The Washington PostDemocracy Dies in Darkness

Opinion How the insurance industry could reform American policing


In a fascinating paper made public earlier this month, University of Chicago law professor John Rappaport points to a powerful but unlikely resource that is quietly reforming police departments across the country: the insurance industry. Many cities take out policies, either off the private market or by pooling with other cities, to cover any damages that may result from lawsuits accusing police officers of excessive force or other misconduct. Rappaport explains how to guard against such awards, the insurers are in turn demanding, among other things, better training, better use of force policies, better screening in the hiring process, and even the firing of bad cops. Cities that comply get lower premiums, lower deductibles, and other incentives. Cities that don’t may lose coverage altogether.

Rappaport doesn’t claim that liability insurance is a panacea. Most of the country’s larger cities “self-insure,” meaning they simply pay out awards and settlements from public funds. He also delves into concerns about moral hazard, and explores potential danger areas where an insurance company’s bottom line might come into conflict with the public good.

But this isn’t a fanciful academic idea. It’s already happening, and has been for some time. While researching a paper for the Cato Institute on police militarization back in 2006, I found several examples in which insurers had demanded changes to policies regarding the use of SWAT teams, usually after one or more incidents that resulted in a payout to someone shot or injured during a police raid. More recently, the city of Irwindale, California had to implement a series of reforms after the city’s insurer threatened to revoke coverage. The town of King City, California, had to rebuild its police department from scratch after reports of cops operating a towing scheme against low-income Latino drivers got the town ousted from its insurance pool. It’s probably no coincidence that after the city of Waukeegan, Illinois faced multiple lawsuits from former city insurers and with those insurers had paid out over $26 million in police brutality claims, city officials are finally talking about reform. Municipal insurers’ concerns about liability have caused city officials to disband police departments altogether in Maywood, California and Lincoln Heights, Ohio.

Rappaport shows that the financial incentives insurers can offer to cities and towns for good policing are powerful. Not only are many municipal insurers already requiring better policies and training, he argues that the government entities that regulate these insurance companies could bring even more positive change.

I sent Rappaport some questions over email. Here are those questions, and his responses.

The Watch: I’d imagine that even many in the criminal justice reform movement would be surprised to learn just how involved insurers already are in crafting policing policy. Is there a good estimate of the percentage of municipalities that currently have insurance protection against police abuse, either through the private market or through pools with other cities?

Rappaport: I tried very hard but was unable to pin down this number. It seems to vary a fair amount from state to state. For example, I was informed that all but four cities in Oregon are in a pool together; the four that are not in the pool self-insure. There is no private competition. I’m told that in Indiana, by contrast, there is no pooling of law enforcement risk. California municipalities are covered by many different pools. The best generalization I can give is that the vast majority of cities other than the “big cities” have coverage from either a commercial carrier or a pool, but opinions differed as to whether the cutoff for a “big city” was 100,000, 500,000, or 750,000 residents.

What are some common ways in which insurers use coverage to alter policing policies, training, and procedures?

When a police department applies for insurance coverage, it has to attach copies of its most important policies, like its policy on the use of force. If the insurer doesn’t like what it sees, or if too many critical policies are missing or incomplete, it can deny coverage or charge a higher rate until the agency brings its policies up to snuff. Insurers also work with agencies to improve their policies, or even offer packaged model policies and procedures. They provide access to online and video training materials or connect police departments with veteran consultants who teach classes on how to do the job in a way that’s both effective and minimizes financial exposure. They also subsidize the use of otherwise prohibitively expensive training equipment like use-of-force virtual reality simulators. Some insurers also offer discounts for agencies that obtain accreditation, and the accreditation process itself requires the agency to satisfy a large number of standards in all different areas of practice and administration.

What are some of the problems in policing that you believe could be improved by a bigger role for insurers?

