Recently, the New York Times ran an op‐​ed by Professor Daniel Epps, titled “Abolishing Qualified Immunity Is Unlikely to Alter Police Behavior.” I’m fully aware that op‐​ed authors generally don’t get to pick the titles of their pieces, and I suspect this isn’t the title Epps himself would have chosen, because the actual op‐​ed is much more thoughtful and nuanced than the headline. Indeed, Epps pretty clearly comes out in favor of eliminating qualified immunity, but argues also that doing so “is no surefire solution to police misconduct.” I agree with this, insofar as there are no “surefire solutions” to police misconduct. But eliminating qualified immunity is one of the most practical, promising ways to encourage police to conform their behavior to constitutional limitations. And to the extent Professor Epps disagrees with that, he’s mistaken.

As a threshold matter, I want to note several important points that Epps makes about qualified immunity, which I’m in total agreement with:

Qualified immunity shields government officials from personal liability in federal lawsuits unless they violate “clearly established” federal law.… In theory, this requirement protects government defendants from unexpected liability when law changes. In practice, courts apply the doctrine aggressively to shield officers from lawsuits unless plaintiffs can point to other cases declaring essentially identical conduct unconstitutional — a difficult hurdle, even when police conduct appears clearly wrong.

Epps describes the doctrine exactly correctly. While the “clearly established law” standard may sound reasonable enough in the abstract (even if plainly contrary to the text of Section 1983), when you look at the way courts actually apply it, the practical effect is that whether or not victims of official misconduct can get redress for their injuries turns not on whether their rights were violated, or whether the defendants were acting in good faith, or even on how egregious the violation was, but simply on the happenstance of the particular fact patterns of prior cases. Put simply, this standard makes a mockery of justice, law, and logic.

Indeed, even if the former police officer Derek Chauvin is convicted of murdering Mr. Floyd, it’s quite plausible that a court could refuse to hold him liable for violating Mr. Floyd’s constitutional rights if his lawyers were unable to point to an earlier case making clear that the specific action Mr. Chauvin took — kneeling on a restrained person’s neck for more than eight minutes — was unconstitutional.

Also completely correct. Qualified immunity doctrine has reached the point where a police officer could be criminally convicted of murder, but where the victim’s family would still be unable to get damages in a civil suit, just because no court had previously held that this particular kind of murder was unconstitutional.

In recent years, an unlikely coalition seeking to end qualified immunity has emerged.

Epps is referring in part to the vast, cross‐​ideological coalition of public‐​policy groups that Cato has helped to assemble, all of whom have called for the elimination of qualified immunity. Epps also discusses how, on the judiciary, both Justice Thomas and Justice Sotomayor have criticized the doctrine.

There are compelling arguments against qualified immunity. One is compensation: People who are harmed by the police (or those people’s families, in cases of police killings) should have a way to obtain money for medical bills and for pain and suffering. Whatever one’s approach to legal interpretation, it’s hard to justify letting judges make up rules to deny people remedies for serious violations of their constitutional rights.

Epps correctly point out that qualified immunity regularly results in denying justice to individuals whose rights have been violated, and that the doctrine itself was made up by the judiciary, under no reasonable theory of legal interpretation. If the argument against qualified immunity were “only” as strong as this, that alone would be sufficient grounds to eliminate it.

Nevertheless, we turn now to what I take to be the core of Epps’ argument, and the focus of our disagreement—how much of an impact the elimination of qualified immunity is likely to have on police behavior. Epps thinks that because of (1) Fourth Amendment jurisprudence that is itself very deferential to police, and (2) the widespread practice of indemnifying officers who are held liable in civil rights suits, eliminating qualified immunity, on its own, is unlikely to actually change police behavior on the ground.

Before explaining why I think Epps is wrong, it’s worth pointing out that this is the exact opposite of the most common defense of qualified immunity out there—namely, that if you eliminate qualified immunity, it will have such a big impact on police decision‐​making that officers will just stop doing their jobs entirely. (This is the spurious argument advanced by the International Association of Chiefs of Police, which I already addressed here.) So, even if Epps were correct that abolishing qualified immunity is “unlikely to alter police behavior,” that actually directly rebuts the primarily line of defense raised by the doctrine’s few proponents.

