MIT's Michael Greenstone and co-authors have published an important paper in PNAS documenting that "coal kills". This finding is based on a natural experiment from China where cheap winter heating was made available in certain geographic areas. This creates a natural treatment group and control group. This well done study is also politically shrewd. As natural gas has grown cheaper, any research documenting the social cost of coal will be welcomed by the "green cities" coalition and those who want sharp cuts in global greenhouse gas emissions.
Comment #1: What is the elasticity of demand for winter heating in China?
"The quasi-experimental empirical approach is based on China’s Huai River policy, which provided free winter heating via the provision of coal for boilers in cities north of the Huai River but denied heat to the south. "If the elasticity of demand equals zero, then this policy should have induced no differential pollution effect across the policy border. The more elastic the demand for winter heating then the larger the environmental impact.
Comment #2: What was the efficiency of the coal boilers used by the Chinese heat providers located in the North of the Huai River? If they had very efficient boilers then they would have needed less coal to make the heat and thus less TSP would have been released.
Comments #1 and #2 suggest that if either the boilers had been efficient or the elasticity of demand was low then there would have been no differential TSP at the border. A "natural experiment" existed because these conditions were not satisfied.
Comment #3: Coal burning bundles many nasty environmental treatments including TSP. Lucas Davis discusses the wide set of environmental problems caused by burning coal. TSP is just one of them. It is possible that TSP floats more uniformly around power plants and coal heating systems than these other toxics but Davis' results suggest that the Greenstone team may over-state the role of TSP in causing death in China because the coal burning simultaneously raises TSP and another vector of toxics that they don't measure. Since they can measure TSP, they attribute all of the deaths to TSP but this implicitly zeroes out all of the other marginal health effects caused by other pollutants that increase because of coal burning.
Comment #4: How much money did local governments and households north of the river save because they spent nothing on winter heating? What did they do with this extra money? If they bought cigarettes then they further injured their health? If the local government invested in cleaner water then this would improve health. A future study should investigate differential consumer expenditure across the border.
UPDATE: Towards the end of the published paper, the authors discuss this point. Here is a quote. "Further, the free provision of coal is an in-kind transfer that increases households’ disposable income, and this may cause northern households to alter their consumption patterns in ways that are protective
(e.g., medical care) or harmful (e.g., tobacco or alcohol) for health." They acknowledge that if the policy changes household or government behavior on health related inputs that the policy does not isolate the causal effect of TSP on mortality holding all else equal.