The Effect of Low Levels of Blood Lead on Children's Test Scores

The recent discovery of high levels of lead in the drinking water in Flint, Michigan, and in several urban school districts has focused public attention on the potential harms of blood lead, particularly for children's development. While child blood lead levels (BLLs) have declined dramatically over the past several decades, estimates suggest that there are still a half a million preschool-aged children in the U.S. with elevated blood lead levels.

Although there is strong epidemiological evidence linking early lead exposure to negative child out-comes, much of this evidence relies on correlations. As children with high BLLs are more likely to come from minority groups, be poor, live in single-parent homes, and have less-educated mothers, it is difficult to distinguish the effect of elevated BLLs from these other potentially confounding factors. In addition, much of the existing evidence is based on children with much higher BLLs than are common today, raising the question of whether exposure to lower levels of lead also negatively affects child outcomes.

In Do Low Levels of Blood Lead Reduce Children's Future Test Scores? (NBER Working Paper No. 22558), researchers Anna Aizer, Janet Currie, Peter Simon, and Patrick Vivier provide new evidence on this question.

The authors construct a dataset that includes all children born in Rhode Island between 1997 and 2005 whose BLL was tested at least once before age six. Due to the state's rigorous testing regime, most of the children in the sample had multiple BLL tests, allowing for more precise measurement of lead exposure than in many previous studies. The researchers match the information on BLLs to data on third grade test scores.

To surmount concerns about confounding factors, the authors make use of variation in lead exposure that resulted from state lead control policies. These policies included requiring owners of buildings where any child had an elevated BLL to mitigate the lead hazard and requiring all landlords to obtain "lead-safe" certificates in order to rent their properties. As the policies were rolled out in a targeted way, neighborhoods with a higher share of older housing—which were also predominantly African American and poor—experienced the most lead mitigation and the sharpest declines in BLLs. The researchers predict a child's BLL based on the probability that the child's home was certified lead-safe given the child's birth year, neighborhood, and family characteristics, and use this predicted level in the analysis to avoid concerns about the actual BLL being correlated with confounding factors.

The researchers first document that BLLs of children in Rhode Island dropped dramatically between 1997 and 2005, with the largest gains to minority children. The geometric mean BLL of African American and Hispanic children fell by 2.8 and 2.4 micrograms per deciliter, respectively, while the value for white children fell by 1.5 micrograms. This represents a decline of nearly 50 percent from the 1997 value for African American and Hispanic children.

Next, the authors examine the relationship between BLLs and third grade test scores. They find that a one microgram per deciliter increase in a child's mean BLL measurement is associated with a 3.1 percentage point increase in the probability of being "substantially below proficient" in reading. This effect represents an increase of about 25 percent relative to the average probability of reading at this level (12 percent). The association between BLL and math scores is about two-thirds as large and somewhat less precisely estimated.

Taken together, these results suggest that the lead-safe certificate program, which was targeted at poor and minority census tracts, led to a narrowing of the achievement gap between white and minority children. The researchers estimate that the program accounted for one-third of reading gains for African American children and one-fifth of reading gains for Hispanic children between 1997 and 2005. The gap in test scores between white children and minority children narrowed during this period, and the researchers' findings suggest that the program was responsible for nearly 60 percent of the convergence in the case of African Americans and about 20 percent in the case of Hispanics.

The health and educational gains arising from the lead-safe certification program came at a cost of ap-proximately $1,900 per child, although the cost per child will fall over time, as future cohorts continue to benefit from the one-time investment in creating lead safe environments. Noting that lead poisoning may be one of the causes of the continuing achievement gap between white and minority children, the researchers conclude, "environmental regulations targeted at hazards that disproportionately impact minority children may have advantages beyond improvements in health."

The researchers acknowledge funding from the John D. and Catherine T. MacArthur Foundation.