Are Sex Offender Laws Backfiring?

Photo: pena2(A map of registered sex offenders in L.A.)

A pair of new studies raise questions as to whether sex offender registries and community notification laws actually reduce recidivism of sex offenders, or even lead to lower sex crime rates overall. Both are published in the University of Chicago’s Journal of Law and Economics.

The first study by Jonah Rockoff of Columbia Business School, and J.J. Prescott, a law professor at the University of Michigan, parses out the effectiveness of the two basic types of sex offender laws. While they find that the registration of released sex offenders is associated with a 13% decrease in crime from the sample mean, public notification laws proved to be counterproductive, and led to slightly higher rates of sex crime because of what the authors refer to as a “relative utility effect”:

Our results suggest that community notification deters first-time sex offenders, but may increase recidivism by registered offenders by increasing the relative attractiveness of criminal behavior. This finding is consistent with work by criminologists showing that notification may contribute to recidivism by imposing social and financial costs on registered sex offenders and, as a result, making non-criminal activity relatively less attractive.
…[C]onvicted sex offenders become more likely to commit crime when their information is made public because the associated psychological, social, or financial costs make crime more attractive.

The second study comes from Amanda Y. Agan, a PhD student at the University of Chicago, who throws water on the whole notion that sex offender registries work in the first place.

Agan compared arrest rates for sex crimes in each U.S. state before and after registry laws were implemented and found no appreciable changes in crime trends following the introduction of a registry. As for recidivism, she looked at data on over 9,000 sex offenders released from prison in 1994. About half were released into states where they needed to register, while the other half did not need to register. Agan found no significant difference in the two groups’ propensity to re-offend, and that those released into states without registration laws were actually slightly less likely to re-offend. Analyzing Washington D.C. census data, Agan went block by block and found that crime rates in general, and sex crimes in particular, do not vary according to the number of sex offenders in the area.


If registries don't protect our kids, eliminate them. Parents opposing their closure don't have their kids' best interests in mind. It wouldn't be the first time (See: Vaccines). The problem is pride. They will ask themselves "If all my hard work actually put my child in danger, then what meaning was there to my suffering?"

Never let someone ask this of themselves, for their conclusion will be self serving. In this case, she will conclude that her suffering wasn't meaningless, and that the scientists are simply biased.

See this coming, and stamp it out immediately.

Enter your name

Are kids the only consideration? If checking a registry stops an adult woman from dating a man previously convicted of aggravated rape, or a nursing home from hiring a janitor who has been convicted of molesting people, wouldn't that also be useful?


No, it would only be useful if it prevented said convict from COMMITTING another crime. People are allowed to date people or getting jobs once they've served their sentence.


Was there really anyone under the delusion these laws were intended to make us safer?

Joshua Northey

Ditto, there were people who thought these served some function? They are transparently sops to parents insecurities.


The point isn't to make communities safer, it's to give us information on our neighbors. I certainly want to know if Mohammed from across the street is just friendly, or if he 'likes' kids and has a record of 'liking' kids. Lets not protect criminal rights (privacy) at the expense of law-abiding decent citizen's 'right to know' their neighbor is a creep and they better keep a closer eye on the kids.

Joshua Northey

Your wife, brother, or coworker is almost certainly a greater risk to children then "Mohammed". I would worry a lot more about that. When is "Mohammed" having unsupervised access to your children anyway?


To back up Joshua's comment, the US Department of Health and Human Services has a few graphs documenting who's most likely to be responsible for the maltreatment of children (including sexual abuse) - see here and here.

Enter your name

I wonder if there are differences between lifetime registries and short-term registries. Imagine the stupidest application of sex offender registries: the boy, who is barely past his 18th birthday, has 100% consensual sex with his long-time girlfriend from school, who happens to be one month under whatever age limit the state has set. (I've met someone whose mentally disabled [below average IQ, but not actually mentally retarded] son is serving a 10-year prison sentence for exactly this. These cases aren't usually prosecuted, but they are occasionally.)

If you put someone like this on a sex offender registry for life, then the cost of further offenses, in terms of the registry, is zero. (What are you going to do? Really, really put him on the registry for life this time?) But if it expired after five or ten years, then the convict might be motivated to avoid re-offending so that he could get off the list.

Different states have different approaches to these registries. It would be interesting to know whether some approaches were more effective than others. For example, if minor, first-time offenses (e.g., drunken students going streaking) don't result in registration, but second offenses do, might we see fewer second offenses? Might we see greater willingness to prosecute minor cases, or greater willingness to plea bargain? (I suspect that now, we have people going through expensive jury trials because a plea bargain guarantees lifetime registration, and there's always a tiny chance that the jury will not convict you.)



