Playing by the Numbers

Article excerpt

My friend L., a magistrate in Britain, is appalled by American-style sentencing, which has taken hold there recently. "So you have this young man who's been in and out of minor trouble before--public drunk, throwing rocks--but he's grown up a lot in the last few years, has a wife and a baby and a job. But he falls off the wagon, gets drunk, hassles a neighbor. He needs alcohol counseling. He needs it to keep his job and feed the baby. He needs it, his family needs it, society would be better off, clearly. But I can't order counseling. The guidelines say we're to imprison him for six months. What nonsense is that?"

L. was referring not just to the kind of determinate sentencing guidelines that have helped swell our own prison ranks to the most populous in the world. He was also concerned about the particular variation that exists in Virginia and whose ethical ramifications were recently the subject of a New York Times Magazine article titled "Sentencing by the Numbers." Punishment is premised on probable propensity for crime, which in turn is figured by taking into account age, gender and socioeconomic class--a sentence more severe if you're young and male, less so if you're an old lady. Criminal sentencing, in other words, begins to resemble the kinds of actuarial tables used by insurance companies.

Trials are always governed by some measure of probability--burdens of proof are all but labeled as such, i.e., preponderance of the evidence, clear and convincing, beyond a reasonable doubt. But the guidelines to which my friend was referring are much broader: Penalties are doled out according to a fixed scale assigning outcomes based on sociological and psychological predictions that this one is a bad egg, that one relatively harmless. This trend toward limiting discretion in sentencing is designed, ostensibly, to root out the softies in the ranks of the judiciary, but whatever the imagined merits of that, it is troubling.

First, it resembles a form of prior restraint, which is illegal. Second, this removes an essential component of what makes the law just. A judge's discretion to sentence within a given range of maximum and minimum sentences is where mercy resides. Yes, we could punish all bad acts exactly in the same manner, no matter what the circumstances, but that would make justice not just blind but a little stupid as well. We have historically trusted judges enough to allow them to consider mitigating circumstances for even the admittedly bad actor. But it is a matter of trust, and we do seem to have become a society in which trust is in rather short supply. Third, this sort of econometric reasoning literally imprisons individuals in status stereotypes. I am curious, too, about why so many reject the use of sociological or other group data for inclusive purposes like affirmative-action plans that enhance economic or racial mobility, yet seem unfazed by requiring it in criminal law, where the exercise of state force ought always to be most particularly scrutinized, narrow and cautious.

There seem to be whole bevies of bushy-tailed number crunchers in high places these days, most of them economists--people so obsessed with the elegance of their models that they sometimes forget morality and norms, to say nothing of common sense. …