138 F.3d 552
Elmo HUMPHREY, III, Petitioner-Appellant,
Burl CAIN, Warden, Louisiana State Penitentiary, Respondent-Appellee.
United States Court of Appeals,
April 3, 1998.
David John Schenck, Hughes & Luce, Dallas, TX, for Petitioner-Appellant.
Alison Wallis, Gretna, LA, for Respondent-Appellee.
Appeal from the United States District Court for the Eastern District of Louisiana.
Before POLITZ, Chief Judge, and KING, JOLLY, HIGGINBOTHAM, DAVIS, JONES, SMITH, DUHE, WIENER, BARKSDALE, EMILIO M. GARZA, DeMOSS, BENAVIDES, STEWART, PARKER and DENNIS, Circuit Judges.
PATRICK E. HIGGINBOTHAM, Circuit Judge:
We took this case en banc to resolve conflicting panel decisions whether Cage-Victor error fits within the second Teague exception, making it available in federal habeas to a state prisoner whose criminal conviction was final when those cases were decided. For the reasons stated in the panel opinion, we today give an affirmative answer to that question. We overrule all prior decisions of this court holding to the contrary. For the reasons stated in the panel opinion,1 we hold that the jury instructions defining reasonable doubt lowered the State's burden of proof below the constitutional minimum. We reverse the district court's judgment denying the writ of habeas corpus and remand the case to the district court with instructions to order the State of Louisiana to either try Elmo Humphrey, III, again or release him from custody within 90 days of the date of the district court's order on remand.
REVERSED and REMANDED with instructions.
EDITH H. JONES, Circuit Judge, with whom JOLLY, DUHE and PARKER, Circuit Judges, join in dissenting:
My colleagues have constitutionally condemned what is surely one of the longest reasonable doubt instructions in recent reported history based on four catch phrases, even though three of those phrases have, in proper context, survived Supreme Court scrutiny. What provoked the majority to ignore the rich forest of the whole instruction for the sake of a few sickly trees is not clear. But their analytical method ignores that we must scrutinize the instructions taken as a whole, not in isolated tidbits. And their emphasis on the court's statement that reasonable doubt "is a serious doubt for which you could give good reason" is misguided. I do not believe Humphrey's jury convicted him of repeated sexual molestation of a then-nine-year old victim on a constitutionally infirm standard. I respectfully dissent.1
On habeas review, "our inquiry is whether the instruction is constitutional, not whether it is exemplary." Vargas v. Keane, 86 F.3d 1273, 1280 (2d Cir.1996). The test of constitutionality is whether taken as a whole, the instruction correctly conveyed the law of reasonable doubt to the jury. See Victor v. Nebraska, 511 U.S. 1, 5, 114 S.Ct. 1239, 1243, 127 L.Ed.2d 583 (1994). Further, a court must determine not whether the instruction could have been applied unconstitutionally, but "whether there is a reasonable likelihood that the jury understood the instructions to allow conviction based on proof insufficient to meet the Winship standard." Id. at 6, 114 S.Ct. at 1243. As the Supreme Court itself pointed out, it has found in only one case that a reasonable doubt definition actually violated the due process clause. Id. at 5, 114 S.Ct. at 1243. (citing Cage v. Louisiana, 498 U.S. 39, 111 S.Ct. 328, 112 L.Ed.2d 339 (1990)). We do not minimize the importance of furnishing a correct jury instruction on reasonable doubt, but neither will we trivialize the importance of carefully reviewing jury instructions as a whole before declaring that the jury was reasonably likely to have interpreted them unconstitutionally.
Two general vices may render a reasonable doubt instruction unconstitutional: that it urges the jury to convict on a standard of proof lower than "beyond a reasonable doubt," or that it permits the jury to go outside the record evidence to convict. See id. at 5-6, 22, 114 S.Ct. at 1243, 1251. Mindful of these pitfalls, we may review the Humphrey reasonable doubt instruction, with each sentence numbered for the convenience of this discussion. The Humphrey jury was instructed:
(1) The burden is upon the State to prove the defendant's guilt beyond a reasonable doubt. (2) In considering the evidence, you must give the defendant the benefit of every reasonable doubt arising out of the evidence or out of the lack of evidence. (3) If you are not convinced of the guilt of the defendant beyond a reasonable doubt, you must find him not guilty.
