Qualified Immunity After Ziglar v. Abbasi: The Case for a Categorical Approach

68 Am. U. L. Rev. 379 (2018).

* Professor, University of Georgia Law School. The Author wishes to thank Nathan Chapman, Dan Coenen, Michael Coenen, Tom Eaton, and Richard Fallon for helpful comments on a draft of this Article.

Qualified immunity protects officers from liability for damages unless they have violated clearly established rights, on the ground that it would be unfair and counterproductive to impose liability without notice of wrongdoing. In recent years, however, the Supreme Court has increasingly applied the doctrine to cases in which it serves little or no legitimate purpose. In Ziglar v. Abbasi, the rights were clearly established but the Court held that the officers were immune due to lack of clarity on other issues in the case. Because cases like Ziglar undermine the vindication of constitutional rights and the deterrence of violations, critics of immunity have called for its abolition. This Article rejects both of these approaches. This Article’s thesis is that the availability of qualified immunity should depend on an assessment of costs and benefits, which vary depending on context. A better approach is to retain the basic doctrine but to identify categories of cases in which immunity should be denied, and others in which it should be strengthened.

Introduction

After the September 11, 2001 attacks, federal officials arrested and detained over seven hundred undocumented aliens.¹ Many detainees were held for several months before being cleared of involvement in the attacks.² They sued the Attorney General and other federal officials over their treatment in detention. These complaints asserted violations of Fourth Amendment unreasonable seizure rights and Fifth Amendment substantive due process and equal protection claims. Because no federal statute authorizes lawsuits of this sort, the plaintiffs sought damages under the implied cause of action theory recognized in Bivens v. Six Unknown Named Agents of Federal Bureau of Narcoticsp³ for certain constitutional violations.⁴ In Ziglar v. Abbasi,⁵ a 4–to–2 majority declined to apply the Bivens doctrine to most of the plaintiffs’ claims.⁶ Ziglar may become a landmark ruling based on its highly restrictive approach to judge-made damages remedies for constitutional wrongs. Although the majority consisted of only four Justices,⁷ the opinion seems designed to bury the “implied cause of action” doctrine. Nor did the outcome in Ziglar come as a surprise. Indeed, the Court’s ruling was the ninth case in a row that narrowed the scope of permissible Bivens actions.⁸ One commentator has suggested that, “for the sake of judicial candor and litigative efficiency[,] [the Court] should hold that the Bivens cause of action is limited to the facts of Bivens” and of two other early post-Bivens cases.⁹

This Article focuses on a separate aspect of Ziglar: over the longer term, this case may attract as much interest for its holding on a separate issue—namely, the qualified immunity of executive officials. This defense blocks the recovery of damages even when the plaintiff can get past the “cause of action” hurdle by establishing an actionable violation of constitutional rights.¹⁰ Qualified immunity shields an official engaged in executive or administrative functions from having to pay damages for constitutional wrongdoing unless the plaintiff can show that the official has violated “clearly established statutory or constitutional rights of which a reasonable person would have known.”¹¹ The Court has left no doubt that qualified immunity is a powerful defense, which “protects ‘all but the plainly incompetent or those who knowingly violate the law.’”¹² It not only knocks out many Bivens claims, but, also, as a practical matter, its greatest impact is in the much larger number of suits, brought under 42 U.S.C. § 1983,¹³ against state and local officials.¹⁴

The qualified immunity issue arose in Ziglar because the detainees did not rely solely on Bivens. They also sued under 42 U.S.C. § 1985(3),¹⁵ which imposes statutory liability when two or more persons “conspire . . . for the purpose of depriving . . . any person or class of persons of the equal protection of the laws.”¹⁶ The Civil Rights Act of 1871 included § 1985(3), which is similar to the better known and far more widely used § 1983.¹⁷ Unlike § 1983, this statute does not operate only when the defendant has acted “under color of state law”; it can be used to sue both federal officials and private actors.¹⁸ The § 1985(3) count of the plaintiffs’ complaint asserted that federal officers “detained them in harsh conditions because of their actual or apparent race, religion, or national origin, in violation of the equal protection component of the Fifth Amendment.”¹⁹ Absent a compelling state interest, it is clear that the government may not discriminate on these grounds.²⁰ Thus, the complaint alleged violations of clearly established law. Even so, the Court held that the officials were protected by qualified immunity, reasoning that the § 1985(3) doctrine was unclear as to whether an actionable conspiracy can exist among officials of the same department of government and that the case involved high level executive decision making about national security.

Ziglar is an especially officer-protective qualified immunity decision because it seems to treat qualified immunity as a general no-liability rule whenever the law bearing on liability is unsettled and no matter why liability is uncertain. The holding thus implicitly rejects an alternative view, under which the official avoids liability only when the disputed element of the plaintiffs’ claims involves the defensibility of the official’s conduct. In other words, the Court in Ziglar treats qualified immunity as a general rule of “no liability when any element is uncertain,” even when the policies that gave rise to the immunity defense do not apply in the case at hand. The Court’s approach rests on a preference for laying down bright-line rules, as opposed to formulating a more refined doctrine rooted in the reasons that underlie the qualified immunity defense. This preference for rules is not inherently objectionable, because rules can serve worthy goals independent of the substantive interests they are designed to further. These goals include promoting stability in the law, predictability of outcomes, control of lower level decision makers, and efficiency in resolving disputes.²¹ The Court’s predilection for bright-line rules over substance-driven rules, so vividly illustrated by Ziglar, permeates the Court’s qualified immunity doctrine.²² This Article focuses on the institutional interests served by framing qualified immunity doctrine in the form of rules. It argues that “ruleness” values deserve to play a role in the qualified immunity context, but not nearly as much of a role as the Court now accords them. The universe of constitutional tort cases is not monolithic. One can make categorical distinctions between substantive contexts in which the policies favoring immunity are comparatively strong or weak.

This Article’s argument rests on three key points: First, recognizing the existence of any immunity defense is arguable. Second, there are nonetheless defensible, substantive justifications for retaining some form of immunity defense. Third, the best accommodation of the competing considerations involves marking off manageable categories of cases in which these substantive grounds for applying the immunity doctrine are especially weak or strong. As Richard Fallon has explained, “official immunity doctrines perform an equilibrating function by diminishing the social costs that constitutional rights would have if officers who violated them were always strictly liable in suits for damages.”²³ Thus, the case against the Court’s broad rule is not strong enough to warrant wholesale abandonment of the defense, despite the empirical arguments marshalled by the Court’s anti-immunity critics. The Court, in addressing the arguments of anti-immunity critics, should embrace a readily available and modest alternative to the current law that falls short of wholesale abandonment of the immunity doctrine.

The analysis in this Article honors the Court’s overarching framework for resolving immunity issues. That framework treats § 1983 and § 1985(3) as “common law statutes,” which authorize the Court to act freely in the manner of a common law court in shaping of official immunity doctrine.²⁴ The Court “seeks a proper balance between two competing interests.”²⁵ One set of interests involves vindicating constitutional rights and deterring constitutional violations. On the other side of the ledger lies an interest in ensuring fairness to government officials²⁶ and in lowering the “social costs” that constitutional tort litigation can generate.²⁷ The Court’s cost-benefit model rejects both the notion that a remedy should be available for every violation of a constitutional right,²⁸ and the proposition that executive officers should always enjoy immunity from suits for damages.²⁹

A key feature of the doctrine is that the Court refuses to balance interests on a case-by-case basis—for example, by evaluating in each case the precise degree of official wrongfulness, the nature and importance of the constitutional rights at stake, and other considerations presented by that particular case. Instead, the Court aims for “the best attainable accommodation of competing values,”³⁰ by applying the “clearly established law” rule to all qualified immunity cases.³¹ That rule thus “reflects a balance that has been struck across the board.”³²

This Article does not take issue with the Court’s rejection of case-by-case balancing in the qualified immunity context, particularly because a repudiation of case-by-case balancing marks much of our constitutional law. In the First Amendment context, for example, the Court has generally rejected such balancing, because it would “lead to unpredictable results and uncertain expectations.”³³ To say that the Court has wisely embraced a non-case-by-case balancing approach in qualified immunity cases is not to say, however, that the particular approach it has embraced makes good sense. Close examination suggests that the Court’s current approach gives too much weight to the pro-immunity side of the balancing ledger. Cost-benefit balancing is best employed, as it is in First Amendment doctrine, as a means for making rough, but useful, comparisons among categories of fact patterns.³⁴ A category-by-category approach would accommodate the competing policies at work in qualified immunity cases more fully than the Court’s current across-the-board rule. And it would do so without giving up the benefits of rule-based doctrine.

Part I of this Article focuses on Ziglar and makes two points, one in Section I.A and the other in Section I.B. Section I.A shows that the application of qualified immunity in Ziglar does not square with the aims of the doctrine, but only with the values served by bright-line rules. The Court’s premise, albeit unarticulated and unexamined, is that qualified immunity must operate as a sweeping rule that is equally applicable in each and every case. Section I.B argues that the case for a rule of this kind is unpersuasive, because its costs outweigh its benefits. Part II addresses critics of qualified immunity who favor its abolition by one means or another, in favor of a general rule of liability for constitutional torts.³⁵ Section II.A defends qualified immunity against objections that it rests on faulty empirical premises and Section II.B rejects an alternative approach that would greatly expand recoveries by shifting liability to local governments. This Article’s defense of qualified immunity, however, is not a defense of the current regime. Building on this idea, Part III proposes a new approach that gives significant—but not always controlling—weight to the values of rule-based doctrine. The alternative this Article calls for is judicial recognition of a limited number of categories in which the defense is unavailable because the benefits of immunity are significantly outweighed by its costs. Section III.A identifies six such categories. Section III.B then identifies two categories of cases in which the benefits of immunity are especially high in comparison to its costs. In these two sets of cases, according to the analysis offered here, a thoughtful build-out of categorical balancing rules might well support the displacement of the Court’s existing clear-error approach with a truly bright-line rule of absolute immunity.

1. Qualified Immunity in Ziglar

Official immunity protects all officials engaged in “discretionary” functions from liability for damages, at least to some extent. Judges,³⁶ legislators,³⁷ and prosecutors³⁸—or, more precisely, persons engaged in judicial, legislative, and prosecutorial functions³⁹—are absolutely immune from liability for damages, no matter how egregious their conduct is. Executive and administrative acts are covered by qualified immunity. Under Harlow v. Fitzgerald,⁴⁰ officers who engage in these acts are “are shielded from liability for civil damages insofar as their conduct does not violate clearly established statutory or constitutional rights of which a reasonable person would have known.”⁴¹ Thus, plaintiffs may lose even when they can prove a violation of constitutional rights and damages resulting from the violation.

