The doctrine of qualified immunity protects public officials from civil liability, even when they break the law, unless a civil rights plaintiff can show that the defendant violated “clearly established law.” This rule is nominally an interpretation of our primary federal civil rights statute, Section 1983, but that statute says nothing about any immunity, qualified or otherwise. And the recent scholarship of William Baude, which Justice Thomas himself has relied on in calling for qualified immunity to be reconsidered, argues that the background common law against which this statute was passed did not include anything like the across-the-board defense for all public officials that characterizes qualified immunity today. Rather, the general rule was strict liability for public officials who committed constitutional violations, with “good faith” only relevant to the extent that it was an element of particular torts; there was not, however, a generally applicable good-faith defense for state actors who acted unlawfully.
But a forthcoming article by Scott Keller, partner and appellate practice chair at Baker Botts, challenges this conclusion. His article, titled Qualified and Absolute Immunity at Common Law, focuses on four 19th century tort treatises, as well as various 19th century Supreme Court decisions, and concludes that “19th century common law did recognize a freestanding qualified immunity protecting all government officers’ discretionary duties—like qualified immunity today.” He does note, however, that the Court’s modern immunity doctrines nevertheless depart in several ways from the 19th-century common law, most notably in that “qualified immunity at common law could be overridden by showing an officer’s subjective improper purpose, instead of ‘clearly established law.'”
This article is an important and ambitious piece of scholarship, and at first glance, it certainly seems to at least complicate the question of whether there were any free-standing immunities around the time that Section 1983 was passed in 1871. For example, he cites Wilkes v. Dinsman, 48 U.S. 89, 129 (1849), which states that “a public officer, invested with certain discretionary powers . . . cannot be made answerable for any injury, when acting within the scope of his authority, and not influenced by malice, corruption, or cruelty.” Various 19th century tort treatises seem to express a similar point — that executive officials, acting in a “quasi-judicial” capacity (what we would today call “discretionary functions”) generally cannot be sued unless acting with some improper motive.
An in-depth historical evaluation of Keller’s article is beyond my capabilities here, so I’m not going to offer a definite conclusion on how much to take from this. It certainly appears that Keller has mustered a strong case, and I’ll admit substantial confusion on how to reconcile some of his evidence with the historical sources discussed in Will Baude’s article, which seem to suggest just as persuasively the opposite conclusion.
On a first look, one possible explanation for this apparent contradiction may be the fact that 19th-century constitutional law included far fewer limitations on public officials than exists today. The Bill of Rights, of course, did not even apply against state officials until the Fourteenth Amendment was passed, and it would take until the mid-20th century for the Supreme Court to incorporate most of those rights against the states. Therefore, in this context, saying that executive officers could not be sued unless they acted “with malice” doesn’t necessarily imply that they were immune for constitutional violations — because at least for non-federal actors, there were few federal constitutional limitations on them in the first place.
But in any event, I am not a historian, and I will definitely be curious to see if a response from Will Baude or other academics can unravel this confusion with more scholarly acumen. Nevertheless, I will offer two reasons for skepticism about the article’s more general conclusion — that modern qualified immunity doctrine has far more historical justification than has previously been believed (even if it’s not a perfect match):
First, whatever the general state of 19th-century common law, the relevant legal question here is whether Section 1983, which says only that any state actor who violates someone’s rights “shall be liable to the party injured,” nevertheless was understood to incorporate by reference the full scope of these supposed common-law immunities. And on this point, it’s absolutely essential to recognize that the Supreme Court, in a 1915 decision called Myers v. Anderson, rejected the application of any good-faith defense to Section 1983 itself.
Myers involved a Section 1983 suit against city officials who refused to register three black voters under a Maryland “grandfather clause” statute. In a related case, Guinn v. United States, the Supreme Court held that such grandfather-clause exemptions to literacy tests violated the Fifteenth Amendment. So the ultimate question in Myers was whether these particular state officers could be personally liable for damages under Section 1983 for enforcing this unconstitutional statute. And the defendants in Myers made exactly the sort of good-faith, lack-of-malice argument that Keller says was well established in 19th-century common law. They argued before the Court that:
The declarations filed in these cases are insufficient in law, because they fail to allege that the action of the defendants in refusing to register the plaintiffs was corrupt or malicious. Malice is an essential allegation in a suit of this kind against registration officers at common law.
But the Supreme Court rejected this argument, holding as follows:
The nonliability, in any event, of the election officers for their official conduct is seriously pressed in argument . . . . But we do not undertake to review the considerations pressed on these subjects, because we think they are fully disposed of by the ruling this day made in the Guinn case and by the very terms of [Section 1983], when considered in the light of the inherently operative force of the Fifteenth Amendment as stated in the case referred to.
In other words, the Myers Court held simply that the statute was unconstitutional under the Fifteenth Amendment, the city officials were enforcing this unconstitutional statute, so they were liable under Section 1983, QED. Whether or not the officers were acting with “malice” was irrelevant, given the plain terms of the statute. So, even if Keller is right about the general state of 19th-century common law, that itself doesn’t make it obvious that this particular statute was meant to subsume those common-law defenses. And in Myers, the Supreme Court held exactly to the contrary.
