US CARNELL
NO. AP-77,038
IN THE COURT OF CRIMINAL APPEALS OF TEXAS
May 12, 2021
ON REHEARING ON THE COURT‘S OWN MOTION FROM CAUSE NO. 2012-2331-C1 IN THE 19TH DISTRICT COURT McLENNAN COUNTY
KELLER, P.J., filed a dissenting opinion in which Yeary, Keel and SLAUGHTER, JJ., joined.
At trial, the jury rejected Appellant‘s intellectual disability claim. In this direct appeal, the Court finds that decision to be manifestly unjust. I see instead a manifestly reasonable response to the evidence. The jury could have found against Appellant on the intellectual-disability issue for two separate reasons. First, there was evidence that IQ testing placed Appellant above the range of intellectual disability even when the standard error of measurement was taken into account. While Appellant presented conflicting evidence in the form of test scores that placed him within the range of intellectual disability, the jury could have rejected some of those tests because no evidence of their reliability was presented, and the jury could have believed that the remaining tests were unreliable because Appellant failed to exert adequate effort or malingered on them. Second, the jury could have believed that Appellant did not show deficits in adaptive functioning that would render him intellectually disabled. Most of the evidence of adaptive deficits came ultimately from witnesses with a vested interest in saving Appellant from the death penalty and much of what these witnesses said was contradicted by other evidence. And on top of that, the test results were questionable because there was evidence that Appellant faked adaptive deficits.
A. The burden of proof was on Appellant and factual sufficiency requires deference to the jury.
Under the Supreme Court‘s jurisprudence, the Eighth Amendment prohibits the execution of intellectually disabled individuals.1
B. The jury could have reasonably rejected Appellant‘s intellectual disability claim on the basis that IQ testing did not show the requisite deficits in general mental abilities.
1. The existence of low tests scores is not enough; the tests must be reliable.
The DSM-5 retains the traditional three-criteria approach to determining intellectual disability.6 The first of these criteria is a showing of deficits in general mental abilities.7 IQ testing is the typical method of assessing the first criterion, and a test result indicates the requisite deficits if it shows mental abilities that are approximately two standard deviations or more below the population mean. When the mean is measured as a score of 100, this translates into an IQ score of 70 or less, plus or minus the standard error of measurement for the test.8
The Court contends that Appellant‘s intellectual disability claim could not be rejected on the basis of IQ testing alone because only one of five tests yielded a score outside the range of intellectual disability. Actually, there were six tests, but that is a minor quibble since it is true that only one of the six yielded a score outside
2. Five of the six tests could have been reasonably rejected as unreliable or as not shown to be reliable.
“Little is known about the circumstances surrounding the IQ test administered in 1991 or the IQ test administered by TDCJ” in 1993.11 As to those two tests, then, because Appellant has the burden of proof, a jury could reject the scores because Appellant produced no evidence to show that the tests were reliable.
Regarding the 1992 tests, Dr. Coxe testified that he did not consider the result of the child version of the IQ test to be reliable because Appellant had failed to put forth sufficient effort on the test.12 In fact, Dr. Coxe‘s doubts about the reliability of that test spurred him to have Appellant take the adult version of the test, on which he scored 74.13
As for the last two tests, taken about twenty years later, Appellant had strong incentives to try to produce a low score. The 2012 test was done as part of a Social Security disability assessment. As the Court notes, one of Appellant‘s letters to the victim involved explaining to her how to get $1200 a month in social security benefits for having a mental disability.14 The 2013 test was conducted in the wake of capital murder proceedings.15 And in both tests, Appellant‘s effort level in taking the test was a serious concern. Dr. Correia, the psychologist who conducted the 2012 test, stated that Appellant “put forth an effort to cooperate minimally and superficially only.”16 In connection with 2013 test, Appellant tested positive on two tests designed to detect malingering and on one of the three subtests for malingering
