Skip to main content
Photo of a Law Library

The Role of Fault in § 1983 Municipal Liability

Michael L. Wells[1]*


Local governments can be sued under 42 U.S.C. § 1983[2] for some but not all constitutional violations that their employees commit. In Monell v. Department of Social Services of New York,[3] the Supreme Court rejected the common law respondeat superior doctrine, which imposes vicarious liability for torts committed by an organization’s employees in the course of the employment.[4] “Instead, it is when execution of a government’s policy or custom . . . inflicts the injury that the government as an entity is responsible under § 1983.”[5] Monell left open two important issues: (a) whether governments would have immunity from liability for damages,[6] by analogy to the qualified immunity defense available to individual defendants unless they violate “clearly established” rights,[7] and (b) how “policy or custom” would be defined.[8] Two years later, it addressed the first of these questions: Owen v. City of Independence flatly denied governments any immunity, even when their officers’ actions complied with clearly established law.[9] Monell, coupled with Owen, thus imposes a kind of strict liability, for both damages and prospective relief,[10] when a local government’s policy or custom is the “moving force” of a constitutional violation.[11]

Later cases on the policy or custom issue have not produced a similarly crisp rule. The Court has distinguished between two types of cases, depending on whether the government involvement is “direct” or “indirect.”[12] In the “direct-effect” cases, a law-making body promulgates a general unconstitutional rule or a single “policymaker” makes an unconstitutional decision.[13] This type of case is illustrated by Monell, which involved a challenge to New York City’s requirement that pregnant employees must take leave at five months of the pregnancy.[14] In Pembaur v. City of Cincinnati, the Court recognized liability in the “single decision” fact pattern.[15] In that case, the local district attorney had directed police officers to enter a doctor’s office by force, in violation of his Fourth Amendment rights.[16] In both Monell and Pembaur, government responsibility is direct in the sense that no actor intervenes between the policymaker and the violation.[17] Under Owen’s “no immunity” rule, the plaintiff wins this type of case when he proves a violation of his constitutional rights and a causal connection between that violation and the plaintiff’s harm.[18] For example, the City was liable in Pembaur even though the Supreme Court did not decide the illegality of the search until years later.[19] Though some analysts object to Owen’s denial of immunity,[20] none of the Justices have questioned this regime.

This Article focuses on indirect-effect cases, in which policymakers do not violate constitutional rights. Rather, a street-level officer, or some other subordinate, makes the key decision that involves a constitutional violation. When plaintiffs sue such officers directly, the officers benefit routinely from the qualified immunity defense.[21] Plaintiffs thus search for ways to secure damages from the government itself in indirect-effect cases.[22] Sometimes they succeed. In particular, governments are subject to liability in these indirect cases if, but only if, policymakers have acted or failed to act with “deliberate indifference” to constitutional rights.[23] When is this liability-triggering standard met? Courts have declared that “a municipal policymaker cannot exhibit . . . deliberate indifference to a constitutional right when that right has not yet been clearly established.”[24] As a practical matter, the deliberate indifference requirement thus limits governmental liability in much the same way that qualified immunity limits officer liability.[25] The law governing these cases is especially important because, in practice, the indirect cases are far more numerous than the direct ones.[26] There are far more underlings than policymakers and, if only for that reason, far more constitutional violations by underlings.

Courts distinguish among several indirect-effect fact patterns. The litigation may involve delegation of authority to subordinates,[27] ratification of decisions by underlings,[28] inaction in the face of a widespread custom of street-level employees,[29] or faulty training, hiring, or supervision of employees.[30] The key feature of all of these cases is that, as with certain common law torts, courts distinguish between objective and subjective liability rules.[31] The Supreme Court has explicitly rejected the objective negligence test,[32] ruling that “a showing of simple or even heightened negligence will not suffice.”[33] Showing deliberate indifference requires proof of the policymaker’s “conscious disregard for [constitutional] rights.”[34] The Court’s rejection of even heightened negligence indicates that deliberate indifference is a subjective test akin to common law recklessness, which focuses on the attitudes and intentions of policymakers, rather than an objective test like common law negligence, which would evaluate the policymaker’s choices against an external standard.[35]

Some critics of the Court’s regime would remove most of the obstacles to local government liability by reversing Monell’s “no respondeat superior” holding.[36] Others would eliminate the officer’s immunity from damages liability.[37] These reformers would, in effect, impose a type of liability without fault—either on governments, officers, or both—in indirect-effect cases. For these analysts, the key goals involve vindicating constitutional rights and deterring constitutional violations, at the expense of the goals—federalism, government resource conservation, and officer inhibition avoidance—that drive the current doctrine’s constraints on recovery.[38] Most discussions of § 1983 municipal liability have treated the issue as a binary choice between the Court’s restrictive doctrine and the critics’ strict liability alternative, which are near-polar extremes.[39]

The problem with this winner-take-all battle is that important values will be sacrificed no matter who wins, and perhaps needlessly so. In devising constitutional remedies, the goal is to achieve “the best attainable accommodation of competing values,”[40] or as Richard Fallon has put it, “the best overall bundle of rights and correspondingly calibrated remedies.”[41] Neither the Court’s existing municipal liability doctrine nor the critics’ proposed alternative is normatively attractive. The Court’s approach puts too many hurdles in the path of plaintiffs seeking recovery for constitutional violations because deliberate indifference, to say the least, is a stringent standard.[42] But the critics’ strict liability solution goes too far in the other direction. The costs of strict liability might well outweigh its benefits.[43] In any event, such a change in the law is not a realistic option for the foreseeable future. Other, more modest styles of reform thus need to be considered.[44]

This Article proposes a new approach to indirect municipal liability cases. This approach would discard the subjective deliberate indifference standard and replace it with an objective test similar to the reasonableness rule of common law negligence. Respondeat superior aside, ordinary tort principles hold employers liable for negligent hiring and supervision.[45] Under this approach, local governments would owe a duty of reasonable care to prevent constitutional violations by their employees. Municipalities would remain shielded from vicarious liability but would be liable for faulty hiring and supervision. Liability would not depend on proof of a policymaker’s conscious disregard of obvious risks but on whether policymakers should have known of constitutional risks, whether they acted reasonably in the circumstances to diminish those risks, and whether the breach of the duty of reasonable care was the proximate cause of the plaintiff’s constitutional injury.[46]

Part II describes types of cases that comprise the indirect branch of § 1983 municipal liability in which a subordinate committed the constitutional violation. Part III makes the case for replacing the subjective deliberate indifference standard with an objective reasonable care standard in these cases. The argument, in a nutshell, is that the overall goal of constitutional tort law is to balance the costs and benefits of liability by formulating rules that accommodate values that favor broad liability and those values that largely close the door to injured plaintiffs who assert these claims. The Court’s current jurisprudence shortchanges this central goal by (unconvincingly) relying on the claimed need to avoid a “collapse[] into respondeat superior.”[47] While limiting liability to “acts for which the municipality is actually responsible”[48] is a credible goal, the deliberate indifference requirement is an overreaction to the problem it addresses. Part IV discusses the implementation of the reasonable care approach, with particular attention to the application of negligence principles to local government liability litigation. Part V defends the fault-based approach against the charge that it would not provide sufficient protection against the collapse into vicarious liability. Under the traditional reasonableness-based tort law standard, the Court’s goal of avoiding vicarious liability can be achieved at far less cost to the remedial goals of constitutional tort law.

Indirect-Effect Local Government Liability

Monell, Owen, and Pembaur are direct-effect cases, in which the policymaker commits the constitutional violation, either by formally enacting an unconstitutional rule, informally engaging in an unconstitutional practice, or making a single unconstitutional decision.[49] But these cases, especially Pembaur, lay the foundation for broader liability, in which the subordinate has committed the constitutional violation and policymakers’ involvement is indirect in the sense that they fail to exercise sufficient oversight to prevent the violation.[50] Though the government’s responsibility is more attenuated in these indirect-effect cases, policymakers’ inattention, passivity, or implicit complicity in violations still may trigger government liability.

From Pembaur to Praprotnik

The leading case in the development of this indirect-effect doctrine is City of St. Louis v. Praprotnik,[51] decided two years after Pembaur. Praprotnik sued the City under § 1983 after losing his job as an architect. A jury found that Praprotnik’s supervisors had retaliated against him on several occasions for speech protected by the First Amendment.[52] Though the culpable supervisors were not policymakers, their superiors up the hierarchical chain had approved the personnel decisions.[53] The case thus differed from Monell, Owen, and Pembaur because Praprotnik did not show that policymakers themselves made the key decisions that resulted in violations of his constitutional rights.

Even so, Pembaur’s holding that a policymaker’s single act can establish liability will, in some cases, uphold liability when that single act involves faulty oversight rather than a direct violation. There was no majority opinion, and in the end, Praprotnik failed to establish municipal liability.[54] Nonetheless, Praprotnik laid the foundation for the indirect-effect doctrine because all eight participating Justices agreed on the basic principle that governments could be held liable where the policymaker’s involvement is indirect.[55] In particular, the separate opinions of Justices O’Connor and Brennan were joined by a total of seven members of the Court, and these two opinions reflected agreement on some basic points with regard to municipal liability for the acts of subordinates.[56]

One of those points is that in a no respondeat superior regime there are good reasons for distinguishing between direct-effect and indirect-effect cases. Justice O’Connor’s plurality opinion noted the structural difference between this indirect case and its earlier rulings in Monell and Pembaur.[57] As she observed, “special difficulties arise when it is contended that a municipal policymaker has delegated his policymaking authority to another official,”[58] and the bar against respondeat superior implies that local governments should not be liable for “the mere exercise of discretion by [a street-level] employee.”[59] Justice O’Connor, however, did not categorically reject municipal liability in these cases, concluding instead that liability would be appropriate for some tasks delegated to underlings.[60] As she pointed out, “[I]f . . . a city’s lawful policymakers could insulate the government from liability simply by delegating their policymaking authority to others, § 1983 could not serve its intended purpose.”[61] Thus, “if the authorized policymakers approve a subordinate’s decision and the basis for it, their ratification would be chargeable to the municipality because the decision is final.”[62] Justice Brennan’s opinion echoes Justice O’Connor on these key points.[63] Justice Stevens, in the lone dissent, would have upheld the jury verdict in Praprotnik’s favor on the ground—rejected by both Justice O’Connor and Justice Brennan—that the plaintiff had shown sufficient involvement by policymakers in the challenged government actions.[64]

Praprotnik provides little guidance to lower courts. Justices O’Connor and Brennan focus their attention on the specific facts of the case and, in particular, on the degree of the constitutional wrong. Neither opinion lays out a general principle for distinguishing between instances of delegation or ratification that will trigger municipal liability and those that will not. For example, Justice O’Connor implies that a city should be liable not only when policymakers ratify decisions by subordinates but also when they “simply . . . delegate[e] their policymaking authority to others.”[65] At the same time, “Simply going along with discretionary decisions made by one’s subordinates . . . is not a delegation to them of the authority to make policy.”[66] Additionally, “[M]ere failure to investigate the basis of a subordinate’s discretionary decisions does not amount to a delegation of policymaking authority.”[67] Neither Justice O’Connor nor Justice Brennan examined the doctrinal implications of these references to delegation and ratification.

Praprotnik Liability in the Lower Courts

In the thirty years since Praprotnik, the Court has not returned to these issues. Lacking clear Supreme Court guidance, the federal circuit courts have taken disparate approaches to ratification and delegation.[68] Though courts often treat these as distinct grounds for exposing governments to § 1983 liability,[69] the line between delegation and ratification is blurry at best. For example, when policymakers “condone” misconduct by subordinates, the policymaker’s attitude may be described by either term.[70] Part IV will argue that a reasonable care standard should govern both.


Some courts have taken a broad view of municipal liability, citing Justice O’Connor’s warning that “[i]f . . . a city’s lawful policymakers could insulate the government from liability simply by delegating their policymaking authority to others, § 1983 could not serve its intended purpose.”[71] These courts hold that policymakers cannot exempt local governments from liability merely by delegating decision-making authority.[72] In King v. Kramer, for example, the county government had hired Health Professionals Ltd. (HPL), a private company, to manage the health care of inmates in a local jail.[73] When an inmate sued the county, asserting a violation of his right to adequate care, the county argued that the private company was not a government policymaker, in part, because of its non-governmental status.[74] The court, however, denied summary judgment to the county. It was enough that “[t]he evidence presented for summary judgment purposes shows that the County’s policy was to entrust final decision-making authority to HPL over inmates’ access to physicians and medications.”[75] Other cases similarly hold that delegation can turn the subordinate into a policymaker[76] or that the higher official has delegated authority to the subordinate by condoning violations.[77]

Other courts have taken a far more cautious approach to the delegation theory. An illustrative case is Bolderson v. City of Wentzville.[78] Diane Bolderson, the city’s building commissioner, “criticized changes to the city’s building code,” questioned purchasing practices, and accused city officials “of fraud and acting with conflicts of interest.”[79] The city administrator fired her, citing these actions on her part.[80] Bolderson sued the city under § 1983, claiming that her dismissal violated her First Amendment rights.[81] One of her grounds for naming the City as a defendant was that the mayor had delegated “authority to the city administrator to address Bolderson’s criticisms” and had “tacit[ly] approve[d] . . . the city administrator’s decision to terminate her.”[82] The court did not reject this characterization of the facts. Yet it rejected Bolderson’s claim,[83] relying on Justice O’Connor’s observation in Praprotnik that “[s]imply going along with discretionary decisions made by one’s subordinates . . . is not a delegation to them of the authority to make policy.”[84] Penley v. McDowell County Board of Education[85] seems to take an even dimmer view of the delegation theory. It suggests that a local government cannot be liable for a decision by a nonpolicymaker (the school superintendent) so long as it “did not retain final review authority”[86] over the office’s action.


If ratification is a distinct category, the feature that sets it apart is that the policymaker’s involvement comes after the subordinate’s constitutional violation. As with delegation, the circuit court cases are not uniform when the plaintiff claims policymaker post-act approval of a subordinate’s unconstitutional conduct. Some cases are easy, as when the plaintiff asserts that the policymaker chose, in a focused and deliberate way, to approve the decision and the unconstitutional reason for it, thereby meeting both prongs of Justice O’Connor’s test.[87] Harder issues arise and disagreements surface when the plaintiff’s theory is that the policymaker should have investigated the subordinate’s earlier decision-making. Montero v. City of Yonkers[88] illustrates a plaintiff-friendly view of municipal liability under these circumstances, as the court recognized a duty on the part of the policymaker to examine why a subordinate acted as he did. The court concluded that liability can attach due to ratification even without proof that any policymaker approved of the unconstitutional reasons for an underling employee’s acts, so long as the plaintiff can show “that the policymaking official was aware of the employee’s unconstitutional actions and consciously chose to ignore them.”[89]

Other courts have ruled for the local government in similar circumstances, finding that no ratification can occur unless there is no review at all;[90] instead, ratification can arise only when the higher official approves both the decision and the unconstitutional grounds for it.[91] In Waters v. City of Chicago,[92] for example, the court declared that “mere failure to investigate the basis of a subordinate’s discretionary decisions does not amount to a delegation of policymaking authority, especially where (as here) the wrongfulness of the subordinate’s decision arises from a retaliatory motive or other unstated rationales.”[93] Though the court uses the term “delegation,”[94] the after-the-fact timing of policymaker involvement suggests that the case more properly falls into the “ratification” category.

Training, Supervision, and Hiring

Relations between policymakers and underlings are many and varied. Besides delegation and ratification, which tend to involve relatively high-up subordinates, policymakers have responsibility for the hiring, training, and supervising of relatively low-level government workers, including police officers and jailors who deal with potential constitutional tort victims on a daily basis. Praprotnik opened the courthouse door to local government liability for constitutional violations by these street-level officers. A year after Praprotnik, the Court endorsed that theory of recovery.

Training: City of Canton v. Harris

Building on Praprotnik, the Court in City of Canton v. Harris endorsed the idea that policymakers’ inadequate training of street-level officers can support local government liability.[95] In Praprotnik the Court had avoided respondeat superior by limiting liability to situations involving delegation or ratification, although it did not define what those limits were. Harris provided more guidance. In training cases, municipal liability would extend to situations in which the plaintiff could prove, first, that the training was deliberately indifferent to constitutional rights[96] and second, that the inadequate training was “closely related to the ultimate injury.”[97]

After her arrest by Canton police officers, Harris fell ill while in custody.[98] Neither the police nor jail officers provided relief or called for medical assistance.[99] In her § 1983 case, Harris claimed that the officers violated her Fourteenth Amendment rights because pretrial detainees are entitled to necessary medical treatment.[100] Harris sued the City, but she could not show that the failures of the jail personnel fit into any of the categories the Court had recognized in the line of cases from Monell through Praprotnik.[101] The Fourteenth Amendment violation did not result from a formal or informal top-down policy or a single act of a policymaker, and it did not involve a delegation of authority or ratification by a policymaker. Harris proposed, and the Court recognized, a new theory of municipal liability.[102] Her theory was that “a city can be liable for inadequate training of its employees,”[103] when the inadequate training causes a violation of constitutional rights by subordinates.

