We have your argument this morning in Case 12, 1088-2, Hall v. Florida. Mr. Waxman. Mr. Chief Justice, and may it please the Court. In Atkins v. Virginia, this Court held that the Constitution bars executing persons with mental retardation, that is, persons with significantly sub-average intellectual function concurrent with deficits in adaptive behavior with an onset before the age of 18.
Because of the standard error of measurement that's inherent in IQ tests, it is universally accepted that persons with obtained scores of 71 to 75 can and often do have mental retardation when those three prongs are met. The Statistical Error of Measurement, or SEM, is a— Mr. Waxman, a line has to be drawn somewhere. And we did say in Atkins that we would leave it up to the states to determine the standards for this issue. So what's the rule we announce today? We tell them 70 is not okay, but 75 would be? I'm not quite sure. How would you announce the rule? Let me first take some issue with all due respect with your characterization of Atkins. What this Court said in Atkins is not that we leave it to the states to establish the standards for the clinical condition of mental retardation. What you said, quoting Ford, is we leave it to the states to
We leave to the states the task of developing appropriate ways to enforce the constitutional restriction that we announce. The rule that we advocate is, and the only real question presented in this case, is just this. If a state conditions the opportunity to demonstrate mental retardation on obtained IQ test scores, it cannot ignore the measurement error that is inherent in those scores. It is a feature, statistical feature of the test instrument itself. But we didn't base R.
decision in Atkins upon a study of what the American Psychiatric Association and other medical associations considered to be mental retardation, we based it on what was the general rule that states had adopted. And a large number of states had adopted 70 as the criterion. I mean, the criterion is, what do the American people think is the level of mental retardation that should make it impossible to impose the death penalty? We didn't look for the answer to that question to the APA or any of the other medical associations. We looked to what the states did. Now, what has changed in what the states do? Anything? Justice Scalia, I have — I would like to respond with four points, and I hope desperately I'll remember them. First of all, what this Court said was this Court —
Number one, it made clear, as it is reiterated in Miller v. Alabama and Graham v. Florida, that while a consensus or a perceived consensus among the states is important, the ultimate test is this Court's conclusion about what the Eighth Amendment does or doesn't allow. In making that determination at page 318 of this Court's opinion in Atkins, this Court, after reciting in footnote 3 the virtually identical clinical definitions, of mental retardation, and in footnote 5, pointing out that 70 to 75 is the established cutoff for mental retardation, this court said the following, quote, clinical definitions of mental retardation require and it recited the three tests because of their impairments mentally retarded persons by definition that is by the clinical definition have diminished capacities to understand and it recited all the other disabilities that made the imposition of the death penalty for persons with that excuse me with that clinical condition unconstitutional now as to what the states did
The court did refer to, I believe, 18 state statutes. Not a single one of those state statutes and not a single decision of the highest court of any state or any court in any state applied 70 or two standard deviations from the mean without reference to the SEM. The only statute that addressed it in 2002, when this court decided Atkins, was Arizona, which expressly provided that the SEM must be taken into account in evaluating and obtained IQ test. The SIM being what? And established by whom? The standard error of measurement, which is established by the creators of the test. It is not something that clinicians dream up. It's not something that is decided by the AAIDD or the American Psychiatric Association. It is inherent in the test, and all clinicians are told, both professional associations make clear, because it is simply a statistical fact.
it must be taken into account such that an obtained IQ test score is actually the result of an obtained IQ test score is a test band that accounts for the standard. For what purpose do they establish these scores? Is it for the purpose of determining who is so incapable of controlling his actions that he shouldn't be subject to the death penalty? Is that what they're looking for when they establish 70 to 75? What are they looking for? Well, what they are looking for, I mean, intelligence tests supply a ...
I mean, they weren't created for the definition of – clinical definition of mental retardation. They were created as – in order to determine a proxy for true intellectual function and, therefore, a true IQ test score. I mean, the general clinical – I'm not talking about IQ tests in general. I'm talking about why do they pick – they used to pick 70. Now they pick between 70 and 75 as the upper limit. Upper limit for what? I assume it is for people who would profit from medical treatment. Isn't that it? There are many reasons why –
A person's IQ, that is, a person's intellectual functioning, may be important for a whole variety of reasons, medical, psychological, developmental, and as a component of the clinical condition of mental retardation, the Eighth Amendment. Now, what intellectual- Mr. Watson, could we just clarify one thing, that what you refer to as the SEM, that is not limited to IQ of 70, 75. That's across the board. I mean, the concept, the statistical concept of a standard error of measurement applies to all forms of testing. So it has nothing to do with the death penalty?
