reduce it by intervening on criminogenic risk factors that are dynamic and manipulable. This strategy focuses on identifying people at the highest risk of recidivism and targeting them for supervision and treatment. The apparent success of this approach for recidivism reduction has been interpreted as evidence that criminogenic risk assessment taps into the causes of criminal behavior more broadly, and that targeting these factors can therefore also reduce criminal behavior and correctional supervision rates overall. Indeed, an explanatory framework has emerged around criminogenic risk factors as fundamental to the origins of criminal behavior and the roots of crime itself, and a model for organizing and applying this knowledge, the risk-need-responsivity model of correctional assessment and rehabilitative programming, is widely accepted and promoted. Proponents suggest that criminogenic risk assessment can improve sentencing procedures, facilitate jail diversion, reduce prison populations, help scale down mass incarceration, improve policing, reduce violence and corrections spending, increase resources for community development, and ultimately, prevent crime altogether” (p. 477-478).
“Yet, confidence in expanding risk assessment in criminal justice policy and practice may be outpacing the theory and evidence to support it. For example, Andrews and Bonta go so far as to argue that ‘the prediction of criminal behavior is perhaps one of the most central activities of the criminal justice system [because] from it stems community safety, prevention, treatment, ethics, and justice.’ They advocate that criminogenic risk assessment’s ‘. . . theoretical and empirical base . . . should be disseminated widely for purposes of enhanced crime prevention throughout the justice system and beyond.’ This is indeed happening: The use of criminogenic risk assessment is expanding from the back-end of the criminal justice system to the front, in pretrial processing, sentencing, and even policing” (p. 478).
“But statements such as Andrews and Bonta raise concerns about potential conceptual and empirical overreach. For example, should we transport prediction instruments designed for tertiary prevention (i.e., recidivism reduction) for use in primary or secondary prevention (i.e., first arrest, pretrial detention, sentencing)? Such an expansion assumes that the populations in which criminogenic risk factors for recidivism were identified are interchangeable with populations facing the onset and duration of exposure to the criminal justice system. Moreover, such statements imply that interventions successful at reducing recidivism will also be successful at reducing the onset and duration of criminal behavior and exposure to the criminal justice system. This potential overreach matters for legal and correctional professionals and policymakers, who frequently do not understand the actuarial technologies upon which they base their decisions, such as mistaking probabilities for administrative certainties. As concerns about prediction and risk assessment in the criminal justice system enter the popular dis- course, it is crucial that researchers, policymakers, and advocates are clear on not only the ethical, but also the methodological and empirical dimensions of the debate” (p. 478).
“This article tests the question: Is it possible that exposure to the criminal justice system increases criminogenic risk levels? If the answer is yes, and the propensities and dispositions our risk instruments are intended to assess are in fact influenced by the criminal justice system itself, what does that mean for expanding the use of criminogenic risk assessment beyond recidivism prediction?” (p. 478).
“This article’s hypothesis is that exposure to the criminal justice system increases some of the risk factors used to predict recidivism and rearrest; therefore, risk factors for recidivism and onset/duration of exposure to the criminal justice system are not interchangeable” (p. 480).
“In a community-based sample of 503 boys followed from childhood into early adulthood, exposure to the criminal justice system increased subsequent criminogenic risk factors. Each arrest, and to a lesser extent conviction, an individual experienced increased their subsequent antisocial characteristics. Arrests likely showed greater effects than convictions because arrest is a more visceral experience than conviction. Furthermore, because not all arrests resulted in conviction, and single arrests sometimes resulted in multiple convictions, the latter may lack precision” (p. 484).
“[T]he populations on which criminogenic risk assessments are validated are systematically different from populations on which they might be applied at the front end of the criminal justice system. The closely related conceptual problem is with the suggestion that risk factors for individual differences in recidivism are the causes of crime more broadly. This cannot be the case if the risk instruments used to support this claim do not fully distinguish between individual-level propensities for criminal behavior and the criminalizing effects of the criminal justice system itself. Moving further upstream, this conceptual problem can be rearticulated: In the era of mass incarceration, the suggestion that risk factors for individual differences in recidivism are the causes of crime more broadly ignores structural causes of exposure to the criminal justice system that are not fully mediated by individual-level factors” (p. 484).
“The analysis presented here shows that arrests and convictions result in subsequently higher levels of antisocial attitudes, behaviors, and peers among boys followed into young adulthood. Results caution against the wholesale expansion of criminogenic risk assessment from community corrections to policing, pretrial decision-making, and sentencing, as the causes of recidivism may not be the same as the causes of the onset and duration of exposure to the criminal justice system. Future research should engage with the social conditions that put people at risk of criminogenic risks and consider the criminalizing effect of contact with the criminal justice system. Researchers and policymakers should more cautiously communicate the scope of reform that criminogenic risk assessment can deliver, and be more open to population approaches to prevention that account for structural influences on the risk distributions of crime and exposure to the criminal justice system” (p. 487).
Translating Research into Practice
“This analysis shows that it may be inappropriate to apply evidence for the predictive and intervention utility of risk factors for recidivism to broader questions about the onset and duration of criminal activity. This is because exposure to the criminal justice system causes some portion of the risk used to predict subsequent involvement in the criminal justice system. These findings suggest that criminogenic risks identified in samples under correctional supervision may be different than more distal risks that occur prior to first exposure to the criminal justice system” (p. 486).
“For example, post hoc analysis revealed that the strongest predictors of first arrest in this sample, controlling for age, were marijuana use, race, and alcohol use. That said, overlapping confidence intervals suggest that many of the top 10 factors may be statistically indistinguishable. Many of the factors … may be measured, and appropriately treated, at the individual level (e.g., substance use disorders). Indeed, when operationalized at the individual level, many are considered criminogenic risk factors in the risk-need-responsivity framework. However, many can also be conceptualized as social factors that are not fully mediated by individual-level characteristics. In many cases, these social factors may be more appropriate targets for primary prevention, for example, the criminalization of all forms of drug use, racial disparities in arrests, and institutional racism in policing” (p. 486).
Other Interesting Tidbits for Researchers and Clinicians
“The issue of “reverse causation” uncovered in this analysis also raises important questions about the efficiency and effectiveness of criminogenic risk assessment. If exposure to the criminal justice system increases the very risk factors used to predict recidivism or future criminal activity, intervening on these risk factors will not reduce overall levels of criminal justice system involvement unless we also reduce the absolute number of people flowing into criminogenic risk. Population prevention strategies, versus population management strategies, would aim to reduce first exposure to the criminal justice system, not merely deploy criminogenic risk assessments during or after first exposure. True prevention would focus on shifting the risk distribution’s mean, not merely truncating its right tail. Indeed, the originators of criminogenic risk assessment, and many of its proponents, have been among the first to advise against conflating recidivism reduction with primary prevention. But they may be failing to heed their own warning in their calls for the sweeping expansion of criminogenic risk assessment throughout the criminal justice system” (p. 487).
Join the Discussion
As always, please join the discussion below if you have thoughts or comments to add!