with whom Justice Ginsburg joins, dissenting.
In 2009, in Arizona v. Gant, 556 U.S. 332, 129 S. Ct. 1710, 173 L. Ed. 2d 485, this Court held that a police search of an automobile without a warrant violates the Fourth Amendment if the police have previously removed the automobile’s occupants and placed them securely in a squad car. The present case involves these same circumstances, and it was pending on appeal when this Court decided Gant. Because Gant represents a “shift” in the Court’s Fourth Amendment jurisprudence, ante, at 1, we must decide whether and how Gant’s new rule applies here.
I
I agree with the Court about whether Gant’s new rule applies. It does apply. Between 1965, when the Court decided Linkletter v. Walker, 381 U.S. 618, 85 S. Ct. 1731, 14 L. Ed. 2d 601, and 1987, when it decided Griffith v. Kentucky, 479 U.S. 314, 107 S. Ct. 708, 93 L. Ed. 2d 649, that conclusion
[564 U.S. 253]
would have been more difficult to reach. Under Linkletter, the Court determined a new rule’s retro-activity by looking to several different factors, including whether the new rule represented a “clear break” with the past and the degree of “reliance by law enforcement authorities on the old standards.” Desist v. United States, 394 U.S. 244, 248-249, 89 S. Ct. 1030, 22 L. Ed. 2d 248 (1969) (internal quotation marks omitted) (also citing “the purpose to be served by the new standards” and “the effect on the administration of justice” as factors (internal quotation marks omitted)). And the Court would often not apply the new rule to identical cases still pending on appeal. See ibid.
After 22 years of struggling with its Linkletter approach, however, the Court decided in Griffith that Linkletter had proved unfair and unworkable. It then substituted a clearer approach, stating that “a new rule for the conduct of criminal prosecutions is to be applied retroactively to all cases, state or federal, pending on direct review or not yet final, with no exception for cases in which the new rule constitutes a ‘clear break’ with the past.” 479 U.S., at 328, 107 S. Ct. 708, 93 L. Ed. 2d 649. The Court today, following Griffith, concludes that Gant’s new rule applies here. And to that extent I agree with its decision.
II
The Court goes on, however, to decide how Gant’s new rule will apply. *305And here it adds a fatal twist. While conceding that, like the search in Gant, this search violated the Fourth Amendment, it holds that, unlike Gant, this defendant is not entitled to a remedy. That is because the Court finds a new “good faith” exception which prevents application of the normal remedy for a Fourth Amendment violation, namely, suppression of the illegally seized evidence. Weeks v. United States, 232 U.S. 383, 34 S. Ct. 341, 58 L. Ed. 652 (1914); Mapp v. Ohio, 367 U.S. 643, 81 S. Ct. 1684, 6 L. Ed. 2d 1081 (1961). Leaving Davis with a right but not a remedy, the Court “keep[s] the word of promise to our ear” but “break[s] it to our hope.”
[564 U.S. 254]
A
At this point I can no longer agree with the Court. A new “good faith” exception and this Court’s retroactivity decisions are incompatible. For one thing, the Court’s distinction between (1) retroactive application of a new rule and (2) availability of a remedy is highly artificial and runs counter to precedent. To determine that a new rule is retroactive is to determine that, at least in the normal case, there is a remedy. As we have previously said, the “source of a ‘new rule’ is the Constitution itself, not any judicial power to create new rules of law”; hence, “[w]hat we are actually determining when we assess the ‘retroactivity’ of a new rule is not the temporal scope of a newly announced right, but whether a violation of the right that occurred prior to the announcement of the new rule will entitle a criminal defendant to the relief sought.” Danforth v. Minnesota, 552 U.S. 264, 271, 128 S. Ct. 1029, 169 L. Ed. 2d 859 (2008). The Court’s “good faith” exception (unlike, say, inevitable discovery, a remedial doctrine that applies only upon occasion) creates “a categorical bar to obtaining redress” in every case pending when a precedent is overturned. Ante, at 243, 180 L. Ed. 2d, at 298.
For another thing, the Court’s holding recreates the very problems that led the Court to abandon Linkletter’s approach to retroactivity in favor of Griffith’s. One such problem concerns workability. The Court says that its exception applies where there is “objectively reasonable” police “reliance on binding appellate precedent.” Ante, at 232, 249-250, 180 L. Ed. 2d, at 290, 302. But to apply the term “binding appellate precedent” often requires resolution of complex questions of degree. Davis conceded that he faced binding anti-Gant precedent in the Eleventh Circuit. But future litigants will be less forthcoming. Ante, at 248, 180 L. Ed. 2d, at 301. Indeed, those litigants will now have to create distinctions to show that previous circuit precedent was not “binding” lest they find relief foreclosed even if they win their constitutional claim.
