Basics of qualified immunity
History of Section 1983 and qualified immunity
As federal statutes go, Section 1983 is pretty straightforward, once you parse the nineteenth century legalese. The operative language of the statute says:
Every person who, under color of any statute, ordinance, regulation, custom, or usage, of any State or Territory or the District of Columbia, subjects, or causes to be subjected, any citizen of the United States or other person within the jurisdiction thereof to the deprivation of any rights, privileges, or immunities secured by the Constitution and laws, shall be liable to the party injured in an action at law, suit in equity, or other proper proceeding for redress
The phrase “under color of [state law]” just means everyone acting under state authority—in other words, all government agents. So the statute says, in essence, that any state actor who causes the violation of someone’s federally protected rights—either constitutional rights or federal statutory rights—“shall be liable to the party injured.” Note that this language is clear and unequivocal, and doesn’t mention any immunities whatsoever, qualified or otherwise.
The Supreme Court has offered a few different legal rationales for qualified immunity, but by far the most prevalent is the idea that nineteenth century common law (i.e., general principles of law announced through judicial decisions, rather than statutes) permitted such a defense for government agents. Chief Justice Roberts stated that “[a]t common law, government actors were afforded certain protections from liability,” and that “the [qualified immunity] inquiry begins with the common law as it existed when Congress passed § 1983 in 1871.” Filarsky v. Delia, 566 U.S. 377, 383-84 (2012).
To be clear, there’s nothing wrong with this sort of argument in principle. For example, suppose a legislature passed a statute that said: “Any person who willfully discharges a firearm at another shall be guilty of a felony, and imprisoned for not more than 10 years.” This statute doesn’t say anything about an exception for self-defense, but self-defense is such a well-established part of our legal tradition that courts would probably assume the legislature did not intended to negate that defense. So the relevant question we need to ask is whether there were, in fact, well-established common-law defenses for government agents at the time when Section 1983 was first passed.
Professor Baude’s article discusses this in more detail, but two Supreme Court cases are especially helpful in understanding this point. The first is an 1804 case called Little v. Barreme. This case involved a claim against an American naval captain—George Little—who had captured a Danish ship in 1799, during the Quasi-War with France. Federal law at the time authorized seizure of ships going to a French port (which this ship was not), but President Adams had issued broader instructions to also seize ships coming from French ports. The question was whether Captain Little’s reliance on these instructions was a defense against liability for the unlawful seizure.
The Court’s decision in Little makes clear that it seriously considered, but ultimately rejected, the very rationales that would come to support the doctrine of qualified immunity. Chief Justice Marshall explained that “the first bias of my mind was very strong in favour of the opinion that though the instructions of the executive could not give a right, they might yet excuse from damages”—that is, even though the captain broke the law, maybe he should be immune from damages. Nevertheless, the Court found him liable, and held that “the instructions cannot change the nature of the transaction, or legalize an act which without those instructions would have been a plain trespass.” In other words, the officer’s only defense was legality, not good faith.
The second is a 1915 case called Myers v. Anderson, which involved Section 1983 itself. In this case, the Court held that a Maryland statute violated the Fifteenth Amendment’s ban on racial discrimination in voting. The defendants (state election officers) then argued that, even if the statute was unconstitutional, they should not be liable for money damages, because they acted on a good-faith belief that the statute was lawful. The Court noted that “[t]he non-liability . . . of the election officers for their official conduct is seriously pressed in argument,” but it ultimately rejected any such good-faith defense. While the Myers Court did not elaborate much on this point, the lower court decision it affirmed was much more explicit:
[A]ny state law commanding such deprivation or abridgment [of constitutional rights] is nugatory and not to be obeyed by any one; and any one who does enforce it does so at his known peril and is made liable to an action for damages by the simple act of enforcing a void law to the injury of the plaintiff in the suit, and no allegation of malice need be alleged or proved.
In other words, half a century before it even invented qualified immunity, the Supreme Court had affirmed that there was no good-faith defense to suits under Section 1983. If state agents violated your constitutional rights, they were liable for any injury, period.
Yes, but only in limited circumstances. There were certain kinds of cases where a lack of “good faith” was part of what you needed to prove to win a lawsuit. For example, suppose you tried to sue a sheriff who arrested you for a crime you didn’t commit. If the sheriff had probable cause to believe you committed the crime—whether or not you actually did—then the sheriff would have a defense to your lawsuit. Another way to say this is that good faith and probable cause is a defense to a claim of “false arrest.”