Although it’s hard to believe given what we’ve been seeing on the news lately, I think insurers could do a lot of good in reducing police violence. There are a couple of reasons for this. First, the use of excessive force tends to cause costly physical injuries that are highly salient to insurers. (Contrast this with, say, Miranda violations, which usually don’t result directly in highly compensable injuries.) Second, we know a decent amount about steps that can be taken to reduce police violence. Good use-of-force policies are one thing, and insurers seem to be on top of that. They can also help police departments train their officers on expensive use-of-force simulators that have yielded some promising results. And third, at least some insurers seem to pay attention to “softer” considerations that traditional legal regulators might neglect, like making sure that officers know how to cope with stress.

Can you give some specific examples in which insurers have effected a change in policy or police leadership?

I talked to a veteran consultant who told me that, on several occasions, city managers had fired police chiefs after the consultant had written a critical report in a management study. The mayor of Rutledge, Tennessee reportedly fired the city’s police chief, who had been accused of misconduct, because the city was at risk of losing its liability coverage if the chief remained. And in Calumet City, Illinois, the city’s insurance cancellation was reportedly the final straw that led the mayor to pick a new police chief from outside the department.

Did you find any significant regional differences in how insurers affect policing policies? What about differences between big cities and smaller towns?

That’s a great question. I found significant variation from state to state, but I can’t say I identified any regional patterns. The size of the city is hugely important, though. One thing that can’t get lost here is that the biggest cities — which are the ones we read about in the newspaper most of the time — almost always self-insure. So there’s no insurer affecting their policies and practices. It may also be the case, although it’s a little harder to pin down, that insurers spend less time with the smallest towns because misconduct there is infrequent and their premiums aren’t sufficiently large to justify a lot of individualized attention.

The cities and counties that are home to some of the more notorious police agencies — Chicago PD, Prince George’s County PD in Maryland, etc. — have paid out millions, tens of millions, even hundreds of millions of dollars in settlements and judgments in recent years. But the misconduct goes on, the internal affairs departments continue to clear cops at rates approaching 100 percent and we’ve seen little in the way of reform. If those costly payouts haven’t brought change, either through the city’s insurer or otherwise, why would a larger insurance premium or a larger deductible?

As I said above, but it bears repeating, the type of large municipalities you mention are self-insured. So they don’t have any insurer looking over their shoulder. They may have a person or department assigned to do risk management, but when risk management is in-house, I think there’s a greater concern that it’ll be subordinated to what are perceived as the pressing operational concerns of the day. But there’s another question there, which is whether money will ever lead to reform in some of these major cities. Joanna Schwartz at UCLA has a great new paper on how exactly municipalities finance these types of judgments and settlements. The answer is that it varies — in some localities, the police department feels absolutely nothing when a major settlement is paid out, but in other localities, they do. We need to pay more attention to these details of public financing, which aren’t sexy but may be really impactful.

You address the moral hazard question in your paper — the idea that insurance for a city coupled with indemnification of individual officers takes away any real consequences for misconduct. Can you summarize why you think moral hazard isn’t as much of a danger as some might think?

Moral hazard is always a potential problem with insurance, theoretically speaking. The question is whether the insurer does enough, through loss-prevention initiatives, premium pricing and so on, to counteract the moral hazard. So, for example, Tom Baker and Sean Griffith have a book about corporate directors and officers [liability] insurance, in which they find, surprisingly, that insurers do very little to neutralize moral hazard. My research makes me think that insurers are doing better in the policing context, but I should caution that my findings are qualitative and that I haven’t been able to verify that we don’t see some negative moral hazard effects. My hope is that future research will clarify this point.

There’s good evidence that private prison companies have been pretty active in lobbying for legislation that fills their cells, particularly in the area of immigration. It seems like there’s some danger that increasing the role of insurers in police misconduct could bring more money to lobby for laws that make it more difficult to prove abuses. One example might be the “police officer’s bill of rights” laws now in effect in several states, which critics say help cover up misconduct and perpetuate the “Blue Wall of Silence.” Do you share those concerns?

Yes. At the end of the day, what insurers really want to reduce is not misconduct, but liability for misconduct. Reducing misconduct is one (good) way to reduce liability for misconduct, but it’s not the only way. Making misconduct hard to prove could be another way. Now, the calculus turns out to be complicated — although the “Blue Wall of Silence” may make it harder to prove an abuse in a particular case (which would benefit the insurer), if it weakens deterrence of misconduct in the aggregate, insurers may oppose it. But this is definitely a question that deserves attention.

