At what point does more punishment reduce rather than enhance public safety?

I teach a couple of courses on the sociology of punishment and I usually end them with some variant of the above question. There are a number of answers to it — punishment policy with respect to sex offenders is an obvious example. I gave this lecture a few nights ago and was deeply saddened to see this story this morning. A young girl was allegedly assaulted and killed by a homeless sex offender living in a field near her home. What could we have done to prevent this?

We could have required him to register as a sex offender so that the girl’s parents knew his whereabouts. We could have monitored him with a GPS tracking unit. We could better assess the likelihood that he was dangerous and monitored him more closely. We could pass a law restricting where sex offenders can live and work to keep them away from children.

Except… all of these things were done.

Darrin Sanford, who has reportedly confessed to the crime, was registered as a Level 3 sex offender (denoting him most likely to re-offend). He was listed as homeless in the sex offender registry. The young woman’s parents were aware of the transients spending time in the field nearby and had warned her of the danger, specifically highlighting sexual assault as a risk. Sanford was required to check in daily with a probation agent (and did so in the days surrounding the girl’s murder). Perhaps most disturbing, Sanford was wearing a GPS monitor when he committed the crime.

Sanford was subject to all of the punishment and control we heap on sex offenders. But, really, what do these sorts of policies do? Do they make us any safer?

GPS tracking is notoriously difficult to implement and, while it may help convict offenders after a crime has been committed, there is not a lot of evidence that it prevents crime in the first place. Laws restricting where sex offenders can live have been successful in increasing the number of sex offenders who are homeless and concentrating them in particular apartment buildings. Homeless sex offenders are harder to monitor and, at the very least, a bunch of sex offenders living together is not a good strategy for public safety. What of the public registries? There is good evidence that sex offender registries are filled with inaccuracies because law enforcement has too few agents and too few dollars to keeps tabs on so many sex offenders. I also worry that our emphasis on labeling sex offenders publicly makes it more likely that they will re-offend.

All of this adds up to the illusion of control over sex offenders, not control itself, and a public that merely feels more safe in the presence of such policies. Feeling safe is good — being safe probably a bit better.