But this kind of good-faith defense is very different from qualified immunity. In the example above, the lack of good faith is an element of a particular tort (in this case, false arrest). If the sheriff was acting in good faith, then he wasn’t breaking the law in the first place. But qualified immunity is an across-the-board defense. It applies to every public official, in every kind of case. When a defendant receives qualified immunity, it doesn’t mean they were acting lawfully; it just means that they can’t be held liable for damages. And there was definitely nothing like this across-the-board defense at common law.
The doctrine of qualified immunity has changed substantially over the years, but it was first articulated in a 1967 Supreme Court case called Pierson v. Ray. That case involved a Section 1983 suit against police officers who had arrested several people under an anti-loitering statute ultimately found to violate the First Amendment. The Supreme Court held that because the common-law tort of false arrest allowed the defense of “good faith and probable cause,” defendants should have that same defense in an “analogous” suit under Section 1983.
Note that Pierson presents exactly the same kind of issue as Myers v. Anderson—both involved state officials who violated individuals’ rights by enforcing unconstitutional statutes. But the Court flipped positions in Pierson, and allowed for the first time a good-faith defense to liability under Section 1983, without even acknowledging that they were doing so.
After Pierson, it looked like the Supreme Court might be charting a narrow course for qualified immunity, limiting it to a subjective, good-faith defense in cases in which the civil rights lawsuit resembled a common-law tort with a good-faith defense—like in the case of false arrest. But the Court soon expanded the doctrine beyond both of these limitations. First, in a 1974 case called Scheur v. Rhodes, the Court abandoned the analogy to common-law torts with a good-faith defense, and essentially held that qualified immunity would apply to all suits under Section 1983.
Second, in a 1982 case called Harlow v. Fitzgerald, the Supreme Court said that what mattered was not the defendant’s actual good-faith belief in the legality of their action, but “the objective reasonableness of an official’s conduct, as measured by reference to clearly established law.” In other words, qualified immunity would be available in all cases, to all defendants, unless the rights they violated were “clearly established.” The Supreme Court has continued to tighten this standard over the years, and clarified that qualified immunity protects “all but the plainly incompetent or those who knowingly violate the law.”
Modern operation of qualified immunity
Courts today generally characterize qualified immunity as a two-step inquiry. First, has the plaintiff established a violation of his or her federally protected rights? And second, were those rights “clearly established” at the time of the violation? If the answer to either of those questions is “no,” then the defendant receives qualified immunity, and the lawsuit against them cannot proceed.
If qualified immunity is resolved at step one, this simply means the defendant did not, in fact, violate the individual’s rights in the first place. This is just the merits question, and courts would have to answer it even if we didn’t have qualified immunity. Where qualified immunity is really doing the work is step two—the “clearly established law” inquiry. If the court resolves qualified immunity at step two, it generally means that the defendant was breaking the law, but that, in the court’s view, the law wasn’t sufficiently clear to justify holding the defendant liable.
This is the million-dollar question, and there’s no clear way to answer it. Since the Supreme Court first announced the “clearly established law” standard in 1982, it has taken dozens of qualified immunity cases to try to clarify the issue, but lower courts still remain quite confused. The Court has repeatedly emphasized that lower courts should not define “clearly established law” at a high level of generality; in other words, courts cannot say “it’s clearly established that you have a right to be free of unreasonable searches and seizures, and this was an unreasonable search, so no qualified immunity.” Rather, prior case law must define the right in a “particularized” manner.
But how “particularized” do the prior cases have to be? The theoretical problem is that all cases have factual distinctions, so how do you decide which ones matter, and which ones don’t? In practice, courts tend to treat this as an exacting standard, but there’s frequent disagreement within and between courts on exactly how to draw this line.
For example, consider Latits v. Phillips, a Sixth Circuit case from 2017. On June 24, 2010, Laszlo Latits was stopped by police for turning the wrong way on a divided boulevard. The dashboard camera showed the officer shining his flashlight into the car and raising his gun to Latits’s head; Latits then drove away, and the police pursued. Another officer, Lowell Phillips, repeatedly rammed Latits’ car, and eventually drove Latits off the road. Phillips then jumped out of his car, ran toward Latits, and shot him three times in the chest, killing him.
The Sixth Circuit unanimously held that Phillips’s use of deadly force violated Latits’s Fourth Amendment rights, because “considering the totality of the circumstances in the light depicted by the video . . . Latits did not present an imminent or ongoing danger” and therefore “the shooting was not objectively reasonable.”