3. Malingering is an especially big concern in Appellant‘s case.
Malingering is an especially big concern in Appellant‘s case because of evidence of his deception and manipulation. Dr. Mayfield, who conducted the 2013 test, acknowledged that Appellant had a documented history of deception and manipulation from an early age,19 and Dr. Coxe testified that he did not think that appellant was an honest individual.20 Dr. Mayfield believed that her testing was accurate, despite the implausibly low scores, but she also believed that most of capital murder defendants that she had tested had given her valid scores.21 This, in spite of the fact that she acknowledged on cross-examination that studies had demonstrated a combined rate of probable and definite malingering for pre-sentenced males referred for neuropsychological assessment to be as high as 54 percent.22
More critically, a fellow inmate testified that Appellant intentionally faked low intelligence or mental incapacity for the purpose of avoiding the death penalty. On original submission, this Court pointed to the testimony of Mubarak Ali, an inmate incarcerated with Appellant, who testified that Appellant had told him of an aunt in Louisiana who was a bail bondsman and could help Ali bond out of jail.23 When Ali asked Appellant to contact his aunt on Ali‘s behalf, Appellant replied that he could not use the phone because he was “telling the Court that he‘s mentally sick.”24 Ali also testified, “Mr. Petetan told me, he say if he act like he‘s mentally sick, he cannot get the death penalty.”25 At Ali‘s prodding, Appellant called his mother and told her about Ali‘s bond situation.26 Ali bonded out of jail the next day.27 Malingering was not just a theoretical possibility in Appellant‘s case, nor was it something that could only be inferred from some test results. The State had evidence that Appellant knew that a deficiency in mental abilities could spare him from the death penalty and that he was actively trying to manipulate the courts into thinking that he was mentally deficient.
Appellant‘s deviousness was also apparent from the capital murder offense he committed. On original submission, we observed that he engaged in a relatively sophisticated scheme to murder Kimberly and frame Miller for the murder.28 Appellant used a stranger to help him scout out Kimberly‘s apartment, and he tricked a neighbor into knocking on Kimberly‘s door so that she would open it without knowing that Appellant was there.29 Appellant schemed to frame Miller for the murder by using the false promise of drugs to lure
Also, the jury could have reasonably inferred from the testimony of psychological experts32 and education personnel33 that appellant was generally uninterested in exerting full effort on testing and that he became easily discouraged when questions became somewhat difficult. As a result, the jury could have reasonably believed that, even when he had no specific motive to lower his score, his test scores understated his intelligence.
The Court says that evidence of malingering on some tests is not alone sufficient to reject his intellectual disability claim because “sources of imprecision in administering the test to a particular individual . . . cannot narrow the test-specific standard-error range.” But malingering is not a “source of imprecision in administering the test.” It goes to the heart of an IQ test‘s reliability.34 And it has nothing to do with the standard-error range. The standard error of measurement assumes that someone diligently tries to score well, not that he fails to try or, worse, purposely tries to skew the results of the test in an effort to appear intellectually disabled.35 It should come as no surprise that Appellant‘s lowest scores are on the last round of testing, when his motivation to appear to be intellectually disabled is at its highest.
4. The only reliable test yielded a score above the intellectual disability range.
Dr. Coxe testified that the 1992 adult test better reflected Appellant‘s intelligence because of the effort he put forth on that test. That test yielded an IQ score of 74, with a standard error of measurement of only three points. Even accounting for the standard error of measurement, then, that IQ test placed Appellant above the intellectual-disability range.36 And the jury could have believed that test was the only test shown to be reliable.
5. The IQ scores are not consistent over different tests.
The Court says that, unlike in many Atkins cases, where the IQ scores start out high and get lower once the incentive to score low arises, “Appellant‘s scores started out low and stayed there.” But the distribution of test scores shows otherwise.
