Snap Shot
- The question: How do intellectual disability and related cognitive conditions bear on culpability and sentencing in sexual offense cases, and what do the Supreme Court’s capital decisions contribute outside capital cases?
- What the research shows: Intellectual disability involves deficits in reasoning, learning, and adaptive functioning that can affect judgment and susceptibility to influence. Experimental and archival studies show that evidence of a mental or neurological condition can both reduce and increase punishment, often because it raises concern about future dangerousness. In an unpublished New Jersey thesis of 3,066 people released from prison or civilly committed after sexual offenses (IQ below 80 or an intellectual disability diagnosis, so including borderline functioning), those classified as having intellectual disability scored higher on two actuarial tools and were more often civilly committed (21.7% vs. 9.6%) without differing in new charges or reincarceration; the study excluded committed individuals from recidivism analyses and should be cited cautiously. A peer-reviewed New Jersey study (Callahan et al., 2025) found low reincarceration in both groups.
- What it does not show: No study shows that intellectual disability causes sexual offending, that it reduces harm to the person victimized, or that a particular defendant’s condition explains his offense. Mock-juror and judge studies use hypothetical cases, few involving sexual offenses against children.
- The law: Atkins, Hall, and Moore bar execution of people with intellectual disability and require courts to follow current clinical standards; they do not govern noncapital sentencing. Federal courts consider disability under 18 U.S.C. § 3553(a); Amendment 836 (effective Nov. 1, 2025) moved most departures, including the Chapter 5H specific-offender-characteristic policy statements, to Appendix B, and the Commission described the change as outcome neutral. Arizona lists significantly impaired capacity as a mitigating circumstance (A.R.S. § 13-701(E)), and R.C.M. 1001 and 1002 govern military sentencing.
- Issues for each side: Prosecutors should ask whether the condition is reliably established and connected to the offense, and should keep the victim’s harm and every sentencing purpose in view; defense counsel should establish the diagnosis by current clinical standards, explain how it relates to the conduct, and answer dangerousness concerns with a concrete treatment and supervision plan.
When a defendant convicted of a sexual offense has an intellectual disability or another cognitive condition, the condition often becomes central at sentencing. Defense counsel may argue that it reduces blameworthiness, that the defendant needs treatment more than prolonged confinement, or that a standard sentence would be disproportionate. Prosecutors may respond that the condition does not lessen the harm to the child or adult victimized, that it may increase the risk of reoffending, or that it has not been established. Both positions draw on real findings.
Two premises frame this paper. First, intellectual disability does not by itself excuse sexual offending and does not reduce the harm done to victims; the Supreme Court said in Atkins v. Virginia (2002) that the deficits associated with the condition do not warrant an exemption from criminal sanctions, although they diminish personal culpability. Second, group research cannot establish how a condition affected a particular defendant’s conduct; that question requires individual evaluation. This paper uses person-first language (“people with intellectual disability”); the term “mental retardation” appears only in older case language and study titles. Competency and Miranda issues are addressed in “Competency, Miranda Waiver, and Accommodations,” and autism is addressed in the Autism and Online Sexual Offense hub.
What the Supreme Court Decided and What It Did Not
In Atkins v. Virginia (2002), the Court held that the Eighth Amendment bars execution of people with intellectual disability. Its reasoning has two parts that are often quoted in noncapital sentencing. First, the Court described diminished capacities to understand and process information, learn from experience, reason logically, control impulses, and understand others’ reactions, and it concluded that these deficits reduce culpability enough that retribution and deterrence do not justify death. Second, the Court noted that such defendants face special risks of wrongful execution, including false confessions, difficulty assisting counsel, and demeanor that may appear to show lack of remorse. It also observed that evidence of intellectual disability can be a two-edged sword, because it may make a jury more likely to find future dangerousness.
