WWP: Blowin' in the wind

Berkeley ARE's resident econometrician (of FWER/FDR fame!), Michael Anderson, has a neat new NBER working paper out this (last) week. Titled "As the Wind Blows: The Effects of Long-Term Exposure to Air Pollution on Mortality", this paper tries to get at an important gap in the air pollution and health literature. We know a lot about the short-run effects of air pollution exposure: it's bad for you. But the long run is harder to pin down. Michael gets at this question using one of my all-time favorite things: a clever identification strategy. 

In order to get at the causal effect of long-term exposure to pollution, Michael exploits the Los Angeles (#beatLA) highway system. He performs a battery of tests to demonstrate that people who live upwind of highways look very similar on observables to people who live downwind of highways. He then uses angle to the highway as an instrument for time spent downwind (constructed using high-frequency wind data) to look at the effect of pollution on mortality among the elderly. He further shows that mobility is low in these Census tracts among the population of interest: the median 75-year-old has lived at her current address for 25 years. It's because of this that Michael can really get at the long-run here. 

Living next to this stuff is bad for your (long-run) health. Who knew? Photo stolen from these guys.

Living next to this stuff is bad for your (long-run) health. Who knew? Photo stolen from these guys.

The cleverness doesn't stop there: there are two other really nice features of this paper. The first is the estimator: In conjunction with the 2SLS strategy, Michael actually uses a modified spatial fixed effects estimator (SFE), such that each observation i is demeaned relative to observations within a radius r along the dimension parallel to the highway - think of this as a highway-segment fixed effect. Instead of a regular fixed effects estimator, the SFE estimator's demeaning group changes for each observation. Running the estimation this way allows for demeaning independently along the highway and in terms of distance away from the highway. Secondly, this paper contains a nice set of falsification tests. Like other papers in this literature, Michael tests for (and finds no evidence of) effects on "placebo" health outcomes. He also shows that highway bearing has no effect on property values, which suggests that people aren't moving away to avoid these negative health consequences.

Michael's abstract describes the paper's results nicely:

There is strong evidence that short-run fluctuations in air pollution negatively impact infant health and contemporaneous adult health, but there is less evidence on the causal link between long-term exposure to air pollution and increased adult mortality. This project estimates the impact of long-term exposure to air pollution on mortality by leveraging quasi- randomvariation inpollution levels generated by wind patterns near major highways. We combine geocoded data on the residence of every decedent in Los Angeles over three years, high-frequency wind data, and Census Short Form data. Using these data, we estimate the effect of downwind exposure to highway-generated pollutants on the age-specific mortality rate byusing bearing to the nearest major highway as an instrument for pollution exposure. We find that doubling the percentage of time spent downwind of a highway increases mortality among individuals 75 and older by 3.6 to 6.8 percent. These estimates are robust and economically significant.

So, next time you're thinking about buying a house, it's probably worth adding yet another thing to the list of deal-breakers: distance to, and direction from, the nearest busy roadway.