Ramos v. Louisiana, 590 U.S. ___ (2020)

NOTICE: This opinion is subject to formal revision before publication in the preliminary print of the United States Reports. Readers are requested to notify the Reporter of Decisions, Supreme Court of the United States, Washington, D. C. 20543, of any typographical or other formal errors, in order that corrections may be made before the preliminary print goes to press.

SUPREME COURT OF THE UNITED STATES

_________________

No. 18–5924

_________________

EVANGELISTO RAMOS, PETITIONER v. LOUISIANA

on writ of certiorari to the court of appeal of louisiana, fourth circuit

[April 20, 2020]

Justice Gorsuch announced the judgment of the Court and delivered the opinion of the Court with respect to Parts I, II–A, III, and IV–B–1, an opinion with respect to Parts II–B, IV–B–2, and V, in which Justice Ginsburg, Justice Breyer, and Justice Sotomayor join, and an opinion with respect to Part IV–A, in which Justice Ginsburg and Justice Breyer join.

Accused of a serious crime, Evangelisto Ramos insisted on his innocence and invoked his right to a jury trial. Eventually, 10 jurors found the evidence against him persuasive. But a pair of jurors believed that the State of Louisiana had failed to prove Mr. Ramos’s guilt beyond reasonable doubt; they voted to acquit.

In 48 States and federal court, a single juror’s vote to acquit is enough to prevent a conviction. But not in Louisiana. Along with Oregon, Louisiana has long punished people based on 10-to-2 verdicts like the one here. So instead of the mistrial he would have received almost anywhere else, Mr. Ramos was sentenced to life in prison without the possibility of parole.

Why do Louisiana and Oregon allow nonunanimous convictions? Though it’s hard to say why these laws persist, their origins are clear. Louisiana first endorsed nonunanimous verdicts for serious crimes at a constitutional convention in 1898. According to one committee chairman, the avowed purpose of that convention was to “establish the supremacy of the white race,” and the resulting document included many of the trappings of the Jim Crow era: a poll tax, a combined literacy and property ownership test, and a grandfather clause that in practice exempted white residents from the most onerous of these requirements.[1]

Nor was it only the prospect of African-Americans voting that concerned the delegates. Just a week before the convention, the U. S. Senate passed a resolution calling for an investigation into whether Louisiana was systemically excluding African-Americans from juries.[2] Seeking to avoid unwanted national attention, and aware that this Court would strike down any policy of overt discrimination against African-American jurors as a violation of the Fourteenth Amendment,[3] the delegates sought to undermine African-American participation on juries in another way. With a careful eye on racial demographics, the convention delegates sculpted a “facially race-neutral” rule permitting 10-to-2 verdicts in order “to ensure that African-American juror service would be meaningless.”[4]

Adopted in the 1930s, Oregon’s rule permitting nonunanimous verdicts can be similarly traced to the rise of the Ku Klux Klan and efforts to dilute “the influence of racial, ethnic, and religious minorities on Oregon juries.”[5] In fact, no one before us contests any of this; courts in both Louisiana and Oregon have frankly acknowledged that race was a motivating factor in the adoption of their States’ respective nonunanimity rules.[6]

We took this case to decide whether the Sixth Amendment right to a jury trial—as incorporated against the States by way of the Fourteenth Amendment—requires a unanimous verdict to convict a defendant of a serious offense.[7] Louisiana insists that this Court has never definitively passed on the question and urges us to find its practice consistent with the Sixth Amendment. By contrast, the dissent doesn’t try to defend Louisiana’s law on Sixth or Fourteenth Amendment grounds; tacitly, it seems to admit that the Constitution forbids States from using nonunanimous juries. Yet, unprompted by Louisiana, the dissent suggests our precedent requires us to rule for the State anyway. What explains all this? To answer the puzzle, it’s necessary to say a bit more about the merits of the question presented, the relevant precedent, and, at last, the consequences that follow from saying what we know to be true.

I

The Sixth Amendment promises that “[i]n all criminal prosecutions, the accused shall enjoy the right to a speedy and public trial, by an impartial jury of the State and district wherein the crime shall have been committed, which district shall have been previously ascertained by law.” The Amendment goes on to preserve other rights for criminal defendants but says nothing else about what a “trial by an impartial jury” entails.

Still, the promise of a jury trial surely meant something—otherwise, there would have been no reason to write it down. Nor would it have made any sense to spell out the places from which jurors should be drawn if their powers as jurors could be freely abridged by statute. Imagine a constitution that allowed a “jury trial” to mean nothing but a single person rubberstamping convictions without hearing any evidence—but simultaneously insisting that the lone juror come from a specific judicial district “previously ascertained by law.” And if that’s not enough, imagine a constitution that included the same hollow guarantee twice—not only in the Sixth Amendment, but also in Article III.[8] No: The text and structure of the Constitution clearly suggest that the term “trial by an impartial jury” carried with it some meaning about the content and requirements of a jury trial.

One of these requirements was unanimity. Wherever we might look to determine what the term “trial by an impartial jury trial” meant at the time of the Sixth Amendment’s adoption—whether it’s the common law, state practices in the founding era, or opinions and treatises written soon afterward—the answer is unmistakable. A jury must reach a unanimous verdict in order to convict.

The requirement of juror unanimity emerged in 14th- century England and was soon accepted as a vital right protected by the common law.[9] As Blackstone explained, no person could be found guilty of a serious crime unless “the truth of every accusation . . . should . . . be confirmed by the unanimous suffrage of twelve of his equals and neighbors, indifferently chosen, and superior to all suspicion.”[10] A “ ‘verdict, taken from eleven, was no verdict’ ” at all.[11]

This same rule applied in the young American States. Six State Constitutions explicitly required unanimity.[12] Another four preserved the right to a jury trial in more general terms.[13] But the variations did not matter much; consistent with the common law, state courts appeared to regard unanimity as an essential feature of the jury trial.[14]

It was against this backdrop that James Madison drafted and the States ratified the Sixth Amendment in 1791. By that time, unanimous verdicts had been required for about 400 years.[15] If the term “trial by an impartial jury” carried any meaning at all, it surely included a requirement as long and widely accepted as unanimity.

Influential, postadoption treatises confirm this understanding. For example, in 1824, Nathan Dane reported as fact that the U. S. Constitution required unanimity in criminal jury trials for serious offenses.[16] A few years later, Justice Story explained in his Commentaries on the Constitution that “in common cases, the law not only presumes every man innocent, until he is proved guilty; but unanimity in the verdict of the jury is indispensable.”[17] Similar statements can be found in American legal treatises throughout the 19th century.[18]

Nor is this a case where the original public meaning was lost to time and only recently recovered. This Court has, repeatedly and over many years, recognized that the Sixth Amendment requires unanimity. As early as 1898, the Court said that a defendant enjoys a “constitutional right to demand that his liberty should not be taken from him except by the joint action of the court and the unanimous verdict of a jury of twelve persons.”[19] A few decades later, the Court elaborated that the Sixth Amendment affords a right to “a trial by jury as understood and applied at common law, . . . includ[ing] all the essential elements as they were recognized in this country and England when the Constitution was adopted.”[20] And, the Court observed, this includes a requirement “that the verdict should be unanimous.”[21] In all, this Court has commented on the Sixth Amendment’s unanimity requirement no fewer than 13 times over more than 120 years.[22]

There can be no question either that the Sixth Amendment’s unanimity requirement applies to state and federal criminal trials equally. This Court has long explained that the Sixth Amendment right to a jury trial is “fundamental to the American scheme of justice” and incorporated against the States under the Fourteenth Amendment.[23] This Court has long explained, too, that incorporated provisions of the Bill of Rights bear the same content when asserted against States as they do when asserted against the federal government.[24] So if the Sixth Amendment’s right to a jury trial requires a unanimous verdict to support a conviction in federal court, it requires no less in state court.

II

A

How, despite these seemingly straightforward principles, have Louisiana’s and Oregon’s laws managed to hang on for so long? It turns out that the Sixth Amendment’s otherwise simple story took a strange turn in 1972. That year, the Court confronted these States’ unconventional schemes for the first time—in Apodaca v. Oregon[25] and a companion case, Johnson v. Louisiana.[26] Ultimately, the Court could do no more than issue a badly fractured set of opinions. Four dissenting Justices would not have hesitated to strike down the States’ laws, recognizing that the Sixth Amendment requires unanimity and that this guarantee is fully applicable against the States under the Fourteenth Amendment.[27] But a four-Justice plurality took a very different view of the Sixth Amendment. These Justices declared that the real question before them was whether unanimity serves an important “function” in “contemporary society.”[28] Then, having reframed the question, the plurality wasted few words before concluding that unanimity’s costs outweigh its benefits in the modern era, so the Sixth Amendment should not stand in the way of Louisiana or Oregon.

The ninth Member of the Court adopted a position that was neither here nor there. On the one hand, Justice Powell agreed that, as a matter of “history and precedent, . . . the Sixth Amendment requires a unanimous jury verdict to convict.”[29] But, on the other hand, he argued that the Fourteenth Amendment does not render this guarantee against the federal government fully applicable against the States. In this way, Justice Powell doubled down on his belief in “dual-track” incorporation—the idea that a single right can mean two different things depending on whether it is being invoked against the federal or a state government.

Justice Powell acknowledged that his argument for dual-track incorporation came “late in the day.”[30] Late it was. The Court had already, nearly a decade earlier, “rejected the notion that the Fourteenth Amendment applies to the States only a ‘watered-down, subjective version of the individual guarantees of the Bill of Rights.’ ”[31] It’s a point we’ve restated many times since, too, including as recently as last year.[32] Still, Justice Powell frankly explained, he was “unwillin[g]” to follow the Court’s precedents.[33] So he offered up the essential fifth vote to uphold Mr. Apodaca’s conviction—if based only on a view of the Fourteenth Amendment that he knew was (and remains) foreclosed by precedent.

B

In the years following Apodaca, both Louisiana and Oregon chose to continue allowing nonunanimous verdicts. But their practices have always stood on shaky ground. After all, while Justice Powell’s vote secured a favorable judgment for the States in Apodaca, it’s never been clear what rationale could support a similar result in future cases. Only two possibilities exist: Either the Sixth Amendment allows nonunanimous verdicts, or the Sixth Amendment’s guarantee of a jury trial applies with less force to the States under the Fourteenth Amendment. Yet, as we’ve seen, both bear their problems. In Apodaca itself, a majority of Justices—including Justice Powell—recognized that the Sixth Amendment demands unanimity, just as our cases have long said. And this Court’s precedents, both then and now, prevent the Court from applying the Sixth Amendment to the States in some mutated and diminished form under the Fourteenth Amendment. So what could we possibly describe as the “holding” of Apodaca?

Really, no one has found a way to make sense of it. In later cases, this Court has labeled Apodaca an “exception,” “unusual,” and in any event “not an endorsement” of Justice Powell’s view of incorporation.[34] At the same time, we have continued to recognize the historical need for unanimity.[35] We’ve been studiously ambiguous, even inconsistent, about what Apodaca might mean.[36] To its credit, Louisiana acknowledges the problem. The State expressly tells us it is not “asking the Court to accord Justice Powell’s solo opinion in Apodaca precedential force.”[37] Instead, in an effort to win today’s case, Louisiana embraces the idea that everything is up for grabs. It contends that this Court has never definitively ruled on the propriety of nonunanimous juries under the Sixth Amendment—and that we should use this case to hold for the first time that nonunanimous juries are permissible in state and federal courts alike.

III

Louisiana’s approach may not be quite as tough as trying to defend Justice Powell’s dual-track theory of incorporation, but it’s pretty close. How does the State deal with the fact this Court has said 13 times over 120 years that the Sixth Amendment does require unanimity? Or the fact that five Justices in Apodaca said the same? The best the State can offer is to suggest that all these statements came in dicta.[38] But even supposing (without granting) that Louisiana is right and it’s dicta all the way down, why would the Court now walk away from many of its own statements about the Constitution’s meaning? And what about the prior 400 years of English and American cases requiring unanimity—should we dismiss all those as dicta too?

Sensibly, Louisiana doesn’t dispute that the common law required unanimity. Instead, it argues that the drafting history of the Sixth Amendment reveals an intent by the framers to leave this particular feature behind. The State points to the fact that Madison’s proposal for the Sixth Amendment originally read: “The trial of all crimes . . . shall be by an impartial jury of freeholders of the vicinage, with the requisite of unanimity for conviction, of the right of challenge, and other accustomed requisites. . . .”[39] Louisiana notes that the House of Representatives approved this text with minor modifications. Yet, the State stresses, the Senate replaced “impartial jury of freeholders of the vicinage” with “impartial jury of the State and district wherein the crime shall have been committed” and also removed the explicit references to unanimity, the right of challenge, and “other accustomed requisites.” In light of these revisions, Louisiana would have us infer an intent to abandon the common law’s traditional unanimity requirement.

But this snippet of drafting history could just as easily support the opposite inference. Maybe the Senate deleted the language about unanimity, the right of challenge, and “other accustomed prerequisites” because all this was so plainly included in the promise of a “trial by an impartial jury” that Senators considered the language surplusage. The truth is that we have little contemporaneous evidence shedding light on why the Senate acted as it did.[40] So rather than dwelling on text left on the cutting room floor, we are much better served by interpreting the language Congress retained and the States ratified. And, as we’ve seen, at the time of the Amendment’s adoption, the right to a jury trial meant a trial in which the jury renders a unanimous verdict.

