Blaming Lead, Not Hormones

by Dish Staff

Yglesias flags a study that links declining teen pregnancy rates to the decline of leaded gasoline:

What [researcher Jessica Wolpow Reyes] does is take advantage of the fact that leaded gasoline was phased out unevenly across states in the late-1970s and early-1980s to generate some not-quite-experimental data. You can see the results here:


(Source: Jessica Wolpow Reyes)

Similar results are found for related “risky” behaviors such as the odds of having sex and drinking at an early age.

It’s worth reflecting on the ways in which the political system is rigged to congenitally under-regulate these kind of health hazards. If you, as a politician, take a stand that goes against the financial interests of some group of incumbent industries your reward is that significant social ills are alleviated … Fifteen to 20 years after your proposal is phased into place. No governor or president – and very few senior legislators – sticks around long enough to claim credit for these things.

Kevin Drum, anti-lead advocate, is far from surprised:

This is not a brand-new finding. Rick Nevin’s very first paper about lead and crime was actually about both crime and teen pregnancy, and he found strong correlations for both at the national level. Reyes, however, goes a step further. It turns out that different states adopted unleaded gasoline at different rates, which allows Reyes to conduct a natural experiment. If lead exposure really does cause higher rates of teen pregnancy, then you’d expect states with the lowest levels of leaded gasoline to also have the lowest levels of teen pregnancy 15 years later. And guess what? They do. …

The neurological basis for the lead-crime theory suggests that childhood lead exposure affects parts of the brain that have to do with judgment, impulse control, and executive functions. This means that lead exposure is likely to be associated not just with violent crime, but with juvenile misbehavior, drug use, teen pregnancy, and other risky behaviors. And that turns out to be the case. Reyes finds correlations with behavioral problems starting at a young age; teen pregnancy; and violent crime rates among older children.