Kevan Brumfield v. Burl Cain, Warden

• Court: Court of Appeals for the Fifth Circuit
• Decided: February 10, 2016
• Docket: 12-30256
• Status: Published

Opinion

Case: 12-30256 Document: 00513375522 Page: 1 Date Filed: 02/10/2016

REVISED February 10, 2016

IN THE UNITED STATES COURT OF APPEALS FOR THE FIFTH CIRCUIT United States Court of Appeals Fifth Circuit

FILED No. 12-30256 December 16, 2015 Lyle W. Cayce KEVAN BRUMFIELD, Clerk

Petitioner - Appellee

v.

BURL CAIN, WARDEN, LOUISIANA STATE PENITENTIARY,

Respondent – Appellant

Appeal from the United States District Court for the Middle District of Louisiana

Before KING, CLEMENT, and ELROD, Circuit Judges. KING, Circuit Judge: Petitioner–Appellee Kevan Brumfield was convicted of first degree murder and sentenced to death in 1995. Following state court proceedings, Brumfield filed a petition for a writ of habeas corpus in the district court, arguing that he is ineligible for the death penalty under Atkins v. Virginia, 536 U.S. 304 (2002), because he is intellectually disabled. The district court found that the state court erred by not holding an Atkins hearing on whether Brumfield was intellectually disabled. Following a multi-day hearing in 2010, the district court granted Brumfield a writ of habeas corpus, finding that he was intellectually disabled under Louisiana’s statutory definition of Case: 12-30256 Document: 00513375522 Page: 2 Date Filed: 02/10/2016

No. 12-30256 intellectual disability. Without reaching the merits of Brumfield’s claim that he is intellectually disabled, this court reversed the district court’s judgment. This court held that because Brumfield had failed to satisfy the requirements of 28 U.S.C. § 2254(d), the district court should not have reached the merits of his Atkins claim. The Supreme Court reversed and remanded, holding that Brumfield had indeed satisfied the requirements of 28 U.S.C. § 2254(d) and that he was thus entitled to have his claim of intellectual disability under Atkins evaluated on the merits. On remand, we review for clear error the district court’s determination that Brumfield is, in fact, intellectually disabled. Because the district court’s determination that Brumfield is intellectually disabled is plausible in light of the record as a whole, its determination is not clearly erroneous. Accordingly, we AFFIRM the ruling of the district court. I. FACTUAL AND PROCEDURAL BACKGROUND The facts and procedural history of this case are recounted exhaustively in prior opinions. See Brumfield v. Cain, 135 S. Ct. 2269 (2015) [hereinafter Brumfield (S. Ct.)]; Brumfield v. Cain, 744 F.3d 918 (5th Cir. 2014) [hereinafter Brumfield (5th Cir.)]; Brumfield v. Cain (Brumfield II), 854 F. Supp. 2d 366 (M.D. La. 2012); Brumfield v. Cain (Brumfield I), No. CIV.A.04-787JJB-CN, 2008 WL 2600140 (M.D. La. June 30, 2008); State v. Brumfield, 737 So. 2d 660 (La. 1998) [hereinafter Brumfield (La.)]. We recount the facts and procedural history as relevant to the limited question before us today. A. The Original Crime and State Court Proceedings On January 7, 1993, Petitioner–Appellee Kevan Brumfield and an accomplice, Henri Broadway, opened fire on a Baton Rouge Police Department vehicle driven by Corporal Betty Smothers. Smothers was escorting Kimen Lee, an assistant manager at the grocery store where Smothers worked part time as a security guard, as Lee made the grocery store’s nightly bank deposit. Brumfield fired seven rounds from the left side of the police cruiser, and 2 Case: 12-30256 Document: 00513375522 Page: 3 Date Filed: 02/10/2016

No. 12-30256 Broadway fired five rounds from the right side. Lee survived, but Smothers did not. Baton Rouge police officers arrested Brumfield for Smothers’ murder on January 11, 1993. When police interrogated Brumfield, he initially denied any involvement in Smothers’ murder and claimed that he had been with his brother at the time. After Brumfield’s brother denied that claim, Brumfield gave a videotaped statement admitting that he drove the getaway car but denying that he murdered Smothers. Later, Brumfield gave another videotaped statement where he admitted to being in the bank parking lot and firing shots at the police car. Following a multi-week trial in June and July of 1995, a jury found Brumfield guilty of first degree murder. He was subsequently sentenced to death on July 3, 1995. Brumfield appealed his conviction, but the Supreme Court of Louisiana affirmed the state trial court. Brumfield (La.), 737 So. 2d at 662, 671. And the Supreme Court of the United States denied his petition for certiorari thereafter. Brumfield v. Louisiana, 526 U.S. 1025 (1999). In March 2000, Brumfield filed for postconviction relief with a state trial court before the Supreme Court of the United States issued its decision in Atkins, 536 U.S. at 321, prohibiting the execution of intellectually disabled criminals. 1 Brumfield later amended his state petition to assert an Atkins claim and argued that he was entitled to an evidentiary hearing on his intellectual disability claim. 2 Brumfield requested funds to develop his claim,

1 Consistent with the Supreme Court’s guidance, we use the term “intellectually disabled” instead of “mentally retarded.” The two terms describe “identical phenomen[a].” Hall v. Florida, 134 S. Ct. 1986, 1990 (2014). 2 Brumfield provided the following evidence of his intellectual disability:

1) his IQ score, obtained prior to trial, of 75; 2) his slow progress in school; 3) his premature birth; 4) his treatment at multiple psychiatric hospitals; 5) various medications he was prescribed; and 6) testimony that he exhibited slower responses than “normal babies,” suffered from seizures, and was hospitalized for months after his birth. 3 Case: 12-30256 Document: 00513375522 Page: 4 Date Filed: 02/10/2016

No. 12-30256 but the state trial court denied his petition in its entirety on October 23, 2003. Brumfield then filed a writ with the Supreme Court of Louisiana, alleging, inter alia, that the trial court erred by failing to hold an Atkins hearing. That court denied the writ without explanation. Brumfield v. State, 885 So. 2d 580, 580 (La. 2004). B. Federal District Court Proceedings Following the Supreme Court of Louisiana’s dismissal of his appeal, Brumfield petitioned the United States District Court for the Middle District of Louisiana for a writ of habeas corpus, asking the court “to declare him [intellectually disabled] and ineligible for the death penalty under Atkins.” Brumfield II, 854 F. Supp. 2d at 370. Brumfield filed an amended petition in 2007 re-raising his Atkins claim, supported by expert findings developed with federal funding. A magistrate judge recommended that, although the state court’s refusal to grant an Atkins hearing was “reasonable and in accordance with clearly established federal law,” the district court should consider the additional evidence Brumfield presented in his amended habeas petition. The magistrate judge explained that Brumfield had demonstrated cause for failing to provide the state court with expert evidence because the state court denied him funding to develop this evidence. The magistrate judge further reviewed the additional evidence submitted by Brumfield and concluded that he had established a prima facie case of intellectual disability and was thus entitled to an Atkins hearing. The district court adopted the magistrate judge’s report and recommendation and held an Atkins hearing July 12–16 and August 3–4, 2010, discussed in detail below. Brumfield II, 854 F. Supp. 2d at 370.

Brumfield (5th Cir.), 744 F.3d at 921 (footnotes omitted).

4 Case: 12-30256 Document: 00513375522 Page: 5 Date Filed: 02/10/2016

No. 12-30256 In its opinion granting Brumfield a writ of habeas corpus, the district court first addressed the legal prerequisites to a federal habeas hearing before addressing the substance of Brumfield’s Atkins claim. Brumfield II, 854 F. Supp. 2d at 373, 384. Under the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), Brumfield could obtain federal habeas relief only if, in rejecting his claim, the state court’s decision “was either ‘contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States,’ or was ‘based on an unreasonable determination of the facts in light of the evidence presented in the State court proceeding.’” Brumfield (S. Ct.), 135 S. Ct. at 2275 (quoting 28 U.S.C. § 2254(d)(1), (2)). The district court found that denying Brumfield an evidentiary hearing without providing him with the funds to develop his Atkins claim “represented an unreasonable application of then-existing due process law as determined by the Supreme Court” and therefore satisfied § 2254(d)(1). Brumfield II, 854 F. Supp. 2d at 383–84. The district court also concluded that the state trial court’s denial of an Atkins hearing “suffered from an unreasonable determination of the facts in light of the evidence presented . . . in violation of § 2254(d)(2).” Id. at 379. The district court then analyzed the merits of Brumfield’s Atkins claim. In determining whether Brumfield is intellectually disabled—and therefore barred from being sentenced to death under Louisiana law, La. Code Crim. Proc. Ann. art. 905.5.1(A)—the district court relied heavily on the American Association on Intellectual and Developmental Disabilities’ (AAIDD’s) 3 Mental Retardation: Definition, Classification, and Systems of Support (10th ed. 2002) [hereinafter Red Book], which “contains the current, consensus definition of

The AAIDD was formerly known as the American Association on Mental Retardation 3

(AAMR). 5 Case: 12-30256 Document: 00513375522 Page: 6 Date Filed: 02/10/2016

No. 12-30256 [intellectual disability],” as “Louisiana law tracks the clinical definition provided by the [Red Book].” Brumfield II, 854 F. Supp. 2d at 385–86. To establish an intellectual disability, the district court explained, “Brumfield bears the burden of proving by a preponderance of the evidence that he meets the statutory definition.” Id. at 385 (citing La. Code Crim. Proc. Ann. art. 905.5.1(C)(1)). All of the experts who testified in this case agreed on the relevant criteria for diagnosing intellectual disability. 4 Consistent with the guidance from the United States Supreme Court and the Louisiana Supreme Court and La. Code Crim. Proc. Art. 905.5.1, the experts agreed that an intellectual disability diagnosis requires satisfying a three part test: “(1) subaverage intelligence, as measured by objective standardized IQ tests; (2) significant impairment in several areas of adaptive skills; and (3) manifestations of this neuro- psychological disorder in the developmental stage.” Brumfield (S. Ct.), 135 S. Ct. at 2274 (quoting State v. Williams, 831 So. 2d 835, 854 (La. 2002)). Each expert also agreed that the diagnosis of intellectual disability is guided by the same relevant psychological and medical texts authored by the American Psychiatric Association (APA) and AAIDD. See generally AAIDD, Intellectual Disability: Definition, Classification, and Systems of Supports (11th ed. 2010) [hereinafter Green Book]; Red Book; AAMR, User’s Guide: Mental Retardation: Definition, Classification, and Systems of Supports (10th ed. 2002) [hereinafter User’s Guide]; APA, Diagnostic and Statistical Manual of Mental Disorders (rev. 4th ed. 2000) [hereinafter DSM-IV-TR]. While the experts agreed on the criteria for diagnosing intellectual disability, they disagreed on whether Brumfield met those criteria.

