This page was last revised on July 12, 2021

American Association on Intellectual and Developmental Disabilities

The most current definition of an intellectual disability from the American Association on Intellectual and Developmental Disabilities  follows.  Readers are, however, referred to applicable state regulations operationizing the definition as well as the IDEA definition.   The AAIDD definition is not a regulatory or statutory definition although many states have designed their criteria to conform at least in part with the AAIDD definition.   (Added July 12, 2021)  Although the word “retardation” has been eliminated from federal statutory and state regulatory language, it may still be found in relevant criminal caselaw.   Rewording the classification to ID did not and does not change the substantive criteria for classification under the IDEA.  Section 504 eligibility decisions are subject to constraints imposed in state regulatory language implementing the IDEA and, where and if divergent from the definition below, may be safely disregarded by a 504 committee.  

 

The IDEA Definition from Section 300.8 (as amended by Rosa’s Law in 2010.) in the 2006 Part B Regulations is provided below.  The federal definition is broad in scope but, again, state educational agencies generally have operationized the criteria with more specific regulations

Intellectual disability means significantly subaverage general intellectual functioning, existing concurrently with deficits in adaptive behavior and manifested during the developmental period, that adversely affects a child’s educational performance.

For additional resources regarding intellectual and developmental disabilities, see our Sped Resources/Intellectual Disabilities.

For an overview of Intellectual Disability within the context of IDEA, see the Center for Parent Information and Resources web page.

The modifiers used in the definition above are undefined by the federal regulations, leaving it to the states to providing quantifying criteria.  For most states, those criteria can be found in their current state regulations (links to all 50 states are on our State Regulations page.)

And for the legal statutes bringing about an end to the use of “mental retardation” and introducing “intellectual disability into federal laws (including the IDEA and Section 504) see Rosa’s Law on our Spedlaw page.

Although the federal regulations give states the right to establish guidelines defining more subjective terms such as “significantly subaverage,” eligibility groups still a responsibility to thoughtfully and thoroughly consider a student’s test scores within the context of a comprehensive evaluation.  There are some excellent books on the subject; the random thoughts provided below are intended as an overview, not a definitive work on the topic.

John Willis’s Random Thoughts on ID

300.8 Child with a disability.

(a) General.

(1) Child with a disability means a child evaluated in accordance with §§300.304 through 300.311 as having an intellectual disability, a hearing impairment (including deafness), a speech or language impairment, a visual impairment (including blindness), a serious emotional disturbance (referred to in this part as “emotional disturbance”), an orthopedic impairment, autism, traumatic brain injury, an other health impairment, a specific learning disability, deaf-blindness, or multiple disabilities, and who, by reason thereof, needs special education and related services. 

(2)(i) Subject to paragraph (a)(2)(ii) of this section, if it is determined, through an appropriate evaluation under §§300.304 through 300.311, that a child has one of the disabilities identified in paragraph (a)(1) of this section, but only needs a related service and not special education, the child is not a child with a disability under this part. (ii) If, consistent with §300.39(a)(2), the related service required by the child is considered special education rather than a related service under State standards, the child would be determined to be a child with a disability under paragraph (a)(1) of this section.

(b) Children aged three through nine experiencing developmental delays. Child with a disability for children aged three through nine (or any subset of that age range, including ages three through five), may, subject to the conditions described in §300.111(b), include a child-

(1) Who is experiencing developmental delays, as defined by the State and as measured by appropriate diagnostic instruments and procedures, in one or more of the following areas: physical development, cognitive development, communication development, social or emotional development, or adaptive development; and

(2) Who, by reason thereof, needs special education and related services.

(c) Definitions of disability terms. The terms used in this definition of a child with a disability are defined as follows:. . .

