The effect of air quality and children’s educational outcomes has received a growing interest in the past years. This attention is in direct response to the medical literature that shows that children are more vulnerable than adults to the adverse effects of pollutants. The empirical evidence in education has mostly focused on estimating the effect of air pollution on attendance –children who miss a lot of school have lower grades and are less engaged in school. These studies have found that high levels of carbon monoxide increase absences, even when those levels are below the federal standards (Currie et al, 2007; Ransom and Pope, 1992).  However, to the best of our knowledge, no study has estimated the direct impact of air quality on student achievement.

We estimate the causal effect of pollution on the academic performance of 4^th graders in Santiago, Chile. Pollution is a serious environmental problem in Santiago, which is located in an enclosed valley with limited wind and rain. This coupled with the high levels of industrial development and vehicle emissions contribute to the levels of pollution that often exceed the suggested by the WHO. We scraped the web for unique information on air quality from all pollution monitors in Santiago (10) for the past 10 years. We focus our analysis on different levels of particulate matter. Those with less than 10 micrometers in diameter (PM10) can potentially increase respiratory and cardiovascular problems, including asthma. Fine particles (PM2.5) are so small that can get into the lungs, causing serious health problems. We also collected Ozone (O3) levels and other pollutants.

Several endogeneity concerns should be taken into consideration in this type of analysis. Simple cross-sectional comparisons of pollution and student achievement suffer from omitted variable bias. Air pollution is not randomly assigned, and several confounding factors (e.g., urbanization, crime levels, school quality) could bias our results. Furthermore, parents could choose where to live base on their preferences for clean air and school quality. We use two identification strategies to address these issues. Our first model uses school and year fixed effects, that is, we compare students within the same school over time while controlling for unobservable year factors (as well as student, family, and school observable characteristics). In our second strategy we use this same model but instrument the level of pollution by precipitation levels. The idea behind this instrument is that precipitation provides an exogenous change in pollution levels –after a rainy day the levels of particulate matter drop to very low levels.

We find that a 1 sd increase in pollution levels decrease student math achievement in 0.1 sd. We find similar negative effects in science. Given that poor neighborhoods have low air quality levels, the results of this study have important policy implications. We rarely think of air quality distribution as an inequality concern in education, nevertheless, our results suggest that pollution could be an important factor on the academic achievement gap observed across students of different socioeconomic backgrounds.