Sex-Offender Risk Assessment: Static-99R, Stable-2007 and SRM

The Static-99R contains ten items. Five relate to sexual offence history: prior sex offences (non-contact + contact), current non-contact sex offence, number of sentencing dates, any male victim, and any unrelated victim. Five relate to general criminal history and demographic factors: any non-sexual violent offence at the current or prior sentencing dates, any prior non-sexual violent offence, four or more sentencing dates, any stranger victim, and age at release (the age-weighting item added for the -R revision, reflecting the meta-analytic finding that age at release is among the strongest actuarial predictors of sexual recidivism, with older offenders showing substantially lower recidivism rates).

Each item is scored 0 or 1 (with the prior-sex-offences item scored 0, 1, 2, or 3, and the sentencing-dates item scored 0, 1, or 2). Total scores range from -3 to 12. The standardised normative tables, which Public Safety Canada updated in 2019 using a large-scale dataset aggregation across the US, UK, Canada, and Australia, convert total scores to approximate 5-year and 10-year sexual recidivism probabilities and to relative risk labels (Below Average, Average, Above Average, Well Above Average) for comparison against the broader offender population.

The age-weighting adjustment distinguishes the Static-99R from the original Static-99. A person scoring 6 on the underlying items who is released at age 60 receives a meaningful reduction in the adjusted total score compared with the same person released at age 25. This reflects the well-replicated finding that sexual recidivism rates drop substantially with age, a finding robust across Western forensic samples and partially supported by the limited data from Indian forensic case series.

The Static-2002R, a parallel instrument also by Thornton and Hanson, includes different item selections and was intended to improve upon specific Static-99 weaknesses. The two instruments correlate highly (approximately r = 0.75) but differ somewhat in predictive accuracy depending on the subsample. Many jurisdictions and institutions use both and report the convergence or divergence between the two scores.

The Stable-2007 was developed by Karl Hanson and Andrew Harris to measure dynamic risk factors for sexual recidivism: characteristics of the individual that change over months to years and that treatment programmes specifically target. It contains thirteen items organised across five domains: significant social influences (problems with negative social influences, no emotionally intimate relationships, hostility toward women); intimacy deficits (emotional identification with children, capacity for relationship stability); sexual self-regulation (sexual preoccupation, sex as a coping mechanism, deviant sexual interests); general self-regulation (impulsive acts, poor problem solving, negative emotionality); and cooperation with supervision (attitude toward the law and supervision).

Each Stable-2007 item is scored 0 (no problem), 1 (maybe a problem), or 2 (yes, a problem), with a total score range of 0 to 26. Higher scores indicate greater dynamic risk. The Stable-2007 normative data show that it adds incremental predictive validity over Static-99 alone for sexual recidivism, with meta-analytic AUC values in the 0.65 to 0.70 range, slightly below those typically achieved by the Static-99R.

The Acute-2007 measures rapidly fluctuating dynamic risk factors that function as situational triggers rather than enduring dispositional characteristics. It contains seven items: victim access, emotional collapse, collapse of social support, hostility, substance abuse, sexual preoccupation, and rejection of supervision. The Acute-2007 is designed for use at each supervision contact rather than at six-month intervals. A person who is otherwise in a stable low-risk configuration but who shows acute victim access combined with current substance intoxication is presenting a qualitatively different immediate risk profile from their baseline Stable-2007 assessment.

The Static-Stable-Acute integration model, operationalised in the ATSA (Association for the Treatment of Sexual Abusers) practice standards and in the Corrections Services Canada risk management protocol, assigns supervision intensity based on the combined Static-99R level (which sets the floor for supervision intensity) and Stable-2007 change (which can raise or lower the intensity within the range permitted by the static baseline).

The Sex Offender Risk Management (SRM) framework developed by the Association for the Treatment of Sexual Abusers (ATSA) and operationalised through the US Department of Justice's Center for Sex Offender Management (CSOM) integrates assessment, treatment, and supervision into a continuous management cycle. The framework has four components: risk assessment using validated instruments (Static-99R, Stable-2007, Acute-2007); treatment matching to risk level and dynamic factors (higher-risk offenders receive more intensive treatment, per the Risk-Need-Responsivity model); supervision intensity calibrated to current risk level; and community safety planning that addresses victim access, relapse precursors, and crisis protocols.