(6) Intellectual disability means significantly subaverage general intellectual functioning, existing concurrently with deficits in adaptive behavior and manifested during the developmental period, that adversely affects a child’s educational performance.  [added emphasis is this writer’s] 

Please note that no IQ number is offered and, in fact, the definition does not even define “general intellectual functioning” as a score on an IQ test.  Certainly, if you were using a test and could demonstrate that some part of even the most appropriate test you could find still penalized the child for some other disability (e.g., visual, hearing, or motor impairment, etc.), you would be justified in not using that part of the test.

Note also that the definition calls for deficits in adaptive behavior.  It does NOT say adaptive behavior scores as low as the IQ scores.  Some misguided people think, for example, that you need an IQ of 70 or less (which you do not) AND adaptive behavior of 70 or less.  IQ and adaptive behavior are not perfectly correlated (if they were, you would not need to measure both), so the Flynn Effect means that if you measure one and then measure the other, the other will usually be closer to the mean (100).  If you require that the IQ be 70 or below (lowest 2% of the population) AND adaptive behavior also be 70 or below (a different lowest 2%), you would identify much less than 2% of the population.  There are fewer blonde, female psychologists than female psychologists or blond(e) psychologists.

Note that the definition does not say ALL areas of adaptive behavior.  The various domains are not perfectly correlated (otherwise you could assess just one randomly selected domain), so some will be higher.

The definition does say that the child’s educational performance must be adversely affected and that the child must need special education and related services (which really means special education or special education and related services, which is how they correctly worded it in P.L. 94-142).  It does NOT say that academic achievement cannot be higher than IQ.  It probably will be (Flynn Effect).  Compare achievement scores to predicted achievement from the IQ, not to the IQ.  Look at the base rate for the difference between achievement and predicted achievement.  It is normal and expected for the achievement to be closer than the IQ to 100.

Please see  Our Resources/Intellectual and Developmental Abilities  for How Can a Person’s Reading Score be Higher than Their IQ (a super simple explanation of predicting achievement based on IQ) and Kevin McGrew’s Forest Gump and IQ Expectations (Why IQ scores are poor predictors of individual achievement) as well as other relevant information.

IQ TESTS ARE NOT INTERCHANGEABLE

John Willis and Ron Dumont provided a brief PowerPoint showing hos the use of different tests of intelligence can substantially affect high stakes decisions regarding eligibility under IDEA and Section 504 or life and death decisions under Atkins.   Click on IQ TESTS ARE NOT INTERCHANGEABLE to review that 15 page PowerPoint.

Guy McBride’s Random Thoughts on ID Identification and the SSA

Even though we were not responsible for assessing children for SSI eligibility, helping families access additional financial resources available to parents whose children had an intellectual disability could strengthen the family — and stronger families could be better partners in educating their child.  Prior to 2010, bringing up the subject when parents had not sometimes was a dicey proposition.  In North Carolina, at least, we had never used the word “retardation” in the eligibility process . . . early on, those students were classified as mentally handicapped and then mentally disabled.  Parents who could take that label in stride were not always able to be accepting when the word retardation . . . the label then used by the SSA . . . was mentioned.  However, now that the term “intellectually disabled” is now universally accepted, that should no longer be a deterrent.

SSA criteria may however differ substantially from that adopted by a state.   In some cases, it is even possible that children whose scores are not congruent wit state criteria may still qualify using the SSA criteria.  EVEN IF A CHILD DOES NOT MEET STATE CRITERIA for identification under the IDEA, it is still possible if he or she meets other generally accepted criteria for disability identification, s/he could still qualify for services under Section 504. I’ve provided those criteria below along with a link to the current SSA ID criteria.

The following federal criteria illustrate John’s observation that the IDEA criteria don’t require low adaptive functioning in ALL areas when applied to high stakes decisions regarding eligibility for SSI from the SSA.