No, I mean, I'm sure that, you know, when Archimedes announced his principle based on experimental – his experimental observations, he also recognized essentially the standard – May I come back to a question? May I come back to something similar to what Justice Sotomayor started out with? In your view, does the Constitution permit a state to establish any hard cutoff, let's say 76? Can it do that? I think it can because that falls — because the standard definition of prong one, that is intellectual functioning, is two or more standard deviations below the mean. All right. If it can do that — oh, I'm sorry. Oh, I'm sorry. Let me just explain. And because — if a state is using an obtained IQ test score as a proxy for true intellectual function —
It has to take into account the standard error of measurement, and therefore states like Mississippi and Oklahoma that in fact establish a cutoff of 75, in our view, is constitutional as this Court announced the class of individuals in Atkins. So that's just saying what — When you say the standard error of measurement, you're talking about a degree of confidence, right? Correct. And your submission is that you need to have a 95 percent degree of confidence. That's what the five gives you, or the number's wrong? Well, on a test that is normed at 100, 70 is two standard deviations below the mean. If there is a standard error of measurement, and this is not my submission, this is the universal. I'm just trying to figure out what it means. What it means is that someone, for example, with an obtained IQ of test score of 71, as Mr. Hall received.
has a 95 percent probability that his score will be between 76. So why is 95 percent? Where does that come from? Why are you picking 95 percent? Why isn't it 90 percent?
I'm not doing any picking. Why did the organizations pick 95 percent? It's been two SEMs, which is 95 percent, for decades and decades, and this Court recognized that consensus, that universal consensus in footnote 5 in its opinion. Which party has the burden of persuasion on the issue of IQ, and what is the standard? So it varies from state to state. What does the Eighth Amendment require? Does the Eighth Amendment permit a state to assign to the defendant the burden of persuasion on IQ?
IQ, IQ above 75, can they assign that burden of persuasion, or above 70? Can they assign that to the defendant? And if they can, what is the standard of proof that the defendant has to meet? So the short answer is, I believe, well, I will come to a yes so that you see where I'm going, but we believe that it is entirely constitutional for the state to assign the burden of proving mental retardation on the defendant. And insofar as the clinical definition recognized by this court in Atkins is a three-part conjunctive test.
I think it's fair to say that a logical consequence of that is that as to every component, the burden may constitutionally be placed on the defendant. Now, the burden with respect to prong one is the burden of proving significantly sub-average intellectual functioning, of which a true IQ score is a probabilistic piece of evidence. I don't think— Why can't the state—you told me that a state can establish a hard cutoff, and you told me that a state can assign the burden to the defendant. Now, in the case of someone who scores 75, is it not the case that there's roughly— there's no more than a 2.5 percent chance that that person's real IQ is 70?
So how does that square with any burden of proof that might be – any standard of proof that might be assigned on that point? That's what I don't understand about your argument. I think – let me see if I can explain this. First of all, we're talking – I mean, this is a man who has a 71. No, I understand. I'm talking about the general issue. As to the general issue, let me say it this way. The whole idea behind measurement error is that you can't make a valid judgment that somebody doesn't have a true score of 70 or below if the obtained score is within the measurement error. And even more fundamental than that –
The – your question suggests, and the state's suggestion suggests, that diagnosing mental retardation, which is the constitutional inquiry, is just a probabilistic inquiry into a person's, quote, true IQ score. But true IQ scores themselves are a statistical concept. It's a score that you would get on a hypothetical test that had no measurement error. But true – and this is another point. True IQ is not the same as intellectual function, and IQ tests themselves, however perfect they may be, don't perfectly capture a person's intellectual function, which is why— I understand that argument, but that doesn't seem to be consistent with your point that a state can establish a hard cutoff, 76. That's the end of—you get 76 on an IQ test. That's the end of the inquiry. The person does not—does not qualify under Atkins. So I—what I'm—this would not be a standard I would endorse, but I believe that in light of the consensus test that all professional organizations apply that was recognized in Atkins,
a score that is above the standard error of measurement of two standard deviations above the mean would be okay. But the point, the converse point, it seems to me, is not true, which is we know for a fact. that many, many people who obtain test scores of 71 to 75, in fact, have mental retardation. And if I just may point out that in this case, there were six experts who fully examined Mr. Hall or supervised a full examination of Mr. Hall. They were cognizant of the IQ test scores that he had received. And each one of them opined without hesitation that he had mental retardation, functional mental retardation, significant —
He did not show adaptive behavior, and the district report said that that was so, because all of those experts that you referred to were speaking retrospectively. There was no evidence of what... The defendant's current condition was. That was, I think it's in the joint appendix. That is correct, Justice Ginsburg. Now, the state trial court ruled that it would not accept evidence as to prongs two and three, but it did allow Mr. Hall's lawyers to make a proffer pursuant to the state's agreement that there could be a proffer in some expeditious manner, and that's at joint appendix 158.
we one of the two grounds that we appealed to the florida supreme court on in addition to the hard cutoff at 70 was the fact that in fact an expeditious proffer did not in fact permit us to put on all of our evidence about prongs two and three and the florida supreme court and this is page 125 of the joint appendix said we don't need to consider that question because we uphold the rule in charity there was nothing that limited you to the retrospective proof the trial judge asked a simple question How did the defendant adapt in prison? In quotes, one expert is saying, well, I didn't test for that. I don't know why I didn't do it. And that same expert said that he had, in fact, done it in other cases.
