BOARD OF COMM'RS OF BRYAN CTY. v. BROWN

Justice O'Connor, delivered the opinion of the Court.

Respondent Jill Brown brought a claim for damages against petitioner Bryan County under Rev. Stat. § 1979, 42 U. S. C. § 1983. She alleged that a county police officer used excessive force in arresting her, and that the county itself was liable for her injuries based on its sheriff's hiring and training decisions. She prevailed on her claims against the county following a jury trial, and the Court of Appeals for the Fifth Circuit affirmed the judgment against the county on the basis of the hiring claim alone. 67 F.3d 1174 (1995). We granted certiorari. We conclude that the Court of Appeals' decision cannot be squared with our recognition that, in enacting § 1983, Congress did not intend to impose liability on a municipality unless deliberate action attributable to the municipality itself is the "moving force" behind the plaintiff's deprivation of federal rights. Monell v. New York City Dept. of Social Servs., 436 U.S. 658, 694 (1978).

I

In the early morning hours of May 12, 1991, Jill Brown (hereinafter respondent) and her husband were driving from Grayson County, Texas, to their home in Bryan County, Oklahoma. After crossing into Oklahoma, they approached a police checkpoint. Mr. Brown, who was driving, decided to avoid the checkpoint and return to Texas. After seeing the Browns' truck turn away from the checkpoint, Bryan County Deputy Sheriff Robert Morrison and Reserve Deputy Stacy Burns pursued the vehicle. Although the parties' versions of events differ, at trial both deputies claimed that their patrol car reached speeds in excess of 100 miles per hour. Mr. Brown testified that he was unaware of the deputies' attempts to overtake him. The chase finally ended four miles south of the police checkpoint.

After he got out of the squad car, Deputy Sheriff Morrison pointed his gun toward the Browns' vehicle and ordered the Browns to raise their hands. Reserve Deputy Burns, who was unarmed, rounded the corner of the vehicle on the passenger's side. Burns twice ordered respondent from the vehicle. When she did not exit, he used an "arm bar" technique, grabbing respondent's arm at the wrist and elbow, pulling her from the vehicle, and spinning her to the ground. Respondent's knees were severely injured, and she later underwent corrective surgery. Ultimately, she may need knee replacements.

Respondent sought compensation for her injuries under 42 U. S. C. § 1983 and state law from Burns, Bryan County Sheriff B. J. Moore, and the county itself. Respondent claimed, among other things, that Bryan County was liable for Burns' alleged use of excessive force based on Sheriff Moore's decision to hire Burns, the son of his nephew. Specifically, respondent claimed that Sheriff Moore had failed to adequately review Burns' background. Burns had a record of driving infractions and had pleaded guilty to various driving-related and other misdemeanors, including assault and battery, resisting arrest, and public drunkenness. Oklahoma law does not preclude the hiring of an individual who has committed a misdemeanor to serve as a peace officer. See Okla. Stat., Tit. 70, § 3311(D)(2)(a) (1991) (requiring that the hiring agency certify that the prospective officer's records do not reflect a felony conviction). At trial, Sheriff Moore testified that he had obtained Burns' driving record and a report on Burns from the National Crime Information Center, but had not closely reviewed either. Sheriff Moore authorized Burns to make arrests, but not to carry a weapon or to operate a patrol car.

In a ruling not at issue here, the District Court dismissed respondent's § 1983 claim against Sheriff Moore prior to trial. App. 28. Counsel for Bryan County stipulated that Sheriff Moore "was the policy maker for Bryan County regarding the Sheriff's Department." Id., at 30. At the close of respondent's case and again at the close of all of the evidence, Bryan County moved for judgment as a matter of law. As to respondent's claim that Sheriff Moore's decision to hire Burns triggered municipal liability, the county argued that a single hiring decision by a municipal policymaker could not give rise to municipal liability under § 1983. Id., at 59-60. The District Court denied the county's motions. The court also overruled the county's objections to jury instructions on the § 1983 claim against the county. Id., at 125-126, 132.

To resolve respondent's claims, the jury was asked to answer several interrogatories. The jury concluded that Stacy Burns had arrested respondent without probable cause and had used excessive force, and therefore found him liable for respondent's injuries. It also found that the "hiring policy" and the "training policy" of Bryan County "in the case of Stacy Burns as instituted by its policymaker, B. J. Moore," were each "so inadequate as to amount to deliberate indifference to the constitutional needs of the Plaintiff." Id., at 135. The District Court entered judgment for respondent on the issue of Bryan County's § 1983 liability. The county appealed on several grounds, and the Court of Appeals for the Fifth Circuit affirmed. 67 F.3d 1174 (1995). The court held, among other things, that Bryan County was properly found liable under § 1983 based on Sheriff Moore's decision to hire Burns. Id., at 1185. The court addressed only those points that it thought merited review; it did not address the jury's determination of county liability based on inadequate training of Burns, id., at 1178, nor do we. We granted certiorari, 517 U.S. 1154 (1996), to decide whether the county was properly held liable for respondent's injuries based on Sheriff Moore's single decision to hire Burns. We now reverse.

II

Title 42 U. S. C. § 1983 provides in relevant part:

We held in Monell v. New York City Dept. of Social Servs., 436 U. S., at 689, that municipalities and other local governmental bodies are "persons" within the meaning of § 1983. We also recognized that a municipality may not be held liable under § 1983 solely because it employs a tortfeasor. Our conclusion rested partly on the language of § 1983 itself. In light of the statute's imposition of liability on one who "subjects [a person], or causes [that person] to be subjected," to a deprivation of federal rights, we concluded that it "cannot be easily read to impose liability vicariously on governing bodies solely on the basis of the existence of an employeremployee relationship with a tortfeasor." Id., at 692. Our conclusion also rested upon the statute's legislative history. As stated in Pembaur v. Cincinnati, 475 U.S. 469, 479 (1986), "while Congress never questioned its power to impose civil liability on municipalities for their own illegal acts, Congress did doubt its constitutional power to impose such liability in order to oblige municipalities to control the conduct of others " (citing Monell, supra, at 665-683). We have consistently refused to hold municipalities liable under a theory of respondeat superior. See Oklahoma City v. Tuttle, 471 U.S. 808, 818 (1985) (plurality opinion); id., at 828 (opinion of Brennan, J.); Pembaur, supra, at 478-479; St. Louis v. Praprotnik, 485 U.S. 112, 122 (1988) (plurality opinion); id., at 137 (opinion of Brennan, J.);Canton v. Harris, 489 U.S. 378, 392 (1989).

Instead, in Monell and subsequent cases, we have required a plaintiff seeking to impose liability on a municipality under § 1983 to identify a municipal "policy" or "custom" that caused the plaintiff's injury. See Monell, supra, at 694; Pembaur, supra, at 480-481; Canton, supra, at 389. Locating a "policy" ensures that a municipality is held liable only for those deprivations resulting from the decisions of its duly constituted legislative body or of those officials whose acts may fairly be said to be those of the municipality. Monell, supra, at 694. Similarly, an act performed pursuant to a "custom" that has not been formally approved by an appropriate decisionmaker may fairly subject a municipality to liability on the theory that the relevant practice is so widespread as to have the force of law. 436 U. S., at 690-691 (citing Adickes v. S. H. Kress & Co., 398 U.S. 144, 167-168 (1970)).

