1. Chay and Greenstone (2003): Air Quality and Infant Mortality
Citation: Chay and Greenstone [2003]. The Impact of Air Pollution on Infant Mortality:Evidence from Geographic Variation in Pollution Shocks Induced by a Recession. QuarterlyJournal of Economics, 118(3):1121-1167.
Research question: What is the causal effect of particulate matter (total suspendedparticulates, TSP) on infant mortality?
Identification strategy: The 1981-1982 recession caused large, sharp reductions inmanufacturing activity and therefore in TSP emissions, but the reductions varied sub-stantially across counties depending on their industrial composition. Counties with morerecession-sensitive industries experienced larger TSP declines an instrumental variable strat-egy where the interaction of county industrial composition and the national recession gen-erates exogenous variation in TSP. This is a forerunner of the Bartik shift-share approachapplied to an environmental question.
Key results: A 1 μg/m³ reduction in TSP reduces the infant mortality rate by approx-imately 0.5 per cent. The elasticity of infant mortality with respect to TSP is estimated at0.35, implying that the Clean Air Act's TSP reductions over 1970-1990 saved approximately1,300 infants per year in the United States.
Takeaway: The effects of particulate pollution on infant health are large and causallyidentified. The study established the recession-variation IV as a credible tool for environ-mental health effects, influencing a generation of subsequent work.
2. Currie and Walker (2011): E-ZPass and Traffic Pollution
Citation: Currie and Walker [2011]. Traffic Congestion and Infant Health: Evidence fromE-ZPass. American Economic Journal: Applied Economics, 3(1):65-90.
Research question: Does traffic-related air pollution (primarily from idling vehicles attoll plazas) affect infant health outcomes?
Identification strategy: The introduction of E-ZPass electronic toll collection in NewJersey and Pennsylvania dramatically reduced vehicle idling at toll plazas and thus localcarbon monoxide and particulate emissions near plazas but left traffic volumes unchanged. The study compares birth outcomes for mothers living within 2 km of a toll plaza (treated)to mothers living 2-10 km from a plaza (control), before and after E-ZPass introduction. This is a DiD design where the treatment is the sharp, local reduction in pollution causedby E-ZPass.
Key results: E-ZPass introduction reduced prematurity rates by 10.8 per cent and lowbirth weight rates by 11.8 per cent for mothers within 2 km of toll plazas, relative to mothersfarther away. No pre-existing trends are found for mothers near plazas before E-ZPass; andno effects are found for traffic accidents (a placebo), supporting the causal interpretation.
Takeaway: Local traffic pollution has large effects on birth outcomes. The E-ZPassdesign is nearly as clean as a randomised trial: E-ZPass was introduced on a known sched-ule, affected pollution at a precisely defined geographic radius, and left confounding factors(traffic, income of drivers) unchanged.
3. Deryugina, Heutel, Miller, Molitor, and Reif (2019): Acute Effects of Ozone on Mortality
Citation: Deryugina et al. [2019]. The Mortality and Medical Costs of Air Pollution:Evidence from Changes in Wind Direction. American Economic Review, 109(12):4178-4219.
Research question: What are the short-run and long-run mortality effects of acuteozone exposure among the elderly?
Identification strategy: The instrument is daily wind direction: when the wind blowspollution from upwind urban sources toward a county, ozone levels increase; when the windblows clean air from less polluted areas, ozone decreases. Wind direction is plausibly exoge-nous to local economic conditions and health behaviour. The study uses Medicare adminis-trative data on daily mortality for elderly individuals to track acute responses to pollutionspikes.
Key results: A 1 ppb increase in daily ozone increases same-day mortality by 0.67per cent and total 30-day mortality by 4.56 per cent. Crucially, the authors show that the"harvesting" hypothesis that pollution accelerates the deaths of individuals who would havedied shortly anyway accounts for only part of the effect; a substantial fraction of deathsare "pulled forward" by more than 30 days.
Takeaway: Acute ozone exposure has large short- and medium-run mortality effectsamong the elderly. The wind direction instrument provides remarkably clean identificationof a causal effect that would be confounded in any cross-sectional or regression analysis.
4. Isen, Rossin-Slater, and Walker (2017): In Utero Exposure to Clean Air Act Benefits
Citation: Isen et al. [2017]. Every Breath You Take Every Dollar You'll Make: The Long-Term Consequences of the Clean Air Act of 1970. Journal of Political Economy, 125(3):848-902.
Research question: Does in utero exposure to improved air quality as a result of the1970 Clean Air Act affect adult earnings?
Identification strategy: The Clean Air Act imposed regulatory constraints based oncounty-level attainment designations (as in Greenstone 2002). The authors compare adultearnings of individuals born in nonattainment counties (which experienced larger TSP reduc-tions) to those born in attainment counties, exploiting variation in the timing and magnitudeof pollution reductions induced by the CAA. This is a DiD-style comparison in which the"treatment" is pollution reduction in early life.
Key results: A one-unit decrease in average annual TSP exposure in the first year oflife increases adult earnings (at ages 25-40) by approximately 1 per cent, with the effectconcentroned among men. The study estimates that the 1970 CAA's TSP reductions raisedthe present discounted value of lifetime earnings by $4,300 per affected cohort member-asubstantial long-run benefit not captured in contemporaneous health cost-benefit analyses.
Takeaway: The long-run effects of in utero pollution exposure on human capital ac-cumulation are large and persist into adulthood. This finding fundamentally changes thebenefit-cost accounting of early environmental regulation: the gains extend far beyond im-mediate mortality reductions to include lifetime earnings.
References
- Chay, K. Y. and Greenstone, M. The impact of air pollution on infant mortality: Evidencefrom geographic variation in pollution shocks induced by a recession.
- Quarterly Journalof Economics, 118(3):1121-1167, 2003.
- Currie, J. and Walker, R. Traffic congestion and infant health: Evidence from E-ZPass.American Economic Journal: Applied Economics, 3(1):65-90, 2011.
- Deryugina, T., Heutel, G., Miller, N. H., Molitor, D., and Reif, J. The mortality and medicalcosts of air pollution: Evidence from changes in wind direction.
- American EconomicReview, 109(12):4178-4219, 2019.
- Isen, A., Rossin-Slater, M., and Walker, W. R. Every breath you take every dollar you'llmake: The long-term consequences of the Clean Air Act of 1970. Journal of PoliticalEconomy, 125(3):848-902, 2017.