Hall v. Florida (2014) held that a state may not treat an IQ score above 70 as conclusive without accounting for the standard error of measurement, and Moore v. Texas (2017) held that courts must be informed by the medical community’s current diagnostic framework. Moore rejected reliance on perceived adaptive strengths, such as improved behavior in prison, when clinicians focus on adaptive deficits and caution against relying on strengths developed in controlled settings, and it recognized that intellectual disability often co-occurs with other conditions.
These decisions govern eligibility for the death penalty. The Supreme Court has extended a categorical Eighth Amendment rule outside capital cases for juveniles (Graham v. Florida, 2010), but it has not done so for intellectual disability. In a noncapital sexual offense case, Atkins does not require any particular sentence. Its description of reduced culpability, and the clinical standards that Hall and Moore require, are nonetheless commonly cited in mitigation, and courts and experts in any forum are expected to diagnose the condition according to current standards.
The Clinical Basis for Mitigation
Intellectual disability is defined by deficits in intellectual functioning and in adaptive functioning, with onset during the developmental period (American Psychiatric Association, 2022). A diagnosis therefore requires more than an IQ score: it requires evidence of how the person functioned in conceptual, social, and practical domains, ideally from school records, earlier testing, and informants who knew the person before the charges. Some defendants offered as having an intellectual disability function in the borderline range, which is not a diagnosis of intellectual disability but may involve some of the same limitations.
The mitigating relevance of the condition depends on its connection to the offense. Evidence may show that the defendant misunderstood social or legal rules, was influenced or directed by another person, acted impulsively, or had limited sexual knowledge. The counterfeit-deviance hypothesis proposes that some sexual behavior by men with intellectual disability reflects limited knowledge or opportunity rather than deviant sexual interest, but tests of it have been mixed: in two cohorts, men with intellectual disability who had offended sexually had equal or greater sexual knowledge than men with intellectual disability who had not (Michie et al., 2006), and in another study differences disappeared when only men who had received sex education were compared (Lunsky et al., 2007). The hypothesis must be tested against the individual’s history and the offense, not assumed.
The connection can also cut the other way. A defendant with intellectual disability may plan, conceal, and select victims, and may have sexual interest in children, as other offenders do. Evidence of planning or concealment does not disprove the diagnosis, but it bears on whether the condition explains the offense. Because poor performance on intellectual and competency measures can be produced deliberately (Everington et al., 2007), and because standard validity tests can misclassify people with genuine intellectual disability (Salekin et al., 2010), the reliability of the diagnosis must be established before its relevance is argued.
Mitigation Evidence and Perceived Dangerousness
The “two-edged sword” that Atkins described has been examined empirically, mostly in hypothetical cases. In a national experiment with 181 state trial judges, expert testimony describing a biological mechanism for a convicted man’s psychopathy reduced the sentence from an average of 13.93 to 12.83 years and increased the proportion of judges who listed mitigating factors, although most judges listed aggravating factors as well (Aspinwall et al., 2012). In a study of 330 participants deciding a hypothetical sexual assault case, describing the defendant’s disorder as neurobiological rather than psychological produced shorter prison sentences but longer terms of involuntary hospitalization, which the authors reported as the first quantitative demonstration of the double-edged effect (Allen et al., 2019).
Archival data point in a similar direction. In 835 North Carolina capital cases that reached a penalty phase from 1990 to 2009, a jury’s acceptance of mental-disorder mitigators generally did not reduce the likelihood of a death sentence, with the exception of a learning disability diagnosis, and a jury’s rejection of mental-health mitigators was associated with substantially higher odds of a death sentence (Miley et al., 2020). In mock-trial experiments involving a juvenile defendant, jurors treated intellectual disability as mitigating and discounted a coerced confession by a juvenile with a disability (Najdowski et al., 2009).
