http://www.washingtonpost.com/wp-dyn/content/article/2007/07/07/AR2007070701073.html?hpid=topnews
The gist is that US and other countries’ crime rates have a remarkable correlation with the use and banning of lead in paint and gasoline. A bit more persuasive than the Levitt abortion theory.
A bit more persuasive than the Levitt abortion theory.
Really? So you think it’s more likely that children were turned into criminals because of exposure to lead-based paints than children became criminals because they unwanted and unable to be well-cared for by their mothers or parents?? I’m surprised no one’s blamed it on fast food yet…
It’s more persuasive because the R2 was between .66 and .90. Those are pretty remarkably high correlations for a social science study. The fact that the results were echoed across different countries at different times suggests that its not a spurious correlation.
I wasn’t unconvinced by the Levitt abortion theory either. It’s just that the numbers here are pretty remarkable.
I thought crime dropped in the 1990s due to the COPS program. Isn’t violent crime on the rise nationwide? It sure is here…must be a lot of lead being uncovered in those old houses…
Correlation does not prove causation. So for any 1990’s crime reduction theory they have to go past just an R^2.
I think it was Barney. That stupid purple dinosaur had to be good for something, right?
That’s an interesting theory. I’ve had a chance to work with HUD’s Healthy Homes and Lead Hazard Control group a few times. While there are always opportunities for improvement, that group actually does a pretty good job identifying and taking care of lead hazards in homes.
I didn’t see it in the article you posted, but even small amounts of lead ingested by infants and toddlers results in a 5% or larger reduction in IQ later. Larger doses have larger impacts on IQ and other development. And, my own bias is that a lot of criminal behavior is often the result of lack of intelligence (except of course when you have those evil geniuses in Bond movies).
The article mentioned lead in gasoline, which is a good thing to get rid of. But a key thing is getting rid of lead in older homes. As paint degrades it chips, especially in areas like windowsills that toddlers grab onto and then put into their mouths. Unfortunately, many of these homes are lower income and often rental properties. So you add poverty, likelihood of less involved parenting, and then throw in the mix lead poisoning that reduces natural IQ. Just a bad mix all around, but there are a lot of folks dedicating their lives to fixing the problem.
Um, sure, but when you have well-known lead impacts on brain development and behavior, a high R^2, and similar statistical results generated from different countries during different time periods, its certainly a lot more convincing than anything a politician may say.
Well if the claims of a politician is the only measure/threshold we have for verifying a theory, then let’s have at it. But that’s not how we test out theories, let’s not muddy the waters on that front.
Lead exposure is clearly a horribly negative thing, but before we start funding lead clean-ups to prevent crime we should have a better understanding than an R^2.
I don’t think the point is that the study should have us fund cleanups, or at least that’s not how I read it. Lead is no longer used in paints or gas and hasn’t been for 30 years. I think the point is that if you want to understand why crime patterns have gone as they have, this is a pretty interesting hypothesis. And it is certainly applicable to other less developed but emerging economies.
While a strong, or extraordinary in this case, R^2 is certainly not enough to qualify a theory, there hasn’t been anything else approached with this level of rigor that has come even close to generating similar results. This, on top of the geographical and time diversity of similar results is pretty interesting. But the fact of the matter is that it is impossible to prove or disprove an allegation because the number of moving pieces is so large and gives opportunity to those who want to deny, or for that matter, those who wish to take credit. So Giuliani can take credit if he wants. I think his problem is that he would have to explain why crime went down precipitous in the country overall and in almost every large American city, which doesn’t really follow from the fact that he was mayor of NYC.
Lead’s still used in children’s toys and many other things… Remember, cheaper is better, always low prices…
http://www.google.com/search?source=ig&hl=en&q=lead+recalls&btnG=Google+Search
I don’t think the point is that the study should have us fund cleanups, or at least that’s not how I read it. Lead is no longer used in paints or gas and hasn’t been for 30 years.
Fatmouse’s excellent point was that lead still remains in many older homes, which in urban areas tend to be where the low income live, and they don’t have the means to clean up their houses. So if lead exposure were shown to be one of the most important factors contributing to crime levels, I’d say we might want to explore a public policy approach to this. Maybe not, but it’d be worth exploring.
And in general, I’m all for studies of this nature, because if we can at least start determining what didn’t lower crime, we can focus on the potential solutions/mitigations. And as you point out, the issue is so complex and there are probably so many different factors it is hard to really get our hands around.
I agree - let’s remove lead. I live in NYC, so my general impression was that lead had been removed, certainly from private housing, and most if not all public housing. I found the study that violent crime in Chicago was hugely prevalent in public housing buildings abutting or over the expressways to be particularly interesting.
I think the political reality of determining the “causes” of crime makes it extremely difficult to handle. You’ve got a phenomena that has multiple causes, although not in equal weight, and certain constituencies that want certain attributes to be prevalent, despite what the facts may or may not say. Giuliani made great hay about Compstat and the broken windows theory, but I think Levitt/Dubner did a pretty thorough job wrecking that. Likewise, they wrecked the general economic state theory as well. However, given that they are statisticians, I strongly suspect they would support the lead theory, which has more than just a correlation going for it, I think.
But in the end its more of a political football than something people can agree on. Unfortunately for Giuliani, he doesn’t have anything in his corner remotely approaching the analytical rigor of what this guy did.
It’s a factor of many things. I believe it a combination of abortion and what the article states.
Why the Rudy Guliani angle? I started looking for the original report and discovered it was first published in the journal Environmental Research in 2000. (2000!!??) It is just making its way into the Washington Post now?
Gotta love politics.
Yeah - I didn’t really understand why it showed up now. I assume it’s because Giuliani is the current Rep frontrunner, and his entire pitch is “I’m the mayor of 9/11 and I singlehandedly lowered crime in New York.” It was never a particularly credible claim, but I live in NYC, so I’m probably relatively privileged in terms of my knowledge vis a vis his typical stump audience.
Not necessarily new info, but Greg Mankiw (Professor of Economics at Harvard) noticed the WaPo article on crime rates as you did: http://gregmankiw.blogspot.com/2007/07/why-crime-has-fallen.html
There are some comments posted, not sure what they say, but there are usually some interesting debates following his posts.