What qualified immunity does (and doesn't) do for police officers

Totally abolishing qualified immunity would cripple good policing

I got a question in the open thread (and have gotten questions before) about qualified immunity. I wanted to address that, because it seems like there’s a lot of bad information out there about what qualified immunity actually is, why it matters to police, and how it gets applied in practice.

First: A short housekeeping note. I’m trying to find the rhythm of how of I can, and should, write here. My post about the Prude investigation was an experiment in writing something shorter, quicker, and topical. But I also had to write it quickly, and I think the quality of my writing and analysis suffered some as a result. This post is an effort to give one subject a much more in-depth discussion but took a lot longer to write. I welcome feedback on what you prefer as well as on the question of how often you want to receive emails from me.

Back to qualified immunity. Before we get into what qualified immunity is and does, we kind of have to start with what qualified immunity is not. I’m also going to mostly abbreviate qualified immunity as “QI” from here on out.

QI only applies to civil lawsuits

Sometimes people have asserted to me that qualified immunity makes it harder to fire bad cops or criminally prosecute police officers who commit crimes. This is legally incorrect and not how qualified immunity works.

Let’s take a step back and talk about ways police officers can be held accountable for misconduct. There are three distinct and overlapping systems:

  1. Employer discipline: Departments can suspend, demote, or terminate officers

  2. Civil liability: Injured people can file a civil lawsuit against an officer

  3. Criminal liability: A prosecutor can charge an officer with a crime

As one might expect, misconduct of greater severity can bring down more of (and sometimes all of) these consequences. You can’t sue a cop for failing to write a police report about something, but his department can suspend him for neglect of duty. On the serious end of the spectrum, Derek Chauvin got fired by MPD, sued by George Floyd’s family, and is being criminally prosecuted for murder.

Qualified immunity only applies to the second disciplinary system here - civil liability. If you try to sue a police officer for violating your constitutional rights, he or she can assert QI as a defense.1 But QI does not prevent a prosecutor from bringing criminal charges against a police officer who commits a crime. And it does not prevent a police department from firing or otherwise disciplining an officer for violating policy.

And it’s also important to note qualified immunity only applies to lawsuits against police officers for violating a person’s constitutional rights.2 If you’re walking on the street and a police officer is driving negligently and runs you over by accident, you can sue just like you could for any other car accident - because a car accident has nothing to do with your constitutional rights. It’s just a regular tort. But this kind of thing isn’t usually what people are talking about when they say “police misconduct.”

There are other ways police can challenge discipline imposed by a department, such as by filing a grievance under their police union’s contract (if they have one). Some states or cities also have statutes which give officers accused of misconduct a “trial board” that is kind of like a court martial. But these procedures have nothing to do with qualified immunity, which generally has no bearing on the question of employment-related discipline.

QI does not prevent lawsuits aimed at systemic issues

In a civil lawsuit, a plaintiff can generally demand injunctive relief, money damages, or both. QI does not offer any defense if a plaintiff is seeking to have a law or policy enjoined or declared unconstitutional. So if your local police department has announced that they intend to begin stopping cars and people for no reason at all, you can sue to have a federal court enjoin the department from doing that on the basis that it violates the Fourth Amendment and Fourteenth Amendments. Qualified immunity provides no defense to police departments or municipal governments in these cases.

QI also only protects individual defendants - in the case of policing, it protects officers, but not the cities that employ them. If a police department has a policy or custom that results in an officer violating someone’s rights, the department or city can be sued directly for money damages. This is called Monell liability, and institutional defendants cannot assert QI as a defense in these cases. Monell cases are hard for plaintiffs to prove because police departments generally do not announce that they are habitually doing unconstitutional stuff, but that has nothing to do with QI.

QI does not make it impossible to sue police

Even in cases where it applies, QI does not mean police can never be sued. There is a legal doctrine whereby some government employees are totally immune from civil liability - it is called absolute immunity. Only judges and prosecutors get absolute immunity. Also, legislators (Congress and state lawmakers) get legislative immunity and cannot be personally sued if the laws they enact are declared unconstitutional. (Can you tell that the lawyers made all these rules yet?)

Nonetheless people love to complain that QI is a total shield making it basically impossible to sue police. For instance:

  • An NPR podcast features someone saying QI means “you see a situation where nobody's liable” for police misconduct.

  • An article on the American Bar Association’s website says QI is the primary factor in creating a culture of “near-zero accountability” for police.

  • The Appeal, a corporate-funded advocacy website, has an “explainer” about QI which claims “qualified immunity means that victims of brutality or harassment by law enforcement generally get no relief in court.”

The problem is that this is “no relief” idea has no empirical support at all. It turns out that in reality, QI barely helps cops who get sued. Professor Joanna Schwartz - a longtime critic of QI - studied this issue in 2017:

I reviewed the dockets of 1,183 Section 1983 cases filed against state and local law enforcement defendants in five federal court districts over a two-year period and measured the frequency with which qualified immunity motions were brought by defendants, granted by courts, and dispositive before discovery and trial. I found that qualified immunity rarely served its intended role as a shield from discovery and trial in these cases. Across the five districts in my study, just thirty-eight (3.9%) of the 979 cases in which qualified immunity could be raised were dismissed on qualified immunity grounds.

