TCS Daily

Muddy Statistics Dirty Air

By Joel Schwartz - May 1, 2002 12:00 AM

According to "State of the Air 2002," a report released today by the American Lung Association (ALA), "more than 142 million Americans live in areas where the air they breathe puts them at risk."

If that were true, air pollution would be one of the most serious health challenges in the United States. Fortunately, it's not. "State of the Air" vastly exaggerates Americans' exposure to air pollution, and misleads the public into believing that air pollution is getting worse, when in fact it has been improving.

How did one of the nation's foremost public health charities get the numbers so wrong? Rather than basing its study on actual air pollution levels as directly measured by pollution monitors, and upon research-based assessments of who is actually harmed by a given level and frequency of air pollution, ALA used statistical legerdemain to cook the books.

  • Clean counted dirty: Ozone can affect your health only if it's high where you're located. Because air pollution varies from place to place, many counties monitor pollution levels at several locations. Yet, ALA tallied a countywide pollution violation for each day that at least one monitor in a county registered ozone greater than 0.084 parts per million (ppm) during an 8-hour averaging period. As a result, ALA counts more ozone violation days for a county as a whole than occurred at any single location in the county. For example, ALA claims Los Angeles County averaged 37 days per year with elevated ozone, even though the most polluted location (Glendora) averaged 22, and most areas far fewer. Indeed, dozens of populous areas, including Los Angeles, Chicago, Phoenix and Sacramento, have only one or a few areas with a substantial number of elevated-ozone days, while most of their counties rarely or never have elevated ozone. Nevertheless, by applying its inflated ozone value to the county as whole, ALA counts clean areas as dirty.

  • Extreme grading standards: The Environmental Protection Agency (EPA) has two ozone health standards. The current standard, known as the "1-hour standard," has been in place for more than two decades. EPA plans to implement soon a newer more stringent standard known as the "8-hour standard." EPA's independent Clean Air Science Advisory Committee, a panel of health experts and atmospheric scientists, found that EPA's proposed new standard would protect human health with an adequate margin of safety. ALA's grading standard is much more stringent. It is akin to requiring students to get 95 or better on a test in order to pass. For example, areas that average only four elevated-ozone days per year get the same F grade as areas with dozens of annual ozone violations - ozone violations, as noted above, that are based on an inflated number of ozone days. By inflating the number of elevated ozone days, and going beyond the standard necessary to protect health, ALA was able to give dozens of areas an undeserved failing grade. But it also blurred the distinction between serious and minor air pollution problems.

  • Outdated data: ALA used pollution data from 1998 to 2000, even though data from 2001 are available. Since air pollution has been declining in many areas, using older data makes air pollution look higher than current actual levels.

  • "Sensitive" people: Health research has shown that in many people, high ozone levels - 0.12 ppm and above - especially combined with exposures longer than three hours, pre-existing respiratory disease, and exercise, can cause both substantial decreases in measured lung function and increases in subjective symptoms, such as coughing and pain while breathing deeply. ALA doesn't stop there. It assumes 40% of the population - including all children under 14, all adults over age 65, and all people with a respiratory disease - are "sensitive" to ozone, and suffer serious and permanent harm even when ozone levels are in the range of 0.085 to 0.105 ppm on just three or four days per year. A wide range of research, including research on children and the elderly, shows that most people don't experience measurable reductions in lung function and even fewer experience subjective respiratory symptoms when ozone levels are this low. Furthermore, the vast majority of the effects of low-level, infrequent ozone exposure are temporary and don't harm long-term health.

The fight against smog is a great success story in environmental protection. According to EPA, ozone levels decreased by an average of about 24% nationwide between 1980 and 2000. Southern California, the region with the worst air in the country, reduced its annual violations of EPA's one-hour ozone standard by about 80% between 1980 and 2001. Houston, the second most polluted area in the country, reduced ozone violations by about 60% during the same period. Most, though not all, metropolitan areas have also achieved significant improvements. And these gains occurred at the same time that Americans increased their driving by 75%.

Readers of the ALA report would never know these facts. Instead, ALA misleads readers into believing air pollution is getting worse.

Only a few metropolitan areas - San Bernardino, Houston, and Fresno - still have serious air pollution problems. The vast majority of other regions have clean air, or have air pollution at a level harmful to only a few percent of the population. Rather than the 142 million claimed by ALA, perhaps 20 million Americans are now at risk from ozone pollution. This is still a large number, and everyone deserves to breathe healthful air. Nevertheless, the real state of the air is far more favorable than ALA's scare-tactics would have Americans believe.

Ironically, ALA's efforts could actually end up reducing Americans' overall health and safety. "State of the Air" will encourage the public to demand unnecessary additional expenditures to clean up air that is already clean. But in a world of limited resources, society can address only some of the many risks people face. When society wastes effort on small or non-existent risks, fewer real problems get the attention they deserve, reducing our health and safety.

"If you torture the data enough, it will confess," goes a cautionary statistics joke. ALA seems to have adopted this maxim without a trace of irony.

Joel Schwartz is senior scientist Reason Public Policy Institute.

TCS Daily Archives