12.05 Intellectual disorder (see 12.00B4), satisfied by A or B:

1. Satisfied by 1, 2, and 3 (see 12.00H):
   1. Significantly subaverage general intellectual functioning evident in your cognitive inability to function at a level required to participate in standardized testing of intellectual functioning; and
   2. Significant deficits in adaptive functioning currently manifested by your dependence upon others for personal needs (for example, toileting, eating, dressing, or bathing); and
   3. The evidence about your current intellectual and adaptive functioning and about the history of your disorder demonstrates or supports the conclusion that the disorder began prior to your attainment of age 22.

OR

2. Satisfied by 1, 2, and 3 (see 12.00H):
   1. Significantly subaverage general intellectual functioning evidenced by a or b:
      1. A full scale (or comparable) IQ score of 70 or below on an individually administered standardized test of general intelligence; or
      2. A full scale (or comparable) IQ score of 71-75 accompanied by a verbal or performance IQ score (or comparable part score) of 70 or below on an individually administered standardized test of general intelligence; and
   2. Significant deficits in adaptive functioning currently manifested by extreme limitation of one, or marked limitation of two, of the following areas of mental functioning:
      1. Understand, remember, or apply information (see 12.00E1); or
      2. Interact with others (see 12.00E2); or
      3. Concentrate, persist, or maintain pace (see 12.00E3); or
      4. Adapt or manage oneself (see 12.00E4); and
   3. The evidence about your current intellectual and adaptive functioning and about the history of your disorder demonstrates or supports the conclusion that the disorder began prior to your attainment of age 22.

Other Considerations in Determining Eligibility (Last updated 1/1/2018)

Consistent with §300.304(b) and section 614(b)(2) of the Act, the evaluation of a child suspected of having a disability, including an SLD, must include a variety of assessment tools and strategies and cannot rely on any single procedure as the sole criterion for determining eligibility for special education and related services. This requirement applies to all children suspected of having a disability, including those suspected of having an SLD.  (Emphasis added)   2006 Part B Regulations, p. 46646  

Having established that state regulations regarding eligibility criteria under the IDEA are guidance only in determining eligibility by eligibility groups and of no binding impact with respect to 504 committees, it is important to keep in mind that there are other definitions of intellectual disabilities than those found in federal state IDEA regulations.  In  states where no hard cut scores have been established in the regulation, of course, those differences may be moot.

Reliability and validity of the measures used:  Sometimes service providers will administer a test like the KBIT-2 in order to get a brief measure of a child’s ability.   When a score from a test like the KBIT-2 is  considered, it is important to take account the test’s reliability and validity, as well as the test producer’s recommendations, before giving it the same weight in an eligiblity decision as was given to a series of tests with documented reliabllity and validity in the past.   In fact, the test authors of the KBIT-2 (mentioned 28 times in the Roland Atkins case below, specifically state that those test results should not be used in any high stakes decision making regarding eligibility.  Even when a screening test is highly correlated with a more reliable test like the Wechsler, individual scores can differ significantly.  For an illustrative example of how frequently they can differ see this article by Ron Dumont and C.L. Boyd How Well Does the KBIT Predict WISC III Results

Some schools may still be using the K-BIT2 as a screening instrument during the multidisciplinary team’s student suppport process.   When used as a tool to reduce the number of children ultimately identified, several problems arise.   First, the parents aren’t given their rights.  Second, they (and in some cases the regular education teachers who brought the child before the referral team) may be given the impression that the school has actually evaluated the child when it has not.  Third, the potential for a K-BIT score giving a false negative in an individual case is not insignificant.  (And if I were a parent, even if there was only a one percent chance that the screening test was “wrong,” I’d want my chlld given a comprehensive evaluation.)  Fourth, even if the scores on the screening test and the evaluation were “spot on,” and not completely congruent with state guidelines, eligibility groups always have the option of exercising their collective clinical judgment if the other test data suggests that a test score is providing a high (or low) estimate of academic ability.

Caveat:  A test’s reliability and validity are of course NOT determined by the length of time it takes to administer, the number of items per subtest, or the number of subtests used to calculate and the overall score. Generally, the longer a test, but better its internal reliability BUT the RIAS II, for example, usually only takes a few more minutes to administer than either the K-BIT or Slosson but yields highly reliable and quite respectable validity coefficients.