Further undermining Louisiana’s inference about the drafting history is the fact it proves too much. If the Senate’s deletion of the word “unanimity” changed the meaning of the text that remains, then the same would seemingly have to follow for the other deleted words as well. So it’s not just unanimity that died in the Senate, but all the “other accustomed requisites” associated with the common law jury trial right—i.e., everything history might have taught us about what it means to have a jury trial. Taking the State’s argument from drafting history to its logical conclusion would thus leave the right to a “trial by jury” devoid of meaning. A right mentioned twice in the Constitution would be reduced to an empty promise. That can’t be right.

Faced with this hard fact, Louisiana’s only remaining option is to invite us to distinguish between the historic features of common law jury trials that (we think) serve “important enough” functions to migrate silently into the Sixth Amendment and those that don’t. And, on the State’s account, we should conclude that unanimity isn’t worthy enough to make the trip.

But to see the dangers of Louisiana’s overwise approach, there’s no need to look any further than Apodaca itself. There, four Justices, pursuing the functionalist approach Louisiana espouses, began by describing the “ ‘essential’ ” benefit of a jury trial as “ ‘the interposition . . . of the commonsense judgment of a group of laymen’ ” between the defendant and the possibility of an “ ‘overzealous prosecutor.’ ”[41] And measured against that muddy yardstick, they quickly concluded that requiring 12 rather than 10 votes to convict offers no meaningful improvement.[42] Meanwhile, these Justices argued, States have good and important reasons for dispensing with unanimity, such as seeking to reduce the rate of hung juries.[43]

Who can profess confidence in a breezy cost-benefit analysis like that? Lost in the accounting are the racially discriminatory reasons that Louisiana and Oregon adopted their peculiar rules in the first place.[44] What’s more, the plurality never explained why the promised benefit of abandoning unanimity—reducing the rate of hung juries—always scores as a credit, not a cost. But who can say whether any particular hung jury is a waste, rather than an example of a jury doing exactly what the plurality said it should—deliberating carefully and safeguarding against overzealous prosecutions? And what about the fact, too, that some studies suggest that the elimination of unanimity has only a small effect on the rate of hung juries?[45] Or the fact that others profess to have found that requiring unanimity may provide other possible benefits, including more open-minded and more thorough deliberations?[46] It seems the Apodaca plurality never even conceived of such possibilities.

Our real objection here isn’t that the Apodaca plurality’s cost-benefit analysis was too skimpy. The deeper problem is that the plurality subjected the ancient guarantee of a unanimous jury verdict to its own functionalist assessment in the first place. And Louisiana asks us to repeat the error today, just replacing Apodaca’s functionalist assessment with our own updated version. All this overlooks the fact that, at the time of the Sixth Amendment’s adoption, the right to trial by jury included a right to a unanimous verdict. When the American people chose to enshrine that right in the Constitution, they weren’t suggesting fruitful topics for future cost-benefit analyses. They were seeking to ensure that their children’s children would enjoy the same hard-won liberty they enjoyed. As judges, it is not our role to reassess whether the right to a unanimous jury is “important enough” to retain. With humility, we must accept that this right may serve purposes evading our current notice. We are entrusted to preserve and protect that liberty, not balance it away aided by no more than social statistics.[47]

IV

A

If Louisiana’s path to an affirmance is a difficult one, the dissent’s is trickier still. The dissent doesn’t dispute that the Sixth Amendment protects the right to a unanimous jury verdict, or that the Fourteenth Amendment extends this right to state-court trials. But, it insists, we must affirm Mr. Ramos’s conviction anyway. Why? Because the doctrine of stare decisis supposedly commands it. There are two independent reasons why that answer falls short.

In the first place and as we’ve seen, not even Louisiana tries to suggest that Apodaca supplies a governing precedent. Remember, Justice Powell agreed that the Sixth Amendment requires a unanimous verdict to convict, so he would have no objection to that aspect of our holding today. Justice Powell reached a different result only by relying on a dual-track theory of incorporation that a majority of the Court had already rejected (and continues to reject). And to accept that reasoning as precedential, we would have to embrace a new and dubious proposition: that a single Justice writing only for himself has the authority to bind this Court to propositions it has already rejected.

This is not the rule, and for good reason—it would do more to destabilize than honor precedent. To see how, consider a hypothetical. Suppose we face a question of first impression under the Fourth Amendment: whether a State must obtain a warrant before reading a citizen’s email in the hands of an Internet provider and using that email as evidence in a criminal trial. Imagine this question splits the Court, with four Justices finding the Fourth Amendment requires a warrant and four Justices finding no such requirement. The ninth Justice agrees that the Fourth Amendment requires a warrant, but takes an idiosyncratic view of the consequences of violating that right. In her view, the exclusionary rule has gone too far, and should only apply when the defendant is prosecuted for a felony. Because the case before her happens to involve only a misdemeanor, she provides the ninth vote to affirm a conviction based on evidence secured by a warrantless search. Of course, this Court has longstanding precedent requiring the suppression of all evidence obtained in unconstitutional searches and seizures. Mapp v. Ohio, 367 U. S. 643 (1961). But like Justice Powell, our hypothetical ninth Justice sticks to her view and expressly rejects this Court’s precedent. Like Justice Powell, this Justice’s vote would be essential to the judgment. So if, as the dissent suggests, that is enough to displace precedent, would Mapp’s exclusionary rule now be limited to felony prosecutions?

Admittedly, this example comes from our imagination. It has to, because no case has before suggested that a single Justice may overrule precedent. But if the Court were to embrace the dissent’s view of stare decisis, it would not stay imaginary for long. Every occasion on which the Court is evenly split would present an opportunity for single Justices to overturn precedent to bind future majorities. Rather than advancing the goals of predictability and reliance lying behind the doctrine of stare decisis, such an approach would impair them.

The dissent contends that, in saying this much, we risk defying Marks v. United States.[48] According to Marks, when “a fragmented Court decides a case and no single rationale explaining the result enjoys the assent of five Justices, ‘the holding of the Court may be viewed as that position taken by those Members who concurred in the judgments on the narrowest grounds.’ ”[49] But notice that the dissent never actually gets around to telling us which opinion in Apodaca it considers to be the narrowest and controlling one under Marks—or why. So while the dissent worries that we defy a Marks precedent, it is oddly coy about where exactly that precedent might be found.

The parties recognize what the dissent does not: Marks has nothing to do with this case. Unlike a Marks dispute where the litigants duel over which opinion represents the narrowest and controlling one, the parties before us accept that Apodaca yielded no controlling opinion at all. In particular, both sides admit that Justice Powell’s opinion cannot bind us—precisely because he relied on a dual-track rule of incorporation that an unbroken line of majority opinions before and after Apodaca has rejected. Still, the dissent presses the issue, suggesting that a single Justice’s opinion can overrule prior precedents under “the logic” of Marks.[50] But, as the dissent itself implicitly acknowledges, Marks never sought to offer or defend such a rule. And, as we have seen, too, a rule like that would do more to harm than advance stare decisis.

The dissent’s backup argument fares no better. In the end, even the dissent is forced to concede that Justice Powell’s reasoning in Apodaca lacks controlling force.[51] So far, so good. But then the dissent suggests Apodaca somehow still manages to supply a controlling precedent as to its result.[52] Look closely, though. The dissent’s account of Apodaca’s result looks suspiciously like the reasoning of Justice Powell’s opinion: “In Apodaca, this means that when (1) a defendant is convicted in state court, (2) at least 10 of the 12 jurors vote to convict, and (3) the defendant argues that the conviction violates the Constitution because the vote was not unanimous, the challenge fails.”[53] Where does the convenient “state court” qualification come from? Neither the Apodaca plurality nor the dissent included any limitation like that—their opinions turned on the meaning of the Sixth Amendment. What the dissent characterizes as Apodaca’s result turns out to be nothing more than Justice Powell’s reasoning about dual-track incorporation dressed up to look like a logical proof.

All of this does no more than highlight an old truth. It is usually a judicial decision’s reasoning—its ratio decidendi—that allows it to have life and effect in the disposition of future cases.[54] As this Court has repeatedly explained in the context of summary affirmances, “ ‘unexplicated’ ” decisions may “ ‘settl[e] the issues for the parties, [but they are] not to be read as a renunciation by this Court of doctrines previously announced in our opinions.’ ”[55] Much the same may be said here. Apodaca’s judgment line resolved that case for the parties in that case. It is binding in that sense. But stripped from any reasoning, its judgment alone cannot be read to repudiate this Court’s repeated pre-existing teachings on the Sixth and Fourteenth Amendments.[56]

B

1

There’s another obstacle the dissent must overcome. Even if we accepted the premise that Apodaca established a precedent, no one on the Court today is prepared to say it was rightly decided, and stare decisis isn’t supposed to be the art of methodically ignoring what everyone knows to be true.[57] Of course, the precedents of this Court warrant our deep respect as embodying the considered views of those who have come before. But stare decisis has never been treated as “an inexorable command.”[58] And the doctrine is “at its weakest when we interpret the Constitution”[59] because a mistaken judicial interpretation of that supreme law is often “practically impossible” to correct through other means.[60] To balance these considerations, when it revisits a precedent this Court has traditionally considered “the quality of the decision’s reasoning; its consistency with related decisions; legal developments since the decision; and reliance on the decision.”[61] In this case, each factor points in the same direction.

Start with the quality of the reasoning. Whether we look to the plurality opinion or Justice Powell’s separate concurrence, Apodaca was gravely mistaken; again, no Member of the Court today defends either as rightly decided. Without repeating what we’ve already explained in detail, it’s just an implacable fact that the plurality spent almost no time grappling with the historical meaning of the Sixth Amendment’s jury trial right, this Court’s long-repeated statements that it demands unanimity, or the racist origins of Louisiana’s and Oregon’s laws. Instead, the plurality subjected the Constitution’s jury trial right to an incomplete functionalist analysis of its own creation for which it spared one paragraph. And, of course, five Justices expressly rejected the plurality’s conclusion that the Sixth Amendment does not require unanimity. Meanwhile, Justice Powell refused to follow this Court’s incorporation precedents. Nine Justices (including Justice Powell) recognized this for what it was; eight called it an error.

Looking to Apodaca’s consistency with related decisions and recent legal developments compounds the reasons for concern. Apodaca sits uneasily with 120 years of preceding case law. Given how unmoored it was from the start, it might seem unlikely that later developments could have done more to undermine the decision. Yet they have. While Justice Powell’s dual-track theory of incorporation was already foreclosed in 1972, some at that time still argued that it might have a role to play outside the realm of criminal procedure. Since then, the Court has held otherwise.[62] Until recently, dual-track incorporation attracted at least a measure of support in dissent. But this Court has now roundly rejected it.[63] Nor has the plurality’s rejection of the Sixth Amendment’s historical unanimity requirement aged more gracefully. As we’ve seen, in the years since Apodaca, this Court has spoken inconsistently about its meaning—but nonetheless referred to the traditional unanimity requirement on at least eight occasions.[64] In light of all this, calling Apodaca an outlier would be perhaps too suggestive of the possibility of company.

When it comes to reliance interests, it’s notable that neither Louisiana nor Oregon claims anything like the prospective economic, regulatory, or social disruption litigants seeking to preserve precedent usually invoke. No one, it seems, has signed a contract, entered a marriage, purchased a home, or opened a business based on the expectation that, should a crime occur, at least the accused may be sent away by a 10-to-2 verdict.[65] Nor does anyone suggest that nonunanimous verdicts have “become part of our national culture.”[66] It would be quite surprising if they had, given that nonunanimous verdicts are insufficient to convict in 48 States and federal court.

Instead, the only reliance interests that might be asserted here fall into two categories. The first concerns the fact Louisiana and Oregon may need to retry defendants convicted of felonies by nonunanimous verdicts whose cases are still pending on direct appeal. The dissent claims that this fact supplies the winning argument for retaining Apodaca because it has generated “enormous reliance interests” and overturning the case would provoke a “crushing” “tsunami” of follow-on litigation.[67]

The overstatement may be forgiven as intended for dramatic effect, but prior convictions in only two States are potentially affected by our judgment. Those States credibly claim that the number of nonunanimous felony convictions still on direct appeal are somewhere in the hundreds,[68] and retrying or plea bargaining these cases will surely impose a cost. But new rules of criminal procedures usually do, often affecting significant numbers of pending cases across the whole country. For example, after Booker v. United States held that the Federal Sentencing Guidelines must be advisory rather than mandatory, this Court vacated and remanded nearly 800 decisions to the courts of appeals. Similar consequences likely followed when Crawford v. Washington overturned prior interpretations of the Confrontation Clause[69] or Arizona v. Gant changed the law for searches incident to arrests.[70] Our decision here promises to cause less, and certainly nothing before us supports the dissent’s surmise that it will cause wildly more, disruption than these other decisions.

2

The second and related reliance interest the dissent seizes upon involves the interest Louisiana and Oregon have in the security of their final criminal judgments. In light of our decision today, the dissent worries that defendants whose appeals are already complete might seek to challenge their nonunanimous convictions through collateral (i.e., habeas) review.

But again the worries outstrip the facts. Under Teague v. Lane, newly recognized rules of criminal procedure do not normally apply in collateral review.[71] True, Teague left open the possibility of an exception for “watershed rules” “implicat[ing] the fundamental fairness [and accuracy] of the trial.”[72] But, as this language suggests, Teague’s test is a demanding one, so much so that this Court has yet to announce a new rule of criminal procedure capable of meeting it.[73] And the test is demanding by design, expressly calibrated to address the reliance interests States have in the finality of their criminal judgments.[74]

Nor is the Teague question even before us. Whether the right to jury unanimity applies to cases on collateral review is a question for a future case where the parties will have a chance to brief the issue and we will benefit from their adversarial presentation. That litigation is sure to come, and will rightly take into account the States’ interest in the finality of their criminal convictions. In this way, Teague frees us to say what we know to be true about the rights of the accused under our Constitution today, while leaving questions about the reliance interest States possess in their final judgments for later proceedings crafted to account for them. It would hardly make sense to ignore that two-step process and count the State’s reliance interests in final judgments both here and again there. Certainly the dissent cites no authority for such double counting.