So qualified immunity shields police in less than 4% of lawsuits. Four percent! That suggests that in 96% of lawsuits, QI makes no difference! Schwartz argues that this still shows QI is bad because it doesn’t actually do a good job weeding out frivilous lawsuits before trial. But that’s a long way from claiming it’s some impenetrable liability shield for police.

A little more recently, Reuters measured the issue differently by looking only at excessive forces cases and only at published appellate opinions. I have my doubts about this methodology: Only about 20% of federal lawsuits make it to an appellate court, and about 80% of appeals court decisions are not published. And most importantly, this data is probably going to exclude all the cases where police officers and cities who got sued decided to settle the case quickly, which they would have a very strong incentive to do if the plaintiff has a really solid claim.

But even Reuters’ data shows that police getting sued still only get QI about half the time on appeal. Reuters still claims this analysis supports the conclusion that qualified immunity is an “often insurmountable” hurdle for people suing police. I’d say it looks more like a coin flip. Either way, it’s nowhere close to total immunity.

Share The Graham Factor

How QI protects good policing

I don’t want to explain the law of QI in detail, because it’s been explained a lot. But QI currently means you cannot sue a police officer for violating your rights if the constitutional right violated was not “clearly established” at the time the officer violated it. In the Supreme Court’s words, QI protects “all but the plainly incompetent or those who knowingly violate the law.” I think this is a good rule, and it’s worth considering the likely consequences if QI were totally abolished.

Right now there is a Supreme Court case pending called Lange v. California. Lange was driving when an California police officer saw him blasting loud music and hitting his horn for no reason. When the officer decided to pull him over, Lange kept driving, turned into his driveway, drove into the garage, and used his remote to try and close the door behind him. The officer used his foot to trigger the garage door sensor, re-open the door, and follow Lange inside. He discovered Lange was driving drunk (he blew a 0.24 BAC) and that this was his second DUI offense.

Lange is arguing that the officer violated his Fourth Amendment rights by entering his garage without a warrant. The officer said he was in hot pursuit because Lange was trying to get away and therefore he did not need a warrant. It turns out, the Supreme Court has never ruled on whether the “hot pursuit” exception to the warrant requirement permits police to enter a home when the suspect is only wanted for a misdemeanor, like a traffic violation, so that is what they will now decide in this case.

If Lange wins in the Supreme Court, he’ll probably be acquitted of DUI.3 But should he also be permitted to sue the officer that arrested him for unlawfully entering his home? Right now, QI prevents him from doing that - because the contours of the Fourth Amendment in this particular situation were not “clearly established” at the time he was arrested. And I think that’s entirely fair, especially when you consider that this officer’s understanding of the law was in line with that of numerous state courts and California’s own Court of Appeal. But if QI were totally abolished tomorrow, this is exactly what could happen.

Some, like Professor Schwartz, argue this is not really a problem because police officers are usually indemnified by departments and don’t personally pay these judgments out of their own pockets. The argument is that police departments forced to pay out judgments for constitutional violations will do a better job supervising and disciplining officers so as to avoid liability, while indemnification means officers won’t be deterred from doing their work for fear of losing their house in a lawsuit.

I don’t buy that argument. First, as a moral matter, I don’t think a drunk driver or a drug dealer should receive a payout from taxpayers just because the courts suddenly recognized a new legal right where one previously did not exist. Unlike other torts, civil rights lawsuits don’t generally allow an officer to raise the suspect’s “comparative fault” as a defense. A world without QI means that every criminal who wins a suppression hearing will also get a payout from the government - effectively rewarding criminal behavior.4

This argument also fails to account for how the law is changing in the real world. Last summer, Colorado lawmakers rushed to pass SB 217 which both eliminates qualified immunity in Colorado state courts and limits indemnification so that officers may be personally liable for up to $25,000 if they are sued. Similar bills are pending in other states. If this were to become the federal rule tomorrow, it would mean the officer in the Lange case would end up paying thousands of dollars out of his own pocket for doing what the law had explicitly allowed him to do.

This is the exact problem the Supreme Court was trying to prevent in 1967 when it created QI. Probable cause, reasonable suspicion, and objective reasonableness are mushy standards and courts disagree about and change them all the time - one day an anonymous tip isn’t good enough to detain someone and then a few years later it is. If officers and police departments are paying out thousands of dollars every time these standards change or get reinterpreted, policing becomes effectively impossible.

How judges apply QI in practice

A common criticism of QI arises from the fact that the Supreme Court has held a constitutional right is not “clearly established” if it is only defined in general terms. What evidence does an officer need to establish “reasonable suspicion” to stop and detain someone? How much force is “objectively reasonable under the totality of the circumstances”? Because these standards are so fact-specific and subjective, the court requires a high degree of fact-specificity to show a right is clearly established. The Supreme Court explained the idea in 2018 in Kisela v. Hughes:

Use of excessive force is an area of the law in which the result depends very much on the facts of each case, and thus police officers are entitled to qualified immunity unless existing precedent squarely governs the specific facts at issue. Precedent involving similar facts can help move a case beyond the otherwise hazy border between excessive and acceptable force and thereby provide an officer notice that a specific use of force is unlawful.