Practice Effect:  In general, practice effect is most likely to be significant when two tests are administered within six months of each other, but when a student has been given the same or similar tests, even when test items are different, during multiple reevaluations, the strategies learned from the first test administration may transfer to the second.  Additionally, if a child has been given the same test multiple times, even when the test administrations have been more than six months apart, there still may be residual practice effect transferring into higher scores.  Kevin McGrew provides a nice summary of anticipated practice effects within the same year in an article entitled IQ test Practice Effects.   (Using a different test, of course, will help to ameliorate practice effect . . . but due to regression effect, users would anticipate that a child scoring low on the first test would score higher on the second and visa versa.)

Parents (and teachers) may also prep their children on how to take an intelligence test using various and sundry on-line materials in order to get them into gifted programs.   While that may also invalidate a intelligence test score, it is not usually an issue when testing for ID.

Flynn Effect and ID cutoff scores:

The Flynn Effect has been defined as “The rise in each successive generation’s average IQ (Intelligence Quotient) test scores, named for the intelligence researcher James Flynn. Example: Due to the Flynn effect, IQ test scoring must occasionally be reworked in order to ‘reset’ the average score to 100.”

This becomes a critical factor when a child has been tested on an older test of intelligence, because intellectual disability is defined by an inter-peer comparison.   However (for example) if a student is tested on an older IQ test, and the cutoff is supposed to be set at the second or fourth percentile, that percentile score might actually be higher than the 69 or 72 suggested by the original test norms.  While school psychologists and eligibility groups have traditionally considered the SEM in determining eligibility, the AAIDD has recommended that both the SEM and Flynn Effect be routinely considered in high stakes decision making..  Taking that recommendation into account,  the Flynn Effect was mentioned in the Roland/Atkins decision below no less than 38 times.

As a general rule of thumb, it would appear defensible to multiply the number of years that have passed since the test was validated and multiply them by .30   subtracting that result from the student’s score (or adding that result to the cut score cutoff established by the state.)  Obviously that is not a requirement but it is always an option and one this writer recommend that at least be considered within the context of the student’s academic achievement, adaptive scores, and social history.   The AAIDD would consider it mandatory, however, if a student was evaluated using a test with outdated norms.

The Flynn effect implies that an individual will likely attain a higher IQ score on an earlier version of a test than on the current version. In fact, a test will overestimate an individual’s IQ score by an average of about 0.3 points per year between the year in which the test was normed and the year in which the test was administered. The ramifications of this effect are especially pertinent to the diagnosis of intellectual disability in high stakes decisions when an IQ cut point is used as a necessary part of the decision-making process.  The Flynn Analysis: A Meta-analysis by Lisa Trahan et. al 2014.

Historically, some have recommended using .33 points per year as a corrective factor, a practice that Kevin McGrew has compared to “slicing butter with a laser beam.”

“Also, I recommend the use of 0.3 points per year as the Flynn effect adjustment value. A value of .33 is frequently cited in the historical Flynn effect literature.  The most recent comprehensive Flynn effect meta-analyses report average values of .293 (Trahan et al., 2014; p. 18) and .28 (Pietschnig & Voracek, 2015, p. 285), which are more consistent with 0.3.  Also, as explained in footnote #22, human IQ scores reported to the second decimal (.00) suggest a degree of measurement precision for IQ scores that is not reality based.  0.3 is more realistic and is used in this declaration.  Using .33 would only produce very minor score differences after rounding.”  (Kevin McGrew)

Although application of the Flynn Effect has been most widely been reported in Atkins’ cases (see below), the effect has been reported in non Atkins literature as Dr. McGrew pointed out below:

The use of a Flynn effect correction in clinical settings is less of a contentious or legal issue given that psychologists in these settings typically have more leeway to interpret scores as ranges, invoke clinical judgment, and incorporate information regarding measurement error in interpretation of the scores when making a diagnosis. In contrast, certain high stakes settings (e.g., Atkins cases; eligibility for Social Security Disability benefits) may have strict point-specific “bright line” cut-scores where examiners, or the recipients of the scores (e.g., the courts), are more constrained in clinical interpretation. Thus, the Flynn effect adjustment has been much more visible and debated in literature and law dealing with Atkins cases.