(4) If you entertain any reasonable doubt as to any fact or element necessary to constitute the defendant's guilt, it is your sworn duty to give him the benefit of that doubt and return a verdict of acquittal. (5) Even where the evidence demonstrates a probability of guilt, yet if it does not establish it beyond a reasonable doubt, you must acquit the accused. (6) This doubt must be a reasonable one, that is, one found[ed] upon a real, tangible, substantial basis, and not upon mere caprice, fancy or conjecture. (7) It must be such a doubt [as] would give rise to a grave uncertainty, raised in your minds by reason of the unsatisfactory character of the evidence; one that would make you feel that you had not an abiding conviction to a moral certain[ty] of the defendant's guilt. (8) If, after giving a fair and impartial consideration to all of the facts in the case[,] you find the evidence unsatisfactory upon any single point indispensably necessary to constitute the defendant's guilt, this would give rise to such a reasonable doubt as would justify you in rendering a verdict of not guilty.
(9) The prosecution must establish guilt by legal and sufficient evidence beyond a reasonable doubt, but the rule does not go further and require a preponderance of testimony. (10) It is incumbent upon the State to prove the offense charged, or legally included in the indictment, to your satisfaction and beyond a reasonable doubt. (11) A reasonable doubt is not a mere possible doubt. (12) It should be an actual or substantial doubt. (13) It is such a doubt as a reasonable man would seriously entertain. (14) It is a serious doubt for which you could give good reason.
The first six sentences pose no constitutional problems. They properly articulate the reasonable doubt standard, and they provide context for the rest of the instruction. Sentences (8)-(10) provide further context and background for every challenged sentence in the instruction. They make clear that the burden of proof rests with the state for each element of the offense charged and that a reasonable doubt must be based upon the evidence. Along with sentences (1)-(6), they mitigate any potential error caused by the challenged language.
Sentence (7) contains the first phrases to which the appellant objects: "grave uncertainty" and "moral certainty." The Supreme Court has held that use of the term "grave uncertainty," while not condoned, presents no constitutional infirmity if it relates to the existence rather than the magnitude of the evidence required. See id. at 20, 114 S.Ct. at 1250. Sentence (7), especially when taken in the context of the preceding two sentences, is rooted in making a comparison between doubts based on the existence (or nonexistence) of evidence versus doubts based on conjecture or fancy. In addition, the Court has held that even if the use of "grave uncertainty" touches upon the magnitude of the evidence, this error can be corrected if the instruction contains alternative, mitigating definitions of reasonable doubt. See id. In this case, sentences (2)-(5) and (8)-(10) serve to make clear that the defendant is to be given the benefit of every reasonable doubt arising from the evidence. Sentence (8) unequivocally requires a not-guilty verdict based upon a reasonable doubt if the evidence is "unsatisfactory upon any single point indispensably necessary" to convict. This sentence makes plain, as the Cage instruction did not, that reasonable doubt arises from any material deficiency in proof by the state.
The Court has also held that use of the term "moral certainty" is not constitutionally improper as long the rest of the instruction gives context to its meaning. See id. at 14-16, 21-22, 114 S.Ct. at 1247-48, 1250-51. The Humphrey instruction as a whole confirms that "moral certainty" is not a vacuous term, but relates to the evidence in the case. As in Victor, other sentences admonished the jury not to be swayed by passion or prejudice, to rule solely on the evidence, and not to indulge in speculation or conjecture. See id. at 16, 114 S.Ct. at 1248. The jury was also required to have an "abiding conviction" of guilt to a moral certainty. In Victor, the Court viewed this phrase as confirming a "subjective state of near certitude" of the defendant's guilt. Id. at 15, 114 S.Ct. at 1247 (internal citation omitted). Thus, any possible misunderstanding of "moral certainty" was fully offset by its context. It is highly unlikely that the jury either misunderstood the burden of proof or strayed beyond the record because of this phrase.