You note that some policing issues aren’t amenable to regulation through liability insurance — racial profiling and wrongful convictions are two you mention specifically. Why is that?

Racial profiling — without some additional, attendant abuse like the use of excessive force — just doesn’t tend to create enough (financial) liability exposure to capture insurers’ concern. These insurance policies do not require the insurers to defend against lawsuits for injunctive relief, which is how a lot of racial profiling issues end up getting raised. The type of police misconduct that leads to wrongful convictions — for example, the use of suggestive eyewitness identification procedures — raises a trickier question. Wrongful convictions can lead to very hefty settlements, particularly if the exoneree was wrongfully incarcerated for a long period of time — as most were. But insurers aren’t very good at managing this risk because wrongful convictions are too rare and the payouts come too long after the misconduct occurs. (The point requires a decent amount of elaboration; I have a separate work in progress addressing this point.)

Today, we often already see officer use a lot of boilerplate language and magical legal phrases drawn straight from court opinions. For example, police officers know to yell “Stop resisting!” while using force, even if a suspect isn’t resisting. Police reports are rife with liability-shielding phrases like “I feared for my life.” I’ve even had frustrated police trainers tell me that much of the lethal force training today has shifted from tactics like negotiation and conflict resolution — tactics on how to avoid using force — to how to justify force after the fact. You mention that to manage risk, some insurers can and do demand better training in the use of force, de-escalation, and conflict resolution. But what if insurers conclude that due to the politics or dynamics of a given municipality, it’s easier to just “teach to the test” — to teach cops how to avoid legal liability without reducing unnecessary force?

I think my answer here relates to what I said above about the “Blue Wall of Silence.” This is definitely a concern. An insurer that was tempted to go down this path would have to make an assessment of the long-term risk of encouraging this kind of “cover up” behavior, which could increase the total amount of misconduct, and determine that it was outweighed by the savings the behavior could garner.

You also address possible criticism from the other direction — that insurers could over-regulate policing in order to reduce losses. Can you explain why you aren’t as concerned about that?

There are two main reasons.  First, to the extent that we’re concerned that loss prevention will hamper good police work, I just haven’t seen the data to substantiate the fear.  It’s also possible that loss prevention facilitates more and better police work by reducing the incidence of distracting lawsuits.  Second, assuming sufficient competition, the market should sort out the problem.  That is, a municipality that wants its police to take more risks than its insurer will tolerate will find a more lenient insurer and pay higher premiums.  This may be why we don’t see insurers trying to disarm the police, for example, even though that would reduce the number of police shootings!

Police unions are an enormously powerful interest group that can influence policing policy in a lot of different ways, from enforcing and perpetuating the Blue Wall to pushing for those aforementioned “police bill of rights” bills. How would unions be affected if insurers were to become more active about regulating the use of force and rooting out misconduct?

I wish I had a stronger basis for answering this question. Sometimes I imagine a world in which insurers act as a beneficial counterweight to unions, but the truth is that I just don’t have enough information (yet) to go on.

Policing has mostly been a state and local issue. The only real federal input comes through grants (and any mandates attached to them) and in extreme cases, the Justice Department’s Office of Civil Rights, although that office’s powers are pretty limited. Could insurance regulation be a path to more federal involvement in developing policing policies and procedures? If so, would that be a good or bad thing?

It seems like you might have some specific federal insurance regulation in mind, but my impression is that, under the McCarran-Ferguson Act, nearly all insurance regulation occurs at the state level. But to answer the second part of your question, I do think that more federal involvement in developing policing policies and procedures would be a good thing. A lot of the model policies or guidance documents I’ve seen the federal government produce — for example, I remember seeing one on eyewitness identification evidence — have been pretty good and have incorporated many of the best practices recommended by the academic community.

Your paper also examines the role of reinsurance — the insurers who back up insurers. Can you summarize that industry’s role in regulating policing?

