Twenty-five years ago, with activist Patty Wetterling by his side, Minnesota governor Arne Carlson signed America's first law establishing a public registry system for sex offenders. The law, which Wetterling helped to write, aimed to give police a starting point for tracking down sexual predators like Danny Heinrich, the man who had kidnapped, molested and murdered Patty's son, Jacob Wetterling.

In the last quarter century all fifty states and the federal government passed laws establishing, expanding and tightening registration requirements for sex offenders. Proposals to implement and improve registries passed legislative bodies overwhelmingly and with little debate. In many ways, the laws appear to have worked: Sexual violence and child abuse declined continuously since registration began while academic analyses show that the registries are effective deterrents.

That said, the registries aren't an unmixed success. Despite huge publicity attached to Jacob's kidnapping and the search for him, Heinrich didn't get caught until earlier this year, when he confessed to the crime after being nabbed on unrelated child pornography charges. Child predators still make their way into schools and daycare centers. Even people already on registries like Phillip Garrido — who held California girl Jaycee Dugard in his backyard for 18 years and raped her continually — go undetected. While eliminating registration would be unwise and unpopular, in short, the registration could use improvements. Three major ones are in order.

First, states should remove low-risk offenders from the registries. While behaviors like indecent exposure, consensual incest between adults, prostitution, soliciting prostitution, and kidnapping ones' own children as part of a custody dispute are undesirable, they don't lead to people becoming child predators and don't indicate incurable assault. Including these and other low-risk offenses clutters registries and makes it harder for law enforcement to track people like Garrido and Heinrich. A too-big registry is useless.

For much the same reason, children adjudicated in juvenile court — many of whom end up in trouble for having a boyfriend or girlfriend just a few years younger — should never have been placed on registries in the first place. Even amongst those that do commit real assault, their mere presence on registries undermines the supposedly short-term consequences of the juvenile justice system. The long-term consequences for a high schooler who gropes a classmate can be greater than those for carjacking. This is nonsensical. Since these children almost never reoffend, they don't belong on registries.

Finally, we should tailor punishments to the actual dangerousness and behavior of sex offenders rather than popular stereotypes. This means that certain punishments should be eased and others should be made more intense. For example, there are widespread laws that forbid all sex offenders from living anywhere near schools and daycare centers. These laws may be popular, but there's almost no evidence that they actually protect children. Most sex offenders are known to their victims' parents — many are relatives — and many residency restrictions actually result in sex offenders becoming homeless and thus harder to track. Even public notification laws tend to reduce home values near where sex offenders live without actually increasing community safety. Indeed, no sex offender has ever been caught solely because of public notification laws.

On the other hand, more severe punishments may make sense for the worst of the worst. Among other things, lifetime civil commitment in hospital-like settings may make sense for certain people with uncontrollable urges to molest children just as it is already common practice for those who have similar urges to commit non-sexual violence. Procedures to allow for civil commitment of sex offenders already exist in 20 states and Washington, D.C. Since a small number of sex offenders are nearly impossible to treat, such commitment deserves expansion. Laws to make it harder for schools to "pass the trash" and transfer sex offenders also need to be implemented.

After 25 years, of sex offender registration, the data are in: Registration of sex offenders is a mixed success. It has made us safer but it also could stand for some improvements.

Eli Lehrer is president and co-founder of the R Street Institute, a free-market think tank. Thinking of submitting an op-ed to the Washington Examiner? Be sure to read our guidelines on submissions.