The parties join issue on whether, under Monell and subsequent cases, a single hiring decision by a county sheriff can be a "policy" that triggers municipal liability. Relying on our decision in Pembaur, respondent claims that a single act by a decisionmaker with final authority in the relevant area constitutes a "policy" attributable to the municipality itself. So long as a § 1983 plaintiff identifies a decision properly attributable to the municipality, respondent argues, there is no risk of imposing respondeat superior liability. Whether that decision was intended to govern only the situation at hand or to serve as a rule to be applied over time is immaterial. Rather, under respondent's theory, identification of an act of a proper municipal decisionmaker is all that is required to ensure that the municipality is held liable only for its own conduct. The Court of Appeals accepted respondent's approach.

As our § 1983 municipal liability jurisprudence illustrates, however, it is not enough for a § 1983 plaintiff merely to identify conduct properly attributable to the municipality. The plaintiff must also demonstrate that, through its deliberate conduct, the municipality was the "moving force" behind the injury alleged. That is, a plaintiff must show that the municipal action was taken with the requisite degree of culpability and must demonstrate a direct causal link between the municipal action and the deprivation of federal rights.

Where a plaintiff claims that a particular municipal action itself violates federal law, or directs an employee to do so, resolving these issues of fault and causation is straight forward. Section 1983 itself "contains no state-of-mind requirement independent of that necessary to state a violation" of the underlying federal right. Daniels v. Williams, 474 U.S. 327, 330 (1986). In any § 1983 suit, however, the plaintiff must establish the state of mind required to prove the underlying violation. Accordingly, proof that a municipality's legislative body or authorized decisionmaker has intentionally deprived a plaintiff of a federally protected right necessarily establishes that the municipality acted culpably. Similarly, the conclusion that the action taken or directed by the municipality or its authorized decisionmaker itself violates federal law will also determine that the municipal action was the moving force behind the injury of which the plaintiff complains.

Sheriff Moore's hiring decision was itself legal, and Sheriff Moore did not authorize Burns to use excessive force. Respondent's claim, rather, is that a single facially lawful hiring decision can launch a series of events that ultimately cause a violation of federal rights. Where a plaintiff claims that the municipality has not directly inflicted an injury, but nonetheless has caused an employee to do so, rigorous standards of culpability and causation must be applied to ensure that the municipality is not held liable solely for the actions of its employee. See Canton, supra, at 391-392; Tuttle, supra, at 824 (plurality opinion). See also Springfield v. Kibbe, 480 U.S. 257, 270-271 (1987) (per curiam) (dissent from dismissal of writ as improvidently granted).

In relying heavily on Pembaur, respondent blurs the distinction between § 1983 cases that present no difficult questions of fault and causation and those that do. To the extent that we have recognized a cause of action under § 1983 based on a single decision attributable to a municipality, we have done so only where the evidence that the municipality had acted and that the plaintiff had suffered a deprivation of federal rights also proved fault and causation. For example, Owen v. Independence, 445 U.S. 622 (1980), and Newport v. Fact Concerts, Inc., 453 U.S. 247 (1981), involved formal decisions of municipal legislative bodies. In Owen, the city council allegedly censured and discharged an employee without a hearing. 445 U. S., at 627-629, 633, and n. 13. In Fact Concerts, the city council canceled a license permitting a concert following a dispute over the performance's content. 453 U. S., at 252. Neither decision reflected implementation of a generally applicable rule. But we did not question that each decision, duly promulgated by city lawmakers, could trigger municipal liability if the decision itself were found to be unconstitutional. Because fault and causation were obvious in each case, proof that the municipality's decision was unconstitutional would suffice to establish that the municipality itself was liable for the plaintiff's constitutional injury.

Similarly, Pembaur v. Cincinnati concerned a decision by a county prosecutor, acting as the county's final decisionmaker, 475 U. S., at 485, to direct county deputies to forcibly enter petitioner's place of business to serve capiases upon third parties. Relying on Owen and Newport, we concluded that a final decisionmaker's adoption of a course of action "tailored to a particular situation and not intended to control decisions in later situations" may, in some circumstances, give rise to municipal liability under § 1983. 475 U. S., at 481. In Pembaur, it was not disputed that the prosecutor had specifically directed the action resulting in the deprivation of petitioner's rights. The conclusion that the decision was that of a final municipal decisionmaker and was therefore properly attributable to the municipality established municipal liability. No questions of fault or causation arose.

Claims not involving an allegation that the municipal action itself violated federal law, or directed or authorized the deprivation of federal rights, present much more difficult problems of proof. That a plaintiff has suffered a deprivation of federal rights at the hands of a municipal employee will not alone permit an inference of municipal culpability and causation; the plaintiff will simply have shown that the employee acted culpably. We recognized these difficulties in Canton v. Harris, where we considered a claim that inadequate training of shift supervisors at a city jail led to a deprivation of a detainee's constitutional rights. We held that, quite apart from the state of mind required to establish the underlying constitutional violation—in that case, a violation of due process, 489 U. S., at 388-389, n. 8—a plaintiff seeking to establish municipal liability on the theory that a facially lawful municipal action has led an employee to violate a plaintiff's rights must demonstrate that the municipal action was taken with "deliberate indifference" as to its known or obvious consequences. Id., at 388. A showing of simple or even heightened negligence will not suffice.

We concluded in Canton that an "inadequate training" claim could be the basis for § 1983 liability in "limited circumstances." Id., at 387. We spoke, however, of a deficient training "program," necessarily intended to apply over time to multiple employees. Id., at 390. Existence of a "program" makes proof of fault and causation at least possible in an inadequate training case. If a program does not prevent constitutional violations, municipal decisionmakers may eventually be put on notice that a new program is called for. Their continued adherence to an approach that they know or should know has failed to prevent tortious conduct by employees may establish the conscious disregard for the consequences of their action—the "deliberate indifference"—necessary to trigger municipal liability. Id., at 390, n. 10 ("It could . . . be that the police, in exercising their discretion, so often violate constitutional rights that the need for further training must have been plainly obvious to the city policymakers, who, nevertheless, are `deliberately indifferent' to the need"); id., at 397 (O'Connor, J., concurring in part and dissenting in part) ("[M]unicipal liability for failure to train may be proper where it can be shown that policymakers were aware of, and acquiesced in, a pattern of constitutional violations . . ."). In addition, the existence of a pattern of tortious conduct by inadequately trained employees may tend to show that the lack of proper training, rather than a one-time negligent administration of the program or factors peculiar to the officer involved in a particular incident, is the "moving force" behind the plaintiff's injury. See id., at 390-391.