4At the Atkins hearing, the district court heard testimony from six expert witnesses— three each for Brumfield and the State—and several other witnesses. 6 Case: 12-30256 Document: 00513375522 Page: 7 Date Filed: 02/10/2016

No. 12-30256 1. Brumfield’s Three Expert Witnesses The asserted role of Brumfield’s first expert, Stephen Greenspan, Ph.D., 5 was to educate the court on intellectual disability. While Greenspan did not evaluate whether Brumfield was intellectually disabled, the district court held that Greenspan “is one of the foremost [intellectual disability] experts in the country.” Brumfield II, 854 F. Supp. 2d at 386. Greenspan testified generally as to the “proper use of the AAIDD’s clinical standards in making diagnoses of [intellectual disability].” Id. Beginning with the subaverage intelligence prong of the intellectual disability test, Greenspan explained that psychologists originally used an IQ score of 70 as the cutoff for determining whether an individual had an intellectual disability, but because of advances in scientific and statistical methods, the AAIDD uses “75 as the upper ceiling now” for a diagnosis of intellectual disability. Commenting on potential factors that may affect the validity of an individual’s IQ score, Greenspan explained that if an individual is “malingering,” which refers to intentionally performing poorly on a test, an IQ test score may not be valid. He further explained that consistently receiving the same IQ score across multiple tests generally rules out malingering by an individual. When Greenspan examined the IQ scores from Brumfield’s previous tests, the scores “[told him] that [the test subject] here . . . clearly me[t] prong one because all of these scores [we]re in the mild [intellectual disability] range.” Greenspan also noted that an individual’s IQ tends to remain stable over time, implying that Brumfield, absent some incident that lowered his IQ, has always had an IQ in the intellectually disabled range.

5Greenspan is a licensed psychologist, obtained his Ph.D. in 1976, and (at the time of the hearing), was employed as a visiting professor at the University of Colorado Medical School. The district court accepted him as an expert in intellectual disability and adaptive behavior. Brumfield II, 854 F. Supp. 2d at 386. 7 Case: 12-30256 Document: 00513375522 Page: 8 Date Filed: 02/10/2016

No. 12-30256 When discussing the second prong of the intellectual disability test— whether an individual has impairments in adaptive behavior 6—Greenspan explained that “adaptive functioning usually would determine whether somebody is really [intellectually disabled]” when a person’s IQ is close to the cutoff for an intellectual disability diagnosis. “Adaptive behavior has to do with how one functions in the real world . . . outside of the testing situation.” Adaptive behavior includes three domains: the practical domain, the social domain, and the conceptual domain. The practical domain concerns daily living skills, the social domain concerns whether an individual can conform to the rules of society, and the conceptual domain concerns quasi-academic skills applicable to the real world, such as telling time. A diagnosis of intellectual disability requires “at least one . . . major domain of a relative impairment.” However, Greenspan was careful to note that an impairment in one domain of adaptive behavior does not require the complete absence of adaptive behaviors in that domain and that it does not preclude the possibility that an individual possesses some strengths in particular areas. To measure adaptive behavior in an individual, psychologists administer tests, such as the Adaptive Behavior Assessment System (ABAS) questionnaires, to people who know or knew the individual being evaluated for an intellectual disability. Greenspan emphasized that “the more people you can talk to, the better picture you get of an individual.” He also noted the importance of interviewing the subject himself. Greenspan testified that when sufficient records are available, reviewing all of the available information can shed light on whether an individual has deficits in adaptive functioning. He further noted that reviewing records is important when evaluating whether an

“Adaptive behavior,” “adaptive functioning,” and “adaptive skills” are used 6

interchangeably in both professional psychology circles and during the district court’s Atkins hearing. 8 Case: 12-30256 Document: 00513375522 Page: 9 Date Filed: 02/10/2016

No. 12-30256 individual satisfies the third prong of an intellectual disability diagnosis— manifestations prior to the age of 18. Additionally, Greenspan explained that the presence or absence of “maladaptive behavior” is not relevant to the diagnosis of intellectual disability. Maladaptive behavior involves a “person act[ing] out” by, for example, “attack[ing] other people” and is “not used diagnostically.” Brumfield’s second expert, Ricardo Weinstein, Ph.D., 7 evaluated Brumfield for intellectual disability. During the course of his evaluation, Weinstein met with Brumfield on at least three separate occasions for between five and seven hours each time, administered psychological tests, and performed a clinical interview with Brumfield. Weinstein also reviewed school records, medical records, and other records relevant to Brumfield’s past. Finally, Weinstein interviewed at least 14 different individuals who knew Brumfield. Based on his evaluation, Weinstein diagnosed Brumfield as intellectually disabled. Focusing on the first criterion for intellectual disability, Weinstein administered two IQ tests to Brumfield in 2007. Id. at 389. Brumfield scored a 72 (95% confidence interval of 69–77) on the Stanford-Binet V and a 70 (95% confidence interval of 65–75) on the C-TONI. Both of these scores fall within the intellectually disabled range and thus meet the first prong of the intellectual disability test. Weinstein also noted that Brumfield’s scores on previous IQ tests were consistent with an intellectual disability diagnosis. 8

7 Weinstein received his Ph.D. in 1971, and at the time of the hearing, he practiced forensic psychology. The district court accepted him as an expert in intellectual disability and forensic neuropsychology. Id. Although the State questioned his credentials, correctly pointing out that he received his Ph.D. from a non-traditional school that is no longer in operation, we note that he is licensed by the State of California and completed a post-doctoral certificate at the Fielding Institute. 8 Based on other psychological testing, Weinstein ruled out malingering as a possible

explanation for Brumfield’s IQ scores. 9 Case: 12-30256 Document: 00513375522 Page: 10 Date Filed: 02/10/2016

No. 12-30256 Specifically, a 1995 administration of the WAIS-R by then-defense expert Dr. Bolter resulted in a score of 75 (95% confidence interval of 70–80), and a 2009 administration of the WAIS-IV by State’s expert Dr. Hoppe yielded a score of 70 (95% confidence interval of 67–75). Because all four of Brumfield’s full-scale IQ scores fell within the intellectually disabled range, Weinstein concluded that Brumfield had satisfied the first requirement for an intellectual disability diagnosis. With respect to his evaluation of Brumfield’s adaptive functioning, Weinstein explained that his job as a psychologist “is to identify deficits,” and not to identify strengths in adaptive behavior, as “the issue . . . of . . . strengths is not relevant.” His evaluation of Brumfield’s adaptive functioning included his interviews with Brumfield and his review of relevant records. Additionally, Weinstein administered ABAS questionnaires to six people who knew Brumfield during his developmental years. However, because Weinstein admitted that the results of the ABAS questionnaires were “not very reliable,” 9 the district court “f[ound] these tests to be of little or no value,” and did not rely on them in reaching its conclusion on Brumfield’s intellectual disability. Id. at 393. The court did, however, consider Weinstein’s interviews with the people to whom he administered the ABAS questionnaires and at least eight other individuals, as well as his review of the records. Based on these interviews and his review of school, hospital, and group home records, Weinstein “identif[ied] very significant deficits in all three domains” of adaptive behavior. First, Weinstein noted that Brumfield was

9 As Weinstein explained, the ABAS was designed to be used contemporaneously while he was “trying to see how Mr. Brumfield functioned prior to the age of 18,” which required him to “ask[] people to remember how [Brumfield] functioned” in the past. Because “these backward-looking questions rely principally upon the memories of the test-takers regarding Brumfield’s abilities dating back 15–20 years,” id. at 393, the scores derived from the ABAS are not, in Weinstein’s opinion, very reliable. 10 Case: 12-30256 Document: 00513375522 Page: 11 Date Filed: 02/10/2016

No. 12-30256 developmentally delayed. For example, Brumfield was “two years behind his chronological age in terms of achievement or even grade levels.” Weinstein also noted impairment in Brumfield’s “visual motor coordination.” In particular, “Brumfield’s writing abilities are severely limited.” According to Weinstein, to write a letter Brumfield “needs to have a guide” and “uses a piece of cardboard that he puts underneath the line” in order to write in a straight line. Brumfield “takes . . . a very long time to write a letter”; in fact, a one page letter “take[s him] several days to write.” When writing, Weinstein noted, Brumfield “gets assistance from people in death row.” With respect to Brumfield’s behavior in the community, Weinstein testified that after “look[ing] at the records [and] talk[ing] to people,” he concluded that Brumfield “had problems with attention” and “with language comprehension.” Weinstein also concluded that Brumfield never learned any skills that could lead to gainful employment. Although Brumfield quit his job in order to sell drugs so that he could make more money, Weinstein stated that this did not suggest that Brumfield was able to obtain or maintain gainful employment. Commenting on the third prong of the intellectual disability inquiry, Weinstein noted that many of the adaptive behavior deficits, such as Brumfield’s academic progress lagging two years behind his age, were present during Brumfield’s developmental years. Although not part of the intellectual disability diagnosis, Weinstein pointed to several risk factors present in Brumfield’s history that support the conclusion that Brumfield manifested symptoms of an intellectual disability before he turned 18. For example, Brumfield’s mother “had psychiatric problems and was being medicated” and did not have “access to prenatal care . . . until she was about six months pregnant.”