A. Ziglar and the Policies Underlying Qualified Immunity

In Ziglar, Justice Kennedy’s majority opinion ordered dismissal of the § 1985(3) claim on the ground that all of the officer-defendants were shielded from liability under the qualified immunity doctrine. According to the opinion, the application of § 1985(3) to these allegations failed the “clearly established law” test, because “the conspiracy recited in the complaint was alleged to have been between or among officers in the same branch of the government (the Executive Branch) and in the same department (the Department of Justice).”⁴² Without addressing whether such “intrabranch” behavior can give rise to a conspiracy for purposes of § 1985(3), the Court held that the law on the question lacked sufficient clarity.⁴³ In order to show why, Justice Kennedy drew an analogy to the antitrust context.⁴⁴ In antitrust, the Court rejected the existence of “intracorporate conspiracies” to restrain trade because the officers of a corporation all act for the same corporate entity.⁴⁵ As such, “agents of the same legal entity are not distinct enough to conspire with one another.”⁴⁶ As for § 1985(3), the Court had never ruled one way or the other on the “intracorporate conspiracy doctrine.” Turning to lower court decisions, Justice Kennedy found a division of authority over the viability of the intracorporate conspiracy doctrine, in cases that involved suits against private businesses and officers sued for violations under § 1985(3).⁴⁷ This division of lower court authority was apparently decisive: “When the courts are divided on an issue so central to the cause of action alleged, a reasonable official lacks the notice required before imposing liability.”⁴⁸ Thus, the officials avoided liability because “[u]nder these principles, . . . reasonable officials in petitioners’ positions would not have known, and could not have predicted, that § 1985(3) prohibited their joint consultations and the resulting policies that caused the injuries alleged.”⁴⁹

Ziglar is an ambitious qualified immunity case in part because the Court ruled on a motion to dismiss the complaint, and thus “accept[ed] as true the facts alleged in the complaint,”⁵⁰ including the allegation that the officers unjustifiably discriminated against the plaintiffs on the basis of race, religion, and national origin. The impermissibility of such discrimination has been settled for decades, and the Court in Ziglar does not suggest otherwise. The only lack of clarity in the case involved the intrabranch conspiracy issue. The question thus arises whether this type of lack of clarity should matter.

1. Fairness to defendants

An important objection to the Court’s reliance on “uncertainty as to intracorporate conspiracies” is that the uncertainty on that point has nothing to do with the either of the two rationales for qualified immunity. First, official immunity is recognized as a matter of fairness when an officer reasonably believes he is acting within the constitutional rules.⁵¹ For example, in Pierson v. Ray,⁵² the first modern case in the development of qualified immunity for constitutional torts, the Court awarded qualified immunity to police officers to avoid the unfairness of imposing liability without fault.⁵³ Writing for the Court, Chief Justice Warren emphasized the unfairness of a strict liability rule when he wrote: “A policeman’s lot is not so unhappy that he must choose between being charged with dereliction of duty if he does not arrest when he has probable cause, and being mulcted in damages if he does.”⁵⁴ the Court added that recognizing qualified immunity avoids “the injustice, particularly in the absence of bad faith, of subjecting to liability an officer who is required, by the legal obligations of his position, to exercise discretion.”⁵⁵ More recently, in Hope v. Pelzer,⁵⁶ the Court said that the point of qualified immunity is that officials are entitled to “fair warning” that their act violates federal law.⁵⁷

This fairness rationale is analogous to the negligence rule in ordinary tort law. A central principle of negligence law is that it is unfair to impose liability on a private actor who cannot reasonably foresee that his act will injure someone.⁵⁸ It seems similarly unfair to impose liability on an officer who cannot reasonably foresee that his act will violate a plaintiff’s constitutional rights. There is a difference between private actors and state officials, in sense that officials typically do intend to touch, threaten, imprison, or otherwise invade the plaintiff’s person or property. But this difference is irrelevant to the “fair warning” issue. Fair warning relates to whether the officer made a choice to risk unconstitutional interference with the plaintiff.⁵⁹ Some intentional interferences are unconstitutional and others are not.⁶⁰ Officers who reasonably but mistakenly believe they have sufficient grounds to make an arrest, fire an employee, or revoke a business license lack “fair warning” that they have crossed constitutional lines.⁶¹ This principle is reflected in the common law of intentional torts, which declines to impose liability on a person who shoots another, even a police officer, in the absence of fair warning that his target is not an aggressor.⁶² It is equally unfair to impose constitutional tort liability on an officer in a similar situation.

2. “Social costs”

Second, constitutional tort litigation generates “social costs [which] include the expenses of litigation, the diversion of official energy from pressing public issues, and the deterrence of able citizens from acceptance of public office.”⁶³ All litigation produces similar costs, yet no immunity defense is generally available to private persons and entities. The difference between private and public actors is that public actors face a different incentive structure. The core of the “social costs” rationale for official immunity is that, at least in theory, private actors capture the benefits of their actions.⁶⁴ When those actions generate risks to others, private actors should take those costs into account.⁶⁵ The aim is to encourage them to (more or less accurately) weigh the real costs against the benefits. Public officers do not capture the benefits of their actions in the same way. Those benefits go to society at large, for example, in the form of effective policing, fire protection, safe products, clean air and water, good teaching, and generally efficient administration of government.⁶⁶

Allowing for official immunity maximizes social utility and, by extension, social welfare by enabling officials to act without limitation toward the public good.⁶⁷ If officers were liable for every constitutional violation, they might hesitate before taking a step that produces a public benefit because an error would lead to personal liability. A police officer may decline to make an arrest that arguably violates the Fourth Amendment, or the supervisor of a government agency may decline to fire an incompetent employee whose job is arguably protected under the due process clause. From the “social costs” perspective, the rationale for the defense is that official immunity will correct the asymmetrical incentives officials otherwise face.⁶⁸ The availability of a defense protects officers who act in the public interest. Thus, the defense encourages the police officer to make the arrest and the supervisor to fire the employee despite the uncertain constitutional issue that would otherwise counsel caution.

3. Fairness, utility, and the conspiracy issue

In Ziglar, neither fairness nor utility were threatened by the imposition of liability. According to the complaint, the officials deliberately discriminated on the basis of race, religion, and national origin in their abusive treatment of the detainees.⁶⁹ There is no lack of clarity in the constitutional doctrine that forbids these types of discrimination, no lack of fair warning that the acts are forbidden, and no danger that officials will be overly cautious in an area in which the constitutional line is uncertain. The application of § 1985(3) conspiracy doctrine is a collateral matter. Lack of clarity on that issue is as irrelevant as lack of clarity on the statute of limitations, or issue preclusion. It is as though the Court, instead of permitting an action for a clear constitutional violation in Bivens itself, should have held that the officials were immune because of lack of clarity as to the previous available Bivens actions. Many questions may determine outcomes, but not all of them have to do with the state of the law as to whether the conduct at issue violated the plaintiff’s constitutional rights. If the officers knew they could be sued under § 1985(3) pursuant to an “intrabranch conspiracy” theory, they understood (according to the complaint) that they were violating constitutional rights.⁷⁰

In this respect, the Ziglar Court seems to enlarge the scope of qualified immunity beyond the Harlow doctrine.⁷¹ Harlow held that executive officials are immune from damages “insofar as their conduct does not violate clearly established statutory or constitutional rights of which a reasonable person would have known.”⁷² The plaintiffs in Ziglar claimed violations of their rights to equal protection and religious liberty. These rights were “clearly established,” no matter how unclear the doctrine may have been on whether intragovernmental conspiracy can trigger § 1985(3) liability.⁷³ Under Ziglar, it appears that the defendant may win without showing lack of clarity as to the right violated, so long as he can show that the law is uncertain as to other aspects of the plaintiff’s case.

B. Benefits and Costs of the Ziglar Rule

In Ziglar, the Court treats “lack of clearly established law” as a broad rule, which hinges on lack of clarity across-the-board, not only as to the core legal obligation owed by the defendant, but also as to whether § 1985(3) reaches collaboration among members of the same Executive Department.⁷⁴ The rule controls the outcome even though there is no link between the rationales for the rule and its application to the plaintiff’s substantive claims. Ziglar presents a typical example of the Supreme Court’s increasing focus on developing sharp edge rules in its qualified immunity cases. The early cases on qualified immunity considered several factors in their application of the defense, including the nature of the defendant’s duties and his state of mind.⁷⁵

The official himself must be acting sincerely and with a belief that he is doing right, but an act violating a student’s constitutional rights can be no more justified by ignorance or disregard of settled, indisputable law on the part of one entrusted with supervision of students’ daily lives than by the presence of actual malice.

Wood, 420 U.S. at 321.

Beginning with Harlow, the Court moved to a more rule-oriented approach, by taking two steps.⁷⁶ First, it abandoned the old distinctions among government jobs in favor of a general norm for all qualified immunity cases.⁷⁷ Second, it eliminated the subjective “malice” or “good faith” prong in favor of what it characterized as an “objective” test.⁷⁸ Some of the reasoning of Harlow seemed, at least to Justice Brennan, to permit recovery when the official “actually knows that he was violating the law . . . even if he could not ‘reasonably have been expected’ to know what he actually did know.”⁷⁹ The Court blocked even that avenue a few years later in Anderson v. Creighton,⁸⁰ holding in the context of a Fourth Amendment challenge to a search that the test is strictly objective.⁸¹ The officer’s “subjective beliefs about the search are irrelevant.”⁸² The Court further extended the rule in Ashcroft v. Al-Kidd,⁸³ at least in its rhetoric, by stating that it would apply unless “every” officer would have been on notice that his action was unconstitutional, and by adding that the constitutional validity of the officer’s action must have been “beyond debate” at that time.⁸⁴ But Ziglar goes further than any of these cases. The innovation in Ziglar is to apply qualified immunity to lack of clarity not only on the content of “constitutional rights,” on which Harlow had focused, but on a collateral matter as to the applicability of a particular statute.

1. Form and substance in qualified immunity doctrine

It is easy to criticize practically any rule, because there is always a gap between the fact patterns to which the rule applies and the policies on which the rule is based. As a result, the rule is overinclusive, underinclusive, or both.⁸⁵ For example, because the Court’s rule for qualified immunity in Ziglar is triggered by “lack of clarity” in judicial interpretation of the conspiracy statute, and not based on fairness and utility rationales, Ziglar sacrifices the plaintiff’s constitutional tort interest in recovery for a violation of settled constitutional rights. The ill-fit of the grounds for a rule and the operation of the rule is a defining feature of rules.⁸⁶ Otherwise, the “rule” would be superfluous. A decision maker would reach the same result by applying the policies underlying the rule. Yet rules are often justified, despite their clumsiness and seeming arbitrariness, because they serve critical values of the legal system. Across many areas of law rules contribute to the resolution of disputes by saving time and effort in decision making,⁸⁷ generating and maintaining stability in the law,⁸⁸ enabling us to predict of outcomes of disputes that may arise,⁸⁹ and controlling the discretion of officers who apply the law to particular cases.⁹⁰

Thus, to evaluate the Ziglar rule fairly, the question of its merit should be framed differently, as a balance between costs and benefits. While the rule may be arbitrary, it has many benefits such as predictability, stability, assuring fairness to officials, and easing concerns about overdeterrence of effective government operations. Given the benefits of rules, what substantive interests on the plaintiff’s side of the case are sacrificed under the Ziglar rule? That is, what are its costs? Notably in Harlow and many other cases, the Court identifies the plaintiff’s interest as “vindication” of rights and “deterrence” of constitutional violations.⁹¹ Can those costs be diminished by an approach that is less sweeping than that of Ziglar, without unduly sacrificing the fairness and social utility benefits that ruling provides? The resolution of this question necessarily depends on value choices. Neither the rigorous champion of remedying constitutional violations nor the hardcore advocate of effective government will have difficulty answering them, though in opposite directions. Many judges, however, will find that their initial value judgments fall somewhere between the two extremes. On the premise that there is value on both sides, the task is to draw distinctions between contexts in which the case for immunity is stronger or weaker.