As noted above, the professional and research literature has recognized the Flynn effect in the assessment and identification of exceptionalities in education (i.e., SLD, gifted and talented, ID), social security, and neuropsychological disorders since the late 1970’s and 1980’s. Dr. Alan Kaufman is arguably the leading scholar, researcher and expert regarding interpretation of the various childhood and adult Wechsler tests.. In his seminal book Intelligent Testing with the WISC-R, Kaufman (1979) stated, regarding score differences between the then out-of-date WISC (Wechsler, 1949) and the newer WISC-R (Wechsler1974), that “…the children in the WISC-R sample performed better than did their WISC counterparts of an earlier generation on the types of tasks included in a Wechsler battery. The end result is a set of WISC-R norms that are steeper than the WISC norms” (p. 128). In addition, “Psychologists and special educators should internalize the practical consequences of using the WISC-R or its predecessor and recognize that any current decision based in part on the 1949 WISC is derived from an instrument whose norms are obsolete and, for practical purposes, meaningless” (Kaufman, 1979; p. 128; italics in original). Kaufman further stated that “By being alert to the expected differences of nearly one-half a standard deviation between the WISC and WISC-R IQs, examiners can make a mental adjustment to the reported WISC IQs” (p. 129; italics added). The issue of the impact of obsolete norms on IQ test scores has also been noted in other school psychology assessment literature since the 1980’s (e.g., Cummings & Moscato, 1984a, 1984b; Kaufman & Kaufman, 1983; McGrew, 1986; Reynolds, 1990; Sattler, 1983; 1992; 2001;[4] Weiner & Kaufman, 1979). S th

The use of a Flynn effect correction in clinical settings is less of a contentious or legal issue given that psychologists in these settings typically have more leeway to interpret scores as ranges, invoke clinical judgment, and incorporate information regarding measurement error in interpretation of the scores when making a diagnosis. In contrast, certain high stakes settings (e.g., Atkins cases; eligibility for Social Security Disability benefits) may have strict point-specific “bright line” cut-scores where examiners, or the recipients of the scores (e.g., the courts), are more constrained in clinical interpretation. Thus, the Flynn effect adjustment has been much more visible and debated in literature and law dealing with Atkins cases

Kevin S. McGrew PhD

Educational Psychologist

Director

Institute for Applied Psychometrics (IAP)

www.themindhub..com

Purpose, Passion & Serendipity

APA Div. 16 Prof. Sponsor (https://tinyurl.com/yd7wftyu)

New Atkins case provided by Ron Dumont  (12/21/2017)
The United States of America v. Farad Roland, District Court (12./18/2017)
Key Words: Atkins Death Penalty Case
Decided for: The Defendant
Published:  Yes

Discussion:

This Atkins death penalty case should be of interest to school psychologists who either (1) live in states where the state education department’s criteria for determining ID eligibility are vaguely defined or (2) are potential witnesses in an adversarial proceeding regarding ID eligibility determinations or, for that matter, ANY eligibility determination.

The court in this decision carefully explains and references the critieria used in determining ID . . . and even discusses stereotypes that neither define nor constitute exclusionary factors for determining ID classification.   While some of the discussion would only apply in the adult world where a retrospective diagnosis is being sought, most of the discussion would be on point at any age level.

