Harvard Law Review Harvard Law Review Harvard Law Review

Capital Punishment

Kansas v. Carr

 

Since deciding in Gregg v. Georgia1×1. 428 U.S. 153 (1976). that the death penalty may be imposed within the confines of the Eighth Amendment, the Supreme Court has held that capital sentencers must weigh mitigating evidence when determining whether to sentence a defendant to death.2×2. See Woodson v. North Carolina, 428 U.S. 280, 304 (1976) (plurality opinion); see also Lockett v. Ohio, 438 U.S. 586, 605 (1978) (plurality opinion). The Court has not, however, required that sentencers receive explicit guidance in this weighing process.3×3. See Weeks v. Angelone, 528 U.S. 225, 232–34 (2000); Buchanan v. Angelone, 522 U.S. 269, 279 (1998). Last Term, in Kansas v. Carr,4×4. 136 S. Ct. 633 (2016). the Court took another step away from explicit guidance, holding that the Eighth Amendment “does not require capital sentencing courts ‘to affirmatively inform the jury that mitigating circumstances need not be proved beyond a reasonable doubt.’”5×5. Id. at 642 (quoting State v. Gleason, 329 P.3d 1102, 1148 (Kan. 2014) (per curiam)). Although the Court’s broad holding regarding mitigation was consistent with precedent, its discussion of the burden of proof applied by the jurors in Carr ignored both contextual empirical evidence and relevant social-scientific data.

In December 2000, Reginald Carr and his younger brother, Jonathan, embarked on a heinous crime spree that ended in the “Wichita Massacre.”6×6. Id. at 638. The spree began on December 7, when Reginald and an unidentified accomplice held up Andrew Schreiber at gunpoint and forced him to withdraw hundreds of dollars from several ATMs.7×7. Id.; State v. Carr, 331 P.3d 544, 575–76 (Kan. 2014) (per curiam). Four days later, Reginald and Jonathan followed Linda Ann Walenta home from orchestra practice and shot her three times as she sat in her car in her driveway.8×8. Carr, 136 S. Ct. at 638; Carr, 331 P.3d at 576–77. One of the bullets severed Walenta’s spine, and she died several weeks later from complications relating to the injury.9×9. Carr, 331 P.3d at 577. The spree culminated on the evening of December 14, when Reginald and Jonathan broke into a residence shared by Aaron S., Brad H., and Jason B.10×10. Id. In addition to the three roommates, Aaron’s friend Heather M. and Jason’s girlfriend, Holly G., were present.11×11. Id. Over the course of the next several hours, the Carrs — both armed with handguns — subjected their victims to unthinkable torture.12×12. See id. at 577–81. In addition to raping the women repeatedly,13×13. Id. at 578–80. Reginald and Jonathan forced the women and men to perform sex acts on each other,14×14. Id. at 578. and they ran several of the victims to local ATMs to withdraw money over the course of the night.15×15. Id. at 578–79. When they were done at the residence, the Carrs drove the victims to a nearby soccer field, lined them up on their knees, and shot them execution-style.16×16. See id. at 580; Carr, 136 S. Ct. at 639. Only Holly — whose hair clip miraculously deflected the bullet fired at her — survived and managed to escape, even after the Carrs ran her over with their truck.17×17. Carr, 136 S. Ct. at 639. Holly found a nearby house and contacted the police, who apprehended Reginald and Jonathan shortly thereafter.18×18. Carr, 331 P.3d at 581–85. The Carrs, in the meantime, had returned to the residence, ransacked it, and killed Holly’s dog. Carr, 136 S. Ct. at 639.

The Carrs were each charged with over fifty counts, including kidnapping, rape, and capital murder.19×19. State v. Carr, 329 P.3d 1195, 1204–05 (Kan. 2014) (per curiam). At trial, Reginald was convicted on fifty counts,20×20. Carr, 331 P.3d at 573. while Jonathan was convicted on forty-three.21×21. Carr, 329 P.3d at 1204. Jonathan was acquitted of all counts in the Schreiber case, id., and he had not been charged with unlawful possession of a firearm, whereas Reginald had been, id. at 1205. The Carrs were sentenced together,22×22. Carr, 136 S. Ct. at 640. and the jury was instructed that, in order to impose the death penalty on the defendants, “[t]he State has the burden to prove beyond a reasonable doubt that there are one or more aggravating circumstances and that they outweigh mitigating circumstances found to exist.”23×23. Brief for Respondent Reginald Dexter Carr, Jr. at 45, Carr, 136 S. Ct. 633 (No. 14-450). For each defendant, the jury returned four death sentences for the murders of Aaron S., Brad H., Jason B., and Heather M., a life sentence for Walenta’s murder, and over forty years of consecutive imprisonment for the remaining convictions.24×24. Carr, 329 P.3d at 1205.