There is no room for argument that a broad immunity rule like that of Ziglar will entirely block vindication of constitutional rights and deterrence of violations in some contexts. When the constitutional violation is in the past, damages constitute the only viable remedy. Courts will deny a request for an injunction or declaratory judgment unless a plaintiff can show a likelihood of recurrence;⁹² City of Los Angeles v. Lyons⁹³ illustrates this theme. In Lyons, the police put the plaintiff in a chokehold.⁹⁴ He sued for both damages and an injunction, on the theory that the officer used excessive force.⁹⁵ The Court held that the suit for an injunction could not go forward, because Lyons could not show that he was likely to be choked again.⁹⁶ Yet there was no assurance that the police would not continue to use chokeholds.⁹⁷ A combination of a broad immunity-from-damages doctrine and this unavailability of prospective relief could eliminate any action that seeks to vindicate constitutional rights and deter constitutional violations.

The problem exists across a range of constitutional tort issues, so long as a plaintiff’s disagreeable encounter with the government is unlikely to be repeated. Many constitutional tort suits involve random encounters with the police,⁹⁸ or complaints by inmates of their treatment by guards on a specific occasion,⁹⁹ or one-time decisions by high-level officers to fire or otherwise disadvantage government employees under their supervision.¹⁰⁰ Typically, the Lyons requirement of likelihood of recurrence is not present, and for these plaintiffs, “it is damages or nothing.”¹⁰¹ For this basic reason, alternatives to the Ziglar rule deserve careful attention.

One simple way to give more weight to vindication and deterrence, at little cost to the fairness and social utility goals served by official immunity, would be for the Court to repudiate Ziglar’s focus on unclear “law” and return to the Harlow formulation, under which immunity is available only for lack of clarity as to the “rights” the defendants have violated. The Ziglar opinion gives no indication that the Court focused on this possibility or even noticed that it was expanding immunity far beyond Harlow.¹⁰² Nor is the Court’s subtle expansion of immunity addressed by Justice Breyer’s dissent.¹⁰³ At the very least, the Court should have to explicitly address the issue of whether immunity extends to cases like Ziglar, in which the lack of clarity has nothing to do with the plaintiff’s substantive rights.

Another step would be to eliminate small but tactically important glosses on the Harlow formulation. For example, the Court has said that immunity would be available unless clearly established law places the “statutory or constitutional question beyond debate,”¹⁰⁴ and that immunity is available unless “every” reasonable officer would understand that the act was unconstitutional.¹⁰⁵ This sort of rhetoric seems to disappear when courts move to deny immunity.¹⁰⁶

2. “Featureless generality” as an alternative to Ziglar

The overinclusiveness objection to the qualified immunity rule is not limited to the Ziglar version of the defense. John Jeffries argues that the misalignment between the qualified immunity policies and the Harlow “clearly established law” rule should be remedied by reformulating the rule, so that it will “adhere more closely to the rationales for limiting liability in the first place.”¹⁰⁷ To this end, he rejects the Court’s “hyper-technical” rule, in favor of “a broader concept of notice derived from existing law.”¹⁰⁸ He would “shift the doctrinal focus from whether the defendant violated a ‘clearly established’ right to whether the defendant’s actions were ‘clearly unconstitutional.’”¹⁰⁹ What he means by this distinction is that the Court should take a holistic approach to the situation presented rather than a technical, precedent-oriented, one. If the police tie a man to a pole in the middle of the night and leave him there, their conduct is unconstitutional even if the case law does not clearly say so.¹¹⁰ Professor Jeffries seems to me to be on the right track, but I am skeptical of this solution. The problem with it is that, without a change in the structure of the doctrine, it seems unlikely a shift in emphasis will necessarily have much impact. As Professor Jeffries admits, “no mere turn of phrase can avoid hard cases or ensure good decisions.”¹¹¹

Professor Jeffries’ proposal accepts the Court’s current immunity framework, in which the doctrine consists of a rule that cuts across all categories of cases.¹¹² That feature of the doctrine makes it “a relatively crude tool” for mediating between rights and remedies.¹¹³ A more radical solution to the overinclusiveness costs of Harlow is to abandon the rule-oriented approach in favor of borrowing from the common law tort of negligence. Instead of a lack-of-clarity rule, the parties would be asked to present an array of evidence for a jury to consider. The judge would instruct the jury to decide whether, in light of all of the evidence it has heard, the officer violated “clearly established statutory or constitutional rights of which a reasonable person would have known.”¹¹⁴ By this means, the plaintiff would have an opportunity to make the case for vindication and deterrence in a wide range of cases,¹¹⁵ and the defendant would be obliged to persuade the jury that the fairness and avoidance of overdeterrence concerns should control in the circumstances of the particular case. In this “ordinary tort” approach, officers would be obliged to rely on reasons specific to the case at hand to support their immunity claims, including but not limited to the lack of clarity of the rights at stake, rather than obtaining dismissal based on judicial application of the “clearly established law” rule.

To see how this approach might work, consider the facts of Brown v. Elliott,¹¹⁶ a typical qualified immunity case. Elliott, a police officer, leaned into the open window of a car at a traffic stop.¹¹⁷ The driver started the car and when it began to move, Elliott fired a shot, killing a passenger.¹¹⁸ The victim’s estate sued for a Fourth Amendment violation, claiming excessive force¹¹⁹ since the relevant precedents had ruled that an officer violates the Fourth Amendment when he shoots at a car that is speeding away from him.¹²⁰ However, Brown held that those earlier cases did not clearly establish that an officer in Elliott’s position should not fire his weapon.¹²¹ According to the court, the difference was that in the earlier cases, “the threat to [the officers’] safety was eliminated and thus could not justify the subsequent shots.”¹²² That reasoning may not hold when the officer “was leaning into the window of a moving truck, not standing off to the side as the truck passed him without veering in his direction.”¹²³ Qualified immunity applied even if Elliott used excessive force,¹²⁴ since the earlier cases did not clearly establish that proposition.

In the hypothesized “ordinary tort” alternative, the judge would initially resolve the constitutional issue. If that ruling favored the plaintiff, the jury would then rule on qualified immunity. Jurors would consider all of the circumstances surrounding the shooting and the state of the law at the time to decide whether the officer acted reasonably in light of “clearly established law” at the time of the shooting.¹²⁵ In principle, this approach would solve the problem of overinclusiveness and underinclusiveness that is endemic to rule-based decision making, just because the trier of fact would have access to a much wider array of information and much more discretion to arrive at the optimum outcome.¹²⁶ But that solution would come at a heavy price. One objection to the case-by-case approach is that the judge-jury issue is one on which constitutional torts differ in a fundamental way from common law torts. The leading case on the role of the jury in adjudicating negligence is Sioux City & Pacific Railroad Co. v. Stout,¹²⁷ in which the Court said that “twelve men know more of the common affairs of life than does one man,” and “they can draw wiser and safer conclusions . . . than can a single judge.”¹²⁸ This rationale does not support jury determinations of whether the constitutional rule was clearly established at the time an officer violated it. Jurors bring no particular wealth of expertise to the task of determining the clarity of constitutional norms.

The viability of the case-by-case approach does not depend solely on whether juries are competent to adjudicate the “clearly established law” issue.¹²⁹ That issue could be left up to the judge. But even if judges decide whether or not there is a “clearly established law” that speaks to the issue of qualified immunity, the case-by-case approach would exacerbate the social costs of constitutional tort litigation, and particularly the danger of overdeterrence of effective government. In ordinary tort law, the reasonable care standard is characterized by “featureless generality,”¹³⁰ which “introduces a large element of uncertainty even into ostensibly routine cases.”¹³¹ Even when judges, rather than juries, adjudicate negligence issues, much uncertainty remains because a range of factors will bear on the decision the judge must consider. This “ordinary tort” approach may produce excessive caution on the part of officials across a range of cases in which the balance between competing factors is uncertain.

A more promising approach is to look for an alternative to Harlow that will keep the advantages of rules without Harlow’s often severe misalignment between the rule and the policies. One alternative, discussed in Part II, is to abandon immunity for a general rule of strict liability, imposed either on officers or governments.¹³² Another alternative, described in Part III, is to reformulate qualified immunity doctrine by carving out categories of fact patterns in which the case for immunity is especially strong or weak, in order to retain the value of immunity while cutting its costs.¹³³ Like Harlow itself, and unlike the “ordinary tort” hypothesis, both of these alternatives give considerable weight to the value of rules. However, the main difference between them is that the abolitionist view gives priority to the substantive goals of vindicating constitutional rights and deterring violations, while the categorical approach attempts to achieve a balance between the two sets of goals.¹³⁴

II. The Costs and Benefits of Abolishing Official Immunity

Some critics of qualified immunity call for its abolition or severe curtailment, either directly or by shifting the liability to local governments.¹³⁵ Three distinct reasons are advanced in support of this view. First, Professor William Baude has argued that qualified immunity is “unlawful,” because neither the text nor the common law background of § 1983 authorizes the defense.¹³⁶ Second, various studies support the conclusion that the empirical foundation on which the Court bases qualified immunity are weak.¹³⁷ Third, critics argue that the Court’s doctrine gives too much weight to the government official’s interest in immunity and too little weight to the plaintiff’s interest in recovery.¹³⁸ This part of the Article uses the “proper balance” framework to challenge the “empirical evidence” and “value choices” aspects of the uncompromising “anti-immunity” position. On both fairness and incentives grounds, the case for an immunity defense remains strong. Section II.A examines the claim that empirical findings about the impact of immunity cast doubt on the supposed benefits of immunity and thereby support radical change in official immunity law. Section II.B acknowledges the argument that the imposition of liability on local governments would enable the courts to capture the vindication and deterrence benefits of liability at little or no cost, since individual officers would not be liable.

A. The Empirical Case Against Qualified Immunity

Harlow’s version of qualified immunity rests, in part, on two empirical assumptions: (1) officials will be overly cautious and thus deterred from acting effectively in the public interest unless they are shielded from liability for damages; and (2) Harlow’s objective test will diminish the “social costs” of constitutional tort litigation by permitting many suits to be dismissed before trial and even without discovery.¹³⁹ Experience under the Harlow regime raised doubts about both of these propositions. If in fact the benefits of immunity are small, a strong case can be made that the resulting reduction in the vindication of constitutional rights and deterrence of violations, are not worth bearing. As a result, the argument goes, immunity should be abandoned.

1. Two studies and their implications

Joanna Schwartz, the author of studies that bear on both of the Court’s empirical assumptions, argues that her work undermines the premises of qualified immunity in just this way. One of her projects examined the practice of government indemnification of police officers held liable for damages for constitutional torts.¹⁴⁰ She gathered information about “litigation payments and indemnification decisions in forty-four . . . large [law enforcement] departments and thirty-seven . . . small and mid-sized departments,” which “employ approximately [twenty percent] of law enforcement officers across the country.”¹⁴¹ She found that “[p]olice officers are virtually always indemnified.”¹⁴² That is, when they are held liable under § 1983, their employers pay the judgments.¹⁴³ To the extent her findings are valid and can be generalized to other government officers, they diminish the concern that officers need immunity or else they will act timidly for fear of liability.¹⁴⁴ After all, if officers do not actually pay damages, they will likely act less cautiously.