Four expert witnesses, including our own Listserv participant Kevin McGrew, testified  for Farad, whereas the government provided three expert witnesses.  Several of the witnesses had school psychology backgrounds, but all had impressive credentials and years of experience in the field.   However, these kinds of decisions are obviously not decided by a vote and, once qualified as an expert (as all were), the judge’s job is to weigh the credibility of competing witnesses.

The court found that all four of Roland’s experts to be credible.  I don’t know how to put this succinctly without seeming to be insulting, but the judge then found the testimony of two of the three government witnesses incredible, whereas the testimony of their third witness was simply at odds with established standards.

It should be noted, I suppose, that the government wasn’t seeking the death penalty because Farad had cognitive problems; he was accused of murdering five people “in aid of racketeering.”  Still, the Supreme Court has prohibited the death penalty for individuals who (in the language of the time) were mentally retarded.

This court’s conclusions were based on a consideration of the three factors (or prongs) for eligibility, but in doing so also took into account standard errors of measurement,  practice effect, Flynn effect, several IQ tests, adaptive  behavior evaluations, his social history, SSA eligibility reports, anecdotal reports of his behavior in various settings, and alleged evidence of malingering. The seeming simplicity of the three prong test below belies the complexity of the analysis that it requires.

“The “generally accepted, uncontroversial intellectual-disability diagnostic definition . . . identifies three core elements: (1) intellectual-functioning deficits (indicated by an IQ score approximately two standard deviations below the mean—i.e., a score of roughly 70—adjusted for the standard error of measurement); (2) adaptive deficits (the inability to learn basic skills and adjust behavior to changing circumstances); and (3) the onset of these deficits while still a minor.” Moore, 137 S. Ct. at 1045.11 Each of these three prongs must be met for a person to be positively diagnosed.”

Almost anticlimactically, the judge concluded “For the reasons above, the Court GRANTS Roland’s motion for a pretrial determination of ID.”

Some online reports of this decision may be found at the links below.  As might be expected, the defense attorney thought the decision was thorough and thoughtful, whereas the U.S. Attorney’s office declined comment.

NJ.com Accused N.J. gang leader ineligible for Death Penalty

USNEWS:  Judge Bar Death Penalty

Guy M. McBride, Ph.D.

The case

Although not particularly relevant to the practice of school psychology, the Adkins debate continues to reflect current best practices (and not so good practices.)  Wikipedia provides a fairly succinct, albeit not authoritative, summary.  However, the issue is far from settled.  (See Atkins Revisited below)

Around midnight on August 16, 1996, following a day spent together drinking alcohol and smoking marijuana, 18-year-old Daryl Atkins and his accomplice, William Jones, walked to a nearby convenience store where they abducted Eric Nesbitt, an airman from nearby Langley Air Force Base. Unsatisfied with the $60 they found in his wallet, Atkins drove Nesbitt in his own vehicle to a nearbyATM and forced him to withdraw a further $200. In spite of Nesbitt’s pleas, the two abductors then drove him to an isolated location, where he was shot eight times, killing him.

Footage of Atkins and Jones in the vehicle with Nesbitt were captured on the ATM’s CCTV camera, which was of the two men with Nesbitt in the middle and leaning across Jones to withdraw money, and further forensic evidence implicating the two were found in Nesbitt’s abandoned vehicle. The two suspects were quickly tracked down and arrested. In custody, each man claimed that the other had pulled the trigger. Atkins’s version of the events, however, was found to contain a number of inconsistencies. Doubts concerning Atkins’s testimony were strengthened when a cell-mate claimed that Atkins had confessed to him that he had shot Nesbitt. A deal of life imprisonment was negotiated with Jones in return for his full testimony against Atkins. The jury decided that Jones’s version of events was the more coherent and credible, and convicted Atkins of capital murder.

During the penalty phase of the trial, the defense presented Atkins’s school records and the results of an IQ test carried out byclinical psychologist Dr. Evan Nelson confirmed that he had an IQ of 59. On this basis they proposed that he was “mildly mentally retarded“. Atkins was nevertheless sentenced to death.