11 Case: 12-30256 Document: 00513375522 Page: 12 Date Filed: 02/10/2016

No. 12-30256 Brumfield’s third expert, Victoria Swanson, Ph.D., also evaluated him for intellectual disability. 10 Swanson initially reviewed Brumfield’s records, particularly his school records, and the reports of other experts. Based on this review, she confirmed the earlier diagnosis of intellectual disability. Although she did not meet with Brumfield prior to confirming his intellectual disability diagnosis, she later met with him for five hours, interviewed people familiar with Brumfield during his developmental years, and broadened her review of the records. Swanson stated that nothing she reviewed or learned after writing her report changed her opinion or diagnosis. After reviewing all of the full-scale IQ scores Brumfield had received, Swanson opined that all of his scores fell within the range of intellectual disability and therefore concluded that Brumfield had satisfied the first prong of the intellectual disability test. Turning to the second prong—adaptive behavior—Swanson discussed Brumfield’s educational history extensively. In 1983, two teachers referred Brumfield for an evaluation within the school system. As part of this evaluation, Brumfield took a number of psychological tests, which indicated that Brumfield was functioning academically between 20 and 41 months behind his chronological age. 11 Based on the results of this evaluation, Brumfield was given the “exceptionality of behavior disorder” and

10 Swanson is a licensed psychologist in the State of Louisiana, and received her Ph.D. from Louisiana State University (LSU) in 1999. She has over 20 years of experience working with intellectually disabled patients. She also assisted the Louisiana legislature in drafting the bill that eventually became the statute governing intellectual disability at issue in this case. The district court accepted Swanson as an expert in intellectual disability and psychology. 11 The Illinois Test for Polylinguistic Abilities indicated that he was functioning at an

age level 41 months behind his chronological age, the Peabody Picture Vocabulary Test indicated Brumfield was 20 months below his chronological age level, and the Woodcock Language Proficiency Battery indicated Brumfield lagged approximately 24 months behind his chronological age. As measured by the Woodcock-Johnson Psychoeducational Battery in 1983, Brumfield’s reading level fell into the seventh percentile. 12 Case: 12-30256 Document: 00513375522 Page: 13 Date Filed: 02/10/2016

No. 12-30256 placed into a classroom setting appropriate for students with this disorder. 12 After spending three years in the special education classroom, Brumfield again took a number of psychological tests. Explaining these tests, Swanson noted that “there hasn’t been any progress academically over the three years that [Brumfield] continued to be in [the behavior disorder] class, and he seems to have plateaued at about the same grade level.” She further opined that Brumfield did not make any progress in the behavior disorder classroom because he, in fact, suffered from an intellectual disability. Explaining that students with behavior disorder typically catch up to their peers once their behavioral needs are met, Swanson stated that Brumfield simply plateaued between a fourth and sixth grade level, which was “consistent with a person with [an intellectual disability] more so than with a person who is just behaviorally disordered.” Swanson also discussed Brumfield’s reading and writing skills at length. She noted that, while in prison, Brumfield possessed both elementary-school- level and collegiate dictionaries, but he was only able to effectively use the elementary-school-level dictionary. Discussing his reading ability more generally based on her interview with Brumfield, she said “he was able to read 60 words a minute, which is extremely low for someone his age, but would be consistent for someone with a fourth grade reading level trying to read at the tenth.” Based on her evaluation of Brumfield, Swanson opined that “a diagnosis of [intellectual disability] would be appropriate for [Brumfield]. He

12 Swanson explained that individuals can have both a behavior disorder and an intellectual disability. Moreover, “[t]here is a high instance of aggression amongst students with [intellectual disability]” because they “are being asked to do things that they can’t do,” which leads to frustration and aggression. When a student has both a behavior disorder and an intellectual disability, she explained, schools often place the student into the behavior disorder classroom. 13 Case: 12-30256 Document: 00513375522 Page: 14 Date Filed: 02/10/2016

No. 12-30256 meets criteria one; he meets criteria two, and . . . there’s evidence of deficits in at least two areas prior to the age of 18.” 2. The State’s Three Expert Witnesses The State’s first expert was Donald Hoppe, Psy.D. 13 Hoppe explained that his primary role in evaluating Brumfield “was the administration of IQ testing in determining an IQ range.” Hoppe administered the WAIS–IV, which is one of the “gold standard” IQ tests, to Brumfield on March 13, 2009. On this test, Brumfield obtained a full-scale IQ score of 70 with a 95% confidence interval of 67 to 75. Hoppe explained that these “results are not that different from the results of Dr. Weinstein’s testing,” suggesting that the IQ scores obtained by both Hoppe and Weinstein are credible. Hoppe noted that he believed “that these scores represent the low end of what Mr. Brumfield’s intellectual range is” because “with more effort, his scores would have been higher.” However, Hoppe explicitly agreed that Brumfield meets the first requirement of an intellectual disability diagnosis based on the IQ test he administered and the previous scores that were consistently between 70 and 75. Although his primary role was to administer IQ testing to Brumfield, Hoppe also reviewed the available records from Brumfield’s past and commented generally on whether Brumfield is intellectually disabled. 14 Hoppe noted that Brumfield had taken an IQ test in 1984, and although no actual score was included in the records concerning the test, a report indicated that Brumfield scored in the “dull normal” range which implied a score between 80

13 Hoppe received his doctorate from Baylor University in 1981 and is a licensed psychologist in the State of Louisiana. He estimated that he has performed “hundreds, if not thousands” of IQ tests over his career. The district court accepted Hoppe as an expert in “clinical and forensic psychology.” 14 Hoppe did not interview anyone familiar with Brumfield. He only reviewed written

records. 14 Case: 12-30256 Document: 00513375522 Page: 15 Date Filed: 02/10/2016

No. 12-30256 and 89. Hoppe further noted that although Brumfield had been evaluated previously by psychologists and psychiatrists, he was only diagnosed with conduct disorder, 15 never with an intellectual disability. Hoppe also discussed Brumfield’s past as it related to the adaptive functioning prong of the intellectual disability test. Discussing Brumfield’s two videotaped confessions to the police following the murder of Corporal Smothers, Hoppe stated that these were “good snapshot[s] of what . . . [Brumfield] was functioning like at the time of the crime.” Hoppe noted that Brumfield appeared to be quick-thinking and gave a “detailed description of the streets in Baton Rouge,” which was not consistent with his having an intellectual disability. With respect to the crime itself, Hoppe agreed that it was fairly complicated, requiring planning and coordination. Hoppe also explained that Brumfield’s previous criminal behavior was important to his conclusion that Brumfield has no intellectual disability. Brumfield appeared to pick “weak victims” in several successive crimes, suggesting that he has the capacity to plan and organize. Discussing earlier details of Brumfield’s life, Hoppe opined that Brumfield’s lack of long-term employment, his lack of a checking account, and the fact that he never entered into a contract, could result from Brumfield being lazy or the fact that he was only 20 years old when he was arrested. He stated that these factors did not necessarily suggest that Brumfield has an intellectual disability. Hoppe also stated that drug dealing is “a form of employment” and that selling drugs requires a skill set that is not necessarily compatible with an intellectual disability diagnosis.

15Conduct disorder is, essentially, the childhood version of antisocial personality disorder. “The essential feature of conduct disorder is a repetitive pattern of behavior in which the basic rights of others or major age-appropriate norms or rules are violated,” i.e., conduct disorder is characterized by aggressive behavior. 15 Case: 12-30256 Document: 00513375522 Page: 16 Date Filed: 02/10/2016

No. 12-30256 The State’s second expert, Robert V. Blanche, M.D., 16 testified primarily as to whether Brumfield had deficits in adaptive functioning. Although Blanche evaluated Brumfield for intellectual disability, he had never heard of the AAMR/AAIDD, Red Book, Green Book, or User’s Guide before his deposition in this case and “was thus unfamiliar with [the AAIDD’s] diagnostic definitions.” 17 Id. at 388. He stated that instead of the AAIDD’s materials and definitions, psychiatrists rely on the DSM-IV instead. In explaining his evaluation of Brumfield, Blanche noted that he was not familiar with the standard adaptive behavior scales used by psychologists and had received no formal training in administering psychological testing. In conducting his evaluation, Blanche did not interview anyone other than Brumfield himself, noting that he did not “feel that [he] would get reliable information” from such interviews. Therefore, beyond his interview with Brumfield, Blanche’s inquiry into Brumfield’s adaptive functioning was limited to the available written records. In the records Blanche reviewed, there was no diagnosis of intellectual disability prior to the Atkins hearing despite multiple evaluations by psychologists and psychiatrists in the past. Blanche explained that Brumfield’s case was “a classic case of conduct disorder” and noted that, while many of the psychologists and psychiatrists who had previously evaluated Brumfield had diagnosed him with some form of conduct disorder, none of them had diagnosed him with an intellectual disability. Reviewing Brumfield’s records from the several group homes where he resided over the years, Blanche recalled a number of reports that Brumfield participated in sports and other group activities. Assessing the two videotaped

16 Blanche received his M.D. from LSU Medical School in 1981 and, at the time of the hearing, worked part time as a psychiatrist in the East Baton Rouge Parish jail, where he identified prisoners in need of mental health care. The district court accepted him as an expert in forensic psychiatry. 17 Blanche admitted this in a deposition that took place in January 2010.

