Ezra notes the forthcoming release of Mark Kleiman’s intriguing book When Brute Force Fails: How to Have Less Crime and Less Punishment. It cuts against our retributionist instincts, but holds out the promise of achieving crime reduction while also having less recourse to prison. Having heard Kleiman talk about his thesis—and the empirical work behind it—I’m eager to dig into the full version. But this also reminds me of a Canadian study I saw mentioned in the press over the weekend, to the effect that juveniles who have been through the juvenile justice system end up much more likely to find themselves on the wrong side of the law as adults—something like seven times more likely—than kids with similar histories but no judicial intervention. The more severe the intervention, they found, the greater the damage.
I should probably go look at the full study, but it seems like there are a handful of different—though not necessarily inconsistent—stories you can tell about what’s going on with. The researchers reportedly go with one of the more obvious ones: Kids grouped with lots of other young delinquents end up enmeshed in a “culture of deviance” where criminality is both taught and valorized. Another possibility has to do with the difficulty of getting accurate reports from kids on their own criminal behaviors. At the very least, one difference in the histories of the kids in the two samples, beyond the intervention of the justice system, is that one group of kids was doing something earlier that attracted the attention of the criminal justice system. Now, you can go out and look for people who didn’t get caught who’ll admit to having committed robbery or assault or drug peddling or what have you as kids. But what I expect it’d be tougher to get from the kids in the system, whatever confidentiality you promise, is an honest report of what other things they were doing up until they got caught. Because you can bet most of them will swear that whatever they got nailed for, it was the first time, honest! But probably not. Even if they were committing precisely the same crimes with precisely the same frequency, getting caught suggests greater recklessness about it, which may in turn suggest deeper impulse-control problems.
More interesting to me is the effect of being branded as a criminal, quite apart from the lessons in criminality one might absorb in juvie. (Martha Nussbaum’s work on shaming punishments may be relevant here.) I’m guessing most of us knew kids back in high school who got in fights or had a penchant for the “five-fingered discount” or dealt a little dope. Some of them went on to more serious crimes. But a lot of them basically grew out of it. And one reason they were able to grow out of it was that, whatever their reasons for acting out in these ways, they didn’t really think of themselves as criminals. They were kids who did some criminal things. Getting away with it means—and requires—preserving a public persona as non-criminal that remains available as the child or teen matures. Formal interaction with the justice system hastens along the shift in self-conception from “I am a person who has done some criminal acts”—this is probably strictly true of the vast majority of us—to “I am a criminal.” This may be the most unsettling possibility, because one can imagine reforming judicial interventions such that, for instance, we didn’t lock up a bunch of young delinquents together to develop a self-reinforcing culture and informal academy for studies in advanced thuggery. But if this “branding” effect plays a significant role, then the fact of intervention itself will tend to cause the problem.
This reminds me of a line from a recent episode of This American Life that provoked a chuckle and a nod at the time. Paraphrasing roughly, it was something like: We routinely accept high school kids doing things to each other that, if you did them as an adult, you would clearly go to jail. And that’s obviously true, but perhaps this study gives us another reason to think that’s not such a bad thing after all.
9 responses so far ↓
1 the teeth // Jul 20, 2009 at 5:40 pm
Interesting post, and I’m pretty much on board with you. I’m a little uncomfortable with one statement:
“Even if they were committing precisely the same crimes with precisely the same frequency, getting caught suggests greater recklessness about it, which may in turn suggest deeper impulse-control problems.”
This is undoubtedly correct, as far as it goes, but I have a strong intuition that frequency is a much better predictor of ‘level of recklessness’ as to whether a particular teen gets caught. And pretty much any sixteen year old who habitually steals or gets in fights will get caught, sooner or later. Drug dealing (and use) is probably a bit different — identical levels of ‘criminality’ can be relatively safe for a levelheaded hooligan, and suicidally risky for one more out of control. In any case, quite separate from getting caught, it seems something like ‘family stability’, followed by affluence, would be the biggest indicators of whether a particular teen goes through the system fully — everybody gets caught, but if you come from a ‘good family,’ you have to be awfully incorrigible to end up incarcerated for any period of time, at least by the state. Which might be almost too obvious and beside the point to deserve explicit mention, in which case: apologies.
2 Julian Sanchez // Jul 20, 2009 at 5:43 pm
I assume (maybe wrongly) they’re controlling for big obvious stuff like that—a kid from a stable, affluent family is both more likely to be shielded if he goes off the rails in his teens and more likely to have opportunities that make a life of adult crime less attractive. But I would be gobsmacked if the researchers hadn’t already factored that in.