Before trial, counsel for Bryan County stipulated that Sheriff Moore "was the policy maker for Bryan County regarding the Sheriff's Department." App. 30. Indeed, the county sought to avoid liability by claiming that its Board of Commissioners participated in no policy decisions regarding the conduct and operation of the office of the Bryan County Sheriff. Id., at 32. Accepting the county's representations below, then, this case presents no difficult questions concerning whether Sheriff Moore has final authority to act for the municipality in hiring matters. Cf. Jett v. Dallas Independent School Dist., 491 U.S. 701 (1989); St. Louis v. Praprotnik, 485 U.S. 112 (1988). Respondent does not claim that she can identify any pattern of injuries linked to Sheriff Moore's hiring practices. Indeed, respondent does not contend that Sheriff Moore's hiring practices are generally defective. The only evidence on this point at trial suggested that Sheriff Moore had adequately screened the backgrounds of all prior deputies he hired. App. 106-110. Respondent instead seeks to trace liability to what can only be described as a deviation from Sheriff Moore's ordinary hiring practices. Where a claim of municipal liability rests on a single decision, not itself representing a violation of federal law and not directing such a violation, the danger that a municipality will be held liable without fault is high. Because the decision necessarily governs a single case, there can be no notice to the municipal decisionmaker, based on previous violations of federally protected rights, that his approach is inadequate. Nor will it be readily apparent that the municipality's action caused the injury in question, because the plaintiff can point to no other incident tending to make it more likely that the plaintiff's own injury flows from the municipality's action, rather than from some other intervening cause.

In Canton, we did not foreclose the possibility that evidence of a single violation of federal rights, accompanied by a showing that a municipality has failed to train its employees to handle recurring situations presenting an obvious potential for such a violation, could trigger municipal liability. 489 U. S., at 390, and n. 10 ("[I]t may happen that in light of the duties assigned to specific officers or employees the need for more or different training is so obvious . . . that the policymakers of the city can reasonably be said to have been deliberately indifferent to the need"). Respondent purports to rely on Canton, arguing that Burns' use of excessive force was the plainly obvious consequence of Sheriff Moore's failure to screen Burns' record. In essence, respondent claims that this showing of "obviousness" would demonstrate both that Sheriff Moore acted with conscious disregard for the consequences of his action and that the Sheriff's action directly caused her injuries, and would thus substitute for the pattern of injuries ordinarily necessary to establish municipal culpability and causation.

The proffered analogy between failure-to-train cases and inadequate screening cases is not persuasive. In leaving open in Canton the possibility that a plaintiff might succeed in carrying a failure-to-train claim without showing a pattern of constitutional violations, we simply hypothesized that, in a narrow range of circumstances, a violation of federal rights may be a highly predictable consequence of a failure to equip law enforcement officers with specific tools to handle recurring situations. The likelihood that the situation will recur and the predictability that an officer lacking specific tools to handle that situation will violate citizens' rights could justify a finding that policymakers' decision not to train the officer reflected "deliberate indifference" to the obvious consequence of the policymakers' choice—namely, a violation of a specific constitutional or statutory right. The high degree of predictability may also support an inference of causation— that the municipality's indifference led directly to the very consequence that was so predictable.

Where a plaintiff presents a § 1983 claim premised upon the inadequacy of an official's review of a prospective applicant's record, however, there is a particular danger that a municipality will be held liable for an injury not directly caused by a deliberate action attributable to the municipality itself. Every injury suffered at the hands of a municipal employee can be traced to a hiring decision in a "but-for" sense: But for the municipality's decision to hire the employee, the plaintiff would not have suffered the injury. To prevent municipal liability for a hiring decision from collapsing into respondeat superior liability, a court must carefully test the link between the policymaker's inadequate decision and the particular injury alleged.

In attempting to import the reasoning of Canton into the hiring context, respondent ignores the fact that predicting the consequence of a single hiring decision, even one based on an inadequate assessment of a record, is far more difficult than predicting what might flow from the failure to train a single law enforcement officer as to a specific skill necessary to the discharge of his duties. As our decision in Canton makes clear, "deliberate indifference" is a stringent standard of fault, requiring proof that a municipal actor disregarded a known or obvious consequence of his action. Unlike the risk from a particular glaring omission in a training regimen, the risk from a single instance of inadequate screening of an applicant's background is not "obvious" in the abstract; rather, it depends upon the background of the applicant. A lack of scrutiny may increase the likelihood that an unfit officer will be hired, and that the unfit officer will, when placed in a particular position to affect the rights of citizens, act improperly. But that is only a generalized showing of risk. The fact that inadequate scrutiny of an applicant's background would make a violation of rights more likely cannot alone give rise to an inference that a policymaker's failure to scrutinize the record of a particular applicant produced a specific constitutional violation. After all, a full screening of an applicant's background might reveal no cause for concern at all; if so, a hiring official who failed to scrutinize the applicant's background cannot be said to have consciously disregarded an obvious risk that the officer would subsequently inflict a particular constitutional injury.

We assume that a jury could properly find in this case that Sheriff Moore's assessment of Burns' background was inadequate. Sheriff Moore's own testimony indicated that he did not inquire into the underlying conduct or the disposition of any of the misdemeanor charges reflected on Burns' record before hiring him. But this showing of an instance of inadequate screening is not enough to establish "deliberate indifference." In layman's terms, inadequate screening of an applicant's record may reflect "indifference" to the applicant's background. For purposes of a legal inquiry into municipal liability under § 1983, however, that is not the relevant "indifference." A plaintiff must demonstrate that a municipal decision reflects deliberate indifference to the risk that a violation of a particular constitutional or statutory right will follow the decision. Only where adequate scrutiny of an applicant's background would lead a reasonable policymaker to conclude that the plainly obvious consequence of the decision to hire the applicant would be the deprivation of a third party's federally protected right can the official's failure to adequately scrutinize the applicant's background constitute "deliberate indifference."

Neither the District Court nor the Court of Appeals directly tested the link between Burns' actual background and the risk that, if hired, he would use excessive force. The District Court instructed the jury on a theory analogous to that reserved in Canton. The court required respondent to prove that Sheriff Moore's inadequate screening of Burns' background was "so likely to result in violations of constitu- tional rights " that the Sheriff could "reasonably [be] said to have been deliberately indifferent to the constitutional needs of the Plaintiff." App. 123 (emphasis added). The court also instructed the jury, without elaboration, that respondent was required to prove that the "inadequate hiring. . . policy directly caused the Plaintiff's injury." Ibid.

As discussed above, a finding of culpability simply cannot depend on the mere probability that any officer inadequately screened will inflict any constitutional injury. Rather, it must depend on a finding that this officer was highly likely to inflict the particular injury suffered by the plaintiff. The connection between the background of the particular applicant and the specific constitutional violation alleged must be strong. What the District Court's instructions on culpability, and therefore the jury's finding of municipal liability, failed to capture is whether Burns' background made his use of excessive force in making an arrest a plainly obvious consequence of the hiring decision. The Court of Appeals' affirmance of the jury's finding of municipal liability depended on its view that the jury could have found that "inadequate screening of a deputy could likely result in the violation of citizens' constitutional rights. " 67 F. 3d, at 1185 (emphasis added). Beyond relying on a risk of violations of unspecified constitutional rights, the Court of Appeals also posited that Sheriff Moore's decision reflected indifference to "the public's welfare." Id., at 1184.