Sexual offense cases may present the sharpest version of the problem. In an experiment with 259 mock jurors, a defendant diagnosed with a paraphilic disorder received harsher punishment than a defendant with no diagnosis or with obsessive-compulsive disorder, even though he was seen as having less control over his behavior (Moon & Nuñez, 2026). The study involved a paraphilia rather than intellectual disability, but it suggests that a diagnosis linked to sexual behavior may be received as aggravating.
Risk evidence itself may reflect the condition. In a New Jersey archival study of 3,066 people released from prison or civilly committed after sexual offenses, the 587 classified as having intellectual disability (a diagnosis or an IQ below 80, so including borderline functioning) did not differ from others in new charges or reincarceration over an average of 6.5 years, but they received higher scores on two actuarial tools and were more than twice as likely to be civilly committed (21.7% versus 9.6%; Callahan, 2022). The study is an unpublished master’s thesis, relies on records, and excluded committed individuals from recidivism analyses, so it should be cited cautiously. A peer-reviewed analysis of New Jersey releases, matching 459 men with intellectual disability or borderline intellectual functioning to 459 men without, (men released between 1996 and 2007 and followed for an average of 5.7 years) likewise found low rates of reincarceration for a sexual offense in both groups and found that the Static-99 and Static-99R did not significantly predict that outcome in the lower-ability group (Callahan et al., 2025). A 2026 meta-analysis of 15 samples (923 men) found that standard risk tools predicted recidivism in men with low intellectual ability with moderate accuracy, similar to that seen in other samples, and that predictive accuracy was higher when recidivism was measured by staff reports than by official records (Hanson et al., 2026). These findings point in more than one direction: findings on risk tools for men with intellectual disability are mixed and come mostly from small samples (moderate accuracy for existing tools in the meta-analysis; no significant prediction by the Static-99 and Static-99R in the largest U.S. sample), and an elevated score may partly reflect disability-related history, such as unemployment or never having lived with a partner, rather than the risk the item was designed to capture. These issues are addressed in “Risk Assessment Tools Outside Their Normative Samples.”
Treatment and Supervision as Mitigation
Because perceived dangerousness is a principal way disability evidence can turn against a defendant, mitigation that addresses risk speaks more directly to that concern than mitigation that addresses culpability alone; no study has tested which approach decision-makers find more persuasive in sexual offense sentencing. A plan that identifies adapted sex-offense-specific treatment, supervision suited to the person’s abilities, supported housing or employment, and the agencies responsible gives the court a way to protect the public while accounting for the impairment. Generic statements that a defendant would “benefit from treatment” give the court little to evaluate, and a plan is only as credible as the evidence that the named programs exist, will accept the person, and can manage the identified risks; a court may reasonably weigh whether the proposed structure is realistic for the offense and the history. An evaluation is most useful when it explains, in concrete terms, how the condition bears on the offense, on risk, and on what the defendant can realistically do under supervision, and when it acknowledges any evidence that cuts the other way.
Limits of the Evidence
The studies of how mitigation is received used hypothetical cases, student or online participants, or capital murder cases; few involved sexual offenses against children, and none tested intellectual disability evidence in a noncapital sexual offense sentencing. The New Jersey data come from a single state and era and from a thesis rather than a peer-reviewed article. Most defendants in these studies were men, and the research located did not examine whether disability evidence is received differently for women who sexually offend or whether decision-makers’ gender affects how such evidence is weighed. Studies of the counterfeit-deviance hypothesis are small and drew on men in intellectual disability services. Group findings cannot show whether a particular defendant’s condition contributed to his offense or whether he is more or less dangerous than others.
Questions to Ask About This Research
When the reduction from 13.93 to 12.83 years in Aspinwall et al. (2012) is cited, ask who was studied and what was tested: judges sentencing a hypothetical man with psychopathy, not a defendant with intellectual disability. When Allen et al. (2019) is cited to show that disability evidence increases punishment, ask what was counted: it measured both prison and hospitalization terms, and the condition was a described neurological or psychological disorder. When the civil commitment rates in Callahan (2022) are cited, ask whether the conclusion follows from the design: the study shows an association in one state’s records, and it grouped borderline intellectual functioning with intellectual disability. When the counterfeit-deviance hypothesis is offered, ask whether it has been tested in this defendant’s history. These questions are explained in “Reading Research Claims: A Guide for Attorneys” on this site.