The Kansas Supreme Court vacated the Carrs’ death sentences and remanded to the trial court for further proceedings.25×25. Carr, 331 P.3d at 574; Carr, 329 P.3d at 1205. First, the court found that “the defendants’ penalty phase cases were at least partially antagonistic to each other”26×26. Carr, 331 P.3d at 719. and that the trial court’s “refusal to sever the defendants’ penalty phase was error that violated both [defendants’] Eighth Amendment right to an individualized capital sentencing determination.”27×27. Id. at 720. Second, the court held that the trial judge erred in failing to instruct the jury that mitigating circumstances need not be proved beyond a reasonable doubt28×28. Id. at 732–33. and that this failure contravened both Kansas law and the Eighth Amendment of the Federal Constitution.29×29. Id.

The Supreme Court reversed and remanded.30×30. Carr, 136 S. Ct. at 646. The Carrs’ cases were consolidated with Kansas v. Gleason, 136 S. Ct. 633 (2016), which also raised the question whether capital jurors must be instructed that mitigating circumstances need not be proved beyond a reasonable doubt. See Brief for Respondent Sidney J. Gleason at i, Carr, 136 S. Ct. 633 (No. 14-452). Writing for the Court, Justice Scalia31×31. Justice Scalia was joined by every member of the Court save Justice Sotomayor. held that the Eighth Amendment “does not require capital sentencing courts ‘to affirmatively inform the jury that mitigating circumstances need not be proved beyond a reasonable doubt,’”32×32. Carr, 136 S. Ct. at 642 (quoting State v. Gleason, 329 P.3d 1102, 1148 (Kan. 2014) (per curiam)). and he rejected the Carrs’ contention that their joint sentencing proceedings violated their constitutional rights.33×33. Id. at 644–46.

Before reaching the merits of the case, the Court dismissed a jurisdictional challenge raised by the defendant in the Carr cases’ companion case, Kansas v. Gleason.34×34. 136 S. Ct. 633; see also id. at 641–42. Sidney Gleason had contended that the Kansas Supreme Court’s decision to vacate his death sentence was beyond review by the U.S. Supreme Court because it relied exclusively on “a rule of Kansas law announced in” two earlier Kansas Supreme Court decisions.35×35. Id. at 641. The Court rejected this claim, ruling that these earlier decisions implicated the Federal Constitution and that the Court had rejected a similar claim a decade prior in Kansas v. Marsh.36×36. 548 U.S. 163 (2006); see Carr, 136 S. Ct. at 641.

Having dispensed with the jurisdictional challenge, the Court proceeded to address whether the Eighth Amendment requires a capital sentencing court to instruct the jury that mitigating factors “need not be ‘proved beyond a reasonable doubt.’”37×37. Carr, 136 S. Ct. at 642 (quoting Gleason, 329 P.3d at 1148). First, the Court expressed uncertainty that “it is even possible to apply a standard of proof to the mitigating-factor determination.”38×38. Id. Whereas the existence of aggravating factors is “purely factual” — and whereas a jury could theoretically be instructed to apply a certain standard of proof to the existence of “the facts establishing mitigating circumstances” — the Court stated that the broader determination whether “mitigation exists . . . is largely a judgment call (or perhaps a value call).”39×39. Id. Furthermore, the Court pointed to prior cases in which it had not required that jurors be given express guidance with respect to mitigation.40×40. Id. (citing Weeks v. Angelone, 528 U.S. 225, 232–33 (2000); Buchanan v. Angelone, 522 U.S. 269, 275 (1998)).

The Court also rejected the defendants’ narrower contention that clarifying instructions were required in their cases because the actual capital sentencing instructions may have confused jurors into believing that the existence of mitigating factors needed to be proved beyond a reasonable doubt.41×41. Id. at 642–43. Citing Boyde v. California42×42. 494 U.S. 370 (1990). for the proposition that “[a]mbiguity in capital-sentencing instructions gives rise to constitutional error only if ‘there is a reasonable likelihood that the jury has applied the challenged instruction in a way that prevents the consideration of constitutionally relevant evidence,’”43×43. Carr, 136 S. Ct. at 642–43 (quoting Boyde, 494 U.S. at 380). the Court found that no such likelihood existed and that “no juror would reasonably have speculated that mitigating circumstances must be proved by any particular standard, let alone beyond a reasonable doubt.”44×44. Id. at 643.

