DSMatticus wrote:Bit late, but I have to comment on this. I hope I'm misunderstanding you in some critical way, but I'm reading that as saying:
1) X has property Y.
2) X is a Z.
3) So maybe all Z's have property Y!
I'm pretty baffled by your confusion. What part of the causal chain do you dispute or not understand?
A.) Some classes of chemicals, lead in particular, are called neurotoxins.
B.) Neurotoxins cause human brain damage which leads to among other things lowered intelligence and emotional control.
C.) Lowered intelligence and emotional control increases the chance of someone causing crime.
D.) For two brief periods of time -- widespread use and then abandonment of lead paint and leaded gasoline -- humans were exposed to much higher rates of lead than normal.
E.) Post-WW2 international crime waves have a strong positive correlation with widespread use of the particular neurotoxin of lead.
F.) Therefore other chemicals identified as neurotoxins in the environment such as mercury and cadmium may be artificially driving crime rates upwards.
Grek wrote:Alternative hypothesis: Being poor growing up makes it more likely that you'll become a criminal as an adult AND more likely to live in the part of town where the swing sets still have lead paint as a child. Lead paint exposure and violent crime are caused by poverty.
DSMatticus wrote:On the specific issue of lead poisoning, poverty is correctly associated with poor mental development, lead exposure, and crime. Lead exposure is similarly and correctly associated with poor mental development. And poor mental development is also linked with more crime. Lead makes people stupid and stupid people make bad decisions. That's pretty straightforward. I'm reasonably sure Lago's studies are exaggerating the effect of lead by way of failing to properly separate from the shitsoup that frequently accompanies it, but the effect almost certainly isn't nonexistent.
Yes, the standard liberal response is that having high levels of lead toxicity is an outcome of poverty. So things like lowered education, IQ, and heightened crime and teen pregnancy rates are due to poverty.
HOWEVER Rick Nevin is claiming -- and I am convinced by -- that lead exposure in particular increases crime above and beyond typical bog-standard poverty. In particular, this section is what convinced me:
http://pic.plover.com/Nevin/Nevin2007.pdf
It is striking that preschool blood lead is highly significant at best-fit lags consistent with peak offending ages for each crime category. Burglary and other property crime arrests peak at ages 15–20, and the best-fit for burglary is 18 years in combined nation regressions and 16–19 years in separate regressions for the USA, Canada, Britain, France, Finland, West Germany, and New Zealand. Aggravated assault peaks from age 18 to the late-20s, and the best-fit is 22–24 years for aggravated assault in the USA and Britain and for violent and sexual assault in Canada and New Zealand. Robbery arrests peak from age 15 to the mid-20s, and the best-fit lag is 23 years in a combined regression and 20–21 years for the USA, Canada, West Germany, and New Zealand. The bestfit lag for index crime is 18–21 years in the USA, Britain, Canada, Italy, Finland, and New Zealand. Some nations show longer best-fits for some crimes, but blood lead is generally still highly significant at the international best-fit for that category.
Although time series comparisons can result in coincidental correlations, no nation shows any correlation between burglary and blood lead at lags of less than 10 or over 38 years—the blood lead coefficient in such regressions is insignificant. No nation shows any significant relationship between robbery or violent and sexual assault versus blood lead with a lag of less than 11 years, between aggravated assault and blood lead with a lag of less than 14 years, or between rape and blood lead with a lag of less than 13 years. Changes in R2 when unemployment is added are also consistent with other evidence that unemployment has a substantively small effect on property crime (burglary and most index crime) and no clear relationship with violence.
The very high significance of blood lead at lags consistent with peak offending ages is especially striking in light of divergent crime rate trends. Canada’s index crime rate was 60% higher than the rate in Britain in the early-1970s, but 20% lower in 2001. The USA index rate was 22% higher than the French rate and 40% higher than Australia’s rate in 1980, but the USA rate was 39% below the French rate and 45% below Australia’s rate in 2001. The 1974 USA burglary rate was 50% and 98% higher than rates in Britain and Australia, respectively, but the 2002 USA rate was 56% and 63% lower than rates in Britain and Australia. The Canadian robbery rate was five times the rate in Britain in 1962, but the 2002 Canadian rate was less than half the rate in Britain. The 1960 USA aggravated assault rate was almost three times the rate in Britain, but the 2002 USA rate was half the rate in Britain. The 1960 USA rape rate was eight times the British rate, but the 2002 USA rape rate was just 50% higher than the British rate.
