 |
|||||
|
|
|
|
|
||
|
|
|||||
|
|
|
|
|
||
SCHRIRO v. SUMMERLIN
United States Supreme Court No. 03-526
1. Is Ring v. Arizona a new rule of criminal procedure, subject to the retroactivity limit of Teague v. Lane, or is it a rule of substantive law and therefore fully retroactive under Bousley v. United States?
2. Does Ring qualify for the second exception to the Teague rule?
ORAL ARGUMENT DATE: April 19, 2004
Summary of facts and case
Summary of argumentI. Ring is a procedural rule, not a substantive one
A. Substance and procedure
B. The Arizona capital sentencing systemII. Ring does not qualify for the second Teague exception
A. The second exception
B. Accuracy and jury trial
C. Fundamental fairness and jury trialThere is a grave danger that the "annually improvised" jurisprudence of capital punishment is killing innocent people through lost deterrence
IN THE |
SUPREME COURT OF THE UNITED STATES |
|
|
| Dora B. Schriro, Director, Arizona Department of Corrections, |
|
Petitioner,
|
| vs. |
|
|
|
Respondent.
|
|
|
BRIEF AMICUS CURIAE OF THE |
The Criminal Justice Legal Foundation (CJLF) (1) is a non-profit California corporation organized to participate in litigation relating to the criminal justice system as it affects the public interest. CJLF seeks to bring the constitutional protection of the accused into balance with the rights of the victim and of society to rapid, efficient, and reliable determination of guilt and swift execution of punishment.
In Ring v. Arizona, 536 U. S. 584 (2002), this Court overturned the capital sentencing systems of at least five States, and possibly as many as nine. See id., at 608, n. 6. It did so despite those States' reliance on a precedent of this Court expressly approving the very system in question against precisely the same attack. The opinion dismissed stare decisis in a single paragraph, id., at 608, and made no mention of the States' massive reliance interest in the precedent. Cf. Dickerson v. United States, 530 U. S. 428, 443 (2000) (importance of stare decisis, even in constitutional cases); Planned Parenthood of Southeastern Pa. v. Casey, 505 U. S. 833, 855-856 (1992) (importance of reliance); Vasquez v. Hillery, 474 U. S. 254, 266 (1986) (departure only "for articulable reasons").
The question now is how much damage this about-face is going to do to the cause of justice. Will the well-deserved sentences of scores of America's most vicious murderers be overturned because of a practice that had been expressly approved at the times of their trials and direct appeals and which is extremely unlikely to have made any difference in the outcome? Such a result would be contrary to the interests of victims of crime and law-abiding society that CJLF was formed to protect.
Nearly 23 years ago, Brenna Bailey went to the home of Warren Summerlin to speak to his wife about a delinquent account. See State v. Summerlin, 675 P. 2d 686, 689 (Ariz. 1983); Summerlin v. Stewart, 341 F. 3d 1082, 1084 (CA9 2003) (en banc). Summerlin bashed in her skull, wrapped her body in a bedspread, and left her in the trunk of a car in a store parking lot. See 675 P. 2d, at 689-690, 696. There is also evidence that he raped her. See id., at 690, 696.
A jury convicted Summerlin of murder. Id., at 690. The trial judge found two aggravating circumstances: a prior violent felony conviction and commission of the present offense in an especially heinous, cruel, or depraved manner. Id., at 696. The trial judge imposed the death penalty upon finding these aggravating circumstances and no mitigating circumstances sufficient to call for leniency. Ibid. The Arizona Supreme Court unanimously affirmed after independent review of the circumstances.
Twenty years of state and federal collateral review followed. See 341 F. 3d, at 1091. Last September, the U. S. Court of Appeals for the Ninth Circuit, sitting en banc, held that this Court's decision in Ring v. Arizona, 536 U. S. 584 (2002) applied retroactively so as to require relief from the death penalty. All guilt-phase claims were rejected unanimously. Summerlin, F. 3d, at 1121. Judges Rawlinson, O'Scannlain, and Tallman dissented from the judgment on penalty. Id., at 1125. This Court granted certiorari on December 1, 2003.
Ring v. Arizona is a straightforward application of Apprendi v. New Jersey, and as such it has the same character. Apprendi is universally agreed to be a procedural decision subject to the retroactivity limit of Teague v. Lane, and it follows that Ring is also. Ring did not change any of the substantive criteria that Arizona law prescribes for which murderers will be sentenced to death. It changed only who decides one of those criteria: whether at least one aggravating circumstance is true. This is a quintessentially procedural rule.
Ring also does not qualify for the second exception to the Teague rule. Previous expansions of the right to have juries rather than judges determine guilt were held nonretroactive under the much looser standard of Stovall v. Denno, because the change did not affect the reliability of the verdict. The principles of these cases necessarily require that the further expansion in Ring cannot qualify under the more restrictive requirements of Teague.
Having one or all of the elements of an offense determined by a jury rather than a judge is not essential to the fundamental fairness of the proceeding. Such an argument is precluded by this Court's cases on juvenile proceedings and plain error.
Recent studies on deterrence provide substantial empirical support for what common sense has always told us--the death penalty will save innocent lives if it is actually enforced. Conversely, the obstruction of the death penalty through constant churning of the rules and retroactive application of the changes will diminish deterrence and cost innocent lives.
Anyone who doubts whether Apprendi v. New Jersey, 530 U. S. 466 (2000) established a procedural rule or a substantive rule need look no further than Apprendi itself for the answer. Defining the scope of the issue before it, the Apprendi Court said,
"although the constitutionality of basing an enhanced sentence on racial bias was argued in the New Jersey courts, that issue was not raised here. The substantive basis for New Jersey's enhancement is thus not at issue; the adequacy of New Jersey's procedure is. The strength of the state interests that are served by the hate crime legislation has no more bearing on this procedural question than the strength of the interests served by other provisions of the criminal code." Id., at 475 (footnote omitted; emphasis added).