You're correct. Part of the expeditious proffer, the expeditious proffer was limited to the testimony of two, I believe actually only one of the experts who examined him and did the adaptive testing function. And that expert did say that he didn't test in prison. Now, there is, again, a universal professional consensus that adaptive functioning is tested by adaptive functioning in the real world, not adaptive functioning that occurs after 35 years on death row. And, in fact, we also know to a clinical certainty that Because mental retardation is a condition that is both developmental and not transient, that is, there has to be an onset, demonstrated onset during the developmental period, but one doesn't emerge from the condition of mental retardation, unlike, for example, mental illness. If you talk about the condition of mental disability that's involved here, I want to go back to something you've said in response to Justice Scalia. The question was along the lines of what does it mean to have a disorder under the DSM? Obviously, one thing that means is that the scholars can talk about it, that they can all focus on the same subject.
Does it have any meaning other than that, that it is an objective index, an objective characterization, that certain people have certain mental conditions? Is that what it means? That's exactly what it means, Justice Kennedy. What it means is, as this Court recognized, it is a clinical condition, unlike, for example, insanity or competence. Is there any evidence that society, in general, gives substantial deference to the psychiatric profession in this respect? Are there any studies on that?
I'm actually not aware of anything that suggests that society doesn't look to professional evaluations to do this. And, in fact, if one looks only at Florida's system, Florida uses mental retardation as a determinant for things other than the death penalty. It uses the existence of the condition for educational remediation, vocational rehabilitation, and everything. And in those instances, as we point out in our brief, Florida does apply the standard error of measurement. We have later in the week an argument about economic theories, and it's a little different because in that case it's the Court's own jurisprudence, and we have not said, as we have in Atkins, that it's up to the State. But do you think we defer to psychiatrists and psychologists and psychiatrists any more or any less than we do to economists?
Oh, I think it has to be much, much more, because as this Court has pointed out, this is a clinical condition. It's a condition that can only be appropriately diagnosed by professionals. They changed their mind, counsel. This APA is the same organization that once said that homosexuality was a mental disability and now says it's perfectly normal. They changed their minds. And they changed their minds as to whether 70 or 75 is the new test for mental retardation. The latter is not true.
The standard two things that are not in dispute in this case. We're only here talking about prong one, which is significantly sub-average intellectual functioning and nothing else. And everyone agrees. All the states agree. They all agreed at the time Atkins was decided that the clinical condition is defined by three elements and that the first element, significantly sub-average intellectual functioning, is defined as a person whose intellectual function is two or more standard deviations below the mean intellectual functioning of contemporary society. Can I take you back to a question that the Chief Justice asked? Because the Chief Justice said, you know, where does this SEN come from? And it is the test maker's determination that this is the margin of error that gives you a 95 percent confidence. I guess the question here, or one question here, is why do we have a 95 percent, why do we need a 95 percent confidence level? And you could say it either way. You could say, gosh, we're putting somebody to death. We should have a 100 percent confidence level. Or you could say, as I take it, Justice Alito,
Justice Alito's point was, well, look, the burden of proof is on the defendant here anyway, so a 95 percent confidence level seems awfully high. We should, you know, ratchet it down to 80 percent. So why for this purpose do we have to go with the testmaker's determination that five is what gives you a 95 percent confidence level? So the fact that. Two SEMs gives you a 95 percent confidence level. It's just a statistical fact. I take your question to be, well, why does, you know, why do clinicians and professional associations use that? Well, no, that's not really my question. I understand why they might use it for a wide variety of purposes. The question is why does their determination that it's useful for a wide variety of purposes to have a 95 percent confidence level, why is the state stuck with that for this purpose? Because the whole, and this goes to the reason that they use it. The reason that they use it is because of the inherent
imprecision in testing in general, but in particular testing for the presence of something like relative intellectual functioning. There are so many, it is so common for people who for a variety of reasons obtain a 71 or 72, in fact, to have mental retardation. And because evidence of intellectual evaluation of intellectual function. involves clinically much more than a test score. I mean, look what happened in this case. All of the IQ tests that were administered, all of the Wechsler tests, were accompanied, because they fell within the standard error of measurement, they were accompanied by the administration of further intelligence testing for confirmatory tests. Is that what you want? That is, I go back to Justice Sotomayor's question. Start mucking around with 95 percent. It's all over the law. I mean, 95 percent is a classical measure by scientists of when they have confidence that the fact that the regression analysis seems to establish is, in fact, a fact. Yes, it is. That is in tort law.