A recent study, conducted by Professor Schwartz, examines whether the Court in Harlow had a valid concern when it eliminated the subjective “bad faith” prong of official immunity.¹⁴⁵ According to Harlow, “substantial costs attend the litigation of the subjective good faith of government officials.”¹⁴⁶ These include “the general costs of subjecting officials to the risks of trial,” the difficulty of deciding “subjective intent” on summary judgment, and “broad-ranging discovery and the deposing of numerous persons,” inquiries which “can be peculiarly disruptive of effective government.”¹⁴⁷ Schwartz “reviewed the dockets of 1183 lawsuits filed against state and local law enforcement defendants over a two-year period in five federal district courts” from Texas, Florida, Ohio, Pennsylvania, and California,¹⁴⁸ and found that “just 0.6% of cases [against law enforcement defendants] were dismissed at the motion to dismiss stage and 2.6% were dismissed at summary judgment on qualified immunity grounds.”¹⁴⁹ Though immunity still blocks relief in many cases and discourages prospective plaintiffs from bringing suits, Schwartz’s findings suggest that the Harlow test may well not diminish the costs of constitutional tort litigation by as much as the Court had expected.

Professor Schwartz’s findings challenge the viability of the current qualified immunity doctrine. The Court has said that the “‘clearly established’ standard protects the balance between vindication of constitutional rights and government officials’ effective performance of their duties.”¹⁵⁰ Put in terms of the cost-benefit model the Court has endorsed, the indemnification study suggests that the “avoidance of overdeterrence” benefits of immunity are smaller than had been anticipated. That goal is already served by government reimbursement. Even before Schwartz’s indemnification study, the Court acknowledged the relation between indemnification and the anti-overdeterrence policy.¹⁵¹ It took account of the anti-immunity implications of indemnification when it rejected qualified immunity for private prison officers sued under § 1983, in part because they could be indemnified by their employers.¹⁵² Professor Schwartz’s second paper on the impact of Harlow on social costs, suggests that the “avoidance of trial and discovery” benefits of the elimination of the subjective prong are small.¹⁵³ Professor Schwartz’s empirical studies thus cast doubt on the Court’s assessment of the cost and benefits of qualified immunity and suggest that the costs imposed override what little benefits qualified immunity provides.

2. Skepticism about empiricism

Though empirical studies of constitutional tort litigation deserve attention, they cannot by themselves resolve the qualified-immunity issue. Professor Schwartz focuses on law enforcement. Studies of other contexts, or scholars using other methodologies, may produce different outcomes. In addition, it is important to distinguish between the question of whether qualified immunity provides a cost or benefit and the question of how great that cost or benefit might be. Showing that indemnification decreases timidity does not suffice to show that timidity is an insignificant concern.¹⁵⁴ Showing that social costs remain despite Harlow does not suffice to show that they would not be greater without Harlow. Even if Schwartz’s findings are valid across the whole field of constitutional tort litigation, as I suspect they are, other rationales for immunity remain valid. Professor Jeffries, in particular, has argued that immunity from damages “fosters the development of constitutional law . . . by reducing the costs of innovation.”¹⁵⁵ Professor Fallon posits an “Equilibration Thesis,” which holds that “courts charged with implementing constitutional rights or values sometimes view justiciability doctrines, merits doctrines, and remedial doctrines as an integrated unit.”¹⁵⁶ An implication of this thesis is that eliminating qualified immunity could result in dilution of the content of substantive constitutional rights.¹⁵⁷

In addition, the idea that empirical studies might justify massive doctrinal change ignores the institutional role judges play in amending doctrine. Qualified immunity is a judge-made doctrine¹⁵⁸ and is certainly as subject to judicial reexamination as any other common-law doctrine. But it is also intimately connected to the law of constitutional remedies. Its abrupt termination would threaten the process values that inhere in the common law tradition, which favors incremental rather than abrupt change in judge-made law.¹⁵⁹ In most instances, courts can and should respond to new information and new conditions with incremental adaptation.¹⁶⁰ That is the point of Holmes’ aphorism that “[t]he life of the law has not been logic: it has been experience.”¹⁶¹ The Court in Harlow acted in the spirit of this precept, even if it acted mistakenly, when it dropped the bad-faith prong of the qualified immunity doctrine based on its sense that “[t]he subjective element . . . has proved incompatible with our admonition
. . . that insubstantial claims should not proceed to trial.”¹⁶² It was “now clear that substantial costs attend the litigation of the subjective good faith of government officials.”¹⁶³ Taking the incremental approach of the common law, the Court retained other components of the qualified immunity defense while eliminating the subjective prong.¹⁶⁴ Professor Schwartz’s findings suggest that the Court in Harlow may have been mistaken in thinking it had found a remedy for the problem it tried to solve. But the more realistic, and prudent, goal is not to eliminate qualified immunity—it is to undo the specific holding of Harlow and reinstate the bad faith prong. That is one element of the categorical approach described in Part III. In this and other respects, the categorical approach is more consistent with the norms of common law adjudication, as it modifies rather than uproots the current doctrine.

3. Empirical studies and the fairness rationale

Professor Schwartz’s findings have no bearing on the fairness rationale for qualified immunity. From a fairness perspective, it does not matter whether the defense actually achieves the goal of reducing social costs. The fairness rationale bases the defense on “the injustice, particularly in the absence of bad faith,”¹⁶⁵ of imposing liability on an officer in the absence of “fair warning” that his act violates a constitutional right.¹⁶⁶ The core of this rationale is that when the law or the application of law to fact is unclear, it is unfair to impose liability on an official who, as a result, cannot reasonably foresee that his act violates the plaintiff’s constitutional right.¹⁶⁷ As Holmes put the point in his defense of the negligence principle, “[a] choice which entails a concealed consequence is as to that consequence no choice.”¹⁶⁸ This objection to liability does not depend on whether the “social costs” of constitutional tort litigation are diminished by qualified immunity. Nor is it necessarily met by findings that officers are typically indemnified. Liability may have other consequences for the officer’s future employment prospects or personal life. Even if it does not, the vindication goal applies to both sides of the constitutional tort suit. An officer who is charged with a constitutional violation has a legitimate interest in vindication of his actions. The fact that he is indemnified does not eliminate the unfairness of liability without fault.

If Professor Schwartz’s findings are confirmed, the likely (and, in my view, the appropriate) impact will not be abandonment, but rather reconstruction of qualified immunity, with greater emphasis on the dynamics of Professor Jeffries’ “right-remedy gap” and on the fairness rationale. Framed in cost-benefit terms, what she has shown is that that the benefits of immunity in cutting social costs are not as great as the Court has thought them to be.¹⁶⁹ She has made the case for a less defendant-friendly doctrine, perhaps along the lines discussed in Part III.¹⁷⁰

B. Local Government Liability

Many proponents of liability unfettered by immunity sidestep the fairness-to-officials objection. They propose that local governments¹⁷¹ be held liable for constitutional violations committed by their employees.¹⁷² This proposal would extend liability well beyond current practice as well as current law. First, § 1983 does not entitle any officer to indemnification.¹⁷³ Second, under the current indemnification practice, governments typically pay for some of their employees’ constitutional torts, but not all of them.¹⁷⁴ Under the current regime, they do not pay for violations of constitutional rights as to which the officer successfully asserts immunity.¹⁷⁵ However, the “broad liability” argument holds that local governments should be liable for all violations.¹⁷⁶

A defense of the categorical approach advocated in this Article cannot be complete without an examination of this alternative. Indemnification aside, the current doctrine imposes damages liability on local governments only in a limited range of circumstances.¹⁷⁷ The leading case is Monell v. Department of Social Services,¹⁷⁸ which holds local governments liable for constitutional violations caused by their “official policy” or “customs,”¹⁷⁹ categories that the courts generally read narrowly. Under the Court’s interpretation of § 1983 in Monell, local governments are not vicariously liable in suits for constitutional torts committed by their employees.¹⁸⁰ In so holding, the Court focused on the language of § 1983 as originally passed, which held liable persons who “shall subject, or cause to be subjected any person” to the deprivation of rights.¹⁸¹ According to the Court, this language “cannot be easily read to impose liability vicariously on governing bodies solely on the basis of the existence of an employer-employee relationship with a tortfeasor.”¹⁸² This is the rule that would need to be changed in order to expand liability for constitutional torts without lifting official immunity.¹⁸³ Given that Monell is a statutory precedent, not federal common law, the “institutional role” case against overruling the no vicarious liability holding may be stronger than with other constitutional tort issues.¹⁸⁴

If Monell’s “no vicarious liability” doctrine was overruled, local governments would be generally liable for their employees’ violations of constitutional rights. However, Owen v. City of Independence¹⁸⁵ held that the municipal government has no immunity defense—so the combined effect of Owen and Monell is that local governments are liable without fault, but they are liable only when the plaintiff can meet the “policy or custom” requirement.¹⁸⁶ To see how the proposal to institute vicarious liability would alter the current regime, it is helpful to briefly sketch in a rough way the distinction drawn by Monell and its progeny between “policy and custom,” for which governments are liable, and the other constitutional torts, for which governments are not liable.

Consider a hypothetical case in which a municipal police officer uses a taser on a motorist at a traffic stop, and the tasing is found to be excessive force in violation of the motorist’s Fourth Amendment rights.¹⁸⁷ Under the current approach, is the officer’s employer, the city, is liable if its written or unwritten policy authorizes police officers to use tasers without special need,¹⁸⁸ but not if it has no such policy and an officer uses a taser without the necessary justification.¹⁸⁹ The city may be liable in the absence of a written policy allowing indiscriminate use of tasers, if the plaintiff can show that a “final policymaker,” such as the police chief, has directed officers to use them,¹⁹⁰ or if the chief knows about a “custom” among police officers and tolerates it.¹⁹¹ The local government may also be liable under Monell on an “inadequate training” theory if the plaintiff can show that officers have not received minimally adequate training and that this lack of training caused the violation.¹⁹² Unless the plaintiff can bring his case within one of these categories, he has no legal right to relief from the local government.

The anti-immunity school would combine retention of Owen with reversal of the “no vicarious liability” rule. The city would be liable in all of the above scenarios, for any unconstitutional use of a taser, whether or not the law on taser use was clearly established, and whether or not the tasing was authorized by an ordinance, pursuant to a custom, approved by a policymaker, or the result of bad training. Put in terms of costs and benefits, the major advantages of this approach are, first, that rights are generally vindicated, more or less across the board, at least so long as the case involves acts by local government employees and not by state government officers.¹⁹³ Second, governments will internalize the costs of constitutional violations, and will thus have proper incentives to minimize them.¹⁹⁴ With respect to costs, the Court said in Owen that the “inhibiting effect” of the threat of liability “is significantly reduced, if not eliminated . . . when the threat of personal liability is removed.”¹⁹⁵ Unfairness “is simply not implicated when the damages award comes not from the official’s pocket, but from the public treasury.”¹⁹⁶

1. Incentives

Professor Alan Sykes’ economic analysis of ordinary tort law provides some support for the imposition of vicarious liability on governments for all constitutional violations committed by their employees in the course of employment.¹⁹⁷ Summarizing Sykes’ study, Professors Richard Epstein and Catherine Sharkey state that “[a]s between the employer and employee . . . the employer is usually the superior risk-bearer because of its greater access to insurance markets,” that “vicarious liability . . . reduces the risk that the insolvency of a particular employee will impose an uncompensated risk on a third party.”¹⁹⁸ Additionally, they found that “the doctrine reduces the need to have an extensive network of voluntary contracts between employees and employers,” and that “vicarious liability protects third parties who know that some firm employee is responsible for the loss, but cannot determine which one is responsible.”¹⁹⁹ These rationales for vicarious liability seem equally applicable to constitutional torts, since the employer-employee relationship is similar to that of a private business. Still, the adoption of government strict liability would not eliminate all constitutional violations by local government officers, even if the incentives created by the liability rule work perfectly. Even in a strict liability regime, private businesses will take precautions only when the gain in accident prevention is greater than the cost of the precaution.²⁰⁰ In the same way, a rational, wealth maximizing government would only stop the constitutional violations that cost the city more (in damages and litigation expenses) than the cost of preventing them, even if it were liable for all of violations. From an economic perspective, some constitutional violations, like some accidents, are not worth preventing.²⁰¹

In addition, there are three incentive-based reasons to doubt whether the private law model can simply be transplanted into constitutional torts with no allowance for the officer’s fault or lack of it. One is that the problem of officials’ self-protective behavior does not vanish just because officials are not personally liable. Larry Kramer and Professor Sykes have examined municipal liability from an incentives-based perspective.²⁰² Strict liability would have both pluses and minuses, because “[t]he prospect of liability will discourage actions that would ultimately prove constitutional as well as actions that would ultimately prove unconstitutional.”²⁰³ Given the tradeoff, it is “impossible to know whether the inefficiencies of self-protective behavior by municipalities will exceed the efficiencies associated with a reduced number of constitutional torts.”²⁰⁴ Consequently, “the efficiency of municipal liability for good faith constitutional torts is very much in doubt.”²⁰⁵

The second reason for skepticism about the private-to-public transplant is that the kinds of incentives that matter to private actors may differ from the ones that count the most for government officials. In theory, if not always in practice, a private business bears the costs and captures the benefits of precautions and therefore will give those costs the weight they deserve in decision making. Government policymakers may not respond to the incentives created by liability rules in the same way businesses and individuals do.²⁰⁶ Among other differences, “[g]overnment does not behave like a wealth-maximizer . . . [r]ather, government internalizes only political incentives.”²⁰⁷ For example, a police commissioner may further his own interests in getting reelected by unduly vigorous policing, with the constitutional tort costs paid for by taxpayers.