Procedural rule changes have, for many years, had limited retroactive effect unless they are fundamental. See Part II, infra. In the past, expansions of the right of jury trial were not deemed so fundamental as to mandate retroactivity on habeas, even under earlier, more expansive views of retroactivity. Hence, it is not surprising that courts have been nearly unanimous that Apprendi is not retroactive. "No Court of Appeals, let alone this Court, has held that Apprendi has retroactive effect." Harris v. United States, 536 U. S. 545, 581 (2002) (Thomas, J., dissenting); infra, at 14.
The universal recognition that Apprendi is not retroactive, combined with the statement in Ring v. Arizona, 536 U. S. 584, 607 (2002) that "[w]e see no reason to differentiate capital crimes from all others in this regard," i.e., that Ring is simply an application of Apprendi, would seem to be sufficient to resolve the question of the retroactivity of Ring. To get around this obvious conclusion, the Ninth Circuit majority opinion in the present case advanced the astonishing proposition that, even though Apprendi is unquestionably a rule of procedure, Ring is nonetheless a rule of substantive law. See Summerlin v. Stewart, 341 F. 3d 1082, 1101, and n. 9 (CA9 2003). (2) This assertion has no merit.
The Ninth Circuit majority opinion relies on Erie R. Co. v. Tompkins, 304 U. S. 64 (1938) for the proposition that the distinction between substance and procedure "is not always easy to divine." Summerlin, 341 F. 3d, at 1099. Indeed, no one who has wrestled with the complexities of the Erie doctrine could deny that the classification is "sometimes a challenging endeavor." Gasperini v. Center for Humanities, Inc., 518 U. S. 415, 427 (1996) (emphasis added). At other times, though, the endeavor is clear, simple, and straightforward. This case falls in the latter category.
Gasperini illustrates when the problem is difficult and, by contrast, why this case is not. That case involved a New York statute, N. Y. Civ. Prac. Law and Rules (CPLR) §5501(c) (McKinney 1995), which established a limit on damages and assigned the task of applying that limit to the Appellate Division. "CPLR §5501(c), appraised under [Erie and its progeny], is both 'substantive' and 'procedural': 'substantive' in that § 5501(c)'s 'deviates materially' standard controls how much a plaintiff can be awarded; 'procedural' in that § 5501(c) assigns decisionmaking authority to New York's Appellate Division." Gasperini, 518 U. S., at 426 (emphasis added). The dual nature of this statute required reference back to the Erie doctrine's purpose for separating substance from procedure to answer the question of how to apply this statute in a diversity case. See id., at 429-431. In contrast, the single holding of Ring is that the Sixth Amendment requires that Arizona's enumerated aggravating factors be found by a jury. See 536 U. S., at 609; see also id., at 597-598, n. 4 (no broader claim made or considered). This holding falls squarely within Gasperini's description of a procedural rule: assignment of decisionmaking authority.
Although Gasperini comes from a different body of law, where the substantive/procedural line is not necessarily drawn in the same place, its description of what is substantive and what is procedural makes a good starting point. The substantive law is the set of rules that determine what judgment will follow from a given set of facts, putting to one side the questions of how those facts are determined and by whom.
"The substantive criminal law . . . is mostly concerned with what act and mental state, together with what attendant circumstances and consequences, are necessary ingredients of the various crimes. Criminal procedure . . . is concerned with the legal steps through which a criminal proceeding passes, from the initial investigation of a crime through the termination of punishment." 1 W. LaFave, Substantive Criminal Law §1.1(a), p. 4 (2d ed. 2003).
Bousley v. United States, 523 U. S. 614, 619-620 (1998) held that the limit on retroactivity established in Teague v. Lane, 489 U. S. 288 (1989) did not apply to a petitioner seeking to benefit from the rule of Bailey v. United States, 516 U. S. 137 (1995), because that rule is substantive rather than procedural. (3) Bailey held that "use" of a firearm for the purpose of 18 U. S. C. §924(c)(1) requires active employment, rather than the previous broad definition. See Bousley, supra, at 616. Bousley quoted Teague's definition of the first exception to its rule to support the proposition that applying Teague to a substantive rule would be "inconsistent with the doctrinal underpinnings of habeas review . . . ." See id., at 620-621. If the Bousley Court were revolutionizing Teague jurisprudence, rather than restating the first exception in different terms, it would surely have devoted more than two paragraphs to the discussion. Whether the principle involved here is an exception to Teague or simply outside the scope of Teague is not material. The situation is analogous to the old debate over whether a declaration of a party opponent is an exception to the hearsay rule or not hearsay. Compare Cal. Evid. Code §1220 (exception), with Fed. Rule Evid. 801(d)(2) (not hearsay). Either formulation gets us to the same result. The Ninth Circuit majority appears to have considered it obvious that the first exception does not apply, brushing it off in two sentences. See 341 F. 3d, at 1109. It is just as obvious that Bousley's equivalent sub-stantive/procedural distinction does not apply.
Some of the cases in this Court's modern capital punishment jurisprudence are undisputedly substantive. That is, they hold that a certain result must follow from certain facts, regardless of how or by whom those facts are determined. Enmund v. Florida, 458 U. S. 782, 797 (1982) held that a minor accomplice swept up in the felony-murder rule could not be executed. The procedural question of who decides whether a particular defendant qualifies for that rule was not resolved by Enmund and was settled years later in Cabana v. Bullock, 474 U. S. 376, 386 (1986), overruled on other grounds, Pope v. Illinois, 481 U. S. 497, 503, n. 7 (1987). Cabana recognized the Enmund rule as a "substantive limitation" and referred to the issue of "who makes the determination" as a "form of procedure." Ibid. Similarly, a defendant who is mentally retarded is ineligible for capital punishment under Atkins v. Virginia, 536 U. S. 304, 321 (2002). This is " 'a substantive restriction.' " Ibid. (quoting Ford v. Wainwright, 477 U. S. 399, 405 (1986)).