Next, the Court rejected the Kansas Supreme Court’s conclusion that the Carrs’ joint sentencing proceeding violated their Eighth Amendment rights.45×45. Id. at 644–46. The Court stated that the defendants’ objections were essentially procedural in nature and that the Eighth Amendment was therefore “inapposite.”46×46. Id. Such claims should be brought instead under the Due Process Clause, and there they should succeed only if the improperly admitted evidence “render[ed] the jury’s imposition of the death penalty a denial of due process.”47×47. Id. at 644–45 (quoting Romano v. Oklahoma, 512 U.S. 1, 12 (1994)). Given the horrific nature of the Carrs’ crimes and the fact that the jury was specifically instructed to consider only evidence applicable to each defendant, the Court found that the joint sentencing proceedings could not have influenced the jury’s death verdicts.48×48. See id. at 645–46. Furthermore, the Court stated, “[j]oint proceedings are not only permissible but are often preferable when the joined defendants’ criminal conduct arises out of a single chain of events.”49×49. Id. at 645.

Justice Sotomayor wrote a lone dissent, arguing that the Court should not have heard the cases and that she would have “dismiss[ed] the writs as improvidently granted.”50×50. Id. at 646 (Sotomayor, J., dissenting). When state courts grant criminal defendants more protection than the Federal Constitution requires, Justice Sotomayor wrote, concerns for federalism — and for the limited impact a grant of certiorari might have — militate against Supreme Court intervention.51×51. See id. at 647–48. Furthermore, when the Court intervenes because a state has overprotected a federal right, the Court often supports its “conclusion by explaining why that right is not very important.”52×52. Id. at 648. Such explanations dissuade “States from adopting valuable procedural protections even as a matter of their own state law,”53×53. Id. and they also “risk[] turning the Federal Constitution into a ceiling, rather than a floor.”54×54. Id. at 649. For example, in “call[ing] into question the logic of specifying any burden of proof as to mitigating circumstances,” the Court stepped beyond its holding as to what the Eighth Amendment requires, and it “denigrate[d] the many States that do specify a burden of proof for the existence of mitigating factors as a matter of state law.”55×55. Id. at 648. Similarly, the Court was “not content to hold that the Eighth Amendment does not, strictly speaking, require severance of capital penalty proceedings,” but it instead opined on the relative demerits of severed proceedings, notwithstanding the fact that several states allow for severance.56×56. Id. at 649 (noting that Ohio, Georgia, and Mississippi allow for severance). These costs were especially unjustified in light of the limited benefits of intervention; on Justice Sotomayor’s account, the Kansas Supreme Court’s decision “turn[ed] on specific features of Kansas’ sentencing scheme,”57×57. Id. and it was thus “particularly unlikely to undermine other States or the Federal Constitution.”58×58. Id. at 650.

Although the Court’s broad holding regarding mitigation was consistent with precedent, its application of Boyde — and its corresponding discussion of the burden of proof applied by jurors in the Carrs’ case — was disconnected from relevant evidence. Specifically, the Court ignored contextual evidence from the trial record, evidence regarding the history and evolution of the contested jury instruction, and a significant body of empirical data on juror comprehension of sentencing instructions. In so doing, the Court deprived itself of information necessary to comprehensively determine whether there was a “reasonable likelihood”59×59. Boyde v. California, 494 U.S. 370, 380 (1990). that the jury applied the wrong burden of proof to mitigating evidence.

The Supreme Court has long held that capital sentencers cannot be constrained in their ability to consider mitigating factors and that they must instead be capable of “giving independent mitigating weight to aspects of the defendant’s character and record and to circumstances of the offense.”60×60. Lockett v. Ohio, 438 U.S. 586, 605 (1978) (plurality opinion) (emphases added); see also Penry v. Lynaugh, 492 U.S. 302, 327–28 (1989); Eddings v. Oklahoma, 455 U.S. 104, 110 (1982). The Court has been less zealous, however, in requiring trial courts to guide sentencers in the weighing process: In Buchanan v. Angelone,61×61. 522 U.S. 269 (1998). the Court held that there is no Eighth Amendment requirement “that a capital jury be instructed on the concept of mitigating evidence generally, or on particular statutory mitigating factors.”62×62. Id. at 270. Similarly, in Weeks v. Angelone,63×63. 528 U.S. 225 (2000). the Court stated that it had “never held that the State must structure in a particular way the manner in which juries consider mitigating evidence.”64×64. Id. at 233 (citing Buchanan, 522 U.S. at 276). Given these precedents, the Carr Court’s broad holding — that the Eighth Amendment does not require courts to inform jurors that mitigating factors need not be proved beyond a reasonable doubt — seems doctrinally straightforward.