Culpability, Mitigation, and the Law
Federal and Military Standards
Federal courts impose a sentence sufficient, but not greater than necessary, to serve the purposes in 18 U.S.C. § 3553(a), which requires consideration of the history and characteristics of the defendant along with the seriousness of the offense, deterrence, protection of the public, and the need for treatment. No limitation is placed on the information a court may consider about the defendant’s background and character (18 U.S.C. § 3661; USSG §1B1.4). Amendment 836 (effective Nov. 1, 2025) moved most departures, including the Chapter 5H specific-offender-characteristic policy statements, to Appendix B; the Commission described the change as outcome neutral, and the same facts may be considered for a variance under 18 U.S.C. § 3553(a). Arguments based on intellectual disability are therefore made as requests for a variance. Appellate courts have vacated sentences far below the range that relied heavily on a developmental condition without adequately addressing other sentencing purposes, such as the seriousness of the offense and general deterrence (United States v. Zuk, 2017; United States v. Crosby, 2024), so disability evidence is weighed alongside, not instead of, the harm caused.
In courts-martial, R.C.M. 1001 permits the defense to present matters in extenuation, which explain the circumstances of the offense, and in mitigation, which are introduced to lessen the punishment, and it allows the military judge to relax the rules of evidence for those matters. R.C.M. 1002 directs the court-martial to consider the history and characteristics of the accused, the impact of the offense on the victim and on the command, and the need to reflect seriousness, promote deterrence, protect others, and rehabilitate. For offenses committed after December 27, 2023, the military judge sentences within the applicable sentencing parameter unless specific facts warrant a sentence outside it, and must then state the factual basis in writing. The rules cited here come from the 2024 Manual for Courts-Martial; later amendments should be checked.
Arizona
A.R.S. § 13-701(E) requires the court to consider mitigating circumstances, including the defendant’s age, a capacity to appreciate the wrongfulness of the conduct or to conform it to the law that was significantly impaired but not so impaired as to constitute a defense, and any other factor relevant to the defendant’s character or background. Arizona generally does not permit expert evidence of a mental condition short of insanity to negate the mental state element of an offense (State v. Mott, 1997; Clark v. Arizona, 2006, left room for observation evidence); how those decisions apply to a particular case is a question for counsel and the court and should be verified before testimony, so evidence of intellectual disability is ordinarily presented at sentencing rather than on intent. Many sexual offenses against children are dangerous crimes against children under A.R.S. § 13-705, which prescribes enhanced sentencing ranges, generally bars release before the sentence is served, and generally requires consecutive sentences, limiting how far mitigation can move a sentence. Arizona’s constitution, criminal rules, and sentencing code allow release and sentencing decisions to take dangerousness into account (see A.R.S. § 13-705, above; Carroll, 2020, pp. 769–770, 787, a law-review article describing these provisions), which is where the double-edged effect described above is most likely to arise. Intellectual disability determinations in capital cases are governed separately by A.R.S. § 13-753.
Special Issues in Legal Practice
The principles described above apply with equal force to every party in a legal proceeding. The science operates the same way regardless of who is asking the questions, and the research summarized here does not by itself favor either side. What differs between the parties is not the science but the vantage point — the particular places where these issues tend to surface, and the particular errors each party is best positioned to guard against. The two summaries below identify those recurring issues. They are offered for orientation, not as litigation strategy; in every case the controlling question is what the evidence in that case will, and will not, support, and the expert’s role is to inform the trier of fact rather than to advocate for either side.