Penry v. Lynaugh, 492 U. S. 302 (1989) makes clear the equivalence between substantive status and the first Teague exception. Penry was the first case to hold that a proposed rule actually qualified for a Teague exception, and, with the possible exception of Bousley, it is the only Supreme Court case ever to find a rule that qualified. The rule proposed by Penry was the categorical exemption of the retarded from execution, i.e., the rule later adopted in Atkins. See id., at 328-329. In defining its first exception, Teague had quoted Justice Harlan's original proposal that "a new rule will be retroactive if it places ' "certain kinds of primary, private individual conduct beyond the power of the criminal law-making authority to proscribe." ' " Id., at 329 (quoting Teague, 489 U. S., at 307, in turn quoting Mackey v. United States, 401 U. S. 667, 692 (1971) (Harlan, J., concurring in part and dissenting in part)). In defining the exception, Justice Harlan distinguished the substantive due process cases, such as Griswold v. Connecticut, 381 U. S. 479 (1965), from procedural due process cases, with the limitation on retroactivity not applying to the substantive cases. See Mackey, supra, at 692, and n. 7.
Penry picks up on this substantive versus procedural distinction to expand the first exception beyond the literal words of Justice Harlan's definition and beyond cases which make the underlying conduct not criminal. "Justice Harlan did speak of in terms of substantive categorical guarantees accorded by the Constitution, regardless of the procedures followed." 492 U. S., at 329 (emphasis added). Examples include cases holding "that the Eighth Amendment, as a substantive matter, prohibits imposing the death penalty on a certain class of defendants because of their status, Ford v. Wainwright, supra, at 410 (insanity), or because of the nature of their offense, Coker v. Georgia, 433 U. S. 584 (1977) (rape) (plurality opinion)." Id., at 329-330. In criminal law, a substantive constitutional rule is a rule that a particular punishment cannot be imposed on a particular defendant for a particular offense regardless of how scrupulously fairly the facts supporting the judgment may be determined. Cf. Washington v. Glucksberg, 521 U. S. 702, 719-720 (1997). The Bousley rule, that Teague applies only to procedural rules and not substantive ones, and Penry's explanation of the first exception are simply two different ways of expressing the same principle.
Whether framed under Bousley or Penry, the question is the same. Did Ring find that the Arizona capital sentencing system was inconsistent with the Sixth Amendment on a substantive basis or a procedural basis? That is, did Ring find that the sentence could not be imposed on the facts found, "regardless of the procedures followed," Penry, 492 U. S., at 329, or did it find that the procedures for determining the facts were wanting.
Constrained by the long line of cases from Gregg v. Georgia, 428 U. S. 153 (1976), to the present, the sentencing systems of the states using capital punishment all operate within the same broad outline. That outline was described in Tuilaepa v. California, 512 U. S. 967 (1994). (4) "[T]he trier of fact must convict the defendant of murder and find one 'aggravating circumstance' (or its equivalent) . . . ." Id., at 972. That circumstance must narrow the class of murderers eligible for the death penalty, i.e., "it must apply only to a subclass of defendants convicted of murder." Ibid. The circumstance must not be too vague. See ibid. To perform their function, the eligibility factors must "require an answer to a question with a factual nexus to the crime or the defendant . . . ." Id., at 973. For this reason, they have always resembled the elements that distinguish a higher degree of an offense from a lower degree, and some states do the required narrowing in precisely that way, by defining a higher degree of murder. See id., at 972. Others have a factfinder enter a separate finding, variously named "aggravating circumstances" or "special circumstances." See ibid.; id., at 975.
Constrained by the case law, all states follow the eligibility decision with a selection decision, "where the sentencer determines whether a defendant eligible for the death penalty should in fact receive that sentence." Id., at 972. This decision is different in kind from the eligibility decision, see id., at 971, and hence it has separate requirements. See id., at 972. In the sentencing decision, the sentencer is not limited to factual questions but instead may consider open-ended factors. See id., at 977-978. Regardless of whether the question is phrased as "weighing," determining the "appropriate" penalty, or answering "special issues," the decision is inherently discretionary. See id., at 979 (even "unbridled discretion" is permitted at this stage). In contrast to the eligibility decision, which functions in all states like a higher degree of murder, whether designated as such or not, the selection decision is very much a traditional sentencing function. It does involve factfinding, since the sentencer must find whether defendant's proffered mitigating facts are true and, in some jurisdictions, may find additional aggravating facts beyond the eligibility factors. See id., at 975 (California); Jones v. United States, 527 U. S. 373, 377-378 (1999) (federal). The final decision, though, inevitably involves the exercise of discretion and a "reasoned moral response to the defendant's background, character, and crime." Penry v. Johnson, 532 U. S. 782, 788 (2001) (internal quotation marks omitted; emphasis in original).
Considered in this context, there really was no mystery how Arizona's pre-Ring system worked. In its initial opinion in Ring, the Arizona Supreme Court quoted with approval the Apprendi dissent's description of the operation of the Arizona capital sentencing system. State v. Ring, 25 P. 3d 1139, 1151 (Ariz. 2001). The state court then continued:
"In Arizona, a defendant cannot be put to death solely on the basis of a jury's verdict, regardless of the jury's factual findings. The range of punishment allowed by law on the basis of the verdict alone is life imprisonment with the possibility of parole or imprisonment for 'natural life' without the possibility of release. A.R.S. §13-703.A-E. It is only after a subsequent adversarial sentencing hearing, at which the judge alone acts as the finder of the necessary statutory factual elements, that a defendant may be sentenced to death. A.R.S. §13-703.B ('The hearing shall be conducted before the court alone. The court alone shall make all factual determinations required by this section or the constitution of the United States or this state.'). And even then a death sentence may not legally be imposed by the trial judge unless at least one aggravating factor is found to exist beyond a reasonable doubt. State v. Gretzler, 135 Ariz. 42, 54, 659 P.2d 1, 13 (1983); see also A.R.S. §13-703.E ('the court . . . shall impose a sentence of death if the court finds one or more of the aggravating circumstances enumerated'). Thus, when the state seeks the death penalty, a separate evidentiary hearing, without a jury, must be held; the death sentence becomes possible only after the trial judge makes a factual finding that at least one aggravating factor is present. The judge makes that finding on the basis of the evidence presented at trial and any other evidence presented at the aggravation/mitigation hearing. A.R.S. §13-703.C. If the judge finds an aggravating circumstance, he must then proceed to determine if there are any mitigating circumstances. A.R.S. §13-703.E. If the judge finds mitigating circumstances, he must then weigh them against the aggravators and decide by 'special verdict' whether a death sentence is appropriate. A.R.S. §13-703.D & E." Ibid.