Instead, the dissent suggests that the feeble reliance interests it identifies should get a boost because the right to a unanimous jury trial has “little practical importance going forward.”[75] In the dissent’s telling, Louisiana has “abolished” nonunanimous verdicts and Oregon “seemed on the verge of doing the same until the Court intervened.”[76] But, as the dissent itself concedes, a ruling for Louisiana would invite other States to relax their own unanimity requirements.[77] In fact, 14 jurisdictions have already told us that they would value the right to “experiment” with nonunanimous juries.[78] Besides, Louisiana’s law bears only prospective effect, so the State continues to allow nonunanimous verdicts for crimes committed before 2019.[79] And while the dissent speculates that our grant of certiorari contributed to the failure of legal reform efforts in Oregon, its citation does not support its surmise. No doubt, too, those who risk being subjected to nonunanimous juries in Louisiana and Oregon today, and elsewhere tomorrow, would dispute the dissent’s suggestion that their Sixth Amendment rights are of “little practical importance.”

That point suggests another. In its valiant search for reliance interests, the dissent somehow misses maybe the most important one: the reliance interests of the American people. Taken at its word, the dissent would have us discard a Sixth Amendment right in perpetuity rather than ask two States to retry a slice of their prior criminal cases. Whether that slice turns out to be large or small, it cannot outweigh the interest we all share in the preservation of our constitutionally promised liberties. Indeed, the dissent can cite no case in which the one-time need to retry defendants has ever been sufficient to inter a constitutional right forever.

In the final accounting, the dissent’s stare decisis arguments round to zero. We have an admittedly mistaken decision, on a constitutional issue, an outlier on the day it was decided, one that’s become lonelier with time. In arguing otherwise, the dissent must elide the reliance the American people place in their constitutionally protected liberties, overplay the competing interests of two States, count some of those interests twice, and make no small amount of new precedent all its own.

V

On what ground would anyone have us leave Mr. Ramos in prison for the rest of his life? Not a single Member of this Court is prepared to say Louisiana secured his conviction constitutionally under the Sixth Amendment. No one before us suggests that the error was harmless. Louisiana does not claim precedent commands an affirmance. In the end, the best anyone can seem to muster against Mr. Ramos is that, if we dared to admit in his case what we all know to be true about the Sixth Amendment, we might have to say the same in some others. But where is the justice in that? Every judge must learn to live with the fact he or she will make some mistakes; it comes with the territory. But it is something else entirely to perpetuate something we all know to be wrong only because we fear the consequences of being right. The judgment of the Court of Appeals is

Reversed.

SUPREME COURT OF THE UNITED STATES

_________________

No. 18–5924

_________________

EVANGELISTO RAMOS, PETITIONER v. LOUISIANA

on writ of certiorari to the court of appeal of louisiana, fourth circuit

[April 20, 2020]

Justice Thomas, concurring in the judgment.

I agree with the Court that petitioner Evangelisto Ramos’ felony conviction by a nonunanimous jury was unconstitutional. I write separately because I would resolve this case based on the Court’s longstanding view that the Sixth Amendment includes a protection against nonunanimous felony guilty verdicts, without undertaking a fresh analysis of the meaning of “trial . . . by an impartial jury.” I also would make clear that this right applies against the States through the Privileges or Immunities Clause of the Fourteenth Amendment, not the Due Process Clause.

I

I begin with the parties’ dispute as to whether the Sixth Amendment right to a trial by jury includes a protection against nonunanimous felony guilty verdicts. On this question, I do not write on a blank slate. As the Court acknowledges, our decisions have long recognized that unanimity is required. See ante, at 6–7. Because this interpretation is not demonstrably erroneous, I would resolve the Sixth Amendment question on that basis.

A

This Court first decided that the Sixth Amendment protected a right to unanimity in Thompson v. Utah, 170 U. S. 343 (1898). The Court reasoned that Thompson, a Utah prisoner, was protected by the Sixth Amendment when Utah was still a Territory because “the right of trial by jury in suits at common law appl[ied] to the Territories of the United States.” Id., at 346. The Court then stated that this right “made it impossible to deprive him of his liberty except by [a] unanimous verdict.” Id., at 355; see also id., at 351, 353.

The Court has repeatedly reaffirmed the Sixth Amendment’s unanimity requirement. In Patton v. United States, 281 U. S. 276 (1930), the Court stated that the Sixth Amendment protects the right “that the verdict should be unanimous,” id., at 288. In Andres v. United States, 333 U. S. 740 (1948), the Court repeated that “[u]nanimity in jury verdicts is required” by the Sixth Amendment, id., at 748. And in Apodaca v. Oregon, 406 U. S. 404 (1972), five Justices agreed that “the Sixth Amendment’s guarantee of trial by jury embraces a guarantee that the verdict of the jury must be unanimous,” id., at 414 (Stewart, J., joined by Brennan and Marshall, JJ., dissenting); see also Johnson v. Louisiana, 406 U. S. 356, 371 (1972) (Powell, J., concurring) (explaining views in Apodaca and its companion case); id., at 382–383 (Douglas, J., joined by Brennan and Marshall, JJ., dissenting) (same). We have accepted this interpretation of the Sixth Amendment in recent cases. See Southern Union Co. v. United States, 567 U. S. 343, 356 (2012); Blakely v. Washington, 542 U. S. 296, 301 (2004); Apprendi v. New Jersey, 530 U. S. 466, 477 (2000).

B

The question then becomes whether these decisions are entitled to stare decisis effect. As I have previously explained, “the Court’s typical formulation of the stare decisis standard does not comport with our judicial duty under Article III because it elevates demonstrably erroneous decisions—meaning decisions outside the realm of permissible interpretation—over the text of the Constitution and other duly enacted federal law.” Gamble v. United States, 587 U. S. ___, ___ (2019) (concurring opinion) (slip op., at 2). There is considerable evidence that the phrase “trial . . . by . . . jury” in the Sixth Amendment was understood since the founding to require that a felony guilty verdict be unanimous. Because our precedents are thus not outside the realm of permissible interpretation, I will apply them.

1

Blackstone—“the preeminent authority on English law for the founding generation,” Alden v. Maine, 527 U. S. 706, 715 (1999)—wrote that no subject can “be affected either in his property, his liberty, or his person, but by the unanimous consent” of a jury, 3 W. Blackstone, Commentaries on the Laws of England 379 (1772); see also 4 id., at 343. Another influential treatise author, Hale, wrote that “the law of England hath afforded the best method of trial, that is possible, . . . namely by a jury . . . all concurring in the same judgment.” 1 M. Hale, Pleas of the Crown 33 (1736) (emphasis deleted). Such views continued in scholarly works throughout the early Republic. See, e.g., 2 J. Story, Commentaries on the Constitution of the United States §777, p. 248 (1833); 6 N. Dane, Digest of American Law, ch. LXXXII, Art. 2, §1, p. 226 (1824); 2 J. Wilson, Works of the Honourable James Wilson 349–350 (1804).

The uniform practice among the States was in accord. Despite isolated 17th-century colonial practices allowing nonunanimous juries, “unanimity became the accepted rule during the 18th century, as Americans became more familiar with the details of English common law and adopted those details in their own colonial legal systems.” Apodaca, supra, at 408, n. 3 (plurality opinion). In the founding era, six States explicitly mentioned unanimity in their constitutions. See Del. Declaration of Rights §14 (1776); Md. Declaration of Rights, Art. XIX (1776); N. C. Declaration of Rights §IX (1776); Pa. Declaration of Rights, Art. IX (1776); Vt. Const., Art. XI (1786); Va. Declaration of Rights §8 (1776). Four more States clearly referred to the common-law jury right, which included unanimity. Ky. Const., Art. XII, §6 (1792); N. J. Const., Art. XXII (1776); N. Y. Const., Art. XLI (1777); S. C. Const., Art. IX, §6 (1790). Some States did not explicitly refer to either the common law or unanimity. See, e.g., Ga. Const., Art. LXI (1777); Mass. Declaration of Rights, Art. XII (1780). But there is reason to believe that they nevertheless understood unanimity to be required. See, e.g., Rouse v. State, 4 Ga. 136, 147 (1848).

In light of the express language used in some State Constitutions, respondent Louisiana argues that the omission of an express unanimity requirement in the Sixth Amendment reflects a deliberate choice. This argument fails to establish that the Court’s decisions are demonstrably erroneous. The House of Representatives passed a version of the amendment providing that “[t]he trial of all crimes . . . shall be by an impartial jury of freeholders of the vicinage, with the requisite of unanimity for conviction, of the right of challenge, and other accustomed requisites,” 1 Annals of Cong. 435 (1789), but the final Amendment contained no reference to vicinage or unanimity. See Amdt. 6. I agree with Justice Harlan and the Court that “the meaning of this change is wholly speculative” and that there is “no concrete evidence” that the Senate rejected the requirement of unanimity. Baldwin v. New York, 399 U. S. 66, 123, n. 9 (1970) (Harlan, J., dissenting); see also ante, at 11–12; Letter from J. Madison to E. Pendleton (Sept. 14, 1789), in 1 Letters and Other Writings of James Madison 491 (1867). There is thus sufficient evidence to support this Court’s prior interpretation that the Sixth Amendment right to a trial by jury requires unanimity.

2

There is also considerable evidence that this understanding persisted up to the time of the Fourteenth Amendment’s ratification. State courts, for example, continued to interpret the phrase “trial by jury” to require unanimity in felony guilty verdicts. The New Hampshire Superior Court of Judicature expounded on the point:

“The terms ‘jury,’ and ‘trial by jury,’ are, and for ages have been well known in the language of the law. They were used at the adoption of the constitution, and always, it is believed, before that time, and almost always since, in a single sense.

“A jury for the trial of a cause . . . must return their unanimous verdict upon the issue submitted to them.

“All the books of the law describe a trial jury substantially as we have stated it. And a ‘trial by jury’ is a trial by such a body, so constituted and conducted. So far as our knowledge extends, these expressions were used at the adoption of the constitution and always before, in these senses alone by all classes of writers and speakers.” Opinion of Justices, 41 N. H. 550, 551–552 (1860).

Other state courts held the same view. The Missouri Supreme Court in 1860 called unanimity one of the “essential requisites in a jury trial,” Vaughn v. Scade, 30 Mo. 600, 603, and the Ohio Supreme Court in 1853 called it one of “the essential and distinguishing features of the trial by jury, as known at common law, and generally, if not universally, adopted in this country,” Work v. State, 2 Ohio St. 296, 306.

Treatises from the Reconstruction era likewise adopted this position. A leading work on criminal procedure explained that if a “statute authorizes [a jury] to find a verdict upon anything short of . . . unanimous consent,” it “is void.” 1 J. Bishop, Criminal Procedure §761, p. 532 (1866). A widely read treatise on constitutional law reiterated that “ ‘by a jury’ is generally understood to mean” a body that “must unanimously concur in the guilt of the accused before a conviction can be had.” G. Paschal, The Constitution of the United States 210 (1876) (capitalization omitted). And a volume on the jury trial was in agreement. See J. Proffatt, Trial by Jury §77, p. 112 (1877).

*  *  *

Based on this evidence, the Court’s prior interpretation of the Sixth Amendment’s guarantee is not demonstrably erroneous. It is within the realm of permissible interpretations to say that “trial . . . by . . . jury” in that Amendment includes a protection against nonunanimous felony guilty verdicts.

II

The remaining question is whether that right is protected against the States. In my view, the Privileges or Immunities Clause provides this protection. I do not adhere to this Court’s decisions applying due process incorporation, including Apodaca and—it seems—the Court’s opinion in this case.

The Privileges or Immunities Clause provides that “[n]o State shall make or enforce any law which shall abridge the privileges or immunities of citizens of the United States.” Amdt. 14, §1. At the time of the Fourteenth Amendment’s ratification, “the terms ‘privileges’ and ‘immunities’ had an established meaning as synonyms of ‘rights.’ ” McDonald v. Chicago, 561 U. S. 742, 813 (2010) (Thomas, J., concurring in part and concurring in judgment). “[T]he ratifying public understood the Privileges or Immunities Clause to protect constitutionally enumerated rights” against abridgment by the States. Id., at 837. The Sixth Amendment right to a trial by jury is certainly a constitutionally enumerated right. See Maxwell v. Dow, 176 U. S. 581, 606–608 (1900) (Harlan, J., dissenting).

The Court, however, has made the Due Process Clause serve the function that the Privileges or Immunities Clause should serve. Although the Privileges or Immunities Clause grants “United States citizens a certain collection of rights—i.e., privileges or immunities—attributable to that status,” the Court has interpreted the Clause “quite narrowly.” McDonald, 561 U. S., at 808 (opinion of Thomas, J.). Perhaps to compensate for this limited view of the Privileges or Immunities Clause, it has incorporated individual rights against the States through the Due Process Clause. Id., at 809.