The Andrews-Bonta Risk-Need-Responsivity (RNR) model provides the underlying treatment-assignment logic: treatment intensity should match risk level (higher static risk gets more hours of structured treatment), treatment should target criminogenic needs (the Stable-2007 items identify the specific dynamic needs that treatment must address), and treatment delivery must be responsive to learning style, cognitive capacity, and cultural background. The RNR model is implemented in the UK Horizon Programme (which replaced the discontinued Sex Offender Treatment Programme in 2017 after the Ministry of Justice's own meta-analysis found the previous SOTP produced null-to-negative outcomes), in Australian State and Territory sex-offender treatment programmes, and in Canadian federal corrections.

The India-specific context adds significant complexity. The Protection of Children from Sexual Offences Act 2012 (POCSO) established criminal liability for a range of child sexual abuse offences but does not create a sex-offender registration or management system. The BNS 2023 § 74 (assault or criminal force to woman with intent to outrage modesty) and § 64-70 (rape and aggravated rape provisions) define the primary sexual offending categories but do not mandate post-sentence risk assessment or management. Several Indian state governments have introduced sex-offender registration obligations by executive order, but there is no national registry, no standardised post-release management protocol, and no mandated risk assessment instrument. NIMHANS and IHBAS use Static-99R and Stable-2007 in research and clinical contexts but have not produced Indian normative supplements.

The treatment gap this creates is significant: sex offenders released from Indian prisons receive minimal structured rehabilitation, the absence of a national registry means supervision is fragmented, and the tools that Western jurisdictions use to differentiate high-risk from lower-risk offenders are applied without local normative data. The Mental Healthcare Act 2017's provisions on treatment of persons with mental disorders do not specifically address sex-offender-specific treatment programming.

In England and Wales, the Multi-Agency Public Protection Arrangements (MAPPA) framework, established under the Criminal Justice Act 2003, is the operational infrastructure for sex offender management post-release. MAPPA convenes police, probation, and prison services around individual high-risk offenders, using Static-99R and Stable-2007 scores as inputs to the risk categorisation that determines the level of multi-agency resource deployment. Scotland has an equivalent framework under the Management of Offenders etc. (Scotland) Act 2005.

The Static-99R normative sample, as reported in Helmus et al. (2012) in Psychological Assessment, aggregates data from twenty-three independent samples totalling 8,106 adult male sex offenders from primarily North American and Western European settings. The normative distribution that converts a raw score to a relative-risk label (Below Average through Well Above Average) is drawn from this combined sample. The 5-year and 10-year recidivism probability estimates are conditional on the base rates in those samples.

The cross-racial validity of Static-99R has been examined within the United States using samples that include Black, Hispanic, and Indigenous American offenders. Findings by Flores et al. (2009) and by Lehmann et al. (2016) suggest that the instrument performs comparably (similar AUC values, no systematic over- or underestimation) across racial groups within the US criminal justice system, when offenders are drawn from US correctional populations. However, this within-US finding does not automatically extend to populations from different national justice systems with different reporting, conviction, and case-selection processes.

The Australian Institute of Criminology's 2019 review found that Static-99R AUC performance was somewhat lower for Indigenous Australian offenders than for non-Indigenous Australian offenders, a pattern consistent with differences in victim access structures, recidivism detection rates, and demographic characteristics of the Indigenous forensic population. The New Zealand Department of Corrections developed a Static-99R normative supplement for Maori offenders.

For India, no published normative validation study for Static-99R or Stable-2007 exists as of the date of writing. Indian sex-offence conviction and reconviction data face additional complexity: a large proportion of Indian sexual offending does not result in FIR registration, cases involving lower-caste or economically vulnerable victims are disproportionately underreported, and the IPC-to-BNS transition has altered the statutory definitions of relevant offences. These structural factors mean that a normative sample drawn from Indian prison populations would not have the same representational properties as Western samples even if one were collected.

The practical recommendation endorsed by the ATSA practice standards 2014 for cross-cultural application is: use the instrument with the best available empirical record (Static-99R), disclose the normative sample's demographic profile and its differences from the assessed individual, report the raw score and its relative-risk band rather than specific probability figures, and supplement with structured professional judgement about contextual factors the instrument does not capture.

The United States has produced by far the largest body of case law on sex-offender risk instrument admissibility, driven by the sexually violent predator (SVP) civil commitment statutes enacted by nearly all states following the Kansas v. Hendricks (1997) Supreme Court decision. Under these statutes, the state must demonstrate by clear-and-convincing evidence (some states require beyond-a-reasonable-doubt) that a convicted sex offender is likely to commit additional sexually violent offences if released. Actuarial risk testimony is the primary vehicle for establishing that likelihood.