But a divided panel of the court nevertheless found that Phillips was entitled to qualified immunity. Even though there were several cases specifically holding that it was unconstitutional to use deadly force on a fleeing driver who posed no ongoing threat, the court found that the law was not “clearly established,” because those prior cases “did not involve many of the keys facts in this case, such as car chases on open roads and collisions between the suspect and police cars.” The dissenting judge noted that “[i]t is a truism that every case is distinguishable from every other,” but that “the degree of factual similarity that the majority’s approach requires is probably impossible for any plaintiff to meet.”
A 2015 study examining 844 federal appellate qualified immunity decisions from 2009-2012 (encompassing 1460 distinct claims), found that qualified immunity was granted for 72% of the claims. See Aaron L. Nielson & Christopher J. Walker, The New Qualified Immunity, 89 S. Cal. L. Rev. 1 (2015). Since 2012, the Supreme Court has continued to tighten up on the doctrine, deciding almost every qualified immunity case against civil rights plaintiffs and encouraging lower courts to define “clearly established law” in a more particularized manner. It therefore stands to reason that qualified immunity would be granted even more frequently today.
With respect to district courts, it’s difficult to give a precise figure, but a 2017 study by Joanna Schwartz provides some guidance. See Joanna C. Schwartz, How Qualified Immunity Fails, 127 Yale L.J. 1 (2017). Schwartz looked at five particular districts, and analyzed qualified immunity motions, rather than qualified immunity decisions. Of these motions (i.e., requests by defendants to dismiss the case on the basis of qualified immunity), she found that:
- 31.6% were denied (meaning the case was allowed to proceed),
- 29.3% were granted on the basis of qualified immunity (the case was dismissed),
- 22.0% were granted on other grounds, or granted qualified immunity in the alternative, and
- 17.0% were unresolved (usually, because the case settled before a decision was issued)
In other words, district courts firmly rejected qualified immunity less than a third of the time. Of the last two categories—the 22.0% and 17.0%—it’s unclear exactly how to count them in answering the more general question of how often courts grant immunity. But in light of the frequent rate at which qualified immunity is granted on appeal—and the incentive that qualified immunity has to induce settlements—it seems reasonable to assume that a major chunk of these 39.0% of cases would have resulted in a grant of immunity, if the question had been squarely resolved.
No. At one point, the Supreme Court did require lower courts to make such a determination. In 2001, in Saucier v. Katz, the Court held that lower courts must resolve the two-step qualified immunity question sequentially—that is, first determine whether a legal violation occurred, and only then determine whether the law was “clearly established.” While that protocol still preserved the worst aspects of qualified immunity (i.e., denying justice to victims whose rights were violated), it at least had the virtue of encouraging development in the law for future cases.
But in 2008, in Pearson v. Callahan, the Court changed course, reversed Saucier, and gave lower courts discretion to resolve qualified immunity questions on the “clearly established law” prong, without deciding the fundamental question of whether the conduct was actually lawful. In other words, a court can say “whether or not this conduct was constitutional, illegality was not clearly established, so no liability.” There is, of course, a question-begging nature to this practice: if courts refuse to resolve legal claims because the law was not clearly established, then the law will never become clearly established.
Responses to arguments in favor of qualified immunity
First, even if qualified immunity could be defended on this basis, it’s worth noting what an incredible double standard this represents between government agents on the one hand, and ordinary citizens on the other. For regular people, it’s a well-known legal maxim that “ignorance of the law is no excuse.” Even in cases with serious criminal penalties, courts routinely permit the prosecution and conviction of defendants who had no idea they were breaking the law. If anything you would expect law enforcement—public officials specifically charged with knowing and enforcing the law—to be held to a higher standard of care than ordinary citizens. But in fact, they’re held to a far lower standard. Ignorance of the law is no excuse—unless you wear a badge.
Nevertheless, putting aside the double standard, this concern seems reasonable in the abstract. And if qualified immunity were limited to instances where public officials were genuinely acting in good faith, or were relying on judicial precedent specifically authorizing their conduct (even if courts later revised that precedent), then it would not be as serious a problem—though it would still be inconsistent with the relevant legal history and the plain terms of Section 1983.
However, qualified immunity is not at all limited to such sympathetic circumstances. A cursory glance at the case law reveals that it frequently is used to shield defendants who commit even egregious misconduct—especially unnecessary and unlawful police shootings. Defendants in these cases are not excused from liability because they were reasonably acting in good faith, but because there just didn’t happen to be a particular prior case in the relevant jurisdiction with functionally similar facts. Whatever the abstract merits of the argument above, it just doesn’t reflect the reality of most qualified immunity litigation.
Finally, even if this were a reasonable policy argument, it’s not appropriate for the courts to impose such a policy by judicial fiat. The intent of Congress—as clearly expressed in the plain terms of Section 1983—was to hold state actors liable when they violate people’s constitutional rights. If there are good arguments for modifying or limiting that liability, it is for the people’s legislative representatives to weigh and consider them, not the courts.
