The leading national measurement of toxic substances is the Toxics Release Inventory (TRI), which the Environmental Protection Agency produces every year. The TRI is an unwieldy measure because it tracks "releases" of more that 600 chemicals that vary widely in their hazardous character. In addition, the EPA refines the TRI from year to year, changing the number of chemicals tracked, the threshold for reporting, and the kind of commercial operations that must report. In the early years, the TRI reported on about 300 chemicals, but starting in 1995, the EPA doubled the inventory to more than 600. These changes make it difficult to discern trends.
The EPA has helpfully broken out the data against a 1988 baseline that includes only the chemicals included in the original inventory (shown in Chart 27). This measure shows a 42-percent decline in toxics releases since 1988, a reduction of nearly 1.5 billion pounds. The chemical industry, not surprisingly, has shown the largest decrease of all industries included in the TRI, with a 50.8 percent reduction in releases since 1988.51 These industry reductions reflect mostly productivity gains and technological improvements, as well as concerted efforts to reduce "releases" for public relations purposes.
Chart 28 shows the last three years according to an updated (1995) baseline. Both Chart 27 and 28 show a slight increase in "releases" in 1997 over 1996.
The TRI has significant limitations as a measure of toxic risk and environmental quality, and is not a very useful indicator of future trends. The TRI cannot be compared, for example, to the trendlines for air or water quality. First, the term "release" is misleading. While it does include compounds that are released into the air, it also includes chemicals that are disposed of in hazardous waste facilities, and even chemicals that are recycled on the premises. It would be more accurate to call the TRI the "Toxics Use Inventory."
Second, the TRI is incomplete and arbitrary in many ways. It does not include any data from government facilities, such as military bases. It does not capture data from many kinds of small businesses, such as "mom and pop" auto body shops. Third, not all chemicals are created equal in their toxic properties.
Measuring chemicals simply by the pound can be highly misleading, since an ounce of one chemical may be more toxic than 10 pounds of another. Hence, the TRI is of limited use in making judgments about environmental quality and toxic risk. The EPA itself has repeatedly cautioned readers about these limitations. The most recent TRI, for 1997, warns: "TRI reports reflect releases and other waste management activities of chemicals, not exposures of the public to those chemicals. Release estimates alone are not sufficient to determine exposure or to calculate potential adverse effects on human health and the environment."52 (Emphasis added.)
Professor George Gray of Harvard University's Center for Risk Analysis puts the problem more bluntly: "Chemical use does not equal chemical risk . . . . Simply knowing how many pounds are used provides no information about health or environmental risks."53 Gray points out that supermarkets can be required under the TRI to report their use of acetic acid as a "toxics release," even though the acetic acid may be in the form of vinegar used in salad dressing. Better to replace chemical use reporting, argues Gray, with chemical risk reporting. "Risk evaluations should be certified by independent agents, just like financial data are certified by accounting firms," Gray says. "This would provide firms with strong incentives to reduce risk and would provide firms and citizens both with useful information . . . . We must focus on how chemicals are used, not whether they are used."
Chart 29 shows the TRI results for Michigan on the EPA's 1988 baseline: a 48.7 percent reduction.