On Tax Day, it’s appropriate to note this new student note in the NYU Law Review: We Tried to Make Them Offer Rehab, but They Said, “No, No, No!”: Incentivizing Private Prison Reform Through the Private Prisoner Rehabilitation Credit, by Cassandre Monique Davilmar. (Apparently the title is an Amy Winehouse reference.) Here’s the abstract:

Mass incarceration in the United States has led many state governments to hand over the management and construction of prisons to private corporations, which are able to meet demand more quickly and are perceived as more cost-effective. There are approximately 100 private prisons housing about 62,000 inmates today, and this number is expected to increase to 360,000 in the coming decade. Unfortunately, private prisons have failed to effectively address many of the issues pervasive in public prisons—namely recidivism, violence, and poor living conditions. Furthermore, the government-customer has failed to effectively hold private prisons accountable for their failures. As a solution this Note proposes the Private Prisoner Rehabilitation (PPR) credit: a performance-based, refundable tax credit that incentivizes private prisons to address some of the key issues plaguing the criminal justice system.

This note cites my own student note at n.126. But most importantly, it deals with similar issues to my own recent Emory Law Journal article, Prison Accountability and Performance Measures, which I blogged about here recently (click here for the last blog post in that series, which links to the other posts in the series).

I’m glad when anyone writes about this important issue. Here are some thoughts I have:

1. Why focus exclusively on private prisons as the source of the problem? Part I of the paper has subparts devoted to cost, prisoner quality of life, and recidivism. On cost, the author notes that cost comparisons are disputed, but after all there’s more to life than cost: “the focus on short-term monetary savings and the pressing need for prison services and facilities shifts the focus away from other externalities.” Fair enough.

But on prisoner quality of life, she similarly hedges: “According to some observers, quality of life in private prisons is worse than in public prisons, due to lack of training and cost cutting. Specifically, there may be more violence in for-profit prisons than in public prisons.” (My emphasis.) And while she cites a small handful of studies, she also (commendably) cites a source (Perrone and Pratt) “finding inconclusive results on safety when comparing private and public prisons.”

Then she gets to recidivism, where she comes down on the side of finding that private prisons are worse on recidivism. She relies on two recidivism studies in nn.51-52, one of which finds (Lanza-Kaduce et al.) that private prisons are better on recidivism. The Spivak & Sharp study does find that private prisons are worse, but the authors themselves note reason for skepticism about this result (pp. 503-04 of their study). The Lanza-Kaduce et al. study admittedly has various flaws, but what about some of the other papers, like Farabee & Knight, or Bales et al., which found no statistically significant difference between recidivism rates at public and private prisons? For a discussion of the recidivism studies, I’d recommend pp. 357-60 of my paper, which concludes that there’s not much we can say definitively on the subject.

Bottom line: I’m all in favor of using incentives to improve prison quality, though I don’t see the need to focus specifically on private prisons: public prisons could use improvement too. Accordingly, my own paper talks about how to improve prisons as a whole — though the specific incentives I suggest might be easier to implement in the private sector. The comment author’s suggestions, too, might work better in private firms, to the extent they work with the profit motive. But that’s not a reason to single out private prisons as the unique (or even the greater) offenders, which is probably more than the data allows us to say.

2. Part III of the paper proposes a specific incentive mechanism to improve private prisons: refundable tax credits for Private Prisoner Rehabilitation (PPR). (A refundable tax credit is essentially a subsidy — though the author says that tax subsidies are more politically palatable than direct expenditures.) Here’s what the author says about PPR credits:

The PPR credit consists of the following features: (1) funding through the tax system; (2) delayed conditional payment; (3) performance goals; and (4) third-party assessors. It would function as follows: Through the tax codes of participating governments, private prison companies could claim the various PPR credits upon meeting specific, tangible, state-mandated benchmarks. For example, a benchmark might consist of a five percent annual decrease in rape incidents, a five percent annual decrease in prison assault incidents, increased employee training through a state-certified program, implementation of transitional programming that sixty percent of inmates attend with eighty percent of them meeting certain educational goals, or other specific goals. Thus, the PPR credit allows the government to influence a private prison company’s “double bottom line.”