The Court adopted this approach,[104] but with modifications. In particular, plaintiffs could not succeed in invoking this theory unless “the failure to train amounts to deliberate indifference to the rights of persons with whom the police come into contact.”[105] Only when this heightened culpability standard is met “can such a shortcoming be properly thought of as a city ‘policy or custom’ that is actionable under § 1983.”[106] Also, “the identified deficiency in a city’s training program must be closely related to the ultimate injury.”[107] These requirements were necessary because “lesser standards of fault and causation would open municipalities to unprecedented liability under § 1983.”[108] According to the Court, “a less stringent standard of fault for a failure to train claim ‘would result in de facto respondeat superior liability on municipalities.’”[109] Deliberate indifference meets this concern because “a policy of inaction in light of notice that its program will cause constitutional violations is the functional equivalent of a decision by the city itself to violate the Constitution.”[110] In this way, the deliberate indifference rule serves the role of identifying the set of violations by street-level workers for which governments may fairly be held responsible.[111]

Harris sets up a formidable obstacle for the plaintiff. In Connick v. Thompson,[112] for example, Thompson had been convicted in New Orleans of crimes in trials at which prosecutors had violated his constitutional right under Brady v. Maryland to be provided with exculpatory evidence.[113] His Monell theory of recovery was that (a) the New Orleans district attorney was the city’s policymaker for prosecutions, (b) the district attorney had failed to train assistant district attorneys on this constitutional rule, and (c) failure-to-train was the moving force of the constitutional violation.[114] The evidence at trial was that “[n]o prosecutor remembered any specific training session regarding Brady” during the relevant time period.[115] Other evidence included four prior instances over the past decade in which the New Orleans district attorney’s office had seen convictions overturned on account of violations of this rule,[116] and he obtained a jury verdict in his favor.[117] The Supreme Court found no fault with the jury instructions on deliberate indifference but overturned the verdict all the same because the earlier Brady violations could be distinguished on their facts.[118] Despite the jury’s verdict, the plaintiff had failed to show “that failing to train the prosecutors amounted to conscious disregard for [his] rights.”[119] Thus, the Supreme Court’s holding seems to be that no reasonable jury could find deliberate difference on these facts.[120]

Hiring: Board of County Commissioners v. Brown

Later cases extended this theory beyond the facts of Harris. Of particular importance, in Board of County Commissioners v. Brown,[121] the Court held that a hiring decision may produce municipal liability if the plaintiff can show that “through its deliberate conduct, the municipality was the moving force behind the injury alleged.”[122] But, having recognized the theory, the Court applied it narrowly to the facts of the case. A sheriff’s deputy, Stacy Burns, had injured Jill Brown at a traffic stop.[123] The county sheriff, B.J. Moore, the policymaker for his department, had hired Burns, the son of the sheriff’s nephew,[124] as a reserve deputy.[125] Burns had a history of misdemeanors and the sheriff admitted that he had “not closely reviewed” Burns’ background.[126] Despite a verdict for the plaintiff, rendered as in Connick by a jury that had been properly instructed on deliberate indifference, the Court deemed this single incident insufficient to meet the test.[127] In relying on Harris to affirm the verdict, the Court explained that Brown:

[I]gnore[d] the fact that predicting the consequences 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 duties.[128]


Lower courts have upheld Monell recovery in cases involving a failure to supervise or to discipline employees, rather than a lack of training.[129] These courts in effect have concluded that it makes no sense to limit Harris only to bad training and faulty hiring. Policymaker involvement with subordinates does not stop once the employees are hired and trained. Lax supervision can be no less culpable than hiring or training and can be closely connected to constitutional violations by subordinates. In fact, no circuit seems to have rejected this ground for recovery. The Court itself has never indicated that it would distinguish between training and hiring on the one hand and supervision on the other, so long as the challenged policymaker conduct involves deliberate indifference to constitutional rights. And given the logic of the Court’s existing precedents, it is hard to see how the Court could draw such a distinction in the future.

Bottom-up Customs

Monell said that municipal liability could be established based on either a formal policy or a custom.[130] The term custom, however, is far from self‑defining, and the Court has not provided further guidance. As the case law has developed in the lower federal courts, there are two distinct lines of custom‑based liability. One is closely related to Monell’s direct liability ground for recovery. In this scenario, the custom is a practice instituted but never formalized by policymakers. Suppose, for example, that policymakers routinely obliged pregnant employees to take leave at five months but had never put the rule on paper. This type of custom has a top-down quality because it was generated and enforced through the actions of policymakers. Once proven, it is equivalent to a formal policy.[131]

But lower courts have imposed liability on a quite distinct custom theory. In this set of cases, the custom originates among mid-level subordinates[132] or even street-level employees such as police officers.[133] It is bottom-up rather than top-down, and policymakers may or may not be aware of it. The basic difference between the two situations is that top-down customs involve a type of direct-effect Monell liability, while bottom-up customs involve indirect‑effect liability. In the former type of case, the plaintiff is required only to prove the custom and a causal link between the custom and the constitutional violation. In a bottom-up custom case, the plaintiff must prove not only the custom but some level of policymaker acceptance or toleration of the unconstitutional practice. In Brown, the Supreme Court in dicta seemed to approve in general terms of liability based on this type of bottom-up custom.[134] The Court, however, has never clarified how far this theory of local government liability reaches.

Lower courts have sought to fill this gap. They typically require plaintiffs to show two things: (a) a widespread practice among municipal employees, and (b) some level of awareness of the practice on the part of policymaking officials.[135] For example, in Matusick v. Erie County Water Authority,[136] a former county employee prevailed by proof of repeated harassment by coworkers due to his interracial relationship.[137] His evidence established that “the acts of discrimination and harassment [he] alleged . . . were frequent and severe.”[138] Moreover, “many human resources personnel, including the director of human resources, were aware of his complaints well before he was terminated. . . . [But they] failed to act.”[139]

On the other hand, in Worldwide Street Preachers v. Town of Columbia, preachers who claimed the police had unconstitutionally interfered with their activities lost their suit against the city because they proved only “a few isolated incidents” of unconstitutional conduct by lower-level officers, and a few incidents do not suffice to prove a custom.[140] On the question of policymaker awareness, no clear rule has emerged. But most courts seem to require proof that policymakers knew enough about the practice to be charged with deliberate indifference to the custom[141] or with “actual or constructive knowledge” of its existence.[142]

An Objective, Negligence-Based Approach to § 1983 Municipal Liability

The foregoing discussion shows that the Monell doctrine distinguishes among at least nine types of cases, including (a) formal rules of general application, (b) top-down custom, (c) single unconstitutional acts of a policymaker, (d) delegation, (e) ratification, (f) bottom-up custom, (g) inadequate training, (h) inadequate hiring, and (i) inadequate supervision. But the seeming complexity of this array of doctrines is misleading. The legally significant distinction in municipal liability cases is between situations in which a policymaker commits the constitutional violation, which include (a) through (c), and those in which a subordinate does so, which include (d) through (i). In the (a) through (c) group of cases, municipal liability is direct, while it is indirect for (d) through (i).[143] The liability for direct policymaker involvement is strict in the sense that the plaintiff wins by presenting proof of a constitutional violation and resulting injury without regard to policymaker fault. But the liability rule for indirect-effect cases is far harder for the plaintiff to meet. Following the Supreme Court’s “inadequate training” decision in Harris, and its applications of that test in Connick and Brown, most of the lower court cases across all indirect-effect contexts apply a subjective test.[144] The discussion in Part II has shown that courts require plaintiffs to show deliberate indifference by one or more policymaking officials to recover for indirect government involvement.

The Monell Court’s Anti-Respondeat Superior Principle

The Court has justified its disparate treatment of these two sets of cases by reference to its decision in Monell to reject the respondeat superior theory of recovery.[145] The anti-respondeat superior principle implies that governmental liability exists only when governments are responsible for constitutional violations.[146] The policymaker’s violation establishes the government’s responsibility in direct-effect cases. By contrast, indirect-effect cases, like the failure-to-train claim in Harris, demand a stringent showing of policymaker involvement to avoid de facto respondeat superior.[147] Since this rationale applies in the same way in all sorts of indirect-effect cases—whether they involve delegation, ratification, bottom-up custom, or whatever—many lower courts have adopted the deliberate indifference test in those indirect-effect cases as well.[148]

This entire edifice of law, however, is built on an unstable foundation. The Court’s endorsement of the deliberate indifference test skips over a key question about means and ends in constitutional tort law. To be sure, the Court’s reasoning in these cases begins from a valid premise: Indirect fact patterns differ in legally significant ways from direct ones.[149] More specifically, as Justice O’Connor explained in Brown,[150] the difference is that governmental responsibility is attenuated in the indirect-effect cases because, in these cases, policymakers merely fail to stop constitutional violations by others from occurring. These inattentive or passive policymakers do not intentionally act in ways that violate constitutional rights.

Concluding that indirect-effect cases differ from direct-effect cases, however, does not logically support the conclusion that liability in indirect‑effect cases should attach only when plaintiffs can satisfy the extremely demanding deliberate indifference test. In particular, there exist alternative approaches available to the Court that would satisfy its stated goal of avoiding both formal and de facto respondeat superior. The most plausible is the standard, familiar, and generally applied tort-law negligence test of objective reasonableness. Indeed, it turns out that the Court in Praprotnik, Harris, Brown, and Connick has never even asked—far less, carefully worked through—the most salient and fundamental legal question: Should municipal liability in these cases be governed by an objective negligence test or a subjective deliberate indifference test?

Even a moment’s reflection suggests that the Court’s stated rationale for adopting the deliberate indifference touchstone for liability has a question‑begging quality. It assumes without analysis that this stringent test is necessary in order to limit the scope of liability appropriately. The problem with that assumption is that, in Brown and Connick, the Court treated anti‑vicarious liability as the only relevant consideration.[151] The Court’s preoccupation with that goal in indirect-effect cases, and its choice of the deliberate indifference test to squelch the perceived danger, has diverted the Justices’ attention away from the core concerns that lay at the heart of § 1983. In fact, the Court has developed an expansive body of § 1983 case law built around the objective of “achieving the best overall bundle of rights and correspondingly calibrated remedies within our constitutional system.”[152]

More specifically, the key goal of the damages remedy has been to obtain “the best attainable accommodation of competing values.”[153] This principle has driven § 1983 doctrine in an across-the-board way, including rulings on officer immunity, causation, damages computations, and attorney’s fees. Within this framework, the Court has recognized that constitutional tort law should take into account a range of pro- and anti-liability values.[154] The main problem with the Court’s municipal liability doctrine is that the deliberate indifference principle for indirect municipal liability gives too much weight to the defendant’s side of the balance, on account of the Court’s fear of liability under respondeat superior.

This fixation has diverted the Court’s attention from the need to clearly articulate the normative foundation of § 1983 municipal liability. As in other contexts, the Court should adopt an approach to indirect-effect cases that involves a smarter accommodation of competing values. The ensuing sections of this Part suggest how this sort of approach might unfold. Section III.A describes the key features of the accommodation principle that the Court has developed and worked within past cases. Section III.B explains why a balanced application of this accommodation principle supports reappraisal of the deliberate indifference test. The reason, in brief, is that the accommodation principle would be better served by holding municipalities to a duty of reasonable care, under which the city would be liable if, but only if, the plaintiff could establish a policymaker’s negligent failure to protect the plaintiff from a constitutional violation by a subordinate. With the negligence model in place, Part IV discusses the practicalities of applying negligence doctrine in § 1983 municipal liability cases. Part V rebuts the objection, suggested in Brown and Connick, that a shift away from the deliberate indifference standard would risk a collapse into strict liability.

Accommodation of Competing Values in Constitutional Tort Law

Across the range of constitutional tort topics, the law-making task is to achieve the “proper balance” among policies that support broader or narrower liability.[155] In its official immunity cases, for example, the Court has recognized that “[t]he resolution of [official] immunity questions inherently requires a balance between the evils inevitable in any available alternative.”[156] Courts must weigh the vindication of rights and the dissuading of constitutional violations against fairness to defendants[157] and the need to limit the “social costs” of liability, which include “the expense of litigation, the diversion of official energy from pressing public issues, and the deterrence of able citizens from acceptance of public office.”[158]

In the official immunity context, the Court has sought to achieve the needed accommodation by adopting a rule under which officials engaged in administrative or executive functions may be sued for damages only when they have violated “clearly established” rights.[159] This test, as applied in recent cases, affords considerable protection,[160] except to “the plainly incompetent and those who knowingly violate the law.”[161] These administrative and executive officials have no immunity from prospective relief, a remedy that does not involve the especially heavy personal and social costs entailed by awards of damages. Conversely, damages may be sought in circumstances in which prospective relief is not available because the plaintiff cannot show a threat of future constitutional injury.[162] Another illustration of the accommodation principle in official immunity law involves the Court’s distinction between publicly-managed and privately-managed prisons.[163] In the latter context, the policies underlying immunity are too weak to support a defense.[164] Yet another illustration concerns a distinction the Court draws between administrative and executive officials, on the one hand, and certain other officers. In a few narrow categories—connected with especially sensitive judicial, prosecutorial, and legislative functions—the balance tips in favor of absolute immunity from damages. And legislators, but only legislators, are fully shielded from prospective injunctive relief.[165]

The accommodation model is most prominent in official immunity cases, in part because the Supreme Court has grappled over the years with many § 1983 immunity cases. But the accommodation principle is a recurring feature of most constitutional tort doctrines. Examples of the Court’s efforts to find a middle ground between competing values in this field include the following:

  • When a government employee is fired and claims that the motive for the dismissal was her protected speech, a causation issue arises as to whether the firing was caused by the speech. The Court does not merely apply the common law but-for test, which puts the burden of proof exclusively on the plaintiff. In this area, the plaintiff is required to prove that the speech was a substantial factor in the dismissal. But then the burden of proof shifts to the defendant to show that the plaintiff would have been fired for some other reason. In other words, the Court has struck a balance rooted in specialized burden-of-proof rules, a balance that takes account of the importance of First Amendment rights and the contrasting abilities of each party to marshal evidence in support of that party’s contentions.[166]
  • Under the Court’s § 1983 “compensation principle,” damages may not be awarded “based on the abstract ‘value’ or ‘importance’ of constitutional rights.”[167] The purposes of damages for constitutional torts are to compensate plaintiffs for injuries caused by violations of constitutional rights and to deter violations. But “there is no evidence that [Congress] meant to establish a deterrent more formidable than that inherent in the award of compensatory damages,”[168] unless the conduct is egregious.[169]
  • Local governments may not be sued for punitive damages, because the balance of interests is different in this context than it is for compensatory damages. Such an award is a “windfall to a fully compensated plaintiff, and [is] likely to be accompanied by an increase in taxes or a reduction of public services for the citizens footing the bill.”[170]
  • Similarly, the Court has declared more generally that common law principles on damages may need to be modified, at least sometimes, in the constitutional tort context because “the interests protected by a particular constitutional right may not also be protected by an analogous breach of the common law of torts.”[171]
  • State law ordinarily governs the issue of whether a § 1983 suit survives the death of the plaintiff. But, as the Court has recognized, the case for survival would be stronger, and an exception to that deference to state law may be needed, if the violation causes the death.[172]
  • To vindicate rights and deter violations, Congress has authorized the award of a “reasonable” attorney’s fee to a plaintiff who is a “prevailing party” in § 1983 litigation.[173] At the same time, if the plaintiff invests substantial resources to obtain a large award, yet obtains only token recovery, the appropriate fee will be zero.[174] Vindication and deterrence may require a substantial award even if the plaintiff only obtains nominal damages, so long as “the significance of the legal issue” warrants that result, or if the litigation served “some public goal other than occupying the time and energy of counsel, court, and client.”[175] In any event, the vindication and deterrence goals carry enough weight to foreclose any rule that the fee must be reduced merely because it is higher than the damages awarded.[176] Very different, and far less accommodating, attorney’s fees rules apply when a defendant seeks them.[177] Again, this difference in treatment reflects a carefully thought-through evaluation of the proper incentives and disincentives to build into § 1983 in light of a focused assessment of competing policy concerns.

Each of these doctrines may or may not reflect the optimal accommodation of relevant § 1983 policies. The key point is that, however debatable each of these rules may be, all of them reflect a common theme. They support the proposition that as a routine practice in constitutional tort litigation the Court works hard to accommodate the competing interests of injured plaintiffs and state actors. There exists, in short, a deep-rooted, accommodation-based approach to constitutional tort law. That approach has long guided the resolution of the issues presented in § 1983 litigation and should also guide the development of municipal liability doctrine.

The Role of Fault in the Accommodation of Values

Under a negligence approach, governments would owe a duty of reasonable care to avoid constitutional violations by their employees and would be held liable if, but only if, their policymakers fail to meet that duty. They would not escape liability, as they can today, just because the plaintiff cannot prove policymakers’ “conscious disregard” of violations by subordinates.[178] Liability would turn on whether the government’s policymakers have met an external standard, that of the reasonably prudent persons in the circumstances. This objective approach serves the accommodation goal of constitutional tort law because it strikes a balance between competing interests.[179] The advantage of negligence as a norm is that it distinguishes, in broad but generally defensible ways, between cases in which the values favoring liability are more likely to be especially strong and cases in which the social costs of liability are more likely to be especially strong.[180] The decisive objection to the subjective deliberate indifference doctrine of Harris and Brown is that it strikes the balance too far in the direction of shielding governments from lawsuits. Unlike “reasonable care,” it does not align with the “moral norms of the community.”[181] It permits local governments to win due to insufficient evidence of conscious disregard, even when the application of community standards favors the plaintiff.[182]

Ordinary Tort Law in § 1983 Litigation

Borrowing from private law to resolve a key problem of public law may seem inappropriate because common law tort doctrine does not usually try to take constitutional values into account.[183] As Richard Fallon has persuasively argued, it would be a mistake to treat “the common law of tort as a paradigm of official accountability and liability for constitutional violations.”[184] But it is no less true that tort concepts can help in the resolution of certain constitutional tort issues. Because many of the values at stake are similar across all situations in which someone sues to recover for a non-contractual past injury, the Court often has looked to ordinary tort concepts as a starting point in adjudicating constitutional tort issues.[185] The Court refers to common law insights, resources, and guidance for two reasons. First, the common law is relevant under general principles of statutory construction because § 1983 contains few specifics and was enacted “against the background of tort liability that makes a man responsible for the natural consequences of his actions.”[186] The implication is that Congress intended to fill in the gaps by reference to the common law. By contrast, the Monell Court’s policy-or-custom formulation has no basis in the statute, the legislative history, or the pre-Monell legal background.

