Schlenker and Walker (2016)

Question 1

paper’s data

Answer 1

the paper uses four data sources. First, it uses data on airport traffic from the Bureau of Transportation Statistics. Second, it uses daily and hourly pollution data from the California Air Resources Board. Third, it uses weather data on temperature, precipitation, and wind speed. Fourth, it uses data on hospital admissions and emergency room visits from the California state government.

Question 1

Correlation between health and air pollution

Answer 1

There are various potential reasons. As an example, poor people are often exposed to pollution and are also less healthy due to worse nutrition, more stress, etc… Conversely, people who are healthy may be more willing to expose themselves to pollution

Question 2

empirical strategy

Answer 2

Schlenker and Walker measure the effect of air pollution on contemporaneous health
outcomes, such as hospital visits. They use variations in air pollution due to congestion at California airports. When there is more congestion, airplanes idle longer and emit more pollution in the area surrounding the airport. Further, they use only the variation in congestion due to network delays that originate outside of California

Question 3

why is local congestion problematic

Answer 3

Local congestion could be problematic for identification because it could be correlated
with a third factor that impacts health. For instance, heavy rain in the area could cause delays, while also impacting health on its own (such as by causing people to stay indoors, be more likely to have car accidents, catch colds, etc…). By contrast, delays based on bad weather outside of the state are much less likely to have distinct impacts on health
in California—other than via their effect on airport congestion in California

Question 4

how they validate their strategy

Answer 4

The authors run a number of tests to validate their strategy. Most prominently, the authors run falsification tests where they calculate effects on health outcomes that should not be affected by air pollution. These outcomes include stroke, bone fractures, and appendicitis

Question 5

paper’s main findings

Answer 5

• First, a one standard deviation increase in airport congestion causes a 0.28 standard deviation increase in carbon monoxide in areas around the airport.
• Second, a one standard deviation increase in air pollution causes hospital admissions for asthma, total respiratory problems, and heart problems to increase by 17%, 17%, and 9% of the baseline average. Magnitudes are larger for infants and the elderly but are also significant for adults age 20-64.
• Third, a one standard deviation increase in air pollution is responsible for 540 thousand dollars in hospitalization costs for the people living within 10km of the airports included in the study