Association for Public Policy Analysis & Management

Exposure to air pollution has been shown to harm the health and development of children. Research on these relationships has not distinguished between the long-run effects of exposure in utero and infancy and the effects of exposure during early childhood. These effects are hard to separate in part because of the difficulty in measuring relevant outcomes for a child from gestation through kindergarten. A second problem is the challenge inherent in identifying impacts of exposure to pollution that occurs contemporaneous to when cognitive tests are administered in early grades, separate from lingering effects of earlier exposure. Consequently, we have limited understanding of the impacts of air quality on children’s school readiness, as distinguished from impacts on achievement and learning once in school.

In this paper, we attempt to disentangle the impacts of air pollution on school readiness, as measured by cognitive skills before school entry, from the impacts of continued exposure on performance in school. We use data on daily measures of air quality at the county level—including levels of fine particulate matter (PM2.5), ozone, and sulphur dioxide (SO2)—collected by the U.S. Environmental Protection Agency (EPA), merged to longitudinal data on young children from the restricted-use Early Childhood Longitudinal Survey – Birth Cohort (ECLS-B) data. Data were collected on children in the ECLS-B at 9 months (appended to information from children’s birth certificates), two years old, preschool, and kindergarten. At 9 months and two years, children were assessed using the Bayley Short Form-Research Edition, which evaluates young children’s cognitive development. In preschool, children were assessed using a battery of standardized instruments. These assessments provide measures of school readiness. We will use ECLS-B data on problem solving, math, reading, and science skills in kindergarten to measure impacts of air quality on achievement.

This merged dataset permits us to measure exposure to air pollution in utero and throughout early childhood. Previous work investigating air quality and school readiness has relied on assumptions that children have resided since birth in the county where they enroll in kindergarten. We will use the panel features of the ECLS-B, along with information on birth dates and cognitive assessment dates, to study the contemporaneous and long-term impact of exposure to poor air quality among young children. We will exploit within-county variation in exposure due to differences in when ECLS-B children in the same communities were born to identify impacts of pollution on birth outcomes. Then, using variation in dates of assessments of sample children in the same county, we will estimate long- and near-term effects of poor air quality on cognitive ability.

Advancing our knowledge of the role of air pollution on children’s school readiness and achievement is important for understanding the impact of poor air quality on development. This will help better establish the real costs of human-made pollution. It will also help researchers and policy makers assess the near- and long-term effects of different threats to air quality and be useful in determining where efforts aimed at abatement or limiting exposure are most valuable.