But while the Court has been hesitant to prescribe general procedures for presenting mitigating evidence, it has not shied away from evaluating contested jury instructions. Since Boyde, the Court has upheld specific instructions as consistent with the Eighth Amendment only if there is no “reasonable likelihood that the jury has applied the challenged instruction in a way that prevents the consideration of constitutionally relevant evidence.”65×65. Boyde v. California, 494 U.S. 370, 380 (1990). In assessing whether such a reasonable likelihood exists, the Court has looked to the language of the instructions,66×66. See id. at 381–82; see also Weeks, 528 U.S. at 232–33; Buchanan, 522 U.S. at 277. as well as to case-specific details, such as “the context of the proceedings”67×67. Boyde, 494 U.S. at 383. and the length of jury deliberations.68×68. See Weeks, 528 U.S. at 235.

The Carr Court’s application of Boyde, by contrast, and its conclusion that “no juror would reasonably have speculated that mitigating circumstances must be proved by any particular standard, let alone beyond a reasonable doubt,”69×69. Carr, 136 S. Ct. at 643. were disconnected from relevant evidence regarding the jury’s comprehension. To begin with, the Court reached its conclusion about juror speculation exclusively through reliance on the text of the jury instruction, without any analysis of the broader courtroom proceedings.70×70. See id. at 642–44. Indeed, Justice Scalia spent only a single paragraph explaining that the jury could not reasonably have applied the wrong burden of proof to mitigating circumstances be-cause the sentencing instructions made clear that “mitigating circumstances . . . must merely be ‘found to exist.’”71×71. Id. at 643.

This text-centric approach is at odds with the Court’s traditional methodology, which takes into account broader contextual evidence from the trial record. In Boyde, for example, the Court held that the defendant’s reading of the instruction at issue was unreasonable, but it nevertheless proceeded to analyze the amount and kind of evidence introduced at trial, as well as the points made by counsel on both sides.72×72. See Boyde, 494 U.S. at 383–86. Similarly, in Buchanan, the Court found the language of the contested jury instruction to be unproblematic, but it still went on to examine the nature of the testimony and arguments at sentencing.73×73. Buchanan v. Angelone, 522 U.S. 269, 278 (1998). And in Weeks, the Court explicitly bolstered its Boyde analysis with “several empirical factors,” such as the fact that “each of the jurors affirmed in open court the verdict”74×74. Weeks v. Angelone, 528 U.S. 225, 234 (2000). and that “the jurors deliberated for more than two hours after receiving the judge’s answer to their question.”75×75. Id. at 235. This inquiry into contextual evidence was absent from the Court’s analysis in Carr.

In addition, the Carr Court looked past relevant evidence regarding the history and evolution of the contested jury instruction. In 2001 — the year prior to the Carrs’ trial76×76. See State v. Carr, 331 P.3d 544, 732 (Kan. 2014) (per curiam). — the Kansas Supreme Court had specifically held that “any instruction dealing with the consideration of mitigating circumstances should state . . . they need to be proved only to the satisfaction of the individual juror in the juror’s sentencing decision and not beyond a reasonable doubt.”77×77. State v. Kleypas, 40 P.3d 139, 268 (Kan. 2001) (per curiam). In response to this admonition, the Kansas pattern jury instruction was revised (though not until a decade later78×78. Carr, 136 S. Ct. at 650 (Sotomayor, J., dissenting). ) to explicitly state that “[m]itigating circumstances need not be proved beyond a reasonable doubt.”79×79. State v. Gleason, 329 P.3d 1102, 1146 (Kan. 2014) (per curiam) (alteration in original). These developments offer persuasive authority to conclude that the instruction at issue in Carr was “sufficiently confusing to demand higher clarity going forward.”80×80. Carr, 136 S. Ct. at 650 (Sotomayor, J., dissenting).