The third problem with transplanting the private law respondeat superior model is based on Professor Jeffries’ insight that a right-remedy gap serves the systemic value in constitutional innovation.²⁰⁸ A static cost-benefit calculation cannot take the long run gain into account. A dynamic model is required. Over time, the effect of blanket local government liability would increase the costs of recognizing new constitutional protections and of maintaining the current ones. Professor Fallon’s “Equilibration Thesis” suggests that rights may be defined in ways that produce less rather than more constitutional protection.²⁰⁹ According to Professor Jeffries, a limit on the availability of damages lowers the costs of recognizing new rights, and thereby “facilitates constitutional innovation and favors the future.”²¹⁰ Some form of qualified immunity and some limits on local government liability may be justified even though “underdeterrence of constitutional violations is a cost and must be counted as such.”²¹¹ If Fallon and Jeffries are correct, the costs of a shift to strict government liability may be far larger than the critics of immunity suppose. The Court might define constitutional rights more narrowly if the monetary costs of rights were higher, and the result could be a weaker rather than a stronger set of constitutional guarantees.²¹²

2. Fairness

Fairness is not at issue when beneficiaries’ money damages is awarded from the public treasury. In Owen, the Court dismissed the notion that the value of fairness should constrain government liability.²¹³ The “public at large” gets “the benefits of the government’s activities,” and “the public treasury” should pay the constitutional tort costs of those activities.²¹⁴ He feared that harmful § 1983 judgments could endanger local governments and consequently could impose “severe limitation[s] on their ability to serve the public.”²¹⁵ In a long dissenting opinion, Justice Powell briefly addressed this topic, suggesting that “ruinous judgments” under § 1983 could be harmful to local governments, and could inflict “severe limitations” on their ability to serve the public.”²¹⁶

Justice Powell’s brief rejoinder relates only to the scope of “policy and custom” liability. As intended, he does not fully and precisely articulate the fairness objection to governmental liability for constitutional torts when unlimited by either policy or custom or any immunity defense. The point is not just that liability will threaten government finances. If a government acts in disregard for constitutional rights often enough and flagrantly enough, it may deserve to collapse under the weight of constitutional tort judgments. The stronger rationale for a fault rule is that the ethical objection to liability without fault applies to municipalities as well as to officers. Government is not an abstraction. It is the sum of all the persons who contribute to it and benefit from it. Thus, Owen’s strict liability rule, even in its current limited form, favors the successful constitutional tort plaintiff over all other claims on government monies, including not only taxpayers but also the beneficiaries of government programs. Call this group “ordinary citizens” for short. The issue is which of these two groups—victims of constitutional violations or ordinary citizens—should, as a matter of fairness, bear the losses due to actions that are arguably proper at the time they are taken but are found to be constitutional violations with the benefit of hindsight.

Professor Daryl Levinson has asserted that “taxpayers do not ‘cause’ constitutional violations in any intuitive sense of causation, nor are they morally responsible for constitutional wrongdoing.”²¹⁷ One need not go that far in order to defend a fault rule. When government, through its employees, has fair notice that the acts of those employees may cross a constitutional line, the pro-plaintiff policies—vindication of constitutional rights and deterrence of violations—provide strong support for municipal liability. Holding taxpayers, as well as the beneficiaries of government, accountable for the faulty actions of government officials is fair despite the taxpayers’ and beneficiaries’ lack of responsibility for those acts. But when notice is lacking, it seems as unfair to impose liability on the ordinary citizens who finance and benefit from government as it is on the officer who unknowingly commits the violation. The facts of Owen are illustrative of such inequity. The City Council fired Owen, the chief of police, at a time when Supreme Court doctrine appeared to allow it to do so.²¹⁸ A few months later, the Supreme Court held, for the first time, that an employee like Owen may hold “property” and “liberty” interests protected by the Due Process Clause, and therefore may be entitled to a “due process” hearing in connection with dismissal.²¹⁹ The City of Independence violated Owen’s rights, but the City and those it represented did not have fair notice that it had done so. The fairness objection to liability is that, incentive effects aside, it is not fair to hold a government liable for an act, just as it is unfair to hold an individual officer liable, without notice that the act may violate constitutional norms. It is not fair because the government is nothing more than the sum of all the people who pay for it and receive services from it, and strict liability holds all of them liable despite the lack of notice that they, through their government, were doing anything wrong.²²⁰

III. A Categorical Approach

This Part of the Article describes and defends a middle ground between the Court’s broad pro-immunity rule and the abolitionist view—what I call the “categorical approach.” The foundations of the argument lie in the principles developed in Parts I and II. One premise is that fairness and utility favor an immunity defense, but not necessarily the current regime’s interpretation of the immunity doctrine, within which defendants win even if the unconstitutionality of their acts is clear. Another premise of the categorical approach is that the predictability and stability benefits of rules counsel against borrowing “ordinary tort” principles, but they do not necessarily require an immunity rule as broad as the rule of Harlow. The better path would be to retain the value of rule-based decision making by a categorical approach, in which distinctions would be drawn among fact patterns on the basis of one or a few features that justify putting them in a “no immunity” category or in a “high immunity” category. Qualified immunity is analogous to the categorical approach of First Amendment doctrine, in which some categories of speech receive a higher level of protection than others. In both areas, “[c]ategorization has the attraction of clarity and of providing guidance to judges and other government officials.”²²¹ Thus, the aim of making contextual distinctions is to carve out categories in which immunity would be denied, as well as categories in which the policies favoring immunity are powerful enough to justify a stronger, even an absolute, immunity.²²²

The categorical approach gives weight to the specific factual and legal context in which an immunity issue arises.²²³ The question one asks in setting up categories is whether the policies on either side of the cost-benefit equation are comparatively strong or weak in a particular context. Section III.A discusses categories of cases in which the cost-benefit case for qualified immunity is especially weak. Section III.B turns to fact patterns in which the cost-benefit calculus may be so one-sided in the other direction that it supports even the award of absolute immunity.

This Article’s claim is not that this categorical approach would resolve all of the current objections to qualified immunity. Some cases, for example Brown v. Elliott,²²⁴ do not fall within any of the special categories this Article identifies. As a result, Harlow would continue to govern those cases. In some fact patterns, issues would arise as to which, if any, of the categories may apply. Some fact patterns will raise hard issues about which category the case falls into. Some fact patterns may push for inclusion in both an expansive and a restrictive category. In these ways, the categorical approach adds complexity to the doctrine. But it would also provide useful guidance and structure to the analysis of qualified immunity issues. The benefits of a more fine-tuned approach are, on balance, worth the added complications.

A. Six Weak Immunity Categories

This Section describes six fact patterns in which the pro-immunity policies of fairness to defendants and avoiding overdeterrence are comparatively weak. In terms of the “proper balance” between the costs and benefits of immunity, this pattern of weakness does most of the work of justifying exceptions to Harlow. The anti-immunity goals of vindicating constitutional rights and deterring subsequent violations seem more or less constant across most of these categories.²²⁵ Thus, the relative strength of vindication and deterrence is greater in these fact patterns.

1. Collateral issues

Ziglar falls into a category in which the case for immunity is unconvincing. When the lack of clarity relates to an issue other than the constitutional or statutory validity of the officer’s conduct, the fairness and “social costs” arguments are weak, if they exist at all. The complaint in Ziglar alleged equal protection violations that were clearly established at the time the officials acted.²²⁶ They had the “fair warning” to which defendants are entitled, and their conduct did not put them to a choice between excessive caution and bold, effective government action.²²⁷ The uncertainty related only to whether the defendants could be sued under the conspiracy statute for conduct that they and everyone else know is unconstitutional.²²⁸ It is not far-fetched to compare immunity based on that lack of certainty to lack of certainty as to whether the plaintiff can obtain personal jurisdiction, or as to whether “state action” is present. The argument for a categorical rule that excludes a defense is that the cost-benefit balance favors the plaintiff’s side across the whole set of such “collateral issue” cases. Given the weak pro-immunity policies, the pro-plaintiff goals of vindication of rights and deterring constitutional violations should control the outcomes in this set of cases.

The facts of Hernandez v. Mesa,²²⁹ decided a week after Ziglar, illustrate the aforementioned situation where pro-plaintiff goals should control the outcome. In that case, Jesus Mesa, a United States border guard standing on the United States side of the Mexican border, shot and killed Jesus Hernandez, an unarmed fifteen-year-old Mexican boy, who was standing on the Mexican side of the border.²³⁰ The boy’s parents brought a Bivens suit, in which they alleged violations of his Fourth and Fifth Amendment rights.²³¹ In the per curiam opinion, the Court remanded for consideration of the Bivens issue in light of Ziglar.²³² Additionally, the Court reviewed the lower court’s ruling “that Mesa was entitled to qualified immunity,”²³³ on the Fifth Amendment claim because of lack of clarity as to the applicability of the Fifth Amendment to a cross-border shooting of a non-U.S. citizen.²³⁴

Uncertainty on that issue is collateral to the duties imposed on Mesa by the Fifth Amendment. The outcome should turn on whether the Fifth Amendment rules are sufficiently clear. The Fifth Amendment claim is that the shooting was a deprivation of Hernandez’s liberty without due process of law, in violation of the substantive component of due process.²³⁵ The leading case is Sacramento County v. Lewis.²³⁶ In that case, the Court denied liability because the injury was not intentional, but it also said that “conduct intended to injure in some way unjustifiable by any government interest is the sort of official action most likely to rise to the conscience-shocking level” that would offend substantive due process.²³⁷ Even if Fifth Amendment protections are less stringent at the border,²³⁸ that rule may be sufficiently clear and broad to enable Hernandez’s parents to overcome a qualified immunity defense, provided they can prove that the shooting was unprovoked.²³⁹ For present purposes, however, that issue can be set aside. The thrust of this Article’s argument is that the lack of clarity on which the lower court relied—as to whether the Fifth Amendment applies at all to cross-border shootings—is collateral to the fairness and avoidance of overdeterrence rationales for immunity.

