Monday, September 17, 2007

Crim & Pub Policy and the Sentencing Policy Institute

A week ago I posted some ideas for a Sentencing Policy Institute, a body preferably housed at a university and composed of a range of scholars pooling their knowledge of sentencing, policy, corrections, administration, communications, and such to produce and sponsor ideas and proposals that would combine all these fields rather than leave them disparate and unaware of each other. Some places, like the Stanford Law Center or the Vera Institute of Justice, whose staffs and product are admirable and accomplished, do parts of what I'm talking about, but for a more direct model, you might want to check out the Santa Fe Institute which does amazing work bringing together disciplines and understanding on complexity theory and its implications. Meanwhile, now what I'd like to do in this post is suggest some possible research and discussion areas for its initial agenda. And, very conveniently, just such an agenda has been offered in the latest issue of Criminology & Public Policy (August 2007, Vol. 6, No.3).

I doubt that the Crim & Pub Policy folks realized that they were making this offering with this issue featuring articles and responses on sentencing reform, sex offenders and sexual victimization, crime & incarceration, and development of a science of prevention, but they did. The first might seem obvious for such an institute, but the latter three only tangential. The first is the more important, but let me note that the latter three sections, with their careful analyses of the inability of juvenile sex offending to predict adult offending, of the truly criminogenic effects of prison on many offenders, and of the payoffs of investment in school-based, family-based, and community-based prevention are exactly the types of research, both methodology and content, that the Institute would promote and disseminate. In the latter, we are talking about research that directly undercuts much of the "thought" behind the major criminal justice and corrections sentencing policies of the day. For some reason, despite all the evidence in history (and our own lives) we act as if things we do as kids are incontrovertibly indicative of how we will act as adults. As if we can take offenders who are yet uncommitted to criminal careers, put them in "crime school" with hard core folks, and then release them "cured" of their criminality. As if we have no experience with prevention programs that works and deny society only actual victimization and harm. Only people resolutely committed to ignoring reality would insist on these paths. But we do because the mechanisms and institutions to challenge conventional "wisdom" are too disjointed and lack authoritativeness. The Institute would be created to overcome this gap between scholars and scholarship and practitioners and practice.

(A good friend reported to me this dialogue during a roundtable at the recent NASC conference. A researcher was insisting that good researchers could never report research that hadn't passed the highest standards and peer review and accepted confidence levels, whatever, despite pleas to get policymakers something better to apply, even if it wasn't ready for publication. It just wasn't done. To which one of the legislators participating apparently replied something along the line of "well, prepare to find yourself ignored." Which is, of course, the answer to the long-term lament of academe on why policymakers still do things so counterproductive and so counter to what research has demonstrated. The Institute would bring in scholars on all aspects of corr sent policy, including its implementation and communication, and link them to the policy world, receiving input directly from policymakers on how to make the findings relevant, understandable, and, most importantly, saleable to the broader community. Among other things the Institute would do, like websites, providing spokespeople to media, applied and theoretical research, maybe my dream of simulating crime, sentencing, corrections, and their consequences in virtual or other dynamic computer modeling.)

The lead section of articles on sentencing reform are, of course, the most immediately relevant to the work of such an Institute, as I indicated. Interestingly, I think the most value comes from the intro and the responses to the actual articles. Not that articles on the impact, still frankly formative, of Apprendi, Blakely, and Booker aren't needed or important. They are exactly the kinds of legal analyses that the policy community, practitioners and scholars, tends to overlook. But these articles tended to cover old ground for those who have been keeping up.

An article by Shawn Bushway and Anne Morrison Piehl, with whom I had the pleasure of working in MD years ago, throws more light on the sort of thing the Institute could do to challenge and move the agenda and thought of our community. They've been doing intriguing work for a while now, along the lines of what their abstract states:

. . . Illustrations from several jurisdictions show that the presumptive sentence guidelines may not have been as effective at reducing discretion and disparity as policy makers seem to believe. In addition, the policy choices embedded in the sentencing grid itself can contribute to disparity, and the grid cand and should be analyzed to assess its role.
. . . these analyses suggest that policy makers need to be more cautious with any conclusion about the relative benefits of presumptive versus voluntary guidelines. Analyses of sentencing variation must vary with the institutional setting. In a presumptive sentencing state, more of the total sentencing variation is likely to be contained in the grid or exercised in the charge bargaining stage (or earlier) compared with a less-structured sentencing system, where appreciable variation may remain at the sentencing stage.