Even assuming without deciding that proof of a single instance of inadequate screening could ever trigger municipal liability, the evidence in this case was insufficient to support a finding that, in hiring Burns, Sheriff Moore disregarded a known or obvious risk of injury. To test the link between Sheriff Moore's hiring decision and respondent's injury, we must ask whether a full review of Burns' record reveals that Sheriff Moore should have concluded that Burns' use of excessive force would be a plainly obvious consequence of the hiring decision.1 On this point, respondent's showing was inadequate. To be sure, Burns' record reflected various misdemeanor infractions. Respondent claims that the record demonstrated such a strong propensity for violence that Burns' application of excessive force was highly likely. The primary charges on which respondent relies, however, are those arising from a fight on a college campus where Burns was a student. In connection with this single incident, Burns was charged with assault and battery, resisting arrest, and public drunkenness.2 In January 1990, when he pleaded guilty to those charges, Burns also pleaded guilty to various driving-related offenses, including nine moving violations and a charge of driving with a suspended license. In addition, Burns had previously pleaded guilty to being in actual physical control of a vehicle while intoxicated.

The fact that Burns had pleaded guilty to traffic offenses and other misdemeanors may well have made him an extremely poor candidate for reserve deputy. Had Sheriff Moore fully reviewed Burns' record, he might have come to precisely that conclusion. But unless he would necessarily have reached that decision because Burns' use of excessive force would have been a plainly obvious consequence of the hiring decision, Sheriff Moore's inadequate scrutiny of Burns' record cannot constitute "deliberate indifference" to respondent's federally protected right to be free from a use of excessive force.

Justice Souter's reading of the case is that the jury believed that Sheriff Moore in fact read Burns' entire record. Post, at 426-427. That is plausible, but it is also irrelevant. It is not sufficient for respondent to show that Sheriff Moore read Burns' record and therefore hired Burns with knowledge of his background. Such a decision may reflect indifference to Burns' record, but what is required is deliberate indifference to a plaintiff's constitutional right. That is, whether Sheriff Moore failed to examine Burns' record, partially examined it, or fully examined it, Sheriff Moore's hiring decision could not have been "deliberately indifferent" unless in light of that record Burns' use of excessive force would have been a plainly obvious consequence of the hiring decision. Because there was insufficient evidence on which a jury could base a finding that Sheriff Moore's decision to hire Burns reflected conscious disregard of an obvious risk that a use of excessive force would follow, the District Court erred in submitting respondent's inadequate screening claim to the jury.

III

Cases involving constitutional injuries allegedly traceable to an ill-considered hiring decision pose the greatest risk that a municipality will be held liable for an injury that it did not cause. In the broadest sense, every injury is traceable to a hiring decision. Where a court fails to adhere to rigorous requirements of culpability and causation, municipal liability collapses into respondeat superior liability. As we recognized in Monell and have repeatedly reaffirmed, Congress did not intend municipalities to be held liable unless deliberate action attributable to the municipality directly caused a deprivation of federal rights. A failure to apply stringent culpability and causation requirements raises serious federalism concerns, in that it risks constitutionalizing particular hiring requirements that States have themselves elected not to impose. Cf. Canton v. Harris, 489 U. S., at 392. Bryan County is not liable for Sheriff Moore's isolated decision to hire Burns without adequate screening, because respondent has not demonstrated that his decision reflected a conscious disregard for a high risk that Burns would use excessive force in violation of respondent's federally protected right. We therefore vacate the judgment of the Court of Appeals and remand this case for further proceedings consistent with this opinion.

It is so ordered.

Justice Souter, with whom Justice Stevens and Justice Breyer join, dissenting.

In Pembaur v. Cincinnati, 475 U.S. 469, 480 (1986), we held a municipality liable under 42 U. S. C. § 1983 for harm caused by the single act of a policymaking officer in a matter within his authority but not covered by a policy previously identified. The central question presented here is whether that rule applies to a single act that itself neither violates nor commands a violation of federal law. The answer is yes if the single act amounts to deliberate indifference to a substantial risk that a violation of federal law will result. With significant qualifications, the Court assumes so, too, in theory, but it raises such skeptical hurdles to reaching any such conclusion in practice that it virtually guarantees its disposition of this case: it holds as a matter of law that the sheriff's act could not be thought to reflect deliberate indifference to the risk that his subordinate would violate the Constitution by using excessive force. I respectfully dissent as much from the level of the Court's skepticism as from its reversal of the judgment.

I

Monell v. New York City Dept. of Social Servs., 436 U.S. 658 (1978), overruled Monroe v. Pape, 365 U.S. 167 (1961), insofar as Monroe had held § 1983 inapplicable to governments beneath the state level ("municipal," for short). Monell, supra, at 663. At the same time that we decided Congress meant municipalities to be persons subject to § 1983, however, we also concluded that municipal liability under the statute could not be based on the traditional theory of respondeat superior. 436 U. S., at 691. We said that for purposes of § 1983 an act could not be attributed to a municipality merely because it was an act of a municipal agent performed in the course of exercising a power delegated to the municipality by local law, and we reasoned instead that "it is [only] when execution of a government's policy or custom, whether made by its lawmakers or by those whose edicts or acts may fairly be said to represent official policy, inflicts the injury that the government as an entity is responsible under § 1983." Id., at 694; see Pembaur, supra, at 480.

In assigning municipal liability under Monell, we accordingly distinguish an act of a municipal agent without independent authority to establish policy from the act of one authorized to set policy under local law, and we likewise distinguish the acts of lower level employees depending on whether they do or do not implement or manifest a policy set by those with the authority to set it. The act of the municipality is the act only of an authorized policymaker or of an employee following the policymaker's lead. "The `official policy' requirement was intended to distinguish acts of the municipality from acts of employees of the municipality, and thereby make clear that municipal liability is limited to action for which the municipality is actually responsible." Pembaur, supra, at 479-480.

While this overview indicates that the policy requirement may be satisfied in more than one way, there are in fact three alternatives discernible in our prior cases. It is certainly met when the appropriate officer or entity promulgates a generally applicable statement of policy and the subsequent act complained of is simply an implementation of that policy. Monell exemplified these circumstances, where city agencies had issued a rule requiring pregnant employees to take unpaid leaves of absence before any medical need arose. Monell, supra, at 660-661.