Furthermore, the Court has previously recognized such instructional revisions as probative in deciding whether to vacate a death sentence on account of juror confusion. In Mills v. Maryland,81×81. 486 U.S. 367 (1988). the Court vacated the defendant’s death sentence after finding “a substantial probability that reasonable jurors” may have believed juror unanimity was required for consideration of mitigating evidence.82×82. Id. at 384. Significant to the Court’s calculus in Mills was the fact that, subsequent to the defendant’s conviction, the Maryland Court of Appeals promulgated a new verdict form clarifying that unanimity was not required in order to consider mitigating evidence.83×83. See id. at 381–82. Although it made clear that the change in verdict form was not dispositive to its decision, the Court stated: “[W]e cannot avoid noticing these significant changes effected in instructions to the jury. We can and do infer from these changes at least some concern on the part of that court that juries could misunderstand the previous instructions . . . .”84×84. Id. at 382. Although Mills predates Boyde, its “substantial probability” inquiry is sufficiently analogous to Boyde’s “reasonable likelihood” test that the Mills Court’s analysis can inform future applications of Boyde. The Carr Court, by contrast, assigned no weight to — and, indeed, made no mention of — the fact that the Kansas pattern jury instruction had been revised at the state supreme court’s urging in order to inform jurors that mitigating circumstances need not be proved beyond a reasonable doubt.85×85. The Carr Court also failed to note that myriad other death penalty states specify a burden of proof relating to mitigation. See Brief for Respondent Sidney J. Gleason at 27, Carr, 136 S. Ct. 633 (No. 14-452) (“Of the thirty-one states that have the death penalty, at least twenty-four explicitly announce the defendant’s burden of proof for mitigating factors via statute or case law, and most of those call for an express instruction on this point . . . .” (footnote omitted)).

Perhaps most important, the Carr Court’s application of Boyde ran orthogonal to a significant body of empirical data suggesting generally that jurors tend to misunderstand mitigation, and finding particularly that jurors often believe mitigating circumstances must be proved beyond a reasonable doubt. Although this evidence is not specific to the Carrs’ trial in the same way that the courtroom proceedings and instructional revisions are, the empirical data nevertheless inform the baseline likelihood that jurors applied the correct burden of proof to mitigating circumstances.86×86. See generally Stephen E. Fienberg & Mark J. Schervish, The Relevance of Bayesian Inference for the Presentation of Statistical Evidence and for Legal Decisionmaking, 66 B.U. L. Rev. 771 (1986). One key source of such data is the Capital Jury Project (CJP), a multidisciplinary research program that began in 1991 with funding from the National Science Foundation.87×87. Capital Jury Project, U. Albany, http://www.albany.edu/scj/capital_jury_project.php [https://perma.cc/DXF3-LFFH]. See generally William J. Bowers, The Capital Jury Project: Rationale, Design, and Preview of Early Findings, 70 Ind. L.J. 1043 (1995). Since 1991, the CJP has “conducted extensive structured interviews of 1201 actual jurors in 354 capital trials in 14 states.” Leona D. Jochnowitz, Does Mental-Health Mitigating Evidence of Personality Disorders Make a Difference to Jurors in Capital-Sentencing Decisions?, 50 Crim. L. Bull. 344, 350 (2014). These interviews have shed especially bright light on the ways in which capital jurors think about aggravation and mitigation. See generally Stephen P. Garvey, Aggravation and Mitigation in Capital Cases: What Do Jurors Think?, 98 Colum. L. Rev. 1538 (1998). These data were also before the Court in Carr. See Brief for Respondent Sidney J. Gleason at 26–27, Carr, 136 S. Ct. 633 (No. 14-452). Among other questions, CJP researchers asked interview participants: “For a factor in favor of a life or lesser sentence to be considered, did it have to be proved ‘beyond a reasonable doubt,’ by ‘a preponderance of the evidence,’ or only to ‘a juror’s personal satisfaction’”?88×88. William J. Bowers et al., The Capital Sentencing Decision: Guided Discretion, Reasoned Moral Judgment, or Legal Fiction, in America’s Experiment with Capital Punishment 413, 437 (James R. Acker et al. eds., 2d ed. 2003). Out of the 1085 responses recorded across thirteen states, 49.2% wrongly answered “beyond a reasonable doubt.”89×89. Id. at 438 tbl.3. Furthermore, confusion persisted at the individual state level: even in the least confused state, Pennsylvania, roughly a third of jurors thought mitigating circumstances needed to be proved beyond a reasonable doubt,90×90. Id. while in the most confused state, Texas (which has performed over a third of the country’s executions since Gregg91×91. See Death Penalty Info. Ctr., Facts About the Death Penalty 3, http://www.deathpenaltyinfo.org/documents/FactSheet.pdf [https://perma.cc/HTA4-WEP2]. ), roughly two-thirds of jurors thought the reasonable doubt standard applied.92×92. Bowers et al., supra note 88, at 438 tbl.3. These responses were admittedly shaped by the question format, and the data did not include surveys with Kansas jurors. But the results are still at odds with the Carr Court’s holding that “no juror would reasonably have speculated that mitigating circumstances must be proved by any particular standard, let alone beyond a reasonable doubt.”93×93. Carr, 136 S. Ct. at 643.

