Last November, a team of sex offense researchers, led by Kristen Zgoba, director of research at the New Jersey Department of Corrections, published a study. It looked at the accuracy of the system used under federal law (called the Adam Walsh Act) to determine the risk that an offender will commit another sex crime. In brief, offenders are put into one of three tiers according to their presumed risk level: those in tier I are presumed to be at the lowest risk for committing a new crime, and tier III are the highest risk.
Zgoba et al. looked at a sample of sex offenders in four states—New Jersey, South Carolina, Minnesota, and Florida—and linked each person’s tiering level to their likelihood to commit another sex crime. The results are disturbing. Here’s the language from a press release about the study.
“Tier level was not significantly associated with recidivism in New Jersey, Minnesota and South Carolina and was inversely associated with recidivism in Florida—the only state in the study’s sample that has been certified as substantially compliant with the federal Adam Walsh Act.”
The italics are mine. If that sentence about Florida doesn’t bother you, it should—it means that the offenders that Florida classified in tier I had the highest rates of re-offense, and those classified in tier III had the lowest rates.
That’s a problem, because in many states it’s the tier II and tier III offenders whose names, photos, and addresses are put on the public sex offender website and the tier I offenders who aren’t—they’re in a database accessible only to police. If Levenson and company are right, parents who use the public sex offender website to check who’s in their neighborhood are seeing photos of the lower risk offenders only; the people most likely to commit another crime don’t even show up.
This isn’t the first time that researchers have found that sex offender registries get it backward. Zgoba and company point to the results of a 2009 peer-reviewed study in New York that found something similar. In their words: “In that study, lower-tiered individuals had higher recidivism rates than those who were assigned into ostensibly higher-risk tiers.”
The bottom line? “The offense-based classification system adopted by the Adam Walsh Act,” says researcher Jill Levenson says in the release, was developed without empirical validation.”
And why is that? As The City Beat blog points out in this excellent summary of the argument over how to classify sex offenders, it may be because it’s just cheaper and easier to run the system that the feds have set up—a system that classifies offenders based on the crime they were convicted for rather than a risk evaluation done by professionals.
And here’s an obvious question some of you might be asking–isn’t it possible that the tier II and III offenders in Florida were found to have lower rates of committing new crimes because they know they’re under more public scrutiny since they’re on the registry? In other words, don’t the results actually show that the public sex offender websites are keeping the most dangerous offenders from committing new crimes?
Not likely. In Florida, where the federal system worked the worst of the four states studied, all sex offenders, regardless of tier, are put on the public website. So being on the website can’t have been the “active ingredient” that prevented the tier II’s and III’s from committing more crimes.