Intellectual Disability and Uncertainty in Hall v. Florida
I’ve been meaning to post about the Supreme Court’s decision last week in Hall v. Florida—the case in which the Court struck down as unconstitutional Florida’s law for determining whether an offender is intellectually disabled and thus cannot be executed. In its 2002 case of Atkins v. Virginia, the Court concluded that it is unconstitutional under the Eighth Amendment to execute a “mentally retarded” individual. (Thankfully, in Hall, the Court switched over to the term “intellectually disabled.” I’ll be using the terms interchangeably in this post.) In Atkins, the Court stated that it was leaving it up to individual legislatures to determine when a person is “mentally retarded”—in the Court’s words, it was “leav[ing] to the States the task of developing appropriate ways to enforce the constitutional restriction upon their execution of sentences.” Now, other states and the medical community generally agree with Florida that a defendant is intellectually disabled if he has (1) “significantly subaverage intellectual functioning,” (2) “deficits in adaptive functioning,” (3) and “onset of these deficits during the developmental period” (by age 18). The first prong—the one at issue in Hall—is ordinarily determined by a defendant’s IQ score. States have concluded that an IQ score that is 70 or lower meets the “significantly subaverage intellectual functioning” standard. The Hall case raises the issue as to whether uncertainty in obtained IQ scores (or confidence intervals) ought to be included in determining the defendant’s true IQ score for the purpose of this first prong of the intellectual disability test.
In a 6-3 decision, the Hall Court concluded that Florida’s approach—of finding that an obtained IQ score greater than 70 may be determinative of the fact that the defendant is not intellectually disabled—is unconstitutional. In reaching this conclusion, the Court took a detour from its ordinary Eighth Amendment analysis, focusing heavily on the opinions of professional organizations. As in prior opinions, the Court was loose with the numbers in the state-counting aspect of its Eighth Amendment analysis, concluding that a “significant majority of States” have adopted procedures contrary to Florida’s approach. The dissent explains that, of the death penalty states, nine have adopted an approach similar to Florida, nine have not addressed the issue, and twelve take the approach that the Court finds to be constitutionally required. It is difficult to find a national consensus in these numbers. In finding a consensus, though, the majority includes the eighteen states that have abolished capital punishment. Whether to include non-death-penalty states in this calculus is an issue that the Justices have debated before. But the Court’s approach to finding a consensus in this case is especially interesting because of the metric it uses in doing so. Instead of looking at the number of states that have categorically prohibited a punishment—such as tallying the number of states that have banned executing the “mentally retarded,” the “insane,” or juveniles—the Court is counting the number of states that take into account standard errors of measurement (SEMs) in determining whether a defendant is intellectually disabled. This metric accounts for the uncertainty inherent in obtained IQ scores and provides a range in which it’s likely the defendant’s true IQ score falls based upon his obtained score. In examining this metric, the Court frames the question as whether it is unconstitutional for a state to not take into account SEMs in determining whether a defendant is intellectually disabled. But is it really this procedural matter that’s at issue here? Or do we instead care about whether individuals who actually have true IQ scores of 70 or below are being executed? For example, if a state were to conclude that a defendant is intellectually disabled if he has an obtained IQ score of 90 or below, and if the test used in the state has a SEM of 2.5—suggesting that it is quite unlikely that a defendant scoring above 90 on an IQ test would have a true IQ score of 70 or below—would it be unconstitutional for that state’s courts not to take into account the SEM in determining whether a defendant is intellectually disabled? The dissent suggests that another way to probe the uncertainty is to admit multiple obtained IQ scores—a practice the Florida procedures in question allowed. While multiple obtained IQ scores are relevant to determining the reliability of the obtained scores, using this evidence, alone, means working with a fairly small sample size. In Hall, the defendant submitted nine obtained IQ scores, and two were excluded by the sentencing court.
The Court’s decision in this case continues to chip away at the death penalty, albeit quite slowly. The majority’s departure from its traditional Eighth Amendment framework for analysis—a step that is far from new for the Court—injects further uncertainty into the limits on punishments under the Constitution. The Court’s willingness to think more deeply about the methodologies, math, and science underlying some of its decisions, though, furthers the understanding that the meaning of the prohibition on cruel and unusual punishments is evolving. Unfortunately, uncertainty remains about how the Court gathers information about these complicated aspects of law and fact, and how adept the Court is at understanding and employing these concepts.
There is much more that could be said about the Hall case, the Eighth Amendment, and judges’ uses of science and technology, but it has come time for me to sign off of Concurring Opinions for now. Thanks again to the Co-Op gang for asking me to visit, and I look forward to the next time!