though I had no idea what the red vertical bars at the bottom of the chart meant? The others are horizontal black bars.

Thanks for this.

The third seems to be the most powerful, but it suffers from a severe unacknowledged observer's paradox. It has striking results, but the intervention that gets blood lead levels down to where they say they get to produces an outcome better than for kids who have 50% as much blood lead concentration.

Those are striking results: Want low crime? Then have really low blood level levels for 30-40 years or have high lead levels for the first 10 years of a kid's life then moderate lead levels for the next 10. This is not only counterintuitive, they really need to have a cause for that particular kind of effect.

What's left are the "CDC-approved interventions", which includes upgrading living conditions, nutrition, monitoring the family and counseling. They could account for the "moderate lead has lower crime than lower levels of lead" results, being, as they are, the kinds of interventions usually used for kids with behavior issues. Sadly there's no way of knowing from that article how much of the effect is from the interventions and how much is from blood lead levels' drop. In other words, this particular study suffers from some of the same problems that the first two studies try to avoid, and that renders the results fuzzier than the researchers claim. It shows a drop, it provides really convincing sounding numbers, but then in the text are unexamined assumptions that undermine the accuracy of the numbers. (Then again, that's the direction the researcher's motivation would lead them: Good, strong claim that constitutes a clarion call to the action they recommend, underplaying factors that might weaken their results and their call to action--and therefore the publishability and the social importance of the study. It also makes those who have done things not responsible, and there's a strong urge to deny responsibility to people that you're in solidarity with.)


I like the idea of lead levels being responsible for a lot of crime and anti-social behavior. It's been a long-standing solution to a long-standing problem, even if a lot of the publications in this area are advocacy-based research. As the MJ article points out, the basics are clear but precise quantification is the problem. I'd really like to see the first two articles. But I'm not sure everybody else wants to follow where this particular argument leads, however. For much of recent history, progressives have hated the actions of the wealthy. One problem: The wealthy were more likely to have lead pipes and lead crystal, both of which produce higher concentrations of lead in the blood.

I don't mind going there because I simply don't find that this is all that exculpatory. It doesn't justify actions, it merely explains them. 90% of the people exposed to high BLLs didn't commit the aberrant behavior, and it makes the victims no less victimized. Those with very low lead levels outside of poverty still have non-zero levels of aberrant behavior, and in areas with very low lead levels you get higher levels of aberrant behavior. There's no way to distinguish the two groups in behavior terms, and some of the high-lead cohort would still engage in aberrant behavior.