That isn't whether jury trials are discriminating because they don't have black people on the jury. It's all over the law. So I assume that you're not asking us to muck around with that number, because I don't know what the consequences would be. And if you're not, here's how we reduce it. You give the same test six times, and now we've reduced it from 5 percent, if he's above 70 all the time, to maybe 1 one-hundredth of 1 percent. Is that what you want to have happen? Well, let me just ask to your latter point. Am I right? Am I right in what I said?
You are not right in some of the things you said. The last thing you said is not right, which is— No, let's go before the last thing. The last thing is important. I'm not saying it isn't important, but I want my thinking to the last thing is dependent on my being right on everything before the last thing. So am I right before the last thing? About how 95 percent— You are right that 95 percent is just a feature—is generally speaking a feature that is widely adopted as a confidence level, and it is particularly important here because the constitutional guarantee announced in Atkins is against the execution of persons with mental retardation. On Justice Breyer's last point before your time expires, because I do think this is important, is there not another way of proving reliability? I suppose—what about multiple—
test. Suppose someone is given 25 Wexler tests, and 24 times the person scores 76, and one time the person scores 72. What would you deal with? How would you deal with that in a state that has a hard cutoff? So this is the last point that I wanted to get to. And I think if you – the best thing I can – before my time runs out, I just want to point you to page 10, footnote 3 of our reply brief, which cites the Oxford Handbook of, I don't know, clinical diagnosis or something. And we've given you the pages. And on those pages, it explains why when you have a situation of somebody who takes more than one test –
The appropriate determinant is very much not the average. It is what's called the composite score. And the composite score is different and, in fact, for people below the mean, below the average, because you have to take into account the fact that regression towards the mean and also the fact that a person who takes two, three, or four tests, multiple tests, changes the bell curve of standard deviation. So the example that's given in the Oxford Handbook is very similar to this case. There were four tests. They averaged at 72. The composite score.
and there's a statistical explanation for how it's arrived. The composite score is 69, and the standard error of measurement is actually larger using a composite score. So that's why, as to Justice Breyer's last point, simply averaging obtained scores does not, in fact, give you a better handle. Because there are so few people who score significantly below the mean on multiple tests, what clinicians use is a statistical analysis that takes into account the different calculation of what a standard deviation below the mean is. That's not consistent with my understanding of it, but I don't claim that I have a deep understanding of it. But what would be your answer to my hypothetical, where there are multiple scores that are above the hard cutoff, but one that's below? And I'll ask the state the opposite question. What would you do there? Well, we know what Florida does. What does the Eighth Amendment require, in your view? Well, in our view, the Eighth Amendment requires that if a state chooses to use IQ test scores as a proxy for intellectual functioning.
rather than a full inquiry into intellectual functioning, it cannot refuse to employ the standard error of measurement that is inherent in the test. But if it were 76, you would not need to go on to adoptive behavior. Is that your view? Our view is that a state consistent with Atkins could say that if you have no obtained score on a valid, properly administered, up-to-date test that is below 76, you may constitutionally be precluded. I think Many clinicians would go ahead and do adaptive functioning and other intellectual testing, but our view is that states like Mississippi and Oklahoma that set 76 as the cutoff do in fact comply with Atkins. May I say the balance? Thank you, counsel.
General Windsor. Mr. Chief Justice, and may it please the Court. This Court should affirm the decision of the Florida Supreme Court because it represents a reasonable legislative judgment and one that's fully consistent with Atkins and the Eighth Amendment. I'd like to start by responding to your question, Justice Alito, about what do you do with multiple scores. And in fact, in this case, we're not talking about someone who had one or two IQ scores. When you look at the Weschler test, which is what the petitioner contends is the gold standard, he had test scores of 71, 72, 73, 74, and 80. And as we understand what the petitioner would have this Court do is to take some of those lower scores and simply subtract five points from them.
That's not consistent with the materials that he cited in the footnote in his brief. If you look at the example there, they do apply some statistical principles to a range of scores, but they do not simply take the lowest score and subtract five points from it. And the logic of that, I would submit, is fairly obvious. You couldn't have a situation where, take in this case, you have a low IQ on the Weichler of 71 and a high IQ on the Weichler of 80, and say at the same time that there's a 95 percent chance his score is between 75 and 85, and also a 95 percent chance that his score is between 66 and 76. So you want us to decide this case and establish the principle?
the very significant principle that where you have a criminal defendant condemned to death for murder, whose scores are 71, 72, 73, 74 and 80, that's okay. That's all you're trying to persuade us of? I mean, I'm not very happy having to go through this in all future cases where you have somebody who has 69, 73, 74, 75, and 81. I mean, don't you have some more general principle other than the particular scores in this case are good enough? Well, we certainly think the particular scores in this case are good enough, but we do. We have a broader principle, which is that when you're dealing with things like mental diagnosis or things in the medical field generally, that there is good reason for this Court to do as it has historically, which is to defer to reasonable legislative judges. Well, let me ask you this.
that the American Psychiatric Association and all other professional associations do use the SEM, suppose that. It seems to me what the State is saying here in declining to use that is that it declines to follow the standards that are set by the people that design and administer and interpret the tests. Well, I have two responses to that. One, if the constitutional rule, which we submit it's not, but if there were a constitutional rule that the Eighth Amendment required Florida to adopt all kinds of clinical criteria that the APA or the AAI ... This is not clinical. This is statistical criteria with the tests they're relying on. Well, there's two parts. I keep saying clinical, but the SEM is not a clinical judgment. It's a standard error of measurement. That's the test makers. Well, that's right. But Justice Kennedy's question, as I understood it, was how can Florida deviate from what the ...
