Pay for Performance in Prison

This week I’m headed out to the Law and Society meeting in New Orleans, where I will be participating in two panels and avidly attending many more. The title of this post refers to a draft of my most recent article, Pay for Performance in Prison: Using Healthcare Economics to Improve Criminal Justice. I’m presenting this paper as part of a roundtable discussion on Friday at 4:45, Rationing Criminal Justice (I’m also participating in a roundtable Thursday at 12:45 on Marijuana Federalism).

Here is the abstract of the paper, as well as some thoughts on the future of the project:

For much of the last seventy-plus years, healthcare providers in the United States have been paid under the fee-for-service system, where providers are reimbursed for procedures performed, not outcomes obtained. Providers, insurers, and consumers are motivated by different individual and organizational incentives; costs and burdens of patient care are shifted from one part of the system to another. The result has been a system that combines exploding costs without concomitant increases in quality. Healthcare economists and policymakers have reacted by proposing a number of policies designed to reign in costs without sacrificing quality. One approach is to focus on the ultimate goal—improving health outcomes—by measuring those outcomes and reconfiguring incentives and structures to deliver healthcare in ways that are both efficacious and efficient. Pay for performance remunerates providers on the basis of health improvements via whatever medically-appropriate method they choose. This means providers are no longer paid for simply doing a given “something” but, rather, are paid for doing “something effective.”

In this Article, I argue that the criminal justice system is similarly fragmented, expensive, and inefficient, marked by many of the same distorted individual and organizational incentives that have plagued health care. Most significantly, in all but a handful of jurisdictions, states wholly subsidize commitments to prison—the fee-for-service model of doing “something”—without tying any of these subsidies to outcomes obtained in prison. This means prison is paid for even if it is neither effective nor efficient. These similarities with the healthcare system suggest that an outcome-oriented, pay-for-performance framework borrowed from healthcare economics might, if applied to criminal justice, improve its efficacy and efficiency. I envision this Article as the first of several applying healthcare economics to criminal justice. It will focus on the similarities of the two systems, the ways in which an outcome orientation might provide a useful framework for controlling costs without making quality subservient, and the suggestion that we begin considering sentencing choices within that framework.

I should note that the medical/criminal analogy itself has a long lineage: in the 19th century and the 1950’s, correctional professionals saw criminality as a disease to be cured. This analysis, however, looks at the recent policies in healthcare economics as a means of suggesting how the criminal justice system might be refocused on quality, not just quantity. I envision this paper as part of a larger series applying the medical model to criminal justice. The paper I plan to write next will some insights from healthcare economics to suggest how we might create incentives to invest in prevention, not just treatment. I welcome any and all feedback on the draft.

Author: W. David Ball

W. David Ball is an Associate Professor at Santa Clara School of Law. He writes and teaches primarily in the fields of criminal law and criminal procedure, with a special focus on sentencing and corrections. He also serves as the Co-Chair of the Corrections Committee of the American Bar Association.

16 thoughts on “Pay for Performance in Prison”

  1. Is the criminal justice system like the healthcare system, where there's apparently enough direct control over outcomes (and there exist sensible, widely-agreed metrics) that pay for performance can affect outcomes in fairly clear ways? Or is it more like the educational system, where massive countervailing influences and questionable metrics have made a lot of pay-for-performance programs into number-manipulation factories?

    1. Great question. It's much more like the latter than the former, and my paper is really about establishing the former as a goal rather than suggesting we're there.

      Having said that, my reading of Paul Starr's excellent Social Transformation of American Medicine suggests that the shift to outcomes was, itself, very contested in medicine. We've reached the point where we have widely-agreed metrics only after a long battle over them. I get that there is a huge incentive towards cooking the books (and we've seen some of that with hotspot policing, with pressure to downgrade serious offenses so that clearance rates aren't too high), and I also understand that we need buy-in from practitioners (I spend a great deal of time talking about this in the draft). But my main point is that the problems in medicine and criminal justice are isometric: you have discretion, you have diagnosis as–at least in part–an art, and there are treatment and selection effects. Nevertheless, medicine has moved in an outcome-based direction. It doesn't seem to me that criminal justice is so much harder that we can't do the same.

  2. FYI, Samuel Butler's "Erewhon" (1872) described a "utopia" where "offenders are treated as if they were ill, whereas ill people are looked upon as criminals" (wiki quote). For some reason, my parents had this around and I read it as a teenager. James, I thought that you would have caught this!

    1. Funny–that was one of the earliest responses I got to this draft! I've bought it but not yet had the chance to read it. It's on the pile, though…

  3. Well, I wish you luck, but I must say this sounds very neoliberal. A major turnoff, I'm afraid. (I will keep reading your posts though, in hopes of being wrong.)

    The other thing is, you may run into the buzzsaw contradiction of American politics, at least the one that exists inside my head. And funnily enough, it exists in healthcare too. There is no societal agreement on what you and I would probably agree on as the purpose of these systems. The right does not care about the health of people who can't afford to pay for insurance, and it does not care what happens to people after they are released from prison. (I am simplifying and maybe also projecting, but you see what I mean i trust.) The part where the convict is put into prison? That *is* the point for them. If you ever really want to get depressed, read this paper called "Twelve Questions," or something like that. (I don't remember the exact title right now. But it's about this divide between right and left.) So this stuff will sound good to other academics, but in the "real world," I'm not so sure it will go very far. And I myself have enough purple in me that I don't think a healthcare approach is going to fix what ails us, though from an economic standpoint it may yield insights. (As is plain though, economics is interesting to me but generally not important as a way of looking at problems, or trying to solve them. As a group I don't think much of economists. This is starting to sound like education "reform." Data here, data there, data everywhere… along with badly hidden agendas and bad intentions. Not yours, *theirs.*)

    1. I think part of my new approach, though–which is very much rooted in cost-benefit analysis–is to extend, not replace, traditional arguments about why our prison system is too big. I think there's a lot of wasted human potential involved in mass incarceration, and my "ounce of prevention" paper (which I'm writing next) explicitly talks about how social justice _is_ public safety. I want to reach a neoliberal audience because they haven't been sufficiently engaged on mass incarceration. These other arguments don't seem to work for them. The truth is, our incarceration of more people than any other society in human history is a failure on so many fronts: economic, moral, social, racial, you name it. The economic model isn't the only way to get people–but my point here is really to underscore that it's also a waste of money that isn't justified by its efficacy. That, I think, is a valuable contribution. Maybe not to someone like you who is already convinced on other grounds, but I hope it expands the number of people who are interested in unwinding mass incarceration.

      1. " a waste of money that isn't justified by its efficacy"

        Of course, to get to the point of efficacy you have to have agreement on what the system is trying to accomplish. Are we trying to minimize (some kinds of) crime? Maximize (some kinds of) local revenue? Something else entirely? I think your work will be good just in terms of teasing out some of those possibly conflicting metrics.

        To what extent are you looking at the effects of the criminal justice system on the people who are part of the enforcement/punishment side? Sometimes those are considered desirable jobs, but is it socially desirable that a lot of people undergo the stresses of those jobs?

        1. The issue about the purpose of the criminal justice system is part of the discussion I want to provoke, so you've hit on a really central point. In the essay I suggest that we use "public safety" while acknowledging that it's malleable. The same is true of "health"–is it important to extend life span or is there a quality-of-life issue? We can see life in a coma might not be "healthy" but giving someone palliative care towards the end of life might be. We might also imagine different metrics for someone living with a chronic disease with no cure versus someone who has a curable disease. Part of what I want to introduce is that the concept of public safety means different things in different contexts just like desirable health outcomes mean different things in different contexts.

          Your focus on the providers is also really important. I see my article occupying a new space in the economic literature. The existing literature mostly focuses on deterrence effects punishment has on the supply of crime/criminals. I want to focus on the incentives that providers face–although I haven't yet thought about the hedonic losses to providers. That's a good point.