16 Case: 12-30256 Document: 00513375522 Page: 17 Date Filed: 02/10/2016

No. 12-30256 confessions Brumfield gave to the police following Corporal Smothers’ murder, Blanche noted that Brumfield had no problems explaining himself to the police even in the face of complex questions. Based on Brumfield’s description of the events leading up to Smothers’ murder, Blanche concluded that the crime clearly involved planning, as Brumfield “scoped out [the] situation.” Additionally, Blanche explained that Brumfield’s other behaviors in the community, though often illegal, also demonstrated his adaptive behavior. For example, Brumfield chose to deal drugs instead of working a typical job not because he was unable to work a typical job but because dealing drugs was more lucrative. Similarly, Brumfield was able to “rent” a car by offering its owner drugs in exchange for the use of the car. Based on his review of the available records, Blanche concluded that, “to a reasonable medical certainty, [Brumfield] is not [intellectually disabled].” Despite this conclusion, Blanche admitted, on cross-examination, that “[Y]eah. I think he has some weaknesses. And in adaptive functioning that there are some—there are some, I will call it deficient. But to how significant they are, is, I think, a question.” He further agreed that Brumfield possesses weaknesses in several domains of adaptive functioning. Identifying specific weaknesses, Blanche stated that Brumfield’s impulsivity fits into the social domain of adaptive behavior and his inability to follow rules fits into the practical domain of adaptive behavior. The State’s final expert, John Bolter, Ph.D., 18 had previously evaluated Brumfield in 1995, written a report based on that evaluation, and testified in the penalty phase of Brumfield’s original trial. However, all of Bolter’s original records and raw data from his 1995 evaluation were destroyed. Bolter stated

18 Bolter received his Ph.D. from the University of Memphis, and at the time of the hearing was a practicing clinical neuropsychologist. 17 Case: 12-30256 Document: 00513375522 Page: 18 Date Filed: 02/10/2016

No. 12-30256 he remembered little about Brumfield’s 1995 evaluation and did not independently recall which materials he reviewed as part of that evaluation. Over Brumfield’s objection, the court accepted Bolter as an expert but restricted his testimony to the scope of his 1995 report. Id. at 388. In preparing his report, Bolter administered “a standard neuropsychological battery of tests to explore . . . brain function, assessing things such as visual spatial skills, language functioning, memory abilities, conceptual or executive functions, motor functions, and basic sensory perception functions.” Based on the tests he ran, Bolter “didn’t see any clear evidence of organic brain dysfunction.” He “saw that [Brumfield] had what [Bolter] thought was an attention deficit hyperactivity disorder and . . . nonspecific learning difficulties . . . borderline intellectual functioning, and . . . an antisocial personality.” Bolter also administered the WAIS-R to Brumfield to measure his IQ. His full-scale IQ score was “in the range of 75” which put Brumfield in the “borderline mentally defective range.” Based on this test and all of the information available to him in 1995, Bolter did not diagnose Brumfield with an intellectual disability. 3. Other Witnesses and Expert Materials In addition to its three experts, the State called five other witnesses to testify at Brumfield’s Atkins hearing. Warrick Dunn was Corporal Smothers’ oldest son. Dunn met with Brumfield on October 23, 2007. Commenting on Brumfield’s verbal abilities, Dunn stated that the two of them “had a conversation like two adults” and agreed that Brumfield was able to express himself well. Jerry Callahan, a retired Baton Rouge Police Department lieutenant, was the lead investigator of Corporal Smothers’ murder. Callahan interrogated Brumfield and was responsible for videotaping Brumfield’s two confessions. Callahan stated that during the five hours he spent with Brumfield, Brumfield never had any problems communicating and, in fact, 18 Case: 12-30256 Document: 00513375522 Page: 19 Date Filed: 02/10/2016

No. 12-30256 “communicated easily.” None of the State’s final three witnesses testified substantively on Brumfield’s intellectual disability. In addition to his three testifying experts, Brumfield also relied on a report compiled by James Merikangas, M.D. 19 In his report, Merikangas stated that a neurological examination of Brumfield revealed no acquired brain damage or ongoing disease. The district court recognized that the implication of this report is that Brumfield’s cognitive deficiencies stem from an underlying disability, as no physical damage to Brumfield’s brain explains his problems. Additionally, the report implies that these deficiencies have been present for the entirety of Brumfield’s life, as no physical damage occurring after his developmental years explains his problems. 4. The District Court’s Conclusion on Intellectual Disability Beginning with the first prong of the intellectual disability test, the district court found that, based on its analysis of Louisiana law and the mental health literature, “an IQ score of 75 or below does not preclude a finding of mild [intellectual disability] for Atkins purposes.” Brumfield II, 854 F. Supp. 2d at 389. After listing Brumfield’s scores on previous IQ tests, the court explained that his “scores consistently show him scoring between 70 and 75 on various IQ tests, a range which falls squarely within the upper bounds of mild [intellectual disability] according to the AAIDD's clinical definition.” Id. at 389–90. Further, the court noted that “[e]very expert that has testified in this matter has admitted that Brumfield meets the intellectual functioning prong of the [intellectual disability] test as set forth in La.C.Cr.P. art. 905.5.1(H)(1).” Id. at 390. Turning to the second prong and relying on the Red Book, the court explained that “Prong Two involves an assessment of Brumfield's adaptive

19 Merikangas received his M.D. in 1969 and is board certified in neuropsychiatry. 19 Case: 12-30256 Document: 00513375522 Page: 20 Date Filed: 02/10/2016

No. 12-30256 skills in the areas of conceptual, social, and practical skills” and that “[h]e must show a significant limitation in at least one of those three domains to satisfy the adaptive skills prong.” Id. at 392 (citing Red Book, supra, at 14). “Without reliable standardized measures available, the [district c]ourt [relied] on the testimony of the expert witnesses and their reports, the [c]ourt’s independent evaluation of Brumfield’s social, educational, medical, and criminal histories, and a common sense appraisal of Brumfield’s actions and abilities.” Id. at 393. In doing so, the district court remained cognizant that an intellectual disability “is ruled in by areas of impairment but is not ruled out by areas of competence” and that “‘people with [intellectual disabilities] are complex human beings’ who may have ‘strengths in one aspect of an adaptive skill in which they otherwise show an overall limitation.’” Id. (quoting Red Book, supra, at 8). The court further noted that it “must take into account the retrospective diagnostic guideline admonishing practitioners to ‘not use past criminal behavior or verbal behavior to infer [a] level of adaptive behavior.’” Id. (quoting Red Book, supra, at 22). However, the court recognized the “propensity of Louisiana courts to take such maladaptive criminal behavior into account when discussing the adaptive skills prong of the [intellectual disability] test.” Id. at 394. “With these important precepts in mind,” the district court evaluated each of the three domains of adaptive behavior under the AAIDD guidelines. Id. at 396. The court began with the conceptual skills, or “functional academics,” domain. Id. First, the court found that “Brumfield's writing abilities are severely limited,” as he “cannot write freehand” and “takes an inordinate amount of time to write a simple, one-page letter.” Id. Second, Brumfield does not have adequate reading abilities, as he reads at “a fourth grade level.” Id. Third, “Brumfield has a dismal record of academic accomplishments in the classroom.” Id. And Brumfield “reached a plateau 20 Case: 12-30256 Document: 00513375522 Page: 21 Date Filed: 02/10/2016

No. 12-30256 somewhere between the fourth and sixth grade, which is where mildly [intellectually disabled] individuals generally fall.” Id. Based on the procedural posture of this case, the district court noted that it was required to “view, more or less in isolation, whether Brumfield [met] the clinical criteria.” Id. at 401. In weighing the credibility of the experts in this case, the district court ultimately found the testimony of Weinstein and Swanson more credible than the testimony of Blanche on the second prong of the intellectual disability test. 20 Id. Blanche “lacked basic knowledge about the AAIDD’s standards until he was deposed in this case shortly before the hearing.” Id. Blanche also “failed to conduct interviews with anyone other than Brumfield himself, which [ran] afoul of the basic guidelines for retrospective diagnoses.” Id. Beyond the expert testimony, the court held that it could not “accord great weight to the facts of the crime, even though they must be taken into account, because the diagnostic guidelines for assessing maladaptive behavior as a part of adaptive skills ha[d] not been sufficiently shown to be present in th[e] case.” Id. “Ultimately, the [district c]ourt f[ound] that, based on the credibility of petitioner’s witnesses combined with the documented problems with the bases of testimony by the State’s experts, Brumfield [showed] by a preponderance of the evidence that he ha[d] significantly limited conceptual skills.” Id. “[O]n balance, the evidence [demonstrated that Brumfield met] the AAIDD’s definition of [intellectual disability] with respect to the conceptual domain of adaptive behavior.” Id. “Because Brumfield’s deficit in conceptual skills

20 The State’s other expert, Hoppe, did not make any determinations on whether Brumfield had significant limitations in adaptive behavior. 21 Case: 12-30256 Document: 00513375522 Page: 22 Date Filed: 02/10/2016

No. 12-30256 satisfie[d] Prong Two of the [intellectual disability] test, the [district c]ourt [conducted] only a brief review of the other two domains.” 21 Id. The district court next addressed the final prong of the intellectual disability test: whether the disability manifested prior to age 18. Id. at 403. The court credited Weinstein’s unrebutted testimony that “there is no question that [Brumfield] had very serious problems from very early on in life.” Id. Swanson reached a similar conclusion in her report. Id. Merikangas evaluated Brumfield in 2007 and concluded that he had no “acquired brain damage or ongoing disease that might negate the existence of an organic reason for Brumfield’s [intellectual disability].” Id. While Brumfield was evaluated during his youth by “no less than six doctors,” none of whom diagnosed him as intellectually disabled, “Swanson [gave] the [c]ourt a compelling reason to not draw a negative inference due to the lack of childhood diagnosis.” Id. at 403– 04. Additionally, “[e]tiological factors appear[ed] to bolster the conclusion that Brumfield was and is [intellectually disabled].” 22 Id. at 404. Weinstein testified that Brumfield’s mother “took psychotropic medication during her pregnancy” and that Brumfield weighed only “three and a half pounds” and suffered fetal distress at birth. Id. “The etiological risk factors, along with

