The healthcare costs of air pollution in France
利用工具变量法估计短期暴露于NO2、O3和PM对法国医疗成本的因果影响，发现即使低于欧盟标准，污染每年造成的医疗成本也比以往估计高十倍，且影响所有年龄层。
I estimate the causal impact of short-term exposure to nitrogen dioxide (NO2), ground-level ozone (O3), and particulate matter (PM) on healthcare costs in France. I construct a large-scale dataset by linking administrative healthcare expenditures for a nationally representative sample with high-resolution air pollution and meteorological data. To address endogeneity concerns related to economic activity, I implement an instrumental variable (IV) strategy that exploits weekly variations in altitude atmospheric conditions—such as thermal inversions, wind speed, and the height of the planetary boundary layer—that predict local pollutant concentrations yet are unlikely to affect healthcare utilization except through pollution. My findings reveal that air pollution, even at concentrations below current European air quality standards, imposes annual healthcare costs that exceed earlier estimates by a factor of ten. Heterogeneity analyses show that pollution affects multiple medical specialties, including cardiology, pulmonology, and ophthalmology, while placebo specialties, such as trauma surgery, exhibit no significant effects. Contrary to prior work focusing on children and the elderly, I find that adverse health outcomes extend across all age groups, demonstrating broader population vulnerability. Moreover, marginal effects prove larger at lower pollution levels, implying a concave doseresponse function that underscores the potential for substantial cost savings from even modest pollution abatement in relatively clean areas. These results suggest that earlier cost-benefit analyses likely undervalue the societal gains from stricter environmental regulation.