The Carr Court’s Boyde analysis appears further disconnected from relevant evidence once the broader results from the CJP and other studies are taken into account. Of particular note, CJP interviewers found that “a number of . . . jurors . . . obviously did not understand the term [mitigation], at times actually confusing it with aggravation.”94×94. Ursula Bentele & William J. Bowers, How Jurors Decide on Death: Guilt Is Overwhelming; Aggravation Requires Death; and Mitigation Is No Excuse, 66 Brook. L. Rev. 1011, 1044 (2001) (emphasis added). These results are consistent with non-CJP interviews of capital jurors,95×95. See Peter Meijes Tiersma, Dictionaries and Death: Do Capital Jurors Understand Mitigation?, 1995 Utah L. Rev. 1, 18. as well as with studies of mock jurors.96×96. See Craig Haney & Mona Lynch, Comprehending Life and Death Matters, 18 Law & Hum. Behav. 411, 418–20 (1994). In one study conducted with college students, for example, researchers explained the California capital trial process before reading the California penalty instruction aloud three times.97×97. Id. at 419. Students were then asked to define the terms “aggravation” and “mitigation,” as those terms were used in the instruction.98×98. Id. Out of nearly 500 students, only 12% offered a legally correct definition of “mitigation,” while 30% offered a completely incorrect definition, and 11% were unable to respond at all.99×99. Id. at 421 tbl.1. The remaining 47% offered a partially correct definition. Id.

Instead of acknowledging this empirical reality, the Court stuck to its presumption that jurors follow their instructions,100×100. See Carr, 136 S. Ct. at 645 (citing Weeks v. Angelone, 528 U.S. 225, 234 (2000)); see also Weeks, 528 U.S. at 234 (citing Richardson v. Marsh, 481 U.S. 200, 211 (1987)). and it maintained its longstanding suspicion of social science.101×101. See J. Alexander Tanford, The Limits of a Scientific Jurisprudence: The Supreme Court and Psychology, 66 Ind. L.J. 137, 138 (1990). This wariness has been notable in the death penalty context, where the Court has rejected studies purporting to show that death-qualified juries are more likely to convict defendants,102×102. See Lockhart v. McCree, 476 U.S. 162, 168–73 (1986). During death qualification, prospective jurors are excluded if their opposition to the death penalty is sufficiently strong. See id. at 167 & n.1. as well as data suggesting that a victim’s race strongly influences whether a defendant receives the death penalty.103×103. See McCleskey v. Kemp, 481 U.S. 279 (1987). As it relates to juror comprehension, such empirical skepticism could have a number of motivations, including disagreement with the underlying survey methodology,104×104. See Nancy S. Marder, Bringing Jury Instructions into the Twenty-First Century, 81 Notre Dame L. Rev. 449, 470–71 (2006). or even concern that juror confusion will persist under any set of instructions.105×105. See id. at 472–73; cf. Gacy v. Welborn, 994 F.2d 305, 311 (7th Cir. 1993) (“[A]ctual levels of comprehension must be compared with achievable levels of comprehension, not . . . ideal levels.”).

Whatever the Court’s reasons for ignoring the CJP and related studies, these social-scientific data — like the record from the Carrs’ sentencing hearing and the history and evolution of the Kansas pattern jury instruction — seem undeniably relevant to the probability that jurors applied the wrong burden of proof to mitigating factors. By failing to take them into account, the Court deprived itself of information necessary to determine whether there was a “reasonable likelihood”106×106. Boyde v. California, 494 U.S. 370, 380 (1990). of constitutional error. Given the horrific facts of the case, the jury likely would have reached the same sentencing determination regardless of the burden of proof it applied, but the same may not be true in future cases in which the crimes at issue are less “[s]hocking.”107×107. Carr, 136 S. Ct. at 651 (Sotomayor, J., dissenting) (“The standard adage teaches that hard cases make bad law. I fear that these cases suggest a corollary: Shocking cases make too much law.”).