So far, so good — this looks similar to what I advocated in my own article, though I focused on putting performance measures into private prison contracts (or, in the case of public prisons, making them part of public prison officials’ compensation). What does the author say about why this is better than contracts? (The author also discusses the advantages of her proposal over “social impact bonds”, which I also discussed in my paper as an alternative financing mechanism.) Here’s what she says:

Some academics have argued that the contractual negotiation process is an adequate tool to remedy the ills of privatization. In essence, payment can be conditioned on private prisons meeting certain goals, such as lowering the number of sexual assaults. Unfortunately, this solution has failed to meaningfully reduce recidivism or improve prisoner quality of life.

Contractual agreements have been inadequate for two main reasons: First, contracts tend to be incomplete, and second, there is a lack of adequate oversight. Private prison contracts typically describe the general services sought, but do not dictate how that service should be provided. This latitude may allow private prison companies to take shortcuts in order to provide services at the lowest possible cost. In addition to lack of enforceable terms in contracts, agency capture and political pressure may prevent the government-customer from effectively monitoring compliance. In contrast, the PPR credit shifts the burden of proving satisfactory performance onto the private prison company; while the government-customer sets the benchmarks for acceptable performance, the private prison company must show that it has earned the PPR credit.

Stating that “[c]ontractual agreements have been inadequate” (in the past tense) is a bit problematic: as I discuss in my own article, performance-based compensation has barely been tried. A fairly small performance-based compensation component was recently used in a federal demonstration project at the Taft facility, and the UK has recently started experimenting with the concept. It seems a bit early to declare this a failure.

Moreover, that contracts are incomplete is true — but this is similarly true of the benchmarks that will be written into the author’s proposed tax code. Whatever performance-based measures are used there could be exactly duplicated in a contract: why not instead have a state or federal statute mandating the performance measures and form of performance-based compensation? Such a statute will be just as incomplete as the author’s proposed PPR credit.

And what of monitoring? Because of “agency capture and political pressure”, there may not be effective monitoring. True, but what about the PPR credit system? The author proposes the following third-party system: “For the PPR credit to function effectively, assessments must be conducted by an independent third party who is not subject to capture and who would provide accurate information.” Nice work, if you can get it. Why won’t that independent third party be subject to capture? Alternatively, why can’t that independent third party monitor be written into contracts?

This is the lesson of the “incomplete contracts” theory in economics: contracts are usually incomplete because of the pressures imposed by reality, like an inability to perfectly observe, verify in court, describe, or enforce. Changing the locus of regulation from private-prison contracts to tax-code incentives doesn’t solve the problem, unless there’s some reason to believe that the IRS and its third-party monitors are more trustworthy than the federal Bureau of Prisons or state Departments of Corrections. But I didn’t see any discussion of this in the paper.

To my mind, the main advantage (or disadvantage, depending on how you look at it) of a tax-code-based solution is something I hadn’t thought of before. If this is written into the Internal Revenue Code, this means the federal government can incentivize performance even at state private prisons in ways that the states aren’t doing. In fact, even if states are doing performance-based compensation, the federal government can use the tax code to undo state incentives and encourage completely different behavior. Of course, there’s nothing that says this has to be done through the tax code — any subsidy or tax sensitive to performance measures will do. The point is that I had assumed that state agencies would use their contracts with private prison firms to set those firms’ incentives, while the Bureau of Prisons would do the same with the firms it contracts with. But there’s no reason the federal government couldn’t monkey with everyone’s incentives, state or federal. (Whether this would be a good idea is another matter.)

Bottom line: This is an interesting paper on an important topic, which can usefully be read in conjunction with mine. I’m not convinced that the pure private-prison focus is appropriate, nor am I convinced that the tax-code approach is better than the contractual approach, but these are interesting issues to think about.