2. The relevance of “bad faith”

Harlow held that officers are immune from damages unless they violate “clearly established statutory or constitutional rights of which a reasonable person would have known.”²⁴⁰ This canonical statement of the rule leaves open an important question: What if an officer believes he is violating the Constitution, and in fact he is violating the Constitution, even though a reasonable officer in his position would not have known? Two sentences in the majority opinion suggest that the officer cannot assert immunity if he “knew or reasonably should have known” his act violates the plaintiff’s constitutional right.²⁴¹ Justice Brennan, in a concurring opinion joined by Justices Marshall and Blackmun, cites those sentences in support of a rule that the Harlow “standard would not allow the official who actually knows that he was violating the law to escape liability.”²⁴² On the other hand, some aspects of Justice Powell’s opinion for the Court pointed in the other direction. Its main theme was that the bad faith prong increased the cost of constitutional tort litigation and must be eliminated, since it authorized a fact-intensive inquiry into the actual state of mind of the defendant.²⁴³ That inquiry made the discovery process especially burdensome and raised an obstacle to disposition of cases on summary judgment.²⁴⁴ Justice Brennan’s concurrence did nothing to allay these concerns, as he found it “inescapable . . . that some measure of discovery may sometimes be required to determine exactly what a public official did ‘know’ at the time of his actions.”²⁴⁵ The issue remained open after Harlow, as Justice Powell did not respond to Justice Brennan on this point.

Five years later, however, the Court definitively rejected Justice Brennan’s position in Anderson. The issue in this case was whether an FBI agent could be held liable for an illegal search, when an officer could have reasonably believed the search was valid, but plaintiffs could show that this defendant believed it violated the Fourth Amendment, as indeed it did. The Court said that Harlow had “completely reformulated qualified immunity . . . replacing the inquiry into subjective malice . . . with an objective inquiry into the legal reasonableness of the official action.”²⁴⁶ The officer’s “subjective beliefs about the search [were] irrelevant.”²⁴⁷ The Harlow ruling became a hardline rule that qualified immunity inquiries focus solely on the state of the law at the time of the constitutional violation, thus excluding evidence of a defendant’s impermissible motives or reckless disregard of a plaintiff’s constitutional rights.²⁴⁸ This rule blocks plaintiffs with valid constitutional claims from recovering from defendants who do not deserve protection.²⁴⁹

From the perspective of achieving the proper balance between the costs and benefits of immunity, this holding is doubtful. Justice Brennan’s alternative seems to accommodate the competing interests better than the Anderson approach. The pro-liability “vindication of rights” and “deterrence of violations” policies are at their strongest when the officer acts in bad faith, that is, without a sincere belief in the validity of his action.²⁵⁰ In cases where officers act in bad faith, the defendant has no standing to complain about unfairly imposed liability, even if the law is sufficiently unclear enough that others acting in good faith could have believed they were acting properly. The holding is consistent with Harlow’s concerns about discovery and trial costs.²⁵¹ Those concerns are valid, especially since the costs must be incurred even when the defendant has not acted in bad faith. But developments since Harlow should be considered in evaluating its current status. The concerns that gave rise to the deletion of the subjective prong may not be as great now as they were in 1982. A few years after Harlow, in Celotex Corp. v. Catrett,²⁵² the Court heightened the standard for defeating summary judgment.²⁵³ In particular, it held that a party opposing summary judgment must “make a showing sufficient to establish the existence of an element essential to that party’s case, and on which that party will bear the burden of proof at trial,” whether or not the summary judgment movant has produced an affidavit or other proof.²⁵⁴ After Celotex, a plaintiff charging bad faith cannot avoid summary judgment without offering proof at that stage of the case. Justice Kennedy has suggested that this development has “alleviated” the problem that had carried so much weight in Harlow.²⁵⁵ Qualified immunity is a judge-made rule. Acting in the common law tradition, the Court is fully authorized to modify it in the light of changed conditions.

Experience with Harlow suggests that, even if trial and discovery costs persist after Celotex, the Court’s rule is not effective at reducing those costs. As we have seen,²⁵⁶ Schwartz’s empirical study of § 1983 litigation found that just 3.2% of § 1983 cases were dismissed before trial on qualified immunity grounds.²⁵⁷ In an earlier study of Bivens cases, Alexander Reinert reported that qualified immunity led to dismissals of only two percent of them over a three year period.²⁵⁸ Professor Schwartz found that “[q]ualified immunity is raised infrequently before discovery begins.”²⁵⁹ Defendants prevailed on some claims at summary judgment, but “additional claims or defendants regularly remained and continued to expose government officials to the possibility of discovery and trial.”²⁶⁰ Evidently, plaintiffs’ lawyers responded to Anderson by drafting complaints that cannot be disposed of on summary judgment under the standards set forth in Celotex.²⁶¹ In addition, they may sue for prospective relief, or they may name local governments as defendants.²⁶² Qualified immunity is not a defense in either context.²⁶³

3. Two bites at the apple

In Anderson v. Creighton, the Court held that an officer could successfully assert qualified immunity for a search that violated the Fourth Amendment.²⁶⁴ The officer may argue, first, that the search was reasonable, and second, that he made a reasonable mistake as to the reasonableness of the search, even if the search was not reasonable. As Justice Stevens put it in dissent, he is allowed “two bites at the apple.”²⁶⁵ This rule was extended to all Fourth Amendment cases in Saucier v. Katz.²⁶⁶ The issue in that case was how qualified immunity should be applied when persons arrested by the police assert that the officers used excessive force.²⁶⁷ Graham v. Connor²⁶⁸ had held that the general Fourth Amendment test is whether the officer’s actions were “objectively reasonable.”²⁶⁹ The qualified immunity test is “whether it would be clear to a reasonable officer that his conduct was unlawful in the situation he confronted.”²⁷⁰ The issue in Saucier was whether the officer is entitled to argue “reasonableness” twice, or only once.²⁷¹ Thus, the plaintiff argued that Graham “sets forth an excessive force analysis indistinguishable from qualified immunity, rendering the separate immunity inquiry superfluous.”²⁷² But the Court, following Anderson, ruled that the two inquiries are distinct.²⁷³ Across the whole range of Fourth Amendment issues, the officer may prevail on either of two distinct grounds: (1) that the force, arrest, or search was “reasonable,” and therefore not a Fourth Amendment violation, or (2) that the force, arrest, or search was improper, but he made a reasonable mistake in determining the activity was permissible.²⁷⁴

The Court’s reasoning is unassailable insofar as it reflects the logical structure of a constitutional tort case.²⁷⁵ Since the substantive constitutional “reasonableness” issue and the immunity “reasonableness” issue are in fact distinct, the resolution of the constitutional issue cannot, by itself, resolve the immunity issue. Under the Fourth Amendment, unreasonableness of the conduct does not exclude an argument that the mistake as to the reasonableness of the conduct was itself a reasonable mistake.²⁷⁶ One criticism of this test is that, logic aside, it is hard for judges and juries to keep the two questions separate, at least in cases in which the Fourth Amendment and the immunity issues both seem to turn on “reasonableness.”²⁷⁷ For present purposes, the more important objection is that the “cumulation of messages . . . has led many lower courts to reject civil liability for excessive force in circumstances where such liability seems fully justified.”²⁷⁸

As a matter of costs and benefits, the case for “two bites at the apple” is strong in some contexts but weak in others. In Anderson’s case, the search was objectively reasonable given the information the officer had to establish probable cause and thus provides a strong example for the aforementioned approach. It is strong in Anderson’s “probable cause” fact pattern. The body of Fourth Amendment doctrine bearing on probable cause to search or to arrest includes a consideration of a myriad factors, including, for example, the reliability of informants, the type of crime, and the existence of “exigent circumstances” that would obviate the need to a warrant, among others.²⁷⁹ The evaluation of all of these factors by a police officer who does not have a law degree (or even one who does) may not yield a clear answer in the circumstances of a given search.²⁸⁰ If an officer gets it wrong, but makes a reasonable mistake, it is unfair to second-guess the decision, and it is likely to lead officers to steer clear of searches that are close to the constitutional line. Thus, both the fairness and “avoiding overdeterrence” policies are in play.

But the rationale for a second bite at the apple is weaker in excessive force cases, in which both the Fourth Amendment issue and the qualified immunity issue turn on “reasonableness.” If the goal of immunity is to achieve a “proper balance” between competing policies, the core problem with Saucier is not the confusion it creates, nor is it simply the impact it has on outcomes. These are symptoms of a more basic objection: to a substantial degree, showing sufficient lack of reasonableness satisfies demonstrating fairness to the officer in establishing a Fourth Amendment violation.²⁸¹ In an excessive force case like Saucier, that showing cannot be made unless the officer has not only invaded the plaintiff’s interests in bodily integrity and personal security, but has acted unreasonably in doing so. Fairness to the officer is largely achieved without proof of a second layer of unreasonableness, i.e., that the officer not only acted unreasonably in violation of the Fourth Amendment, but made an unreasonable mistake in thinking he was allowed to use force. As a matter of avoiding overdeterrence, the policy argument against official immunity is that the officer already has the protection he needs in order to use force, in the form of the Fourth Amendment requirement that the force be “unreasonable” in order to trigger Fourth Amendment protection. For these reasons, “qualified immunity would impart only a very slight addition to the protection built into the constitutional standard for excessive force.”²⁸²

This theme is illustrated by the Supreme Court’s award of immunity in Kisela v. Hughes.²⁸³ In that case, an officer shot at a woman who was carrying a knife because he believed she was threatening someone else.²⁸⁴ Suppose the belief was mistaken. If the mistaken belief were nonetheless reasonable, there would be no Fourth Amendment violation.²⁸⁵ But suppose the belief was unreasonable, so that the shooting was a violation of the woman’s Fourth Amendment rights. The officer still escapes liability if he can show that he reasonably believed the shooting was justified. In Kisela, the defense may yield a bit of “avoiding overdeterrence” benefit, and a bit more protection for officers against imposition of liability without fair warning. If the provision of an extra shield against liability were costless, the marginal benefit added by immunity might tip the scale in favor of a defense, even in “two bites at the apple” excessive force cases. But that marginal benefit comes at the expense of the pro-liability policies of vindication of constitutional rights and deterrence of violations. The case against Saucier is that in this “two bites at the apple” context, the benefits of a qualified immunity defense are comparatively small, while the costs are as great as they always are when plaintiffs are denied a victory on the merits on account of official immunity.