Okay, but with all that prefacing out of the way, onto the actual problems with Epps’ position. Why does Professor Epps think eliminating qualified immunity is unlikely to be effective?

Courts interpret constitutional rights against police violence quite narrowly, and it is unlikely they will provide redress for a great deal of troubling police conduct even without qualified immunity. Supreme Court doctrine permits police officers to use deadly force when they have “probable cause” to believe someone “poses a threat of serious physical harm.” The standard is highly deferential.… Even where smarter tactics could have prevented death, courts will find no violation so long as the officer can plausibly argue that he feared he was under threat at the moment he used deadly force. In these cases, ending qualified immunity won’t make a difference.

As my former law professor Charles Fried used to say, “I agree with everything but the ‘therefore.’ ” Epps is definitely right that Fourth Amendment case law is already highly deferential to on‐​the‐​spot police decision‐​making. Indeed, I have made exactly this point myself to explain why qualified immunity is “entirely unnecessary to ensure that police can make quick, split‐​second decisions, because that protection is already baked into our Fourth Amendment jurisprudence.” And Epps is similarly correct that there’s a gap between “best‐​practices policing” and “police conduct that minimally satisfies constitutional standards.” Eliminating qualified immunity will not, on its own, ensure that police conform to “best practices.”

But… there’s also a ton of police misconduct that does fail to meet constitutional standards, but which nevertheless gets excused under the doctrine of qualified immunity (see, for example, the three examples I discuss in this post). Epps himself already acknowledged that Derek Chauvin, even if found guilty of murder, might nevertheless receive qualified immunity. So sure, it’s correct that “minimally meeting constitutional standards” is not the same as “best‐​practices policing” — but law enforcement today regularly fails to satisfy even the most basic constitutional standards! If eliminating qualified immunity “only” gets us to the point where police officers more diligently avoid committing constitutional violations — even if they fall short of best practices — that itself will represent a massive change in police behavior.

Epps next discusses how, even today, police officers are nearly always indemnified for any liability in civil rights cases, meaning that it’s actually their municipal employers, rather than the officers themselves, who have to pay.

If police departments are largely footing the bill, perhaps the increased liability risk would encourage them to take more steps to prevent abuses. But governments are not profit‐​maximizing entities, and they do not respond to costs the way private businesses do. Sometimes, politicians may conclude it’s easier to just keep paying judgments rather than change police culture in meaningful ways. The City of Chicago, for example, over the past 15 years has spent many hundreds of millions of dollars in payouts and legal fees in civil cases involving police. Yet that high bill doesn’t seem to have prompted the city to fundamentally rethink its approach to policing.

But the mere fact that governments are not “profit‐​maximizing entities” doesn’t mean they don’t respond to financial incentives. Indeed, it’s extraordinarily well‐​documented that financial incentives cause cities like Chicago to aggressively raise revenues through criminal fines, fees, and forfeitures, often in ways that violate people’s constitutional rights. For example, the DOJ’s Ferguson Report concluded that “FPD supervisors and line officers have undertaken the aggressive code enforcement required to meet the City’s revenue generation expectations” and that “FPD officers routinely conduct stops that have little relation to public safety and a questionable basis in law.” Or take a look at the Institute for Justice’s “Policing for Profit” report, which focuses on the especially egregious — but lucrative — practice of civil forfeiture.

So, going back to Epps’ example, the Chicago Tribune reports that in 2018, Chicago paid $97.9 million in settlements and judgments involving police misconduct. That certainly sounds like a lot in the abstract, but is it the “right” amount? Is that actually the sum total of all the harm caused by all the misconduct of police officers in such a huge city? How much higher would that total have been in the absence of qualified immunity? And how much higher would it need to be before the city was forced to make meaningful changes? I can’t say for sure, but here’s another figure to put those questions in context — according to Chicago’s 2020 budget, in 2019, the city collected $345 million in fines, forfeitures, and penalties — 7.7% of the city’s total revenue for the year.