To oversimplify a bit, reinsurers regulate the insurers to make sure the insurers are regulating the police, and to support the insurers’ efforts to do so. The lower the “attachment point” — that is, the sooner the reinsurer’s liability kicks in — the more active the reinsurer is likely to be. But more generally, in everything they do, insurers operate against the backdrop of reinsurance underwriting. The better the insurer is doing at risk management, the lower reinsurance rates will be.

Is there any danger of insurers reinforcing existing biases? My sense in about a decade of following these stories is that the cases in which victims get paid — either through a settlement or the rare jury award — the victim tends to be sympathetic (no priors, etc.), and that white victims are more likely to get paid than black ones. That’s changed a bit with some of the high-profile incidents caught on camera. But assuming it’s true that a payout is more likely against victims from, say, white or middle-class victims, is there a danger that insurers may be more aggressive about curbing police abuses against those populations, where’s there’s more risk of liability, and more willing to defer to unions or existing policies with respect to communities of color?

I can’t say that any of my research bears on this question directly, but yes, it seems like a (troubling) possibility. Hopefully that’s changing with recent high-profile settlements in cases involving black victims, but I don’t think we really know yet. It’s a good question.

Your paper also addresses how insurers may seek to influence case law on policing. You make the interesting and perhaps not commonly considered point that insurers actually want some liability, or there would be no need for insurance. So let’s say we let municipal insurance companies write Fourth Amendment case law. What would it look like?

This part of the argument is a little bit speculative, but I think it’s worth considering. My claim isn’t that insurers have some consciously developed master plan for the evolution of Fourth Amendment law, but that their natural pursuit of their long-term interests may, over time, tend to push the law in certain directions. But to entertain the spirit of your question and speculate a little: I think insurers would favor clear rules rather than vaguer standards like “reasonable under the totality of the circumstances.” Fourth Amendment doctrine contains a fair amount of the latter, but it’s tempered by qualified immunity doctrine, which shields the police (and thus their insurers) from the financial consequences of Fourth Amendment violations that weren’t “clearly established” before the fact. So I would expect insurers to be big proponents of qualified immunity.

One proposal I’ve often seen in the police reform debate is to end both qualified immunity and indemnification for individual police officers, and instead requiring them to purchase liability insurance. Based on your research for this paper, what do make of that idea?

I’ve seen these proposals too, and I’m not totally sure what to make of them. One thing that worries me is this: Individual police officers would be really small-time customers for insurance companies; their premiums would be too small to justify much hands-on loss-prevention work. (To see the point, think about how your homeowner’s insurer doesn’t really do very much — they don’t come inspect your home and recommend changes or anything like that. But the company that insures a skyscraper definitely does. I worry that the same will be true in policing.) I also wonder just how much financial burden we’re really willing to have police officers take on. I understand the point to be that officers who commit misconduct will have to pay higher rates. But what will we do when officers start quitting because their rates are too high? And what steps will we take to ensure that municipalities are not essentially compensating officers for their premiums under the table, and thereby neutralizing whatever positive incentives we’d hope the insurance premiums would create?

Incidentally, there is a market for individual policies for federal law enforcement officers, but only federal. I’m not completely sure why — it may have something to do with the absence of unions at the federal level, but I’m not sure. The federal government pays half of the premiums.

It seems like the main lessons here are for smaller and medium-sized cities. Is there any way to create the same incentives for insurance-driven reform in the larger cities that mostly self-insure?

The lessons about “what is” do concern smaller and medium-sized cities — which, by the way, does encompass most American municipalities. As one insurer put it to me, the United States is a “country of small towns.” But the lessons for “what ought to be” extend to big cities too. Policing is an inherently risky activity. We know some things about how to manage that risk from an insurer’s perspective. But scholars who have studied in-house risk management in big city police departments, like Joanna Schwartz and Carol Archbold, have found that many cities aren’t doing much. Perhaps states could require municipalities that fall short on risk management to purchase liability coverage on the market — a sort of “soft mandate” or “defeasible mandate.” But politicians and the public need to understand that spending some money up front on risk management could save a lot of money — not to mention lives — in the long run.