4. Unwarranted reliance on circuit precedent

In Lane v. Franks,²⁸⁶ the substantive issue arose when Edward Lane, a government employee, was fired for giving truthful testimony pursuant to a subpoena at a criminal trial.²⁸⁷ That criminal prosecution concerned a public corruption scandal that Lane discovered at work. The Eleventh Circuit had dismissed his § 1983 suit on the ground that Lane’s testimony was not protected speech, and that, even if it were protected speech, the constitutional rule was unclear at the time Lane was fired.²⁸⁸ On the First Amendment issue, the Court held unanimously that the testimony was protected speech.²⁸⁹ The leading cases in the area are Pickering v. Board of Education²⁹⁰ and Connick v. Myers.²⁹¹ Taken together, these cases held that an employee could not be fired for speech on a matter of public concern unless the disruptive impact of the speech outweighed its value.²⁹² As applied to Lane’s case, the testimony “clearly” met the “speech as a citizen on a matter of public concern” prong.²⁹³ Ever since Pickering, the Court has recognized that a public employee’s speech concerning his employment holds special value because of the insight gained regarding public concern through his profession.²⁹⁴ Moreover, public employee speech is evidently important in cases such as a public corruption scandal.²⁹⁵ Accordingly, holding that that the speech required to prosecute a public official for corruption may never be the basis for a retaliation claim under the First Amendment is “antithetical to our jurisprudence.”²⁹⁶ On the balance of interests prong, the employer’s interests on the Pickering scale are completely absent, as the employer neither asserted nor demonstrated any government interest to the contrary.²⁹⁷

However, this case also presents a qualified immunity issues as Lane sought damages in a § 1983 suit against the supervisors.²⁹⁸ Despite the unanimous holding that the testimony was protected by the First Amendment, the Court held, again unanimously, that the officials were immune.²⁹⁹ The stumbling block was Garcetti v. Ceballos,³⁰⁰ which had held that “when public employees make statements pursuant to their official duties, the employees are not speaking as citizens for First Amendment purposes.”³⁰¹ Ceballos, an assistant district attorney had been denied a promotion and had suffered other adverse actions for statements he made to other prosecutors and law enforcement officials, in which he found fault with an affidavit and urged dismissal of a criminal case.³⁰² This type of “pursuant to official responsibilities” statement, the Court held, was not covered by the Court’s precedent in Pickering and Connick.³⁰³ After Garcetti, the Eleventh Circuit applied the holding to situations in which the employee did not speak as part of the job, but rather about information learned on the job.³⁰⁴ On the First Amendment merits, the Court in Lane decisively rejected this reading of Garcetti because “[t]he sworn testimony in this case is far removed from the . . . internal memorandum [at issue in Garcetti].”³⁰⁵ For that matter, “Garcetti said nothing about speech that simply relates to public employment.”³⁰⁶ In a concurring opinion joined by Justices Scalia and Alito, Justice Thomas likewise gave short shrift to the Eleventh Circuit’s reasoning.³⁰⁷ He noted that Garcetti’s “unprotected speech” holding applied “when a public employee speaks ‘pursuant to’ his official duties.”³⁰⁸

Despite the Court’s repudiation of the Eleventh Circuit’s precedents, it held that the defendants were entitled to rely on them.³⁰⁹ The Court devoted a single sentence to explain why. It said that at the time of the firing, “no decision of this Court was sufficiently clear to cast doubt on the controlling Eleventh Circuit precedent.”³¹⁰ The Court did not attempt to reconcile this ruling with its earlier observation that “the Eleventh Circuit read Garcetti far too broadly,”³¹¹ or with Justice Thomas’s view that the substantive issue in Lane “require[d] little more than a straightforward application of Garcetti.”³¹² Instead, the Court merely recounted the holdings of the Eleventh Circuit cases and concluded that “[a]t the time of Lane’s termination, Eleventh Circuit precedent did not provide clear notice that subpoenaed testimony concerning information acquired through public employment is speech of a citizen entitled to First Amendment protection.”³¹³

The Court shifted its tone more than once. Having found, initially, that “the Eleventh Circuit read Garcetti far too broadly,” and then having found a lack of “clear notice,” the Court next asserted that “[t]here [was] no doubt that the Eleventh Circuit incorrectly concluded that Lane’s testimony was not entitled to First Amendment protection.”³¹⁴ Despite the lack of doubt, “the question was not ‘beyond debate’ at the time Franks, the supervisor who fired Lane, acted.”³¹⁵

The explanation for these equivocations is that the Court gives weight to circuit precedent in applying the “clearly established law” principle,³¹⁶ and does so even in the face of Supreme Court rulings to the contrary. Consequently, even if a circuit rule is “no doubt” mistaken, it will protect the officer from liability for damages. Officers are certainly entitled to look to the law of the circuit on matters the Supreme Court has not addressed. But the Court gives too much deference to the circuits in cases like Lane. The Supreme Court’s balancing test set forth by Pickering and Connick has roots in the fundamental First Amendment principle that speech bearing on self-government receives a high level of protection.³¹⁷ The application of that principle will occasionally present hard questions, but Lane’s case was not one of them. Few, if any, competent lawyers could have doubted that the First Amendment protected Lane’s truthful, subpoenaed trial testimony on public corruption. Despite the Eleventh Circuit precedent that read Garcetti broadly, that point was sufficiently clear to provide the notice necessary to satisfy the qualified immunity policies. Garcetti’s distinction between speech “pursuant to” official duties and other speech by government employees may not be sufficiently clear to mark the protected/unprotected border for every instance of government employee speech. But it is clear from Garcetti’s “pursuant to official duties” rationale that testimony compelled under subpoena in a criminal case is not covered by Garcetti’s holding.

Lane may belong to a category of cases in which reliance on circuit court precedent is unreasonable.³¹⁸ The Court should recognize an exception to Harlow for cases that fall into this category, if only to discourage reliance on Lane in situations in which Supreme Court precedent clearly establishes unconstitutionality but defendants can cite dubious lower court decisions.³¹⁹ Under the categorical approach, reliance on circuit court precedent would be unavailing in such a case no matter how strongly the circuit court cases support the defense. To put the rationale for this rule in cost-benefit terms: there is no occasion to accommodate the competing constitutional tort policies³²⁰ because there is no uncertainty at all as to the place of the Supreme Court at the top of the hierarchy of courts. Thus, the fairness concern is weak when the Supreme Court doctrine is sufficiently clear to provide “fair warning”³²¹ even if circuit precedent is to the contrary. For the same reason, a clear Supreme Court ruling eliminates the gray area that sometimes produces overly cautious government action, and it does so even if circuit court precedent seems to allow the officer to go further. In this type of case, the plaintiff-friendly policies of vindication of rights and deterrence of violations should control the outcome.

5. Distinctions without differences

In many constitutional tort cases, the “clearly established law” issue is more accurately described as “clearly established application of law to fact,” because “the right cannot be defined at a high level of generality; instead, the key is whether the specific conduct has been clearly established as a constitutional violation.”³²² For example, the Fourth Amendment authorizes police officers to use “reasonable” force to subdue suspects.³²³ In “excessive force” cases, the issue is whether the officer has gone too far.³²⁴ Since the reasonableness of the use of force depends on the circumstances of each particular encounter, and since the law developed in connection with one circumstance may not clearly establish the law for another circumstance, officers often have grounds for an assertion of qualified immunity.³²⁵ The strength of those grounds will vary from one case to the next because some factual variations have legal significance while others do not. The Supreme Court’s formulation of the qualified immunity rule invites courts to tilt the analysis in favor of immunity in close cases. It requires that “every reasonable official would have understood that what he is doing violates [the] right,”³²⁶ and that “existing precedent must have placed the . . . question beyond debate.”³²⁷

Some courts award qualified immunity even when the legal or factual distinction is slight. For example, in A.M. v. Holmes,³²⁸ a thirteen-year-old seventh grader disrupted class by burping, laughing, and refusing to stop when told to do so.³²⁹ His classmates were entertained but the teacher was not.³³⁰ She called the school resource officer—a police officer assigned to the school—who arrested the boy, put him in handcuffs, and took him to a juvenile detention facility.³³¹ The boy and his parents sued under § 1983 for a Fourth Amendment violation but a divided Tenth Circuit panel granted the officer qualified immunity.³³² Though earlier cases had held that an arrest would be improper in similar circumstances, the majority distinguished those cases, as they involved a different criminal statute.³³³ In his dissent, then-Judge Gorsuch pointed out that “the relevant language of the two statutes is identical.”³³⁴ Similarly in Coffin v. Brandau,³³⁵ a police officer came to the plaintiff’s house with a summons but without an arrest or search warrant.³³⁶ The law was clear that he could not enter the front door without a warrant or consent.³³⁷ However, he entered the door of the attached garage instead.³³⁸ The Eleventh Circuit held that the entry violated the Fourth Amendment but granted qualified immunity anyway, on the ground that under Fourth Amendment principles a reasonable person might distinguish a garage door from a front door.³³⁹ Lastly in Gilmore v. Hodges,³⁴⁰ a hearing-impaired prisoner sued for failure to provide him with batteries for his hearing aids.³⁴¹ Although Eighth Amendment precedents clearly established that denial of dentures, eye glasses, and prostheses amounted to cruel and unusual punishment, the Eleventh Circuit found no clearly established rule against denial of hearing aids.³⁴²

While pro-immunity values are comparatively weak across cases that differ from earlier plaintiff-friendly rulings, the constitutional norms remain the same and therefore do not support a finding of surprise.³⁴³ The Supreme Court could recognize this systematic imbalance by adopting a rule that immunity should not be awarded when there is a factual distinction but the distinction does not reflect a legally significant difference.³⁴⁴ In Hope v. Pelzer, it seemed to do just that. An earlier binding case had held that chaining a prisoner to a fence for a long period of time as punishment would violate the Eighth Amendment right against cruel and unusual punishment.³⁴⁵ In Hope, prison guards had chained the prisoner to a hitching post, and the Eleventh Circuit applied qualified immunity on the theory that precedent cases did not clearly establish an Eighth Amendment violation for chaining a prisoner to a hitching post.³⁴⁶ The Supreme Court rejected the Eleventh Circuit’s “rigid overreliance on factual similarity”³⁴⁷ and held that the key issue is “fair warning.”³⁴⁸ But the Court soon turned away from the plaintiff-friendly approach it had taken in Hope. Its “subsequent decisions veered back toward requiring precedential specificity.”³⁴⁹

“Fair warning” is too vague a concept to guide and constrain decision making. Its content depends on a decision maker’s conception of fairness, which can depend on a wide array of contextual factors. “Distinctions without differences” is more rule-like, because it directs courts to deny qualified immunity unless they are persuaded that a later case differs in a legally significant way from the earlier one. The retreat from Hope might have been resisted more effectively if the Court had articulated a “distinctions without difference” rule instead of putting all of its weight on the norm of “fair notice.” As Frederick Schauer has pointed out, “rule-bound decisionmaking is inherently . . . conservative, in the nonpolitical sense of the word.”³⁵⁰ Judges inclined to stray would have at least been obliged to advance arguments as to why the rule should not apply to the case at hand.

This is a weaker constraint than the first four, because it does not identify a circumstance that would resolve the immunity issue by itself. It may not compel different outcomes in cases like Kisela, A.M., Coffin, and Gilmore, but it would at least furnish the plaintiff’s lawyer a means for drawing the court’s attention to the relevant question. Kisela, A.M., Coffin, and Gilmore are cases in which the prior law lays down a clear constitutional line, which provides both fair warning and the guidance needed to allay undue caution. In cases of this type, courts may have to decide, based on the facts of that particular case, whether a given distinction actually reflects a difference. But the articulation of a category still has value, because it provides crucial guidance to the adjudicator.³⁵¹ It focuses the argument on the legally significant feature of the situation. For example, in a case like Coffin, the officer may argue that a garage door is not really like a front door, because the garage is not central to the dwelling. The plaintiff’s answer to this line of reasoning is that the settled law on front doors gives the fair warning to which the officer is entitled, as there is no legally significant difference, for Fourth Amendment purposes, between front doors and attached garage doors. The rule would direct the court to concentrate on that issue, and to refrain from speculation of the kind invited by Ashcroft v. Al-Kidd³⁵² and Reichle v. Howards³⁵³ as to whether “every” officer would have understood that the entry violated the Fourth Amendment, or whether the issue is “beyond debate.” The “distinctions without differences” rule would remind the court that the immunity issue is not whether a distinction can be drawn but whether the distinction is warranted by the fair warning goal of official immunity. The mere availability of a front door/garage door distinction does not cast doubt on the conclusion that the officer had sufficient notice to satisfy the fairness and social costs goals served by qualified immunity.

