A Framework for Incentive-Oriented Police Reform


The tumult over needed police reform reminded me of an idea I had a long time ago, but never wrote up. It's a preliminary sketch of a framework for discussion, I don't claim to have a detailed proposal that would answer all the questions the proposal raises.

Here goes: In a big city like New York, why not divide the police department into, say, five separate operating units, run by separate private companies under contract with the city (with the officers still officers of the state). Judge them annually via an independent consulting firm on various measures, like crime rates, crimes solved, complaints of misbehavior, etc. The company that performs best each year gets to extend its jurisdiction by, say, 5%, the worst performing company loses five percent. Over time, the best company patrols more and more of the city, the worst less and less. Maybe the latter eventually disappears; maybe it improves. The key, though, where unlike today police departments are primarily beholden to political bosses and union leaders, the folks running the operating units would be accountable to objective standards of performance.

NEXT: Is the Future of Freedom of Speech in Jeopardy?

Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of Reason.com or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

  1. Does the winning company get more money, either for profit or to hire more officers? Otherwise, the incentives seem backwards.

  2. Sort of like charter schools? The police union will love that just as much as the teachers union loves charter schools.

    1. Which is why reform #1 should be to abolish public servant unions. (Of course, ‘they’ think they are public masters, so assume it won’t apply to them. BWAHAHA.)

  3. Over time, nobody wants to police the black people because they don’t want murders or complaints.

    1. Perhaps, but on the other hand some police entrepreneurs may want to tackle the challenge and reduce crime rates in troubled parts of town.

      I also suspect that bids on high crime areas might be submitted jointly and those areas patrolled by several of the companies, spreading out the investment and the risk.

  4. Why not? Because eventually one company will become so big that it’s political influence will be so great that even negative evaluations will not lead to any consequences. Then we wold be stuck with an even less accountable police force.

    1. It is impossible to conceive of a less accountable policing system than what we have now.

      1. No, not really. Were the police really less accountable, and less corrupt than in prohibition Chicago and he Jim Crow south and pre-Miranda? It is more, like a lot of things, that we are more aware and focused on a smaller problem. The amount of news distorts the understanding of the underlying truth of what it really happening (say, childhood safety / risk, or if it is not too soon, disease risk (compare 19th century polio with today’s coronavirus, both ‘pre’ vaccine)).

      2. Miranda was a tremendous step forward but remains just a drop in the ocean. Civil Forfeiture, pre-trial detention, plea bargaining, discovery and sentencing are all unimaginably corrupt. Not to mention Sovereign Immunity. The criminal justice system is a cesspool.

        The primary issue is that the police are, in effect, accountable to no one. If a government cop commits a crime, it is the government who investigates, prosecutes, adjudicates and punishes the offender. There is no way that could work well.

        A private, corporate police would At least be subject to a government judicial system and a hostile appropriations system as well as the discipline of the marketplace.


        1. 1. Glad you’re conceding that there were, in fact, less accountable police systems at certain points in American history—which would a fortiori make it possible to conceive of less accountable ones than we have now.

          1. 2. Prof. Bernstein is expressly proposing that these police officers would “still [be] officers or the state”.

  5. This is a bad idea for any number of reasons. Let’s hit on a few.

    1. It actively discourages cooperation between police departments.

    -Generally speaking, you want police departments (outside of a couple select areas like internal affairs) to cooperate with one another. When a murder suspect flees from one jurisdiction to another, it’s usually a good idea if the local cops help out. This type of “competition” discourages that. Why use your resources to help someone else’s crime…?

    2. It paradoxically lowers the police resources in high crime areas.

    Some areas have higher crime than other areas. By lowering the amount of resources that go into the high crime areas (because they’re high crime areas, and score “poorly”), it encourages more crime.

    3. It concentrates crime into poorer areas.

    -This “let’s take 5% of your district” idea sounds great…until you realize that the “5%” taken will invariably be the “good” area of the district. No police department is going to want to expand into a high-crime area that will lower its numbers. They’ll get the medium and good areas, leaving the crappy department with the highest crime areas and the lowest budget. In addition, criminals will flock to the area with poor enforcement (because they are less likely to be caught).

    There are more reasons, but this is a poorly designed concept.