We have also held the policy requirement satisfied where no rule has been announced as "policy" but federal law has been violated by an act of the policymaker itself. In this situation, the choice of policy and its implementation are one, and the first or only action will suffice to ground municipal liability simply because it is the very policymaker who is acting. See Pembaur, supra, at 480-481; cf. Newport v. Fact Concerts, Inc., 453 U.S. 247, 250-252 (1981) (implicitly assuming that a policymaker's single act can sustain § 1983 action); Owen v. Independence, 445 U.S. 622, 625-630 (1980) (same). It does not matter that the policymaker may have chosen "a course of action tailored [only] to a particular situation and not intended to control decisions in later situations," Pembaur, 475 U. S., at 481; if the decision to adopt that particular course of action is intentionally made by the authorized policymaker, "it surely represents an act of official government `policy' " and "the municipality is equally responsible whether that action is to be taken only once or to be taken repeatedly," ibid.

We have, finally, identified a municipal policy in a third situation, even where the policymaker has failed to act affirmatively at all, so long as the need to take some action to control the agents of the government "is so obvious, and the inadequacy [of existing practice] so likely to result in the violation of constitutional rights, that the policymake[r] . . . can reasonably be said to have been deliberately indifferent to the need." Canton v. Harris, 489 U.S. 378, 390 (1989). Where, in the most obvious example, the policymaker sits on his hands after repeated, unlawful acts of subordinate officers and that failure "evidences a `deliberate indifference' to the rights of [the municipality's] inhabitants," id., at 389, the policymaker's toleration of the subordinates' behavior establishes a policy-in-practice just as readily attributable to the municipality as the one-act policy-in-practice described above. Such a policy choice may be inferred even without a pattern of acts by subordinate officers, so long as the need for action by the policymaker is so obvious that the failure to act rises to deliberate indifference. Id., at 390, n. 10.

Deliberate indifference is thus treated, as it is elsewhere in the law,1 as tantamount to intent, so that inaction by a policymaker deliberately indifferent to a substantial risk of harm is equivalent to the intentional action that setting policy presupposes. Compare Pembaur, supra, at 483 (plurality opinion of Brennan, J.) ("deliberate choice" by policymaker), and Oklahoma City v. Tuttle, 471 U.S. 808, 823 (1985) (plurality opinion of Rehnquist, J.) ("`policy' generally implies a course of action consciously chosen"), with Canton, supra, at 389 ("Only where a municipality's failure to train its employees . . . evidences a `deliberate indifference' to the rights of its inhabitants can . . . a shortcoming be . . . city `policy or custom' . . . actionable under § 1983").

Under this prior law, Sheriff Moore's failure to screen out his 21-year-old great-nephew Burns on the basis of his criminal record, and the decision instead to authorize Burns to act as a deputy sheriff, constitutes a policy choice attributable to Bryan County under § 1983. There is no serious dispute that Sheriff Moore is the designated policymaker for implementing the sheriff's law enforcement powers and recruiting officers to exercise them, or that he "has final authority to act for the municipality in hiring matters." Ante, at 408. As the authorized policymaker, Sheriff Moore is the county for purposes of § 1983 municipal liability arising from the sheriff's department's exercise of law enforcement authority. As I explain in greater detail below, it was open to the jury to find that the sheriff knew of the record of his nephew's violent propensity, but hired him in deliberate indifference to the risk that he would use excessive force on the job, as in fact he later did. That the sheriff's act did not itself command or require commission of a constitutional violation (like the order to perform an unlawful entry and search in Pembaur ) is not dispositive under § 1983, for we have expressly rejected the contention that "only unconstitutional policies are actionable" under § 1983, Canton, supra, at 387, and have never suggested that liability under the statute is otherwise limited to policies that facially violate other federal law. The sheriff's policy choice creating a substantial risk of a constitutional violation therefore could subject the county to liability under existing precedent.2

II

At the level of theory, at least, the Court does not disagree, and it assumes for the sake of deciding the case that a single, facially neutral act of deliberate indifference by a policymaker could be a predicate to municipal liability if it led to an unconstitutional injury inflicted by subordinate officers. See ante, at 412. At the level of practice, however, the tenor of the Court's opinion is decidedly different: it suggests that the trial court insufficiently appreciated the specificity of the risk to which such indifference must be deliberate in order to be actionable; it expresses deep skepticism that such appreciation of risk could ever reasonably be attributed to the policymaker who has performed only a single unsatisfactory, but not facially unconstitutional, act; and it finds the record insufficient to make any such showing in this case. The Court is serially mistaken. This case presents no occasion to correct or refine the District Court's jury instructions on the degree of risk required for deliberate indifference; the Court's skepticism converts a newly demanding formulation of the standard of fault into a virtually categorical impossibility of showing it in a case like this; and the record in this case is perfectly sufficient to support the jury's verdict even on the Court's formulation of the high degree of risk that must be shown.

A

The Court is certainly correct in emphasizing the need to show more than mere negligence on the part of the policymaker, for at the least the element of deliberateness requires both subjective appreciation of a risk of unconstitutional harm, and a risk substantial enough to justify the heightened responsibility that deliberate indifference generally entails. The Court goes a step further, however, in requiring that the "particular" harmful consequence be "plainly obvious" to the policymaker, ante, at 411, a characterization of deliberate indifference adapted from dicta set forth in a footnote in Canton, see 489 U. S., at 390, n. 10. Canton, as mentioned above, held that a municipal policy giving rise to liability under § 1983 may be inferred even when the policymaker has failed to act affirmatively at all, so long as a need to control the agents of the government "is so obvious, and the inadequacy [of existing practice] so likely to result in the violation of constitutional rights, that the policymake[r] . . . can reasonably be said to have been deliberately indifferent to the need." Id. , at 390. While we speculated in Canton that "[i]t could . . . be that the police, in exercising their discretion, so often violate constitutional rights that the need for further training must have been plainly obvious to the city policymakers, who, nevertheless, are `deliberately indifferent' to the need," see id. , at 390, n. 10, we did not purport to be defining the fault of deliberate indifference universally as the failure to act in relation to a "plainly obvious consequence" of harm. Nor did we, in addressing the requisite risk that constitutional violations will occur, suggest that the deliberate indifference necessary to establish municipal liability must be, as the Court says today, indifference to the particular constitutional violation that in fact occurred.

The Court's formulation that deliberate indifference exists only when the risk of the subsequent, particular constitutional violation is a plainly obvious consequence of the hiring decision, see ante, at 411, while derived from Canton, is thus without doubt a new standard. See post, at 433-434 (Breyer, J., dissenting). As to the "particular" violation, the Court alters the understanding of deliberate indifference as set forth in Canton, where we spoke of constitutional violations generally.3 As to "plainly obvious consequence," the Court's standard appears to be somewhat higher, for example, than the standard for "reckless" fault in the criminal law, where the requisite indifference to risk is defined as that which "consciously disregards a substantial and unjustifiable risk that the material element exists or will result . . . [and] involves a gross deviation from the standard of conduct that a law-abiding person would observe in the actor's situation." See American Law Institute, Model Penal Code § 2.02(2)(c) (1985).

That said, it is just possible that our prior understanding of the requisite degree of fault and the standard as the Court now states it may in practice turn out to amount to much the same thing, but I would have preferred an argument addressing the point before ruling on it. There was, however, no such argument here for the simple reason that petitioner never asked that deliberate indifference be defined to occur only when the particular constitutional injury was the plainly obvious consequence of the policymaker's act. Petitioner merely asked the District Court to instruct the jury to determine whether Sheriff Moore acted with "conscious indifference," see 2 Record 342, and made no objection to the District Court's charge that "Sheriff B. J. Moore would have acted with deliberate indifference in adopting an otherwise constitutional hiring policy for a deputy sheriff if the need for closer scrutiny of Stacy Burns' background was so obvious and the inadequacy of the scrutiny given so likely to result in violations of constitutional rights, that Sheriff B. J. Moore can be reasonably said to have been deliberately indifferent to the constitutional needs of the Plaintiff." 10 Record 800-801. If, as it appears, today's standard does raise the threshold of municipal liability, it does so quite independently of any issue posed or decided in the trial court.

B

The Court's skepticism that the modified standard of fault can ever be met in a single-act case of inadequate screening without a patently unconstitutional policy, ante, at 412-414, both reveals the true value of the assumption that in theory there might be municipal liability in such a case, and dictates the result of the Court's review of the record in the case before us. It is skepticism gone too far.

It is plain enough that a facially unconstitutional policy is likely to produce unconstitutional injury, see, e. g., Pembaur, 475 U. S., at 480-481; Monell, 436 U. S., at 660-661, and obvious, too, that many facially neutral policy decisions evince no such clear portents. Written standards for hiring law enforcement personnel might be silent on the significance of a prior criminal record without justifying much worry about employing axe murderers (who are unlikely to apply) or subjecting the public to attacks by someone with a 30-year-old assault conviction (who has probably grown up). But a policymaker need not mandate injury to be indifferent to its risk when obvious, and, because a particular hiring decision may raise a very high probability of harm down the line, it simply ignores the issue before us to lump together in one presumptively benign category every singular administrative act of a policymaker that does not expressly command or constitute unconstitutional behavior. Thus, a decision to give law enforcement authority to a scofflaw who had recently engaged in criminal violence presents a very different risk from hiring someone who once drove an overweight truck. While the decision to hire the violent scofflaw may not entail harm to others as unquestionably as an order to "go out and rough-up some suspects," it is a long way from neutral in the risk it creates.

While the Court should rightly be skeptical about predicating municipal or individual liability merely on a failure to adopt a crime-free personnel policy or on a particular decision to hire a guilty trucker, why does it extend that valid skepticism to the quite unsound point of doubting liability for hiring the violent scofflaw? The Court says it fears that the latter sort of case raises a danger of liability without fault, ante, at 408. But if the Court means fault generally (as distinct from the blame imputed on classic respondeat superior doctrine), it need only recall that whether a particular violent scofflaw is violent enough or scoffing enough to implicate deliberate indifference will depend on applying the highly demanding standard the Court announces: plainly obvious consequence of particular injury. It is the high threshold of deliberate indifference that will ensure that municipalities be held liable only for considered acts with substantial risks. That standard will distinguish single-act cases with only a mild portent of injury from single-act cases with a plainly obvious portent, and from cases in which the harm is only the latest in a series of injuries known to have followed from the policymaker's action. The Court has fenced off the slippery slope.

A second stated reason of the skeptical majority is that, because municipal liability under Monell cannot rest on respondeat superior, ante, at 410, 415, "a court must carefully test the link between the policymaker's inadequate decision and the particular injury alleged," ante, at 410. But that is simply to say that the tortious act must be proximately caused by the policymaker. The policy requirement is the restriction that bars straightforward respondeat superior liability, and the need to "test the link" is merely the need to apply the law that defines what a cognizable link is. The restriction on imputed fault that saves municipalities from liability has no need of categorical immunization in singleact cases.

In short, the Court's skepticism is excessive in ignoring the fact that some acts of a policymaker present substantial risks of unconstitutional harm even though the acts are not unconstitutional per se. And the Court's purported justifications for its extreme skepticism are washed out by the very standards employed to limit liability.

C

For demonstrating the extreme degree of the Court's inhospitality to single-act municipal liability, this is a case on point, for even under the "plainly obvious consequence" rule the evidence here would support the verdict. There is no dispute that before the incident in question the sheriff ordered a copy of his nephew's criminal record. While the sheriff spoke euphemistically on the witness stand of a "driving record," the scope of the requested documentation included crimes beyond motor vehicle violations and the sheriff never denied that he knew this. He admitted that he read some of that record; he said he knew it was "long"; he said he was sure he had noticed charges of driving with a suspended license; and he said that he had taken the trouble to make an independent search for any outstanding warrant for Burns's arrest. As he put it, however, he somehow failed to "notice" charges of assault and battery or the list of offenses so long as to point either to contempt for law or to incapacity to obey it. Although the jury might have accepted the sheriff's disclaimer, no one who has read the transcript would assume that the jurors gave any credit to that testimony,4 and it was open to them to find that the sheriff was simply lying under oath about his limited perusal. The Court of Appeals noted this possibility, see 67 F.3d 1174, 1184 (CA5 1995), which is more likely than any other reading of the evidence. Law enforcement officers, after all, are not characteristically so devoid of curiosity as to lose interest part way through the criminal record of a subject of personal investigation.

If, as is likely, the jurors did disbelieve the sheriff and concluded he had read the whole record, they certainly could have eliminated any possibility that the sheriff's decision to employ his relative was an act of mere negligence or poor judgment. He did not even claim, for example, that he thought any assault must have been just a youthful peccadillo magnified out of proportion by the criminal charge, or that he had evaluated the assault as merely eccentric behavior in a young man of sound character, or that he was convinced that wild youth had given way to discretion. There being no such evidence of reasonable but mistaken judgment, the jury could readily have found that the sheriff knew his nephew's proven propensities, that he thought the thrust of the evidence was so damaging that he would lie to protect his reputation and the county treasury, and that he simply chose to put a family member on the payroll (the third relative, in fact5) disregarding the risk to the public.

At trial, petitioner's expert witness stated during crossexamination that Burns's rap sheet listed repeated traffic violations, including driving while intoxicated and driving with a suspended license, resisting arrest, and more than one charge of assault and battery. The witness further testified that Burns pleaded guilty to assault and battery and other charges 16 months before he was hired by Sheriff Moore.6 Respondent's expert witness testified that Burns's arrest record showed a "blatant disregard for the law and problems that may show themselves in abusing the public or using excessive force," 7 Record 316, and petitioner's own expert agreed that Burns's criminal history should have caused concern. When asked if he would have hired Burns, he replied that it was "doubtful." 9 Record 537. On this evidence, the jury could have found that the string of arrests and convictions revealed "that Burns had [such] a propensity for violence and a disregard for the law," see 67 F. 3d, at 1184, n. 20, that his subsequent resort to excessive force was the plainly obvious consequence of hiring him as a law enforcement officer authorized to employ force in performing his duties.

III

The county escapes from liability through the Court's untoward application of an enhanced fault standard to a record of inculpatory evidence showing a contempt for constitutional obligations as blatant as the nepotism that apparently occasioned it. The novelty of this escape shows something unsuspected (by me, at least) until today. Despite arguments that Monell `s policy requirement was an erroneous reading of § 1983, see Oklahoma City v. Tuttle, 471 U. S., at 834 (Stevens, J., dissenting), I had not previously thought that there was sufficient reason to unsettle the precedent of Monell. Now it turns out, however, that Monell is hardly settled. That being so, Justice Breyer's powerful call to reexamine § 1983 municipal liability afresh finds support in the Court's own readiness to rethink the matter.

I respectfully dissent.

Justice Breyer, with whom Justice Stevens and Justice Ginsburg join, dissenting.

In Monell v. New York City Dept. of Social Servs., 436 U.S. 658 (1978), this Court said that municipalities cannot be held liable for constitutional torts under 42 U. S. C. § 1983 "on a respondeat superior theory," but they can be held liable "when execution of" a municipality's "policy or custom. . . inflicts the injury." 436 U. S., at 691, 694. That statement has produced a highly complex body of interpretive law. Today's decision exemplifies the law's complexity, for it distinguishes among a municipal action that "itself violates federal law," ante, at 404, an action that "intentionally deprive[s] a plaintiff of a federally protected right," ante, at 405, and one that "has caused an employee to do so," ibid. It then elaborates this Court's requirement that a consequence be "so likely" to occur that a policymaker could "reasonably be said to have been deliberately indifferent" with respect to it, Canton v. Harris, 489 U.S. 378, 390 (1989) (emphasis added), with an admonition that the unconstitutional consequence must be "plainly obvious," ante, at 411. The majority fears that a contrary view of prior precedent would undermine Monell `s basic distinction. That concern, however, rather than leading us to spin ever finer distinctions as we try to apply Monell `s basic distinction between liability that rests upon policy and liability that is vicarious, suggests that we should reexamine the legal soundness of that basic distinction itself.

I believe that the legal prerequisites for reexamination of an interpretation of an important statute are present here. The soundness of the original principle is doubtful. The original principle has generated a body of interpretive law that is so complex that the law has become difficult to apply. Factual and legal changes have divorced the law from the distinction's apparent original purposes. And there may be only a handful of individuals or groups that have significantly relied upon perpetuation of the original distinction. If all this is so, later law has made the original distinction, not simply wrong, but obsolete and a potential source of confusion. Cf., e. g., Continental T. V., Inc. v. GTE Sylvania Inc., 433 U.S. 36, 47-49 (1977) (reexamining Sherman Act's interpretation set forth in United States v. Arnold, Schwinn & Co., 388 U.S. 365 (1967)); Hubbard v. United States, 514 U.S. 695, 697-715 (1995) (reexamining interpretation of 18 U. S. C. § 1001 set forth in United States v. Bramblett, 348 U.S. 503 (1955)); Monell, supra, at 664-690, 695-701 (reexamining interpretation of 42 U. S. C. § 1983 set forth in Monroe v. Pape, 365 U.S. 167 (1961)). See also United States v. Gaudin, 515 U.S. 506, 521-522 (1995).

First, consider Monell `s original reasoning. The Monell "no vicarious liability" principle rested upon a historical analysis of § 1983 and upon § 1983's literal language—language that imposes liability upon (but only upon) any "person." Justice Stevens has clearly explained why neither of these rationales is sound. Oklahoma City v. Tuttle, 471 U.S. 808, 834-844 (1985) (dissenting opinion); Pembaur v. Cincinnati, 475 U.S. 469, 489-491 (1986) (opinion concurring in part and concurring in judgment). Essentially, the history on which Monell relied consists almost exclusively of the fact that the Congress that enacted § 1983 rejected an amendment (called the Sherman amendment) that would have made municipalities vicariously liable for the marauding acts of private citizens. Monell, supra, at 666-667, 694. Cf. Jett v. Dallas Independent School Dist., 491 U.S. 701, 726-729 (1989) (plurality opinion). That fact, as Justice Stevens and others have pointed out, does not argue against vicarious liability for the act of municipal employees —particularly since municipalities, at the time, were vicariously liable for many of the acts of their employees. See Tuttle, supra, at 836, n. 8 (Stevens, J., dissenting) (citing cases); Pembaur, supra, at 489-490 (Stevens, J., concurring in part and concurring in judgment). See also, e. g., Kramer & Sykes, Municipal Liability Under § 1983: A Legal and Economic Analysis, 1987 S. Ct. Rev. 249, 256-265; Mead, 42 U. S. C. § 1983 Municipal Liability: The Monell Sketch Becomes a Distorted Picture, 65 N. C. L. Rev. 517, 535-537 (1987). But see Welch & Hofmeister, Praprotnik, Municipal Policy and Policymakers: The Supreme Court's Constriction of Municipal Liability, 13 S. Ill. U. L. J. 857, 881 (1989) (adopting Monell `s reading of the legislative history).

Without supporting history, it is difficult to find § 1983's words "[e]very person" inconsistent with respondeat superior liability. In 1871 "bodies politic and corporate," such as municipalities, were "person[s]." See Act of Feb. 25, ch. 71, § 2, 16 Stat. 431 (repealed 1939); Monell, supra, at 688-689. Section 1983 requires that the "person" either "subjec[t]" or "caus[e]" a different person "to be subjected" to a "deprivation" of a right. As a purely linguistic matter, a municipality, which can act only through its employees, might be said to have "subject[ed]" a person or to have "cause[d]" that person to have been "subjected" to a loss of rights when a municipality's employee acts within the scope of his or her employment. See Restatement (Second) of Agency § 219 (1957); W. Landes & R. Posner, The Economic Structure of Tort Law 120-121 (1987). Federal courts on occasion have interpreted the word "person" or the equivalent in other statutes as authorizing forms of vicarious liability. See, e. g., American Telephone and Telegraph Co. v. Winback and Conserve Program, Inc., 42 F.3d 1421, 1429-1434 (CA3 1994), cert. denied, 514 U.S. 1103 (1995) (Lanham Act); American Soc. of Mechanical Engineers, Inc. v. Hydrolevel Corp., 456 U.S. 556 (1982) (Sherman Act); United States v. A & P Trucking Co., 358 U.S. 121, 124-125 (1958) (criminal statute). See also Tuttle, supra, at 835 (Stevens, J., dissenting).

