This is just how dramatically removing lead from the environment can improve education

In the wake of the crisis in Flint, MI, it is clearer than ever that elevated lead levels appear to be blot many American water supplies—and that poor communities and people of color — especially children — are most at risk from being exposed.

What hasn’t been clear is just how much removing lead can help, especially when it comes to children’s academic performance. We know that it’s harmful, of course, but many previous studies have not done a great job weeding out other factors, like poverty or parents’ education, that could also be impacting their test scores, according to the authors of a new working paper that purports to fix this problem.

The researchers, led by a team from Brown University, looked at all children in Rhode Island born between January 1997 and September 2005 whose blood lead levels (BLL) were measured at least once before age six. This turns out to be a particularly large group, thanks to an aggressive state-sponsored testing program Rhode Island ran during this period.

The state instituted a lead-inspection certificate program that required all landlords to mitigate sources of lead (paint, soil, etc.) in the homes they owned. The researchers came up with a probability model to guess which Census tracts were likely to have been inspected. They found that the program targeted poor and minority areas and that it did not result in gentrification of these areas, meaning there likely wasn’t much turnover in the residents of these homes.

It is through using the Census tract data, which contains information about things like who receives a free lunch in addition to race, and controlling for other state test score trends over time, that the authors are able to cancel out the factors that produced noise in non-controlled lead impact studies.

The authors found that removing lead from the environment can have a much greater impact on educational performance than was previously understood. They calculated that, for every point the amount of lead in a child’s blood increased, there was a 1-point decline in reading scores, and a 3% increase in the probability of being “substantially below proficient” in reading. Using this metric, they were able to work backwards to calculate how much a decline in lead levels would impact learning.

That effect is three-times as great as a non-controlled study. It translates into the lead-certificate program explaining 44% of the improvement in black test scores, and explaining 16% of the improvement in Latino ones, the authors write. And it explains 57% of the improvement in black children substantially below proficient, and 19% of the share improvement for Latino ones.

Here’s the table showing the improvements:

The mean third-grade reading score for a black child in 1997 was 37.8. By 2005, it had improved to 41.54. For Latinos, the jump was from 36.20 to 40.47. Similar gains were seen in math scores.

And the share of “not reading proficient” students fell to 17% from 23% for black children, and to 21% from 30% for Latinos. The drops in this category were not as great in math, but still statistically significant, the authors say.

And it didn’t even cost much to do so: they estimate that over the course of eight years the cost of lead-safe certification was just $4 million.

“The most surprising thing about the study was that Rhode Island was able to reduce children’s blood levels very quickly through a relatively inexpensive public policy targeting housing in high lead areas,” co-author Janet Currie of Princeton University said in an email.

“So what officials should take away from the results is that lead remediation is doable and does not have to be enormously expensive.”

Rob covers business, economics and the environment for Fusion. He previously worked at Business Insider. He grew up in Chicago.