Air Pollution Externalities: Evidence from U.S. Electricity Generation over the Twentieth Century

Carnegie Mellon University, Pittsburgh PA

Investigators

Abstract

This project examines the effect that expansion of electricity generation had on health and other outcomes over the twentieth century. By digitizing new data on power plant-level consumption of bituminous coal and generation of electricity, it becomes possible to address two important issues. The first is how the health effects of air pollution have changed over time with changes in income, nutrition, access to health services, and medical care. This research can then help policymakers understand how future changes in these measures will affect the health of effects of air pollution. The second is the extent to which air pollution makes influenza pandemics more severe. New experimental evidence suggests that air pollution exacerbates the health effects of influenza in mice. The analysis will examine the Spanish Influenza pandemic of 1918-1919 and later pandemics in 1957-1958, 1968-1969, 1977 (in children), and 2008-2009 to provide new evidence on the extent to which pollution exacerbates the health effects of infectious disease in humans. Influenza pandemics are unpredictable but recurring events that create substantial economic disruption and pose a major threat to U.S. and global health. Even with modern antiviral and antibacterial drugs, a pandemic virus with similar pathogenicity to the 1918 virus would quickly overwhelm the existing medical infrastructure, and would likely kill more than 100 million people worldwide. Thus, preventative strategies and targeted distribution of medical resources are likely to be critical in mitigating the severity of an outbreak in the United States. Because of the sector?s reliance on bituminous coal, electricity was an important contributor to air pollution over the twentieth century. The proposed project will improve knowledge about the externalities of air pollution by studying it at a level of detail that has not previously been possible. A key aspect of the project is the expansion of a new detailed electric power plant-level dataset on capacity, generation, fuel consumption, and other production expenses for 1963-1993. Data for 1915 and 1938-1962 have been digitized, and beginning in 1994, these data are available in digital form. This project uses the new dataset to advance knowledge in two main areas of inquiry. First, how have the marginal health effects of air pollution varied over time and as a function of environmental characteristics such as access to electricity and levels of pollution? Existing studies often examine short time frames that have little variation in access to electricity or levels of pollution. The proposed research overcomes this limitation by using power-level data on openings and coal consumption. Monitor data do not become available until 1953, the network of monitors remained sparse into the 1960s, and even today the network is not very dense. Using power-plant level data permits advances in two dimensions. It makes it possible to examine the period prior to the collection of monitor data. Further, it makes it possible to examine the effects of air pollution for a much bigger geographic area. This research allows the PIs to better understand the shape of the pollution-infant mortality concentration-response function and how it varies with levels of pollution, and shifts in the concentration-response function over time due to income, changes in nutrition, access to health services, and advances in medical care. Second, new experimental evidence suggests that air pollution exacerbates the health effects of influenza in mice. Are humans affected in a similar way? This research will provide new evidence from the Spanish Influenza pandemic of 1918-1919 and later pandemics on the extent to which pollution exacerbates the health effects of infectious disease in humans. Understanding such effects is important for three reasons. The current benefits of pollution reduction are understated, because they do not account for the effects on deaths from infectious disease. Pollution abatement becomes a long run and possibly a short run (emergency shutdowns) tool for addressing pandemics and infectious disease more broadly. Finally, during pandemics it may be efficient for polluted locations to receive more prevention and treatment resources than less polluted locations.