Second, as a normative matter, the Court has found that ordinary tort rules provide viable solutions for constitutional tort issues. Common law principles are well-entrenched in § 1983 doctrine, including the rules on official immunity,[187] damages,[188] proximate cause,[189] and release-dismissal agreements.[190] The Court has emphasized that common law rules should not be mechanically applied to constitutional torts.[191] Tort concepts must be adapted to the constitutional tort context in order to serve the distinctive needs of § 1983 liability.[192] Even so, the Court has routinely looked to tort law for guidance in resolving § 1983 issues.[193] This pattern of decisionmaking comports with an even broader pattern of constitutional interpretation rooted in recurring or traditional conventions of subconstitutional law.[194]

The anomaly in § 1983 doctrine is the one highlighted here—that is, the all-but-complete lack of resort to ordinary tort principles in local government liability cases. For some, this absence may be justified by conventional means of statutory interpretation. In particular, the Court in Monell rejected vicarious liability based on its reading of legislative history.[195] But the Court’s adoption of policy or custom had no basis either in legislative history or in background tort law. Furthermore, the no respondeat superior rule does not tell us the extent to which other background common law doctrines should bear on municipal liability. In particular, nothing in the legislative history of § 1983 suggests that courts should also reject the common law duty of reasonable care. On that issue, the 42nd Congress was silent.

(Objective) Negligence vs. (Subjective) Deliberate Indifference

As defined by Brown and Connick, deliberate indifference is a subjective test.[196] It requires a showing of “conscious disregard of an obvious risk.”[197] In § 1983 municipal liability cases, “[a] showing of simple or even heightened negligence will not suffice” for deliberate indifference.[198] Thus, it is not enough that the policymaker should know because a reasonable person in his position would know and act on that knowledge. The Court’s description of deliberate indifference in these cases resembles the Restatement of Tort’s definition of recklessness, which requires proof that “the person knows of the risk of harm created by the conduct or knows facts that make the risk obvious to another in the person’s situation.”[199] In § 1983 municipal liability cases, courts have used this subjective and objective distinction as a basis for rejecting liability under the deliberate indifference test even when the plaintiff has produced proof of policymaker fault.[200]

A different approach better serves accommodation. To achieve the proper balance among competing values,[201] municipal liability doctrine should give due weight to all of the competing interests and not merely the interest in avoiding any risk of vicarious liability. Following the model of ordinary tort law, adjudication of the negligence issue for municipal liability would be an objective test. It would require evaluation of the policymaker’s conduct by the standard of a reasonable policymaker, a test that is external to that officer.[202] Negligence turns not only on the subjective knowledge or motives of policymakers, but also on their access to information and their ability to act on it.[203] Negligence is especially well-suited to the task—often presented by indirect-effect cases—of determining whether a policymaker’s failure to act is culpable. It is almost incoherent to talk about a conscious decision to disregard a known risk in a failure-to-train case or a failure-to-address, bottom-up-custom case. Omissions of this kind are seldom the product of highly focused and deliberate thought, but such omissions may still reflect failure to foresee events a reasonable person would have foreseen, or failure to take remedial steps a reasonable person would have taken.

A rule that governments are liable if their policymakers fail to exercise reasonable care to safeguard constitutional rights in indirect-effect cases would align municipal liability doctrine with the accommodation principle. Negligence-based liability would serve all of the competing policies, at least to some extent. Plaintiffs would obtain vindication when they could show a constitutional violation, causation of harm, and cognizable fault on the part of government policymakers. In the same way, the negligence principle would provide incentives for taking appropriate precautions against constitutional violations by establishing the benchmark of unreasonableness. On the other side of the balance, from the perspective of municipal defendants, the fault principle would protect against the unfairness of imposing liability without adequate warning. In addition, it would amply protect against the overly burdensome liability, and attendant costs, associated with a strict liability respondeat superior rule. As with any effort to accommodate competing values, the negligence approach would fully satisfy the interests of neither § 1983 plaintiffs nor defendants. In particular, many victims of constitutional violations will not be able to recover any damages at all. Even so, under a reasonableness approach, plaintiffs will prevail in that set of cases in which their interests are most pressing, and the same is true of local government defendants.

From the perspective of a plaintiff who will have already proven the violation of his constitutional right, the goals of vindicating and deterring constitutional violations point to negligence over current doctrine for a simple reason: negligence produces liability more often and thus produces more vindication and a stronger deterrent. On the other hand, ordinary tort law principles dictate that fairness to defendants is adequately served by absolving them from liability when either (a) they cannot reasonably foresee that one alternative is less risky than another,[204] or (b) there is no safe choice because all of the alternatives carry substantial risk.[205] The negligence principle protects governments from liability for their policymakers’ reasonable responses to constitutional risk, but not when the policymaker has sufficient knowledge, time, and resources to make safer choices.

In contrast to the negligence approach, deliberate indifference saves governments from liability in cases in which policymakers do have the information and the means to minimize constitutional risks but fail to take the necessary curative steps based on their own inadequacy or laxity in failing subjectively to recognize those risks. Governments are sure to argue, however vaguely, that shifting from deliberate indifference to a reasonable care standard would come at too high a price. But at the least, traditional tort principles stretching back to Vaughan v. Menlove,[206] the leading case on negligence being an objective standard, suggest that such a claim is off the mark. The durability of the negligence principle surely counts in its favor.

Reasonable Care in the § 1983 Municipal Liability Context

One advantage of a negligence-based approach is that it is a familiar concept with a long history. As a result, there exist many cases that draw lines between sets of facts as to which liability is or is not appropriate and set out the respective roles of judge and jury in drawing those lines. The resolution of policy or custom issues would benefit from the attention courts have given to the development of negligence doctrine over the centuries. As we have seen, the common law should not serve as a set of “prefabricated components” for determining § 1983 liability.[207] Rather, “in applying, selecting among, or adjusting common law approaches, courts must closely attend to the values and purposes of the constitutional right at issue.”[208] Put another way, it makes perfect sense for the Court to borrow—and modify—common law negligence principles as it deals with indirect-effect Monell actions. The ensuing sections show how courts may fine-tune negligence principles to work well in the municipal liability context without risking the collapse into respondeat superior that the Court so greatly fears.


Under a negligence model, municipal liability in indirect-effect cases turns on whether the local government’s policymakers have exercised reasonable care. For this purpose, the relevant actors are not those employees who handle the problem at hand but, instead, include a narrow band of higher‑ups who fall within the description of “some official or body that has the responsibility for making law or setting policy in any given area of a local government’s business.”[209] The Court distinguishes between the policymaker and the actor who has violated the plaintiff’s rights: “Once those officials who have the power to make official policy on a particular issue have been identified, it is for the jury to determine whether their decisions have caused the deprivation of rights at issue.”[210] No matter how badly a subordinate may have behaved, the government is not liable unless the plaintiff shows the requisite culpability on the part of the policymaker.

The workability of the reasonable care approach depends on adherence to the distinction between policymakers and other officers. Lower courts have not always maintained that distinction or have blurred the line between the two.[211] Some courts treat delegation cases as instances in which “final policymakers . . . create [municipal] liability by . . . delegating policymaking authority to a subordinate.”[212] Some courts hold that officers who are not constrained by municipal policies are themselves policymakers.[213] In Webb v. Sloan,[214] for example, police in Carson City, Nevada, had mistakenly arrested Webb. When the error was discovered, Webb refused to sign a release‑dismissal agreement,[215] by which he would have given up his false arrest claim in exchange for dismissal of the criminal charge. In retaliation for this, a deputy district attorney prosecuted Webb without probable cause. Upon acquittal, Webb sued Carson City and the police officer and won a jury verdict against the city.[216] The court affirmed on the ground that the deputy district attorney was a municipal policymaker.[217] The court laid weight on a Nevada statute that provided that deputy district attorneys “may transact all official business relating to the officer to the same extent as their principals.”[218] The court reasoned that the statute dictated that “if principal district attorneys are final policymakers, then so are their deputies.”[219] Furthermore, “Carson City presented no evidence that its principal district attorney actually has constrained the deputies’ authority,”[220] thus suggesting that the district attorney’s casualness in asserting his own authority created in his underlings a sort of “pinch hitter” policymaker status.[221]

Within a duty-of-reasonable-care framework for municipal liability, the issue in Webb would be framed differently: The Carson City district attorney would be the sole relevant policymaker for prosecutorial decisions. Municipal liability would turn on whether the district attorney paid sufficient attention to the selection, training, and supervision of deputies. Evidence that the district attorney gave them free rein would help establish negligence on his part. On proof that negligence caused a violation of Webb’s constitutional rights, Carson City would be liable under § 1983. But evidence of the district attorney’s casualness would not transform the deputy into a policymaker.

The general principle illustrated by this example is that all such cases should depend on whether the higher-up officer has exercised reasonable care in supervising the subordinate. If he has done so, the delegation of decision making to the subordinate would not create liability, no matter how much discretion the subordinate has exercised in the case at hand. By contrast, under Webb every case turns on the vagaries of state law and highly contextual assessments of whether and how much a recognized policymaker has assigned to others his policymaking status.[222] In Webb, for example, the court distinguished an earlier case in which it had held that “deputy prosecutors in Hawaii did not have final policymaking authority,”[223] on the ground that Hawaii law operated differently.[224] The reasonable care approach does not require this kind of state-by-state (and case-by-case) analysis. Under a negligence approach, the issue is whether policymakers who delegate authority to underlings should have foreseen the risk of constitutional violations by the subordinates and should have taken steps to head off the violation that occurred. Whether they have done so would depend, as it should, on the application of negligence principles borrowed from tort law and adapted to this purpose in specialized ways that are detailed below.

Negligence Principles: Costs, Benefits, and Foreseeability

The negligence framework directs fact finders to balance the risks and benefits of a precaution.[225] That approach has already influenced municipal liability doctrine, even in the deliberate indifference regime. Some lower court cases, though they frame the issue in terms of deliberate indifference,[226] identify circumstances that deserve consideration in a fault system. In Dunn v. City of Elgin,[227] an inadequate training case, the court described the relevant factors:

Deliberate indifference may be shown in one of two ways. First, a municipality shows deliberate indifference when it fails to train its employees to handle a recurring situation that presents an obvious potential for a constitutional violation and this failure to train results in a constitutional violation. Second, a municipality shows deliberate indifference if it fails to provide further training after learning of a pattern of constitutional violations by the police.[228]

Translated into the vocabulary of reasonable care, these two prongs identify ways in which the likelihood of constitutional violations may become sufficiently foreseeable to justify further training. The “recurring situation” and the “pattern of constitutional violations” relate to the foreseeability of violations and they are equally significant in fault-in-hiring, bottom-up custom, ratification, and delegation cases.[229] The move to a negligence standard would only modify the policymaker’s standard, holding them to the objective standard of a reasonable person. Liability may be appropriate even when the potential for a constitutional violation is not quite as obvious, and a pattern of constitutional violations may not be needed if other evidence supports a finding of foreseeability.

Cost-benefit analysis provides useful guidance in the law, but it is an abstract concept and thus lacks the concrete specificity needed to guide the adjudication of most tort issues, whether in common law negligence cases or under § 1983. Ordinary tort law has generated a body of context-specific principles for resolving negligence issues, and these principles furnish useful analogs for use in § 1983 municipal liability cases. Four of these principles seem especially pertinent: One involves “customary practice.” Suppose, for example, local governments follow some customary practices to minimize constitutional violations. Either the plaintiff or the defendant may find it helpful to introduce evidence of that practice and the policymaker’s adherence to or deviation from it.[230] Similarly, policymaker fault sometimes will be based on whether she followed a professional standard, as is the practice in common law medical and legal malpractice litigation.[231] The second principle is that of “negligence per se” for statutory violations. Under it, courts may deem state or federal statutes or regulations relevant to the determination of fault, especially if their purpose is to diminish the likelihood of constitutional violations.[232] Third, the principle of res ipsa loquitur applies when fault is apparent from the event that caused injury: Some constitutional violations may be so egregious that the policymaker in charge of preventing them can be held liable without further evidence of fault, under the principle, that “the thing speaks for itself.”[233] Fourth, an actor may avoid liability even when his action carries with it unmistakable dangers, provided the risk of not acting is even more dangerous.[234] In Raimondo v. Harding, for example, the plaintiff darted into traffic to escape attackers.[235] Though he could foresee danger from the traffic, the court found that he exercised reasonable care under the principle that his obligation is to act only “as a reasonably prudent person would act under the same emergency circumstances.”[236] Municipalities sued under § 1983—particularly when sued for policymakers’ responses to emergencies—may benefit from this principle.

Reshaping Municipal Liability Doctrine

Adoption of a municipal duty of reasonable care to prevent constitutional violations would treat delegation, ratification, bottom-up custom, inadequate training, faulty hiring, and slipshod supervision as factual variations of a single unitary theory of recovery. All of these fact patterns are tied together by the same cords: first, a policymaker whose duties include oversight of a subordinate; second, a constitutional violation by the subordinate; and third, a causal connection between inadequate oversight and the constitutional violation. Delegation, ratification, inadequate training, and bottom-up custom all describe ways in which policymakers interact with subordinates. In any large institution with a hierarchical structure, policymakers assign many duties to subordinates, who then assign tasks to their subordinates, and on down the line. When policymakers review and approve of decisions made by underlings, their action seems to fit Justice O’Connor’s ratification fact pattern. When policymakers do not intervene, the policymaker’s act can be described as “delegation.” When policymakers do not intervene against widespread street-level practices, the facts may be described as bottom-up custom. When the policymaker-subordinate relationship involves training, hiring, or supervision, the fact pattern fits Harris or Brown. Some fact patterns will involve more than one of these features. For example, acquiescence in decisions by subordinates may be described either as bottom-up custom, or as delegation to subordinates.[237] Cursory review may be labeled as either delegation or ratification.[238]

It is useless to try to put fact patterns into pigeonholes when the pigeonholes have no significance for the liability rule. This point is illustrated in common law tort by Rowland v. Christian.[239] In that case, the California Supreme Court rejected the old categorical approach to land occupier liability, with distinctions between trespassers, licensees, and invitees. That framework did “not reflect the major factors which should determine whether immunity should be conferred upon the possessor of land.”[240] Instead of the categorical approach, “the proper test . . . [was] whether in the management of his property he has acted as a reasonable man in view of the probability of injury to others.”[241] As with land occupiers, a municipal liability doctrine that distinguishes among the categories does not help to identify instances in which the case for liability is stronger or weaker. A more promising approach would focus on whether the policymaker gives appropriate weight to the protection of constitutional rights in interacting with subordinates under that policymaker’s supervision, regardless of whether the fact pattern reflects delegation, ratification, bottom-up custom, inadequate training and hiring, or some combination of them. In all of these situations, liability should turn on whether policymakers exercised reasonable care.

Even so, there are reasons to pay attention to the fact patterns that shape the current doctrine. Again, Rowland provides a helpful analogy. The court observed that “the plaintiff’s status as a trespasser, licensee, or invitee may in the light of the facts giving rise to such status have some bearing on the question of liability.”[242] Thus, for example, the variations among the delegation, ratification, bottom-up custom and failure to train fact patterns guide judges, jurors, and lawyers, simply because the variations identify the specific questions that these individuals must address in determining fault in a particular context and the evidence that bears on those questions. Tracking Rowland’s approach to common law land occupier liability, a fault-based § 1983 municipal liability doctrine would recognize that the differing contexts in which policymakers act may well afford those policymakers different levels of notice of constitutional danger and greater or lesser means to minimize those dangers.


As we have seen, the issue in § 1983 delegation cases should not be whether policymakers have delegated authority to make a decision but whether policymakers are at fault in supervising subordinates to whom authority has been given. Liability will often depend on the tasks that the policymaker delegates and to whom they delegate those tasks. Key factors are foreseeability of harm in a given context and opportunity to minimize attendant risks. Policymakers, for example, would often be found at fault when they fail to oversee subordinates’ decisions that have an impact on many people, but not so often when the foreseeable consequences of the subordinate’s decision are highly limited. Fault would be found more often with regard to policymaker neglect of subordinates’ decisions when those decisions have constitutionally relevant consequences, such as oversight of the local police, the jail, and the schools—as opposed, for example, to file clerks or janitors. Policymakers also would be more often found at fault when they have reason to know that a given subordinate has in the past acted in unconstitutional, or arguably unconstitutional, ways.