21 Analyzing the social skills domain, the court found that “[o]n balance, this domain [was] a close call, but [it] d[id] not find Brumfield [met] the criteria for a significant overall deficit in the domain of social skills.” Id. at 402. Considering the practical skills domain, the court found that “Brumfield ha[d] not met his burden of showing he ha[d] significant deficits in practical skills.” Id. at 403. 22 As Greenspan explained, “[e]tiology has to do with cause and effect or things that

put the person at risk that could explain why he became [intellectually disabled].” Greenspan further explained that “for the most part, when we talk about etiology, we are talking about something biological,” such as “an infection or a brain malformation that came about in utero . . . [or] some physical cause that organically places the person at risk” of developing an intellectual disability. Environmental causes of intellectual disability also exist, such as severe child abuse; and some etiological risk factors are both environmental and biological such as “oxygen deprivation at birth, or a low birth weight.” 22 Case: 12-30256 Document: 00513375522 Page: 23 Date Filed: 02/10/2016

No. 12-30256 Brumfield’s school and medical records, indicate[d] that his mental health problems and developmental delays occurred prior to adulthood.” Id. at 405. “Based on the showing of substantial intellectual functioning and adaptive behavior deficiencies detailed above, the [district c]ourt credit[ed] the testimony of Brumfield’s experts and f[ound that] Brumfield ha[d] met his burden to show by a preponderance of the evidence that those deficits occurred before he turned 18.” Id. Because the district court concluded that Brumfield was intellectually disabled, it granted his petition for a writ of habeas corpus, rendering him ineligible for execution. Id. at 405–06. C. Proceedings in the Fifth Circuit and Supreme Court The State timely appealed the district court’s grant of the writ to this court. Brumfield (5th Cir.), 744 F.3d at 922. This court reversed the district court, concluding that Brumfield’s habeas petition did not satisfy either of § 2254(d)’s requirements. Id. at 927. First, because this court determined that none of the Supreme Court’s precedents required a state court to grant an Atkins petitioner funds to develop his claim, it rejected the district court’s conclusion that the state court had unreasonably applied clearly established federal law. Id. at 925–26. Second, because this court’s “review of the record persuade[d it] that the state court did not abuse its discretion when it denied Brumfield an evidentiary hearing,” it held that the state court’s decision did not rest on an unreasonable determination of the facts. Id. at 926. Having concluded that Brumfield failed both of the requirements of 28 U.S.C. § 2254(d), this court did not review the district court’s determination that Brumfield was intellectually disabled. Id. at 927. However, in a footnote, this court noted that “[e]ven if we were to consider the new evidence presented to the district court, we likely would hold that Brumfield failed to establish an Atkins claim.” Id. at 927 n.8.

23 Case: 12-30256 Document: 00513375522 Page: 24 Date Filed: 02/10/2016

No. 12-30256 The Supreme Court granted certiorari and vacated this court’s decision on June 18, 2015, in a 5–4 decision. Brumfield (S. Ct.), 135 S. Ct. at 2283. The Court explained that to obtain an Atkins evidentiary hearing, a defendant in Louisiana must “put forward sufficient evidence to raise a ‘reasonable ground’ to believe him to be intellectually disabled.” Id. at 2274 (citing Williams, 831 So. 2d at 861). The Court held that the state court’s refusal to grant Brumfield an Atkins hearing rested on two unreasonable factual determinations that related directly to the three-part test for intellectual disability. Id. at 2276– 82. First, the Court noted that “the state court apparently believed” that Brumfield’s IQ score of 75 and an expert witness’s testimony that he “may have scored higher on another test . . . belied the claim that Brumfield was intellectually disabled because they necessarily precluded any possibility that he possessed subaverage intelligence.” Id. at 2277. However, the Court explained, “this evidence was entirely consistent with intellectual disability.” Id. The Court further explained—relying on its prior decision in Hall v. Florida, 134 S. Ct. 1986 (2014), and Louisiana statutory law and caselaw— that “Brumfield’s reported IQ test result of 75 was squarely in the range of potential intellectual disability.” Id. at 2278. “To conclude . . . that Brumfield’s reported IQ score of 75 somehow demonstrated that he could not possess subaverage intelligence therefore reflected an unreasonable determination of the facts.” Id. Second, the Court held that the state court unreasonably determined that “the record failed to raise any question as to Brumfield’s ‘impairment . . . in adaptive skills.’” Id. at 2279. Even under the interpretation of the second prong of the intellectual disability test “most favorable to the State,” the Court held that it was unreasonable for the state court to conclude that Brumfield lacked deficits in adaptive behavior. Id. at 2279–81. The Court noted a number of examples of Brumfield’s deficits in the state trial court record. Id. 24 Case: 12-30256 Document: 00513375522 Page: 25 Date Filed: 02/10/2016

No. 12-30256 at 2279–80. For example, when Brumfield was born, he had a low birth weight and “slower responses than other babies.” Id. at 2279. Brumfield was placed “in special classes in school and in multiple mental health facilities.” Id. One report from one of these facilities “questioned his intellectual functions,” and Dr. Bolter noted that Brumfield had only a “fourth-grade reading level . . . with respect to ‘simple word recognition,’” and did not even reach that level with respect to “comprehension.” Id. at 2280. “All told,” the Court concluded, “the evidence in the state-court record provided substantial grounds to question Brumfield’s adaptive functioning” because “[a]n individual, like Brumfield, who was placed in special education classes at an early age, was suspected of having a learning disability, and can barely read at a fourth-grade level, certainly would seem to be deficient in both ‘[u]nderstanding and use of language’ and ‘[l]earning.’” 23 Id. (alteration in original) (citation omitted). Finally, with respect to the third prong of the test, the Court noted that “the state trial court never made any finding that Brumfield had failed to produce evidence suggesting he could meet this age-of-onset requirement.” Id. at 2282. Therefore, there was no “determination on that point to which a federal court had to defer in assessing whether Brumfield satisfied § 2254(d).” Id. The Court noted that “[i]f Brumfield presented sufficient evidence to suggest that he was intellectually limited, as we have made clear he did, there is little question that he also established good reason to think that he had been

23 The Court also noted that:

An individual who points to evidence that he was at risk of “neurological trauma” at birth, was diagnosed with a learning disability and placed in special education classes, was committed to mental health facilities and given powerful medication, reads at a fourth-grade level, and simply cannot “process information,” has raised substantial reason to believe that he suffers from adaptive impairments.

Id. at 2281. 25 Case: 12-30256 Document: 00513375522 Page: 26 Date Filed: 02/10/2016

No. 12-30256 so since he was a child.” Id. at 2283. Based on its conclusion that the state trial court decision “was based on an unreasonable determination of the facts in light of the evidence,” 28 U.S.C. § 2254(d)(2), the Supreme Court held that “Brumfield ha[d] satisfied the requirements of § 2254(d).” Brumfield (S. Ct.), 135 S. Ct. at 2283. Accordingly, the Court reversed the judgment of this court and remanded the case for further proceedings. Id. The sole remaining issue on remand is whether the district court clearly erred when it found Brumfield was intellectually disabled, as the Supreme Court held that Brumfield had satisfied § 2254(d) and that Brumfield “was therefore entitled to have his Atkins claim considered on the merits in federal court.” Id. at 2273. II. STANDARD OF REVIEW “[T]he determination of whether a defendant is [intellectually disabled] is inherently an intensively factual inquiry.” State v. Williams, 22 So. 3d 867, 887 (La. 2009); see also State v. Turner, 936 So. 2d 89, 98 (La. 2006). Because intellectual disability is a factual finding, this court reviews a district court’s determination that an individual is intellectually disabled for clear error. 24 Rivera v. Quarterman, 505 F.3d 349, 361 (5th Cir. 2007). “A finding is clearly erroneous only if it is implausible in the light of the record considered as a whole.” Id. (quoting St. Aubin v. Quarterman, 470 F.3d 1096, 1101 (5th Cir. 2006)); see also Anderson v. City of Bessemer City, 470 U.S. 564, 573 (1985) (“[A] finding is ‘clearly erroneous’ when although there is evidence to support it, the reviewing court on the entire evidence is left with a definite and firm conviction that a mistake has been committed.” (quoting United States v. United States Gypsum Co., 333 U.S. 364, 394–95 (1948))). “If

24The State never mentions the standard of appellate review in its brief, and despite direct questions at oral argument, the State refused to acknowledge the appropriate standard of review. In its brief and also at oral argument, the State argued that the district court refused to introduce the state trial court record into evidence when, in fact, the district court allowed the State to introduce the vast majority of the state court record into evidence. 26 Case: 12-30256 Document: 00513375522 Page: 27 Date Filed: 02/10/2016

No. 12-30256 the district court’s account of the evidence is plausible in light of the record viewed in its entirety, the court of appeals may not reverse it even though convinced that had it been sitting as the trier of fact, it would have weighed the evidence differently.” Anderson, 470 U.S. at 573–74. “Where there are two permissible views of the evidence, the factfinder’s choice between them cannot be clearly erroneous.” Id. at 574. The Supreme Court has explained that: [W]hen a trial judge’s finding is based on his decision to credit the testimony of one of two or more witnesses, each of whom has told a coherent and facially plausible story that is not contradicted by extrinsic evidence, that finding, if not internally inconsistent, can virtually never be clear error.

Id. at 575. This court “cannot second guess the district court’s decision to believe one witness’ testimony over another’s or to discount a witness’ testimony,” and is thus “reluctant to set aside findings that are based upon a trial judge’s determination of the credibility of witnesses.” Canal Barge Co. v. Torco Oil Co., 220 F.3d 370, 375 (5th Cir. 2000). III. INTELLECTUAL DISABILITY Although the determination of whether an individual has an intellectual disability under Atkins is necessarily a question for the court to decide, this determination is heavily informed by clinical standards and guidelines. In Atkins, when the Supreme Court left to states the task of implementing its holding that intellectually disabled individuals may not be executed, it cited with approval the clinical standards of the AAIDD and APA. Atkins, 536 U.S. at 308–09, 317. The Supreme Court of Louisiana first implemented the Atkins mandate in Williams, 831 So. 2d at 835. Noting that the Atkins Court adopted a “‘clinical definition’ of [intellectual disability],” the Supreme Court of Louisiana explicitly relied on the definition of intellectual disability developed by the AAIDD and the APA in crafting the test for intellectual disability. Id. at 852. 27 Case: 12-30256 Document: 00513375522 Page: 28 Date Filed: 02/10/2016

No. 12-30256 Following Atkins and Williams, Louisiana enacted a statute providing that “[n]otwithstanding any other provisions of law to the contrary, no person who is [intellectually disabled] shall be subjected to a sentence of death.” La. Code Crim. Proc. Ann. art. 905.5.1(A). The statute defining intellectual disability at the time of the Atkins hearing provided as follows: (1)“[Intellectual disability]” means a disability characterized by significant limitations in both intellectual functioning and adaptive behavior as expressed in conceptual, social, and practical adaptive skills. The onset must occur before the age of eighteen years.