On appeal, the Supreme Court of Virginia affirmed the conviction but reversed the sentence after finding that an improper sentencing verdict form had been used. At retrial, the prosecution proved two aggravating factors under Virginia law—that Atkins posed a risk of “future dangerousness” based on a string of previous violent convictions, and that the offense was committed in a vile manner. The state’s witness, Dr. Stanton Samenow, countered the defense’s arguments that Atkins was mentally retarded, by stating that Atkins’s vocabulary, general knowledge and behavior suggested that he possessed at least average intelligence. As a result, Atkins’s death sentence was upheld. The Virginia Supreme Court subsequently affirmed the sentence based on a prior Supreme Court decision,Penry v. Lynaugh, 492 U.S. 302 (1989). Justice Cynthia D. Kinser authored the five-member majority. Justices Leroy Rountree Hassell, Sr. and Lawrence L. Koontz, Jr. each authored dissenting opinions and joined in each other’s dissent.

Due to what it perceived to be a shift in the judgments of state legislatures as to whether the mentally retarded are appropriate candidates for execution in the thirteen years since Penry was decided, the Supreme Court agreed to review Atkins’s death sentence. The Court heard oral arguments in the case on February 20, 2002.

The ruling

The Eighth Amendment to the United States Constitution forbids cruel and unusual punishments. In the ruling it was stated that, unlike other provisions of the Constitution, the Eighth Amendment should be interpreted in light of the “evolving standards of decency that mark the progress of a maturing society.” The best evidence on this score was determined to be the judgment of state legislatures. Accordingly, the Court had previously found that the death penalty was inappropriate for the crime of rape in Coker v. Georgia, 433 U.S. 584 (1977), or for those convicted of felony murder who neither themselves killed, attempted to kill, or intended to kill in Enmund v. Florida, 458 U.S. 782 (1982). The Court found that the Eighth Amendment forbids the imposition of the death penalty in these cases because “most of the legislatures that have recently addressed the matter” have rejected the death penalty for these offenders, and the Court will generally defer to the judgments of those bodies.

The Court then described how a national consensus that the mentally retarded should not be executed had emerged. In 1986, Georgia was the first state to outlaw the execution of the mentally retarded. Congress followed two years later, and the next year Maryland joined these two jurisdictions. Thus, when the Court confronted the issue in Penry in 1989, the Court could not say that a national consensus against executing the mentally retarded had emerged. Over the next twelve years, nineteen more states exempted the mentally retarded from capital punishment under their laws, bringing the total number of states to twenty-one, plus the federal government. While there are 50 states, 19 don’t allow the death penalty under any circumstance, making 21 out of 31 a clear majority of the death penalty states. In light of the “consistency of direction of change” toward a prohibition on the execution of the mentally retarded, and the relative rarity of such executions in states that still allow it, the Court proclaimed that a “national consensus has developed against it.” The Court, however, left it to individual states to make the difficult decision regarding what determines mental retardation.

Also, the “relationship between mental retardation and the penological purposes served by the death penalty” justifies a conclusion that executing the mentally retarded is cruel and unusual punishment that the Eighth Amendment should forbid. In other words, unless it can be shown that executing the mentally retarded promotes the goals of retributionand deterrence, doing so is nothing more than “purposeless and needless imposition of pain and suffering”, making the death penalty cruel and unusual in those cases. Being mentally retarded means that a person not only has substandard intellectual functioning but also significant limitations in adaptive skills such as communication, self-care, and self-direction. These deficiencies typically manifest before the age of eighteen. Although they can know the difference between right and wrong, these deficiencies mean they have a lesser ability to learn from experience, engage in logical reasoning, and understand the reactions of others. This means that inflicting the death penalty on one mentally retarded individual is less likely to deter other mentally retarded individuals from committing crimes. As for retribution, society’s interest in seeing that a criminal get his “just desserts” means that the death penalty must be confined to the “most serious” of murders, not simply the average murder. The goal of retribution is not served by imposing the death penalty on a group of people who have a significantly lesser capacity to understand why they are being executed.