6. Ineluctable inferences

Closely related to “distinctions without differences” is the situation in which well-settled general principles support liability. But there is no prior case with similar facts, and thus no occasion for the court to draw a distinction, even one that does not reflect a legally significant difference. Many courts tend to grant qualified immunity in this type of case.³⁵⁴ For example, in Robles v. Prince George’s County,³⁵⁵ the Fourth Circuit granted immunity to police officers who left an arrestee tied to a pole in a parking lot in the middle of the night when they could not find a place to take him.³⁵⁶ The court acknowledged that earlier cases involving unjustified physical restraints on pretrial detainees put the officers on notice that “they were acting inappropriately.”³⁵⁷ Unfortunately for Robles, none of them involved tying someone to a pole.³⁵⁸ Thus, for the Fourth Circuit, the officers’ Fourth Amendment violation was not “clearly established.”³⁵⁹ Professor Jeffries cites this case as an illustration of the current misalignment between the immunity policies and the application of Harlow.³⁶⁰ The specific problem it illustrates is that, “[i]n searching for clearly established law, courts seek precedents similar on the facts.”³⁶¹ They do this despite the Supreme Court’s admonition in Hope that “officials can still be on notice that their conduct violates established law even in novel factual circumstances.”³⁶²

An advantage of a categorical approach is that the articulation of a rule to that effect would elevate the status of inference-drawing. Assuming Robles is the typical case, the question of whether there is fair warning is often ignored in practice, but nonetheless appears to be the question lower courts ought to ask.³⁶³ The creation of a special category would heighten its salience. As with “distinctions without differences” it would channel the debate over immunity more effectively than the current broad rule. The plaintiff’s lawyer would be enabled to call the court’s attention to the “ineluctable inferences” rule and the officer would lose unless he has a convincing rebuttal.

Consider Mirabella v. Villard³⁶⁴ as an illustration of how this rule may influence qualified immunity litigation. Maureen and John Mirabella had quarreled with a neighbor over the neighbor’s use of property that belonged to Montgomery Township.³⁶⁵ After complaining to the local authorities without success, the Mirabellas sent the Board of Supervisors an email, informing them that they would sue the neighbor and would name the Township as an indispensable party.³⁶⁶ Joseph Walsh, Chairperson of the Board of Supervisors, responded with an email in which he told the Mirabellas to have no further contact with him or other Township employees, on any topic or for any reason.³⁶⁷ The Mirabellas then sued Walsh and other officials under § 1983.³⁶⁸ They alleged, among other things, that Walsh’s email amounted to retaliation for the exercise of their First Amendment rights.³⁶⁹ The defendants did “not dispute that the Mirabellas . . . engaged in constitutionally protected speech . . . when they protested the Township’s failure to protect the open space and threatened litigation.”³⁷⁰ Nor was there any dispute over the settled rule that retaliation for the exercise of a constitutional right is a constitutional violation.³⁷¹ Furthermore, the “no contact” email, which “barred the Mirabellas from communicating directly with their local government, for any reason, . . . was retaliatory.”³⁷²

Nonetheless, the Mirabellas lost on qualified immunity grounds.³⁷³ The court could not find any cases that clearly established the “right to be free from a retaliatory restriction on communication with one’s government, when the plaintiff has threatened or engaged in litigation against the government.”³⁷⁴ Under the “ineluctable inference” rule, the inquiry would focus on fair warning. Failure to find a case on point would not be fatal to the plaintiff’s case.³⁷⁵ In light of the clearly established right against retaliation for the exercise of the clearly established free speech right, the issue would be whether those principles gave Walsh fair warning that his email was retaliatory. Walsh may still have a good defense, but he will not have a good defense just because the Mirabellas cannot cite a prior case.

B. Two Strong Immunity Categories

The logic of the cost-benefit model implies that it may, in some circumstances, favor broader, rather than narrower, official immunity. In the two contexts discussed in this section, the pro-defendant “avoiding overdeterrence” and “fairness to defendants” interests are especially strong, and thus may justify absolute immunity for officials, rather than Harlow’s qualified immunity. The two contexts discussed here involve cases in which: (1) high officials are sued for actions related to national security; and (2) cases in which the plaintiff obtains prospective relief. As with the six “anti-immunity” categories, my aim is to show that there are grounds for distinctive treatment of these fact patterns. Whether those grounds are strong enough is a separate question, and one for the Supreme Court to resolve. In any event, the Court, with the aid of history, has already identified some categories in which the cost-benefit calculation supports an absolute immunity, namely those in which the actor exercises a prosecutorial, judicial, or legislative function,³⁷⁶ as well as official actions taken by the President of the United States.³⁷⁷ Thus, the issue explored in this Section is whether the national security and prospective relief contexts are sufficiently analogous to these previously established categories to warrant similar treatment.

1. Absolute immunity for “executive officers” in national security cases

In Harlow, the Court left open the possibility that it would make an exception to its blanket qualified immunity rule for executive officers, including Presidential aides.³⁷⁸ It said that “[f]or aides entrusted with discretionary authority in such sensitive areas as national security or foreign policy, absolute immunity might well be justified to protect the unhesitating performance of functions vital to the national interest.”³⁷⁹ Since Harlow, that question has remained open in the development of qualified immunity doctrine. In Ziglar, the Court returned to this theme, but it did so in a somewhat different context. It took the step of giving three of the defendants a special label, calling them “Executive Officials.”³⁸⁰ These included former Attorney General John Ashcroft, former FBI Director Robert Mueller, and former Immigration and Naturalization Service Commissioner James Ziglar.³⁸¹ In explaining its rejection of a Bivens suit, the Court emphasized its role in high-level policy making. The special costs of authorizing a damages cause of action in this context included “inquiry and discovery into the whole course of the discussions and deliberations that led to the policies and governmental acts being challenged.”³⁸² Courts would be required “to interfere in an intrusive way with sensitive functions of the Executive Branch.”³⁸³ In this particular case, Abbasi’s complaint related to “sensitive issues of national security,” which are “the prerogative of Congress and the President.”³⁸⁴ When Justice Kennedy turned to the qualified immunity issue raised by the § 1985(3) claim, he referenced this earlier “Executive Officials” discussion and added that “open discussion among federal officers is to be encouraged, so that they can reach consensus on the policies a department of the Federal Government should pursue.”³⁸⁵ If private suits for damages were allowed in this context, “the result would be to chill the interchange and discourse that is necessary for the adoption and implementation of governmental policies.”³⁸⁶ As the Court cautioned, “[t]he risk of personal damages liability is more likely to cause an official to second-guess difficult but necessary decisions concerning national security policy.”³⁸⁷

In its opinion, the Court raised this concern to support the denial of a Bivens cause of action on most of the claims, as well as the award of qualified immunity on the § 1985(3) claim.³⁸⁸ The logic of this rationale may not be so limited. The Court’s reasoning, with its emphasis on the distinctive disadvantages of allowing suits against high level “Executive Officials” in national security cases, may have implications beyond the Bivens and § 1985(3) contexts.³⁸⁹ Instead of qualified immunity, the cost-benefit calculus underlying official immunity doctrine may justify absolute immunity, at least for some acts by Executive Officials. Both as a matter of fairness and effective government, the costs of liability may be especially great in this context.

The Court’s reasoning on the § 1985(3) claim seems to treat the “Executive Officials” theme as a freestanding ground for immunity, separate from the “lack of clarity on intrabranch conspiracies” prong of the analysis. The “Executive Officials” prong makes no reference to lack of clarity in the law, and thus suggests that this concern exists even if the legal doctrine is clearly established. In that event, the general Harlow rule would be unhelpful in protecting high officers from personal liability, yet the pro-immunity case for “Executive Officials and national security” would still be strong. The Court does not call it an absolute immunity, but as a functional matter it seems to be accurately characterized as one. In the Court’s absolute immunity doctrine, the key issue is not whether the law is clearly established, but whether the officer’s function is one that needs special protection. Ziglar’s rationale for the immunity of executive officers in national security decision-making may be best understood as a ruling that high-level executive policy making is one of those functions.

The need to shield federal “Executive Officials” from liability for damages for some aspects of high level policy making may well apply to actions by state officials as well. The Court’s anti-Bivens holding does not address that concern, as that holding has no force in suits against state officers. Since those officers can be sued pursuant to a statutory cause of action under § 1983, the whole Bivens doctrine is irrelevant in such actions. If the “high level national security policy making” objection to liability is valid in the context of suits against federal officials, it should be presumptively available in suits against state officials as well, provided they can show that the activity in question is sufficiently similar to national security decision making. As the Court pointed out in another context, “[t]he pressures and uncertainties facing decisionmakers in state government are little if at all different from those affecting federal officials.”³⁹⁰ Immunity doctrine, not Bivens doctrine, is the means by which that principle can be implemented.

Should the Court recognize this absolute immunity rule, a question would arise as to its scope: would it apply only to national security decisions and similar decisions by high ranking state officers, or would it cover all high-level decisions? The Court’s language in Ziglar, encouraging wide-reaching immunity for high ranking government officials dealing with sensitive issues of national security, suggests a broader meaning. But two Supreme Court cases hold otherwise. In Scheuer v. Rhodes, the Court rejected absolute immunity for a state governor for decisions made in connection with deploying the National Guard to suppress campus unrest.³⁹¹ Citing Scheuer, the Court rejected absolute immunity for the Secretary of Agriculture in Butz v. Economou.³⁹² The Court explained that absolute immunity “would seriously erode the protection provided by basic constitutional guarantees.”³⁹³ The cost of immunity is especially great when the official holds a high rank because “the greater power of such officials affords a greater potential for a regime of lawless conduct.”³⁹⁴ Butz and Scheuer are solid precedents that courts would need to be overcome if they do not want to confine the hypothesized absolute immunity to the national security context. In my view, Butz and Scheuer were rightly decided. Thus, the immunity doctrine suggested by Ziglar’s reasoning should be confined to an absolute immunity for national security matters and not allowed to spill over into other forms of high-level decision making.

2. Vindication and deterrence by prospective relief

In some constitutional tort suits the vindication and deterrence benefits of liability are comparatively small, because those goals can be obtained, at least to some extent, without an award of damages. In this set of fact patterns, the Lyons rule will not preclude prospective relief,³⁹⁵ because the violation is continuing or a recurrence is sufficiently likely.³⁹⁶ The plaintiff who succeeds on the merits obtains an especially powerful remedy—an injunction or a declaratory judgment that should put a stop to the illegality, rather than merely putting a price on it.³⁹⁷ That ruling will not only affirm the plaintiff’s rights but also deter future constitutional violations. “Defendants will likely be better deterred from continuing to commit constitutional violations through the use of injunctions or declaratory judgments, as opposed to the damages remedy, because such judgments can be enforced without bringing entirely new lawsuits. An example of this is through contempt judgments against officials who do not comply.³⁹⁸

Vindication of rights is compromised when damages are denied, as they are when immunity is successfully asserted. But the Court’s rules on damages do not favor large awards in any event. Those rules follow the common law model. Under them, plaintiffs must prove damages, and the Court has explicitly rejected jury instructions that would authorize recovery for the value of the constitutional right.³⁹⁹ Since denials of constitutional rights often produce few tangible losses, few plaintiffs are able to obtain very large awards.