Due process incorporation is a demonstrably erroneous interpretation of the Fourteenth Amendment. As I have explained before, “[t]he notion that a constitutional provision that guarantees only ‘process’ before a person is deprived of life, liberty, or property could define the substance of those rights strains credulity for even the most casual user of words.” Id., at 811. The unreasonableness of this interpretation is underscored by the Court’s struggle to find a “guiding principle to distinguish ‘fundamental’ rights that warrant protection from nonfundamental rights that do not,” ibid., as well as its many incorrect decisions based on this theory, see Obergefell v. Hodges, 576 U. S. 644 (2015); Roe v. Wade, 410 U. S. 113 (1973); Dred Scott v. Sandford, 19 How. 393 (1857).

I “decline to apply the legal fiction” of due process incorporation. Timbs v. Indiana, 586 U. S. ___, ___ (2019) (Thomas, J., concurring in judgment) (slip op., at 3) (internal quotation marks omitted). As a result, I part ways with the Court on both its affirmative argument about the Fourteenth Amendment and its treatment of Apodaca, in which five Justices agreed the Sixth Amendment included a right to unanimity but a different majority concluded that the right did not apply to the States. See ante, at 7–11.

I would accept petitioner’s invitation to decide this case under the Privileges or Immunities Clause. The Court conspicuously avoids saying which clause it analyzes. See, e.g., ante, at 3, 7. But one assumes from its silence that the Court is either following our due process incorporation precedents or believes that “nothing in this case turns on” which clause applies, Timbs, supra, at ___ (Gorsuch, J., concurring) (slip op., at 1).

I have already rejected our due process incorporation cases as demonstrably erroneous, and I fundamentally disagree with applying that theory of incorporation simply because it reaches the same result in the case before us. Close enough is for horseshoes and hand grenades, not constitutional interpretation. The textual difference between protecting “citizens” (in the Privileges or Immunities Clause) and “person[s]” (in the Due Process Clause) will surely be relevant in another case. And our judicial duty—not to mention the candor we owe to our fellow citizens—requires us to put an end to this Court’s due process prestidigitation, which no one is willing to defend on the merits.

I would simply hold that, because all of the opinions in Apodaca addressed the Due Process Clause, its Fourteenth Amendment ruling does not bind us because the proper question here is the scope of the Privileges or Immunities Clause. I cannot understand why the Court, having decided to abandon Apodaca, refuses to correctly root its holding in the Privileges or Immunities Clause.[1]

III

There is no need to prove the original meaning of the

Sixth Amendment right to a trial by jury in this case.[2] The evidence that I have recounted is enough to establish that our previous interpretations of the Sixth Amendment are not demonstrably erroneous. What is necessary, however, is a clear understanding of the means by which the Sixth Amendment right applies against the States. We should rely on the Privileges or Immunities Clause, not the Due Process Clause or the Fourteenth Amendment in some vague sense. Accordingly, I concur only in the judgment.

SUPREME COURT OF THE UNITED STATES

_________________

No. 18–5924

_________________

EVANGELISTO RAMOS, PETITIONER v. LOUISIANA

on writ of certiorari to the court of appeal of louisiana, fourth circuit

[April 20, 2020]

Justice Kavanaugh, concurring in part.

In Apodaca v. Oregon, this Court held that state juries need not be unanimous in order to convict a criminal defendant. 406 U. S. 404 (1972). Two States, Louisiana and Oregon, have continued to use non-unanimous juries in criminal cases. Today, the Court overrules Apodaca and holds that state juries must be unanimous in order to convict a criminal defendant.

I agree with the Court that the time has come to overrule Apodaca. I therefore join the introduction and Parts I, II–A, III, and IV–B–1 of the Court’s persuasive and important opinion. I write separately to explain my view of how stare decisis applies to this case.

I

The legal doctrine of stare decisis derives from the Latin maxim “stare decisis et non quieta movere,” which means to stand by the thing decided and not disturb the calm. The doctrine reflects respect for the accumulated wisdom of judges who have previously tried to solve the same problem. In 1765, Blackstone—“the preeminent authority on English law for the founding generation,” Alden v. Maine, 527 U. S. 706, 715 (1999)—wrote that “it is an established rule to abide by former precedents,” to “keep the scale of justice even and steady, and not liable to waver with every new judge’s opinion.” 1 W. Blackstone, Commentaries on the Laws of England 69 (1765). The Framers of our Constitution understood that the doctrine of stare decisis is part of the “judicial Power” and rooted in Article III of the Constitution. Writing in Federalist 78, Alexander Hamilton emphasized the importance of stare decisis: To “avoid an arbitrary discretion in the courts, it is indispensable” that federal judges “should be bound down by strict rules and precedents, which serve to define and point out their duty in every particular case that comes before them.” The Federalist No. 78, p. 529 (J. Cooke ed. 1961). In the words of The Chief Justice, stare decisis’ “greatest purpose is to serve a constitutional ideal—the rule of law.” Citizens United v. Federal Election Comm’n, 558 U. S. 310, 378 (2010) (concurring opinion).

This Court has repeatedly explained that stare decisis “promotes the evenhanded, predictable, and consistent development of legal principles, fosters reliance on judicial decisions, and contributes to the actual and perceived integrity of the judicial process.” Payne v. Tennessee, 501 U. S. 808, 827 (1991). The doctrine “permits society to presume that bedrock principles are founded in the law rather than in the proclivities of individuals, and thereby contributes to the integrity of our constitutional system of government, both in appearance and in fact.” Vasquez v. Hillery, 474 U. S. 254, 265–266 (1986).

The doctrine of stare decisis does not mean, of course, that the Court should never overrule erroneous precedents. All Justices now on this Court agree that it is sometimes appropriate for the Court to overrule erroneous decisions. Indeed, in just the last few Terms, every current Member of this Court has voted to overrule multiple constitutional precedents. See, e.g., Knick v. Township of Scott, 588 U. S. ___ (2019); Franchise Tax Bd. of Cal. v. Hyatt, 587 U. S. ___ (2019); Janus v. State, County, and Municipal Employees, 585 U. S. ___ (2018); Hurst v. Florida, 577 U. S. ___ (2016); Obergefell v. Hodges, 576 U. S. 644 (2015); Johnson v. United States, 576 U. S. 591 (2015); Alleyne v. United States, 570 U. S. 99 (2013); see also Baude, Precedent and Discretion, 2020 S. Ct. Rev. 1, 4 (forthcoming) (“Nobody on the Court believes in absolute stare decisis”).

Historically, moreover, some of the Court’s most notable and consequential decisions have entailed overruling precedent. See, e.g., Obergefell v. Hodges, 576 U. S. 644 (2015); Citizens United v. Federal Election Comm’n, 558 U. S. 310 (2010); Montejo v. Louisiana, 556 U. S. 778 (2009); Crawford v. Washington, 541 U. S. 36 (2004); Lawrence v. Texas, 539 U. S. 558 (2003); Ring v. Arizona, 536 U. S. 584 (2002); Agostini v. Felton, 521 U. S. 203 (1997); Seminole Tribe of Fla. v. Florida, 517 U. S. 44 (1996); Planned Parenthood of Southeastern Pa. v. Casey, 505 U. S. 833 (1992);[1] Payne v. Tennessee, 501 U. S. 808 (1991); Batson v. Kentucky, 476 U. S. 79 (1986); Garcia v. San Antonio Metropolitan Transit Authority, 469 U. S. 528 (1985); Illinois v. Gates, 462 U. S. 213 (1983); United States v. Scott, 437 U. S. 82 (1978); Craig v. Boren, 429 U. S. 190 (1976); Taylor v. Louisiana, 419 U. S. 522 (1975); Brandenburg v. Ohio, 395 U. S. 444 (1969) (per curiam); Katz v. United States, 389 U. S. 347 (1967); Miranda v. Arizona, 384 U. S. 436 (1966); Malloy v. Hogan, 378 U. S. 1 (1964); Wesberry v. Sanders, 376 U. S. 1 (1964); Gideon v. Wainwright, 372 U. S. 335 (1963); Baker v. Carr, 369 U. S. 186 (1962); Mapp v. Ohio, 367 U. S. 643 (1961); Brown v. Board of Education, 347 U. S. 483 (1954); Smith v. Allwright, 321 U. S. 649 (1944); West Virginia Bd. of Ed. v. Barnette, 319 U. S. 624 (1943); United States v. Darby, 312 U. S. 100 (1941); Erie R. Co. v. Tompkins, 304 U. S. 64 (1938); West Coast Hotel Co. v. Parrish, 300 U. S. 379 (1937).

The lengthy and extraordinary list of landmark cases that overruled precedent includes the single most important and greatest decision in this Court’s history, Brown v. Board of Education, which repudiated the separate but equal doctrine of Plessy v. Ferguson, 163 U. S. 537 (1896).

As those many examples demonstrate, the doctrine of stare decisis does not dictate, and no one seriously maintains, that the Court should never overrule erroneous precedent. As the Court has often stated and repeats today, stare decisis is not an “inexorable command.” E.g., ante, at 20.

On the other hand, as Justice Jackson explained, just “because one should avoid Scylla is no reason for crashing into Charybdis.” Jackson, Decisional Law and Stare Decisis, 30 A. B. A. J. 334 (1944). So no one advocates that the Court should always overrule erroneous precedent.

Rather, applying the doctrine of stare decisis, this Court ordinarily adheres to precedent, but sometimes overrules precedent. The difficult question, then, is when to overrule an erroneous precedent.

To begin with, the Court’s precedents on precedent distinguish statutory cases from constitutional cases.

In statutory cases, stare decisis is comparatively strict, as history shows and the Court has often stated. That is because Congress and the President can alter a statutory precedent by enacting new legislation. To be sure, enacting new legislation requires finding room in a crowded legislative docket and securing the agreement of the House, the Senate (in effect, 60 Senators), and the President. Both by design and as a matter of fact, enacting new legislation is difficult—and far more difficult than the Court’s cases sometimes seem to assume. Nonetheless, the Court has ordinarily left the updating or correction of erroneous statutory precedents to the legislative process. See, e.g., Kimble v. Marvel Entertainment, LLC, 576 U. S. 446, 456–457 (2015); Patterson v. McLean Credit Union, 491 U. S. 164, 172–173 (1989); Flood v. Kuhn, 407 U. S. 258, 283–284 (1972). The principle that “it is more important that the applicable rule of law be settled than that it be settled right” is “commonly true even where the error is a matter of serious concern, provided correction can be had by legislation.” Burnet v. Coronado Oil & Gas Co., 285 U. S. 393, 406 (1932) (Brandeis, J., dissenting) (emphasis added).[2]

In constitutional cases, by contrast, the Court has repeatedly said—and says again today—that the doctrine of stare decisis is not as “inflexible.” Burnet, 285 U. S., at 406 (Brandeis, J., dissenting); see also ante, at 20; Payne, 501 U. S., at 828; Scott, 437 U. S., at 101. The reason is straightforward: As Justice O’Connor once wrote for the Court, stare decisis is not as strict “when we interpret the Constitution because our interpretation can be altered only by constitutional amendment or by overruling our prior decisions.” Agostini, 521 U. S., at 235. The Court therefore “must balance the importance of having constitutional questions decided against the importance of having them decided right.” Citizens United, 558 U. S., at 378 (Roberts, C. J., concurring). It follows “that in the unusual circumstance when fidelity to any particular precedent does more to damage this constitutional ideal than to advance it, we must be more willing to depart from that precedent.” Ibid. In his canonical opinion in Burnet, Justice Brandeis described the Court’s practice with respect to stare decisis in constitutional cases in a way that was accurate then and remains accurate now: In “cases involving the Federal Constitution, where correction through legislative action is practically impossible, this Court has often overruled its earlier decisions.” 285 U. S., at 406–407 (dissenting opinion).

That said, in constitutional as in statutory cases, to “overrule an important precedent is serious business.” Jackson, 30 A. B. A. J., at 334. In constitutional as in statutory cases, adherence to precedent is the norm. To overrule a constitutional decision, the Court’s precedents on precedent still require a “special justification,” Allen v. Cooper, 589 U. S. ___, ___ (2020) (slip op., at 9) (internal quotation marks omitted); Arizona v. Rumsey, 467 U. S. 203, 212 (1984), or otherwise stated, “strong grounds,” Janus, 585 U. S., at ___ (slip op., at 34).

In particular, to overrule a constitutional precedent, the Court requires something “over and above the belief that the precedent was wrongly decided.” Allen, 589 U. S., at ___ (slip op., at 9) (internal quotation marks omitted). As Justice Scalia put it, the doctrine of stare decisis always requires “reasons that go beyond mere demonstration that the overruled opinion was wrong,” for “otherwise the doctrine would be no doctrine at all.” Hubbard v. United States, 514 U. S. 695, 716 (1995) (opinion concurring in part and concurring in judgment). To overrule, the Court demands a special justification or strong grounds.

But the “special justification” or “strong grounds” formulation elides a key question: What constitutes a special justification or strong grounds?[3] In other words, in deciding whether to overrule an erroneous constitutional decision, how does the Court know when to overrule and when to stand pat?

As the Court has exercised the “judicial Power” over time, the Court has identified various stare decisis factors. In articulating and applying those factors, the Court has, to borrow James Madison’s words, sought to liquidate and ascertain the meaning of the Article III “judicial Power” with respect to precedent. The Federalist No. 37, at 236.

The stare decisis factors identified by the Court in its past cases include:

the quality of the precedent’s reasoning;

the precedent’s consistency and coherence with previous or subsequent decisions;

changed law since the prior decision;

changed facts since the prior decision;

the workability of the precedent;

the reliance interests of those who have relied on the precedent; and

the age of the precedent.