I have three weird thoughts about all of this....

1) There is a difference between actually re-offending and GETTING ARRESTED for re-offending. I'm thinking that if a sexual offender is in a state where he must be registered with the police department, then if/when certain sexual offenses take place, they are going to be looking for him first.

But in a state where he does not have to register (and if I were a sexual offender, I would move to one of those states--as I'm sure many others already have--which may mean there are more sexual offenders there per a square mile than elsewhere), it might be that a sexual offense does not trigger any particular alarm about HIM...and so he doesn't get the same scrutiny.

2) It MIGHT be that offenders, believing they have suffered enough by being identified, the target of gossip, and considered the town pervert, believe they "deserve" a "treat," and so act out. It happens in corporate offices, as well. A person who didn't get promoted feels they deserve better...and so feels that it's quite alright to work a little less hard, to leave a little earlier than normal, to take shortcuts, etc. That just seems to be a possibility.

3) It is indeed wrong to lump an evil, wicked predator with the teenage boy who, "in love" with his 16-year-old girl friend, doesn't wait. Or to lump the child molester with or rapist with the lonely man who tried to pick up the wrong prostitute. It's all sexual. It might even all be immoral. But some of it is dark and evil. THAT is the person I want to know about. I don't care if my neighbor tried to hook up with a woman in town one night. That's not my business. But the one who just might harm my family, I WANT TO KNOW.

And do not tell me that ME knowing about the pervert next door won't make me safer. It's has GOT to be safer than not knowing.


Joshua Northey

It cannot make you safer because you are not really at risk.

Don't tell me a registry identifying everyone who ever hit someone else in grade school won't make me safer. It has got to make me safer.

Don't tell me a registry identifying everyone who doesn't attend church won't make me safer. It has got to make me safer.

Don't tell me a registry identifying everyone who owns a gun won't make me safer. It has got to make me safer.

Personally I would be more interesting in a gun owner registry then a pedophile registry. I suspect my children are in more danger from gun owners (through accidents) then they are pedophiles.

How about we just skip the registries unless there is a clear benefit?


"Don’t tell me a registry identifying everyone who doesn’t attend church won’t make me safer."

Someone already has this idea: (Although I suspect they are just publicity seeking.)

Going back to the main topic: I assert than any law named after a person (e.g. "Terence's Law"*) is a bad law.

* "Terence" was chosen at random, and does not specifically refer to an actual "Terence's Law", if there is one.


I'm a mom. I assume everyone is a sex offender and potential molestor. I mean, not really, but I don't allow unsupervised access to my children. Most people are not out to harm my kids, but if my mother married a man whose own children didn't have anything to do with him, you better believe my mother wouldn't have her grandkids in her house without me. If I were widowed, I probably wouldn't seriously date until my children were grown. My husband will be the boy scout co-leader when the day comes. And so on...we're not paranoid, but we're cautious. Just because someone is not on a registry doesn't mean he or she is not a molestor who just hasn't been caught.


The problem with sex-offender registries is that they do not (to the best of my knowledge) differentiate between different types of sex offenses. As another commenter noted, because so many different things can be consdered to be sex offenses in different states, many people end up on sex offender lists for actions which clearly pose no threat to other members of a community.

Probably the only sex offenders who are truly threats to other members of a community are those who have been convicted of offenses involving non-consensual sex with total strangers.


Sadly, politicians do not consider factual information before forwarding these laws that the facts prove either do not work or actually do more harm than good.

Lesley Hatch

Actually, the statistical signifigance was not given in this write up, and unless the significance is very high, which is of course correlational, not causal, these studies are irrelevant for the following reasons: 1) it's well established that sex offenders are afflicted with a pathology that is for the most part not curable, and their rates of recividism are high no matter what treatment they receive, and 2) the registries are not for the offenders. They are for the nighborhood, for parents and individuals to react/alter their behavior if they choose. No one who has children in the neighborhood is concerned with the outcome for the sex offender. But like all things related to parenting, it's up to the parents whether or not they do anything about keeping their child from being the next victim. For the parent or individual who alters their behavior and consequently is not the next victim, then the registry worked- and this is the only thing for which registries were implemented. The consequence(s) to the sex offender is irrelevant, unless there is an overwhelming amount of evidence that they increase their behavior significantly, *as a result of* the registries, they should continue to be implemented nation- wide.