So You Want to Structure Sentencing? Part I

The previous seven part series discussed ingredients of a good commission. Structuring the sentencing decision, a separate consideration, is outlined here.

Mike’s previous series “So You Want to Start a Sentencing Commission?” emphasizes the many functions of sentencing commissions in addition to, or even “instead of,” the construction and maintenance of sentencing guidelines. Nonetheless, judging from the “flood” of comments from our reader(s), we need to spend a little more time discussing the decision to structure sentences.

“Structured Sentencing”
What is commonly called Structured Sentencing (sentencing the world over has some structure, right?) is a relatively new development, and still not the norm in the 50 US states and the District of Columbia. The indeterminate sentencing model -- broad authorized sentencing ranges with great judicial discretion, substantial parole discretion, and case-by-case decision making – was nearly universal in the US in 1975 and is still common. Judge Marvin Frankel in Criminal Sentencing: Law Without Order argued for a system of sentencing based on the rule of law, including some form of “detailed chart or calculus” to reduce or channel judicial discretion in individual cases. So what we will call Structured Sentencing seeks to channel judicial discretion in individual cases by providing a recommended disposition for each offense (e.g. prison, probation, or intermediate sanction) and a proposed sentence length or range, at least where a prison term is the recommended sentence.

Not all structured sentencing systems are sentencing guidelines (usually thought of as judicial limits fashioned with input from a broadly-representative sentencing commission). Some state legislatures refined statutory limits on sentences rather than adopting sentencing guidelines. In these jurisdictions, legislatures narrowed the sentencing range within which a judge can sentence an offender by 1) setting a new maximum sentence, 2) setting a new sentencing range, including both a minimum sentence and a maximum sentence, or 3) setting a specific presumptive sentence with departure principles. These statutory limits serve to narrow judicial discretion in individual cases, and are mandatory for judges to follow. As recently as 2002, at least five states had this form of structured sentencing without instituting guidelines: Arizona, California, Illinois, Maine, and New Mexico. (Anybody want to help update this? Are you out there?) However, these systems run the risk of being rather clumsy, and not channeling discretion in important ways.

“Sentencing Guidelines”
But sentencing guidelines are the most common form of structured sentencing in the US. Many states and the federal system use presumptive or mandatory guidelines, which limit or take away the judicial discretion to impose a sentence outside of the guidelines. Other states and the District of Columbia opted for voluntary or advisory guidelines, which guide the judicial selection of a sentence, but do not require the judge to sentence within the recommended range. Guideline systems may also be descriptive (and deliberately mirror past sentencing practices) or prescriptive (and rely more on normative principles to set sentence ranges without rigid adherence to historical practice). In a 2002 report, the DC Sentencing Commission identified twenty-two states and the federal government that have a sentencing guidelines system in effect (18 states) or under study (4 states): Alabama, Alaska, Arkansas, Delaware, Iowa, Kansas, Louisiana, Maryland, Massachusetts, Michigan, Minnesota, Missouri, Nevada, North Carolina, Ohio, Oklahoma, Oregon, Pennsylvania, South Carolina, Utah, Virginia, and Washington. (Again, a little help with an update, please) Similarly, in a fine 2005 review of state guideline systems, Prof. Richard Frase identified 21 past and present guideline systems. See M. Tonry and R. Frase, Sentencing and Sanctions in Western Countries, for a more global perspective.

“Where do I start?”
Whether descriptive or prescriptive, advisory or presumptive, most structured sentencing systems are well advised to acquire (by whatever means necessary - usually involves begging) data on current practice. This falls under the category “You can’t know where you’re going, if you don’t know where you’ve been.” There is no substitute for accurate or even semi-accurate data on prevailing sentencing practice with respect to the in-out decision (Is a prison sentence the norm for similar cases?) and the sentence length decision (If prison, for how long?). Probably nothing will jeopardize a fledgling guideline system faster than naïve sentence recommendations that unconsciously and radically alter the mix of probationers and prisoners while ignoring prevailing practice (“the going rate”) – I am not talking here about prescriptive guidelines that attempt to consciously and with near-surgical precision alter prevailing practice for selected crimes, such as armed crimes of violence.

Some folks have tried to make-do with hypothetical exercises in place of actual data on past practice, in place of actual data. Bad idea! D. Kahneman and A. Tversky became famous (Nobel-prize famous) for their work on decision making under uncertainty, among other things. They found that even experts make intuitive predictions and judgments with severely flawed strategies if they don’t base their strategies on existing data and probabilities. For example, people tend to ignore the typical case, and base their judgments on their recall of the atypical case, which stand out in their mind precisely because it is out-of-the-ordinary(and thus a poor general guide)….

In addition to a baseline providing by good data on prevailing practice, the purposes of sentencing must be discussed, and a mission statement developed. I will not revisit this issue, so ably covered by colleague in “So You Want to Start a Sentencing Commission? Part V.”

Next time: To Grid or Not to Grid.