A fault principle of this sort would not collapse into vicarious liability. Many constitutionally relevant duties are delegated to street-level officers, but supervision of their work is typically far removed from policymakers. Policymakers would not be at fault for failing to oversee every act of every police officer or school official because they will typically have little opportunity to oversee that officer’s actions. For example, in Davison v. Randall, the policymakers were the Loudon County, Virginia Board of Supervisors.[243] In retaliation for Davison’s protected speech, Louise Randall, the chair of the board, banned Davison from the chair’s Facebook page.[244] Loudon County was not liable for Randall’s violation of Davison’s First Amendment rights, because this was “a one-off unilateral decision . . . in the heat of the moment.”[245] Absent a pattern of such behavior by Randall, this case seems rightly decided under the negligence approach. One-off events are often unforeseeable.[246] On the other hand, delegation is not an all-purpose shield against municipal liability, as it seems to be for a few courts. For example, Penley v. McDowell County Board of Education held that “a county board of education cannot be held liable for personnel decisions over which it did not retain final review authority.”[247] From a negligence perspective, the issue is not whether the Board of Education retained final review authority, but whether it acted reasonably in turning over all of its authority to subordinates, especially without any follow-up supervision.

The negligence principle holds that it is sometimes reasonable to delegate decision making to subordinates without review, when the circumstances indicate the unlikelihood of constitutional violations. This may be true, for example, of the fire department or the municipal water and trash collection office, at least until complaints begin to surface. Other instances of delegation may present a greater likelihood of constitutional violations and thus a need for closer supervision in a reasonable care regime. In Gschwind v. Heiden,[248] for example, an Illinois teacher sued a school district on free speech grounds when, on the recommendation of the principal, the school board fired him. According to the court, “[i]n Illinois the school board is the ultimate policymaking body with regard to personnel decisions.”[249] But the school board “allow[ed] principals and assistant principals to make evaluation and employment decisions as they s[aw] fit . . . and the Board of Education follow[ed] these decisions and recommendations.”[250] The court found that the plaintiff stated a good claim against the school district because “[t]his was evidence of a policy of . . . condoning unconstitutional terminations.”[251]

Under a negligence approach, this outcome could not rest solely on the delegation of authority to the principal, as it evidently did for the court. Instead, the issue would be whether the school board acted with reasonable care in giving principals such broad discretion. The result may well be the same, since teachers often charge school authorities with free speech violations, and not infrequently prevail on the merits. A similar broad delegation, however, would probably not justify a finding of fault on the policymaker’s part if the aggrieved employee worked for the municipal water department, a line of work that produces much less First Amendment litigation.


When policymakers more actively oversee decisions by subordinates, the degree of oversight may justify characterization of the case as “ratification” rather than “delegation.”[252] In such cases, policymakers who undertake to monitor their subordinates’ work should be obliged to take reasonable care to minimize the risk of constitutional violations. It should not be necessary to show that the policymaker “consciously chose to ignore” violations.[253] The specific facts of a given instance of ratification bear on the determination of what a reasonable policymaker, considering available resources, should know about the incident and what a reasonable policymaker could do to limit constitutional risk. In a case like Praprotnik, liability would depend on facts that were never established, because they were not relevant, under the Court’s approach. In a negligence regime, it would be important to determine whether St. Louis policymakers knew or should have known of the risk that Hamsher, Praprotnik’s immediate supervisor, would violate First Amendment rights. If similar incidents had occurred in the past, or if Hamsher had indicated a particular hostility to Praprotnik’s protected speech, these warning signs would have required heightened attention from policymakers, and their failure to at least investigate the circumstances of Hamsher’s treatment of Praprotnik might justify a finding of fault.


Though the policymaker’s role in bottom-up custom cases is factually distinct from delegation and ratification, it should be governed by the same obligation of reasonable care. When policymakers know enough about street‑level practice to make choices about further investigation, their negligence in ignoring the risk of violations is an adequate ground for imposing liability on the municipality, just as if they had delegated authority without sufficient oversight or implicitly ratified an underling’s decision despite reasons to doubt the constitutional validity of the grounds for that decision.[254] Custom implies a widespread practice.[255] Under a negligence approach, however, the key factor is foreseeability. In some cases, the custom may arise out of only a few instances of unconstitutional conduct. In Baron v. Suffolk County Sheriff’s Department,[256] the plaintiff successfully proved a custom of informal punishment of jail employees who informed authorities of constitutional violations by coworkers, despite an inability to show a pattern of such punishments of any officers other than the plaintiff. In situations like this one, a custom—here, of improper punishment—results from only a few instances in which a plaintiff shows that he was abused. In upholding his claim, the court explained that a custom of informal abuse for snitching would probably dissuade snitching and thus lead to few instances of abuse.[257] By contrast, the city might well prevail if it can show that other officers who snitched on their colleagues were not punished, as policymakers on these facts would have less reason to suspect a pattern of harassment was afoot.

Inadequate Training, Hiring, and Supervision

In cases of inadequate training, hiring, and supervision, a reasonable care approach would probably produce different outcomes on the liability issue in Brown[258] and Connick.[259] Both of these cases overturned jury verdicts for plaintiffs under the stringent deliberate indifference standard,[260] which is not satisfied by “[a] showing of simple or even heightened negligence.”[261] In both cases, the jury instructions were proper. The Court focused on the sufficiency of the evidence to support the juries’ verdicts.[262] To be sure, these verdicts may be questioned under the deliberate indifference test. The Court in Brown may have correctly rejected the jury’s finding that Sheriff Moore was deliberately indifferent to constitutional rights in hiring Deputy Burns.[263] Even if that were so, however, Burns’ history of minor criminal-law violations provided notice to the sheriff that Burns might endanger constitutional rights, thus justifying a jury verdict that Moore violated the duty of reasonable care.[264]

In Connick, courts had overturned four convictions for Brady violations by New Orleans’ district attorney’s office over the past decade, though the factual backgrounds of those rulings differed from the plaintiff’s case.[265] The Court agreed with District Attorney Connick that “Thompson did not prove that he was on actual or constructive notice of, and therefore deliberately indifferent to, a need for more or different Brady training.”[266] Under a reasonable care approach, the four earlier violations could well have supported a jury verdict that the district attorney should have understood the risk of future violations and taken action to prevent them. A reasonable jury thus could determine that Connick should have foreseen this type of future constitutional violation, even if earlier violations differed in some respects.[267] At the least, negligence doctrine would not have dictated, as the Court’s ruling did, that a “pattern of similar constitutional violations by untrained employees is ‘ordinarily necessary’ to demonstrate deliberate indifference for purposes of failure to train.”[268] Under the negligence test, reasonable foreseeability is the key to liability, and such foreseeability may be shown by other means in addition to a history of on-all-fours past violations.[269]

Cases That Fall Outside the Categories

Some lower court cases do not fit neatly into any of these categories, yet the case for liability is strong, because the vindication and deterrence policies are strong. The reasonable care principle provides a useful framework for adjudicating them. For example, in Hill v. Cundiff, policymakers had adopted the “catch in the act” rule under which public schools could not discipline students for sexual harassment absent eyewitness testimony.[270] A teacher’s aide devised a scheme by which a female student would serve as “rape bait” and the male aggressor would be caught in the act.[271] The scheme backfired, however, and the student was tragically raped.[272] The court dismissed the student’s § 1983 suit against the school district, evidently on the ground that the plaintiff had not shown deliberate indifference.[273] Taking its cue from the failure-to-train cases, the court said that “[t]he Board could not have foreseen a rape-bait scheme that required an eighth-grade student to voluntarily subject herself to sexual harassment as a ‘known or obvious consequence’ of the ‘catch in the act’ policy or its training policies.”[274]

In a reasonable care regime, the Board’s conduct would be evaluated under an objective standard. The outcome may have been different, as the risk of such a scheme may be one the Board should have considered even if it was not a “known or obvious consequence.”[275] In particular, some judges might conclude that the school system’s draconian and unorthodox catch in the act rule may well lead to dangerous enforcement efforts of some kind, thus at least raising a duty to secure approval from the school principal before schemes to catch wrongdoers were launched.

While the facts of Hill are unlikely to recur, the case illustrates a larger point about governmental responsibility for constitutional violations, a point made by Christina Whitman long ago. As Professor Whitman observed: “[I]njuries can be brought about quite inadvertently through the workings of institutional structures—through the massing or fragmentation of authority, or by the creation of a culture in which responses and a sense of responsibility are distorted.”[276]

The categorical approach that currently dominates § 1983 municipal liability doctrine does not provide a neat doctrinal pigeonhole for cases such as Hill in which, at least arguably, a systemic institutional breakdown of one kind or another has occurred, but no specific policymaker has demonstrated deliberate indifference to constitutional rights. A strength of the reasonable care model is that it provides a framework that identifies the issues that need to be resolved to adjudicate such cases.

Does Negligence-Based Liability Collapse into Respondeat Superior?

Much of the Supreme Court’s § 1983 indirect municipal liability doctrine is shaped by the Court’s determination to avoid vicarious liability. In Brown, for example, the Court warned that “[w]here a court fails to adhere to rigorous requirements of culpability and causation, municipal liability collapses into respondeat superior liability.”[277] This line of reasoning is difficult to understand because the only formal step needed to avoid respondeat superior is to reject respondeat superior, as the Court did in Monell. Put simply, a properly applied negligence rule would not impose respondeat superior any more than does the deliberate indifference test.

Formal doctrine aside, the Court has feared the practical consequences of lowering the hurdle for plaintiffs in real-world litigation. It said in Connick that “[a] less stringent standard of fault for a failure-to-train ‘claim would result in de facto respondeat superior liability’ . . . .”[278] But why does the Court fear that a fault-based system would collapse into vicarious liability? This question does not receive any answer in the Court’s opinions. Perhaps the Justices worry that juries, in the application of a less stringent standard of liability, would too readily find fault and that lower courts would fail to control juries. Under a more stringent standard, such as deliberate indifference, plaintiffs will win less often, and thus they will rarely win cases that they deserve to lose.

But the Court’s position has an important corollary that the majority is understandably reluctant to articulate in its opinions: To the extent the imposing obstacle of deliberate indifference is justified solely by the need for an extra layer of insulation from liability to ward off the menace of a collapse into vicarious liability, the merit of the deliberate indifference test turns on whether the extra layer is actually needed. If that layer is in fact superfluous, or largely so, the logical implication is that plaintiffs who face deliberate indifference also lose cases that they may deserve to win.

Seen in this light, the Court’s stance rests on two unarticulated, unexamined, and unsupported assumptions. First, it seems to assume that judges and juries charged with applying the negligence principle will often drift into respondeat superior, a strict liability approach.[279] Second, the Court seems to ignore other means for controlling overly broad liability, such that a super strict, state-of-mind requirement is deemed necessary to achieve the proper constraint. The Court’s approach, however, gives too little credit to the ability of judges and juries to maintain the integrity of negligence law—just as the law calls upon them to do in countless numbers of ordinary tort cases, including cases brought against cities. It is not necessary to sacrifice the benefits of an accommodation-oriented approach to municipal liability for the sake of avoiding a hypothetical problem. Even if the danger has some possible grounding in reality, its speculative and uncertain quality must be weighed against the undeniable fact that under current law the victims of constitutional violations automatically lose their cases when they can only show that the municipal defendant has acted unreasonably.

The ensuing subsections discuss two reasons to question the validity of the Court’s concern that a reasonable care approach would collapse into vicarious liability: First, courts have successfully maintained the viability of the distinction between negligence and respondeat superior in other contexts. Second, the tort law concepts of cause-in-fact and proximate cause provide effective means for addressing the Court’s “no liability without responsibility” concern in a more focused way than does deliberate indifference, all without a wholesale sacrifice of injured plaintiffs’ legitimate interest in vindication and deterrence of constitutional violations.

The Distinction Between Negligence and Respondeat Superior in Other Contexts

In Monell, the Court moved from rejection of no respondeat superior to adoption of policy or custom in two sentences, as though no other alternatives were available.[280] And yet, in the context of negligent hiring and supervision, common law courts routinely distinguish between vicarious liability and other theories of liability.[281] That distinction is illustrated by the large and complex body of law on the “course of employment” issue. Thus, the Restatement of Agency declares that an employee’s act occurs in the course of the employment only if “it is actuated, at least in part, by a purpose to serve the master.”[282] Illustrations of the body of law devoted to distinguishing between respondeat superior and other theories of liability include cases that distinguish between “frolic and detour,”[283] cases that focus on employer fault when the activity is clearly not within the scope of the employment,[284] and cases that ask whether the defendant has given “apparent authority” to the tortfeasor.[285] The very persistence of these doctrines demonstrates their workability and suggests that courts are fully capable of maintaining the integrity of rules that base liability on employer fault without falling into vicarious liability.

Developments in public law also support the viability of negligence-based liability. Indeed, the Court itself recently adopted an approach that distinguishes between respondeat superior and negligence in the context of sexual harassment claims brought against employers under Title VII of the Civil Rights Act of 1964. In Vance v. Ball State University,[286] the Court held that the employer would not be liable on a respondeat superior basis for such claims unless the harasser was someone whom “the employer has empowered . . . to take tangible employment actions against the victim.”[287] More particularly, it went on to conclude that “[n]egligence provides the better framework for evaluating an employer’s liability when a harassing employee lacks the power to take tangible employment actions.”[288] Justice Alito’s majority opinion framed the issue as a choice between negligence and vicarious liability, contained nothing to suggest any worries that negligence may collapse into respondeat superior, and did not consider adoption of a deliberate indifference test as an alternative approach.[289] Justice Alito noted that the “negligence standard . . . is thought to provide adequate protection for tort plaintiffs in many other situations”[290] and expressed confidence that negligence could handle “the variety of situations that will inevitably arise” in non-supervisor harassment cases.[291] An implicit premise of the Court’s reasoning is that judges and juries will be able to maintain the negligence and vicarious liability distinction. Vance, to say the least, casts doubt on the Court’s apparent distrust of the ability or willingness of federal courts to maintain the exact same distinction in § 1983 municipal liability cases.

Other Constraints on Liability: Causation in Fact and Proximate Cause

The Court has paid too little attention to the other means courts have at their disposal to ensure that governments are held liable only for those constitutional injuries fairly attributable to municipal policies and customs. To begin with, neither a government nor an official can be held liable without proof of actual harm because “the abstract value of a constitutional right may not form the basis for § 1983 damages.”[292] The Court also has provided local governments with an extra level of protection by banning punitive damages against municipalities in § 1983 suits.[293] Although adoption of the negligence approach would lower the hurdles that block access to relief for some plaintiffs, it would also bring into focus the far-from-unlimited nature of municipal liability. Among other things, the negligence approach would demand reasonable care on the part of high-level officers and would cut back existing opportunities for recovery by disapproving of cases in which lower courts have named low-level officials as policymakers.[294] A negligence-based approach also would reject the analytical framework of cases like Gschwind v. Heiden,[295] in which courts have treated delegation as a stand-alone for liability, without regard to the reasonableness of the policymakers’ choice to reassign decisionmaking to a subordinate.[296]

Ordinary tort principles accord considerable weight to the goal of assuring that liability is not imposed unless an actor is responsible for the plaintiff’s harm. Besides fault, the burden of proof is on the plaintiff to show that the defendant’s negligent act is both the cause in fact of the harm and the proximate cause of the injury. Indeed, as the ensuing discussion highlights, both Brown[297] and Connick[298] overturned jury verdicts for insufficient evidence of deliberate indifference, while ignoring opportunities to employ cause-in-fact and proximate-cause principles to assure that municipalities are not held liable unless they are truly responsible for constitutional injuries.[299]

Cause-in-Fact in Connick v. Thompson

The issue in Connick was whether New Orleans was subject to liability for supervisors’ failure to train assistant district attorneys about their Brady obligation to provide exculpatory evidence to criminal defendants.[300] Despite a jury verdict for the plaintiff, the Court held that the plaintiff had not shown deliberate indifference on the part of District Attorney Connick, the final policymaker.[301] Brady is a well-known criminal procedure rule. In his opinion for the Court, Justice Thomas pointed out that assistant district attorneys have plenty of opportunities to learn about it, not only in law school and in preparing for the bar but also in continuing legal education courses as well.[302] Justice Thomas linked Brady’s prominence to the deliberate indifference standard.[303] He reasoned that, in light of these many opportunities to learn about Brady, “recurring constitutional violations are not the ‘obvious consequence’ of failing to provide prosecutors with formal, in-house training about how to obey the law.”[304]

Close inspection reveals, however, that Justice Thomas’s fancy footwork with the deliberate indifference standard diverted attention from the strongest objection to recovery in the Connick case: Even if Thompson could establish a violation of the City’s duty of care, the city would probably have had a causation-based defense—a defense that would have been available even if the liability rule were negligence rather than deliberate indifference.[305] The black-letter doctrine on cause in fact is that “[c]onduct is a factual cause of harm when the harm would not have occurred absent the conduct.”[306] Thompson’s prosecutors failed to comply with Brady despite many opportunities to learn about that rule and the consequences of violating it. Their conduct in the face of the opportunities belies the notion that a lack of in-house training made a difference.[307] To be sure, a court might have concluded that Connick’s failure to train was one of several causes of the Brady violation, triggering application of the principle that any one of a set of multiple sufficient causes may be held liable.[308] Yet New Orleans might well have won despite this principle. Just as a defendant who negligently starts a small fire may not be held liable when it joins with a much larger one, New Orleans’ negligent failure to provide adequate in-house training may have made only a trivial contribution to the causal set of circumstances that led to the Brady violation.[309]

There is a further point to be made on the topic of cause-in-fact as a tool for limiting municipal liability. In the common law, “policy considerations frequently exert their influence when we attempt to resolve a simple cause inquiry . . . .”[310] That being so, the policy of limiting municipal governments’ exposure to liability without responsibility can be implemented by requiring especially strong proof of cause-in-fact for the plaintiff to avoid summary judgment or judgment for the defendant “as a matter of law” in a § 1983 municipal liability case.[311]

Proximate Cause in Board of County Commissioner of Bryan County v. Brown

Sheriff B.J. Moore, the county policymaker responsible for hiring deputies, employed Burns without first checking his background.[312] Burns then used excessive force in pulling Brown from her car.[313] Although investigation would have revealed Burns’ previous misdemeanor convictions for a number of crimes, including one for assault and battery,[314] the Court overturned a jury verdict for Brown against Bryan County.[315] Applying the deliberate indifference test, the Court found “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 use of excessive force would follow.”[316] Under a negligence standard, the culpability threshold would be lower and the Court would have been hard-pressed to find the evidence insufficient to support a verdict for Brown.

Even so, Bryan County might well have won the case on proximate cause grounds, regardless of the plaintiff’s demonstration of fault and causation in fact. The relevant black-letter rule provides that “[a]n actor’s liability is limited to those harms that result from the risks that made the actor’s conduct tortious.”[317] The risk that made Sheriff Moore’s conduct tortious is that Burns would use excessive force, but the main evidentiary basis for Burns’ propensity for violence was a “single incident” in which he had pleaded guilty to misdemeanor assault and battery, public drunkenness, and resisting arrest.[318] These charges arose “from a fight on a college campus where Burns was a student.”[319] From a proximate cause perspective, Bryan County might well prevail even if Moore was negligent in hiring Burns because of the difference between the rowdiness one could foresee from Burns and the serious misuse of power on his part that occurred during the traffic stop. Indeed, Justice O’Connor’s opinion for the Court gestures toward this very rationale. Thus, “culpability . . . must depend on a finding that this officer was highly likely to inflict the particular injury suffered by the plaintiff.”[320] The difference between a fault-based regime and the Court’s approach is that, within the former, culpability and liability would be distinct issues. And while culpability would be established by showing negligence, liability would depend on demonstrating a meaningful causal link between the policymaker’s negligence and the constitutional violation.[321]

Whether a particular link is sufficiently close to produce liability would depend on whether, in the particular context, recovery would further the vindication and deterrence aims of constitutional tort enough to justify the costs of municipal liability. In ordinary tort law, courts use proximate cause doctrine to answer the question of whether fault and injury relate closely enough to justify recovery.[322] Brown illustrates the relevance of these common law principles to a particular set of facts.[323] In certain types of § 1983 litigation, proximate cause principles may produce categorical limits on liability. For example, the main common-law proximate cause principle limits liability “to those harms that result from the risks that made the actor’s conduct tortious.”[324] John Jeffries has argued, by analogy, that “compensation for violations of constitutional rights should encompass only constitutionally relevant injuries—that is, injuries within the risks that the constitutional prohibition seeks to avoid.”[325] He suggests, for example, that municipal liability for shutting down an adult bookstore in violation of the owner’s First Amendment rights may not extend to the economic loss from fewer customers and purchases because “the damage to [the] business enterprise is not the sort of injury that the first amendment is concerned to prevent.”[326] The point, for present purposes, is not to endorse Jeffries’ (admittedly tentative) resolution of the hypothetical.[327] The point is that limitations rooted in proximate cause principles under common law can be employed to narrow municipal liability within a fault system, without obliging the plaintiff to meet a stringent deliberate indifference test.


Critics of the Court’s Monell doctrine include several Supreme Court Justices. Writing in dissent for himself and two others in Brown,[328] Justice Breyer directed his fire at the “basic distinction [drawn in Monell] between liability that rests upon policy and liability that is vicarious.”[329] As Justice Breyer points out, that distinction “has generated a body of interpretive law that is so complex that the law has become difficult to apply.”[330] Justice Breyer pointed out that the confusion results from the extension of the doctrine to include indirect-effect municipal liability for the acts of underlings. The result is that “[f]actual and legal changes have divorced the law from the distinction’s apparent original purposes,”[331] such that “later law has made the distinction, not simply wrong, but obsolete and a potential source of confusion.”[332] Even Justice Breyer, however, stopped short of advocating adoption of a generalized respondeat superior approach, though some academic critics have done so.[333] Instead, he asserted only that “the case for reexamination is a strong one.”[334]

The inadequacy of current doctrine does not mean that generalized respondeat superior is the right solution. Among other objections, vicarious liability does not even attempt to achieve accommodation. Its increased burdens on governments may siphon resources from other, arguably more socially productive uses,[335] and those costs may, in the long run, hinder the growth of constitutional protections if judges worry that new guarantees may cost too much in terms of payments for past injuries.[336] Also, the benefit of the imposition of those increased costs may be small, as vicarious liability may not deter any better than negligence.[337] As a practical matter, it is unrealistic to think that the Court as currently constituted, or that any politically-attuned group of federal legislators, would expand municipal liability to this furthest point.

A danger of polarized debates is that the very terms of the discussion deflect attention from solutions that can satisfy some, if not all, of the interests of both sides. Such is the case with § 1983 municipal liability. This Article has identified a middle-ground approach between (1) the Court’s largely incoherent current case law[338] and (2) vicarious liability for all constitutional torts committed by government employees in the course of their employment. This alternative starts with the proposition, drawn from other § 1983 cases, that the overall point of the doctrine is to accommodate competing interests. Adoption of a municipal duty of reasonable care, under which cities would be held liable for unreasonable policymaker failings in overseeing subordinates’ work, would serve the accommodation principle by vindicating constitutional rights and deterring violations more effectively than current law, all without incurring the heavy costs that would attend across-the-board adoption of the respondeat superior approach.

  1. * Carter Professor, University of Georgia Law School. Thanks to participants in a workshop at the University of Tennessee Law School for helpful comments on a draft. The author also wishes to thank his colleagues, Dan Coenen and Tom Eaton, for their help.
  2. . The statute provides, in part, that “[e]very person who, under color of [state law] subjects, or causes to be subjected, any . . . person . . . to the deprivation of [constitutional rights] shall be liable to the party injured in an action at law, suit in equity, or other proper proceeding for redress.” 42 U.S.C. § 1983 (2018).
  3. . Monell v. Dep’t of Soc. Servs., 436 U.S. 658, 690, 694 (1978) (explaining that local governments are “persons” subject to suit under § 1983). State governments are not “persons” within the meaning of the statute and thus may not be sued under § 1983. See Will v. Michigan Dep’t of State Police, 491 U.S. 58, 64-67 (1989); cf. John C. Jeffries, Jr., The Liability Rule for Constitutional Torts, 99 Va. L. Rev. 207, 236 (2013) (criticizing this distinction and arguing that “states and cities should be treated the same”).
  4. . See, e.g., Gaston v. Ghosh, 920 F.3d 493, 494–95, 497 (7th Cir. 2019). But see Ira S. Bushey & Sons, Inc. v. United States, 398 F.2d 167, 171–72 (2d Cir. 1968). For discussion of the fairness and economic rationales for vicarious liability, see Gary T. Schwartz, The Hidden and Fundamental Issue of Employer Vicarious Liability, 69 S. Cal. L. Rev. 1739, 1749–64 (1996).
  5. . Monell, 436 U.S. at 694. See, e.g., Calgaro v. St. Louis Cty., 919 F.3d 1054, 1059 (8th Cir. 2019) (“[O]ne erroneous determination by a county employee . . . does not establish a policy or custom of the County that deprives [plaintiffs] of their constitutional rights.”).
  6. . See Monell, 436 U.S. at 701.
  7. . See Harlow v. Fitzgerald, 457 U.S. 800, 818 (1982).
  8. . See Monell, 436 U.S. at 694–95.
  9. . Owen v. City of Independence, 445 U.S. 622, 622 (1980).
  10. . See L.A. Cty. v. Humphries, 562 U.S. 29, 30 (2010).
  11. . Monell, 436 U.S. at 694.
  12. . See Bd. of Cty. Comm’rs v. Brown, 520 U.S. 397, 404–05 (1997) (distinguishing the two types of cases). The “direct” and “indirect” vocabulary is borrowed from Mann v. County of San Diego, 907 F.3d 1154, 1164 (9th Cir. 2018).
  13. . A policymaker is “some official or body that has the responsibility for making law or setting policy in any given area of a local government’s business.” City of St. Louis v. Praprotnik, 485 U.S. 112, 125 (1988) (plurality opinion). The determination of whether a given official is a policymaker on a particular topic requires “[r]eviewing the relevant legal materials, including state and local positive law, as well as custom or usage having the force of law.” Jett v. Dallas Indep. Sch. Dist., 491 U.S. 701, 737 (1989) (citation and internal quotation marks omitted). In practice, “the location of ‘policymaking’ authority” has been a “conceptually difficult problem” for lower courts. See Brown, 520 U.S. at 435 (Breyer, J., dissenting). In this Article, however, I stipulate the continuing viability of the Court’s “policymaker” approach to municipal liability.
  14. . Monell, 436 U.S. at 661 n.2.
  15. . Pembaur v. City of Cincinnati, 475 U.S. 469, 480 (1986).
  16. . Id. at 469.
  17. . See id. at 481; Monell, 436 U.S. at 661.
  18. . See generally Owen v. City of Independence, 445 U.S. 622, 657 (1980).
  19. . See Pembaur, 475 U.S. at 475 (discussing the holding of Steagald v. United States, 451 U.S. 204 (1981)).
  20. . See, e.g., Richard H. Fallon, Jr., Bidding Farewell to Constitutional Torts, 107 Calif. L. Rev. 933, 978–79 (2019); Edward C. Dawson, Replacing Monell Liability with Qualified Immunity for Municipal Defendants in 42 U.S.C. § 1983 Litigation, 86 U. Cin. L. Rev. 483 (2018).
  21. . See Harlow v. Fitzgerald, 457 U.S. 800, 818 (1982). This defense is available to policymakers as well as subordinates. Those officials who exercise legislative, judicial, or prosecutorial functions are absolutely immune from liability from damages, no matter how clearly established the law may be. See John C. Jeffries, Jr., et al., Civil Rights Actions: Enforcing the Constitution 49–89 (Saul Levmore et al. eds., 4th ed. 2018) [hereinafter Civil Rights Actions].
  22. . Id. at 57 n.a.
  23. . City of Canton v. Harris, 489 U.S. 378, 388 (1989).
  24. . Arrington-Bey v. City of Bedford Heights, 858 F.3d 988, 994 (6th Cir. 2017) (emphasis omitted) (citing Hagans v. Franklin Cty. Sheriff’s Office, 695 F.3d 505, 511 (6th Cir. 2012)) ; see also Bustillos v. El Paso Cty. Hosp. Dist., 892 F.3d 214, 222 (5th Cir. 2018) (“Because Bustillos did not demonstrate a clearly established right, it follows that her claims for deliberate indifference against the District also fail.”).
  25. . See John C. Jeffries, Jr., In Praise of the Eleventh Amendment and Section 1983, 84 Va. L. Rev. 47, 58 (1998) (“[F]ault has been explicitly reintroduced by requiring proof of ‘deliberate indifference’ to constitutional rights.”).
  26. . It should also be noted that the distinction between direct and indirect cases is not always a sharp one. See, e.g., Morgan v. Fairfield Cty., 903 F.3d 553, 566 (6th Cir. 2018) (distinguishing between municipal policy and an employee’s “interpretation of a policy”).
  27. . See, e.g., Gschwind v. Heiden, 692 F.3d 844 (7th Cir. 2012).
  28. . See, e.g., Amnesty Am. v. Town of W. Hartford, 361 F.3d 113, 119 (2d Cir. 2004).
  29. . See Bd. of Cty. Comm’rs v. Brown, 520 U.S. 397, 404 (1997) (“[A]n 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.”); see, e.g., Doe v. Vigo Cty., 905 F.3d 1038, 1044–45 (7th Cir. 2018); Owens v. Baltimore City State’s Attorneys Office, 767 F.3d 379, 402 (4th Cir. 2014) (stressing “condonation”); Griffin v. City of Opa-Locka, 261 F.3d 1295, 1308 (11th Cir. 2001).
  30. . See City of Canton v. Harris, 489 U.S. 378, 388 (1989); see also Connick v. Thompson, 563 U.S. 51, 61–62 (2011) (stating that deliberately indifferent inaction “is the functional equivalent of a decision by the city itself to violate the Constitution” (citing Harris, 489 U.S. at 395)).
  31. . See Civil Rights Actions, supra note 20, at 124–27.
  32. . In so doing, the Court has apparently rejected its dicta in Farmer v. Brennan, 511 U.S. 825, 841–42 (1994). Farmer was not a municipal liability case. It involved the Eighth Amendment rights of prisoners to be protected from assault. The Court held that the Eighth Amendment standard was whether prison officials had shown deliberate indifference in a subjective sense to the prisoner’s plight. In the course of its opinion, the Court characterized municipal liability deliberate indifference as an objective standard. Despite the more recent Supreme Court cases, a few circuit courts continue to follow Farmer and take the view that deliberate indifference (in the municipal liability context) is an objective test. See, e.g., Garza v. City of Donna, 922 F.3d 626, 635–36 (5th Cir. 2019) (citing Farmer, 511 U.S. at 837) (following Farmer, the court adopted “objective deliberate difference”); Doe v. Ft. Zumwalt R-II Sch. Dist., 920 F.3d 1184, 1189 (8th Cir. 2019) (citing Farmer, 511 U.S. at 841) (“This court applies an objective standard of deliberate indifference to [plaintiff’s] claim . . . .”); Robinson v. Pezzat, 818 F.3d 1, 12 (D.C. Cir. 2016) (citing Baker v. District of Columbia, 326 F.3d 1302, 1306 (D.C. Cir. 2003)); Harvey v. District of Columbia, 798 F.3d 1042, 1053 (D.C. Cir. 2015) (citing Jones v. Horne, 634 F.3d 588, 601 (D.C. Cir. 2011)).
  33. . Brown, 520 US. at 407. See, e.g., Kramer v. Wasatch Cty. Sheriff’s Office, 743 F.3d 726, 759 (10th Cir. 2014) (citing Harris, 489 U.S. at 391); Cash v. Cty. of Erie, 654 F.3d 324, 334 (2d Cir. 2011) (citing Amnesty, 361 F.3d at 128); Kelly v. Borough of Carlisle, 622 F.3d 248, 263 (3d Cir. 2010) (plurality opinion) (citing Berg v. Cty. of Allegheny, 219 F.3d 261, 279 (3d Cir. 2000)); Zarnow v. City of Wichita Falls, 614 F.3d 161, 170 (5th Cir. 2010).
  34. . Connick, 563 U.S. at 71; see also Brown, 520 U.S. at 410 (explaining that deliberate indifference is a stringent standard of fault, which imposes liability only for disregard of “plainly obvious” risks).
  35. . Many cases illustrate the subjective test. See, e.g., Doe, 905 F.3d at 1045 (applying the subjective test for custom); Alvarez v. City of Brownsville, 904 F.3d 382, 391 (5th Cir. 2018) (distinguishing a policymaker’s deliberate indifference from negligence); Winkler v. Madison Cty., 893 F.3d 877, 902 (6th Cir. 2018) (applying the subjective test for custom); Kelly, 622 F.3d at 264 (plurality opinion) (citing City of St. Louis v. Praprotnik, 485 U.S. 112, 127 (1988)) (stating subjective test for ratification); Amnesty, 361 F.3d at 126 (applying the subjective test for delegation); Christie v. Iopa, 176 F.3d 1231, 1238–39 (9th Cir. 1999) (plurality opinion) (quoting Praprotnik, 485 U.S. at 127) (applying the subjective test for ratification). Other courts require proof of the subjective awareness that is characteristic of deliberate indifference but do not use the term. See, e.g., Dixon v. Cty. of Cook, 819 F.3d 343, 348 (7th Cir. 2016) (“[A] Monell claim can prevail only if a policy-making official knows about [deficiencies] and fails to correct them”); Salvato v. Miley, 790 F.3d 1286, 1296 (11th Cir. 2015) (“[P]ersistent failure to take disciplinary action against officers can give rise to the inference that a municipality has ratified conduct . . . .” (quoting Thomas v. Roberts, 261 F.3d 1160, 1174 n.12 (11th Cir. 2001), vacated, 536 U.S. 953 (2002), remanded to 323 F.3d 950 (11th Cir. 2003))); Dean v. Cty. of Gage, 800 F.3d 945, 955 (8th Cir. 2015) (holding that liability is appropriate when policymaker “directed, endorsed, and encouraged” subordinates who violated constitutional rights); Jones v. Town of E. Haven, 691 F.3d 72, 81 (2d Cir. 2012) (explaining that the plaintiff can recover if he can show that the “policymaking official was aware of the employee’s unconstitutional actions and consciously chose to ignore them” (citing Amnesty, 361 F.3d at 127)); Love-Lane v. Martin, 355 F.3d 766, 783 (4th Cir. 2004) (explaining that plaintiff must show that the policymaker “was aware of the constitutional violation and either participated in, or otherwise condoned, it”).
  36. . See, e.g., Vodak v. City of Chicago, 639 F.3d 738, 747 (7th Cir. 2011) (asserting that Monell’s no respondeat superior holding was based on “what scholars agree are historical misreadings”); see also Karen M. Blum, Section 1983 Litigation: The Maze, the Muck, and the Madness, 23 Wm. & Mary B. of Rts. J. 913, 963 (2015); Akhil R. Amar, Fourth Amendment First Principles, 107 Harv. L. Rev. 757, 811–13 (1994). For an academic discussion of whether Monell correctly read the legislative history of § 1983, see Civil Rights Actions, supra note 20, at 49–89.
  37. . See, e.g., Joanna C. Schwartz, The Case Against Qualified Immunity, 93 Notre Dame L. Rev. 1797 (2018).
  38. . Id. at 1826.
  39. . Blum, supra note 35, at 913, 963.
  40. . Harlow v. Fitzgerald, 457 U.S. 800, 814 (1982) (discussing the qualified immunity doctrine).
  41. . Richard H. Fallon, Jr., Asking the Right Questions About Officer Immunity, 80 Fordham L. Rev. 479, 480 (2011).
  42. . Connick v. Thompson, 563 U.S. 51, 62 (2011) (quoting Bd. of Cty. Comm’rs v. Brown, 520 U.S. 397, 410 (1997)); see also Brown, 520 U.S. at 415. In Connick and Brown, for example, the Court overturned jury verdicts in the plaintiff’s favor, even though the juries had been properly instructed on deliberate indifference.
  43. . See, e.g., Fallon, supra note 19, at 976–79 (discussing the costs of constitutional tort liability and arguing that defenses should be available to allay those costs); John C. Jeffries, Jr., The Right-Remedy Gap in Constitutional Law, 109 Yale L.J. 87, 90–91 (1999) (discussing the value of a gap between rights and remedies in furthering constitutional innovation).
  44. . This is probably so at least so long as suggestions for law reform are directed to the Supreme Court rather than to Congress because judicial creativity is typically limited to comparatively small reforms. See Robert Keeton, Creative Continuity in the Law of Torts, 75 Harv. L. Rev. 463, 472–73 (1962) (“It is part of the common law tradition . . . that in so far as these breaks with precedent are accomplished by judicial decision rather than by statute they are in the nature of adjustment and adaptation of segments of doctrine. The more comprehensive breaks with the past are accomplished by statute, and the courts exercise their power of overruling precedent under an obligation of restraint.”).
  45. . See, e.g., Richard A. Epstein & Catherine M. Sharkey, Cases and Materials on Torts 631 (11th ed. 2016).
  46. . In this Article, “fault,” “lack of reasonable care,” and “negligence” are used as synonyms.
  47. . Brown, 520 U.S. at 415; see also Connick, 563 U.S. at 62 (“A less stringent standard of fault for a failure-to-train claim ‘would result in de facto respondeat superior liability on municipalities. . . .’” (quoting City of Canton v. Harris, 489 U.S. 378, 392 (1989))).
  48. . City of St. Louis v. Praprotnik, 485 U.S. 112, 123 (1989) (plurality opinion) (citing Pembaur v. City of Cincinnati, 475 U.S. 469, 480 (1986)).
  49. . Pembaur, 475 U.S. at 484; Owen v. City of Independence, 445 U.S. 622, 633 (1980); Monell v. Dep’t of Soc. Servs., 436 U.S. 658, 658 (1978).
  50. . Pembaur, 475 U.S. at 470.
  51. . Praprotnik, 485 U.S. at 112.
  52. . Id. at 117.
  53. . Id.
  54. . See id. at 128; see also id. at 142 (Brennan, J., concurring in the judgment).
  55. . Justice O’Connor wrote the plurality opinion, which was joined by Chief Justice Rehnquist and Justices White and Scalia. See id. at 114 (plurality opinion). Justice Brennan concurred in the judgment, in an opinion joined by Justices Marshall and Blackmun. See id. at 132 (Brennan, J., concurring in the judgment). Justice Stevens dissented, arguing that the jury verdict should have been upheld. See id. at 148 (Stevens, J., dissenting). Justice Kennedy did not participate. See id. at 114.
  56. . See id. at 114–32 (plurality opinion); id. at 132–48 (Brennan, J., concurring in the judgment).
  57. . See id. at 121–27 (plurality opinion).
  58. . Id. at 126.
  59. . Id.
  60. . Id.
  61. . Id.
  62. . Id. at 127.
  63. . See id. at 132–42 (Brennan, J., concurring in the judgment). Justice Brennan’s disagreement with Justice O’Connor centered on the identification of policymakers. Justice O’Connor asserted that “the identification of policymaking officials is a question of state law. . . . [I]t is not a question of fact in the usual sense.” Id. at 124 (plurality opinion); see also id. at 130. In Justice Brennan’s view, “[t]he identification of municipal policymakers is an essentially factual determination ‘in the usual sense,’ and is therefore rightly entrusted to a properly instructed jury.” Id. at 144 (Brennan, J., concurring in the judgment). The Court later resolved this dispute by borrowing elements from both opinions. See Jett v. Dallas Indep. Sch. Dist., 491 U.S. 701, 737 (1989). Writing for a majority on this issue, Justice O’Connor characterized the policymaker issue as “a legal question to be resolved by the trial judge,” but one to which “custom or usage” was also relevant.
  64. . See Praprotnik, 485 U.S. at 147 (Stevens, J., dissenting).
  65. . Id. at 126 (plurality opinion).
  66. . Id. at 130.
  67. . Id.
  68. . Jack C. Hanssen, Note, Municipal Liability Under 42 U.S.C. 1983 and the Ratification Theory of City of St. Louis v. Praprotnik: An Analysis of Federal Circuit Treatment, 27 J. Legis. 361, 367–77 (2001).
  69. . See, e.g., Darchak v. City of Chi. Bd. of Educ., 580 F.3d 622, 630 (7th Cir. 2009) (“Darchak must demonstrate that the Board either delegated final policymaking authority to Acevedo [a public school principal] or ratified Acevedo’s action.”).
  70. . See George M. Weaver, Ratification as an Exception to the § 1983 Causation Requirement: Plaintiff’s Opportunity or Illusion?, 89 Neb. L. Rev. 359, 394 (2010); Shari S. Weinman, Supervisory Liability under 42 U.S.C. Section 1983: Searching for the Deep Pocket, 56 Mo. L. Rev. 1041, 1062 (1991).
  71. . Praprotnik, 485 U.S. at 126 (plurality opinion).
  72. . Id.
  73. . King v. Kramer, 680 F.3d 1013, 1013 (7th Cir. 2012).
  74. . See id. at 1020.
  75. . Id. (emphasis added); see also Gschwind v. Heiden, 692 F.3d 844, 848 (7th Cir. 2012) (holding that the school district delegated the authority to make employment decisions to principals).
  76. . See, e.g., Vogt v. City of Hays, 844 F.3d 1235, 1250–52 (10th Cir. 2017); Liverman v. City of Petersburg, 844 F.3d 400, 413 (4th Cir. 2016); Thompson v. District of Columbia, 832 F.3d 339, 349–50 (D.C. Cir. 2016); Kristofek v. Vill. Orland Hills, 712 F.3d 979, 987 (7th Cir. 2013); Lytle v. Carl, 382 F.3d 978, 982–85 (9th Cir. 2004).
  77. . See Gschwind, 692 F.3d at 848.
  78. . Bolderson v. City of Wentzville, 840 F.3d 982, 986 (8th Cir. 2016).
  79. . Id. at 984.
  80. . Id.
  81. . Id. at 986.
  82. . Id.
  83. . Id.
  84. . City of St. Louis v. Praprotnik, 485 U.S 112, 130 (1988) (plurality opinion).
  85. . Penley v. McDowell Cty. Bd. of Educ., 876 F.3d 646, 646 (4th Cir. 2017).
  86. . Id. at 653 (citing Love-Lane v. Martin, 355 F.3d 766, 782 (4th Cir. 2004)). But cf. Praprotnik, 485 U.S. at 126 (plurality opinion) (“If . . . a city’s lawful policymakers could insulate the government from liability simply by delegating their policymaking authority to others, § 1983 could not serve its intended purpose.”).
  87. . See, e.g., Culbertson v. Lykos, 790 F.3d 608, 622–23 (5th Cir. 2015) (upholding a complaint that made “[a] plausible claim that [policymakers] knew about the reasons for [subordinates’] recommendations and . . . then ratified them”).
  88. . Montero v. City of Yonkers, 890 F.3d 386, 403 (2d Cir. 2018) (quoting Amnesty Am. v. Town of W. Hartford, 361 F.3d 113, 126 (2d Cir. 2004)). See generally Jones v. Town of E. Haven, 691 F.3d 72 (2d Cir. 2012).
  89. . Montero, 890 F.3d at 403.
  90. . See, e.g., Lewis v. Union City, 877 F.3d 1000, 1020 (11th Cir. 2017) (denying municipal liability because “[t]here is no requirement that the administrative review be ideal, simply that it be a meaningful layer of review of an official’s decision”); Soltesz v. Rushmore Plaza Civic Ctr., 847 F.3d 941, 946–47 (8th Cir. 2017). In Lewis, for example, the plaintiff had “not offered any facts to suggest that [the policymaker] was a mere rubber stamp or that he approved any improper motive.” 877 F.3d at 1021.
  91. . See, e.g., Saunders v. Town of Hull, 874 F.3d 324, 331 (1st Cir. 2017) (citing Welch v. Ciampa, 542 F.3d 927, 940 (1st Cir. 2008)) (acknowledging that “knowledge and authorization” are required); Culbertson, 790 F.3d at 622.
  92. . See Waters v. City of Chi., 580 F.3d 575, 584–85 (7th Cir. 2009). While the court uses the term “delegation,” Waters was a case in which the plaintiff’s dismissal was reviewed by a policymaker. That officer’s approval seems better described as “ratification.”
  93. . Id. at 585 (quoting City of St. Louis v. Praprotnik, 485 U.S 112 (1988)).
  94. . Id.
  95. . City of Canton v. Harris, 489 U.S. 378, 379 (1989).
  96. . Id. at 388.
  97. . Id. at 391.
  98. . Id. at 381.
  99. . Id.
  100. . See, e.g., City of Revere v. Mass. Gen. Hosp., 463 U.S. 239 (1983); Harris, 489 U.S. at 397.
  101. . Harris, 489 U.S. at 385–86.
  102. . Id. at 388.
  103. . Id.
  104. . In particular, the Court “reject[ed Canton’s] contention that only unconstitutional policies are actionable under the statute.” Id. at 387.
  105. . Id. at 388. Some lower courts describe this as a “knew or should have known” test. See, e.g., Gray v. Cummings, 917 F.3d 1, 14 (1st Cir. 2019) (quoting Haley v. City of Bos., 657 F.3d 39, 52 (1st Cir. 2011)). That language suggests negligence and is probably out-of-step with the Supreme Court’s current approach, though it is consistent with the thesis of this Article.
  106. . Harris, 489 U.S. at 389.
  107. . Id. at 391.
  108. . Id.
  109. . Connick v. Thompson, 563 U.S. 51, 62 (2011) (quoting Harris, 489 U.S. at 392).
  110. . Connick, 563 U.S. at 61–62 (internal quotation marks omitted) (citing Harris, 489 U.S. at 395 (O’Connor, J., concurring in part and dissenting in part)).
  111. . A corollary of this principle is that no matter how deliberately indifferent the training may be, plaintiff must also establish a violation of his constitutional rights, see City of L.A. v. Heller, 475 U.S. 796 (1986) (per curiam), and a causal link between the deliberate indifference and the constitutional violation, see Harris, 489 U.S. at 378.
  112. . Connick, 563 U.S. at 51.
  113. . Brady v. Maryland, 373 U.S. 83, 91 (1963).
  114. . Connick, 563 U.S. at 57.
  115. . Id.
  116. . See id. at 62.
  117. . Id. at 57.
  118. . See id. at 62–63.
  119. . Id. at 71.
  120. . See id. at 72 (“The District Court should have granted Connick judgment as a matter of law on the failure-to-train claim because Thompson did not prove a pattern of similar violations . . . .”). Some post-Connick cases hold that proof of a pattern is not always necessary. See, e.g., Thomas v. Cumberland Cty., 749 F.3d 217, 223 (3d Cir. 2014) (citing City of Canton v. Harris, 489 U.S. 378, 390 n.10 (1989)); cf. Jackson v. City of Cleveland, 925 F.3d 793, 834 (6th Cir. 2019) (explaining that the evidence as to policy was ambiguous and conflicting, thus “whether Cleveland had a policy of permitting Brady violations” was a jury question).
  121. . Bd. of Cty. Comm’rs v. Brown, 520 U.S. 397 (1997).
  122. . Id. at 404.
  123. . Id. at 401.
  124. . Id.
  125. . Id.
  126. . Id.
  127. . Id. at 412.
  128. . Id. at 410. Plaintiffs rarely win hiring claims. See Young v. City of Providence ex rel. Napolitano, 404 F.3d 4, 30 (1st Cir. 2005) (“It is much harder . . . to succeed on a hiring claim than a failure to train claim.”).
  129. . See, e.g., Rodriguez v. Cty. of L.A., 891 F.3d 776, 803 (9th Cir. 2018) (quoting Hunter v. Cty. of Sacramento, 652 F.3d 1225, 1234 n.8 (9th Cir. 2011)); Velazquez v. City of Long Beach, 793 F.3d 1010, 1027 (9th Cir. 2015); Cash v. Cty. of Erie, 654 F.3d 324, 333–39 (2d Cir. 2011).
  130. . Monell v. Dep’t of Soc. Servs., 436 U.S. 658, 694 (1978).
  131. . See, e.g., Robinson v. Hunt Cty., 921 F.3d 440, 449 (5th Cir. 2019) (observing a custom that constitutes an “official policy” of “viewpoint discrimination” as evidenced by the Facebook page of the Hunt County Sheriff’s Office).
  132. . See, e.g., Walker v. City of Calhoun, 901 F.3d 1245, 1256 (11th Cir. 2018) (affirming district court’s finding that “the City could directly regulate bail if it wished to and so may be held responsible for acquiescing in an unconstitutional policy and practice by its Municipal Court and police”).
  133. . See, e.g., Estate of Roman v. City of Newark, 914 F.3d 789, 799 (3d Cir. 2019); Doe v. Vigo Cty., 905 F.3d 1038, 1045 (7th Cir. 2018).
  134. . See Bd. of Cty. Comm’rs v. Brown, 520 U.S. 397, 404 (1997) (stating that a government can be liable for “an act performed pursuant to a ‘custom’ that has not been formally approved by an appropriate decisionmaker” if “the practice is so widespread as to have the force of law”).
  135. . See, e.g., Owens v. Balt. City State’s Attorney’s Office, 767 F.3d 379, 402–03 (4th Cir. 2014) (discussing the elements of a custom theory); Curtis v. Anthony, 710 F.3d 587, 595 (5th Cir. 2013) (“[A] plaintiff must [show] that the body governing a municipality, or an official to whom the body has delegated its policy-making authority, had actual or constructive knowledge of the custom of policy at issue.” (citing Burge v. St. Tammary Par., 336 F.3d 363, 370 (5th Cir. 2003))).
  136. . 757 F.3d 31 (2d Cir. 2014).
  137. . Id. at 36–37, 62–63.
  138. . Id. at 63.
  139. . Id.
  140. . World Wide St. Preachers v. Town of Columbia, 591 F.3d 747, 755 (5th Cir. 2009). Compare Jackson v. Barnes, 749 F.3d 755, 763 (9th Cir. 2014) (noting that repeated practices of one police officer, with no rebuke by superiors, would amount to “a policy of inaction” by the county), with D’Ambrosio v. Marino, 747 F.3d 378, 388 (6th Cir. 2014) (holding that “allegations regarding the repeated failures of only one prosecutor” are insufficient to state a custom case).
  141. . Palmer v. Marion Cty., 327 F.3d 588, 595–96 (7th Cir. 2003) (citing Jackson v. Marion Cty., 66 F.3d 151, 152 (7th Cir. 1995)). Some opinions suggest, if in an off-handed way, that policymakers must have knowledge of the unconstitutional practice for the plaintiff’s custom claim to succeed. See, e.g., Regains v. City of Chi., 918 F.3d 529, 536 (7th Cir. 2019).
  142. . Curtis v. Anthony, 710 F.3d 587, 595 (5th Cir. 2013) (citing Burge v. St. Tammary Par., 336 F.3d 363, 370 (5th Cir. 2003)).
  143. . See, e.g., Mann v. Cty. of San Diego, 907 F.3d 1154, 1164 (9th Cir. 2018) (drawing the distinction between direct and indirect municipality liability cases).
  144. . See, e.g., Connick v. Thompson, 563 U.S. 51, 61 (2011).
  145. . See Monell v. Dep’t of Soc. Servs., 436 U.S. 658, 691 (1978).
  146. . See, e.g., Connick, 563 U.S. at 60; Bd. of Cty. Comm’rs v. Brown, 520 U.S. 397, 403 (1997); City of Canton v. Harris, 489 U.S. 378, 385 (1989) (quoting Monell, 426 U.S. at 694); City of St. Louis v. Praprotnik, 485 U.S. 112, 121–22 (1988) (plurality opinion) (quoting Springfield v. Kibbe, 480 U.S. 257, 267 (1987) (O’Connor, J., dissenting)).
  147. . Connick, 563 U.S. at 62 (emphasis omitted) (quoting Harris, 489 U.S. at 392).
  148. . See, e.g., Mann, 907 F.3d at 1164; Doe v. Vigo Cty., 905 F.3d 1038, 1045 (7th Cir. 2018); Alvarez v. City of Brownsville, 904 F.3d 382, 391 (5th Cir. 2018).
  149. . See Brown, 520 U.S. at 404–05.
  150. . Id. at 397.
  151. . See id.; Connick, 563 U.S. at 51.
  152. . Fallon, supra note 40, at 480.
  153. . Harlow v. Fitzgerald, 457 U.S. 800, 814 (1982); see also Ziglar v. Abbasi, 137 S. Ct. 1843, 1866 (2017); Anderson v. Creighton, 483 U.S. 635, 638 (1987) (quoting Davis v. Scherer, 468 U.S. 183, 195 (1984)).
  154. . See Harlow, 457 U.S. at 808.
  155. . Ziglar, 137 S. Ct. at 1866.
  156. . Harlow, 457 U.S. at 813.
  157. . Hope v. Pelzer, 536 U.S. 730, 739–40 (2002); Scheuer v. Rhodes, 416 U.S. 232, 240–43 (1974).
  158. . Harlow, 457 U.S. at 814.
  159. . Id. at 818.
  160. . See Fallon, supra note 19, at 956 (noting “the Court’s commitment to a robustly protective doctrine of qualified immunity”).
  161. . Malley v. Briggs, 475 U.S. 335, 341 (1986).
  162. . See City of L.A. v. Lyons, 461 U.S. 95, 105 (1983). In Lyons, for example, the plaintiff sued for damages for an allegedly illegal police chokehold. But he also sought an injunction that would forbid chokeholds in the future. The Court ruled that his suit for prospective injunctive relief could not proceed, because he had not shown a likelihood that he would be subjected to a chokehold in the future.
  163. . Richardson v. McKnight, 521 U.S. 399, 402 (1997) (citing McKnight v. Rees, 88 F.3d 417, 425 (6th Cir. 1996)).
  164. . Id. at 404–05.
  165. . For discussion of these doctrines, see Civil Rights Actions, supra note 20, at 49–89.
  166. . Mt. Healthy City Sch. Dist. Bd. of Educ. v. Doyle, 429 U.S. 274, 285–86 (1976) (“The constitutional principle at stake is sufficiently vindicated if such an employee is placed in no worse a position than if he had not engaged in the conduct.”).
  167. . Memphis Cmty. Sch. Dist. v. Stachura, 477 U.S. 299, 310 (1986).
  168. . Carey v. Piphus, 435 U.S. 247, 256–57 (1978).
  169. . See Smith v. Wade, 461 U.S. 30, 49 (1983).
  170. . City of Newport v. Fact Concerts, Inc., 453 U.S. 247, 267 (1981).
  171. . Carey, 435 U.S. at 258. Thus, some courts have allowed recovery of presumed damages for certain constitutional violations, when the traditional damages rules do not seem adequate to vindicate the constitutional right at stake or to provide adequate deterrence. See, e.g., Siebert v. Severino, 256 F.3d 648, 655 (7th Cir. 2001) (quoting Erwin v. Cty. of Manitowoc, 872 F.2d 1292, 1299 (7th Cir. 1989)); Walje v. City of Winchester, 827 F.2d 10, 13 (6th Cir. 1987).
  172. . Robertson v. Wegmann, 436 U.S. 584, 594 (1986); see, e.g., Wheeler v. City of Santa Clara, 894 F.3d 1046, 1053 (9th Cir. 2018) (citing Carlson v. Green, 446 U.S. 14, 24 (1980)).
  173. . 42 U.S.C. § 1988(b) (2018).
  174. . See Farrar v. Hobby, 506 U.S. 103, 115–16 (1992) (quoting City of Riverside v. Rivera, 477 U.S. 561, 580 (1986) (plurality opinion)).
  175. . Id. at 121–22 (1992) (O’Connor, J., concurring). Justice O’Connor’s views are especially important on these points, as the Court was divided 5–4, and she provided the swing vote for the majority.
  176. . See Rivera, 477 U.S. at 574 (plurality opinion).
  177. . See, e.g., Fox v. Vice, 563 U.S. 826, 828 (2011).
  178. . Connick v. Thompson, 563 U.S. 51, 71 (2011).
  179. . This notion of accommodation of competing interests is reflected in the Restatement’s core definition of negligence. See Restatement (Third) of Torts: Liability for Physical and Emotional Harm § 3 (Am. Law Inst. 2010) (“Primary factors to consider in ascertaining whether the person’s conduct lacks reasonable care are the foreseeable likelihood that the person’s conduct will result in harm, the foreseeable severity of any harm that may ensue, and the burden of precautions to eliminate or reduce the risk of harm.”). It is also manifest in the work of many torts scholars who have studied negligence. See, e.g., Kenneth W. Simons, Negligence, 16 Soc. Phil. & Pol’y 52, 90 (1999) (“Such an instruction invites juries to identify and articulate the moral norms of the community.”); Benjamin C. Zipursky, Reasonableness in and out of Negligence Law, 163 U. Pa. L. Rev. 2131, 2161 (2015) (“[T]he phrases ‘reasonable person’ and ‘reasonably prudent person’ and the concept of a reasonable person connect with a conception of reasonableness as the capacity to constrain one’s behavior and choices in a manner that is not wholly insensitive to others’ need and wishes, but rather displays some sense of mutuality.”); Gregory C. Keating, Reasonableness and Rationality in Negligence Theory, 48 Stan. L. Rev. 311, 312 (1996) (“When we act reasonably, we restrain our pursuit of self-interest by acting in accordance with principles that fix fair terms of cooperation.”).
  180. . Cf. Gertz v. Robert Welch, Inc., 418 U.S. 323, 350 (1974) (adopting negligence as the norm for accommodation of competing interests in reputation and freedom of speech in defamation law).
  181. . Simons, supra note 178, at 90.
  182. . Adoption of “reasonable care” would not wholly banish the concept of subjective deliberate indifference from municipal liability cases. In its due process jurisprudence, the Court has chosen to define the substantive constitutional obligation, owed by both governments and officers, as a duty not to harm persons by deliberate indifference. See, e.g., Lapre v. City of Chi., 911 F.3d 424, 438 (7th Cir. 2018);see also Farmer v. Brennan, 511 U.S. 825, 841 (1994). That use of the term is wholly distinct from its use as in policy or custom doctrine, in which the issue is not the scope of constitutional rights but the scope of governmental liability for constitutional violations, no matter how the substantive violations are defined.
  183. . See, e.g., Sheldon Nahmod, Section 1983 Discourse: The Move from Constitution to Tort, 77 Geo. L.J. 1719, 1720 (arguing that “the Court, by using tort rhetoric, is attempting to marginalize § 1983 and make it less protective of fourteenth amendment rights”).
  184. . Fallon, supra note 19, at 997.
  185. . See, e.g., Filarsky v. Delia, 566 U.S. 377, 384 (2012) (quoting Imbler v. Pachtman, 424 U.S. 409, 418 (1975)) (discussing the role of common law principles in constitutional tort); see also Tower v. Glover, 467 U.S. 914, 920 (1984). On post-1871 developments, see Rehberg v. Paulk, 566 U.S. 356, 361–62 (2012); Kalina v. Fletcher, 522 U.S. 118, 123–24 (1997) (citing Imbler, 424 U.S. at 410); Owen v. City of Independence, 445 U.S. 622, 637–38 (1980).
  186. . Monroe v. Pape, 365 U.S. 167, 187 (1961); see also Larry Kramer & Alan O. Sykes, Municipal Liability Under § 1983: A Legal and Economic Analysis, 1987 Sup. Ct. Rev. 249, 256–57 (discussing the legislative history).
  187. . Harlow v. Fitzgerald, 457 U.S. 800, 818–19 (1982).
  188. . See Carey v. Piphus, 435 U.S. 247, 257 (1978).
  189. . See Cty. of L.A. v. Mendez, 137 S. Ct. 1539, 1548–49 (2017) (citing Paroline v. United States, 572 U.S. 434, 444–45 (2014)).
  190. . See Town of Newton v. Rumery, 480 U.S. 386, 392 (1987).
  191. . See Manuel v. City of Joliet, 137 S. Ct. 911, 920–21 (2017) (quoting Hartman v. Moore, 547 U.S. 250, 258 (2006)).
  192. . See id. at 921; Filarsky v. Delia, 566 U.S. 377, 384 (2012) (citing Imbler v. Pachtman, 424 U.S. 409, 418 (1976)); Rehberg v. Paulk, 566 U.S. 356, 366 (2012); Carey, 435 U.S. at 258.
  193. . Filarsky, 566 U.S. at 384 (citing Imbler, 424 U.S. at 418).
  194. . See Dan T. Coenen, A Constitution of Collaboration: Protecting Fundamental Values with Second-Look Rules of Interbranch Dialogue, 42 Wm. & Mary L. Rev. 1575, 1713 (2001).
  195. . See Monell v. Dep’t of Soc. Servs., 436 U.S. 658, 691 (1978).
  196. . As noted earlier, these later cases reject the objective approach to indirect liability, which the Court had earlier seemed to adopt, albeit in dicta, in Farmer v. Brennan, 511 U.S. 825, 837 (1994). But the shift has not been complete in the lower courts. See, e.g., Garza v. City of Donna, 922 F.3d 626, 637 (5th Cir. 2019) (relying on Farmer for the proposition that the test is “objective deliberate indifference”). The thrust of this Article is that the key distinction is between objective and subjective approaches and that the objective approach is preferable to the subjective one. The terminology is not especially important. Thus, Garza, in substance, follows the approach favored here, despite the Garza court’s use of the term deliberate indifference. From a tort perspective, however, it is worth noting that the “objective” approach to liability is ordinarily called “negligence.”
  197. . Bd. of Cty. Comm’rs of Bryan Cty. v. Brown, 520 U.S. 397, 415 (1997); see also Connick v. Thompson, 563 U.S. 51, 71 (2011) (“Thompson had to show that it was so predictable that failing to train the prosecutors amounted to conscious disregard for defendants’ Brady rights.”). For an illustrative circuit court case, see Schneider v. City of Grand Junction Police Department, 717 F.3d 760, 769 (10th Cir. 2013) (“A local government policymaker is deliberately indifferent when he deliberately or consciously fails to act when presented with an obvious risk of constitutional harm which will almost inevitably result in constitutional injury of the type experienced by the plaintiff.” (quoting Hollingsworth v. Hill, 110 F.3d 733, 745 (10th Cir. 1997))).
  198. . Brown, 520 U.S. at 407; see, e.g., Outlaw v. City of Hartford, 884 F.3d 351, 373 (2d Cir. 2018).
  199. . Restatement (Third) of Torts: Physical & Emotional Harm § 2 (Am. Law Inst. 2010).
  200. . See, e.g., Keith v. DeKalb Cty., 749 F.3d 1034, 1047 (11th Cir. 2014) (quoting Franklin v. Curry, 738 F.3d 1246, 1250 (11th Cir. 2013)) (explaining that deliberate indifference is “conduct that is more than gross negligence”); Mason v. Lafayette City-Par. Consol. Gov’t, 806 F.3d 268, 279 (5th Cir. 2015); Amnesty Am. v. Town of W. Hartford, 361 F.3d 113, 128 (2d Cir. 2004) (noting that liability must be based on a “‘conscious choice’ rather than mere negligence” (quoting City of Canton v. Harris, 489 U.S. 378, 389 (1989))).
  201. . Ziglar v. Abbasi, 137 S. Ct. 1843, 1866 (2017).
  202. . Restatement (Third) of Torts: Physical & Emotional Harm § 3 (Am. Law Inst. 2010).
  203. . See id. § 3 cmt. k.
  204. . See id. § 3 cmt. g; see also Oliver Wendell Holmes, The Common Law 76 (Mark DeWolfe Howe ed., Harvard Univ. Press 1963) (1881) (“A choice which entails a concealed consequence is as to that consequence no choice.”). For an application of this general principle as a rationale for official immunity for constitutional torts, see Barbara E. Armacost, Qualified Immunity: Ignorance Excused, 51 Vand. L. Rev. 583, 588–89 (1998).
  205. . See Restatement (Third) of Torts: Physical & Emotional Harm § 3 cmt. e (Am. Law Inst. 2010); see also Raimondo v. Harding, 341 N.Y.S.2d 679, 681 (N.Y. App. Div. 1973) (quoting Townes v. Park Motor Sales, 180 N.Y.S.2d 553, 559 (N.Y. App. Div. 1958)).
  206. . 132 Eng. Rep. 490, 492 (C.P. 1837); see also Holmes, supra note 203, at 86–87 (defending the objective test on two grounds: (a) “the impossibility of nicely measuring a man’s powers and limitations,” and (b) “when men live in society, a certain average of conduct . . . is necessary to the general welfare”). In the constitutional tort context, these rationales simply reinforce the vindication and deterrence goals of § 1983.
  207. . Manuel v. City of Joliet, 137 S. Ct. 911, 921 (2017) (quoting Hartman v. Moore, 547 U.S. 250, 258 (2006)); see also Nieves v. Bartlett, 139 S. Ct. 1715, 1726 (2019) (first quoting Kalina v. Fletcher, 522 U.S. 118, 123 (1997); then quoting Manuel, 137 S. Ct. at 921).
  208. . Manuel, 137 S. Ct. at 921.
  209. . City of St. Louis v. Praprotnik, 485 U.S. 112, 125 (1988) (plurality opinion); see also Jett v. Dallas Indep. Sch. Dist., 491 U.S. 701, 737 (1989) (citing Praprotnik, 485 U.S. at 124 n.1) (“The trial judge must identify those officials or governmental bodies who speak with final policymaking authority for the local governmental actor concerning the action alleged to have caused the particular constitutional or statutory violation at issue.”).
  210. . Jett, 491 U.S. at 737.
  211. . See Bd. of Cty. Comm’rs v. Brown, 520 U.S. 397, 435 (1997) (Breyer, J., dissenting) (noting that “results have sometimes proved inconsistent” and that “the location of ‘policymaking’ authority” is a “conceptually difficult problem”).
  212. . Soltesz v. Rushmore Plaza Civic Ctr., 847 F.3d 941, 946 (8th Cir. 2017) (“In addition to creating municipal liability for their own actions, final policymakers can also create this liability by . . . delegating policymaking authority to a subordinate.”); see also Webb v. Town of St. Joseph, 925 F.3d 209, 215 (5th Cir. 2019) (“A municipal employee may . . . possess final policymaking authority where the final policymaker has delegated that authority, either expressly or impliedly.”); Liverman v. City of Petersburg, 844 F.3d 400, 413 (4th Cir. 2016) (“[T]he fact that Dixon ‘serves under the direction and control of the city manager’ does not necessarily establish that he lacked final authority to promulgate the policy.”); Thompson v. District of Columbia, 832 F.3d 339, 349 (D.C. Cir. 2016) (noting that a lack of meaningful oversight justifies finding a subordinate to be a policymaker).
  213. . See, e.g., Patel v. Hall, 849 F.3d 970, 979 (10th Cir. 2017). To determine whether an actor is a policymaker, courts consider “whether the employee is meaningfully constrained by policies not of his own making” and “whether his decisions are final, i.e., not subject to any meaningful review.” Id. (quoting Randle v. City of Aurora, 69 F.3 411, 448 (10th Cir. 1995)); see also Kristofek v. Vill. of Orland Hills, 832 F.3d 785, 799 (7th Cir. 2016) (inquiring “whether the official is constrained by policies of other officials or legislative bodies [and] whether the official’s decision on the issue in question is subject to meaningful review”).
  214. . Webb v. Sloan, 330 F.3d 1158, 1162 (9th Cir. 2003).
  215. . Id.; see Town of Newton v. Rumery, 480 U.S. 386 (1987) (upholding the validity of release-dismissal agreements).
  216. . Webb, 330 F.3d at 1162–63. The prosecutor in such a case is absolutely immune from § 1983 liability. See Imbler v. Pachtman, 424 U.S. 409, 409 (1976).
  217. . Webb, 330 F.3d at 1161.
  218. . Id. at 1164 (quoting Nev. Rev. Stat. § 252.070 (1) (2001)).
  219. . Id.
  220. . Id. at 1165. The court then noted that “[i]n fact, Carson City presented evidence to the contrary,” perhaps on the flawed theory that the city could win by showing that the deputy was a renegade. Id.
  221. . Pinch Hitter,, (defining “pinch hitter” as any substitute for another, especially in an emergency).
  222. . Webb, 330 F.3d at 1165.
  223. . Id. (citing Christie v. Iopa, 176 F.3d 1231 (9th Cir. 1999)).
  224. . Id. at 1165–66 (citing Christie, 176 F.3d at 1238).
  225. . See Restatement (Third) of Torts: Physical & Emotional Harm § 3 (Am. Law Inst. 2010) (“[F]actors to consider in ascertaining whether the person’s conduct lacks reasonable care are the foreseeable likelihood that the person’s conduct will result in harm, the foreseeable severity of any harm that may ensue, and the burden of precautions to eliminate or reduce the risk of harm.”).
  226. . Many opinions are not precise in their use of terms. Such courts may be applying objective deliberate indifference, which is, in effect, the substantive norm proposed in this article under the heading of “negligence.” See supra note 31 and accompanying text.
  227. . 347 F.3d 641, 645 (7th Cir. 2003). For other illustrations of cost-benefit balancing, see, for example, Littell v. Houston Independent School District, 894 F.3d 616, 624–25 (5th Cir. 2018), where the court held that a jury could reasonably find deliberate indifference in an intrusive school search case, in which the risk of constitutional violations was a recurring one and the relevant officers received no training. “A jury could conclude that the severity of the consequences of a friendly fire shooting forced the department to take notice of the high risk despite the rarity of such an incident.” Young v. City of Providence ex rel. Napolitano, 404 F.3d 4, 29 (1st Cir. 2005).
  228. . Dunn, 347 F.3d at 646 (first citing Bd. of Cty. Comm’rs v. Brown, 520 U.S. 397, 409 (1997); then citing Robles v. City of Fort Wayne, 113 F.3d 732, 735 (7th Cir. 1997); and then citing Palmquist v. Selvik, 111 F.3d 1332, 1346 (7th Cir. 1997)).
  229. . Id. at 645–46.
  230. . See Restatement (Third) of Torts: Physical & Emotion Harm § 13 (Am. Law Inst. 2010) (describing the role of custom in adjudicating the negligence issue).
  231. . See, e.g., Lama v. Borras, 16 F.3d 473, 478 (1st Cir. 1994).
  232. . See Restatement (Third) of Torts: Physical & Emotional Harm §§ 14–16 (discussing the role of statutory violations in adjudicating negligence).
  233. . Id. § 17 (discussing res ipsa loquitur).
  234. . Id. § 9. (stating that for both plaintiffs and defendants, negligence law takes into account “an unexpected emergency requiring rapid response”).
  235. . Raimondo v. Harding, 41 A.D.2d 63, 63–64 (N.Y. App. Div. 1973).
  236. . Id. (quoting 1 NY PJI 2:40).
  237. . See, e.g., Walker v. City of Calhoun, 901 F.3d 1245, 1256 (11th Cir. 2018).
  238. . See, e.g., Jones v. Town of E. Haven, 691 F.3d 72 (2d Cir. 2012) (citing Amnesty Am. v. Town of W. Hartford, 361 F.3d 113, 126 (2d Cir. 2004)).
  239. . Rowland v. Christian, 443 P.2d 561, 567 (Cal. 1968).
  240. . Id.
  241. . Id. at 568.
  242. . Id.
  243. . Davison v. Randall, 912 F.3d 666, 673 (4th Cir. 2019).
  244. . Id. at 675.
  245. . Id. at 690 (quoting Davison v. Loudoun Cty. Bd. of Supervisors, 267 F. Supp. 3d 702, 715 (E.D. Va. 2017)).
  246. . See, e.g., Larson v. St. Francis Hotel, 188 P.2d 513, 515 (Cal. Ct. App. 1948) (holding that a hotel was not liable for a guest throwing a chair out the window during a sudden celebration of the end of World War II).
  247. . Penley v. McDowell Cty. Bd. of Educ., 876 F.3d 646, 653 (4th Cir. 2017) (quoting Love-Lane v. Martin, 355 F.3d 766, 782 (4th Cir. 2004)).
  248. . Gschwind v. Heiden, 692 F.3d 844, 845–46 (7th Cir. 2012).
  249. . Id. at 847 (citing 105 Ill. Comp. Stat. 5/10-20, 20.7 (1990)).
  250. . Id. at 848 (internal quotation marks omitted).
  251. . Id.
  252. . See, e.g., Amnesty Am. v. Town of W. Hartford, 361 F.3d 113, 126 (2d Cir. 2004) (“Another method of implicating a policymaking official through subordinates’ conduct is to show that the policymaker was aware of a subordinate’s unconstitutional actions, and consciously chose to ignore them.”). When the policymaker does not just oversee subordinates, but makes the decision himself, as in the hypothetical case provided by Justice O’Connor in Praprotnik, the case is one of direct-effect liability, as she recognized. See City of St. Louis v. Praprotnik, 485 U.S. 112, 127 (1988) (plurality opinion). For example, in Pembaur, municipal liability was based on the single act of the district attorney, who told the police to enter the doctor’s office. There is no legally significant difference between that event and a variation in which the police officers recommend entry and the policymaker ratifies that decision. If the ultimate decision is for the policymaker, the subordinate does not ever make a decision, only a recommendation.
  253. . Amnesty Am., 361 F.3d at 126.
  254. . See, e.g., Winkler v. Madison Cty., 893 F.3d 877, 902 (6th Cir. 2018) (distinguishing this case in which plaintiff discussed only one instance “and therefore cannot establish that the County had a custom of deliberate indifference to the serious healthcare needs of all of the inmates”).
  255. . See Monell v. Dep’t of Soc. Servs., 436 U.S. 658, 691 (1978) (quoting Adickes v. S.H. Kress & Co., 398 U.S. 144, 167–68 (1970)). For applications of this principle, see, for example, Regains v. City of Chi., 918 F.3d 529, 536 (7th Cir. 2019); Malone v. Hinman, 847 F.3d 949, 955 (8th Cir. 2017); Velazquez v. City of Long Beach, 793 F.3d 1010, 1027 (9th Cir. 2015) (quoting Hunter v. Cty. of Sacramento, 652 F.3d 1225, 1232–33 (9th Cir. 2011)).
  256. . 402 F.3d 225 (1st Cir. 2005).
  257. . See id. at 237–38.
  258. . See Bd. of Cty. Comm’rs v. Brown, 520 U.S. 397, 414–16 (1997).
  259. . 563 U.S. 51 (2011).
  260. . Id. at 70.
  261. . Brown, 520 U.S. at 407.
  262. . Id. at 412.
  263. . See id. at 415.
  264. . See id. at 419–20 (Souter, J., dissenting).
  265. . Connick, 563 U.S. at 62–63.
  266. . Id. at 59.
  267. . See id. at 62–63 (“None of those cases involved failure to disclose blood evidence, a crime report, or physical or scientific evidence of any kind.”).
  268. . Id. at 62 (citing Brown, 520 U.S. at 409); see, e.g., Westfall v. Luna, 903 F.3d 534, 552 (5th Cir. 2018) (quoting Burge v. St. Tammany Par., 336 F.3d 363, 370 (5th Cir. 2001)); Brown v. Battle Creek Police Dep’t, 844 F.3d 556, 575 (6th Cir. 2016).
  269. . Cf. Thomas v. Cumberland Cty., 749 F.3d 217, 223 (3d Cir. 2014) (quoting City of Canton v. Harris, 489 US. 378, 390 n.10 (1989)) (explaining that even under the deliberate indifference test, foreseeability can sometimes be found without a pattern).
  270. . Hill v. Cundiff, 797 F.3d 948, 957–58 (11th Cir. 2015).
  271. . Id. at 971–72.
  272. . Id. at 963.
  273. . See id. at 977–78.
  274. . Id.
  275. . See id. at 985.
  276. . Christina B. Whitman, Government Responsibility for Constitutional Torts, 85 Mich. L. Rev. 225, 226 (1986). Castro v. County of Los Angeles, 833 F.3d 1060 (9th Cir. 2016), illustrates some of the problems identified by Professor Whitman. In Castro, a pre-trial detainee was beaten while housed in a “sobering cell.” Id. at 1065. The government’s liability was based on the following:[T]he poor design and location of the sobering cell, . . . a custom of housing intoxicated inmates in sobering cells that contained inadequate audio monitoring[,] . . . [and] a policy to check on inmates only every 30 minutes. . . . These routine practices were consciously designed and, together, they amount to a custom or policy.