6. Consequently, the jury‘s finding was not factually insufficient.
The upshot of all this is that the jury could have reasonably believed that only one of the six tests had been shown to be reliable, and that was the 1992 adult test, with a score of 74, which, with a 3-point margin of error, placed Appellant above the range of intellectual disability, even taking standard error of measurement into account. Given Appellant‘s general tendency to not exert full effort on tests, even this score could be an underestimate of his intelligence. In any event, a finding that Appellant has failed to prove the requisite deficits in general mental abilities is eminently reasonable and, contrary to the Court‘s conclusion, is not so against the great weight and preponderance of the evidence as to be manifestly unjust, shocking to the conscience, or clearly biased.
C. The jury could have reasonably rejected Appellant‘s intellectual disability claim on the basis that significant adaptive deficits had not been shown.
1. The only test for adaptive deficits has serious flaws that would make it reasonable to reject the test entirely.
The Court points out that the test Dr. Craig used to assess adaptive functioning was not normed for retrospective assessment, included items that were inappropriate for someone between ages fourteen and sixteen, and had not been normed for individuals who had been incarcerated for 20 years of their life.37 The Court also points out that the test used Appellant and his close relatives (mother, brother, sister, and uncle) as informants and that these people all had an obvious vested interest in the outcome of their interviews, namely, saving Appellant from the death penalty.38 And the Court notes other methodological flaws: interviewing each informant only once and interviewing most informants over the telephone rather than in person.39 On original submission, this Court also pointed out that the normed age range for the test—thirty to thirty-nine years—“far exceeded the age or ages for which appellant was being assessed—ages fourteen to sixteen”40 and that the test was normed after a “major social and technological change” from the time period being reported—namely, “the rapid expansion and widespread use of the internet.”41 In light of these many flaws, it would be entirely reasonable for a jury to believe that this test‘s results were worthless.
The Court recites a great deal of testimony from Appellant‘s relatives on various adaptive deficits. But Appellant‘s relatives had an obvious incentive to say what was needed to spare Appellant from the death penalty. That incentive was the strongest for the close relatives, and they rated Appellant‘s adaptive behavior the worst.
Also, the Court points to a plethora of other evidence, much of which was entirely objective, that contradicts the assertions of these family members, undermining their reliability as honest or accurate in their assessments. For example, Appellant‘s mother and sister said he could not count or manage money, but there was plenty of evidence to the contrary. Appellant testified to and was seen paying for items on his own.42 He did correct math calculations in a letter to the victim, and other letters to her showed that he knew the value of various items and even understood the concept of a sliding scale.43 His uncle said that Appellant did not understand what a DBA (“doing business as“) form was, but another inmate testified that Appellant explained to him how to file one.44 Appellant also explained to this inmate the concept of royalties, saying that an artist receives only a percentage of money from each sale.45
The record included numerous letters to the victim. Appellant‘s sister said she thought someone must have written the letters on Appellant‘s behalf because Appellant‘s own handwriting was comparable to that of a preschooler or kindergartner. The prosecutor was so confident that any one of these letters was far more sophisticated than what the sister described that he told her, “I‘m going to do the Carnell Petetan letter grab bag. Go ahead and pick a letter, any letter.”46 The sister was forced to acknowledge that the letter she chose, at random, was clearer and neater than her description of his writing.47 And it is hard to imagine that kindergarten-grade letters from Appellant could have induced the victim to fall in love with him. A business plan entitled, “Cartel Boxing Promotions Business Plan Prepared By Petetan,” included a well-written biographical section,48 and authorities also found a forty-page document purporting to be Appellant‘s autobiography.49 Appellant denied writing the business plan or the autobiography, but they both purport to be written by him, and a jury could have believed that indeed they were.50
Appellant‘s mother and sister said that he could not use a cell phone, but Appellant testified to using a cell phone, and the victim‘s daughter testified that Appellant and the victim both had cell phones in his
The jury could have reasonably disbelieved all of the testimony of the family members.
3. Little other evidence of adaptive deficits remained. A jury could have reasonably believed that education records did not persuasively establish adaptive deficits given the testimony of education personnel.
The main other evidence of possible adaptive deficits was the record of Appellant‘s bad grades in school and his poor performance on California Achievement tests, but testimony from education personnel indicated that his problem was attitude rather than intelligence.54
4. The jury had evidence that Appellant intentionally faked adaptive deficits.
As I explained earlier, a fellow inmate, Mubarak Ali, testified about Appellant actively trying to fool the courts into thinking that he was “mentally sick” so that “he cannot get the death penalty.” Appellant had at first declined to use the telephone to call his bail-bondsman relative on Ali‘s behalf because of this scheme to fool the courts. So the jury had evidence that Appellant intentionally faked adaptive deficits in an effort to avoid the death penalty. This evidence undermines the reliability of any testing in connection with capital murder proceedings.
5. The Court‘s analysis conflates legal and factual sufficiency.
The Court contends that the evidence was legally sufficient for a jury to reject the existence of adaptive deficits because a rational jury could reject such deficits “in isolation,” but that the evidence was factually insufficient because failing to consider evidence of adaptive deficits in conjunction with evidence of deficits in general mental abilities runs afoul of the Supreme Court‘s decision in Hall. The Court also contends that the evidence is factually insufficient because emphasizing adaptive strengths to undermine expert testimony runs afoul of the Supreme Court‘s pronouncement in the Moore cases. There are several problems with this line of reasoning.
First, although couched as a conclusion that the evidence is factually insufficient, this reasoning, if based on correct premises, would compel a conclusion that the evidence is not just factually insufficient, but legally insufficient. If, as the Court says, the jury‘s rejection of Appellant‘s intellectual disability claim violates Hall, then as a federal constitutional matter, Appellant has established his claim as a
The same conclusion can be drawn if the only way for a jury to reject Appellant‘s claim is to improperly overemphasize Appellant‘s adaptive strengths. If the jury‘s rejection of Appellant‘s intellectual disability claim violates Moore, then as a federal constitutional matter, he has established his claim as a matter of law, and he is entitled to a reformation of his sentence instead of a new punishment hearing.
6. The Court incorrectly concludes that the jury must have engaged in an improper analysis.
But contrary to its reasoning, the Court now holds that the evidence is legally sufficient to support the jury‘s rejection of adaptive deficits. And that leads to the second problem with the Court‘s line of reasoning: the jury did not have to violate Hall and Moore to reject Appellant‘s claim. Even if the jury thought that some IQ testing suggested the requisite deficits in general mental abilities, and considered that testing in conjunction with evidence of adaptive deficits, the jury could have still reasonably found that adaptive deficits were not shown. The jury could have decided that the adaptive-deficits testing was invalid, that the testimony of family members was untrustworthy, and that Appellant‘s poor performance in the school setting (on both grades and testing) reflected obstinacy, willful behavior, and a lack of discipline rather than adaptive deficits. Nor was the jury required to rely on adaptive strengths to reject Applicant‘s intellectual disability claim. His adaptive-deficits claim relied on flawed testing and biased witnesses, and there was evidence (his conversation with Ali) indicating that he was faking adaptive deficits to avoid the death penalty.
7. The Court‘s assessment of the evidence is flawed.
The real question on adaptive deficits is whether the evidence of adaptive deficits is so overwhelming that the jury‘s finding to the contrary, even if minimally rational, was unjust. The Court essentially answers that question “yes” when it says, “Expert after expert diagnosed Appellant with mild intellectual disability,” “the medical community was of one mind,” and he “consistently scored within the range for intellectual disability on intelligence testing administered across decades.” None of those statements are entirely accurate, but even if they were, the jury would have had good reason to discount both the expert assessments and the tests.