Sentence (12) contains the third phrase condemned by the majority, "actual or substantial doubt." As with "grave uncertainty," the use of the phrase "actual or substantial doubt" is, in the context of the full instruction, tied to the existence rather than the magnitude of the evidence. This is particularly evident from the earlier reference to a "substantial basis" for reasonable doubt in sentence (6), which is juxtaposed to a doubt founded in "caprice, fancy or conjecture." The Supreme Court expressly held in Victor that any ambiguity about a "substantial doubt" was cured by its being distinguished from a mere possibility, caprice, or conjecture. See id. at 20, 114 S.Ct. at 1250.
If these three phrases were the only ones involved in this case, the instruction would be upheld under Victor. It would be upheld because the phrases are properly explained by their surrounding passages to enforce the state's high burden of proof beyond a reasonable doubt and to tie conviction only to the evidence or lack of evidence in the record. According to the majority, however, these phrases somehow lose the benefit of their mitigating context and are irretrievably misleading because they are conjoined with the last sentence in the instruction: "It is a serious doubt for which you could give good reason."
The majority's conclusion embodies three criticisms: of the term "good reason," of the "articulation-demanding" quality of the instruction, and of the alleged synergy with the other three undesirable phrases. On analysis, these criticisms, while aesthetically accurate, are constitutionally suspect.
First, the use of the phrase "good reason" in sentence (14) only repeats what the entire instruction has already conveyed to the jury. A reasonable doubt is not one based upon speculation or mere possibility. A "good reason" refers to the actuality of doubt arising from the evidence. Like other courts, see, e.g., Beverly v. Walker, 118 F.3d 900, 903 (2d Cir.1997); Chalmers v. Mitchell, 73 F.3d 1262, 1268 (2d Cir.1996), we do not condone the use of this phrase, but to say that it renders the instruction unconstitutional is to read the tea leaves more closely than is reasonably possible. The instruction as a whole emphasizes that a "reasonable doubt" is the type of doubt required for acquittal, that the benefit of every doubt is to be given to the defendant, and that probability of guilt is not sufficient to convict.
The majority is able to cite only one case in which a sister circuit has held that a reasonable doubt instruction containing the phrase "good reason" is unconstitutional. See Dunn v. Perrin, 570 F.2d 21 (1st Cir.1978).2 The specific offending language in Dunn was as follows: "It is such a doubt as for the existence of which a reasonable person can give or suggest a good and sufficient reason." Id. at 23 n. 1 (emphasis added). To require "good and sufficient reason" clearly ratchets the defense's burden upward. Moreover, the court expressly stated that while this language may have been improper, it was not by itself reversible error. See id. at 23. Rather, the court found that only the cumulative effect of this language plus numerous other impermissible phrases rendered the entire instruction unconstitutional. See id. at 25.3
Second, the majority condemned the phrase "for which you could give good reason" as an "articulation-demanding" standard, and it complains:
A juror favoring guilt would have a powerful tool if he could demand that undecided jurors articulate good reasons for considering an acquittal. Such a juror could use Humphrey's reasonable-doubt instruction to argue that it's not enough to have a reason to acquit--that one must have an especially strong, compelling, or persuasive reason for doubting the defendant's guilt. He could plausibly assert that a paucity of government evidence should not count as a "good" reason. Under this sort of pressure, inarticulate and undecided jurors are less likely to give defendants the benefit of their doubts. Requiring articulation of good reasons, then, skews the deliberation process in favor of the state by suggesting that those with doubts must perform certain actions in the jury room--actions that many individuals find difficult or intimidating--before they may vote to acquit.
Humphrey v. Cain, 120 F.3d 526, 531 (5th Cir.1997).4 This novel exercise in amateur psychology is not only highly speculative, it results in a reductio ad absurdum--any phrase plucked from a reasonable doubt instruction could be used to put a juror in the "uncomfortable" position of having to defend his vote of not guilty.
The flaw in the majority's analysis is illustrated by another--apparently unobjectionable--sentence from the Humphrey instruction: "This doubt must be a reasonable one, that is, one founded upon a real, tangible, substantial basis, and not upon mere caprice, fancy or conjecture." Why do these words not present the same danger as the words "good reason?" Does not a juror favoring guilt have a powerful tool if he can demand that undecided jurors articulate why their doubt is real, tangible, or substantial? Does not this phrase skew the deliberation process in favor of the state by suggesting that those with doubts must--or at least should--be able to explain why their doubts are not based upon mere caprice, fancy, or conjecture? Yet no one questions the propriety of including the sentence from which these questions arise in a constitutional reasonable doubt instruction.
Indeed, the Humphrey instruction does not require, as the majority suggests, that a juror give--or even be able to give--a "good reason." Rather, it merely equates reasonable doubt with "a serious doubt for which you could [not should or must] give good reason."
It should be significant that no other court has so strongly condemned the "doubt for which you could give a reason" instruction.5 Judge Friendly seemed sympathetic to the observation that challenges to such an instruction are "hypercritical," but he went on to agree that such an instruction is "not approved" and "perhaps unwise," yet it is "not erroneous." United States v. Davis, 328 F.2d 864, 867-68 (2d Cir.1964). More recently, the Second Circuit upheld a reasonable doubt instruction that contained the sentence, "It is a doubt for which you can give a reason if called upon to do so by a fellow juror in the jury room." Vargas, 86 F.3d at 1275. The court explained its holding as follows:
The premise of the petitioner's challenge ... is that a juror may hold a reasonable doubt of a defendant's guilt, based on the evidence or lack of evidence offered by the prosecution, and yet be unable to articulate that doubt. Even if we assume that there exists a category of "reasonable" yet not "articulable" doubt, we do not see how the addition of the phrase "if called upon to do so by a fellow juror" alters the substance of the instruction sustained in cases such as Davis and Leecan in any constitutionally significant way. The challenged language does not suggest--any more than the "for which a reason can be given" formulation standing alone--that a juror will be called upon to articulate his doubts, or that he must articulate them if so called upon. Moreover, language surrounding the challenged language renders it unlikely that jurors would interpret the passage to require a higher degree of doubt than "reasonable" doubt for acquittal. The challenged language was immediately preceded and followed by language giving additional content to the phrase "reasonable doubt," emphasizing that the jurors' task was to engage in a careful, reasoned analysis of the evidence--or lack of evidence--presented. Accordingly, the jury was unlikely to have understood the challenged language to bear upon the degree or quantum of doubt necessary for acquittal, but rather, upon the appropriate basis for the formulation of a doubt sufficient for acquittal: as the charge properly stated, a "reasonable" doubt is not "just any kind of doubt, a mere whim or a guess or a feeling, or a surmise that the accused may not be guilty," but a doubt that "arises out of the credible evidence or the lack of credible evidence." Viewed in context, the challenged language simply does not suggest that a doubt formulated within one's own mind--reasonable, but not articulable--is insufficient for acquittal. Rather, it serves to emphasize, in conjunction with surrounding language, that any doubt sufficient for acquittal must be based on a reasoned evaluation of the evidence or lack of evidence, rather than on speculation, emotion, or whim.
Vargas, 86 F.3d at 1278 (emphasis added).
The Humphrey majority attempts to distinguish Vargas, in part, on the ground that it did not require jurors to articulate a "good" reason, as did the Humphrey instruction. See Humphrey, 120 F.3d at 532. Anyone reading the Vargas instruction cannot entertain the notion that because it calls for only "a reason" rather than a "good reason," it is constitutionally less suspect. The entire Vargas instruction reads as follows:
A reasonable doubt is a doubt which you consciously have after the use of your powers of reasoning which arises out of the credible evidence or the lack of credible evidence.
It is a doubt for which you can give a reason if called upon to do so by a fellow juror in the jury room. The standard of reason must be prevailing. When you are convinced that you have no doubt in reason of the defendant[']s guilt, that is when you will be justified in rendering a verdict accordingly. Consequently, after fairly and fully considering all of the evidence and the lack of evidence and using the same power of reasoning and power of thinking that would apply and do apply to weigh the importance of matters related to the [sic] important business affairs, if you then believe that each element of the crime submitted to you[ ] has been established beyond a reasonable doubt, you would be justified in finding the defendant guilty of that crime.