But the Court has articulated and applied those various individual factors without establishing any consistent methodology or roadmap for how to analyze all of the factors taken together. And in my view, that muddle poses a problem for the rule of law and for this Court, as the Court attempts to apply stare decisis principles in a neutral and consistent manner.

As I read the Court’s cases on precedent, those varied and somewhat elastic stare decisis factors fold into three broad considerations that, in my view, can help guide the inquiry and help determine what constitutes a “special justification” or “strong grounds” to overrule a prior constitutional decision.

First, is the prior decision not just wrong, but grievously or egregiously wrong? A garden-variety error or disagreement does not suffice to overrule. In the view of the Court that is considering whether to overrule, the precedent must be egregiously wrong as a matter of law in order for the Court to overrule it. In conducting that inquiry, the Court may examine the quality of the precedent’s reasoning, consistency and coherence with other decisions, changed law, changed facts, and workability, among other factors. A case may be egregiously wrong when decided, see, e.g., Korematsu v. United States, 323 U. S. 214 (1944); Plessy v. Ferguson, 163 U. S. 537 (1896), or may be unmasked as egregiously wrong based on later legal or factual understandings or developments, see, e.g., Nevada v. Hall, 440 U. S. 410 (1979), or both, ibid.

Second, has the prior decision caused significant negative jurisprudential or real-world consequences? In conducting that inquiry, the Court may consider jurisprudential consequences (some of which are also relevant to the first inquiry), such as workability, as well as consistency and coherence with other decisions, among other factors. Importantly, the Court may also scrutinize the precedent’s real-world effects on the citizenry, not just its effects on the law and the legal system. See, e.g., Brown v. Board of Education, 347 U. S., at 494–495; Barnette, 319 U. S., at 630–642; see also Payne, 501 U. S., at 825–827.

Third, would overruling the prior decision unduly upset reliance interests? This consideration focuses on the legitimate expectations of those who have reasonably relied on the precedent. In conducting that inquiry, the Court may examine a variety of reliance interests and the age of the precedent, among other factors.

In short, the first consideration requires inquiry into how wrong the precedent is as a matter of law. The second and third considerations together demand, in Justice Jackson’s words, a “sober appraisal of the disadvantages of the innovation as well as those of the questioned case, a weighing of practical effects of one against the other.” Jackson, 30 A. B. A. J., at 334.

Those three considerations together provide a structured methodology and roadmap for determining whether to overrule an erroneous constitutional precedent. The three considerations correspond to the Court’s historical practice and encompass the various individual factors that the Court has applied over the years as part of the stare decisis calculus. And they are consistent with the Founding understanding and, for example, Blackstone’s shorthand description that overruling is warranted when (and only when) a precedent is “manifestly absurd or unjust.” 1 Blackstone, Commentaries on the Laws of England, at 70.

Taken together, those three considerations set a high (but not insurmountable) bar for overruling a precedent, and they therefore limit the number of overrulings and maintain stability in the law.[4] Those three considerations also constrain judicial discretion in deciding when to overrule an erroneous precedent. To be sure, applying those considerations is not a purely mechanical exercise, and I do not claim otherwise. I suggest only that those three considerations may better structure how to consider the many traditional stare decisis factors.

It is inevitable that judges of good faith applying the stare decisis considerations will sometimes disagree about when to overrule an erroneous constitutional precedent, as the Court does in this case. To begin with, judges may disagree about whether a prior decision is wrong in the first place—and importantly, that disagreement is sometimes the real dispute when judges joust over stare decisis. But even when judges agree that a prior decision is wrong, they may dis- agree about whether the decision is so egregiously wrong as to justify an overruling. Judges may likewise disagree about the severity of the jurisprudential or real-world consequences caused by the erroneous decision and, therefore, whether the decision is worth overruling. In that regard, some judges may think that the negative consequences can be addressed by narrowing the precedent (or just living with it) rather than outright overruling it. Judges may also disagree about how to measure the relevant reliance interests that might be affected by an overruling. And on top of all of that, judges may also disagree about how to weigh and balance all of those competing considerations in a given case.[5]

This case illustrates that point. No Member of the Court contends that the result in Apodaca is correct. But the Members of the Court vehemently disagree about whether to overrule Apodaca.

II

Applying the three broad stare decisis considerations to this case, I agree with the Court’s decision to overrule Apodaca.

First, Apodaca is egregiously wrong. The original meaning and this Court’s precedents establish that the Sixth Amendment requires a unanimous jury. Ante, at 6–7; see, e.g., Patton v. United States, 281 U. S. 276, 288 (1930); Thompson v. Utah, 170 U. S. 343, 351 (1898). And the original meaning and this Court’s precedents establish that the Fourteenth Amendment incorporates the Sixth Amendment jury trial right against the States. See Duncan v. Louisiana, 391 U. S. 145, 149 (1968); id., at 166 (Black, J., concurring); see also Malloy, 378 U. S., at 10–11; see generally Timbs v. Indiana, 586 U. S. ___ (2019); McDonald v. Chicago, 561 U. S. 742 (2010). When Apodaca was decided, it was already an outlier in the Court’s jurisprudence, and over time it has become even more of an outlier. As the Court today persuasively explains, the original meaning of the Sixth and Fourteenth Amendments and this Court’s two lines of decisions—the Sixth Amendment jury cases and the Fourteenth Amendment incorporation cases—overwhelmingly demonstrate that Apodaca’s holding is egregiously wrong.[6]

Second, Apodaca causes significant negative consequences. It is true that Apodaca is workable. But Apodaca sanctions the conviction at trial or by guilty plea of some defendants who might not be convicted under the proper constitutional rule (although exactly how many is of course unknowable). That consequence has traditionally supplied some support for overruling an egregiously wrong criminal-procedure precedent. See generally Malloy, 378 U. S. 1.

In addition, and significant to my analysis of this case, the origins and effects of the non-unanimous jury rule strongly support overruling Apodaca. Louisiana achieved statehood in 1812. And throughout most of the 1800s, the State required unanimous juries in criminal cases. But at its 1898 state constitutional convention, Louisiana enshrined non-unanimous juries into the state constitution. Why the change? The State wanted to diminish the influence of black jurors, who had won the right to serve on juries through the Fourteenth Amendment in 1868 and the Civil Rights Act of 1875. See Strauder v. West Virginia, 100 U. S. 303, 308–310 (1880); T. Aiello, Jim Crow’s Last Stand: Nonunanimous Criminal Jury Verdicts in Louisiana 16, 19 (2015). Coming on the heels of the State’s 1896 victory in Plessy v. Ferguson, 163 U. S. 537, the 1898 constitutional convention expressly sought to “establish the supremacy of the white race.” Semmes, Chairman of the Committee on the Judiciary, Address at the Louisiana Constitutional Convention in 1898, in Official Journal of the Proceedings of the Constitutional Convention of the State of Louisiana 375 (H. Hearsey ed. 1898). And the convention approved non-unanimous juries as one pillar of a comprehensive and brutal program of racist Jim Crow measures against African-Americans, especially in voting and jury service. See Aiello, supra, at 16–26; Frampton, The Jim Crow Jury, 71 Vand. L. Rev. 1593, 1620 (2018).[7]

In light of the racist origins of the non-unanimous jury, it is no surprise that non-unanimous juries can make a difference in practice, especially in cases involving black defendants, victims, or jurors. After all, that was the whole point of adopting the non-unanimous jury requirement in the first place. And the math has not changed. Then and now, non-unanimous juries can silence the voices and negate the votes of black jurors, especially in cases with black defendants or black victims, and only one or two black jurors. The 10 jurors “can simply ignore the views of their fellow panel members of a different race or class.” Johnson v. Louisiana, 406 U. S. 356, 397 (1972) (Stewart, J., dissenting). That reality—and the resulting perception of unfairness and racial bias—can undermine confidence in and respect for the criminal justice system. The non-unanimous jury operates much the same as the unfettered peremptory challenge, a practice that for many decades likewise functioned as an engine of discrimination against black defendants, victims, and jurors. In effect, the non-unanimous jury allows backdoor and unreviewable peremptory strikes against up to 2 of the 12 jurors.

In its 1986 decision in Batson v. Kentucky, the Court recognized the pervasive racial discrimination woven into the traditional system of unfettered peremptory challenges. See 476 U. S., at 85–89, 91. The Court therefore overruled a prior decision, Swain v. Alabama, 380 U. S. 202 (1965), that had allowed those challenges. See generally Flowers v. Mississippi, 588 U. S. ___ (2019).

In my view, Apodaca warrants the same fate as Swain. After all, the “requirements of unanimity and impartial selection thus complement each other in ensuring the fair performance of the vital functions of a criminal court jury.” Johnson, 406 U. S., at 398 (Stewart, J., dissenting). And as Justice Thurgood Marshall forcefully explained in dissent in Apodaca, to “fence out a dissenting juror fences out a voice from the community, and undermines the principle on which our whole notion of the jury now rests.” Johnson, 406 U. S., at 402 (Marshall, J., dissenting in both Johnson and Apodaca).

To be clear, one could advocate for and justify a non-unanimous jury rule by resort to neutral and legitimate principles. England has employed non-unanimous juries, and various legal organizations in the United States have at times championed non-unanimous juries. See, e.g., Juries Act 1974, ch. 23, §17 (Eng.); ABA Project on Standards for Criminal Justice, Trial By Jury §1.1, p. 7 (App. Draft 1968); ALI, Code of Criminal Procedure §355, p. 99 (1930). And Louisiana’s modern policy decision to retain non- unanimous juries—as distinct from its original decision in the late 1800s to adopt non-unanimous juries—may have been motivated by neutral principles (or just by inertia).

But the question at this point is not whether the Constitution prohibits non-unanimous juries. It does. Rather, the disputed question here is whether to overrule an erroneous constitutional precedent that allowed non-unanimous juries. And on that question—the question whether to overrule—the Jim Crow origins and racially discriminatory effects (and the perception thereof ) of non-unanimous juries in Louisiana and Oregon should matter and should count heavily in favor of overruling, in my respectful view. After all, the non-unanimous jury “is today the last of Louisiana’s Jim Crow laws.” Aiello, supra, at 63. And this Court has emphasized time and again the “imperative to purge racial prejudice from the administration of justice” generally and from the jury system in particular. Pena- Rodriguez v. Colorado, 580 U. S. ___, ___–___ (2017) (slip op., at 13–14) (collecting cases).

To state the point in simple terms: Why stick by an erroneous precedent that is egregiously wrong as a matter of constitutional law, that allows convictions of some who would not be convicted under the proper constitutional rule, and that tolerates and reinforces a practice that is thoroughly racist in its origins and has continuing racially discriminatory effects?

Third, overruling Apodaca would not unduly upset reliance interests. Only Louisiana and Oregon employ non-unanimous juries in criminal cases. To be sure, in those two States, the Court’s decision today will invalidate some non-unanimous convictions where the issue is preserved and the case is still on direct review. But that consequence almost always ensues when a criminal-procedure precedent that favors the government is overruled. See Ring, 536 U. S. 584; Batson, 476 U. S. 79. And here, at least, I would “count that a small price to pay for the uprooting of this weed.” Hubbard, 514 U. S., at 717 (Scalia, J., concurring in part and concurring in judgment).

Except for the effects on that limited class of direct- review cases, it will be relatively easy going forward for Louisiana and Oregon to transition to the unanimous jury rule that the other 48 States and the federal courts use. Indeed, in 2018, Louisiana amended its constitution to require jury unanimity in criminal trials for crimes committed on or after January 1, 2019, meaning that the transition is already well under way in Louisiana.

Importantly, moreover, this Court applies a separate non-retroactivity doctrine to mitigate the disruptive effects of overrulings in criminal cases. Under the Court’s precedents, new constitutional rules apply on direct review, but generally do not apply retroactively on habeas corpus review. See Teague v. Lane, 489 U. S. 288, 311 (1989) (plurality opinion); Griffith v. Kentucky, 479 U. S. 314, 328 (1987). Teague recognizes only two exceptions to that general habeas non-retroactivity principle: “if (1) the rule is substantive or (2) the rule is a ‘watershed rul[e] of criminal procedure’ implicating the fundamental fairness and accuracy of the criminal proceeding.” Whorton v. Bockting, 549 U. S. 406, 416 (2007) (internal quotation marks omitted). The new rule announced today—namely, that state criminal juries must be unanimous—does not fall within either of those two narrow Teague exceptions and therefore, as a matter of federal law, should not apply retroactively on habeas corpus review.

The first Teague exception does not apply because today’s new rule is procedural, not substantive: It affects “only the manner of determining the defendant’s culpability.” Schriro v. Summerlin, 542 U. S. 348, 353 (2004).

The second Teague exception does not apply because today’s new rule, while undoubtedly important, is not a “watershed” procedural rule. This Court has flatly stated that “it is unlikely that any such rules” have “yet to emerge.” Whorton, 549 U. S., at 417 (internal quotation marks omitted). In “the years since Teague, we have rejected every claim that a new rule satisfied the requirements for watershed status.” Id., at 418, 421 (rejecting retroactivity for Crawford v. Washington, 541 U. S. 36 (2004)); see, e.g., Beard v. Banks, 542 U. S. 406, 420 (2004) (rejecting retroactivity for Mills v. Maryland, 486 U. S. 367 (1988)); Summerlin, 542 U. S., at 358 (rejecting retroactivity for Ring v. Arizona, 536 U. S. 584 (2002)); O’Dell v. Netherland, 521 U. S. 151, 167–168 (1997) (rejecting retroactivity for Simmons v. South Carolina, 512 U. S. 154 (1994)); Lambrix v. Singletary, 520 U. S. 518, 539–540 (1997) (rejecting retroactivity for Espinosa v. Florida, 505 U. S. 1079 (1992) (per curiam)); Sawyer v. Smith, 497 U. S. 227, 241–245 (1990) (rejecting retroactivity for Caldwell v. Mississippi, 472 U. S. 320 (1985)); see also Allen v. Hardy, 478 U. S. 255, 261 (1986) (per curiam) (rejecting retroactivity for Batson v. Kentucky, 476 U. S. 79 (1986)); DeStefano v. Woods, 392 U. S. 631, 635 (1968) (per curiam) (rejecting retroactivity for Duncan, 391 U. S. 145).