Second, Monell `s basic effort to distinguish between vicarious liability and liability derived from "policy or custom" has produced a body of law that is neither readily understandable nor easy to apply. Today's case provides a good example. The District Court in this case told the jury it must find (1) Sheriff Moore's screening "so likely to result in violations of constitutional rights" that he could "reasonably [be] said to have been deliberately indifferent to the constitutional needs of the Plaintiff" and (2) that the "inadequate hiring . . . policy directly caused the Plaintiff's injury." App. 123a (emphasis added). This instruction comes close to repeating this Court's language in Canton v. Harris. In Canton, the Court said (of the city's failure to train officers in the use of deadly force):

The majority says that the District Court and the Court of Appeals did not look closely enough at the specific facts of this case. It also adds that the harm must be a "plainly obvious consequence " of the "decision to hire" Burns. Ante, at 411. But why elaborate Canton `s instruction in this way? The Court's verbal formulation is slightly different; and that being so, a lawyer or judge will ignore the Court's precise words at his or her peril. Yet those words, while adding complexity, do not seem to reflect a difference that significantly helps one understand the difference between "vicarious" liability and "policy." Cf. ante, at 421-422 (Souter, J., dissenting). Even if the Court means only that the record evidence does not meet Canton `s standard, it will be difficult for juries, and for judges, to understand just why that is so. It will be difficult for them to apply today's elaboration of Canton —except perhaps in the limited context of police force hiring decisions that are followed by a recruit's unconstitutional conduct.

Consider some of the other distinctions that this Court has had to make as it has sought to distinguish liability based upon policymaking from liability that is "vicarious." It has proved necessary, for example, to distinguish further, between an exercise of policymaking authority and an exercise of delegated discretionary policy-implementing authority. See St. Louis v. Praprotnik, 485 U.S. 112, 126-127 (1988) (plurality opinion). Compare Tuttle, 471 U. S., at 817 (plurality opinion), with Canton, supra, at 389-390. Without some such distinction, "municipal liability [might] collaps[e] into respondeat superior, " ante, at 410, for the law would treat similarly (and hold municipalities responsible for) both a police officer's decision about how much force to use when making a particular arrest and a police chief's decision about how much force to use when making a particular kind of arrest. But the distinction is not a clear one. It requires federal courts to explore state and municipal law that distributes different state powers among different local officials and local entities. Praprotnik, supra, at 125-126, 127-131 (plurality opinion); Jett, supra, at 737-738. That law is highly specialized; it may or may not say just where policymaking authority lies, and it can prove particularly difficult to apply in light of the Court's determination that a decision can be "policymaking" even though it applies only to a single instance. Pembaur, 475 U. S., at 481. See also Praprotnik, supra, at 143 (Brennan, J., concurring in judgment); Schnapper, A Monell Update: Clarity, Conflict, and Complications, Practising Law Institute, Litigation and Administrative Practice Series, No. 381, Vol. 2, p. 36 (1989); Schuck, Municipal Liability Under Section 1983: Some Lessons From Tort Law and Organization Theory, 77 Geo. L. J. 1753, 1774-1779 (1989).

It is not surprising that results have sometimes proved inconsistent. Compare ante, at 408 (sheriff was final policymaker in hiring matters), with Greensboro Professional Fire Fighters Assn., Local 3157 v. Greensboro, 64 F.3d 962, 965— 966 (CA4 1995) (fire chief was not policymaker with respect to hiring and firing), and Harris v. Pagedale, 821 F.2d 499, 505-508 (CA8) (municipality was deliberately indifferent to charges of sexual assault), cert. denied, 484 U.S. 986 (1987), with Wilson v. Chicago, 6 F.3d 1233, 1240-1241 (CA7 1993) (municipal policymaker was not deliberately indifferent to charges of abuse of pretrial detainees), cert. denied, 511 U.S. 1088 (1994). See also Auriemma v. Rice, 957 F.2d 397, 400— 401 (CA7 1992) (describing confusion in courts).

Nor does the location of "policymaking" authority pose the only conceptually difficult problem. Lower courts must also decide whether a failure to make policy was "deliberately indifferent," rather than "grossly negligent." Canton, supra, at 388, n. 7. And they must decide, for example, whether it matters that some such failure occurred in the officer-training, rather than the officer-hiring, process. Ante, at 409-410.

Given the basic Monell principle, these distinctions may be necessary, for without them, the Court cannot easily avoid a "municipal liability" that "collaps[es] into respondeat superior. " Ante, at 410. But a basic legal principle that requires so many such distinctions to maintain its legal life may not deserve such longevity. See Mead, 65 N. C. L. Rev., at 542 (describing the "confusion and uncertainty" in the lower courts "caused by the Monell Court's choice of the policy or custom causation requirement"); Schuck, supra, at 1783 (noting the "extraordinary unpredictability of the `official policy' test").

Finally, relevant legal and factual circumstances may have changed in a way that affects likely reliance upon Monell `s liability limitation. The legal complexity just described makes it difficult for municipalities to predict just when they will be held liable based upon "policy or custom." Moreover, their potential liability is, in a sense, greater than that of individuals, for they cannot assert the "qualified immunity" defenses that individuals may raise. Owen v. Independence, 445 U.S. 622 (1980). Further, many States have statutes that appear to, in effect, mimic respondeat superior by authorizing indemnification of employees found liable under § 1983 for actions within the scope of their employment. See, e. g., Conn. Gen. Stat. § 7-465 (1997); Idaho Code § 6-903 (1990); Ill. Comp. Stat., ch. 745, § 10/2-302 (1994); Kan. Stat. Ann. § 75-6109 (1989); Minn. Stat. § 466.07 (1994); Mont. Code Ann. § 2-9—305 (1994); Nev. Rev. Stat. § 41.0349 (1989); N. H. Rev. Stat. Ann. § 29—A:2 (1988); N. D. Cent. Code § 32-12.1— 04(4) (Supp. 1993); Okla. Stat., Tit. 51, § 162 (Supp. 1995); 42 Pa. Cons. Stat. § 8548 (1982); S. D. Codified Laws § 3-19-1 (1994); Utah Code Ann. § 63-30-36 (1993); W. Va. Code § 29-112A-11 (1992); Wis. Stat. § 895.46 (1993-1994). These statutes—valuable to government employees as well as to civil rights victims—can provide for payments from the government that are similar to those that would take place in the absence of Monell `s limitations. To the extent that they do so, municipal reliance upon the continuation of Monell `s "policy" limitation loses much of its significance.

Any statement about reliance, of course, must be tentative, as we have not heard argument on the matter. We do not know the pattern of indemnification: how often, and to what extent, States now indemnify their employees, and which of their employees they indemnify. I also realize that there may be other reasons, constitutional and otherwise, that I have not discussed that argue strongly for reaffirmation of Monell `s holding. See, e. g., Gerhardt, The Monell Legacy: Balancing Federalism Concerns and Municipal Accountability under Section 1983, 62 S. Cal. L. Rev. 539 (1989) (discussing federalism); Nahmod, Constitutional Accountability in Section 1983 Litigation, 68 Iowa L. Rev. 1, 24-25 (1982) (describing Monell as having the "proper approach to local government accountability under section 1983" and describing a fault-based interpretation of § 1983); Welch & Hofmeister, 13 S. Ill. U. L. J., at 883, n. 176 (discussing disadvantages of an "expansive view of municipal liability," including lack of insurance coverage).

Nonetheless, for the reasons I have set forth, I believe the case for reexamination is a strong one. Today's decision underscores this need. Consequently, I would ask for further argument that would focus upon the continued viability of Monell `s distinction between vicarious municipal liability and municipal liability based upon policy and custom.