[564 U.S. 255]
At the same time, Fourth Amendment precedents frequently require courts to “slosh” their “way through the factbound morass of ‘reasonableness.’ ” Scott v. Harris, 550 U.S. 372, 383, 127 S. Ct. 1769, 167 L. Ed. 2d 686 (2007). Suppose an officer’s conduct is consistent with the language of a Fourth Amendment rule that a court of appeals announced in a case with clearly distinguishable facts? Suppose the case creating the relevant precedent did not directly announce any general rule but involved highly analogous facts? What about a rule that all other jurisdictions, but not the defendant’s jurisdiction, had previously accepted? What rules can be developed for determining when, where, and how these different kinds of precedents do, or do not, count as *306relevant “binding precedent”? The Linkletter-like result is likely complex legal argument and police force confusion. See Mackey v. United States, 401 U.S. 667, 676, 91 S. Ct. 1148, 28 L. Ed. 2d 388 (1971) (opinion of Harlan, J.) (describing trying to follow Linkletter decisions as “almost as difficult” as trying to follow “the tracks made by a beast of prey in search of its intended victim”).
Another such problem concerns fairness. Today’s holding, like that in Linkletter, “violates basic norms of constitutional adjudication.” Griffith, 479 U.S., at 322, 107 S. Ct. 708, 93 L. Ed. 2d 649. It treats the defendant in a case announcing a new rule one way while treating similarly situated defendants whose cases are pending on appeal in a different way. See ante, at 248-249, 180 L. Ed. 2d, at 301-302. Justice Harlan explained why this approach is wrong when he said:
“We cannot release criminals from jail merely because we think one case is a particularly appropriate one [to announce a constitutional doctrine] .... Simply fishing one case from the stream of appellate review, using it as a vehicle for pronouncing new constitutional standards, and then permitting a stream of similar cases subsequently to flow by unaffected by that new rule constitute an indefensible departure from [our ordinary] model of judicial review.” Mackey, supra, at 679, 91 S. Ct. 1148, 28 L. Ed. 2d 388.
[564 U.S. 256]
And in Griffith, the Court “embraced to a significant extent the comprehensive analysis presented by Justice Harlan.” 479 U.S., at 322, 107 S. Ct. 708, 93 L. Ed. 2d 649.
Of course, the Court may, as it suggests, avoid this unfairness by refusing to apply the exclusionary rule even to the defendant in the very case in which it announces a “new rule.” But that approach would make matters worse. What would then happen in the lower courts? How would courts of appeals, for example, come to reconsider their prior decisions when other circuits’ cases lead them to believe those decisions may be wrong? Why would a defendant seek to overturn any such decision? After all, if the (incorrect) circuit precedent is clear, then even if the defendant wins (on the constitutional question), he loses (on relief). See Stovall v. Denno, 388 U.S. 293, 301, 87 S. Ct. 1967, 18 L. Ed. 2d 1199 (1967). To what extent then could this Court rely upon lower courts to work out Fourth Amendment differences among themselves— through circuit reconsideration of a precedent that other circuits have criticized? See Arizona v. Evans, 514 U.S. 1, 23, n. 1, 115 S. Ct. 1185, 131 L. Ed. 2d 34 (1995) (Ginsburg, J., dissenting).
B
Perhaps more important, the Court’s rationale for creating its new “good faith” exception threatens to undermine well-settled F ourth Amendment law. The Court correctly says that pre-Gant Eleventh Circuit precedent had held that a Gant-type search was constitutional; hence the police conduct in this case, consistent with that precedent, was “innocent.” Ante, at 240, 180 L. Ed. 2d, at 296. But the Court then finds this fact sufficient to create a new “good faith” exception to the exclusionary rule. It reasons that the “sole purpose” of the exclusionary rule “is to deter future Fourth Amendment violations,” ante, at 236-237, 180 L. Ed. 2d, at 293. The “deterrence benefits of exclusion vary with the culpability of the law en*307forcement conduct at issue,” ante, at 238, 180 L. Ed. 2d, at 295 (internal quotation marks and brackets omitted). Those benefits are sufficient to justify exclusion
[564 U.S. 257]
where “police exhibit deliberate, reckless, or grossly negligent disregard for Fourth Amendment rights,” ibid. (internal quotation marks omitted). But those benefits do not justify exclusion where, as here, the police act with “simple, isolated negligence” or an “objectively reasonable good-faith belief that their conduct is lawful,” ibid. (internal quotation marks omitted).
If the Court means what it says, what will happen to the exclusionary rule, a rule that the Court adopted nearly a century ago for federal courts, Weeks v. United States, 232 U.S. 383, 34 S. Ct. 341, 58 L. Ed. 652, and made applicable to state courts a half century ago through the Fourteenth Amendment, Mapp v. Ohio, 367 U.S. 643, 81 S. Ct. 1684, 6 L. Ed. 2d 1081? The Court has thought of that rule not as punishment for the individual officer or as reparation for the individual defendant but more generally as an effective way to secure enforcement of the Fourth Amendment’s commands. Weeks, supra, at 393, 34 S. Ct. 341, 58 L. Ed. 652 (without the exclusionary rule, the Fourth Amendment would be “of no value,” and “might as well be stricken from the Constitution”). This Court has deviated from the “suppression” norm in the name of “good faith” only a handful of times and in limited, atypical circumstances: where a magistrate has erroneously issued a warrant, United States v. Leon, 468 U.S. 897, 104 S. Ct. 3405, 82 L. Ed. 2d 677 (1984); where a database has erroneously informed police that they have a warrant, Arizona v. Evans, supra, Herring v. United States, 555 U.S. 135, 129 S. Ct. 695, 172 L. Ed. 2d 496 (2009); and where an unconstitutional statute purported to authorize the search, Illinois v. Krull, 480 U.S. 340, 107 S. Ct. 1160, 94 L. Ed. 2d 364 (1987). See Herring, supra, at 142, 129 S. Ct. 695, 172 L. Ed. 2d 496 (“good faith” exception inaptly named).
