Chisholm’s Take on Criminal-Justice Reform

I posted last week on a recent op-ed by Milwaukee County Sheriff David Clarke opposing so-called “smart-on-crime” initiatives.  Clarke may have been responding, at least in part, to Milwaukee County District Attorney John Chisholm’s Feb. 11 address at Marquette Law School.  (Chisholm’s text is here, and video is here.)  Whether or not Clarke meant to take on Chisholm, it is clear their respective approaches differ quite dramatically.  I’ll share some thoughts in this post on Chisholm’s proposals, which I generally find more appealing than Clarke’s call for mass incapacitation.

Chisholm advances three proposals.  First, “local safety officials must continue to adopt evidence based practices at all levels, from the time a person has contact with a police officer until the time they are released back into the community.” 

Stated at this level of generality, it’s hard to argue against “evidence-based decisionmaking” — there’s too much at stake in criminal cases to feel comfortable with decisions based merely on untested intuitions.  But EBDM can encompass a lot of different specific reforms, and I don’t find all of them equally attractive.  For instance, Chisholm seems to tout both universal screening of defendants at booking (which involves the collection of risk-related information that would then be made available electronically to all decisionmakers later in the process) and drug treatment courts.  While universal screening seems quite helpful, I do have my reservations about drug treatment courts.

The more general concern about EBDM — and I think this relates to some of Clarke’s criticisms in the op-ed — is that the “evidence” is unlikely ever to give us clear answers to all of our policy questions.  In a world of great social and institutional complexity and limited research dollars, there will always be large areas of uncertainty in the evidence, which means that presumptions and default positions are crucial in EBDM.  In fact, if you think about it, evidence alone doesn’t really answer anything — you first need to have a system of social values in place in order to weigh the evidence.  In that regard, I’ve criticized some proponents of EBDM for buying into a risk-reduction value system that does not sufficiently recognize the value of propotionality in punishment.  In my view, we should be careful about losing sight of the “justice” in “criminal justice.”  Putting justice values first may sometimes mean relatively harsh punishment for “low-risk” offenders and relatively lenient punishment for “high-risk” offenders.

Chisholm’s second proposal is this: “to allow judges in the State of Wisconsin, based on the best evidence available at the time of disposition, the option of imposing either a determinate or indeterminate sentence.”  If we are going to have a criminal-justice system in which risk-reduction is a primary objective, then this proposal makes a great deal of sense.  It would mean a partial rollback of the “truth in sentencing” principle, but that may be warranted in some cases.  Here’s how Chisholm puts it:

Truth in Sentencing is a powerful tool for incapacitating serious violent offenders. Truth in Sentencing, is not as effective for offenders who are low to medium risk and who are in need of timely effective intervention at the local level because it backloads services to the time shortly before release, creating greater likelihood that supervision in the community will fail. Indeterminate sentencing allowed corrections officials to prioritize resources for re-entry to the community based on parole eligibility at 25% of the maximum sentence, and provided incentives for actors under supervision to participate in programming early in a sentence.

Although the proposal is framed in terms of efficient risk-reduction, I do not think indeterminate sentencing is necessarily incompatible with a proportionality-based approach.  In fact, that is precisely the subject of my most recent article.

Chisholm’s final proposal is what he calls the Community Justice Reinvestment Act.  The basic idea is to provide incentives and resources for local officials to develop programs that reduce the flow of offenders into state prisons:

The current cost per inmate per year in the state correctional system is approximately $30,000. This proposal suggests that participating counties would receive $15,000 for every fewer inmate sent to prison. The state would retain the other $15,000 for either DOC services or for funding non-DOC programs. Imagine what could be done with the resources if we reduced admissions from Milwaukee by 500 to 1000 in a year.

Implementation would be quite tricky, but this does strike me as a truly compelling idea.  Given the current structure of corrections financing, there does seem to be a real incentives problem as local decisionmakers confront the question of whether to send an offender to prison or to try a community-based approach.  In economic terms, sending the offender to prison almost entirely externalizes the costs of dealing with the offender, while keeping the offender around means that he remains a burden on local resources.  Because of the mismatch between who makes the decision and who pays for it, there will be a natural tendency to overuse imprisonment.  Chisholm’s CJRA proposal would at least partially rectify the problem.

However, it might be quite a bit more complicated than Chisholm suggests to calculate a fair amount to be returned to the local level.  For instance, using the average cost per inmate does not accurately indicate the fiscal impact of reduced prison admissions.  Since prisons have large fixed costs, an institution could experience a sizeable decline in its population without benefitting much on the budget side.  On the other hand, if inmate reductions reached a tipping point that would permit closure of an institution, the savings could indeed be enormous.  In principle, the financial benefits to a county of sending fewer inmates should be based on marginal, not average, costs per inmate.  This will be harder and more controversial to calculate, though, especially when it comes to figuring out who should get the “credit” for institutional closures. 

But the practical challenges should not obscure the important point: there should be more balance in the incentives when imprisonment decisions are made at the local level.