Special Issues for Prosecutors
The research supports asking whether the claimed condition was diagnosed by current clinical standards, with records predating the charges, adaptive-functioning information, and effort assessed, and whether the evaluation explains how the condition relates to this offense. Evidence of planning, grooming, concealment, or a pattern of conduct bears on whether the condition explains the behavior, and Atkins itself states that the deficits do not warrant exemption from criminal sanctions. Findings on risk tools for men with low intellectual ability are mixed and come mostly from small samples (a 2026 meta-analysis found moderate accuracy for existing tools, while the largest U.S. sample found no significant prediction by the Static-99 and Static-99R), and the harm to the person victimized, and the protection of other children, remain central sentencing purposes regardless of the defendant’s culpability.
The research also counsels caution. An elevated actuarial score in a man with intellectual disability may partly reflect disability-related history rather than the risk the items were designed to measure, and in one unpublished thesis of one state’s records such men were committed more often without reoffending more. Under Moore, improved behavior in a controlled setting such as jail is not a sound basis for discounting a diagnosis. In 835 North Carolina capital cases, a jury’s rejection of mental-health mitigators was associated with higher odds of a death sentence (Miley et al., 2020); no study has tested rejection of disability evidence in noncapital sexual-offense sentencing. Disputes are best resolved on the evidence about the diagnosis and its connection to the offense.
Special Issues for Defense
The research supports presenting a diagnosis established by current standards, with full intellectual and adaptive-functioning data, an explanation of how the condition relates to the conduct, and a concrete plan for adapted treatment and supervision. Because disability evidence can raise concern about dangerousness, a plan that addresses risk and supervision speaks to that concern; no study has tested which approach decision-makers find more persuasive. Defense counsel are often best positioned to gather records and informants and to identify adapted programs that will accept the client.
The research also counsels caution. Atkins states that the deficits do not warrant exemption from criminal sanctions, no study shows that intellectual disability causes sexual offending, and tests of the counterfeit-deviance hypothesis have been mixed, so evidence of limited knowledge or ability needs support from this defendant’s own history.
The research also counsels caution. Group findings do not show that this client’s condition caused his offense, and the counterfeit-deviance hypothesis has mixed support. A diagnosis based only on testing after arrest, or on an IQ score without adaptive-functioning data, invites challenge. Evidence of a sexual-behavior diagnosis may be received as aggravating, and arguments that rely on the diagnosis without addressing the victim’s harm and every sentencing purpose are vulnerable on appeal, as Zuk and Crosby illustrate.
This paper is for educational purposes only and is not legal advice. The author is a psychologist, not an attorney; summaries of law are general and must be verified against current authority or with a licensed attorney.
References
Allen, C. H., Vold, K., Felsen, G., Blumenthal-Barby, J. S., & Aharoni, E. (2019). Reconciling the opposing effects of neurobiological evidence on criminal sentencing judgments. PLoS ONE, 14(1), Article e0210584. https://doi.org/10.1371/journal.pone.0210584
American Psychiatric Association. (2022). Diagnostic and statistical manual of mental disorders (5th ed., text rev.). https://doi.org/10.1176/appi.books.9780890425787
Aspinwall, L. G., Brown, T. R., & Tabery, J. (2012). The double-edged sword: Does biomechanism increase or decrease judges’ sentencing of psychopaths? Science, 337(6096), 846–849. https://doi.org/10.1126/science.1219569
Callahan, P. A. (2022). Intellectual disability, risk, and recidivism in an American sample of incarcerated sexual offenders [Master’s thesis, John Jay College of Criminal Justice]. CUNY Academic Works. https://academicworks.cuny.edu/jj_etds/225
Callahan, P. A., Jeglic, E. L., & Calkins, C. (2025). Intellectual ability and sexual recidivism risk assessment: Comparing predictive accuracy in an incarcerated U.S. sample. Sexual Abuse, 37(2), 181–214. https://doi.org/10.1177/10790632241268471
Carroll, J. E. (2020). Safety, crisis, and criminal law. Arizona State Law Journal, 52(3), 769–809.