DSM and what the AAIDD suggests are best practices. No, this has nothing to do with best practices. It has to do with what the test givers say is the right way to look at their tests. Well, the test measures publish the error measurement, but it's the DSM and the AAIDD that are suggesting how many deviations that you should have. No, no, they're not challenging the two standard deviations. They're saying if you're going to preclude functioning abilities and the other two factors of your test based on a score of a test that says it has an SEM of five.
then you have to use the FCM. It's very different. They're not saying you have to take that number and declare that person mentally, intellectually challenged. You just have to apply the other factors. Well, it's a three-prong test. So in any instance, you would have to demonstrate the existence of all three prongs. But with respect to the 95% interval, that is... Can I stop you there for a minute? Yes, sir. I thought that you don't have to go to, under your view, you don't have to go to the second and third standards if you, on the first, it's 70 or below.
I thought that adaptive behavior doesn't come into the picture, onset doesn't come into the picture, if the IQ is above 70. That's correct, Your Honor. There's a three-part test, and the medical community doesn't dispute that, and the petitioner doesn't dispute that. To achieve a diagnosis of mental retardation, you would have to demonstrate that you meet each of the three criteria. So what is wrong? There may be agreement among you on this. What Atkins says is there are three parts, as you say.
One part is significantly sub-average intellectual functioning. That's the first part. And so what you say is if it's above a 70 on an IQ test or a couple of them, that's the end of it. We don't go further. What they say is I want to tell the jury something or the judge if the judge is deciding it. Judge, I have an expert here.
Thank you. Expert. I want to tell you, Your Honor, that that number 70 is subject to error. It could be, and indeed the state can do the same thing if it's 68. The number 68 is subject to error. So if somebody measures 68, you could bring in the witness. And he would say, 5% of the time, it's within five points either way. I think that's all they want to do. Now, there could be other ways of going about it. Maybe you give the same test six times with different questions, and that may not eliminate, but it might reduce the possibility of error. Or there may be some other way to do it. You call in a psychiatrist, and he says, okay, or an expert.
72. But he's still, we have other ways. We have other ways, not just tests. Now, I think you would do the same thing if you wanted to on the downside, I guess. And that might lead people not to being executed, you see. And that's their position, though, I think. And they get to do it on the upside.
All right? What's wrong with that? It doesn't sound so terrible. And anyway, the Eighth Amendment, this is a way of enforcing the Eighth Amendment. This doesn't need to be, I don't think, an independent Eighth Amendment violation. But go ahead. That's the kind of question I would love to have some questions. Sure. Well, what's wrong with that is, substantially, if you raise the limit to 75, as Mr. Waxman suggested you could, that doubles — It doesn't raise the limit to 75. What it does is it says just what I said, and I don't want to repeat it. When it's there at 70, they call their expert, who informs the decision-maker just what I said. Now, that would take a little time, maybe 15 minutes, maybe a little longer.
But that's what they want to do, I think. And why not? I mean, what's so terrible about doing it? What's so terrible about doing it is you would end up increasing the proportion of people or the number of people who would be eligible for a mental retardation diagnosis. Only those who, in fact, are mentally retarded. No, Your Honor. Because they're not mentally retarded. There's no disagreement that 70 is the appropriate threshold here. So this is almost an evidentiary matter. It's a matter of what does it take to prove by clear and convincing evidence, which is a standard of proof that they have as a matter of Florida law, and it's a standard of proof they do not challenge in this case. And all Florida recognizes is that the best measure of your true IQ is your obtained IQ test score. And so for someone who... General, the ultimate determination here is whether somebody is mentally retarded. And the IQ test...
is just a part of that. It's a part of one prong of that ultimate determination. And what your cutoff does is it essentially says the inquiry has to stop there. And the question is, how is that at all consistent with anything we ever say when it comes to the death penalty? Because when we have this whole line of cases that says when it comes to meeting out the death penalty, we actually do individualized consideration, and we allow people to make their best case about why they're not eligible for the death penalty. And essentially what your cutoff does is it stops that in its tracks as to a person who may or may not even.
have a true IQ of over 70, and let alone it stops people in their tracks who may not be mentally, who may be mentally retarded. Well, first, with respect to the mitigation, this is, this Atkins hearing in Florida is completely separate from the mitigation phase. And so he does still have individualized decision-making with respect to whether to have a death sentence. And he's still had an opportunity. But he doesn't have it with respect to this critical question, right? We've said you cannot execute somebody who is mentally retarded. And he says, now, you are preventing me from showing you that you're mentally retarded because you have an IQ test, a part of one prong of a three-prong test. You have an IQ test that says that I'm not mentally retarded, but, you know, that IQ test may be wrong. It's not, it's given that you're not using a margin of error.
Well, with respect to the IQ test just being one part of the intellectual functioning prong, that is a very recent development. And one of the problems we have with the idea of constitutionalizing medical criteria is that it is changing. If you look at the DSM-IV, which was in existence at the time of Atkins, the DSM-V replaced it last year. They said that intellectual functioning, the prong, was defined. by IQ as measured on test scores. General Winston, we don't allow all factors to be considered, do we? Would the state have been able to refute his assertion of mental retardation by pointing to the fact that he is the one who seized the young woman, who pushed her into a car, who drove the car with his accomplice following in another car, and who killed her and killed a policeman, too, later, I guess? Could the state bring that in and say...
Somebody who is mentally retarded enough, so mentally retarded as not to be responsible and not to be subject to the death penalty, certainly could not have pulled all of this off. This is not a person who is that mentally retarded, significantly mentally retarded. Could the state show that? Well, the state's in refutation of his mental retardation evidence. On the adaptive functioning portion of the test, there's a three-pronged test, the intellectual functioning, which historically has been all about IQ until very recently. And then adaptive functioning talks about how people react in the ordinary world to difficult situations. And some of what you talked about may or may not be relevant to that. But further responding to the earlier question.
It's not that Florida is not allowing evidence that you meet prong one. It's that Florida is making a finding that you cannot satisfy prong one. And so that's why you don't go to the other. But it seems to me that to follow from Justice Kagan's question, and I think it's a very important question, that we've been talking about here about the inaccuracy to some extent of IQ scores. And your rule prevents us from getting a better understanding of whether that IQ score is accurate or not, because we cannot even reach the adaptive functioning prong. You prevent it.
At the outset. And incidentally, you don't prevent it if it's under 65, under 70. Well, it's a three-pronged test, so you'd have to satisfy all three. But with respect to your question about whether adaptive functioning evidence can affect the reading of the IQ, we submit that's not the case. That's why they're discrete inquiries. And so if you have multiple test boards or if you have one test... But in very close cases, doesn't it illuminate whether or not the IQ test is...
exactly as reported, or if it is subject to some decrease or increase depending on what the evidence of adaptive functioning shows. No, Your Honor. That would be the position of the modern DSM, but that's a radical departure from where it has been historically. Again, it used to define... The intellectual functioning prong is being determined exclusively. I'll read Atkins again, but I thought Atkins did refer to the adaptive function. Oh, no, make no mistake. There is an adaptive functioning inquiry. That's one of the three prongs. And so you have to prove intellectual functioning. You have to prove adaptive functioning. But that was even under DSM-IV, correct? Oh, yes, sir. That's been a part for decades. What is changing is the way the medical community looks at how to measure IQ or what to do with IQ. Mr. General, at the very least, you give somebody an IQ test, he scores a 71. Now, he might actually have an IQ of 71. Or we know from the way these standard margins of error work, he might have an IQ of 69.
And you won't let him go to the adaptive behavior prong of the test and show that, you know, and show that he can't function in society in the ways that Atkins seems to care about, as Justice Kennedy says, notwithstanding that this IQ score number might be accurate or might not be. Well, the adaptive functioning is a critical component, but even the guidelines, the DSM would agree, that no matter what your deficits are in adaptive functioning, you do not qualify for a mental retardation diagnosis without also showing substantial deficits in intellectual functioning. Since when, I know that there's... less emphasis now on the IQ test than there was before. But when the IQ test was used, did they always use it as a fixed number, or did they always include the SEM as informing the clinical judgment? Oh, the SEM has been part of the equation. Yes, I'm not disputing that. Since then, they have not changed. We're not disputing that. But again — That's been the same in all medical diagnoses. Well, I think that the application of the SEM has been a component of this for some time. We don't dispute that. And we do note that the emphasis on IQ is decreasing and that the medical community is now suggesting that you should rely less and less on IQ, and they've changed — They're not arguing for that. They're just arguing that —
we should stay where it's always been, which is using the SEM. Well, I think what they're arguing is that you should do this. You should apply the SEM in the same way the clinicians do because that's the way the clinicians do it. And if you go down that road, then it is very difficult to understand in a principled way where that would stop. Is it the case that those who use IQ tests always require a 95 percent confidence level and always must require a 95 percent confidence level? Let's suppose a school.
on the other end of the IQ scale, wants to identify gifted children. And they say a child is gifted if the child has an IQ of 130 or above. So they say, if you have an obtained score of 130, you're in. You're in the gifted child program, even though there is the same percentage that would be the case with respect to someone with an IQ of 70, that really the person is below 130. Would there be something wrong with their doing that? Are there places that do that?
Oh, certainly. That's up to the decision-maker who's relying on the IQ for whatever the purpose he or she is. There is an SEM that's published that's a part of the test, but the decision-maker who's relying on an IQ test score, to take your example about someone in a school, they can set that as high or low as they want to because they might want to be over-inclusive, they might want to be particularly restrictive. And that's one of the areas where what we're dealing with here in the Atkins context is fundamentally different because we have an adversarial process, at least with respect to contested cases. We have a burden of proof, a clear and convincing evidence burden of proof that's not shared in the clinical setting. And so there are a lot of reasons why it's very different to make a diagnosis in a clinical setting, particularly now.
where the emphasis in the medical community is on providing services or making services available to people, and where you don't have the same distance and have to be over-inclusive. General, could the state change its statute to say we're now using a threshold of 60? Well, the state certainly has substantial leeway. I think the answer to that is yes, although it would be more difficult to defend, because I think what you'd want to do is go back and look at the consensus that was a part of Atkins, the consensus that supported the decision in Atkins. But I think before making a decision on 60 as a threshold or some other number, you'd want to look at the whole picture. Well, I guess I don't understand it. You have to explain that to me a little bit. Because I thought that the 70 was very longstanding. Everybody has agreed that it's 70 for many, many decades, maybe forever. So how could the state – if the state – why could the state say no to that? What would you look at? Well, I think you'd look at, again, the special interest at issue in Atkins and the fact that the state may need to be more restrictive because of the malingering and incentives that inmates would have to score.
lower than they would ordinarily perform at, that you wouldn't have in a clinical setting or you wouldn't have necessarily in a school setting where people are always trying to perform. That's why you have the other two prongs. I'm sorry? That's why you have the other two prongs. Well, you have. You certainly have. And at every juncture, when you have a fixed cutoff, you have the ability to defeat the other two prongs, but you're stopping them on a test based on a test score that has a margin of error.
recognized by the designers of the test. Well, we're not stopping them from putting on – all we're stopping is the consideration of the other prongs when it's clear that the first prong can't be satisfied. So I think there's been, in the briefing, this idea that it necessarily has to be sequenced a certain way, and it doesn't. If someone came in and it were undisputed that he could not satisfy the adaptive functioning prong, for example, then you wouldn't necessarily have to look at IQ. So – No, please. Then I misunderstand the case. I thought the Florida Accord held.
in effect, my words, that the IQ was a threshold in order to make this inquiry. And if you had 70, over 70, you cannot make the showing. But please correct me if I'm wrong. No, that's correct. And what happened in this case was there was a motion eliminated by the state recognizing that the IQ scores that were issued here were all above 70. And so it was sort of an ordinary evidentiary motion, you know, if you had a different case where you had to prove causation and damages, if there was no evidence. So you cannot get, if you do not satisfy prong one, you do not get to prongs two or three, period.
That's right. By the same token, if you don't satisfy prong two, you wouldn't get to prong three and so on. So the evidentiary ruling was certainly a recognition — simply a recognition that you have to satisfy all three prongs, which, again, is a factor that — What happens right now, today, under the law of Florida, a similar case, and there's an IQ score of 71, and the — The prosecutor points out to the judge that that's higher than 70. And the defense lawyer says, Your Honor, I would like to bring in my test expert here who will explain to you that even though this test did show 71, there is some fairly small but significant probability of error, and it could, in fact, be as high as 76. And he would like to explain to you that that's the situation, and therefore can I have him testify. Does the judge have to let him testify or not?
If I understand the hypothetical correctly, you have one test score of 71, and so without an obtained test score of 70 or below, he would not die. So then this is the dispute in the case. They would like to present that expert. You would say no. That's right. That brings me back to my first question, which I won't repeat. And this man has been on death row for over 35 years, I take it. Yes, sir. 1978 was the act. He didn't raise mental retardation until 10 years after his first conviction, isn't that right? That's right, Your Honor. He raised it in the Hitchcock setting in the late 80s and then went back and had some of the same evidence that he's... It's gone on this long. 1978 is when he killed this woman. There have been a number of appeals in this case. There have been a number of issues raised, and there was a... But, yes, there is... General, the last 10 people Florida has executed has been an average of...
24.9 years on death row. Do you think that that is consistent with the purposes of the death penalty, and is it consistent with sound administration of a justice system? Well, I certainly think it's consistent with the Constitution, and I think that there are obvious – I'm sorry, I apologize. Is it consistent with the purposes that the death penalty is designed to serve, and is it consistent with an orderly administration of justice? It's consistent with the – Go ahead. It's consistent with the purposes of the death penalty, certainly. Windsor, maybe you should ask us that question, inasmuch as most of the delay has been because of rules that we have imposed. Well, let me just ask this. Of course, most of the delay is at the end of the event. In this case, it was five years before there was a hearing on the action question.