    1. I’m not sure we should throw stones at him for trying though.

      It’s true that in all government / quasi-government systems that have tried to introduce some kind of quasi market, with targets and scorecards, that the competitors chase the targets not what lies behind them. So if you set a target of 1,000 arrests, the police will make 1,000 easy minor drug offense or traffic violation arrests. Harder cases will be ignored.

      But that doesn’t necessarily make the result worse than the old union run monopoly.

      A different sort of reform that has a better chance – though sadly not in the police area where local monopolies are unavoidable – is in the realm of licensing. Whether that be in licensing business or charter schools or adoptions or doctors etc. For licensing you can avoid a politically entrenched monopoly by allowing more than one licensing board.

      So for example you could allow a school or an adoption agency to be licensed by a regulator approved by any of :

      (a) the State Executive
      (b) either State legislative House
      (c) the municipality
      (d) the federal government (ie an Executive agency

      1. Oh, I’m not throwing stones at him for trying. I’m criticizing the idea, and pointing out why it wouldn’t work in this context and plan.

        Licensing (by multiple entities) DOES work. The clearest example of this is the “licensing” (or incorporation) of corporations. There are 50 different states which allow for corporations to be formed, but you don’t need to be “in” the state to incorporate under that state’s laws. Typically you see Delaware used most frequently. What makes it work is, among other things, it’s decided by the individual.

        But, let’s go back to the police system for a second. I’m not sure that a local monopoly IS unavoidable. Local, county, and state police departments often have overlapping jurisdictions, so there’s a little bit of room there. Let’s take this a step further.

        Imagine instead of there just being a “NYPD” there were three different NYPDs (Let’s call them Yankee NYPD, Mets NYPD and Knicks NYPD) Each of them had jurisdiction over the city, and their funding came from the taxpaying citizens. But critically, the taxpaying citizens on their tax forms had the ability to direct their taxes to one of the three PDs. Moreover, when responding to calls, the taxpayer could specify “their” PD.

        In such a way, the PDs would be uniquely motivated to respond to the needs of “their” citizens and punished by any perceived abuses (by being dropped as a PD that got funding). If “Knicks” PD has a history of doing poorly, and not being able to protect its citizens, its citizens switch to Yankee PD.

        1. That is how fire departments initially were — you subscribed to a private fire company and nailed their metal decal onto your house. The various fire companies would show up at a fire and the competing companies would attempt to prevent the specific one from fighting the fire. Lots of fistfights and such.

          It got so out of hand that we went to municipal departments.

        2. Seems to pose similar cooperation and coordination issues.

    2. Thanks Armchair Lawyer. you said what I was thinking. This whole idea is dumb as a box of rocks. People need to stop looking at making criminality profitable for the state. Or private companies. It is the wrong expectation. Crime doesn’t happen regularly because lots of people just want to be criminals. Crime happens because people are grown into a system that is designed to make you crave wealth you can never achieve if you didn’t have some kind of systemic wealth in the first place.

      As sure as a plant grows to fit the space it’s in, if the only space you have to live your whole life in, is poor, poorly educated, aimed at keeping the poor where they are and specifically using the poor to enrich those who are not poor, that is what will happen (aside from a few serial killers). If your only happiness is a false happiness brought on by a drugged state, you will learn to only seek out a drugged state because you know no other such happiness. And then you feed and water the potted plants with low grade poisons, poor-quality foods for mind and body. That’s not to say a few branches and seeds do not fall out of the pot they were springing from, and manage to achieve some stability in income, education and health, we have a system where the biggest and wealthiest are free to believe those few stragglers may threaten their dominant wealth and privileges; and so they find ways to continue the system that returns the stragglers or the stragglers’ future generations back to similar pots to start over from. And that usually is the start of over policing, of controlling movement, of criminalizing behaviors that shouldn’t be criminalized. After all, it was the only path left to them.

      1. Poetic. Bad sociology and bad economics, but poetic.

    3. You could remedy one defect by keeping a rump police department that just does investigation of crimes, and limit the contractor to patrol and arrests.

      Other obvious reforms are obviously no swat squads public or private, no no-knock warrants, no asset forfeiture.

      But some reforms have already been tried and failed like no cash bail for violent offenders.

    4. #1 is hogwash. Police departments, sheriffs, FBI, DEA, ICE — they all cooperate now. What makes you think these police departments would not cooperate?