Vargas, 86 F.3d at 1275. A recent Second Circuit case upheld an even stronger articulation standard:
A juror who has a reasonable doubt and asserts it ought to first be able to give that reasonable doubt a reason for it to himself, and he should be able to communicate that to his fellow jurors that reasonable doubt in the event they ask him to do so.
Beverly, 118 F.3d at 901. The court rejected the contention that this instruction shifted the burden of proof to the defendant, because the trial court repeatedly instructed the jury that the defendant was presumed innocent and that the prosecution bore the burden of proof for conviction. See id. at 903; see also Vargas, 86 F.3d at 1278-79; Chalmers, 73 F.3d at 1267-68 (criticizing a charge that defined reasonable doubt as "a doubt for which some good reason can be given," but concluding that it was not reasonably likely that the jury misunderstood which party bore the burden of proof, since the court made it clear that the defendant remained innocent until the prosecution proved his guilt beyond a reasonable doubt).
The majority's fear that inarticulate jurors will be prejudiced or intimidated by this sentence is fanciful. The sentence states no requirement of articulation, and it never mentions that word. If there were little more to this instruction than the challenged sentence, its potential for misleading the jury would have been highlighted. But the saving grace of this trial court's instructions is that they said too much, not too little, and much of what they said--eleven out of fourteen sentences on reasonable doubt alone--was entirely correct and unobjectionable.
The majority's final worry is that the "give a good reason" standard somehow magnified the potential errors in the "grave uncertainty," "moral certainty," and "substantial doubt" phrases so that together, they effectively lowered the state's burden of proof. This conclusion is an unreasoned ipse dixit and deserves no rebuttal.6 Yet it is only when all four phrases are taken in combination that the majority are willing to say this instruction is unconstitutional. The majority's conclusion, however, rests only on phrases and flatly ignores the Supreme Court's insistence that we look to the jury instructions as a whole.
Let me reiterate that I do not condone the Humphrey instruction insofar as it contains dubious and potentially misleading phrases. But paraphrasing Judge Friendly, I believe the instruction is "not approved" and "perhaps unwise," but it is not unconstitutional. I respectfully dissent.
I concur in the court's decision that the Cage/Victor standard must be retroactively applied and in the ex post facto discussion
Our research has disclosed one other, most unusual reasonable doubt instruction that a court of appeals held unconstitutional on habeas review. See Lanigan v. Maloney, 853 F.2d 40 (1st Cir.1988). The language in that charge is inapposite to this discussion
Specifically, the court found that two other parts of the instruction were "clearly wrong." Id. at 23. These parts (i) equated reasonable doubt with a "strong and abiding conviction as still remains after careful consideration of all the facts and arguments," and (ii) permitted the jury to convict or acquit on the basis of "information which we do not positively know to be absolutely true and which we have no means of verifying." Id. at 24. Thus, "good reason" became unconstitutional only when combined with other highly damaging phrases. Such is not the case with the Humphrey instruction wherein "good reason" is combined with other phrases found constitutional by the Supreme Court when used in a proper context
I note that the jury did not have the written jury instruction with them in the jury room. Rather, the instruction was read to them once by the judge before their deliberations began
See, e.g., Gilday v. Callahan, 59 F.3d 257, 264-65 (1st Cir.1995) (holding constitutional a reasonable doubt instruction containing the sentence, "When you get all through analyzing this evidence, it has to be a doubt nagging your mind, leaving you with an uncertainty of conviction to that moral certainty which you can stand up and argue in the jury room with principle and integrity and honesty to your fellow jurors."); United States v. Dale, 991 F.2d 819, 852-53 (D.C.Cir.1993) (panel including Wald, Ruth Bader Ginsburg, JJ.) (holding constitutional a reasonable doubt instruction containing the sentence, "Reasonable doubt, as that name implies, is a doubt based on reason, it is a doubt for which you can give a reason.")
Three previous cases from this court have considered instructions substantially similar to the Humphrey instruction, and all three instructions were found to be constitutional. See Schneider v. Day, 73 F.3d 610 (5th Cir.1996); Weston v. Ieyoub, 69 F.3d 73 (5th Cir.1995); Gaston v. Whitley, 67 F.3d 121 (5th Cir.1995)