Has the Attorney General of Florida suggested to the legislature any measures, any provisions, any statutes to expedite the consideration of these cases? Your Honor, there was a statute enacted last session, last spring, that is – it's called the Timely Justice Act that addresses a number of issues that you raised, and it's presently being challenged in front of the Florida Supreme Court. But I'd like to talk about the – Can I just ask why you have this policy? I'm sorry? Why you have the policy? I mean, is it administrative convenience? Just tell me why you have the policy. Well, the people of Florida have decided that the death penalty is an appropriate punishment for the most horrific crimes, like the crime issue. No, no, no. Why you have the 70th threshold? Well, that's what I was getting at. And so Florida has an interest in ensuring that the people who evade execution because of mental retardation are people who are, in fact, mentally retarded. And if we apply the rule that the petitioner has suggested, it would double the number of people who are eligible for the exemption. And that's inconsistent with Florida's purposes of the death penalty. Well, that's just to say that it would...
I mean, double the number of people eligible, but some of them may be mentally retarded. I mean, presumably, we want accurate decision-making with respect to this question, don't we? Well, we do, certainly. And they're not mentally retarded if they don't have an IQ of 70 or below. And that's a position that petitioner doesn't... We're not mentally retarded if they don't have an IQ score of 70 below? I mean, you don't believe that yourself, right? This is a tool to decide whether somebody is mentally retarded. And it's a tool that functions in one prong of a three-prong test. It is the first prong. The IQ threshold is the first prong. So no matter what your adaptive deficits are, you must...
And again, here in this adversary setting, you must demonstrate by clear and convincing evidence that you have an IQ of 70 or below. And what we believe is that if you say, well, there's a 95% chance that my IQ is somewhere between, say, 68 and 78, that you have not satisfied that first crawling. And I'd like to talk about the 95% confidence interval, because it is not the case that you have, say, with a 72, a 95% chance that your IQ is 70 or below. In fact, it's a very small chance. What the confidence interval measures is that you have a 95% chance that your true IQ is within five points of the measured IQ. But it's not that you would have an equal chance of having a 66, a 67, a 68. It falls under the bell curve.
And so if you take the test over and over again, you're going to score near the peak of that bell curve most of the time, which is where your true IQ would be. And at the outer ends of that 95% threshold are very small likelihoods that that's your true IQ. And then with each additional test you take... the odds that's above 70, the odds would go down. And so it's simply not the case that you can say, well, he has a 72, so he has satisfied or even might have satisfied the first prong because as a statistical matter, as a factual matter, every Wechsler test he has taken that was admitted into evidence was over 70. He has a 71, a 72, 73, 74, and an 80. And so if you want to apply statistics to it, you'd have to look and say, well, what are the odds that with that group of testing that his true IQ...
is understand now is it possible certainly it's possible it's over 100 you know you can exceed beyond the 95 confidence interval uh and nobody disputes that um that the true iq is something that is incapable of being measured or incapable of and but the iq test is what the community has and it's the most objective of the three problems which is why we believe it's particularly important to focus on because it's the most objective uh test that we have how many states uh retain that practice with a rigid 70 You know, by our count there are eight states that have both a hard cutoff and a 70 or two standard deviations, which approximates to the same thing, that has been expressly recognized by the states. There are a number of other states that have statutes similar to Florida's but that have not been interpreted one way or the other that we may or may not. Of those eight, how many actually have a fixed cutoff and how many have a SEN? I thought it was only four.
That didn't have consideration. No, no, those eight, Your Honor, all have a fixed cutoff of 70 or two standard, two standard. But by judicial decision, they've considered, well, that's something that. In most of the instances, Your Honor, they have done what Florida has done, which is they've had a statute that then was interpreted by the, by the courts. So we. Exactly. That's what I'm saying. Only four have it interpreted without the SEM. I apologize. I thought only four had interpreted without using the SEM.
had interpreted their statutes without using the SEM? Only four, like Florida. No, we have eight. We have Alabama, Florida, Idaho, Kansas, Kentucky, North Carolina, and Virginia, and Maryland, which has repealed the death penalty, but that was their standard. We would ask respectfully that the court affirm the Florida Supreme Court. Thank you, General. Mr. Waxman, you have a minute remaining.
In State v. Cherry, which is the Florida Supreme Court decision that established this rule, that if your lowest score or your only score is 71, you're out. And that applies whether you take one test or multiple tests. Here I'm quoting from the Supreme Court's decision in Cherry that, quote, it is a universally accepted given, that is, that the SEM is a universally accepted given and as such should logically be considered in determining whether a defendant has mental retardation. What the Court said was, We have to read the plain meaning of the Florida statute, and the Florida statute says two standard deviations. The notion that the Florida legislature or the people – may I finish my sentence? The Florida legislature or the people of Florida have made a considered decision not to account for the SEM is baseless and is belied by the legislative report that accompanies the statute, which said 70 to 75. Thank you. Thank you, Mr. Waxman. Counsel, the case is submitted.