This is based on a study that didn't directly measures the health effects or economic costs of chemical exposure. It wass a modeling exercise that combines observational studies, animal experiments, exposure estimates, economic assumptions, and opinion. Basically very low quality.
There was a fun observation in American science that the effect of lead on IQ seems to have gotten worse over time.
The postulated mechanism is:
1. Some places have lead, and suck.
2. Because they suck, they are inhabited by people without better options.
3. Those people are stupid.
4. You can do a study comparing the IQs of people in relatively high-lead environments and other people, and, if you attribute the entire difference to lead exposure, you'll see that lead has a negative effect on IQ, which you can represent in terms of IQ points per hair ppm or whatever unit you want.
5. If you then mount a big effort to rehabilitate those environments to clean out the lead, they'll have less lead. Still more than usual, but not as much as before. Lead exposure in the people who live there will fall.
6. But those environments are still bad (if not quite as bad) and the people there are still stupid.
7. If you rerun your study from step 4, you'll observe that people in high-lead areas are still dumber than people in low-lead areas, by about the same amount as the earlier study found.
8. But, the high-lead areas have less lead now than they did then.
9. Attributing the entire IQ deficit to lead exposure, we logically find that each unit of lead exposure is now much more harmful than the earlier study found. There's a lot less lead, but the deficit is just as bad - this means that the effect of the lead got worse!