To decide which aspects of this system Ring disapproved, which it left standing, and the procedural or substantive status of each, the statement from Gasperini, quoted supra, at 6, provides a useful template. Ring left untouched the requirement that at least one aggravating factor from the statutory list be found. Ring did not strike down the requirement to determine if there are any mitigating circumstances. Ring did not strike down the requirement that the sentence be determined based on a weighing of the two sets of circumstances and a decision of whether the death penalty is appropriate. Indeed, it could not have overturned any of these requirements without a wholesale uprooting of long-established jurisprudence as outlined in Tuilaepa, supra.
These were the substantive criteria for determining who was sentenced to death in Arizona: murderers for whom at least one aggravating circumstance is true, and for whom the death penalty is appropriate after weighing those circumstances against any mitigating circumstances. Just like the criterion for "how much a plaintiff can be awarded" in Gasperini, 518 U. S., 426, these substantive criteria decided "how much" of a penalty a murderer would receive. Nothing in Ring even remotely questions the validity of these criteria. The Ninth Circuit majority's statement that "Ring rendered Arizona's substantive capital murder statute unconstitutional," 341 F. 3d, at 1101-1102 (emphasis added), is simply incorrect.
The lone flaw in the Arizona system was that "the death sentence becomes possible only after the trial judge makes a factual finding that at least one aggravating factor is present." The fact that the Arizona law "assigns decisionmaking authority," cf. Gasperini, 518 U. S., at 426, to the trial judge was the one and only problem that Ring found with the system. "Because Arizona's enumerated aggravating factors operate as 'the functional equivalent of an element of a greater offense,' Apprendi, 530 U. S., at 494, n. 19, the Sixth Amendment requires that they be found by a jury." Ring, 536 U. S., at 609.
This holding of Ring is no more "substantive" than the holding of Apprendi. Justice Scalia explained this aspect of Ring and Apprendi for the plurality in Sattazahn v. Pennsylvania, 537 U. S. 101 (2003). Apprendi decided that the fact used as an enhancement in that case was "an 'element' of an offense for purposes of the Sixth Amendment's jury trial guarantee . . . no matter how the State labels it . . . ." Id., at 111. In other words, for Sixth Amendment purposes, Apprendi was convicted of the offense of possession of a firearm for an unlawful purpose plus bias-motivated intimidation, of which unlawful-purpose possession was a lesser included offense. See also Apprendi, 530 U. S., at 491-492; id., at 501 (Thomas, J., concurring) ("the core crime and the aggravating facts together constitute an aggravated crime"). In exactly the same way, "for purposes of the Sixth Amendment's jury-trial guarantee, the underlying offense of 'murder' is a distinct, lesser included offense of 'murder plus one or more aggravating circumstances . . . .' " Sattazahn, supra, at 111. For both Apprendi and Ring, the distinguishing feature that made the fact an element of an offense for Sixth Amendment purposes was its effect of exposing the defendant to a sentence greater than the maximum the sentencer could legally impose based on the elements of the underlying offense alone. See ibid.; Apprendi, 530 U. S., at 490.
Curiously, the Ninth Circuit majority relies on this passage of Sattazahn and Justice Thomas's concurrence in Apprendi for the proposition that Ring is substantive, without any explanation of why the same authorities do not also render Apprendi substantive. See Summerlin, 341 F. 3d, at 1105-1106. As noted earlier, the procedural status of Apprendi is conclusively established. See supra, at 4. Despite the most vigorous effort the Ninth Circuit majority can muster, there is no escape from the logic of the simple syllogism:
Apprendi is procedural.
Ring has the same character as Apprendi.
Therefore, Ring is procedural.
A similar syllogism should be sufficient to dispose of the alternative ground of the Ninth Circuit's decision, i.e., that the rule Ring v. Arizona, 536 U. S. 584 (2002) qualifies for the second exception to the nonretroactivity rule of Teague v. Lane, 489 U. S. 288 (1989), see Summerlin v. Stewart, 341 F. 3d 1082, 1121 (CA9 2003) (en banc). The federal courts of appeals and state supreme courts are in universal agreement that Apprendi v. New Jersey, 530 U. S. 466 (2000) is not retroactive. See People v. De La Paz, 791 N. E. 2d 489, 497, and nn. 1 and 2 (Ill. 2003), cert. denied, De La Paz v. Illinois, 157 L. Ed. 2d 221, 124 S. Ct. 320 (2003) (collecting state and federal cases and disapproving contrary intermediate appellate court holding); see also supra, at 4-5. A more detailed look simply confirms this conclusion.
In contrast to the straightforward first exception, the second exception to the rule of Teague has been a font of litigation and confusion from the day it was announced. This is unfortunate, unnecessary, and contrary to the purpose of Teague itself. See Teague, 489 U. S., at 305 (noting disparity of treatment from conflicting decisions). In our brief in Beard v. Banks, No. 02-1603, (5) amicus CJLF suggests some clarifications. The simplest, though, would be to simply recognize what the Court suggested in Tyler v. Cain, 533 U. S. 656, 667, n. 7 (2001), "it is unlikely that any of these watershed rules has yet to emerge." (Internal quotation marks and brackets omitted; emphasis added.) The farther we go down the road of ever-expanding rights for criminal defendants, the smaller becomes the likelihood of discovering any previously unrecognized right that is truly fundamental. The probability is asymptotically approaching zero.