3 the teeth // Jul 20, 2009 at 6:32 pm
Just to be clear, my statement wasn’t meant to be any sort of comment on the study — I’m sure that they controlled for affluence & family stability. I guess my point was — attracting the attention of the justice system and actually experiencing the whole lengthy hell it can produce are two very different things. And I feel like the frequency vs recklessness statement is misleading — if you hold frequency constant, recklessness will be more of a determiner, but if you don’t, I’d expect frequency to be much more important. It seems that in practice, you’d either be equipped hold both constant or neither.
Apart from all that, your followup post points to a good question — to what extent does the (assumed) far greater future success of affluent delinquent teens have to do w/ the opportunities available to them once they’re in condition to reform, and how much does it have to do w/ being shielded from the nastier bits of the criminal justice system?
4 Barry // Jul 21, 2009 at 8:05 am
Julian Sanchez :
“I assume (maybe wrongly) they’re controlling for big obvious stuff like that—a kid from a stable, affluent family is both more likely to be shielded if he goes off the rails in his teens and more likely to have opportunities that make a life of adult crime less attractive. But I would be gobsmacked if the researchers hadn’t already factored that in.”
I’ll check it out when it’s available.
There are a lot of methods which can cope with differing factors (matching and propensity scores). If they didn’t use such methods, they shouldn’t have gotten through peer review.
5 stephen // Jul 21, 2009 at 9:14 am
Well the artilce was way too vague about the details of the study to draw much of a conclusion. It doesn’t give the correlation or the r-squared value, nor does it say anything about the treatment. It seems like it was an observational study so there is bound to be some selection bias.
As you pointed out it doesn’t even consider differing personality types or time preferences, which is bound to have some role, perhaps the most significant.
In addition, and I don’t know how it is in Quebec, but in most American cities the kids that are most likely to contribute to a “culture of deviance” are not equally distributed. It would seem that exposure to a “criminal culture” starts long before juvenile detention, granted it would be less concentrated. But maybe the study controlled for that, I don’t know.
At the end of the day I just don’t know what to do with articles like this. Crime predates juvenile detention and overall crime rates have been falling for some time now. I am willing to bet that if juvy was eliminated that we would not see a %600 reduction in adult crime that could be solely attributed to that particular change.
6 Michael B Sullivan // Jul 21, 2009 at 12:36 pm
I agree! But note that even if we accepted that juvie, totally in and of itself, increased your propensity for adult crimes by 7 times, eliminating juvie would not reduce adult crime to 1/6th of the present amount.
Presumably, some (large) number of adult criminals did not go to juvie. Thus, removing juvie would not affect their crime rate. Thus, removing juvie (even assuming that it is solely and causally responsible for the x7 criminality) would reduce adult crime rate by considerably less than a factor of 7.
That said, I’m on-board with what I presume to be your basic underlying point that it seems prima facia unlikely that simply removing juvie would reduce crime in its alumni by a factor of 7.
If we accept that juvie inducts young people into a culture which valorizes crime, surely there are other ways they could be inducted into the same culture — and perhaps those other ways would become more prominent if juvie were not there to scoop up many obvious candidates.
7 RickRussellTX // Jul 21, 2009 at 1:24 pm
Does it really matter whether juvenile convicts have a “self conception” of being a criminal? It seems to me that the important question is whether *everyone else* perceives them to be a criminal. That’s going to be the most obvious externally perceptible difference between “kids who went to juvie” and “kids who stayed out of juvie”, even if they are otherwise matched in criminal frequency, recklessness and socio-economic conditions. Once their friends, family, teachers, local businesspeople, etc. are aware of the “juvie kids”, it’s going to color all subsequent interactions.
We all know how hard it is for an adult to succeed after a felony or even a high-class misdemeanor. As the Gina Grant case showed (http://en.wikipedia.org/wiki/Gina_Grant), your juvenile record may be “sealed”, but that doesn’t prevent people, including law enforcement, from blabbing about you all over town.
8 stephen // Jul 21, 2009 at 2:59 pm
Michael
You are right, I was sloppy. It was early when I wrote that.
I was also being somewhat contradictory by stating the article was too vague about the study and then proceding to complain about the implications of a study I admitted I had too little information to form an opinion about.
Although my priors lead me to doubt the results, my real beef was with the article itself.
9 Mike // Jul 26, 2009 at 12:35 am
It’s interesting that this occurs across the line of legal maturity. I assume juvenile criminal records are sealed in Canada as they are here…(?)
That slightly cuts against the intuition I want to suggest, but I’ll share it anyway. In my mind, one potential driver of recidivism in the U.S. is the extreme stigma that attends a conviction, or even an arrest of any kind, and the sensitivity of employers show to a record reflecting that history in hiring decisions. As the extent of a person’s record increases, his employment prospects dim in proportion, and crime will increase in attractiveness. Are we extending criminality further in some cases by our oversensitivity to past offenses?