Because the mentally retarded are not able to communicate with the same sophistication as the average offender, there is a greater likelihood that their deficiency in communicative ability will be interpreted by juries as a lack of remorse for their crimes. They typically make poor witnesses, being more prone to suggestion and willing to “confess” in order to placate or please their questioner. Thus, there is a greater risk that the jury may impose the death penalty despite the existence of evidence that suggests that a lesser penalty should be imposed. In light of the “evolving standards of decency” that the Eighth Amendment demands, the fact that the goals of retribution and deterrence are not served as well in the execution of the mentally retarded, and the heightened risk that the death penalty will be imposed erroneously, the Court concluded that the Eighth Amendment forbids the execution of the mentally retarded.

In dissent, Justices Antonin Scalia, Clarence Thomas and Chief Justice William Rehnquist argued that in spite of the increased number of states that had outlawed the execution of the mentally retarded, there was no clear national consensus, and even if one existed, the Eighth Amendment provided no basis for using such measures of opinion to determine what is “cruel and unusual”. Justice Antonin Scalia commented in his dissent that “seldom has an opinion of this court rested so obviously upon nothing but the personal views of its members.” The citing of an amicus brief from the European Union also drew criticism from Chief Justice Rehnquist, who denounced the “Court’s decision to place weight on foreign laws”.^[3]

Subsequent developments for Daryl Atkins

Although Atkins’s case and ruling may have saved other mentally handicapped inmates from the death penalty, a jury in Virginia decided in July 2005 that Atkins was intelligent enough to be executed on the basis that the constant contact he had with his lawyers provided intellectual stimulation and raised his IQ above 70, making him competent to be put to death under Virginia law. The prosecution had argued that his poor school performance was caused by his use of alcohol and drugs, and that his lower scores in earlier IQ tests were tainted. His execution date was set for December 2, 2005, but was later stayed.

In January 2008, however, Circuit Court Judge Prentis Smiley, who was revisiting the matter of whether Atkins was mentally handicapped, received allegations of prosecutorial misconduct. These allegations, if true, would have authorized a new trial for Atkins. After two days of testimony on the matter, Smiley determined that prosecutorial misconduct had occurred. At this juncture, Smiley could have vacated Atkins’s conviction and ordered a new trial. Instead, Smiley determined the evidence was overwhelming that Atkins had participated in a felony murder and commuted Atkins’s sentence to life in prison.^[4] Prosecutors sought writs of mandamus and prohibition in the Virginia Supreme Court on the matter, claiming Smiley had exceeded his judicial authority with his ruling. On June 4, 2009, the Virginia Supreme Court, in a 5-2 decision authored by Chief Justice Leroy R. Hassell, Sr., ruled that neither mandamus nor prohibition was available to overturn the court’s decision to commute the sentence. Justice Cynthia D. Kinser, joined by JusticeDonald W. Lemons, the Court’s two most conservative members, wrote a lengthy dissent that was highly critical of both the majority’s reasoning and the action of the circuit court in commuting the sentence.^[5][6]

Ultimately,  Atkins was not spared because of his mental retardation. Instead, Atkins received his reprieve because an attorney for the co-defendant in the case revealed that prosecutors had not disclosed the whole story of his client’s involvement in the crime.

Atkins Revisited

In response  to and in argument against the criteria established by the Texas Judiciary to determine if a defendant was intellectually disabled (mentally retarded) the APA filed an Amicus Brief with the United States Supreme Court (2016).

A number of other national organizations have also provided briefs, including AAIDD,  ACLU,  and the American Bar Association.  (Actually, links to all of the briefs may be found here.)

An abstract of an article by Cecil Reynolds and others arguing that consideration of Flynn Effect should be mandatory in death penalty cases may be accessed by clicking on the preceding link.