The requirements of this exceedingly narrow exception are well established, if not well understood. In his original proposal for the rule that is now Teague, Justice Harlan suggested an exception for cases in which a person "has been incarcerated under a procedure which creates an impermissibly large risk that the innocent will be convicted." Desist v. United States, 394 U. S. 244, 262 (1969) (dissenting opinion). Later, Justice Harlan reconsidered, in part because he found "inherently intractable the purported distinction" on which this exception was based. Mackey v. United States, 401 U. S. 667, 695 (1971) (opinion concurring in part and dissenting in part). Instead, he proposed that retroactivity on habeas be limited to those rules that met the pre-incorporation standard of due process under Palko v. Connecticut, 302 U. S. 319, 325 (1937), overruled, Benton v. Maryland, 395 U. S. 784, 794 (1969). Mackey, supra, at 693. Teague adopted an even narrower view of the exception. For a new rule to be retroactive on habeas corpus, i.e., to collaterally attack judgments that were correct under the law in effect when they became "final," the rule must meet both tests. 489 U. S., at 312. Ring meets neither.
The course of this Court's retroactivity jurisprudence has not been simple, to put it mildly. See Mackey, 401 U. S., at 676 ("as difficult to follow as the tracks made by a beast of prey in search of its intended victim"). There has, however, been one constant throughout the changes. Every expansion of the Sixth Amendment right to jury trial has been held nonretroactive on collateral review, regardless of the standard used. The only jury case which this Court has applied retroactively on habeas was a due process case dealing with neutrality in jury selection.
Before Teague, retroactivity on habeas was governed by the three-factor test of Stovall v. Denno, 388 U. S. 293, 297 (1967). The first and most important factor was the purpose of the rule. Correction of a practice that "raises serious questions about the accuracy of guilty verdicts in past trials," Williams v. United States, 401 U. S. 646, 653 (1971) (plurality opinion), was the purpose which called most strongly for retroactivity. Such a purpose would trump the other two factors. See ibid. Justice Marshall, agreeing in part with Justice Harlan's approach, said, "if the purpose of a new rule implicates decisively the basic truth-determining function of the criminal trial, then I believe the rule should be given full retroactive application . . . ." Id., at 666 (opinion concurring in part and dissenting in part) (emphasis added).
In essence, Justice Harlan's Desist proposal identified the rules which had the strongest claim to retroactivity under the Stovall test and limited the second exception to those rules. It follows, then, that any rule which did not qualify for retroactivity under the Stovall regime would also not qualify for the accuracy-enhancing prong of the second Teague exception.
De Stefano v. Woods, 392 U. S. 631 (1968) (per curiam) demonstrates that the application of the Stovall purpose test to jury trial was not even a close question. Until Duncan v. Louisiana, 391 U. S. 145 (1968), decided a month earlier, the Sixth Amendment jury trial right had not been held to apply to the States at all, and Louisiana did not provide one unless the defendant faced death or hard labor. See id., at 146. After Duncan, the Court granted certiorari in De Stefano and affirmed the denial of the Sixth Amendment claim without briefing or argument. Denial of jury trial alone does not render a trial unfair. De Stefano, supra, at 633-634. "The values implemented by the right to jury trial would not measurably be served by requiring retrial of all persons convicted in the past by procedures not consistent with the Sixth Amendment right to jury trial." Id., at 634 (emphasis added). Not only does jury trial not "decisively" implicate reliability, to use Justice Marshall's term, the impact is not even measurable.
Although four Justices did not join the decision, not a single one expressed disagreement with the Court's application of the Stovall test. Justices Harlan and Stewart reiterated their disagreement with Duncan, see id., at 635, while Justices Douglas and Black reiterated their disagreement with the whole concept of nonretroactivity. See ibid. Given both Duncan and Stovall, no one thought the question of applying Stovall to Duncan to yield nonretroactivity was a question that even warranted briefing and argument.
The only major jury case to be given full retroactivity on habeas corpus is Witherspoon v. Illinois, 391 U. S. 510 (1968), decided between Duncan and De Stefano. However, Witherspoon is a due process case, not a Sixth Amendment case. See id., at 523. Although the opinion mentions the Sixth Amendment in passing, see id., at 518, a reversal on that basis could not be reconciled with the nearly contemporaneous holding in De Stefano that Duncan's incorporation of the jury trial right in the Fourteenth Amendment, making it applicable to the States, was not retroactive. The issue in Witherspoon was not jury versus judge but rather the long-established due process right to a neutral trier of fact. See id., at 521-522.
The next major expansion of the Sixth Amendment jury trial right came in Taylor v. Louisiana, 419 U. S. 522 (1975), which constitutionalized the requirement that the jury be drawn from a cross-section of the community. See id., at 528; cf. id., at 538-540 (Rehnquist, J., dissenting) (discussing prior contrary authority). As with Duncan, the application of the Stovall test was so straightforward as to not require argument. Daniel v. Louisiana, 420 U. S. 31 (1975) (per curiam) summarily affirmed a conviction obtained contrary to the new rule of Taylor. The cross-section requirement did not implicate the reliability of the result to a degree that would require retrial of cases which had been tried in reliance on the authorities noted in Justice Rehnquist's dissent. See Id., at 32-33. Again, there was no dissent from the conclusion that jury trial expansion did not meet Stovall's criteria for nonretroactivity. Justice Douglas reiterated his dissent from Stovall itself. See id., at 33-34 (dissenting opinion).
Burch v. Louisiana, 441 U. S. 130, 138 (1979), requiring six-member juries to be unanimous, was the first case to have its retroactivity decided along the lines of today's rules of Teague and Griffith v. Kentucky, 479 U. S. 314, 328 (1987). Brown v. Louisiana, 447 U. S. 323 (1980) addressed the retroactivity of Burch but did not produce a majority opinion. The holding is that of the opinion concurring on the narrowest grounds. See Marks v. United States, 430 U. S. 188, 193 (1977). That is the opinion of Justice Powell, limiting retroactivity of Burch to cases then pending on direct review. Brown, supra, at 337. (6)
Although it was not decided on Sixth Amendment grounds, the treatment of Batson v. Kentucky, 476 U. S. 79 (1986) sheds light on the retroactivity of expansion of constitutional jury rules. Batson had the strongest claim to retroactivity of any jury case since Witherspoon. Its prohibition of racially biased peremptory challenges, see id., at 89, was closer to Witherspoon's concern with a neutral decisionmaker than to the issues of judge versus jury in Duncan. The exclusion of black jurors from the trial of a black defendant has far greater implications for the fairness of the trial and reliability of the result than does the exclusion of women from the trial of a man at issue in Taylor. See Taylor, 419 U. S., at 539-540 (Rehnquist, J., dissenting); Allen v. Hardy, 478 U. S. 255, 263-264 (1986) (Marshall, J., dissenting).