The Static-99 and Static-99R have been admitted under Frye (general-acceptance standard) in California, Florida, New York, and other Frye-jurisdictions, with courts finding that the instrument has achieved general acceptance in the forensic psychology community. In Daubert-jurisdictions, the Static-99R has been subject to more searching scrutiny. In US v. Shields (D. Mass. 2013), the court admitted Static-99R testimony but required the evaluator to explain the normative sample, the relative-risk labelling system, and the distinction between the population-level probability and any individual prediction. In re Detention of Bollman (Washington Court of Appeals, 2009) found that Static-99 meets the Frye standard.

The most significant challenge to Static-99 testimony has come from the normative revision itself. When the 2009 normative update reduced the relative-risk labels assigned to several score bands (a re-analysis finding that the earlier normative sample had somewhat higher base rates than are typical), defence attorneys in SVP proceedings successfully argued that offenders previously labelled "high risk" should be re-evaluated using the updated norms. Several state courts required retrospective re-evaluations of committed individuals following the normative update.

In England and Wales, the Static-99R is used operationally in MAPPA risk categorisation and by forensic psychologists preparing reports for the Parole Board. It has not been the primary subject of admissibility challenges comparable to the US SVP proceedings, partly because England does not have an SVP-equivalent civil commitment statute that makes risk-instrument testimony so directly determinative of liberty. The Criminal Procedure Rules Part 19 (expert evidence) require reliability and relevance but do not apply the Daubert four-factor test formally.

Canada, where the instrument was developed, uses the Static-99R in federal corrections and in provincial forensic mental health services. The Supreme Court of Canada's general expert evidence admissibility standard (R v. Mohan, 1994, applied and refined in White Burgess Langille Inman v. Abbott and Haliburton Co., 2015) requires relevance, necessity, absence of exclusionary rule, and proper qualification. Canadian courts have consistently admitted Static-99R testimony on those grounds.

In India, BSA 2023 § 39 governs expert opinion admissibility. As discussed in the companion topic on general violence risk assessment (see also the forensic-psychology-expert-witness-and-daubert-challenges topic for the full admissibility framework), the provision does not establish Daubert-equivalent reliability criteria. A forensic psychologist presenting Static-99R testimony in an Indian court proceeding (likely in the context of a post-conviction sentencing report or a BNSS § 367 mental-state proceeding) should follow the same transparency principles that good practice requires in Western jurisdictions.

The Association for the Treatment of Sexual Abusers (ATSA) publishes practice standards that govern how its member practitioners use risk assessment instruments in clinical and forensic contexts. The 2014 Practice Standards and Guidelines for Members of the Association for the Treatment of Sexual Abusers include specific guidance on Static-99R application: the instrument should be scored from official records where possible, any departures from standard administration should be documented, and the score should be reported as a relative-risk label (with the underlying normative source identified) rather than a raw probability figure.

The ATSA standards also address three common misuses of Static-99R. The first is treating the score as a sufficient basis for a treatment placement decision without considering dynamic risk factors. A person in the Well Above Average static risk band who has completed a structured cognitive-behavioural treatment programme, shows low Stable-2007 scores post-treatment, and has strong community support presents a different management picture than a person with the same static score who refuses treatment and has elevated Stable-2007 items. The second misuse is applying Static-99R to adolescent sex offenders. The instrument was validated exclusively on adult males and its items (prior adult conviction, age at release) do not translate to juvenile populations. Juvenile-specific tools, including the JSORRAT-II and the JRAS, are the appropriate instruments for adolescent risk assessment. The third misuse is applying Static-99R to female sex offenders. The instrument has no female normative data, and female sexual offending pathways differ substantially from male pathways in ways that make the item set theoretically inappropriate.

The Australian Sex Offender Comprehensive Risk Evaluation (SORE) framework, developed by the Australian Institute of Criminology in consultation with state corrective services, adopts the ATSA-aligned Static-99R and Stable-2007 combination as the foundational assessment structure for sentence planning and pre-release decisions. New South Wales, Victoria, and Queensland each have their own legislated high-risk-offender supervision schemes that use SORE-framework assessments as the primary input to court applications for extended supervision orders.

The New Zealand Parole Board uses Static-99R and Stable-2007 assessments routinely in release decisions, operationalising the framework under the Parole Act 2002. The Corrections Act 2004 includes provisions for extended supervision orders for high-risk child sex offenders that require risk assessment evidence meeting the Mohan standard adapted for New Zealand courts (R v. Calder, NZCA, 2008).