The risk of “chilling” public officials in the discharge of their duty—especially in the law enforcement context—is one of the primary policy rationales used to defend qualified immunity. But the whole point behind civil liability is that we want public officials to expect to face legal consequences when they violate constitutional rights. Section 1983 is intended not just to provide redress to individual victims, but to ensure accountability at a structural level. Another way of rephrasing the “chilling” concern is that civil liability may encourage police officers to think twice before taking actions on the edge of violating constitutional liberties. That’s a feature, not a bug, of our civil rights laws.
Police officers do indeed have to operate under dangerous, uncertain, and evolving conditions—which is exactly why our legal standards for determining whether a constitutional violation occurred in the first place are incredibly deferential to on-the-spot police decision-making. If a police officer reasonably believes, in light of the information available to them at the time, that there’s probable cause for an arrest, or that the use of force is necessary in a particular circumstances, then that officer has not committed a constitutional violation in the first place—even if that judgment later turns out to be wrong!
The Fourth Amendment—which governs most on-the-street, police-citizen interactions—prohibits “unreasonable searches and seizures.” This “reasonableness” standard therefore already accounts for the kind of practical deference that police need to do their job effectively. Qualified immunity only comes into play when an officer has acted objectively unreasonably under all the circumstances. In that situation, qualified immunity amounts to an unnecessary and inappropriate “double counting” of this deference.
This is a reasonable policy concern, but in practice, it’s basically a non-issue—because almost all public officials (especially members of law enforcement) are indemnified for their alleged misconduct. That is, even when a public official is sued in their individual capacity, their state employers will almost always pay for their legal defense, and cover the cost of any judgment against them. In a recent study, Joanna Schwartz demonstrated that governments paid approximately 99.98% of all dollars that civil rights plaintiffs recovered in lawsuits against police officers. Joanna Schwartz, Police Indemnification, 89 N.Y.U. L. Rev. 885 (2014).
Therefore, the primary immediate effect of eliminating qualified immunity would not be to subject individual defendants to massive personal liability, but rather to ensure that victims of unconstitutional misconduct obtain a remedy.
This is also a reasonable concern. There are complicated policy questions at play in determining the best overall system for ensuring accountability for law enforcement. A more rational system would probably not just be automatic indemnification, but rather a requirement that law enforcement officers—like basically all other professionals whose jobs entail a risk to the public—carry liability insurance. That approach would have the benefits of (1) ensuring that there will be funds to compensate victims, (2) protecting individual defendants from ruinous judgments, and (3) pricing out those particular officers who regularly commit constitutional violations, by making them uninsurable.
Of course, this is just one proposal, and there’s a larger discussion to be had about the optimal way of structuring such a system. But until and unless qualified immunity is eliminated (or at least, substantially revised), there’s no real incentive for state and municipal governments to even consider these ideas.
“Stare decisis”—Latin for “to stand by things decided”—is the general principle that judicial decisions should be guided by precedent, and that past decisions should not be lightly overruled, even when there’s reason to doubt their correctness. But while stare decisis is an important idea, the Supreme Court has always made clear that it is not an “inexorable command,” and that it can give way to countervailing considerations. And there are especially strong reasons for the Court to reconsider its qualified immunity jurisprudence.
First, the primary justification for stare decisis is that everyone benefits when the law is stable and predictable. But qualified immunity has utterly failed to create any such stability or predictability. “Clearly established law” is an inherently amorphous legal standard, and despite the Supreme Court deciding dozens of qualified immunity cases over the last few decades, lower courts remain persistently confused and divided over how to apply it. Nobody can reliably predict when a civil rights plaintiff’s will be found to be “clearly established” in the first place, so it would hardly be less clear to just get rid of the doctrine, and clarify that Section 1983 means exactly what it says—that when a state actor violates someone’s rights, they “shall be liable to the party injured.”
And second, the Supreme Court has already revised its qualified immunity precedent, on many occasions. For example, the Court originally said that qualified immunity turned on the good faith of the defendant, but then changed course and said the doctrine should be based on the “clearly established law” standard. And the Court created a mandatory sequencing standard in Saucier v. Katz, in 2001—requiring courts to first address the merits and then decide whether the law was clearly established—but then overruled Saucier eight years later in Pearson v. Callahan, which made that sequencing optional. It would be a perverse principle of stare decisis that permitted modifications only as a one-way ratchet in favor of greater immunity, and against the grain of text and history to boot.