Even so, a majority of the Court considered the retroactivity of Batson on habeas under the Stovall test to be straightforward enough to resolve, once again, in a summary per curiam opinion. Batson "may have some bearing on the truthfinding function," Allen, 478 U. S., at 259, but that was not enough to apply retroactively on habeas a decision overruling a Supreme Court precedent on point. It did not have "such a fundamental impact on the integrity of factfinding as to compel retroactive application." Ibid. Griffith, 479 U. S., at 326-328 applied Batson retroactively on direct review only by abandoning Stovall and adopting Justice Harlan's position of full retroactivity on direct review for all new constitutional rules.
A long and consistent line of precedent conclusively establishes that the right to have a fact determined by a jury rather than a judge was not so central to the accuracy of the trial as to require retroactivity under the test in effect before Teague. The accuracy-enhancing prong of the second Teague exception is a more restricted version of the purpose prong of Stovall, and hence the further expansion of the jury trial right in Apprendi and Ring necessarily fails to qualify.
The Ninth Circuit majority opinion in the present case seeks to avoid this result with a long and mostly irrelevant discussion of the purported advantages of jury sentencing versus judge sentencing, and with criticism of the informality of Arizona penalty-phase proceedings. See 341 F. 3d, at 1109-1116. Tellingly, the opinion relies on Justice Breyer's opinion concurring in the judgment in Ring. See id., at 1115. Justice Breyer's opinion was based on his belief that the Eighth Amendment, not the Sixth, requires the jury to determine the sentence, not just eligibility. See Ring, 536 U. S., at 613-614. Whatever the intrinsic merits of this belief, it is not the law. Proffitt v. Florida, 428 U. S. 242 (1976) is the law, and it unequivocally rejected the proposition "that jury sentencing is constitutionally required." Id., at 252 (lead opinion); see also id., at 260-261 (White, J., concurring in the judgment) (Florida procedures were constitutional).
Ring "has nothing to do with jury sentencing. What [Ring] says is that the jury must find the existence of the fact that an aggravating factor existed." Ring, 536 U. S., at 612 (Scalia, J., concurring) (emphasis in original); see also id., at 597, n. 4 (opinion of the Court) ("claim is tightly delineated," no argument on "ultimate determination"). The Ninth Circuit claims accuracy-enhancing status for Ring because of the "role of a jury in capital cases to make the important moral decisions inherent in rendering a capital verdict." 341 F. 3d, at 1113 (emphasis added). The "moral judgment" decision in capital sentencing, though, is the selection decision, not the eligibility decision, see supra, at 10, and Ring does not require a jury to make the selection decision. "Those States that leave the ultimate life-or-death decision to the judge may continue to do so . . . ." Ring, supra, at 612 (Scalia, J., concurring).
This Court has repeatedly explained "the critical difference in the two functions" of the eligibility finding and the sentencing decision. Shafer v. South Carolina, 532 U. S. 36, 50 (2001). Justice Kennedy explained it in his opinion for the Court in Tuilaepa v. California, 512 U. S. 967 (1994), see supra, at 9, and Justice Ginsburg explained it again in Shafer. "[T]he existence of a statutory aggravator [is] a tightly circumscribed factual inquiry . . . . The jury, as aggravating circumstance factfinder, exercises no sentencing discretion itself." Shafer, supra, at 51. The "moral judgment" aspect of the process comes in when the sentencer decides "whether to impose the death penalty . . . ." Ibid. In other words, it comes in when the decision is made on the "ultimate determination" that Ring unequivocally stated was not at issue in that case.
"If there is any place where adherence to evidentiary rules, constitutional restraints, and the defendant's confrontation rights is paramount," the Ninth Circuit majority declares, "it must be when the defendant is exposed to the penalty of death." 341 F. 3d, at 1113. This is a stunningly irrelevant statement. Nothing in Ring even remotely suggests that the formal rules of evidence at trial are constitutionally required for the penalty phase, and the contrary rule is thoroughly established in precedent. See, e.g., Romano v. Oklahoma, 512 U. S. 1, 7-8 (1994). Of course, the defendant's confrontation rights must be respected, as well as more general fairness in refuting the evidence against him, but that rule is Gardner v. Florida, 430 U. S. 349 (1977), not Ring. Ring adds little or nothing on this point to the pre-existing protections of the Gardner line of cases, and it is only the increments over pre-existing protections that matter in Teague analysis. See Sawyer v. Smith, 497 U. S. 227, 244 (1990). As for the blanket term "constitutional restraints," this Court rejected 15 years ago the notion that Teague would not apply to the penalty phase of capital cases, see Penry v. Lynaugh, 492 U. S. 302, 313-314 (1989), and Congress's implicit endorsement of the Teague line in the Antiterrorism and Effective Death Penalty Act of 1996 precludes reopening that question. See Williams v. Taylor, 529 U. S. 362, 412 (2000); id., at 382 (opinion of Stevens, J.) (rules that are "new" for Teague are not "clearly established" for 28 U. S. C. §2254(d)(1)). (7)
Ring requires that the factor which makes the convicted murderer eligible for the death penalty be found by a jury, but it does not increase the reliability of that determination significantly, if at all. Its rule is based on other considerations. See Ring, 536 U. S., at 607. The rule fails the Desist prong of the second Teague exception. Ring is not retroactive on habeas corpus.
The second (Mackey) prong of the second exception has been variously stated, but the phrase "absolute prerequisite to fundamental fairness," Teague, 489 U. S., at 314, will suffice here. Ring unambiguously disclaims fairness as the basis of its holding. See 536 U. S., at 607. Justice Cardozo's opinion in Palko v. Connecticut, 302 U. S. 319 (1937), invoked by Justice Harlan in Mackey, expressly names jury trial as one of the rights that is "not of the very essence of a scheme of ordered liberty." See id., at 325 (emphasis added). "Few would be so narrow or provincial as to maintain that a fair and enlightened system of justice would be impossible without" jury trial. Ibid.
Any contention that jury trial is an "absolute prerequisite to fundamental fairness" simply cannot be reconciled with the fact that jury trial is not required in juvenile proceedings. In McKeiver v. Pennsylvania, 403 U. S. 528 (1971), the plurality noted that jury trial is neither "a necessary component of accurate factfinding," id., at 543, nor "a necessary part of every criminal process that is fair and equitable." Id., at 547. In other words, it passes neither requirement for the second Teague exception. Justice Brennan, concurring in the judgment in one of the joined cases and dissenting in the other, said "the question in these cases is whether jury trial is among 'the essentials of due process and fair treatment,' In re Gault, 387 U. S. 1, 30 (1967) . . . ." Id., at 553. His answer was no, provided other safeguards were observed. Id., at 554-556.
Similarly instructive is the more recent decision in Johnson v. United States, 520 U. S. 461 (1997). Johnson's claim was essentially identical to Summerlin's. An element of the offense of perjury--materiality--was determined by the judge rather than the jury, in accordance with the case law at the time of trial. See id., at 464; cf. Ring, 536 U. S., at 609. Because Johnson had not objected, her appeal on that point was subject to the "plain error" rule. This Court decided unanimously that "there is no basis for concluding that the error 'seriously affect[ed] the fairness, integrity or public reputation of judicial proceedings.' " 520 U. S., at 470.
How could it be seriously maintained that nonobservance of an "absolute prerequisite to fundamental fairness" could ever fail to "seriously affect the fairness . . . of judicial proceedings"? "Absolute prerequisite" is a far more stringent standard than "seriously affect." This case is different from Tyler v. Cain, 533 U. S. 656, 666-667 (2001), where the second Teague exception and the structural error rule, although overlapping, were different in ways such that neither mandates the other. The second Teague exception and the fourth prong of the plain error test both focus on fairness, but Teague does so in much more restrictive terms. See O'Dell v. Netherland, 521 U. S. 151, 167, n. 4 (1997) ("inevitable that . . . 'miscarriage[s] of justice' will occur"). A rule which implicates the fairness of the trial so minimally as to fail that fourth prong is not remotely close to qualifying for Teague.
To qualify for the second exception, Ring would have to meet both the Desist and Mackey criteria. It meets neither. Ring is not retroactive on habeas corpus.
The single greatest difficulty in the modern era of capital punishment has been the constantly changing state of the case law. Practices expressly approved one year are struck down in another, and new, previously unknown restrictions are discovered lurking in the Constitution, seemingly without end. In his dissent in Morgan v. Illinois, 504 U. S. 719, 751 (1992), Justice Scalia referred to "the fog of confusion that is our annually improvised Eighth Amendment, 'death is different' jurisprudence . . . ." The price for this constant churning is not just in judgments, finality, and the cost of retrials, as important as those are. The cost may very well include the lives of innocent people.
In the past several years, there has been a sea change in the scholarly literature of the deterrent effect of capital punishment. Although not unanimous, the preponderance of recent studies indicates that a death penalty which is actually enforced saves innocent lives in large numbers. Conversely, the obstruction of the execution of deserved sentences of guilty murderers for reasons having no bearing on the reliability of the guilt verdict diminishes the deterrent effect. Put more bluntly, retroactivity kills innocent people.
One of the reasons for the change is simply the availability of significant amounts of data from the period after Gregg v. Georgia, 428 U. S. 153 (1976). The early, controversial work of Isaac Ehrlich showing deterrence and the contrary results of his critics were all obtained with data from a system of capital punishment which no longer exists. See H. Dezhbakhsh & J. Shepherd, The Deterrent Effect of Capital Punishment: Evidence from a "Judicial Experiment," Emory U. Econ. Working Paper 03-14, pp. 6-7 (July 2003) (available at http://people.clemson.edu/~jshephe/CaPuJLE_submit.pdf) (summarizing early studies) (cited below as "Judicial Experiment"). Furman v. Georgia, 408 U. S. 238 (1972) declared this system unconstitutional in part because the large number of potentially capital offenses and the standardless imposition of the death penalty on a very small portion of those eligible could not be expected to produce a deterrent effect. See id., at 312 (White, J., concurring).
Yunker, A New Statistical Analysis of Capital Punishment Incorporating U.S. Postmoratorium Data, 82 Social Science Q. 297 (2001) showed the effect of this change by using the methodology of earlier studies on data before and after Gregg without attempting any other improvements. "Whatever advance is achieved over the many previous studies of the question comes simply from the incorporation of additional data." Id., at 298. The additional data alone were sufficient to show a deterrent effect "statistically significant at a high level of confidence (approximately 95 percent)." Id., at 310.
Along with the post-Furman reforms in the death penalty system, more recent studies also have available better data from improved collection systems. Mocan & Gittings, Getting Off Death Row: Commuted Sentences and the Deterrent Effect of Capital Punishment, 46 J. L. & Econ. 453, 455-456 (2003) discuss some of the difficulties with past data sets and the improvements now available.
The third change in recent years is that the science of econometrics, like most sciences, moves forward. Dezhbakhsh, Rubin, & Shepherd, Does Capital Punishment Have a Deterrent Effect? New Evidence from Postmoratorium Panel Data, 5 Am. L. & Econ. Rev. 344, 352-361 (2003) discuss improved data and methods, including adjustments in response to criticisms of earlier studies. They conclude, "our most conservative estimate is that the execution of each offender seems to save, on average, the lives of eighteen potential victims. (This estimate has a margin of error of plus and minus ten)." Id., at 373. Mocan and Gittings estimate a smaller but still very substantial deterrent effect of five innocent lives saved for each murderer executed. See 46 J. L. & Econ., at 474.