The fact that such exceptions are few and far between is understandable. Defendants frequently move to suppress evidence on Fourth Amendment grounds. In many, perhaps most, of these instances the police, uncertain of how the Fourth Amendment applied to the particular factual circumstances they faced, will have acted in objective good faith. Yet, in a significant percentage of these instances, courts will
[564 U.S. 258]
find that the police were wrong. And, unless the police conduct falls into one of the exceptions previously noted, courts have required the suppression of the evidence seized. 1 W. LaFave, Search and Seizure § 1.3, pp. 103-104 (4th ed. 2004) (“good faith” exception has not yet been applied to warrantless searches and seizures beyond the “rather special situations” of Evans, Herring, and Krull). See Valdes, Frequency and Success: An Empirical Study of Criminal Law Defenses, Federal Constitutional Evidentiary Claims, and Plea Negotiations, 153 U. Pa. L. Rev. 1709, 1728 (2005) (suppression motions are filed in approximately 7% of criminal cases; approximately 12% of suppression motions are successful); LaFave, supra, at 64 (“Surely many more Fourth Amendment violations result from carelessness than from intentional constitutional violations”); Stewart, The Road to Mapp v. Ohio and Beyond: The Origins, Development and Future of *308the Exclusionary Rule in Search-and-Seizure Cases, 83 Colum. L. Rev. 1365, 1389 (1983) (“[T]he vast majority of fourth amendment violations ... [are] motivated by commendable zeal, not condemnable malice”).
But an officer who conducts a search that he believes complies with the Constitution but which, it ultimately turns out, falls just outside the Fourth Amendment’s bounds is no more culpable than an officer who follows erroneous “binding precedent.” Nor is an officer more culpable where circuit precedent is simply suggestive rather than “binding,” where it only describes how to treat roughly analogous instances, or where it just does not exist. Thus, if the Court means what it now says, if it would place determinative weight upon the culpability of an individual officer’s conduct, and if it would apply the exclusionary rule only where a Fourth Amendment violation was “deliberate, reckless, or grossly negligent,” then the “good faith” exception will swallow the exclusionary rule. Indeed, our broad dicta in Herring—dicta the Court repeats and expands upon today—may already be leading
[564 U.S. 259]
lower courts in this direction. See United States v. Julius, 610 F.3d 60, 66-67 (CA2 2010) (assuming war-rantless search was unconstitutional and remanding for District Court to “perform the cost/benefit analysis required by Herring” and to consider “whether the degree of police culpability in this case rose beyond mere . . . negligence” before ordering suppression); United States v. Master, 614 F.3d 236, 243 (CA6 2010) (“[T]he Herring Court’s emphasis seems weighed more toward preserving evidence for use in obtaining convictions, even if illegally seized . . . unless the officers engage in ‘deliberate, reckless, or grossly negligent conduct’ ” (quoting Herring, supra, at 144, 129 S. Ct. 695, 172 L. Ed. 2d 496)). Today’s decision will doubtless accelerate this trend.
Any such change (which may already be underway) would affect not “an exceedingly small set of cases,” ante, at 247, 180 L. Ed. 2d, at 301, but a very large number of cases, potentially many thousands each year. See Valdes, supra, at 1728. And since the exclusionary rule is often the only sanction available for a Fourth Amendment violation, the Fourth Amendment would no longer protect ordinary Americans from “unreasonable searches and seizures.” See Wolf v. Colorado, 338 U.S. 25, 41, 69 S. Ct. 1359, 93 L. Ed. 1782 (1949) (Murphy, J., dissenting) (In many circumstances, “there is but one alternative to the rule of exclusion. That is no sanction at all”), overruled by Mapp v. Ohio, supra; Herring, supra, at 152, 129 S. Ct. 695, 172 L. Ed. 2d 496 (Ginsburg, J., dissenting) (the exclusionary rule is “an essential auxiliary” to the Fourth Amendment). It would become a watered-down Fourth Amendment, offering its protection against only those searches and seizures that are egregiously unreasonable.
III
In sum, I fear that the Court’s opinion will undermine the exclusionary rule. And I believe that the Court wrongly departs from Griffith regardless. Instead I would follow Griffith, apply Gant’s rule retroactively to this case, and require
[564 U.S. 260]
suppression of the evidence. Such an approach is consistent with our precedent, and it would indeed affect no more than “an exceedingly small set of cases." Ante, at 247, 180 L. Ed. 2d, at 301.
*309For these reasons, with respect, I dissent.