      #2 needs a damned good citation. Poor folks don’t get anywhere near the policing, and the kind of policing, they want now; what makes you think they do, and what makes you think they’d get less when police would actually have an incentive to police those areas?

      #3 goes right along with #2. Bald assertions do not facts make.

      1. 1. Because under this plan, you provide an incentive for them not to cooperate. If they don’t cooperate, the other department’s numbers look worse, theirs look better, and they get a boost in pay.

        2. Yes, high crime areas get more police attention by virtue of being called to the area more often.

        1. #1 does not work that way. I suppose any government burrocrat would like to design incentives that badly, but it would also take skill, and they lack that.

          #2,3 show your lack of faith in humans. If people had a choice between good and bad cops, they would choose the good cops. If bad cops decided to stay away, that would leave an opening for good cops.

          Market failure is how markets work: someone sees a failure and works to fix it. When governments step in to “fix market failures”, it is burrocratese for not fixing what is not a failure, and instead making government failure worse to expand their burrocracy.

  6. How about: since I’m not close to an expert on this, I’ll suggest I read up on this from those who are, and maybe *then* toss out a ‘thought experiment?’

  7. The devil in the details here will be developing the metrics for what constitutes success without political interference, and then finding an actual disinterested party to apply them.

    Good luck with that.

  8. How about a slight modification.

    The city is divided as you say for purposes of ordinary police services. However, all five organizations have jurisdiction to go after police misconduct anywhere in the city, and when one district makes an arrest of an officer from another, they get some of the other district’s points too.

    If a citizen sees police from 911A kneeling on someone’s neck, at least they can call 911B and hope to get some action.

    1. This. And, maybe, instead, just have regular police like we do now (but no qualified immunity, no no-knock warrants, no choke-holds, etc.), and set up these for-profit policing agencies to police the police, instead of having the police police themselves.

      There is a reason we don’t have a for-profit army as our go-to army. I don’t see how a for-profit police force (which unfortunately, with civil forfeiture, we already have too much of that in the police we do have) is really any better.

      As someone else pointed out above, private fire department associations didn’t work very well. Policing seems even less conducive to a “private” solution.

      And can we please get rid of “for-profit” prisons? It has created a powerful interest group vested in criminalizing more conduct and incarcerating more people. As if we don’t already have an incarceration rate way out of proportion to every other reasonably free country in the world.

  9. But, where would the politicians get their graft?

    1. Have some faith in our public officials, man!

      1. Minneapolis’s public officials:

        Only Democratic Mayors since 1973.
        Almost 100% Democratic City Council.
        State representatives: Democratic.
        Governor: Democratic.
        US representatives: Democratic.

        The standard result of one-party Democratic rule: incompetence, oppression, brutality.

        1. You responded to a joke, Kevin.

          You must be a blast at parties.

            1. You thought I didn’t believe graft was a thing?

              1. You used speculative telepathy to assume that the statement was made in bad faith from the other side.


                1. I may have been stepping upon Sarcastro’s turf, but I was totally sarcastic in my declamation of political graft.

                  1. Flight-ER-Doc and Sarcastro were funny. Why ya’ll ruining the joke? Jeez.

      2. He’s in normal company. PEW surveys say about 20% of folks trust government, down from over 70% in 1958.


        What gets me, is that even though democrats trust government at about 30% and republicans at about 15%, how many democrats still want to expand government. Like they think that Trump won’t get reelected or there own’t be a conservative that ever gets his or her hands on the reins of power again.

        1. I’m not sure it’s that odd. People are mostly comparing the government they see with the government they want, and unsurprisingly the government they see falls short.

          The government they don’t see – ie the hundreds of far worse governments across the world, and the infinite stock of far worse governments that they could have – don’t figure in the comparison. (Except for dark gloomy conservatives, who can imagine a lot worse.)

          The popularity of the idea that government could be a lot better is a sign that most people are very naive – naive enough to vote for more government in the expectation that it will be well run. In reality government couldn’t be a lot better, unless we got rid of electoral politics. Democracy is very inefficient.

          The trouble, though, is that while various forms of authoritarian government could be quite effective, there is no mechanism to stop them becoming very effective at the wrong things.