Three studies not yet published are worth noting here. P. Zimmerman, State Executions, Deterrence, and the Incidence of Murder, Soc. Sci. Res. Network (March 2003) (unpublished manuscript) (available at http://ssrn.com/abstract=354680), estimates 14 murders deterred per execution. Z. Liu, Capital Punishment and the Deterrence Hypothesis: Some New Insights and Empirical Evidence, E. Econ. J. (forthcoming) (manuscript available at http://ssrn.com/abstract=352681), refutes the common, simplistic argument that the mere comparison of murder rates in states that do and do not have capital punishment disproves the deterrent effect. J. Shepherd, Murders of Passion, Execution Delays, and the Deterrence of Capital Punishment, Working Paper (Aug. 2003) (available at http://people.clemson.edu/~jshephe/DPpaper_fin.pdf), finds that long delays on appeal cost innocents lives by reducing the deterrent effect.
One of the difficulties in this field of study is the inability to do controlled experiments to confirm or refute hypotheses in the way that can be done in the physical sciences. However, some studies have been done on the "quasi-experiments" created by suspensions of capital punishment. Cloninger & Marchesini, Execution and Deterrence: A Quasicontrolled Group Experiment, 33 Applied Econ. 569, 575 (2001), found that over 200 lives were lost as the result of a one-year suspension in Texas, although Sorensen, et al., Capital Punishment and Deterrence: Examining the Effect of Executions on Murder in Texas, 45 Crime & Delinq. 481, 489-491 (1999), found no deterrent effect from an examination of the same suspension. (8) Dezhbakhsh & Shepherd, "Judicial Experiment," supra, studied the Furman moratorium as well as individual states' suspension and reimposition of the death penalty. They found that suspension or abolition of the death penalty increases murder rates, and reinstatement lowers them.
The Mocan and Gittings study is the most directly pertinent to the present case. In addition to the usual study of the effect of executions, they studied specifically the effect of removals from death row, i.e., commutations and appellate reversals of sentences. Their stark result is that removals kill people. "[O]ne additional removal from death row generates about one additional homicide." 46 J. L. & Econ., at 474.
We do not, of course, claim these studies as conclusive proof. The difficulty, if not impossibility, of definitively proving either side of this debate has long been known. See, e.g., Furman, 408 U. S., at 395-396 (Burger, C. J., dissenting) ("empirical stalemate"). Even so, there is now a formidable body of empirical evidence to confirm what common sense has always told us--the death penalty will deter murders and save lives if it is actually enforced, and needless obstruction of it will cost lives. Any proposal to wipe out large numbers of sentences validly imposed under the law in effect at the time of trial must be evaluated in light of the very real possibility that innocent people will die as a result.
Weighed against this grave possibility, we must ask what the likelihood is that Summerlin's trial, or the trial of the others similarly situated, would have come out any differently if he had been given a jury trial on eligibility. The trial judge in this case found two aggravating circumstances, either of which would be sufficient for eligibility: A prior violent felony conviction and commission of the present crime in an especially heinous, cruel, or depraved manner. See 341 F. 3d, at 1090. Nothing short of jury nullification could have prevented a finding that the prior conviction circumstance was true. As for the "especially heinous" circumstance, lay jurors are less likely to be jaded and more likely to be shocked by a brutal murder than is a criminal court judge who sees depravity all the time. It would be an exceedingly rare case where having the jury rather than the judge make the eligibility finding would produce a result more favorable to the defendant. (9)
Arizona and the other States had a massive reliance interest in Walton v. Arizona, 497 U. S. 639 (1990), a precedent of this Court squarely on point that held their systems were valid. That reliance has already been grievously betrayed by the reversal of the cases that were pending on appeal when Ring was decided. In the absence of any substantial likelihood that the procedure complained of made any difference in the outcomes of these cases, there is no justification for aggravating the damage by overturning final judgments. There is good reason to believe that innocent people's lives depend on the enforcement of those judgments.
The judgment of the Court of Appeals for the Ninth Circuit should be reversed.
January, 2004
Respectfully submitted,
Kent S. Scheidegger
Attorney for Amicus Curiae
Criminal Justice Legal Foundation
1. This brief was written entirely by counsel for amicus, as listed on the cover, and not by counsel for any party. No outside contributions
were made to the preparation or submission of this brief.
Both parties have given written consent to the filing of this brief.
2. Attempting to distinguish Ring from Apprendi, the opinion says in this footnote, "Ring, conversely, did reach the relevant substantive
basis. See 536 U. S., at 589-90." The cited passage is the recitation of the facts and says nothing that even conceivably supports the
proposition for which it is cited.
3. 28 U. S. C. §2254(d), as added by the Antiterrorism and Effective Death Penalty Act of 1996, has no such limitation. See Lockyer v.
Andrade, 538 U. S. 63, 155 L. Ed. 2d 144, 158, 123 S. Ct. 1166, 1175 (2003) (applying AEDPA to a substantive Eighth Amendment
claim). However, this section of AEDPA does not apply to the present case. See 341 F. 3d, at 1092.
4. For a more thorough review than space permits here, see Oken v. State, 2003 Md. LEXIS 750 (Nov. 17, 2003), reh'g denied, Dec. 18,
2003.
5. This brief is available on our Web site at http://www.cjlf.org/ briefs/BanksMerits.pdf.
6. At this point, Justice Powell had adopted the view of retroactivity previously proposed by Justice Harlan, see Hankerson v. North
Carolina, 432 U. S. 233, 246-248 (1977) (Powell, J., concurring in the judgment), and subsequently adopted in Griffith and Teague.
7. This disposes of Judge Reinhardt's concurring opinion, which is little more than a call to overrule Penry and apply all new
constitutional rules with full retroactivity on habeas corpus in capital cases. See 341 F. 3d, at 1122-1125.
8. This is the only peer-reviewed, direct study cited by Justice Breyer in Ring v. Arizona, 536 U. S. 584, 615 (2002) (opinion concurring
in the judgment). See Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U. S. 579, 593-594 (1993) (importance of peer-review
publication process). The other items cited, a newspaper article and a survey, are entitled to little, if any, weight.
9. There is at least anecdotal evidence that Arizona's decision to switch to the jury for both the eligibility and selection decisions has
increased the number of death sentences. See Greenberger, Death-Penalty Law Backfires, Wall Street Journal, Aug. 6, 2003, p. A4, col.
1.