The “clearly established” requirement sometimes results in libertarian outrage when a court decides (for example) that you can’t sue cops who committed theft because there was no clear precedent saying “don’t steal stuff during a search warrant.”

My take is that outrage over these kinds of highly technical decisions obscures the reality of what is actually happening in most QI cases. For the most part, judges are just deciding QI cases based on what they think about the underlying merits - conservative judges are more likely to think the police were reasonable, and liberal judges are more likely to think they were not.

American judges are generally not allowed to decide factual issues, and when deciding whether a case should proceed to trial or be dismissed, they’re supposed to presume a plaintiff’s factual allegations are true. But they also don’t want a case to go to a jury if they don’t think the officer did something wrong, because QI is supposed to “shield officials from harassment, distraction, and liability when they perform their duties reasonably.” When it decided Kisela, the Supreme Court did not explicitly say that the use of force in question was lawful - but it strongly suggested that it was:

Here, the Court need not, and does not, decide whether [Officer] Kisela violated the Fourth Amendment when he used deadly force against Hughes. For even assuming a Fourth Amendment violation occurred—a proposition that is not at all evident—on these facts Kisela was at least entitled to qualified immunity.

But because the Supreme Court can only review about a hundred cases a year, liberal judges have no problem getting around the “clearly established” requirement if they want to. Take a case from Seattle: Officers conducting surveillance saw a man named Che Taylor with a gun. They recognized Taylor and knew he was a felon, and shortly thereafter tried to arrest him for being in possession of a firearm. Taylor was shot and killed during the arrest. A firefighter treating Taylor found he was wearing a gun holster and a firearm was located inside his vehicle. Taylor’s family sued, claiming (alongside an excessive force argument) that the officers did not have probable cause to arrest Taylor because he might have stopped possessing the firearm prior to arrest.5

This court decided it had been “clearly established” that Taylor’s arrest was unlawful, determining a case about protection orders (somehow) met the standard set in Kisela:

In Beier, which was decided twelve years before the events of this case, the Ninth Circuit admonished officers who arrested a man for violating a protection order without having acquainted themselves with the terms of the order so that they could assess whether the man had actually violated it. Although factually distinct, Beier suggests that officers whose mistakes of fact or mixed law and fact stem from a failure to adequately investigate should not be insulated from liability. The Court is satisfied that, at the time Miller and Spaulding took steps to arrest Taylor, the “clearly established” law required them, in light of the intervening circumstances, to confirm their original basis for probable cause.

Instead of filing an appeal, Seattle gave Taylor’s family $1.5 million to settle the case. This case attracted a lot of heat and light in Seattle, and I think if you read the opinion it’s pretty obvious the judge’s reasoning here was motivated. And this is not the only time blue-state judges worked around the “clearly established” rule. Courts in California and Colorado have denied QI to police even in cases where officers shot armed suspects or the officer himself was shot. Reuters’ analysis of QI cases found big regional variations in how often courts grant QI to police, with judges in California granting it half as often as judges in Texas.

Wrapping up

I could go on about some of the quirkier legal aspects of QI doctrine, but my point is that the basic underlying difficulty here is not QI. The difficulty is the inherently subjective nature of the applicable Fourth Amendment standards and the frequency with which courts re-interpret them. Post-hoc judicial review of police conduct is always going to be somewhat unpredictable. To me, this is an inescapable feature of the fact that virtually every rule of criminal procedure governing police conduct has been invented by the Supreme Court since the 1960s.

Sometimes I think about alternative models. What if everything that police did was governed by statute instead of by the courts? That kind of system, I think, could potentially solve a lot of the dumber problems with American criminal law. But unless we are willing to reverse 70 years of Supreme Court precedent - and to be clear, nobody is - qualified immunity is a necessity for effective policing.


It doesn’t attract much attention, but qualified immunity also protects other government employees against allegations that they violated individual rights. So social workers, school principals, nurses at state-run hospitals, etc. can also assert QI if sued for a constitutional violation, which they sometimes are.


People sometimes call these civil rights lawsuits or “Section 1983 suits” because they are brought under 42 U.S.C. 1983.


Subject to the application of the good faith exception, which doesn’t affect the QI analysis.


Unlike basically every other area of tort law, plaintiffs who file civil rights lawsuits can win attorney’s fees even if they don’t have any substantial injury. This means a $1 jury verdict can cost the government or an officer hundreds of thousands of dollars because they have to pay their own legal fees and the other side’s fees as well.


This was a ridiculous argument. Even if Taylor had given the gun to someone else by the time he was arrested, he was still guilty of being a felon in possession of a firearm. A person who commits a crime and then stops committing it has still committed a crime.