    Id. at 1075; see also Newton v. City of New York, 779 F.3d 140, 153–56 (2d Cir. 2015) (holding that deficiencies in the City’s evidence management system were responsible for plaintiff’s illegal incarceration, thus justifying municipal liability). The court upheld a finding of “recklessness or deliberate indifference” to plaintiff’s rights. Id. at 156. For present purposes, the distinctive feature of these cases is that, like Hill, they are not direct municipal liability cases, yet neither do they fall squarely into any of the categories traditionally associated with indirect liability.

  277. . Bd. of Cty. Comm’rs v. Brown, 520 U.S. 397, 415 (1997); see also id. at 410 (“To prevent municipality 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.”).
  278. . Connick v. Thompson, 563 U.S. 51, 62 (2011) (quoting City of Canton v. Harris, 489 US. 378, 392 (1989)).
  279. . The Justices may be influenced by what they learned in law school decades ago. For a time in the late twentieth century, it appeared that strict liability may replace negligence as the dominant test for ordinary tort liability. See, e.g., George L. Priest, The Invention of Enterprise Liability: A Critical History of the Intellectual Foundations of Modern Tort Law, 14 J. Legal Stud. 461, 516 (1985). However, the negligence standard appears to have successfully withstood the assault. See, e.g., G. Edward White, Tort Law in America: An Intellectual History 244–90 (expanded ed. 2003) (discussing “the unexpected persistence of negligence, 1980–2000”); James A. Henderson, Why Negligence Dominates Tort, 50 UCLA L. Rev. 377 passim (2002).
  280. . See Monell v. Dep’t of Soc. Servs., 436 U.S. 658, 694 (1978).
  281. . E.g., Schechter v. Merchants Home Delivery, Inc., 892 A.2d 415, 431 (D.C. 2006) (“Potential recovery in tort for negligent hiring or retention of an employee is not based on respondeat superior, but rather on proof of negligence on the part of the employer himself.” (citing Fleming v. Bronfin, Inc., 80 A.2d 915, 917 (D.C. 1951)); Freeman v. Bell, 366 So. 2d 197, 199 (La. Ct. App. 1978).
  282. . Restatement (Second) of Agency § 228(1) (Am. Law Inst. 1958).
  283. . See, e.g., Coe v. Carroll & Carroll, Inc., 709 S.E.2d 324, 332 (Ga. Ct. App. 2011).
  284. . See, e.g., Schechter, 892 A.2d at 427, 432 (finding that workers who committed robbery were clearly outside the scope of their employment, but the employer could be held liable based on negligent hiring and supervision).
  285. . See, e.g., Petrovich v. Share Health Plan of Ill., Inc., 719 N.E.2d 756, 765 (Ill. 1999).
  286. . 570 U.S. 421 (2013).
  287. . Id. at 431. Thus, the Court in Vance distinguished between sexual harassment by coworkers and by supervisors. See id. at 432.
  288. . Id. at 439.
  289. . The Court was split 5–4. Id. at 422. Justice Alito wrote for the majority. Id. Justice Ginsburg dissented, in an opinion joined by Justices Breyer, Sotomayor, and Kagan. Id.
  290. . Id. at 446.
  291. . Id.
  292. . Memphis Cmty. Sch. Dist. v. Stachura, 477 U.S. 299, 308 (1986).
  293. . See generally City of Newport v. Fact Concerts, Inc., 453 U.S. 247 (1981).
  294. . See cases cited supra notes 181–89 and accompanying text.
  295. . 692 F.3d 844 (7th Cir. 2012).
  296. . See cases cited supra notes 211–14.
  297. . See generally Bd. of Cty. Comm’rs v. Brown, 520 U.S. 397 (1997).
  298. . See generally Connick v. Thompson, 563 U.S. 51 (2011).
  299. . See generally id.
  300. . Id. at 54–57.
  301. . Id. at 71–72.
  302. . See id. at 64–65.
  303. . See id. at 71.
  304. . Id. at 66 (citing Bd. of Cty. Comm’rs v. Brown, 520 U.S. 397, 409 (1997)).
  305. . See id. at 84 n.5 (noting but not addressing the causation issue).
  306. . Restatement (Third) of Torts: Physical & Emotional Harm § 26 (Am. Law Inst. 2010); see, e.g., Shumpert v. City of Tupelo, 905 F.3d 310, 318 (5th Cir. 2018) (“[Beneficiaries] have failed to present evidence that additional training would have prevented [the decedent’s] injuries.”).
  307. . See also Alvarez v. City of Brownsville, 904 F.3d 382, 390–91 (5th Cir. 2018) (stating that failure to provide plaintiff with evidence was due to a “series of interconnected errors within the Brownsville Police Department that involved individual officers” and this series of errors was “separate from the general policy of non-disclosure of information. . . . The general policy of non-disclosure was not a direct cause of Alvarez’s injury”).
  308. . Restatement (Third) of Torts: Physical & Emotional Harm § 27 cmt. a (Am. Law Inst. 2010).
  309. . Id. § 36.
  310. . Wex S. Malone, Ruminations on Cause in Fact, 9 Stan. L. Rev. 60, 97 (1956).
  311. . Cf. Nieves v. Bartlett, 139 S. Ct. 1715, 1724 (2019) (requiring the plaintiff in a retaliatory arrest case to prove not only that unconstitutional animus motivated the arrest but also that the officer lacked probable cause to make the arrest because in such cases “it is particularly difficult to determine whether the adverse government action was caused by the officer’s malice or the plaintiff’s potentially criminal conduct”).
  312. . Bd. of Cty. Comm’rs v. Brown, 520 U.S. 397, 401 (1997).
  313. . See id. at 400–01.
  314. . Id. at 401.
  315. . Id. at 416.
  316. . Id. at 415.
  317. . Restatement (Third) of Torts: Physical & Emotional Harm § 29 (Am. Law Inst. 2010). For illustrations of this rule to § 1983 municipal liability, see, for example, Eisenhour v. Weber County, 897 F.3d 1272, 1280 (10th Cir. 2018), and Black Earth Meat Market, L.L.C. v. Village of Black Earth, 834 F.3d 841, 849–50 (7th Cir. 2016).
  318. . Brown, 520 U.S. at 401.
  319. . Id. at 413.
  320. . Id. at 412.
  321. . Cf. Wilson v. Cook Cty., 742 F.3d 775 (7th Cir. 2014). In this case, a government employee named Vanaria had, in an earlier government job as a probation officer, coerced female probationers to have sex with him. Id. at 777. He was fired from that job, but seven years later he was rehired for another government job. Id. at 778. He then “used the promise of a phony job to convince” another woman to have sex with him. Id. at 777. According to the court, “[g]iven the passage of time without incident and the fact that Vanaria had aged seven years, it is difficult to conclude that Vanaria’s misconduct with respect to [the second woman] was so obvious that any jury could find causation or deliberate indifference.” Id. at 782. However, even if the policymakers who rehired Vanaria were not deliberately indifferent to the plaintiff’s rights, a jury could surely find that the later misconduct was reasonably foreseeable in light of his earlier behavior, and plaintiff would satisfy both fault and proximate cause under a negligence-centered regime.
  322. . See Albert Levitt, Cause, Legal Cause and Proximate Cause, 21 Mich. L. Rev. 34, 54 (1922).
  323. . See Black Earth Meat Mkt., L.L.C. v. Vill. of Black Earth, 834 F.3d 841, 849 (7th Cir. 2016) (“There were four steps between the Village’s threat of litigation and all the deprivations except [one, as to which there were] . . . . three intermediate causal steps.”).
  324. . Restatement (Third) of Torts: Physical & Emotional Harm § 29 (Am. Law Inst. 2010); see also id. § 30 (“An actor is not liable for harm when the tortious aspect of the actor’s conduct was a type that does not generally increase the risk of that harm.”); id. § 34 (“When a force of nature of an independent act is also a factual cause of harm, an actor’s liability is limited to those harms that result from the risks that made the actor’s conduct tortious.”).
  325. . John C. Jeffries, Jr., Damages for Constitutional Violations: The Relation of Risk to Injury in Constitutional Torts, 75 Va. L. Rev. 1461, 1461 (1989).
  326. . Id. at 1481.
  327. . Jeffries acknowledges that “[a]n instrumental rationale [for liability for the economic loss] might well exist,” in that “[t]he prospect of damages no doubt would help deter unconstitutionality and would encourage government to take great pains to know (or anticipate) the law.” Id. at 1483.
  328. . Bd. of Cty. Comm’rs v. Brown, 520 U.S. 397, 430 (1997) (Breyer, J., dissenting).
  329. . Id.
  330. . Id. at 431.
  331. . Id.
  332. . Id.
  333. . See, e.g., Blum, supra note 35, at 963 (first citing interview with Rosalie B. Levinson, Professor, Valparaiso Univ. Law School; and then citing interview with David Rudovsky, Univ. of Pa. School of Law); see also Civil Rights Actions, supra note 20, at 217 (“[Strict liability under § 1983] is widely endorsed by academic commentators, many of whom favor strict liability either of the officer defendant or the government employer.”); cf. Amar, supra note 35, at 812–12 (focusing on police officers’ liability for Fourth Amendment violations).
  334. . Brown, 520 U.S. at 437 (Breyer, J., dissenting).
  335. . See Daryl J. Levinson, Making Government Pay: Markets, Politics, and the Allocation of Constitutional Costs, 67 U. Chi. L. Rev. 345, 412 (2000) (“Compensating constitutional torts cannot be expected to bring society closer to any just distributive pattern and will, in many cases, exacerbate the injustice of the existing distribution . . . .”).
  336. . See Jeffries, supra note 42, at 90.
  337. . See Kramer & Sykes, supra note 185, at 285 (noting that outcomes will likely be similar and suggesting that “by singling out negligent supervisory officials and identifying the measures that they should have taken, a negligence-based approach to vicarious liability might be more effective than strict vicarious liability at motivating cost-effective monitoring, training, and similar measures: negligence cases would generate a body of information about required precautionary measures for the guidance of other municipalities”).
  338. . See Civil Rights Actions, supra note 20, at 267–68.