So assuming that the Court faithfully applies Teague, today’s decision will not apply retroactively on federal habeas corpus review and will not disturb convictions that are final.[8]

In addition, as to ineffective-assistance-of-counsel claims, an attorney presumably would not have been deficient for failing to raise a constitutional jury-unanimity argument before today’s decision—or at the very least, before the Court granted certiorari in this case. Before today, after all, this Court’s precedents had repeatedly allowed non-unanimous juries in state criminal cases. In that situation, the Courts of Appeals have consistently held that an attorney is not ineffective for failing to anticipate or advocate for the overruling of a constitutional precedent of this Court. See, e.g., Walker v. United States, 810 F. 3d 568, 577 (CA8 2016); United States v. Smith, 241 F. 3d 546, 548 (CA7 2001); Honeycutt v. Mahoney, 698 F. 2d 213, 216–217 (CA4 1983); see also Steiner v. United States, 940 F. 3d 1282, 1293 (CA11 2019) (per curiam); Snider v. United States, 908 F. 3d 183, 192 (CA6 2018); Green v. Johnson, 116 F. 3d 1115, 1125 (CA5 1997).

For those reasons, the reliance interests at stake in this case are not especially substantial, and they do not mandate adherence to Apodaca.[9]

*  *  *

In sum, Apodaca is egregiously wrong, it has significant negative consequences, and overruling it would not unduly upset reliance interests. I therefore agree with the Court’s decision to overrule Apodaca.[10]

SUPREME COURT OF THE UNITED STATES

_________________

No. 18–5924

_________________

EVANGELISTO RAMOS, PETITIONER v. LOUISIANA

on writ of certiorari to the court of appeal of louisiana, fourth circuit

[April 20, 2020]

Justice Sotomayor, concurring as to all but Part IV–A.

I agree with most of the Court’s rationale, and so I join all but Part IV–A of its opinion. I write separately, however, to underscore three points. First, overruling precedent here is not only warranted, but compelled. Second, the interests at stake point far more clearly to that outcome than those in other recent cases. And finally, the racially biased origins of the Louisiana and Oregon laws uniquely matter here.

I

Both the majority and the dissent rightly emphasize that stare decisis “has been a fundamental part of our jurisprudence since the founding.” Post, at 12 (opinion of Alito, J.); see ante, at 20. Indeed, “[w]e generally adhere to our prior decisions, even if we question their soundness, because doing so ‘promotes the evenhanded, predictable, and consistent development of legal principles, fosters reliance on judicial decisions, and contributes to the actual and perceived integrity of the judicial process.’ ” Alleyne v. United States, 570 U. S. 99, 118 (2013) (Sotomayor, J., concurring) (quoting Payne v. Tennessee, 501 U. S. 808, 827 (1991)).

But put simply, this is not a case where we cast aside precedent “simply because a majority of this Court now disagrees with” it. Alleyne, 570 U. S., at 133 (Alito, J., dissenting). Rather, Apodaca v. Oregon, 406 U. S. 464 (1972), was on shaky ground from the start. That was not because of the functionalist analysis of that Court’s plurality: Reasonable minds have disagreed over time—and continue to disagree—about the best mode of constitutional interpretation. That the plurality in Apodaca used different interpretive tools from the majority here is not a reason on its own to discard precedent.

What matters instead is that, as the majority rightly stresses, Apodaca is a universe of one—an opinion uniquely irreconcilable with not just one, but two, strands of constitutional precedent well established both before and after the decision. The Court has long recognized that the Sixth Amendment requires unanimity. Ante, at 11, 20–22. Five Justices in Apodaca itself disagreed with that plurality’s contrary view of the Sixth Amendment. Justice Powell’s theory of dual-track incorporation also fared no better: He recognized that his argument on that score came “late in the day.” Johnson v. Louisiana, 406 U. S. 356, 375 (1972) (concurring opinion).

Moreover, “[t]he force of stare decisis is at its nadir in cases concerning [criminal] procedur[e] rules that implicate fundamental constitutional protections.” Alleyne, 570 U. S., at 116, n. 5. And the constitutional protection here ranks among the most essential: the right to put the State to its burden, in a jury trial that comports with the Sixth Amendment, before facing criminal punishment. See Codispoti v. Pennsylvania, 418 U. S. 506, 515–516 (1974) (“The Sixth Amendment represents a deep commitment of the Nation to the right of jury trial in serious criminal cases as a defense against arbitrary law enforcement” (internal quotation marks omitted)). Where the State’s power to imprison those like Ramos rests on an erroneous interpretation of the jury-trial right, the Court should not hesitate to reconsider its precedents.

II

In contrast to the criminal-procedure context, “[c]onsiderations in favor of stare decisis are at their acme in cases involving property and contract rights.” Payne, 501 U. S., at 828. Despite that fact, the Court has recently overruled precedent where the Court’s shift threatened vast regulatory and economic consequences. Janus v. State, County, and Municipal Employees, 585 U. S. ___ (2018); id., at ___ (Kagan, J., dissenting) (slip op., at 23) (noting that the Court’s opinion called into question “thousands of . . . contracts covering millions of workers”); see South Dakota v. Wayfair, Inc., 585 U. S. ___, ___ (2018) (slip op., at 21) (noting the “legitimate” burdens that the Court’s overruling of precedent would place on vendors who had started businesses in reliance on a previous decision).

This case, by contrast, threatens no broad upheaval of private economic rights. Particularly when compared to the interests of private parties who have structured their affairs in reliance on our decisions, the States’ interests here in avoiding a modest number of retrials—emphasized at such length by the dissent—are much less weighty. They are certainly not new: Opinions that force changes in a State’s criminal procedure typically impose such costs. And were this Court to take the dissent’s approach—defending criminal-procedure opinions as wrong as Apodaca simply to avoid burdening criminal justice systems—it would never correct its criminal jurisprudence at all.

To pick up on the majority’s point, ante, at 23, in that alternate universe, a trial judge alone could still decide the critical facts necessary to sentence a defendant to death. Walton v. Arizona, 497 U. S. 639 (1990), overruled by Ring v. Arizona, 536 U. S. 584 (2002). An officer would still be able to search a car upon the arrest of any one of its recent occupants. New York v. Belton, 453 U. S. 454 (1981), holding limited by Arizona v. Gant, 556 U. S. 332 (2009). And States could still deprive a defendant of the right to confront her accuser so long as the incriminating statement was “reliable.” Ohio v. Roberts, 448 U. S. 56 (1980), abrogated by Crawford v. Washington, 541 U. S. 36 (2004). The Constitution demands more than the continued use of flawed criminal procedures—all because the Court fears the consequences of changing course.

III

Finally, the majority vividly describes the legacy of racism that generated Louisiana’s and Oregon’s laws. Ante, at 1–2, 13–14, and n. 44. Although Ramos does not bring an equal protection challenge, the history is worthy of this Court’s attention. That is not simply because that legacy existed in the first place—unfortunately, many laws and policies in this country have had some history of racial animus—but also because the States’ legislatures never truly grappled with the laws’ sordid history in reenacting them. See generally United States v. Fordice, 505 U. S. 717, 729 (1992) (policies that are “traceable” to a State’s de jure racial segregation and that still “have discriminatory effects” offend the Equal Protection Clause).

Where a law otherwise is untethered to racial bias—and perhaps also where a legislature actually confronts a law’s tawdry past in reenacting it—the new law may well be free of discriminatory taint. That cannot be said of the laws at issue here. While the dissent points to the “legitimate” reasons for Louisiana’s reenactment, post, at 3–4, Louisiana’s perhaps only effort to contend with the law’s discriminatory purpose and effects came recently, when the law was repealed altogether.

Today, Louisiana’s and Oregon’s laws are fully—and rightly—relegated to the dustbin of history. And so, too, is Apodaca. While overruling precedent must be rare, this Court should not shy away from correcting its errors where the right to avoid imprisonment pursuant to unconstitutional procedures hangs in the balance.

SUPREME COURT OF THE UNITED STATES

_________________

No. 18–5924

_________________

EVANGELISTO RAMOS, PETITIONER v. LOUISIANA

on writ of certiorari to the court of appeal of louisiana, fourth circuit

[April 20, 2020]

Justice Alito, with whom The Chief Justice joins, and with whom Justice Kagan joins as to all but Part III–D, dissenting.

The doctrine of stare decisis gets rough treatment in today’s decision. Lowering the bar for overruling our precedents, a badly fractured majority casts aside an important and long-established decision with little regard for the enormous reliance the decision has engendered. If the majority’s approach is not just a way to dispose of this one case, the decision marks an important turn.

Nearly a half century ago in Apodaca v. Oregon, 406 U. S. 404 (1972), the Court held that the Sixth Amendment permits non-unanimous verdicts in state criminal trials, and in all the years since then, no Justice has even hinted that Apodaca should be reconsidered. Understandably thinking that Apodaca was good law, the state courts in Louisiana and Oregon have tried thousands of cases under rules that permit such verdicts. But today, the Court does away with Apodaca and, in so doing, imposes a potentially crushing burden on the courts and criminal justice systems of those States. The Court, however, brushes aside these consequences and even suggests that the States should have known better than to count on our decision.

To add insult to injury, the Court tars Louisiana and Oregon with the charge of racism for permitting non- unanimous verdicts—even though this Court found such verdicts to be constitutional and even though there are entirely legitimate arguments for allowing them.

I would not overrule Apodaca. Whatever one may think about the correctness of the decision, it has elicited enormous and entirely reasonable reliance. And before this Court decided to intervene, the decision appeared to have little practical importance going forward. Louisiana has now abolished non-unanimous verdicts, and Oregon seemed on the verge of doing the same until the Court intervened.[1]

In Part II of this opinion, I will address the surprising argument, advanced by three Justices in the majority, that Apodaca was never a precedent at all, and in Part III, I will explain why stare decisis supports retention of that precedent. But before reaching those issues, I must say something about the rhetoric with which the majority has seen fit to begin its opinion.

I

Too much public discourse today is sullied by ad hominem rhetoric, that is, attempts to discredit an argument not by proving that it is unsound but by attacking the character or motives of the argument’s proponents. The majority regrettably succumbs to this trend. At the start of its opinion, the majority asks this rhetorical question: “Why do Louisiana and Oregon allow nonunanimous convictions?” Ante, at 1. And the answer it suggests? Racism, white supremacy, the Ku Klux Klan. Ante, at 1–2. Non-unanimous verdicts, the Court implies, are of a piece with Jim Crow laws, the poll tax, and other devices once used to disfranchise African-Americans. Ibid.

If Louisiana and Oregon originally adopted their laws allowing non-unanimous verdicts for these reasons,[2] that is deplorable, but what does that have to do with the broad constitutional question before us? The answer is: nothing.

For one thing, whatever the reasons why Louisiana and Oregon originally adopted their rules many years ago, both States readopted their rules under different circumstances in later years. Louisiana’s constitutional convention of 1974 adopted a new, narrower rule, and its stated purpose was “judicial efficiency.” State v. Hankton, 2012–0375, p. 19 (La. App. 4 Cir. 8/2/13), 122 So.  3d 1028, 1038. “In that debate no mention was made of race.” Ibid.; 7 Records of the Louisiana Constitutional Convention of 1973: Convention Transcripts 1184–1189 (La. Constitutional Convention Records Comm’n 1977). The people of Louisiana ratified the new Constitution. The majority makes no effort to show either that the delegates to the constitutional convention retained the rule for discriminatory purposes or that proponents of the new Constitution made racial appeals when approval was submitted to the people. The same is true for Oregon’s revisions and reenactments. Ore. Const., Art. I, §11 (amended May 18, 1934); Ore. Rev. Stat. §136.450 (1997); §136.610 (1971).

The more important point, however, is that today’s decision is not limited to anything particular about Louisiana or Oregon. The Court holds that the Sixth Amendment requires jury unanimity in all state criminal trials. If at some future time another State wanted to allow non-unanimous verdicts, today’s decision would rule that out—even if all that State’s lawmakers were angels.

For this reason, the origins of the Louisiana and Oregon rules have no bearing on the broad constitutional question that the Court decides. That history would be relevant if there were no legitimate reasons why anyone might think that allowing non-unanimous verdicts is good policy. But that is undeniably false.[3]

Some years ago the British Parliament enacted a law allowing non-unanimous verdicts.[4] Was Parliament under the sway of the Klan? The Constitution of Puerto Rico permits non-unanimous verdicts.[5] Were the framers of that Constitution racists? Non-unanimous verdicts were once advocated by the American Law Institute and the American Bar Association.[6] Was their aim to promote white supremacy? And how about the prominent scholars who have taken the same position?[7] Racists all? Of course not. So all the talk about the Klan, etc., is entirely out of place.[8] We should set an example of rational and civil discourse instead of contributing to the worst current trends.

II

Now to what matters.

A

I begin with the question whether Apodaca was a precedent at all. It is remarkable that it is even necessary to address this question, but in Part IV–A of the principal opinion, three Justices take the position that Apodaca was never a precedent. The only truly fitting response to this argument is: “Really?”