          So its best to go for the minimax solution. Settle for the bad government that limited-government democracy offers, to avoid something worse.

          1. Voting for control over women’s bodies, the bloated Republican budgets, criminalization of recreational drug use, civil forfeiture (looking at you Jeff Sessions), harsher criminal punishment (i.e., the government literally taking control over more people’s lives for a longer period of time), and more brutal policing (stop and frisk), is definitely bigger government. Please stop with the “Republicans are for smaller government”. The current Republican party is pushing for bigger government, particularly in ways that directly control people’s bodies and lives. They are in no sense for a smaller government than Democrats. (And the spending spree under unified Republican government from 2017 forward (and only in 2019 did Democrats take control of the House, so Republicans still control 2.5 of the three branches) proves that beyond any doubt.)

            The Democrats’ big government sin according to Republicans: wanting to raise taxes during strong economies to pay for the stuff both Democrats and Republican have chosen to spend on. And don’t say, no, it’s they want to spend more. Democratic Presidents spend less than Republican Presidents. (Congress plays a role, etc., but look at the deficit numbers by President and there is no debate).

          2. Lee, I do agree with your minimax solution. Was it Churchill who said “Democracy is the worst form of government, next to every other system that’s been tried.”

      3. sarcastro : Have some faith in our public officials, man!

        The definition of insanity is doing the same thing over and over and expecting different results.

        The American Republic is not founded on faith but on the ancient wisdom that the first and last question of government is “quis custodiet ipsos custodes.”

      4. I thought it was clear I meant have some faith our government will find ways to graft.

        I’m an optimist, but not a Pollyanna!

        I take it as my error.

        1. In which case I withdraw my brief lecture.

          Though you’re stiill a mug to be optimistic about government. Save your optimism for the potential of humans acting and co-operating voluntarily.

          Government is there for the non voluntary bits and that is seldom pretty.

          1. The thing is that governmental agencies ARE people cooperating voluntarily, just not to maximize profits.

            Markets are great, but sometimes the ends you want are not efficiency.

            1. The thing is that governmental agencies ARE people cooperating voluntarily

              In the same sense that a gang of bank robbers are people co-operating voluntarily, so long as you restrict the set of “people” to those within the gang. The co-operation of those outside the gang is not voluntary.

              It’s possible that there are some bits of government that do nothing at all except disburse goodies to voluntary recipients, but I’m not aware of any such whose co-operation with its financial supporters is any more voluntary than the relationship between bank robbers and banks.

              And I doubt there are many, or any, that have no coercive legal powers over and above what a charity involved in similar disbursements might be.

              But feel free to illustrate your point in fuller color.

              1. I would urge anyone who thinks as kinda of a default “government is that thing we do together so it starts out as a good, even if it may be corrupted” to look up the the idea put forth but Mancur Olson of a “roving bandit” and “stationary bandit” about governments. There is no clear answer which of those a democracy is.

                1. “government is that thing we do together so it starts out as a good, even if it may be corrupted” may be a little on the kool-aidy side, but I am with Mancur Olson that, on the whole, a stationary bandit is generally to be preferred to a roving one.

                  But as we have seen, writ large in the 20th century, some stationary bandits are not that big on economics, and seem to like killing people for the sake of killing people. When it comes to homicidal maniac bandits, perhaps the roving kind is better – at least they may move on before they’ve killed everybody.

                  In any event, on this occasion I wasn’t really arguing with sarcastro about the goodness of government, but whether, good or bad, it could reasonably be characterised as “people co-operating voluntarily.” My view is that it cannot.

        2. You were entirely clear. People need to read with less partisanship if they didn’t get it.

    2. Is it not obvious? Politicians and their staffers will take up post-politics jobs with the consulting industry, just like they do in every other public->private transfer of power. This is an old trick.

      1. It is obvious. People keep falling for it for some reason though…..

  10. Doesn’t go far enough. Each police department should be linked to its own private prison complex. Each private prison complex should in turn be financed by forced-labor contracting to private industry. Then add private organ harvesting operations, with organs of prisoners auctioned to highest bidders world-wide. Develop an advance-scheduling option, for organ customers willing to pay extra for reliable delivery dates and assured quality.

    Use the labor contracting and organ harvesting to finance the whole shebang. Cost-free, maximum-deterrent law enforcement, with no accountability headaches at all. What could be better?

    1. Sigh.

      People are being killed by police and your biggest fear is someone somewhere will make a profit.

      1. I had a sarc-off bit at the end, but apparently it looked too real and got stripped out.

        Or, oh!

        You think sarcasm about the accountability of free markets is sacrilege? What’s your problem, the folks killed by police, or private profit potential withheld by government?

        1. OK, a mindless belief in the perfect efficiency of markets is a fair target for sarcasm. For that matter everything is a fair target for sarcasm.

  11. I’d love to see that kind of objective-measurement system in place, but someone has to be in charge of interpreting the results — and how are you going to keep that person’s job out of corrupt hands?

  12. Lol “objective standards of performance”. Yes, I’m certain those standards will not be games/captured/perverted now that a private company stands to gain a billion dollar contract.

  13. Sometimes I think that libertarians are so much for individualism vs communitarianism that they ignore the fact that they don’t have their own water supply and their own latrines, which they then recycle. Or perhaps they DO recycle it, in the form of columns like this. And David, don’t walk on the sidewalk or drive on the road that your hated taxes paid for.

  14. Policing is too important to be left to the private sector.

    1. Providing people with food, more important than the policing, is left to the private sector. You need to think of another way to express what you’re saying.

      1. He is right. I understand exactly what he’s saying and how food is different from policing.

  15. Look at what happened when Massachusetts privitized ambulance services. It neither reduced costs nor increased service.

    1. Did they privatize them, or did they just install crony capitalism?

      1. Is there a difference?

  16. This makes no sense. Bernstein appears to believe that the management of policing is corrupt and incompetent, and that the public doesn’t know or care about the quality of policing, but that public sector contracting is a bastion of integrity and good judgment, closely monitored by an engaged public. That’s almost the opposite of the truth.

  17. The National hangover from the latest left wing driven moral panic has begun. After a bender of protests, violence, and looting the ramifications of all of this are coming to roost.
    -People are asking why the looted stores are now permanently closed. Of course, the ones in rich downtown districts are re-opening. The ones that aren’t are grocery stores, drug stores, etc. in poorer neighborhoods.
    -“Defund the police” really means wacky ideas like “restorative justice” for thefts and burglaries. If someone breaks into your house and they are poor then you are supposed to meet them in person and try to understand their plight. (Pretty much give them a pass for stealing your property and threatening your safety.)
    -Now that bars and restaurants are opening again, protesting is not the cool thing to put in your social media feed. Last week it was edgy to talk about looting as social change. This week you need to get a selfie at the outdoor beer garden.

    It is time for society to drink a few glasses of water, take some aspirin, nurse its hangover, and get back to real business.

  18. Could the the companies be sued for damages when their personal use excessive force?

  19. “Judge them annually via an independent consulting firm on various measures, like crime rates, crimes solved, complaints of misbehavior, etc.”

    Why grant a private consulting firm a license to steal from the state? If we know enough to centrally plan the criteria for judgment (which I doubt), why would we pointlessly outsource it? And even if we could, it begs the question of how to exercise oversight of the consulting firm.

    These sorts of half-baked policy proposals are not helpful. Anybody who has spent time working for or around government would immediately recoil at this idea, unless they were lobbyists for the law enforcement consulting industry.

    1. https://en.wikipedia.org/wiki/Single_Audit

      We already have this for grants. Dunno how well it works though.

      1. A few issues. First, that’s federal. I haven’t spent enough time around the feds to know how functional their auditing is, but there is a scaling component that would not apply even to large states.

        Second, I’m not worried about the financial component. A third-party CPA evaluating expenditures is just a bean counter looking at objective facts. It’s the compliance component I’m wildly skeptical of. To the extent we even trust the metrics selected by government (at any level)–which I wouldn’t for something as complicated as policing–the ones suggested by Professor Bernstein scream for abuse. From the OP:

        “like crime rates, crimes solved, complaints of misbehavior, etc.”

        Every metric can be gamed. If you place incentives on decreasing crime rates, and make people’s money depend on it, then they will discourage reporting, engage in draconian measures to decrease crime, etc. Crimes are more easily solved if you just plant evidence or accuse the mentally ill who will admit guilt. If the department gets punished because of “complaints of misbehavior” do you think that will increase or decrease self-reporting of misbehavior by police? (Financially incentivizing complaints has the opposite problem; cops will lie about good cops to get money.)

        Lots of states have non-financial auditing mechanisms, sunset laws for existing legislation, etc. Anybody who has been around those processes knows that they are political, run by people who are not subject matter experts about any of the thing being audited (which is both a feature and a bug, for reasons explained below), and those institutions themselves lack rigorous accountability for the same reasons that the entities they evaluate are dysfunctional.

        So why not just have subject-matter experts review? Well you still have to choose the compliance measures. But subject-matter experts have a financial incentive to make sure every problem is the one only they can solve. Regulatory capture is utterly rampant at every level of government.

        These problems are too complicated to be solved with central planning. Privatizing the solution with centrally planned non-financial metrics does the system no favors, since the root problem is the same. It’s probably worse, since it injects the solvers with steroids, because the financial incentives are massive.

        It takes unfathomable hubris to think central planning can effect “police reform” (whatever that means). We should start small and humble. Solve the obvious problems, beginning with the ones created by government. Qualified immunity. Armed beat cops. Police unions. These are all things that only exist because of some centrally planned decision. And those decisions can be reversed. Want to cut down on interactions between police and the citizenry? Make marijuana legal, everywhere.

        Why are we here? Floyd allegedly tried to use a counterfeit bill. The clerk found out. Did the clerk accept the bill? If not, why is this a matter for the police anyway? If so, why would you need four cops to make that arrest? The punishment is <1 year and $3,000 fine. Who fucking cares? It's a god damned $20 bill. Not every fucking crime requires an arrest.

        It's also so strange that this incident spawns international protests now requiring national reform. The problem that Mr. Floyd faced was very simple. A dumpy fucking cop with poor training murdered him. The answer is to not have dumpy fucking cops with poor training. And you do that by firing the cop, prosecuting him for the assault and/or murder, and moving on. There will be a settlement that will cost the police department lots of money. We shouldn't have needed protests to get there.

  20. I liked this thought experiment better when it starred Peter Weller.


    1. Well, meanwhile, the police are slashing the tires of cars belonging to journalists and protestors:


      1. And not for any good reason, by the way, if you don’t read the link. And, of course, there was an initial denial until footage came out proving it. Do the police ever not lie? (The old man tripped, etc.)

  21. This is a great idea! Let me just run it by the police union and get their okay.

  22. 1 – To judge police (definitionally, a reactive agency) by lower crime rates is to produce incentive for corruption and ignore the reality that policing, or any government or psuedo governmental activity can not control crime. Crime is the activity of individuals; unless one is willing to accept a tyrannical form of policing, not going to control crime. 2 – Crime does not respect lines on a map. Policing agencies need incentive to cooperate, not incentives to not cooperate. Setting up 5 agencies that have vested interest in the other four failing sets out a fabric for lack of crime solution.

  23. Do we really need a Rube Goldberg-style police structure? Don’t reform police departments, reform qualified immunity doctrine so cops are held accountable when they obviously violate the constitutional rights of people. Oh, and go ahead and reform police departments too.

  24. This was tried with private prisons, with similar theories and disasterous results. Libertarian theory doesn’t work here. The profit motive simply won’t work for people to put themselves at risk for the benefit of people behaving adversarially towards them. People reduce costs and risks by screwing the people they are ostensibly supposed to benefit and shielding the fact from the people paying.

    Look at for-profit colleges – far more scam artists than non-profit ones.

    Why should we think doing the same thing and imagining we will get different results is a sane policy?

    Like anarchism, strict libertarianism depends on theories of human nature that are sometimes true but sometimes just ain’t so. Just as relying exclusively on the goodness of human nature often won’t work, relying exclusively on the profit motive for circumstances where people have to behave altruistically for the system to work won’t work.

    1. Allow me to point out that the sort of system I’m proposing isn’t libertarianism, it isn’t even free markets. It’s applying economic incentives to public provision of services. The best criticism in the above comments is not that incentives and markets don’t work to improve quality and service (as a rule, they do), but precisely because government will be involved in created an artificial market and developing “objective” metrics, there won’t in practice be any incentive for real improvement, and it may just be an opportunity for graft.

Please to post comments