« Broccoli, Liberalism, and Lifestyle | Main | Kleinfeld on Method in Criminal Law Theory »

Saturday, June 16, 2012

In Defense of Cost-Benefit Analysis: Lessons from Recent Rules for Preventing Prison Rape

Lisa Heinzerling has a thoughtful post objecting to the Obama Administration's use of cost-benefit analysis ("CBA") in its recent rules implementing the Prison Rape Elimination Act ("PREA"), 42 U.S.C. 15601 et seq. in a nutshell, Professor Heinzerling objects to the commodification of sexual abuse, in which the Administration attempts to calculate the benefits of avoiding sexual abuse in terms of the victims' willingness to avoid it. According to Heinzerling, "[r]eading DOJ's analysis itself feels like a violation."

After talking with some of the former commissioners who served on the National Prison Rape Elimination Commission and reading the comments responding to the DOJ's NPRM, I have come to a conclusion about CBA and PREA that is diametrically opposed to Lisa's: The problem with the new prison rape elimination standards is not that the Obama Administration indulged in too much CBA but rather that they did not use CBA enough. In particular, by not construing 42 U.S.C. section 15607(a)(3) to allow the costs of proposed prison reforms to be offset by those reform's medical, legal, and disciplinary benefits, the Obama Administration exaggerated the costs and under-stated the benefits of certain reforms, such as banning cross-gender patdowns of male inmates by female officers.

There is a more general lesson here about CBA: To motivate regulation that has substantial costs, one must highlight the regulation's benefits. Because the costs are often much easier to measure than the benefits, if one slights CBA, then one might actually under-regulate.

My interest in PREA stemmed from my final exam for Administrative & Regulatory State. Every Spring, I concoct an exam question based on a real-life problem from administrative law, and, this term, I used the Department of Justice's then-ongoing rule-making to implement the PREA. (DOJ issued an NPRM back in February of 2011). (For the curious, here is my final exam: Download Hills 2012 Exam. Question I deals with the PREA rule-making. I tinkered with the language of the statute to create some separation-of-powers issues, changed the DOJ's position to endorse application of PREA to ICE detention facilities, and altered the NPRM to endorse a partial ban on cross-gender patdowns, but the summaries of the NPRM and comments are otherwise verbatim drawn from the actual administrative record. On the alteration regarding cross-gender patdowns, I cannot help but boast that my alteration of the NPRM correctly predicted the position adopted in the final rule, which bans male officers' patdowns of female inmates but not vice versa).

The PREA is that rare statute supported by both conservatives and liberals in Congress: It was pressed by an odd coalition of evangelically inclined reformers influenced in part by Chuck Colson as well as purely secular civil rights advocates. To win conservative support, however, PREA included a provision barring any prison rape elimination standards that "would impose substantial additional costs compared to the costs presently expended by Federal, State, and local prison authorities.” 42 U.S.C. section 15607(a)(3). One could construe the entire statute, therefore, as an exercise in futility: "Stop prison rape -- as long as doing so does not cost anything."

The National Prison Rape Elimination Commission ("NPREC," an expert body statutorily charged with issuing a report on which the PREA rules must be based), however, adopted a narrow reading of section 15607(a)(3): NPREC recommended that the "the Department should discount from its calculations any costs necessary to bring a particular facility into compliance with its Eighth Amendment obligations and should only subsume within ‘substantial additional costs’ those expenses that the standards impose over and above this level." (See NPRM at 6272). In effect, NPREC urged that, if a prison would have to install a reform in any case in order to avoid 8th Amendment, then those costs could not be "additional" to the prison's "present[]" expenditures -- because the prison ought to be making those expenditures anyway.

The DOJ's NPRM rejected NPREC's interpretation, stating that the NPREC's position was "in tension with the plain language of the statute and is in any event impractical to apply," because "PREA standards will apply to almost 13,000 facilities across the country, operated by thousands of jurisdictions and entities. It is not possible to determine which facilities are ‘constitutionally compliant’ and which are not, in part because constitutional non-compliance often becomes apparent only after the fact—that is, after a violation." (NPRM at 6273).

Numerous comments on the NPRM took issue with this reading of the statute, urging that not only the benefits of Eighth Amendment compliance (in terms of avoided damages) but also the medical and disciplinary savings of PREA standards ought to be factored into the calculation of whether those standards imposed "substantial additional costs." Avoiding prison rape, after all, saves money in terms of avoided trauma and resulting discipline and recidivism problems, medical expenses, and the costs of disciplining perpetrators (inmate and officer alike).

Note that, had one adopted this reading of "substantial additional costs," then the scale of reforms that could be imposed under PREA would be greatly enlarged. rather than focusing merely on the placement of video cameras, segregation of inmates, and quick prosecution of offenders -- minimally costly reforms -- one could also consider re-configuration of the physical structure of prisons, more elaborate training programs, and inmate rehab and education. Of course, all of these possible reforms would be contingent on showing that their substantial cost was so offset by their benefits that they did not impose "substantial additional costs in violation of the statute. But plenty of comments proposed precisely such a defense of more elaborate reforms than the DOJ was prepared to consider.

In particular, NPREC and other commentators urged that the DOJ ban cross-gender viewing and pat-downs of male inmates by female officers on the ground that such surveillance violated the inmates' dignity and encouraged sexual abuse. The final rule adopted by the Obama Administration (like the NPRM) rejected such a ban on cross-gender surveillance of male inmates by female officers, while imposing a ban on male patdowns and viewing of female inmates.

Why? Largely on cost grounds: There are a lot more male inmates and female officers than female inmates and male officers, such the a ban on female supervision of male inmates could cost a lot of money. (See the Final Rule at pages 56-57). Nowhere did the Final Rule consider the possibility that these costs would not be "additional" costs because they should be offset against the medical, legal, or other benefits of the rule.

Was the Obama Administration right to read "substantial additional costs" to exclude offsets? That is an interesting question of statutory interpretation that I served up to my class on the final exam, getting many thoughtful and lawyerly answers. But, going beyond this question of statutory interpretation, would inmates have benefited from an analysis that compared costs to benefits of avoiding rape? Absolutely: The scale of reforms could have been far greater.

Of course, such CBA would require some effort to quantify the benefits of avoiding prison rape. Admittedly, as Heinzerling notes, this can be an unpleasant business that "feels like a violation." It requires the government to quantify the unquantifiable benefits of avoiding crime; It makes explicit what everyone implicitly recognizes -- that our collective desire to avoid even truly horrific harms is always bound by our selfish desire to avoid imposing very big costs on ourselves.

But I am inclined to believe that it is better to "feel violated" by facing up to the need to choose between costly reforms and substantial benefits than to refrain from adopting any costly reforms at all in order to avoid the comparison. Put more generally, sometimes it is impossible to induce judges, legislators, and the voters at large to bear big burdens unless one makes explicit the benefits that such burdens will produce. By reading the PREA to exclude such a comparison, the Obama Administration avoided the need to defend the burdens and enunciate the benefits -- a conclusion that serves our squeamishness about commodification but, perhaps, did not serve the interests of inmates.

Posted by Rick Hills on June 16, 2012 at 11:51 AM | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference In Defense of Cost-Benefit Analysis: Lessons from Recent Rules for Preventing Prison Rape:


Post a comment