La. Code Crim. Proc. Ann. art. 905.5.1(H). 25 As Swanson stated in her testimony and as the district court noted, this definition tracks the Red Book’s definition of intellectual disability. The Red Book provides that “[intellectual

25The Louisiana legislature amended the statute in June 2014, which currently reads as follows:

A. Notwithstanding any other provisions of law to the contrary, no person with an intellectual disability shall be subjected to a sentence of death. ... H. (1) “Intellectual disability”, formerly referred to as “mental retardation”, is a disability characterized by all of the following deficits, the onset of which must occur during the developmental period: (a) Deficits in intellectual functions such as reasoning, problem solving, planning, abstract thinking, judgment, academic learning, and learning from experience, confirmed by both clinical assessment and individualized, standardized intelligence testing. (b) Deficits in adaptive functioning that result in failure to meet developmental and sociocultural standards for personal independence and social responsibility; and that, without ongoing support, limit functioning in one or more activities of daily life including, without limitation, communication, social participation, and independent living, across multiple environments such as home, school, work, and community.

La. Code Crim. Proc. Ann. art. 905.5.1. The district court relied on the older version of the statute, and we do the same here. However, we note that while the new statute is worded differently, the test for intellectual disability remains largely unchanged. 28 Case: 12-30256 Document: 00513375522 Page: 29 Date Filed: 02/10/2016

No. 12-30256 disability] is a disability characterized by significant limitations both in intellectual functioning and in adaptive behavior as expressed in conceptual, social, and practical adaptive skills,” and that “[t]his disability originates before age 18.” Red Book, supra, at 1. Since this statute was enacted, the Supreme Court has reiterated that “[t]he clinical definitions of intellectual disability . . . were a fundamental premise of Atkins.” Hall, 134 S. Ct. at 1999. In this case, the Supreme Court again cited with approval the clinical guidelines on intellectual disability. Brumfield (S. Ct.), 135 S. Ct. at 2274, 2278. Therefore, the district court properly relied on the clinical guidelines of the AAIDD and APA in assessing whether Brumfield satisfied the statutory test for intellectual disability, and we similarly look to these guidelines in our review of the district court’s decision. In reviewing the district court’s decision, we address seriatim the three prongs of the test for intellectual disability. A. First Prong: Intellectual Functioning The assessment of an individual’s intellectual functioning requires the administration of standardized intelligence testing. “The ‘significant limitations in intellectual functioning’ criterion for a diagnosis of intellectual disability is an IQ score that is approximately two standard deviations below the mean, considering the standard error of measurement for the specific instruments used.” Green Book, supra, at 31; accord Red Book, supra, at 58 (“[T]he ‘intellectual functioning’ criterion for diagnosis of [intellectual disability] is approximately two standard deviations below the mean, considering the [standard error of measurement] for the specific assessment instruments used.”). As Greenspan explained, IQ tests are normalized so that the mean score is 100 and the standard deviation is 15; thus, two standard deviations below the mean equates to a score of 70. This is consistent with the assessment of Louisiana law by the Supreme Court of the United States, as it explained that, “[t]o qualify as ‘significantly subaverage in general intellectual 29 Case: 12-30256 Document: 00513375522 Page: 30 Date Filed: 02/10/2016

No. 12-30256 functioning’ in Louisiana, ‘one must be more than two standard deviations below the mean for the test of intellectual functioning.’” Brumfield (S. Ct.), 135 S. Ct. at 2277 (quoting Williams, 831 So. 2d at 853). Although a score of 70 is two standard deviations below the mean score, both the Supreme Court of the United States and the Louisiana Supreme Court have rejected a bright-line numerical cutoff for intellectual disability. See Hall, 134 S. Ct. at 1996; Williams, 22 So. 3d at 888. As the Supreme Court of the United States explained in Hall, “[t]he concept of standard deviation describes how scores are dispersed in a population,” but “[s]tandard deviation is distinct from standard error of measurement, a concept which describes the reliability of a test.” 134 S. Ct. at 1994. The Court further explained that the standard error of measurement “reflects the reality that an individual’s intellectual functioning cannot be reduced to a single numerical score.” Id. at 1995. Therefore, “an individual’s score is best understood as a range of scores on either side of the recorded score.” Id. Thus, scores higher than 70 can satisfy the first prong of the intellectual disability test. The Supreme Court in Hall explicitly rejected the contention that an IQ score of 75 precludes the possibility of an intellectual disability diagnosis. Id. at 1996. Similarly, the Louisiana Supreme Court in State v. Dunn (Dunn III), 41 So. 3d 454, 470 (La. 2010), stated that “[t]he ranges associated with the two scores of 75 brush the threshold score for [an intellectual disability] diagnosis.” Moreover, the AAIDD recognizes that a score of 75 is consistent with an intellectual disability. Red Book, supra, at 59; see also DSM-IV-TR, supra, at 41–42. In this case, the district court concluded that Brumfield satisfied the first prong of the intellectual disability test based on his IQ scores. As found by the district court, Brumfield’s IQ test scores were as follows:

30 Case: 12-30256 Document: 00513375522 Page: 31 Date Filed: 02/10/2016

No. 12-30256 — In a 1995 WAIS–R test administered by then-defense expert Dr. Bolter, he scored a 75, with a 95% confidence interval of 70–80. — In a 2007 Stanford–Binet V test administered by petitioner’s expert, Dr. Weinstein, he scored a 72, with a 95% confidence interval of 69–77. — In a 2007 C–TONI test administered by Dr. Weinstein, he scored a 70, with a 95% confidence interval of 65–75. — In a 2009 WAIS–IV test administered by the State’s expert, Dr. Hoppe, he scored a 70, with a 95% confidence interval of 67–75.

Brumfield II, 854 F. Supp. 2d at 389–90. All four of the confidence intervals (the range of scores calculated from the standard error of measurement) surrounding Brumfield’s full-scale IQ scores include scores of 70 or below, and therefore satisfy the first prong of the intellectual disability test based on how both the Supreme Court and Supreme Court of Louisiana have analyzed IQ scores in the past. 26 Even ignoring the confidence intervals, no score exceeds 75, and the Supreme Court noted in Atkins, Hall, and Brumfield (S. Ct.), that a score of 75 can satisfy the first prong of the intellectual disability test. Brumfield (S. Ct.), 135 S. Ct. at 2278; Hall, 134 S. Ct. at 1996; Atkins, 536 U.S. at 309 n.5. Moreover, every single expert agreed that Brumfield’s scores satisfied the first prong of the intellectual disability test. 27 As this court noted

26 Weinstein explained that as long as the lower bound of the confidence interval includes a score of 70 or less, an individual can satisfy the first prong of the intellectual disability test. 27 The district court, experts, and parties discussed the import of the “Flynn effect,”

which describes the phenomenon whereby the American public’s score on any given IQ test increases by approximately three points per decade. Brumfield II, 854 F. Supp. 2d at 391. “Thus, when an older test is used to measure a test subject, the subject’s IQ score may be artificially inflated because that test was normalized using a past sample of Americans.” Id. at 391. To correct for the Flynn effect, a test subject’s score may be adjusted downward by 0.30–0.33 for every year that has elapsed since the test was normalized. Id. The State correctly points out that the Fifth Circuit has not recognized the Flynn effect. In re Salazar, 443 F.3d 430, 433 n.1 (5th Cir. 2006); see also In re Mathis, 483 F.3d 395, 398 n.1 (5th Cir. 2007). It is not necessary to decide whether to recognize the Flynn effect in this case, however, as Brumfield’s scores satisfy the first prong of the intellectual disability test without a Flynn effect adjustment. 31 Case: 12-30256 Document: 00513375522 Page: 32 Date Filed: 02/10/2016

No. 12-30256 in Rivera, 505 F.3d at 361, “[a] finding is clearly erroneous only if it is implausible in the light of the record considered as a whole.” Given that all of Brumfield’s reported IQ scores fell at or below 75 and that the experts’ conclusions were based on these scores, the district court’s conclusion that Brumfield met the first criterion for an intellectual disability diagnosis is not implausible and therefore is not clearly erroneous. The State argues that “assessments consistently demonstrated that Brumfield had an IQ in the 70-85 range.” However, the State does not point to specific IQ scores which demonstrate that Brumfield’s IQ fell within this range. Presumably, it refers to the tests administered to Brumfield in the 1980s. As Weinstein explained, no actual IQ scores from these tests were reported anywhere in Brumfield’s records; instead, the reports based on these IQ tests provided only descriptions of the ranges into which Brumfield’s scores fell. For example, Weinstein explained that one report described Brumfield’s IQ score as falling into the “dull normal” range, which Weinstein further explained corresponded to a score between 80 and 89. The district court’s discrediting of this range of scores in favor of reported, full-scale IQ scores was not clear error, as the Supreme Court similarly disregarded supposedly higher IQ scores when no actual score was provided. See Brumfield (S. Ct.), 135 S. Ct. at 2278–79. Moreover, multiple expert witnesses discredited this range of scores in favor of the reported scores, and this court “cannot second guess the district court’s decision to believe one witness’ testimony over another’s or to discount a witness’ testimony.” Canal Barge, 220 F.3d at 375. The State also argues that Brumfield’s scores may be explained by his low effort on the IQ tests. However, the experts in this case—including the State’s expert who administered IQ tests—also administered tests for malingering and found that Brumfield was, in fact, not malingering. Moreover, Greenspan explained that Brumfield’s consistent scores across multiple tests 32 Case: 12-30256 Document: 00513375522 Page: 33 Date Filed: 02/10/2016

No. 12-30256 over multiple years ruled out malingering. We decline the State’s invitation to second guess the district court’s decision to believe the multiple experts who stated that Brumfield’s scores were not a product of malingering. Accordingly, we find no clear error in the district court’s finding that Brumfield satisfied the first prong of the intellectual disability test. B. Second Prong: Adaptive Behavior “Adaptive behavior is the collection of conceptual, social, and practical skills that have been learned and are performed by people in their everyday lives.” Green Book, supra, at 43; see also Red Book, supra, at 73. Under the AAIDD’s definition, 28 a diagnosis of intellectual disability requires that an individual have significant limitations in at least one of the three domains of adaptive skills—conceptual, social, and practical skills. 29 Red Book, supra, at 14. The district court found that Brumfield showed significant limitations in the conceptual domain but not in the social or practical domains. Brumfield II, 854 F. Supp. 2d at 396–403. The first deficit the court found in the conceptual domain was Brumfield’s writing abilities, as Brumfield could not write in a straight line without an aid, took an “inordinate amount of time to write a simple, one-page letter,” and relied on the assistance of other inmates when writing letters. Id. at 396. In coming to this conclusion, the district court relied on Weinstein’s

28 The district court correctly noted that, as with the intellectual functioning prong, the AAIDD prefers that practitioners employ standardized testing to evaluate adaptive functioning. See Red Book, supra, at 76. However, utilizing standardized testing, such as the ABAS questionnaires administered by Weinstein, is difficult in situations requiring a retrospective diagnosis. In these situations, the district court correctly explained that the User’s Guide, supra, at 17–22, calls for additional inquiry into the subject’s past and interviews alongside the types of questionnaires used in situations of contemporaneous diagnosis. That additional inquiry and those interviews were conducted by two of Brumfield’s experts in this case. 29 Neither the State nor Brumfield contests the use of the “three domain” test on

remand. The State structures its argument that Brumfield has no deficits in adaptive skills around this test. 33 Case: 12-30256 Document: 00513375522 Page: 34 Date Filed: 02/10/2016

No. 12-30256 testimony, and in concluding that the State’s reliance on the “quality of his expressions in his prison correspondence is misplaced,” the court credited the testimony of Swanson. Id. The court next found that Brumfield’s reading skills were deficient. Id. The court explained that after listening to Brumfield read some of his letters, Swanson concluded he read at approximately a fourth grade level. Id. Finally, the district court found that “Brumfield has a dismal record of academic accomplishments.” Id. The court relied on the testimony of Weinstein, who stated that Brumfield was always behind in school because of developmental delays, and Swanson, who noted that Brumfield “reached a plateau somewhere between fourth and sixth grade, which is where mildly [intellectually disabled] individuals generally fall.” Id. In reaching its conclusion that Brumfield demonstrated significant limitations in the conceptual skills domain, the district court carefully explained its reasoning, identified the specific evidence it relied upon, and specifically credited the testimony of certain experts. Because nothing in the district court’s reasoning suggests its conclusion “is implausible in the light of the record considered as a whole,” Rivera, 505 F.3d at 361 (quoting St. Aubin, 470 F.3d at 1101), and because this court must give “due regard . . . to the opportunity of the trial court to judge of the credibility of the witnesses,” Anderson, 470 U.S. at 573 (quoting Fed. R. Civ. P. 52(a)), we hold that the district court’s finding is not clearly erroneous. Brumfield was only required to demonstrate significant limitations in one of the three domains of adaptive behavior to satisfy the legal and clinical tests for intellectual disability. Thus, the district court’s finding that Brumfield met “the AAIDD’s definition of [intellectual disability] with respect to the conceptual domain of adaptive behavior,” Brumfield II, 854 F. Supp. 2d at 401, was sufficient for the district court to conclude that Brumfield had satisfied the second prong of the intellectual disability test. 34 Case: 12-30256 Document: 00513375522 Page: 35 Date Filed: 02/10/2016

No. 12-30256 In challenging the district court’s conclusion, the State argues that Brumfield’s academic problems, which led to his being placed in special education classes, stemmed primarily from his behavior problems and conduct disorder, not an intellectual disability. However, the district court credited the testimony of Swanson, who explained that, at the time Brumfield attended school, school systems were urged to substitute diagnoses of conduct disorder for intellectual disability essentially for political reasons. 30 Id. at 397. Moreover, the Supreme Court noted that “[t]he diagnostic criteria for [intellectual disability] do not include an exclusion criterion; therefore, the diagnosis should be made . . . regardless of and in addition to the presence of another disorder.” Brumfield (S. Ct.), 135 S. Ct. at 2280 (quoting DSM-IV-TR, supra, at 47). Both the State and Brumfield tell “coherent and facially plausible stor[ies],” Anderson, 470 U.S. at 575, as either behavioral problems or an intellectual disability could explain all or some of Brumfield’s poor academic record. “When ‘the district court is faced with testimony that may lead to more than one conclusion, its factual determinations will stand so long as they are plausible—even if we would have weighed the evidence otherwise.’” Heck v. Triche, 775 F.3d 265, 284 (5th Cir. 2014) (quoting Nielsen v. United States, 976 F.2d 951, 956 (5th Cir. 1992)); see also Anderson, 470 U.S. at 574

30 The district court explained that:

Swanson [gave] the Court a compelling reason to not draw a negative inference due to the lack of childhood [intellectual disability] diagnosis. She points out that during Brumfield’s school years in the late 1970s, African–Americans males were b[e]ing disproportionately diagnosed with [intellectual disabilities]. School officials, psychologists, and appraisal teams were accordingly cautious not to over-represent black males as being [intellectually disabled] and were instead urged to consider other alternatives that would avoid placing the [intellectually disabled] label on them. Swanson confirmed that East Baton Rouge Parish schools, which Brumfield attended, had received this admonition.

Id. at 404. 35 Case: 12-30256 Document: 00513375522 Page: 36 Date Filed: 02/10/2016

No. 12-30256 (“Where there are two permissible views of the evidence, the factfinder’s choice between them cannot be clearly erroneous.”). The State also points to elements of Brumfield’s past that it argues demonstrate adaptive functioning. For example, Blanche testified that Brumfield “owned” a car, engaged in cash transactions by renting motel rooms, and helped his girlfriend financially. Although the district court acknowledged these activities, among others, it explained that “[m]ildly [intellectually disabled] people generally have mental ages ranging from seven to eleven,” and “[i]t is not inconceivable for someone around the age of ten to have the mental capacity” to engage in these types of activities. Brumfield II, 854 F. Supp. 2d at 398. The State also argues that Brumfield’s activities while in prison belie any intellectual disability, as he wrote letters, possessed books (including two dictionaries), and explained complex tasks to people over the phone. With respect to Brumfield’s writing letters, the district court credited the testimony of Weinstein and Swanson that “Brumfield requires assistance from other death row inmates to write his letters, . . . and thus the reliance by the States’ experts on the quality of his expressions in his prison correspondence is misplaced.” Id. at 396. The court further found that, based on Swanson’s testimony, “[t]he reading materials in his prison cell are targeted to middle school audiences and are consistent with someone who has [an intellectual disability].” Id. Finally, with respect to Brumfield’s phone calls, the district court found that they were “simply not sufficient to show adaptive strength in communication abilities,” and that “one or two instances of him exhibiting oral communication skills expected of adults could hardly be said to outweigh the other documented adaptive weaknesses in the conceptual domain,” as “strengths can coexist alongside weaknesses.” Id. at 399. Although the evidence emphasized by the State tends to undermine the district court’s 36 Case: 12-30256 Document: 00513375522 Page: 37 Date Filed: 02/10/2016

No. 12-30256 conclusion that Brumfield had significant limitations in adaptive functioning, we are “not entitle[d to]. . . reverse the finding of the trier of fact” even if we “would have weighed the evidence differently.” Anderson, 470 U.S. at 573–74. Because nothing the State emphasizes establishes that the district court’s account of the evidence is implausible, we hold that the district court’s finding—that Brumfield’s poor academic performance and his deficiencies in reading and writing constitute deficits in adaptive behavior—is not clearly erroneous. See id. at 573–74 (“If the district court’s account of the evidence is plausible in light of the record viewed in its entirety, the court of appeals may not reverse it even though convinced that had it been sitting as the trier of fact, it would have weighed the evidence differently.”). Furthermore, we note that the district court’s finding is not clearly erroneous because it has more evidentiary support than prior cases in which this court upheld a district court’s intellectual disability determination. In Wiley v. Epps, 625 F.3d 199, 219–22 (5th Cir. 2010), this court found no clear error when a district court held that petitioner Wiley had an intellectual disability based on deficits in functional academic skills, communication, and self-direction. In that case, Wiley was evaluated four separate times with conflicting results. Id. at 219–21. Based on these results and evidence that he struggled academically while in the military, the district court found that he was deficient in the area of functional academic skills. Id. at 221. This court refused to reverse the district court because doing so would essentially substitute the opinion of the State’s expert for Wiley’s experts. Id. at 218. As the district court was in a better position to judge the credibility of the experts, this court declined to reverse the district court. Id. In Rivera, the district court found that Rivera had “adaptive limitations,” including “consistent[] . . . academic problems.” 505 F.3d at 362. After remarking that the district court “is in a better position than this court to judge and weigh the credibility of the 37 Case: 12-30256 Document: 00513375522 Page: 38 Date Filed: 02/10/2016