The Court cites five experts as diagnosing Appellant with intellectual disability. Two of those experts (Dr. Scott and Dr. Correia) explicitly hedged their diagnoses, one (Dr. Coxe) did not use adaptive testing, and the other two (Dr. Mayfield and Dr. Craig) were defense experts who relied on flawed tests. As the Court explains earlier in its opinion, Dr. Scott diagnosed Appellant as having “mild [intellectual disability] versus borderline functioning.” Dr. Scott said that he hedged his intellectual disability diagnosis with a borderline-functioning diagnosis because Appellant‘s passive-aggressive nature and his stubbornness undermined the usefulness of the clinical interview as a measure of intellectual ability.55 On original submission we pointed out that Dr. Scott explained his hedging even further: “I don‘t know that he is as low as he appeared. Stubbornness can contaminate that and someone can look worse than they really are.”56 And the
That leaves the defense experts. I have already detailed the fact that Appellant tested positive for malingering on three different measuring instruments in the course of Dr. Mayfield‘s IQ testing. In addition, I have detailed Dr. Mayfield‘s apparent naivete when it comes to whether capital murder defendants malinger on these sorts of tests, and I have pointed out the fact that the scores obtained were wildly out of line with the score of 74 on the 1992 test. And I have also detailed the numerous, serious flaws in the adaptive-deficits testing conducted by Dr. Craig. Further undermining the reliability of the defense experts’ conclusions was Appellant‘s history of deception and manipulation, his strong incentive to malinger on testing and to fake adaptive deficits, and evidence that he in fact malingered on testing and faked adaptive deficits for the specific purpose of avoiding the death penalty.
The Court‘s conclusion that “expert after expert” found Applicant to be intellectually disabled suggests more consensus than what the record actually shows: the experts who were not aligned with the defense hedged their diagnoses. The experts’ hedging shows that the medical community was not really of “one mind” about whether Appellant was intellectually disabled. But, importantly, even if they were, all of the expert evidence was subject to a systemic problem: malingering and lack of effort on Appellant‘s part, an issue that, at least to some degree, goes all the way back to 1991.
Similarly, the Court‘s contention that Appellant “consistently scored within the range for intellectual disability on intelligence testing administered across decades” suggests more than what the record shows. First, there were basically two clusters of tests: (1) tests taken from 1991 to 1993, and (2) tests taken in 2012 and 2013. Those two sets of tests, when taken together, can literally be said to be “across decades,” but the reality is that only two relatively narrow time periods of testing were involved. Second, of course, the score of 74 was not within the range for intellectual disability.
In any event, the later tests involved substantially lower scores than the earlier
The Court, in conclusion says that “a factfinder cannot substitute its opinion for that of all of the examining doctors.” But the Supreme Court has not said that. To the contrary, it has said that “the legal determination of intellectual disability is distinct from a medical diagnosis” and that “it is informed by the medical community‘s diagnostic framework.”60 It is the Supreme Court itself that uses the word “informed;” it does not say “dictated.” And what is to inform the legal conclusion is not the opinions of medical experts but the medical community‘s “diagnostic framework.” The Supreme Court has never said that the issue of intellectual disability is dictated by the opinions of medical experts, but that is the Court‘s holding today. It is entirely consistent with Supreme Court precedent to hold, instead, that a jury making the legal determination on intellectual disability can indeed reject the opinion of all of the examining doctors.
The Court also suggests that we should treat intellectual disability the way we treat insanity. But if the Court is suggesting that a jury must abide by the opinions of experts who relied on flawed procedures, I disagree. And if the Court is saying that a jury must accept an expert‘s opinion on insanity even if there is evidence that the defendant faked insane behavior in order to appear insane, I disagree with that. Similarly, we should not ignore the fact that all of the medical evidence suggesting intellectual disability in this case is flawed and that the non-medical evidence is, in the main, from biased and unreliable sources. And most of all, we should not ignore the evidence of pervasive malingering, deception, and fakery by this defendant.
I do not believe that the jury‘s verdict is manifestly unjust, shocks the conscience, or is clearly biased, so I would affirm the trial court‘s judgment in its entirety. Because the Court does not, I respectfully dissent.
Filed: May 12, 2021
Publish