This isn't new for these two, but it certainly wasn't a view frequently presented at that NASC conference (and I was in charge of the program!!). Would have been a little like pointing out at the christening that that really isn't that beautiful a baby. But these views need airing and the challenges need to be made, as the intro and two responses make clear. Nicole Leeper Piquero, for example, notes how poorly guidelines, and corr sent generally, deal with institutional actors as opposed to individual offenders. Pretty much all you'd need to know to guess that guidelines weren't originally interested in crime and sentencing, but in particular types of crimes and offenders and outcomes if you'd never known about them before. Not a mortal wound but one, as she notes, that we would do well to address now.

Actually, the two most valuable pieces in the entire issue for me (and you know how much I had to smile when reading the criminogenic effects of incarceration article) were the intro and the other response. Charles Wellford of the University of MD's crim and crim just program and past member of two sentencing commissions (MD and DC) organized the sentencing reform section and did his usual insightful work. Several years ago at the NASC conference in Williamsburg, Charles sort of challenged the rest of us to take commissions and guidelines onto a higher level than the plateau we had all reached with our abilities to report basic data and issue occasional research and recs. This intro and the articles he engineered for the section reissue that call. Here are just a few of his more provocative statements:

. . . [the work of determining guidelines effectiveness and legal foundation of the last 30 years] generated two problems for research on sentencing: (1) It led to estimates of the effects of the impacts of specific factors on poorly specified models of sentencing; and (2) it diverted our attention from sentencing to sentencing reform--as if we could reform sentencing without more fully understanding it.
The limits we have placed on ourselves in doing research on sentencing were appropriate as long as the question was whether the guidelines work. Now the question is, once again, what determines a particular sentence?
. . . The actions of the Supreme Court may cause problems for some sentencing reforms, but they will benefit sentencing research if they result in research that includes the role of charging in the process, expand the range and depth of variables we test as predictors of sentence (including the identification of the judge and her relevant characteristics), and require us to spend more effort in systematizing our research by developing empirically grounded theories of sentencing.
Imagine if we approached sentencing as a more dynamic system in which we allowed judges to select from a list of factors that could be used in sentencing and had been demonstrated to predict desired sentencing outcomes and met the legal standard for proving those factors existed in a particular case, provided judges with historical data on sentences for those with the constellation of factors selected by the judge, allowed the judge to select a sentence from the range established by statute, and monitored sentencing at the judge level to search for disparity--would there be any departures? I think not.

The last point sets up what I think is the most important work in the issue, although we may have some bias here since it was written by Kim Hunt of this blog and reiterates many of the things he has said here, particularly the emphasis on grounding our corr sent policy in actual evidence rather than the stories and myths that we have been flying by. He argues directly that "if sentencing commissions and their guidelines survive, it will be because they can demonstrate that they are a means to worthy ends, and will continue to improve to demonstrate their relevance."

If you've read this blog long, you're familiar with the basic ideas, but here are a few of Kim's choicest comments:

. . . our standards for effective criminal justice policy analysis are too weak. Justice policies are too often enacted in a rush to judgment, witout a theory of how the program should work and without a rigorous plan for data collection and program evaluation once the program is in operation.
It is difficult for commissions to specify goals, which are multiple and sometimes contradictory. Sentencing commissions are typically large bodies that include multiple stakeholders with a morass of competing expectations, which leads them to political negotiations and consensus building and to the blurring of goals.
However, the costs of failing to pursue evidence-based sentencing policy are greater than the costs of increasing our efforts.
Sentencing problem solving requires, among other things, a theory of sentencing policy change--a testable hypothesis about how sentencing policy innovation is expected to alter sentencing practice. . . . In addition to a clear theory of change, sentencing commissions must use rigorous data collection and analytical methodologies that withstand scientific scrutiny and offer compelling evidence of success or failure. Sentencing commissioners, undertaking one of the largest policy experiments of the last 30 years, have not played the part of scientists sufficiently well.

Kim mixes into these thoughtful comments reference to Frankel's original call for commissions, the inadequacy of evidential foundation for "three strikes" laws, the premature conclusions that prescriptive guidelines have been successful in their goals, medicine's increased use of evidence-based practice, and the Campbell Collaboration's leadership in spreading the latter into other fields. Maybe now you can see why I said he should be on the short list for the scholars who would compose the Sentencing Policy Institute. As he and Charles both hammer home, and as all the other pieces in the issue reinforce, corrections sentencing policymaking and philosophy are roiling with new ideas and evidence from a range of different fields, and practice has leavened that mixture with the need to understand how sentencing and sentencing policy are idealized and implemented and communicated. We can keep doing it piecemeal, waiting for Crim & Pub Pol's next issue on these subjects, hoping that connections can occasionally be made at conferences or workshops, uncoordinated and unguided and counting on persimony to bring us light. Or we can step up to the plate, create one or more bodies dedicated to the task, and start bringing that light more quickly to our enterprise.

Did I give away what I think we should do?

No comments: