Friday, September 12, 2008

Lead and crime

Sometime back, there was a Freakonomics post on lead and crime. I came across an old Washingtonian article from August 2006 that writes in more detail about the correlations between lead and behavior. I think the article is more persuasive than the Washington Post articles cited in the Freakonomics post.

1. The time series evidence presented is more convincing than, oh, say the abortion and crime link.
2. I'm not fully convinced because I think it matters how lead exposure is measured: The article in the Washingtonian convinces me that lead paint is a major problem but none of the papers are able to come up with a good measure of lead paint exposure. The time series evidence uses leaded gasoline as a proxy for lead exposure. Steve Levitt says he tried to use air borne lead measurements but did not find any relationship. I am sympathetic to the air borne measure but I think it suffers from the same problems as pollen counts in that exposure can be extremely localized. Again, it also depends on whether the airborne measures are at the state, county, city or census tract level. (I'm only kidding - I don't expect anything more than data at the state and perhaps some city level).
3. The most interesting pointer was to research by Jessica Reyes:
Environmental Policy as Social Policy? The Impact of Childhood Lead Exposure on Crime:
Childhood lead exposure can lead to psychological traits that are strongly associated with aggressive and criminal behavior. In the late 1970s in the United States, lead was removed from gasoline under the Clean Air Act. I use the state-specific reductions in lead exposure that resulted from this removal to identify the effect of childhood lead exposure on crime rates. The elasticity of violent crime with respect to childhood lead exposure is estimated to be 0.8, and this result is robust to numerous sensitivity tests. Mixed evidence supports an effect of lead exposure on murder rates, and little evidence indicates an effect of lead on property crime. Overall, I find that the reduction in childhood lead exposure in the late 1970s and early 1980s was responsible for significant declines in violent crime in the 1990s and may cause further declines in the future. Moreover, the social value of the reductions in violent crime far exceeds the cost of the removal of lead from gasoline.

No comments: