Beyond Mental Illness as a Risk Factor: How Can Forensic Psychologists Refine Parole Suitability Assessments to Address Antisocial Patterns and Criminogenic Needs?

The most cited evidence comes from Skeem and colleagues' matched-sample study of parolees with and without mental illness, which tracked 221 matched parolees for over a year following release. Offenders with mental illness were equally likely to be rearrested as their non-mentally-ill counterparts, although more likely to return to prison custody. Beyond symptom-specific risk factors operationalized through the Historical Clinical Risk Management-20 (HCR-20), parolees with mental illness had significantly more general risk factors for recidivism, measured by the Level of Service/Case Management Inventory (LS/CMI), than offenders without mental illness. The authors concluded that untreated mental illness is, at best, a weak predictor of recidivism. A separate line of research examining the temporal sequencing of psychiatric symptoms and criminal acts reinforces the point: for the majority of mentally ill offenders, psychiatric symptoms did not precede the criminal behavior in time, undermining the assumption of a direct causal pathway.

For parole suitability work, this finding means a formulation cannot rest on an "if symptoms are managed, risk will drop" framework because this does not capture the dynamic for most candidates. That framework applies only to the subgroup whose offending was clearly symptom-driven, and identifying that subgroup is itself an evaluative task rather than a default assumption. The forensic psychologist's job is to specify, for the individual candidate, whether the criminal behavior pattern reflects symptom-driven instability, criminogenic-driven antisociality, or some combination, and to weight the risk assessment accordingly.

The Risk-Need-Responsivity framework, developed by Andrews and Bonta, provides the most empirically supported organization of criminogenic factors. The Big Four (antisocial cognition, antisocial personality pattern, antisocial associates, and a history of antisocial behavior) and the broader Central Eight have been validated across multiple samples and now extend into populations with mental illness. Research examining Big Four and Central Eight factors among psychiatric inpatient populations found that distinct criminal risk factors, not mental illness, were more strongly associated with most criminal behaviors. Forensic psychologists assessing a parole candidate should be able to articulate where the candidate falls on each of these dimensions, with evidence drawn from criminal history, institutional records, collateral interviews, and structured assessment data.

There is also a developmental dimension that bears directly on parole work. A study of New Jersey parole-release decisions identified juvenile criminal onset as a marker that substantially shapes whether criminal risk relates to recidivism among adults with serious mental illness. For early-start offenders, risk level was positively related to recidivism, while for late-start offenders, that relationship was not observed. The practical implication is that two parole candidates with otherwise similar diagnostic profiles can carry meaningfully different risk pictures depending on whether their criminal involvement began in adolescence. Juvenile history provides key information that changes which intervention targets are likely to reduce future violence and which are unlikely to.

Substance use and co-occurring disorders deserve specific attention in the formulation. The criminogenic literature increasingly suggests that the elevated recidivism risk often attributed to serious mental illness is largely accounted for by co-occurring substance use disorders rather than by psychiatric symptoms alone. A formulation that distinguishes the candidate's symptom-driven, substance-driven, and antisocial-pattern-driven risks gives the parole board a more accurate picture of what release conditions and interventions are likely to bear on the outcome.

Forensic psychologists working in parole contexts typically draw on a mix of static and dynamic instruments, and the criminogenic-needs framing changes how the data should be weighted and presented. Static actuarial tools provide an anchor at a point in time, while instruments such as the LS/CMI map directly onto the Central Eight and are well-suited to parole suitability formulations. The HCR-20 v3 incorporates both historical and dynamic items and can isolate the contribution of symptom-specific risk factors from broader patterns. The PCL-R has its own role and its own concerns. A study of the rapid growth of Psychopathy Checklist-Revised (PCL-R) use in U.S. courts found that PCL-R use increased from 87 cases between 1991 and 2004 to 348 cases between 2005 and 2011, with California parole hearings and sexually violent predator proceedings accounting for the majority of the growth. Roughly a quarter of cases involved parole determinations, and PCL-R scores reported by prosecution evaluators averaged seven points higher than scores reported by defense evaluators on the same defendants.