Everington, C., Notario-Smull, H., & Horton, M. L. (2007). Can defendants with mental retardation successfully fake their performance on a test of competence to stand trial? Behavioral Sciences & the Law, 25(4), 545–560. https://doi.org/10.1002/bsl.735
Hanson, R. K., Babchishin, K. M., May, K., Reid, B., & Wilson, R. J. (2026). Risk tools predict recidivism for men with low intellectual ability and a history of sexual offending. Sexual Abuse, 38(4), 386–419. https://doi.org/10.1177/10790632261429124
Lunsky, Y., Frijters, J., Griffiths, D. M., Watson, S. L., & Williston, S. (2007). Sexual knowledge and attitudes of men with intellectual disability who sexually offend. Journal of Intellectual & Developmental Disability, 32(2), 74–81. https://doi.org/10.1080/13668250701408004
Michie, A. M., Lindsay, W. R., Martin, V., & Grieve, A. (2006). A test of counterfeit deviance: A comparison of sexual knowledge in groups of sex offenders with intellectual disability and controls. Sexual Abuse, 18(3), 271–278. https://doi.org/10.1007/s11194-006-9019-y
Miley, L. N., Heiss-Moses, E., Cochran, J. K., Heide, K. M., Fogel, S. J., Smith, M. D., & Bejerregaard, B. J. (2020). An examination of the effects of mental disorders as mitigating factors on capital sentencing outcomes. Behavioral Sciences & the Law, 38(4), 381–405. https://doi.org/10.1002/bsl.2477
Moon, H., & Nuñez, N. (2026). Juror bias and mental illness: The double stigma of paraphilic disorders in sex crime cases. Psychiatry, Psychology and Law, 33(4), 1225–1242. https://doi.org/10.1080/13218719.2025.2521625
Najdowski, C. J., Bottoms, B. L., & Vargas, M. C. (2009). Jurors’ perceptions of juvenile defendants: The influence of intellectual disability, abuse history, and confession evidence. Behavioral Sciences & the Law, 27(3), 401–430. https://doi.org/10.1002/bsl.873
Salekin, K. L., Olley, J. G., & Hedge, K. A. (2010). Offenders with intellectual disability: Characteristics, prevalence, and issues in forensic assessment. Journal of Mental Health Research in Intellectual Disabilities, 3(2), 97–116. https://doi.org/10.1080/19315861003695769
Legal Authorities
Atkins v. Virginia, 536 U.S. 304 (2002).
Clark v. Arizona, 548 U.S. 735 (2006).
Graham v. Florida, 560 U.S. 48 (2010).
Hall v. Florida, 572 U.S. 701 (2014).
Moore v. Texas, 581 U.S. 1 (2017).
State v. Mott, 187 Ariz. 536, 931 P.2d 1046 (1997).
United States v. Crosby, 119 F.4th 1239 (10th Cir. 2024).
United States v. Zuk, 874 F.3d 398 (4th Cir. 2017).
U.S. Const. amend. VIII.
18 U.S.C. §§ 3553(a), 3661.
Ariz. Rev. Stat. Ann. §§ 13-701(E), 13-705, 13-753.
Rule for Courts-Martial 1001, 1002.
U.S. Sentencing Guidelines Manual § 1B1.4 & app. B (U.S. Sentencing Comm’n 2025).
U.S. Sentencing Comm’n, Amendment 836 (effective Nov. 1, 2025) (deleting §§ 5H1.1–5H1.12 and 5K2.0–5K2.24).
Related topic papers: Evaluating Defendants With Developmental and Cognitive Conditions: Standards and Methods · Risk Assessment Tools Outside Their Normative Samples · Competency, Miranda Waiver, and Accommodations · Adolescent Brain Development and Culpability · Young Adults Aged 18 to 25
Related topics from Arizona Forensic Services: Psychosexual Risk Evaluations · Courts-Martial