No. 12-30256 witnesses who testified,” this court declined to find a clear error. Id. at 363. However, neither Wiley nor Rivera involved Louisiana law. Dunn III, on the other hand, did involve Louisiana law, and this court noted previously that, based on this case, it would likely determine that the district court erred in finding Brumfield intellectually disabled, 31 Brumfield (5th Cir.), 744 F.3d at 927 n.8. In that case, the Supreme Court of Louisiana reviewed a trial court’s determination that Dunn was not intellectually disabled following an Atkins hearing. Dunn III, 41 So. 3d at 455–56. Dunn had reported IQ scores of 70, 78, and 78. Id. at 462–63. Multiple experts administered ABAS scales, but like this case, the evidence on Dunn’s adaptive behavior conflicted. Id. at 463–70. After reviewing that evidence, the court noted that “[i]t is also important to consider the defendant’s behavior during the planning and commission of the instant crime as it relates to his adaptive skills functioning.” Id. at 471. In evaluating Dunn’s crime, the court found that “the evidence at trial established defendant engaged in the leadership and planning of a major bank robbery” and held that the defendant’s planning “with its premeditative aspects, clearly lacks the impulsiveness and non-

31 The State argues that the district court failed to consider other Louisiana cases addressing the question of how to factor criminal behavior into an evaluation of an individual’s adaptive functioning. However, the court recognized the “propensity of Louisiana courts to take such maladaptive criminal behavior into account when discussing the adaptive skills prong of the [intellectual disability] test.” Brumfield II, 854 F. Supp. 2d at 394. Addressing this propensity, the district court identified five cases where the Supreme Court of Louisiana “affirmed on direct appeal a jury’s assessment of death in the penalty phase of the trial where the [intellectual disability] issue was actually litigated.” Id.; see generally Williams, 22 So. 3d 867; State v. Anderson, 996 So. 2d 973 (La. 2008); State v. Lee, 976 So. 2d 109 (La. 2008); State v. Scott, 921 So. 2d 904 (La. 2006); State v. Brown, 907 So. 2d 1 (La. 2005). However, the district court found these cases distinguishable because the Supreme Court of Louisiana was required under Jackson v. Virginia, 443 U.S. 307 (1979), to apply a different standard of review than the standard that applies to Atkins hearings. Brumfield II, 854 F. Supp. 2d at 394. The court found that Dunn III “[was] the only Louisiana Supreme Court case on point.” We agree and find no error with the manner in which the district court factored Brumfield’s criminal behavior into its analysis of his adaptive functioning. 38 Case: 12-30256 Document: 00513375522 Page: 39 Date Filed: 02/10/2016

No. 12-30256 leadership interactions associated with [intellectually disabled] persons” based on “the firmly established facts of this case.” Id. at 471–72. The district court carefully considered this case and concluded that it could consider “evidence of the criminal action in the overall assessment if ‘firmly established facts’ show[ed] clear instances of premeditation and leadership.” Brumfield II, 854 F. Supp. 2d at 395. In considering the evidence of Brumfield’s criminal activity, the district court concluded that it was not sufficient to demonstrate an absence of deficits in the conceptual skills domain, id. at 398–401, and that nothing in the record suggested Brumfield “‘led’ this terrible scheme.” Id. at 400. The district court further reasoned that even if the crime involved planning and premeditation by Brumfield, “this particular instance [should not be] sufficient to overwhelm the other demonstrated showings of adaptive deficits in conceptual skills.” Id. Beyond the facts of Smothers’ murder, the State argues that other aspects of Brumfield’s criminal history demonstrate that he does not have significant limitations in adaptive functioning. First, the State contends that Brumfield’s two confession videos show his composure under pressure, ability to lie, and think quickly. However, the district court credited Swanson’s testimony that, in the first tape, Brumfield responded to cues from police and that, in the second tape, Brumfield spoke more quickly because he was more familiar with the topic at that point. Id. Second, the State argues that Brumfield’s history of drug dealing and other criminal behavior demonstrates his ability to plan, his ability to handle complex transactions, and his adaptive functioning generally. 32 Although the State is correct that Brumfield dealt drugs in the past, the court noted that “[t]he record is barren of any testimony

32The State notes that Brumfield demonstrated an ability to choose weak and vulnerable victims for his past crimes. We see nothing in the record concerning this ability that demonstrates clear error on the part of the district court. 39 Case: 12-30256 Document: 00513375522 Page: 40 Date Filed: 02/10/2016

No. 12-30256 regarding his efficacy in drug transactions,” id. at 398, and both Greenspan and Weinstein testified that Brumfield’s drug dealing was not inconsistent with an intellectual disability diagnosis. Third, the State argues that Brumfield’s ability to avoid the police after his crime demonstrates adaptive functioning, but the district court found that “[w]hile evading police and avoiding capture can exhibit raw physical skills, at other times those acts are just as consistent with primal survival instincts as they are with callous, cold- blooded calculation.” Id. at 399. Overall, the district court considered the facts surrounding Smothers’ murder as well as Brumfield’s other criminal activities. Thus, while the district court considered similar evidence as the trial court in Dunn III, it simply reached a different conclusion. Although this difference in findings based on relatively similar evidence certainly weighs against the conclusion that Brumfield is intellectually disabled, it does not necessarily demonstrate that the district court clearly erred based on the record before it. The Dunn III court recognized that trial courts are called on “to make exceedingly fine distinctions” between those who are mildly intellectually disabled and those who are not. Dunn III, 41 So. 3d at 469. We agree with the Dunn III court on this point. Accordingly, we decline to disturb the “exceedingly fine distinctions,” id., the district court made in this “intensively factual inquiry,” Williams, 22 So. 3d at 887. Even if we were to disagree about how to weigh the evidence in this case, the clear error standard “plainly does not entitle a reviewing court to reverse the finding of the trier of fact simply because it is convinced that it would have decided the case differently.” Anderson, 470 U.S. at 573. C. Third Prong: Onset during Developmental Years The final prong of the intellectual disability test requires that the disability manifest before the age of 18. The district court did not clearly err 40 Case: 12-30256 Document: 00513375522 Page: 41 Date Filed: 02/10/2016

No. 12-30256 in finding that Brumfield’s disability manifested during his developmental years. In fact, one of the principal findings of the district court with respect to Brumfield’s deficits in the conceptual skills domain—his poor academic record while in school—necessarily involved finding that the disability manifested before age 18. Brumfield II, 854 F. Supp. 2d at 396. Similarly, the district court credited Swanson’s testimony that while in the eighth grade, Brumfield read at only a third grade level. Id. Although none of the IQ tests was administered to Brumfield prior to the age of 18, Greenspan testified that IQ scores remain stable over time. Additionally, Merikangas evaluated Brumfield and found no physical problems with his brain that would explain his consistent IQ scores between 70 and 75, meaning that Brumfield’s disability stems from some underlying problem he has had all of his life. Finally, the district court pointed to etiological factors such as, inter alia, Brumfield’s low birth weight, fetal distress at birth, and family history of intellectual disability. Id. at 404–05. Although not dispositive, these factors certainly bolster the court’s conclusion that Brumfield’s intellectual disability manifested during his developmental years. Id. at 405. D. Expert Credibility and Brumfield’s Medical History On remand, the State correctly highlights a number of weaknesses in Brumfield’s expert witnesses that undermine their credibility. For example, Greenspan never evaluated Brumfield, Weinstein obtained his Ph.D. from an unaccredited institution, and Swanson diagnosed Brumfield prior to meeting with him. However, the district court explicitly weighed the credibility of different witnesses. Id. at 401. For example, the court pointed out that Blanche “lacked basic knowledge about the AAIDD’s standards until he was deposed in this case shortly before the hearing,” id. at 401, and that Hoppe failed to interview anyone other than Brumfield, id. at 387 n.21. Giving “due 41 Case: 12-30256 Document: 00513375522 Page: 42 Date Filed: 02/10/2016

No. 12-30256 regard” to the “opportunity of the trial court to judge the credibility of the witness[es],” Anderson, 470 U.S. at 573 (quoting Fed. R. Civ. P. 52(a)), we decline to disturb the district court’s findings, see also Dunbar Med. Sys. Inc. v. Gammex Inc., 216 F.3d 441, 453 (5th Cir. 2000) (“The burden of showing that the findings of the district court are clearly erroneous is heavier if the credibility of witnesses is a factor in the trial court’s decision.” (quoting Coury v. Prot, 85 F.3d 244, 254 (5th Cir. 1996))). All of the experts in this case agreed that Brumfield had never been diagnosed with an intellectual disability prior to the Atkins hearing, and the district court was rightly wary about a “made-for-litigation diagnos[i]s.” Brumfield II, 854 F. Supp. 2d at 404. However, Swanson gave the court “a compelling reason to not draw a negative inference due to the lack of childhood diagnosis” by explaining the political incentives in place at the time Brumfield was in school. Id. In doing so, Swanson told a “coherent and facially plausible” story. Anderson, 470 U.S. at 575. Therefore, the district court’s refusal to give preclusive effect to the lack of a previous diagnosis of intellectual disability is not clearly erroneous. Id. Overall, while the State points to evidence that undermines the district court’s conclusion that Brumfield is intellectually disabled, it has not pointed to sufficient evidence to establish that the district court’s finding of intellectual disability was not “plausible in light of the record viewed in its entirety.” Id. at 574. Therefore, we hold that the district court committed no clear error. IV. CONCLUSION In this case, we are called upon to determine whether the district court’s conclusion that Brumfield is intellectually disabled is clearly erroneous, i.e., whether we have a firm and definite conviction that the district court made a mistake here. Both the State and Brumfield present plausible views of the evidence, although, on balance, Brumfield’s witnesses were somewhat stronger 42 Case: 12-30256 Document: 00513375522 Page: 43 Date Filed: 02/10/2016

No. 12-30256 and presented a slightly more compelling view. Given that there are two permissible views of the evidence here and the Supreme Court’s guidance that the choice by a trier of fact between two permissible views of the evidence cannot be clearly erroneous, we find no clear error in the district court’s conclusion that Brumfield is intellectually disabled. Because the State has not demonstrated clear error on the part of the district court, we AFFIRM the ruling of the district court that Brumfield is intellectually disabled and, accordingly, ineligible for execution.