The major pollutants first regulated by the Clean Air Act are still causing substantial damages in the United States, particularly to human health. Specifically, ammonia and the five criteria pollutants - fine and coarse particulates, sulfur dioxide, nitrogen dioxide, and volatile organic compounds - currently cause damages that range from $75 - $280 billion annually. Here we will explain how these damages are estimated, what sources are responsible for the damages, and compare them with estimates of the damages from greenhouse gases (GHGs).
Economists measure the impacts of air pollution using integrated assessment models that logically connect emissions to their final effects on society. Of primary concern are the human health effects associated with air pollution, including premature mortality, chronic illness (such as bronchitis and asthma), and several acute illnesses. However, the models also measure the damages from reduced crop and timber yields, impaired visibility, deterioration of manmade materials, and diminished recreation services.
Integrated assessment models applied to the United States begin with available emissions data and then calculate pollution concentrations across the lower 48 states. These concentrations are then converted to "exposures," using county-level population information. Exposures, in turn, are converted into physical effects using concentration-response functions that capture the number of physical effects a certain exposure is likely to cause. Finally, physical effects are converted to dollar damages through valuation techniques.
To do so, we rely on an integrated assessment model called the Air Pollution Emission Experiments and Policy (APEEP). APEEP resembles other integrated assessment models in the literature. However, the way we are using APEEP is innovative. First, APEEP calculates the damages due to current emissions from all existing sources. One ton of emissions is then added at a single source, and APEEP recalculates the aggregate damage. The change in the aggregate damage is the marginal damage of the additional ton of emissions. By repeating this experiment for the six pollutants and 10,000 source locations, APEEP estimates the marginal damage of all emissions of these pollutants in the United States. Multiplying the tons of emissions from each source location by the source-specific marginal damage and summing across all sources yields the gross annual damage (GAD). This is a measure of the value of air pollution damages just as gross domestic product (GDP) is a measure of the value of economic production.
We find that GAD in 2002 is between $75 billion to $280 billion (0.7 - 2.8 percent of GDP). The estimates vary so widely because of three controversial issues: the value of mortality risks, the age dependency of this value, and the relationship between exposure to air pollutants and mortality rates.
First, although the values of many damages from air pollution are known - reduced crop yields, for instance - the value of human health and longevity (and their inverse, illness and death) is contentious. One approach is to use the extra wages paid to workers in risky jobs. This is problematic, however, because mortality risk in the workplace is often associated with sudden death, whereas mortality from air pollution is usually due to long-term exposure. It is also true that people do not agree on what value to place on a small risk of death and so any single estimate will be contentious no matter how it was estimated. The second controversy is whether the value attributed to mortality risks should be the same for all age groups or decline with expected years of remaining life. That is, should a smaller value be assigned to older age groups? Age-specific values are rational because remaining consumption declines with age. However, American principles of equality as guaranteed by the Constitution may dictate that every person be valued the same, regardless of age. Finally, the magnitude of the physical impact of exposure to pollutants is also uncertain. Because controlled experimentation (intentionally exposing humans) is unethical, epidemiologists must rely on natural experiments and toxicologists must rely on animal experiments to learn about human sensitivity to pollution. Unfortunately, these methods are less precise and so the estimates are "noisy." For all these reasons, the range of GAD values is wide.
Turning from aggregate damage to individual pollutants, we find that not all pollutants are equally harmful. Although emissions of fine particulate matter (PM2.5), ammonia, sulfur dioxide, and volatile organic compounds make up only half of all emissions by weight, these pollutants cause almost 80 percent of total damages. PM2.5, very tiny particles that can lodge in the lungs, account for only 6 percent of total emissions by weight, but cause 23 percent of total damages. In contrast, nitrogen oxides and coarse particulates are responsible for almost half of the total tonnage but only 20 percent of damages.
What fraction of GAD is due to different effects? We find that human health damages account for more than 95 percent of GAD. Loss of visibility is clearly one of the most palpable costs of air pollution but its contribution to GAD is small. The same can be said for crop damage, forest damage, and material damages.
The largest source is still industrial production, which causes 50 percent of air pollution damages. The largest single industrial source of emissions are coal-fired power plants, which cause 20 percent of GAD. Mobile sources are responsible for the next largest share, 35 percent. Light-duty gasoline-powered cars and motorcycles contribute 9 percent, SUVs and light-duty gasoline trucks contribute 7 percent, diesel trucks and heavy-duty gasoline vehicles contribute 15 percent, and rail, aircraft, and marine generate the remaining 4 percent of mobile source damages. Residential combustion of fossil fuels and wood, primarily for heating, produces perhaps more damage than people think - 5 percent. Finally, agricultural sources also cause a surprisingly large share of damages (10 percent), from ammonia from livestock production and fertilizers, and dust from tilling cropland.
The above GAD estimates do not include GHGs. How does their impact compare to GAD? Although they have high current visibility on policymakers' agendas, we believe that at least current GHG emissions are not nearly as harmful as criteria pollutants. The empirical impact literature estimates that current emissions will cause future global damages of between $0.50 and $10 per ton of carbon dioxide. So the current six billion tons of carbon dioxide emitted annually in the United States will likely cause future global damages of between $3-60 billion. In comparison, annual GAD lies between $75 and $280 billion. GHGs do need to be addressed, but the damages that current emissions will cause are relatively small compared to the damages from criteria pollutants. Of course, GHGs are accumulating and future emissions will cause higher damages, so they will become relatively more important to control in the future.
Tighter regulations on emissions of ammonia, fine particulates, sulfur dioxide, and VOCs are clearly called for. Important emissions sources to tighten are cars, light trucks, SUVs, all diesel vehicles (especially marine vessels and heavy-duty trucks) and industrial sources. Two other sources that have generally escaped attention must also be examined: residential homes and farms. Although each farm and each house contributes only a little to GAD, the net effect of all homes and all farms is substantial. Finally, pollution control efforts aimed at reducing solid waste (incineration) and water pollution (waste treatment plants) generate an inordinate amount of air pollution damage. Regulators need to think more carefully about integrated pollution management so that in the effort to reduce one pollution problem they do not create a larger one.
Muller, Nicholas Z. and Robert Mendelsohn. 2007. "Measuring the Damages of Air Pollution in the United States." Journal of Environmental Economics and Management, Vol. 54, No. 1. July. pp 1-14.