Now, Keller does briefly discuss Myers in a footnote, arguing that commentators “frequently overread” this decision “for the proposition that executive officials had no immunity from damages claims even in the early 20th century.” His explanation appears to be that because Myers involved a racial discrimination claim, the defendants here — by enforcing the unconstitutional statute — necessarily had the sort of “discriminatory purpose” that would constitute “common law malice.” In other words, Myers doesn’t stand for the idea that malice is unnecessary to show liability under Section 1983; the defendants here just happened to be acting with malice, given the nature of the claim.
But if you look at what the Court actually said in Myers (the language I quoted above), it’s clear this was not the rationale the Court was adopting. While the Court didn’t go into much detail, note that the defendants explicitly argued that “[m]alice is an essential allegation in a suit of this kind.” The Court could easily have said “yes, this is generally true, but malice is inherent to racial discrimination claims” or “malice is plainly established by the facts alleged.” But that’s not what the Court did. Rather, the Court essentially said this malice argument was irrelevant, given the ruling in the Guinn case (which held that such statutes were unconstitutional) and “by the very terms of [Section 1983].”
I think the only reasonable interpretation of this passage is that the Court was holding that the presence of common-law malice was simply irrelevant to Section 1983 suits — all that matters is whether the defendants violated the plaintiffs’ constitutional rights. Therefore, notwithstanding Keller’s dismissal of this case, I think Myers is powerful evidence that no matter the general state of 19th-century common law, Section 1983 did not incorporate across-the-board good-faith defenses for all public officials.
Second, even if Keller is correct that Section 1983 was meant to incorporate a general, good-faith defense for executive officers, there is a massive difference between this sort of actual good-faith requirement and the “clearly established law” standard that characterizes modern qualified immunity. Keller acknowledges this distinction, of course, and explicitly discusses how it represents a divergence between the 19th-century common law and modern doctrine. But I think he understates just how massive a gulf there is between these two approaches. If we were to replace the “clearly established law” standard with an actual, good-faith defense, that would not be a minor, technical correction — it would be a fundamental reshaping and substantial limitation of what qualified immunity actually is.
Notwithstanding that the Supreme Court has called the “clearly established law” test an “objective good faith” standard, the practical operation of the doctrine has nothing to with what a reasonable person would call “good faith.” Under qualified immunity today, even if officers are explicitly acting in bad faith, even if they are intending to violate someone’s constitutional rights, they can still receive qualified immunity, just because no court has confronted that particular set of facts before.
The clearest example of this point is Jessop v. City of Fresno, in which the Ninth Circuit granted immunity to police officers who were alleged to have stolen over $225,000 in cash and rare coins while executing a search warrant. The court said that while “the theft [of] personal property by police officers sworn to uphold the law” may be “morally wrong,” the officers could not be sued for the theft because the Ninth Circuit had never specifically decided “whether the theft of property covered by the terms of a search warrant, and seized pursuant to that warrant, violates the Fourth Amendment.” Obviously these officers were not acting in good faith, and no one contended otherwise. But because there was no prior case involving such outlandishly illegal misconduct, they received qualified immunity.
Indeed, in the law enforcement context, most police actions that are genuinely carried out in “good faith” will not violate the Constitution in the first place! After all, the touchstone for most Fourth Amendment questions is “reasonableness.” A police officer does not violate the Fourth Amendment just because they arrest someone who turns out to be innocent or use force that, with the benefit of hindsight, was unnecessary. As I discussed in more detail here, the Supreme Court’s decision in Graham v. Connor creates an “objective reasonableness” standard for excessive‐force claims that is highly deferential to on-the-spot police decision-making. Basically by definition, an officer who is genuinely acting in good faith will not be acting “objectively unreasonably,” which means they won’t be violating the Constitution at all.
Another way of expressing this point is that, at least for police officers, there is not much of a gap between reasonableness and lawfulness. The only obvious example I can think of where you might say an officer is genuinely acting in good faith, but nevertheless violating someone’s constitutional rights, is where they are carrying out a statute they reasonably believed to be constitutional (but which is then held unconstitutional), or acting in accord with judicial precedent that was governing at the time (but which is then modified or reversed). There’s a reasonable case for limiting liability in these context, and it’s for exactly this reason that I suggested a legislative fix to qualified immunity could include safe-harbor provisions in these circumstances. This was also the policy judgment reflected in Senator Mike Braun’s “Reforming Qualified Immunity Act,” which included exactly these safe harbors.
But whether or not police officers should receive immunity in those limited circumstances, such cases make up only a very small fraction of Section 1983 suits. Thus, in the mine run of cases involving law enforcement, replacing the “clearly established law” standard with an actual good-faith standard would be close to eliminating qualified immunity entirely.
* * *
In conclusion, I commend Scott Keller for a thoughtful and detailed article, and for contributing to the scholarship on such an important question. As I noted above, he makes a powerful argument that the common-law background on governmental immunity is, at the very least, more complicated than a lot of us previously recognized, and I look forward to the academic responses I’m sure this will provoke.
However, even if Keller is right about the 19th-century common law background, I remain skeptical that Section 1983 itself was meant to incorporate any kind of across-the-board defense for all public officials. And Keller himself acknowledges that the “clearly established law” standard, which is the defining feature of modern qualified immunity doctrine, is fundamentally at odds even with his interpretation of 19th-century common law. In my view, the case for eliminating qualified immunity — or at the very least, eliminating the “clearly established law” test — remains as strong as ever.