That seven-point allegiance gap has a direct bearing on how a PCL-R score will be received in a parole context. A high score is strongly associated with high-risk statements in the testimony that follows, and admissibility challenges have been infrequent and largely unsuccessful. An inflated or under-anchored score, therefore, carries substantial weight in the parole decision. The remedy is to score the PCL-R conservatively, document rater training and reliability checks, and present the score as one input into a broader formulation that integrates criminogenic factors rather than treating psychopathy as the sole risk variable.

The framing question for the parole board is what the instruments collectively suggest about future offending and what would change that picture. Structured professional judgment frameworks integrating static, dynamic, and risk-management items produce more clinically useful conclusions than single-instrument reports. From the criminogenic-needs perspective, dynamic items become more informative when they are tied to identified criminogenic domains rather than to general clinical impressions, and risk-management recommendations become more defensible when they correspond to factors with established links to recidivism.

A parole board reads a violence risk assessment to inform a release decision and, in many jurisdictions, to shape the conditions of supervision and treatment if release is granted. The criminogenic-needs framing improves both. For the release decision, the formulation should specify which factors are driving the candidate's risk, which factors are amenable to change, and which intervention targets have evidence of reducing recidivism rather than merely managing symptoms. The practical takeaway from the Skeem line of research is that mental health treatment that does not address criminogenic factors leaves the actual drivers of recidivism unaddressed. The forensic psychologist's recommendations should reflect this learning, naming criminogenic targets where they apply rather than defaulting to symptom-focused referrals.

For supervision conditions, the same logic applies. Conditions tied to antisocial peer associations, substance use, employment stability, and procriminal attitudes have a stronger evidence base for reducing recidivism than conditions tied solely to symptom monitoring. That does not mean symptom-focused conditions should be removed from the formulation when they are clinically indicated; it means they should be paired with criminogenic conditions where the risk picture warrants. A cognitive-behavioral referral targeting criminal thinking patterns, for example, has different recidivism implications than a psychiatric medication review, and a formulation that recommends both where appropriate gives the supervising agency more leverage than one that names only the latter. Parole-decision research in a large sample of parole-eligible individuals in Pennsylvania found that, despite inconclusive evidence on whether mental illness is itself a risk factor for criminal behavior, individuals with mental health problems generally fare worse on the risk assessment tools used in parole decisions. The forensic psychologist's report can either reinforce that pattern or counter it, depending on how the formulation is structured.

Finally, the forensic psychologist should consider how individuals with mental health issues are treated and monitored in a broader context. The Skeem matched-sample study found that parolees with mental illness were more likely to return to prison custody despite being equally likely to be rearrested, which the authors interpreted as evidence of supervision disparities. Parole officers may respond differently to similar conduct when the parolee carries a mental illness diagnosis. A forensic psychologist whose report frames mental illness as the central risk driver may inadvertently contribute to that dynamic by inviting supervision responses calibrated to symptom presentation rather than to behavior. A formulation that distinguishes symptom-driven from criminogenic-driven risk, and that specifies which supervision triggers correspond to which, is more useful to the board and more protective of release outcomes for the candidate.

Refining parole suitability violence risk assessments around criminogenic needs, incorporating mental illness factors where appropriate, is a correction to the proportional weight that mental illness is typically given. For most parole candidates with serious mental illness, the antisocial pattern, the criminal history, the procriminal cognition, and the social context predict future violence more powerfully than the diagnosis or the symptoms. Forensic psychologists who structure their assessments accordingly produce formulations that the empirical literature supports, that parole boards can act on with greater confidence, and that point toward interventions with evidence of reducing recidivism rather than interventions that address only the most visible features of the candidate's presentation. 

eBook

• The Forensic Psychologist as Risk Assessor
  
  

Training

• Limited-Time Specially Priced Risk Assessment Training Bundle
• Violence Risk Assessment Certificate

Blog Posts

• Mentally Ill Offenders Need Criminogenic Treatment Too
• Psychiatric Symptoms Do Not Precede Criminal Behavior for the Majority of Mentally Ill Offenders
• What Psychotherapeutic Interventions Should Target When Treating People with Mental Illness and Criminal Justice Involvement
• Juvenile Criminal Involvement May Signal the Presence of Heightened Criminogenic Need Among Adults with SMI
• Rapid Growth of PCL-R Use in U.S. Courts Accounted for by Prosecution of Sexually Violent Predators and Parole Hearings