Life Unleaded: Three Essays on The Socio-Economic Effects of Lead Exposure

Abstract

Chapter 1: Does lead pollution increase crime? We perform the first meta-analysis of the effect of lead on crime, pooling 542 estimates from 24 studies. The effect of lead is overstated in the literature due to publication bias. Our main estimates of the mean effect sizes are a partial correlation of 0.16, and an elasticity of 0.09. Our estimates suggest the abatement of lead pollution may be responsible for 7–28% of the fall in homicide in the US. Given the historically higher urban lead levels, reduced lead pollution accounted for 6–20% of the convergence in US urban and rural crime rates. Lead increases crime, but does not explain the majority of the fall in crime observed in some countries in the 20th century. Additional explanations are needed. Chapter 2: How does lead pollution affect birth outcomes? Does a mother’s lead exposure increase the risk of child death? Does it lower the infant’s birthweight (a proxy for later health outcomes)? We test these hypotheses by examining an intervention in the Scottish water supply, which reduced water lead levels and blood lead levels in Scotland’s largest two cities. In our main estimates, we use a staggered difference-in-differences design estimated with two-way Mundlak pooled OLS. We do not find the lead reduction interventions reduced birthweights. However, we do find that they may have had a large effect on infant deaths. Our main estimates suggest lead reduction accounts for 0.3-0.1 percentage points decrease in deaths in Glasgow and a 0.7-0.1 percentage points decrease in deaths in the Alnwickhill water plant supplied area of Edinburgh. However, these results are not robust to alternative specifications, and therefore can only be taken as weak evidence of an effect. Chapter 3: Does lead pollution harm educational achievement? And are the marginal effects greater at low or high levels of lead? We use exogenous variation in lead pollution from water treatment in Glasgow, Scotland, combined with within-household sibling differences, to estimate the effect of lead on education. We compare pre- and post-treatment sibling differences between treated and control areas with difference-in-differences estimation. We find a clear dose-response relationship. Treated areas with low prevalence of lead piping show no change compared to a control group. In contrast, high lead pipe prevalence areas show improvement in educational outcomes. Our findings indicate that countries and areas with very high levels of lead can expect large educational gains from even small amounts of lead abatement, while those with already low levels of lead can expect much lower marginal improvements.