Climate change garners most of the headlines, but the Trump administration is pushing a much larger and broader pro-pollution agenda whose latest manifestation is a push at the EPA to overturn a long-established scientific consensus that fine particulate pollution (colloquially “soot”) kills people.
This is critically important for two main reasons.
One is that for decades the EPA has been regulating various sources of particulate emissions and the science around how harmful they are plays a role in driving how strict those regulations become. The other is that particulate emissions play a key role in the bureaucratic politics of climate change.
Because carbon dioxide emissions are global and the consequences of climate change are also global, it is generally hard to demonstrate that cutting a given source of greenhouse gas emissions will have large benefits to Americans. But most regulations that reduce carbon emissions also reduce much more localized soot — and taking into account the fact that soot has a marked tendency to kill people who live nearby the emissions sources is important to making the cost-benefit analysis math work.
Tony Cox, an industry consultant Trump tapped to lead the EPA’s Clean Air Scientific Advisory Committee, argues that the EPA’s agency’s current standards for evaluating this question does not feature enough causal rigor. He wants to abandon the current “weight of the evidence” standards that allows for a wide range of evidence to be considered, with credibility weighted according to the quality of study design, in favor of a very narrow class of studies that employ “manipulative causation” methods.
But as Francesca Dominici, a biostatistician at Harvard’s T.H. Chan School of Public Health, writes with researcher Gretchen Goldman in a new Science essay on the subject, this loads the dice against demonstrating harms, because “randomized control trials are not possible (or ethical) when studying environmental hazards.”
In other words, you can’t get a group of volunteers together and then randomly assign half of them to have the air they breathe be deliberately poisoned just to see how bad the poison is. You have to rely on ways of exploiting random variation through methods like observational data, time series, and quasi-experiments.
It’s true, of course, that you don’t want to do a completely naive associational study and then leap to big policy conclusions from it. But public health researchers are aware of the limits of the data available and use matching, statistical controls, and other standard techniques to try to get a clearer picture of what’s going on. Barring all that evidence would, as Dominic writes, “place a nearly unattainable burden of proof on the scientific community” and totally undermine the EPA’s legal mandate to protect vulnerable people.
Both the public health research and the EPA debate are largely focused on the question of soot’s ability to kill people (through, e.g., asthma), but there’s a big emerging set of empirical economics studies suggesting that fine particulates have a wider range of adverse cognitive impacts and are significantly underrated.
In other words, the consequences of rolling back the war on soot could be substantially worse than the mortality impacts alone imply.
Particulates are bad for kids, working-age people, and seniors
Economists, like public health researchers, are limited by their inability to deliberately poison people in order to study the impact of the poisoning. Consequently, none of the studies I’ve read on particulate pollution and cognitive impairment quite reaches “gold standard” levels of scientific evidence.
But there are, instead, lots of quasi-experiments and statistical matching studies, and they all find that air pollution is worse than we thought. In addition to killing small numbers of people, air pollution harms large numbers of people in small but meaningful ways:
- Victor Lacy, Avraham Ebenstein, and Sefi Roth study the impact of short-term ambient air pollution on Israeli students’ test scores and find “a robust negative relationship with test scores” which “suggest[s] that the gain from improving air quality may be underestimated by a narrow focus on health impacts.”
- Steffen Meyer and Michaela Pagel look at the impact of particulate pollution on stock trading and find “that the negative effects of pollution on white-collar work productivity are much more severe than previously thought.”
- Tom Chang, Joshua Graff Zivin, Tal Gross, and Matthew Neidell look at a pear-packing plant in Georgia and find that blue -ollar workers are no better off: “An increase in PM2.5 outdoors leads to a statistically and economically significant decrease in packing speeds inside the factory, with effects arising at levels well below current air quality standards.”
- Kelly Bishop, Jonathan Ketcham, and Nicolai Kuminoff link detailed Medicare records to EPA air pollution records and “find that a 1 microgram-per-cubic-meter increase in average decadal exposure (9.1% of the mean) increases the probability of receiving a dementia diagnosis by 1.3 percentage points” and therefore “conclude that regulation of air pollution has greater benefits than previously known, in part because dementia impairs financial decision making.”
- Wes Austin, Garth Heutel, and Daniel Kreisman look at the rollout of school buses in Georgia that have had their engines retrofitted to be cleaner and “find that retrofitting districts see significant test score gains in English and smaller gains in math.”
This is an emerging field of study, and it would obviously benefit from some further inquiry. That said, it appears to be the case that fine particulate pollution induces cognitive impairments in school children, white collar workers, blue collar workers, and the elderly — in other words, all people at all stages of life.
Each of these impacts independently suggests that we are currently allowing too much air pollution, and they also cry out for more research that attempts to assess the long-term cumulative impact of pollutants that appear to induce cognitive problems throughout the life cycle.
The question of exactly how much we are under-regulating particulates seems somewhat open to me, but the sign of the error is very clear.