Consider what it would mean if Apodaca was never a precedent. It would mean that the entire legal profession was fooled for the past 48 years. Believing that Apodaca was a precedent, the courts of Louisiana and Oregon tried thousands of cases under rules allowing conviction by a vote of 11 to 1 or 10 to 2, and appellate courts in those States upheld these convictions based on Apodaca.[9] But according to three Justices in the majority, these courts were deluded.

This Court, for its part, apparently helped to perpetuate the illusion, since it reiterated time and again what Apodaca had established. See Timbs v. Indiana, 586 U. S. ___, ___, n. 1 (2019) (slip op., at 3, n. 1) (Apodaca held “that the Sixth Amendment requires jury unanimity in federal, but not state, criminal proceedings”); McDonald v. Chicago, 561 U. S. 742, 766, n. 14 (2010) ( Sixth Amendment “does not require a unanimous jury verdict in state criminal trials”); United States v. Gaudin, 515 U. S. 506, 511, n. 2 (1995) (Apodaca “conclude[d] that jury unanimity is not constitutionally required”); Schad v. Arizona, 501 U. S. 624, 634, n. 5 (1991) (plurality opinion) (“[A] state criminal defendant, at least in noncapital cases, has no federal right to a unanimous jury verdict”); Brown v. Louisiana, 447 U. S. 323, 330–331 (1980) (plurality opinion) (“[T]he constitutional guarantee of trial by jury” does not prescribe “the exact proportion of the jury that must concur in the verdict”); Burch v. Louisiana, 441 U. S. 130, 136 (1979) (Apodaca “conclude[d] that a jury’s verdict need not be unanimous to satisfy constitutional requirements”); Ludwig v. Massachusetts, 427 U. S. 618, 625 (1976) (“holding” in Apodaca was that “the jury’s verdict need not be unanimous”); see also Holland v. Illinois, 493 U. S. 474, 511 (1990) (Stevens, J., dissenting) (“we have permitted nonunanimous verdicts,” citing Apodaca); McKoy v. North Carolina, 494 U. S. 433, 468 (1990) (Scalia, J., dissenting) (the Court has “approved verdicts by less than a unanimous jury,” citing Apodaca).

Consistent with these statements of the governing law, whenever defendants convicted by non-unanimous verdicts sought review in this Court and asked that Apodaca be overruled, the Court denied those requests—without a single registered dissent.[10] Even the legal academy, never shy about puncturing misconceptions, was taken in.[11] Everybody thought Apodaca was a precedent. But, according to three of the Justices in the majority, everybody was fooled. Apodaca, the precedent, was a mirage. Can this be true?

No, it cannot. The idea that Apodaca was a phantom precedent defies belief. And it certainly disserves important objectives that stare decisis exists to promote, including evenhandedness, predictability, and the protection of legitimate reliance. See, e.g., Gamble v. United States, 587 U. S. ___, ___ (2019); Kimble v. Marvel Entertainment, LLC, 576 U. S. 446, 455–456 (2015); Payne v. Tennessee, 501 U. S. 808, 827 (1991).

B

Under any reasonable understanding of the concept, Apodaca was a precedent, that is, “a decided case that furnishes a basis for determining later cases involving similar facts or issues.” Black’s Law Dictionary 1366 (10th ed. 2014); see also J. Salmond, Jurisprudence 191 (10th ed. 1947); M. Gerhardt, The Power of Precedent 3 (2008); Landes & Posner, Legal Precedent: A Theoretical and Empirical Analysis, 19 J. Law & Econ. 249, 250 (1976).

Even though there was no opinion of the Court, the decision satisfies even the narrowest understanding of a precedent as this Court has understood the concept: The decision prescribes a particular outcome when all the conditions in a clearly defined set are met. See Seminole Tribe of Fla. v. Florida, 517 U. S. 44, 67 (1996) (explaining that, at the very least, we are bound by the “result” in a prior case). In Apodaca, this means that when (1) a defendant is convicted in state court, (2) at least 10 of the 12 jurors vote to convict, and (3) the defendant argues that the conviction violates the Constitution because the vote was not unanimous, the challenge fails. A majority of the Justices in Apodaca expressly agreed on that result, and that result is a precedent that had to be followed in subsequent cases until Apodaca was overruled.

That this result constituted a precedent follows a fortiori from our cases holding that even our summary affirmances of lower court decisions are precedents for “the precise issues presented and necessarily decided” by the judgment below. Mandel v. Bradley, 432 U. S. 173, 176 (1977) (per curiam). If the Apodaca Court had summarily affirmed a state-court decision holding that a jury vote of 10 to 2 did not violate the Sixth Amendment, that summary disposition would be a precedent. Accordingly, it is impossible to see how a full-blown decision of this Court reaching the same result can be regarded as a non-precedent.[12]

C

What do our three colleagues say in response? They begin by suggesting that Louisiana conceded that Apodaca is not a precedent. See ante, at 16–17. This interpretation of the State’s position is questionable,[13] but even if Louisiana made that concession, how could that settle the matter? What about Oregon, the only State that still permits non-unanimous verdicts? Oregon certainly did not make such a concession. On the contrary, it submitted an amicus brief arguing strenuously that Apodaca is a precedent and that it should be retained. Brief for State of Oregon as Amicus Curiae 6–32. And what about any other State that might want to allow such verdicts in the future? So the majority’s reliance on Louisiana’s purported concession simply will not do.

Our three colleagues’ next try is to argue that Apodaca is not binding because a case has no ratio decidendi when a majority does not agree on the reason for the result. Ante, at 19, and n. 54. This argument, made in passing, constitutes an attack on the rule that the Court adopted in Marks v. United States, 430 U. S. 188 (1977), for determining the holding of a decision when there is no majority opinion. Under the Marks rule, “[w]hen a fragmented Court decides a case and no single rationale explaining the result enjoys the assent of five Justices, the holding of the Court may be viewed as that position taken by those Members who concurred in the judgments on the narrowest grounds.” Id., at 193 (internal quotation marks omitted). This rule ascribes precedential status to decisions made without majority agreement on the underlying rationale, and it is therefore squarely contrary to the argument of the three Justices who regard Apodaca as non-precedential.

The Marks rule is controversial, and two Terms ago, we granted review in a case that implicated its meaning. See Hughes v. United States, 584 U. S. ___ (2018). But we ultimately decided the case on another ground and left the Marks rule intact. As long as that rule stands, it refutes the argument that Apodaca is not binding because a majority did not agree on a common rationale.

Finally, our three colleagues contend that treating Apodaca as a precedent would require the Court “to embrace a new and dubious proposition: that a single Justice writing only for himself has the authority to bind this Court to propositions it has already rejected.” Ante, at 16. This argument appears to weave together three separate questions relating to the precedential effect of decisions in which there is no majority opinion. I will therefore attempt to untangle these questions and address each in turn.

An initial question is whether, in a case where there is no opinion of the Court, the position taken by a single Justice in the majority can constitute the binding rule for which the decision stands. Under Marks, the clear answer to this question is yes. The logic of Marks applies equally no matter what the division of the Justices in the majority, and I am aware of no case holding that the Marks rule is inapplicable when the narrowest ground is supported by only one Justice. Certainly the lower courts have understood Marks to apply in that situation.[14]

The next question is whether the Marks rule applies any differently when the precedent that would be established by a fractured decision would overrule a prior precedent. Again, the logic of Marks dictates an affirmative answer, and I am aware of no case holding that the Marks rule applies any differently in this situation. But as far as the present case is concerned, this question is academic because Apodaca did not overrule any prior decision of this Court. At most, what the Court had “recognized,” ante, at 6, in prior cases is that the Sixth Amendment guaranteed the right to a unanimous jury verdict in trials in federal and territorial courts.[15] Whether the same rule applied in state prosecutions had not been decided, and indeed, until Duncan v. Louisiana, 391 U. S. 145, 154–158 (1968), was handed down just four years before Apodaca, the Sixth Amendment had not been held to apply to the States.

The final question is whether Justice Powell’s reasoning in Apodaca—namely, his view that the Fourteenth Amendment did not incorporate every aspect of the Sixth Amendment jury-trial right—is a binding precedent, and the answer to that question is no. When, in the years after Apodaca, new questions arose about the scope of the jury-trial right in state court—as they did in cases like Apprendi v. New Jersey, 530 U. S. 466 (2000), and Blakely v. Washington, 542 U. S. 296 (2004)—nobody thought for a second that Apodaca committed the Court to Justice Powell’s view that the right has different dimensions in state and federal cases. And no one on this Court or on a lower court had any trouble locating the narrow common ground between Justice Powell and the plurality in Apodaca: The States need not require unanimity to comply with the Constitution.

For all these reasons, Apodaca clearly was a precedent, and if the Court wishes to be done with it, it must explain why overruling Apodaca is consistent with the doctrine of stare decisis.

III

A

Stare decisis has been a fundamental part of our jurisprudence since the founding, and it is an important doctrine. But, as we have said many times, it is not an “inexorable command.” Payne, 501 U. S., at 828; Gamble, 587 U. S., at ___–___ (slip op., at 11–12). There are circumstances when past decisions must be overturned, but we begin with the presumption that we will follow precedent, and therefore when the Court decides to overrule, it has an obligation to provide an explanation for its decision.

This is imperative because the Court should have a body of neutral principles on the question of overruling precedent. The doctrine should not be transformed into a tool that favors particular outcomes.[16]

B

What is the majority’s justification for overruling Apodaca? With no apparent appreciation of the irony, today’s majority, which is divided into four separate camps,[17] criticizes the Apodaca majority as “badly fractured.” Ante, at 8. But many important decisions currently regarded as precedents were decided without an opinion of the Court.[18] Does the majority mean to suggest that all such precedents are fair game?

The majority’s primary reason for overruling Apodaca is the supposedly poor “quality” of Justice White’s plurality opinion and Justice Powell’s separate opinion. Ante, at 19–21. The majority indicts Justice White’s opinion on five grounds: (1) it “spent almost no time grappling with the historical meaning of the Sixth Amendment’s jury trial right,”[19] (2) it did not give due weight to the “Court’s long-repeated statements that [the right] demands unanimity,”[20] (3) it did not take into account “the racist origins of [the] Louisian[a] and Orego[n] laws,”[21] (4) it looked to the function of the jury-trial right,[22] and (5) it engaged in “a breezy cost-benefit analysis” that, in any event, did not properly weigh the costs and benefits.[23] All these charges are overblown.

First, it is quite unfair to criticize Justice White for not engaging in a detailed discussion of the original meaning of the Sixth Amendment jury-trial right since he had already done that just two years before in his opinion for the Court in Williams v. Florida, 399 U. S. 78, 92–100 (1970). In Williams, after examining that history, he concluded that the Sixth Amendment did not incorporate every feature of the common-law right (a conclusion that the majority, by the way, does not dispute). And in Apodaca, he built on the analysis in Williams. Accordingly, there was no need to repeat what had been said before.

Second, it is similarly unfair to criticize Justice White for not discussing the prior decisions that commented on jury unanimity. None of those decisions went beyond saying that this was a feature of the common-law right or cursorily stating that unanimity was required.[24] And as noted, Williams had already held that the Sixth Amendment did not preserve all aspects of the common-law right.

Third, the failure of Justice White (and Justice Powell) to take into account the supposedly racist origins of the Louisiana and Oregon laws should not be counted as a defect for the reasons already discussed. See supra, at 4–5.

Fourth, it is hard to know what to make of the functionalist charge. One Member of the majority explicitly disavows this criticism, see ante, at 2 (Sotomayor, J., concurring in part), and it is most unlikely that all the Justices in the majority are ready to label all functionalist decisions as poorly reasoned. Most of the landmark criminal procedure decisions from roughly Apodaca’s time fall into that category. See Mapp v. Ohio, 367 U. S. 643, 654 (1961) ( Fourth Amendment); Miranda v. Arizona, 384 U. S. 436, 444 (1966) ( Fifth Amendment); Gideon v. Wainwright, 372 U. S. 335, 344–345 (1963) ( Sixth Amendment); Furman v. Georgia, 408 U. S. 238, 239 (1972) (per curiam) ( Eighth Amendment).[25] Are they all now up for grabs?

The functionalist criticism dodges the knotty problem that led Justice White to look to the underlying purpose of the jury-trial right. Here is the problem. No one questions that the Sixth Amendment incorporated the core of the common-law jury-trial right, but did it incorporate every feature of the right? Did it constitutionalize the requirement that there be 12 jurors even though nobody can say why 12 is the magic number? And did it incorporate features that we now find highly objectionable, such as the exclusion of women from jury service? At the time of the adoption of the Sixth Amendment (and for many years thereafter), women were not regarded as fit to serve as a defendant’s peers. Unless one is willing to freeze in place late 18th-century practice, it is necessary to find a principle to distinguish between the features that were incorporated and those that were not. To do this, Justice White’s opinion for the Court in Williams looked to the underlying purpose of the jury-trial right, which it identified as interposing a jury of the defendant’s peers to protect against oppression by a “ ‘corrupt or overzealous prosecutor’ ” or a “ ‘compliant, biased, or eccentric judge.’ ” 399 U. S., at 100 (quoting Duncan, 391 U. S., at 156).

