Sex offenders and recidivism

Every day it seems we hear a story about a young girl who was murdered by a sex offenders who had been released from prison. The editorial implication is often that the murderer should never have been released. The problem with this type of analysis, however, is that it rarely discusses actual studies on recidivism. Of course, if we *know* for a fact that a particular offender is highly likely to murder when released, then it would be hard to argue that he should ever be released. However, we never know for sure which offenders will murder and which will never offend again. Some might say err on the side of caution, and incarcerate all sex offenders for as long as possible. The problem, though, is that so many people are classified as sex offenders that we'd be incarcerating a far larger percentage of our population than most people would ever suspect. The vast majority of these people will not murder when they are released, I'm confident of that, and many will be productive members of society.

Eugene Volokh has this interesting post on the RRASOR ("razor") test, which is a simple screening of sex offenders to determine their likelihood of recidivism. Any of the following factors are deemed to increase an offender's likelihood of recidivism: youthfulness, male victims, multiple offenses. What Volokh, a law professor, wonders is, if the parole board takes this test into account in determining who will get parole, does it violate equal protection of law?