  4. Goodhart's Law: any statistic used as a performance indicator becomes unreliable.

    The approach runs into quite difficult problems of priorities. The child killer Mary Bell in England did it seems achieve rehabilitation through the English penal system for young offenders. It took extremely expensive and sustained interventions, in effect substitute parenting by a team of professionals. The same resources could have been thrown at a large number of petty offenders. How do you weigh a reduction in shoplifting by many against the possible prevention of another murder? I don't raise this as an objection but as a difficulty.

    1. That's a really good point and one that is problematic. There's been a lot of interesting work done on cost-benefit analysis in criminal justice (particularly some recent work by Steven Raphael). One of the issues is how you cost out the losses to victims (property is pretty easy, loss of life is pretty hard). And there's always the scope for talking about retribution and punishment for its own sake. So on a purely rehabilitative standpoint you might not care about someone at the end of life killing people because they couldn't be reformed but might want to punish them on other rationales. There's also something to be said for the life cycle of criminality–one might see punishing a juvenile offender as being worth it if you could divert them from criminality in the future (though there is evidence that arresting young people actually pushes them into crime).

  5. " The right does not care about the health of people who can't afford to pay for insurance, and it does not care what happens to people after they are released from prison. "

    I think this is not quite right, but it is perhaps an understandable misunderstanding. If you think there are no bounds to the appropriate reach of government, then there won't appear to be any breathing room between not thinking government should be addressing a problem, and not caring about it. Only the latter could produce the former.

    But, that isn't the right's perspective. Most things are not within the purview of government, far fewer are within the purview of the federal government, and caring about something doesn't imply thinking the government should be doing something about it. If you don't ask your auto mechanic to help your son with his homework, it doesn't imply you don't care about his education.

    From the right's perspective, the left has an unhealthy, dangerous even, obsession with assigning every responsibility to government, and generally the most centralized level of government available. To exaggerate only a little, you seem to think the UN should be helping my son with his homework, and that anyone who doesn't agree must be opposed to learning.

    1. Not getting into the argument you're having, but I should note that those on the right should want to downsize prisons: there is nothing more big government than that.

      1. I see the prison population as a dependent variable. If it's to be minimized, it should happen as a consequence of changes that are justified on their own merits; Abolishing victimless crime laws, scaling back felony inflation, that sort of thing. I have some rather radical views in that regard. (For instance, I think anyone the legal system fails to convict should be made whole in regards to the costs that had been imposed on them.)

        Prison is a cost, but we shouldn't shrink from bearing it for people who should be in prison. And those who shouldn't, shouldn't, even if it were free. Either way, prison isn't for the benefit of those in it, but instead everyone else.

    2. So you're saying that right wingy types *do* care about rehabilitation services during incarceration, and when I say "care," I mean, they are willing to pay for it?

      1. I'm not sure how you got that from what I said.

        Nobody thinks prison is good for the people in it. You basically put people in prison when you've decided that doing so is necessary for the good of other people. Its purpose is incapacitation of people who've been determined to be a danger to those around them. Their own benefit is secondary at best, a side constraint on the main purpose.

        Rehabilitating them? It's the left that thinks people are infinitely malleable, not the right. Rehabilitation might fall into the "nice if you could do it" category, except that the right actually finds the concept a bit scary, smelling as it does of "reeducation" camps and the like. If you could really "rehabilitate" people, what else might you be able to do to them? What else might you try?

        Not the job of government to save people's souls, and things get really ugly when government decides it IS its job.

        1. Well we do seem to be getting our wires crossed a bit. It does sound a bit though as if you and I agree about what would be the right's approach to the question of, what is the purpose of incarceration? Mere restraint, as I originally speculated. So maybe we aren't really arguing. On the larger issue of the whole size of government/world view thing… I think we understand our relative positions okay.

          Anyway, Prof. Ball, I hope I didn't sound tooooo dismissive. Now that I understand that you are aiming at that sizable? chunk of centrists who mostly only care about money… a group which imo includes few of us here… it makes a lot more sense. At any rate your intention is certainly noble.

Comments are closed.