The majority decries this “functionalist” approach but provides no alternative. It does not claim that the Sixth Amendment incorporated every feature of common-law practice, but it fails to identify any principle for identifying the features that were absorbed. On the question of jury service by women, the majority’s only answer, buried in a footnote, is that the exclusion of women was outlawed by “further constitutional amendments,” ante, at 15, n. 47, presumably the Fourteenth Amendment. Does that mean that the majority disagrees with the holding in Taylor v. Louisiana, 419 U. S. 522 (1975)—another opinion by Justice White—that the exclusion of women from jury service violates the Sixth Amendment? Id., at 531, 533–536.[26]

Fifth, it is not accurate to say that Justice White based his conclusion on a cost-benefit analysis of requiring jury unanimity. His point, rather, was that what the Court had already identified as the fundamental purpose of the jury-trial right was not undermined by allowing a verdict of 11 to 1 or 10 to 2.

I cannot say that I would have agreed either with Justice White’s analysis or his bottom line in Apodaca if I had sat on the Court at that time, but the majority’s harsh criticism of his opinion is unwarranted.

What about Justice Powell’s concurrence? The majority treats Justice Powell’s view as idiosyncratic, but it does not merit that derision. Justice Powell’s belief that the Constitution allows the States a degree of flexibility in the interpretation of certain constitutional rights, although not our dominant approach in recent years, McDonald, 561 U. S., at 759–766, has old and respectable roots. For a long time, that was the Court’s approach. See id., at 759–761. Only gradually did the Court abandon this “two-tier” system, see id., at 762–767, and it was not until Duncan, supra, at 154–158, decided just four years before Apodaca, that the Sixth Amendment jury-trial right was held to apply to the States at all. Justice Powell’s approach is also not without recent proponents, including, at least with respect to the Second Amendment, Justices now in the majority.[27]

Even now, our cases do not hold that every provision of the Bill of Rights applies in the same way to the Federal Government and the States. A notable exception is the Grand Jury Clause of the Fifth Amendment, a provision that, like the Sixth Amendment jury-trial right, reflects the importance that the founding generation attached to juries as safeguards against oppression. In Hurtado v. California, 110 U. S. 516, 538 (1884), the Court held that the Grand Jury Clause does not bind the States and that they may substitute preliminary hearings at which the decision to allow a prosecution to go forward is made by a judge rather than a defendant’s peers. That decision was based on reasoning that is not easy to distinguish from Justice Powell’s in Apodaca. Hurtado remains good law and is critically important to the 28 States that allow a defendant to be prosecuted for a felony without a grand jury indictment.[28] If we took the same approach to the Hurtado question that the majority takes in this case, the holding in that case could be called into question.

The majority’s only other reason for overruling Apodaca is that it is inconsistent with related decisions and recent legal developments. Ante, at 21; ante, at 2 (Sotomayor, J., concurring in part). I agree that Justice Powell’s view on incorporation is not in harmony with the bulk of our case law, but the majority’s point about “recent legal developments” is an exaggeration. No subsequent Sixth Amendment decision has undercut the plurality. And while Justice Powell’s view on incorporation has been further isolated by later cases holding that two additional provisions of the Bill of Rights apply with full force to the States, see Timbs, 586 U. S., at ___ (slip op., at 2) ( Eighth Amendment’s Excessive Fines Clause); McDonald, supra, at 791 (plurality opinion) ( Second Amendment), the project of complete incorporation was nearly done when Apodaca was handed down. See McDonald, supra, at 765, n. 13.

While the majority worries that Apodaca is inconsistent with our cases on incorporation, the majority ignores something far more important: the way in which Apodaca is intertwined with the body of our Sixth Amendment case law. As I have explained, see supra, at 15, the Apodaca plurality’s reasoning was based on the same fundamental mode of analysis as that in Williams, 399 U. S. 78, which had held just two years earlier that the Sixth Amendment did not constitutionalize the common law’s requirement that a jury have 12 members. Although only one State, Oregon, now permits non-unanimous verdicts, many more allow six- person juries.[29] Repudiating the reasoning of Apodaca will almost certainly prompt calls to overrule Williams.

C

Up to this point, I have discussed the majority’s reasons for overruling Apodaca, but that is only half the picture. What convinces me that Apodaca should be retained are the enormous reliance interests of Louisiana and Oregon. For 48 years, Louisiana and Oregon, trusting that Apodaca is good law, have conducted thousands and thousands of trials under rules allowing non-unanimous verdicts. Now, those States face a potential tsunami of litigation on the jury- unanimity issue.

At a minimum, all defendants whose cases are still on direct appeal will presumably be entitled to a new trial if they were convicted by a less-than-unanimous verdict and preserved the issue in the trial court. And at least in Oregon, even if no objection was voiced at trial, defendants may be able to challenge their convictions based on plain error. See Ore. Rule App. Proc. 5.45(1), and n. 1 (2019); State v. Serrano, 355 Ore. 172, 179, 324 P. 3d 1274, 1280 (2014). Oregon asserts that more than a thousand defendants whose cases are still on direct appeal may be able to challenge their convictions if Apodaca is overruled. Brief for State of Oregon as Amicus Curiae 12–13.[30] The State also reports that “[d]efendants are arguing that an instruction allowing for non-unanimous verdicts is a structural error that requires reversal for all convictions, even for those for which the jury was not polled or those for which the jury was unanimous.” Id., at 14.

Unimpressed by these potential consequences, the majority notes that we “vacated and remanded nearly 800 decisions” for resentencing after United States v. Booker, 543 U. S. 220 (2005), held that the Federal Sentencing Guidelines are not mandatory. Ante, at 23. But the burden of resentencing cannot be compared with the burden of retrying cases. And while resentencing was possible in all the cases affected by Booker, there is no guarantee that all the cases affected by today’s ruling can be retried. In some cases, key witnesses may not be available, and it remains to be seen whether the criminal justice systems of Oregon and Louisiana have the resources to handle the volume of cases in which convictions will be reversed.

These cases on direct review are only the beginning. Prisoners whose direct appeals have ended will argue that today’s decision allows them to challenge their convictions on collateral review, and if those claims succeed, the courts of Louisiana and Oregon are almost sure to be overwhelmed.

The majority’s response to this possibility is evasive. It begins by hinting that today’s decision will not apply on collateral review under the framework adopted in Teague v. Lane, 489 U. S. 288, 315 (1989) (plurality opinion). Under Teague, “an old rule applies both on direct and collateral review,” but if today’s decision constitutes a new procedural rule, prisoners will be able to rely on it in a collateral proceeding only if it is what we have termed a “watershed rule” that implicates “the fundamental fairness and accuracy of the criminal proceeding.” Whorton v. Bockting, 549 U. S. 406, 416 (2007). Noting that we have never found a new rule of criminal procedure to qualify as “watershed,” the Court hints that the decision in this case is likely to meet the same fate.

But having feinted in this direction, the Court quickly changes course and says that the application of today’s decision to prisoners whose appeals have ended should not concern us. Ante, at 23–24. That question, we are told, will be decided in a later case. Ibid.

The majority cannot have it both ways. As long as retroactive application on collateral review remains a real possibility, the crushing burden that this would entail cannot be ignored. And while it is true that this Court has been chary in recognizing new watershed rules, it is by no means clear that Teague will preclude the application of today’s decision on collateral review.

Teague applies only to a “new rule,” and the positions taken by some in the majority may lead to the conclusion that the rule announced today is an old rule. Take the proposition, adopted by three Members of the majority, that Apodaca was never a precedent. Those Justices, along with the rest of the majority, take the position that our cases established well before Apodaca both that the Sixth Amendment requires unanimity, ante, at 6–7, and that it applies in the same way in state and federal court, ante, at 9. Thus, if Apodaca was never a precedent and did not disturb what had previously been established, it may be argued that today’s decision does not impose a new rule but instead merely recognizes what the correct rule has been for many years.

Two other Justices in the majority acknowledge that Apodaca was a precedent and thus would presumably regard today’s decision as a “new rule,” but the question remains whether today’s decision qualifies as a “watershed rule.” Justice Kavanaugh concludes that it does not and all but decides—without briefing or argument—that the decision will not apply retroactively on federal collateral review and similarly that there will be no successful claims of ineffective assistance of counsel for failing to challenge Apodaca. See ante, at 15–17 (opinion concurring in part).

The remaining Justices in the majority, and those of us in dissent, express no view on this question, but the majority’s depiction of the unanimity requirement as a hallowed right that Louisiana and Oregon flouted for ignominious reasons certainly provides fuel for the argument that the rule announced today meets the test. And in Oregon, the State most severely impacted by today’s decision, watershed status may not matter since the State Supreme Court has reserved decision on whether state law gives prisoners a greater opportunity to invoke new precedents in state collateral proceedings. See Verduzco v. State, 357 Ore. 553, 574, 355 P. 3d 902, 914 (2015).[31]

Whatever the ultimate resolution of the retroactivity question, the reliance here is not only massive; it is concrete. Cf. Dickerson v. United States, 530 U. S. 428, 443 (2000) (reliance weighed heavily in favor of precedent simply because the warnings in Miranda v. Arizona, 384 U. S. 436, had become “part of our national culture”). In my view, it weighs decisively against overruling Apodaca.

In reaching this conclusion, I do not disregard the interests of petitioner and others who were convicted by a less-than-unanimous vote. It is not accurate to imply that these defendants would have been spared conviction if unanimity had been required. In many cases, if a unanimous vote had been needed, the jury would have continued to deliberate and the one or two holdouts might well have ultimately voted to convict.[32] This is almost certainly the situation in Oregon, where it is estimated that as many as two-thirds of all criminal trials have ended with a non-unanimous verdict. See Brief for State of Oregon as Amicus Curiae 12. It is impossible to believe that all these cases would have resulted in mistrials if unanimity had been demanded. Instead, after a vote of 11 to 1 or 10 to 2, it is likely that deliberations would have continued and unanimity would have been achieved.

Nevertheless, the plight of defendants convicted by non-unanimous votes is important and cannot be overlooked, but that alone cannot be dispositive of the stare decisis question. Otherwise, stare decisis would never apply in a case in which a criminal defendant challenges a precedent that led to conviction.

D

The reliance in this case far outstrips that asserted in recent cases in which past precedents were overruled. Last Term, when we overturned two past decisions, there were strenuous dissents voicing fears about the future of stare decisis. See Franchise Tax Bd. of Cal. v. Hyatt, 587 U. S. ___, ___ (2019) (Breyer, J., dissenting); Knick v. Township of Scott, 588 U. S. ___, ___ (2019) (Kagan, J., dissenting). Yet in neither of those cases was there reliance like that present here.

In Franchise Tax Board, the dissent claimed only the airiest sort of reliance, the public’s expectation that past decisions would remain on the books. 587 U. S., at ___–___ (opinion of Breyer, J.) (slip op., at 12–13). And in Knick, the dissent disclaimed any reliance at all. 588 U. S., at ___ (opinion of Kagan, J.) (slip op., at 17). The same was true the year before in South Dakota v. Wayfair, Inc., 585 U. S. ___ (2018), where the dissent did not contend that any legitimate reliance interests weighed in favor of preserving the decision that the Court overruled. Id., at ___–___ (opinion of Roberts, C. J.) (slip op., at 1–2). And our unanimous decision in Pearson v. Callahan, 555 U. S. 223, 233 (2009), found that no reliance interests were involved.

In other cases overruling prior decisions, the dissents claimed that reliance interests were at stake, but whatever one may think about the weight of those interests, no one can argue that they are comparable to those in this case.

In Montejo v. Louisiana, 556 U. S. 778, 793–797 (2009), the Court abrogated a prophylactic rule that had been adopted in Michigan v. Jackson, 475 U. S. 625 (1986), to protect a defendant’s right to counsel during post- arraignment interrogation. The dissent did not claim that any defendants had relied on this rule, arguing instead that the public at large had an interest “in knowing that counsel, once secured, may be reasonably relied upon as a medium between the accused and the power of the State.” Montejo, supra, at 809 (opinion of Stevens, J.). This abstract interest, if it can be called reliance in any proper sense of the term, is a far cry from what is at stake here.

In Citizens United v. Federal Election Comm’n, 558 U. S. 310 (2010), where we overruled precedent allowing laws that prohibited corporations’ election-related speech, we found that “[n]o serious reliance interests” were implicated, id., at 365, since the only reliance asserted by the dissent was the time and effort put in by federal and state lawmakers in adopting the provisions at issue, id., at 411–412 (Stevens, J., concurring in part and dissenting in part). In this case, by contrast, what is at stake is not the time and effort of Louisiana and Oregon lawmakers but a monumental litigation burden and the potential inability to retry cases that might well have ended with a unanimous verdict if that had been required.

Finally, in Janus v. State, County, and Municipal Employees, 585 U. S. ___ (2018), where we overruled Abood v. Detroit Bd. of Ed., 431 U. S. 209 (1977), we carefully considered and addressed the question of reliance, and whatever one may think about the extent of the legitimate reliance in that case, it is not in the same league as that present here. Abood had held that a public sector employer may require non-union members to pay a portion of the dues collected from union members. 431 U. S., at 235–236. In overruling that decision, we acknowledged that existing labor contracts might have been negotiated in reliance on Abood, but we noted that most labor contracts are of short duration, that unions had been on notice for some time that the Court had serious misgivings about Abood, and that unions could have insisted on contractual provisions to protect their interests if Abood later fell. Janus, supra, at ___–___ (slip op., at 44–47).[33]

By striking down a precedent upon which there has been massive and entirely reasonable reliance, the majority sets an important precedent about stare decisis. I assume that those in the majority will apply the same standard in future cases.

*  *  *

Under the approach to stare decisis that we have taken in recent years, Apodaca should not be overruled. I would therefore affirm the judgment below, and I respectfully dissent.