Why Governments Can't Handle Risk
Safety Is a Process of Discovery
MARCH 01, 1995 by RANDY T. SIMMONS
Professor Simmons is the Head of the Political Science Department of Utah State University.
Public opinion surveys indicate that mainstream America is worried about environmental risks. In 1990, for the first time since pollsters began asking the questions, a plurality (46 percent) of American voters believed that the quality of life where they live was worse than it was five years previous, and the number who were pessimistic about the future of the environment (46 percent) exceeded the number who were optimistic (32 percent).
These surveys, reported in The Polling Report, also indicate that Americans expect government to resolve these anxieties. In 1982, one-third of Americans wanted more government regulation of the environment. By 1990, two-thirds wanted more. In 1982, 45 percent agreed with the statement that the environment was so important that requirements and standards could not be too high. In 1990, 80 percent agreed. People apparently remain confident of government’s ability to protect them against risk.
But the truth is that the government is spectacularly ill-suited to anticipate future harms. There are a number of reasons.
First, most of the potential harms we face are low-probability future events about which no one can know very much. By putting protection against these events into the hands of a central authority, almost inevitably a single approach to the harm will be taken. Given such uncertainty, any policy of anticipation is likely to be the wrong one.
The problem with leaving prediction in the hands of a central authority is illustrated by the government’s mineral assessment process (even though geology is a more certain science than assessing risks in an uncertain future). For each proposed wilderness area, the Bureau of Land Management, the Bureau of Mines, and the Geological Survey conduct mineral assessments to determine the potential for finding mineral deposits, based on existing geological theory. The agencies produce probabilistic estimates of mineral potential.
But scientists do not regard these estimates as specific, quantitative data. Even for the areas that appear to offer little mineral promise, a negative assessment report is not absolute. The vast oil and gas deposits in the Overthrust Belt were unknown only a few decades ago; several exploration companies had failed to find anything. But someone with a new geological theory applied a slightly different technology in a previously dry hole and discovered the reserves.
If we didn’t have a variety of people making different assessments—if, instead, everyone relied on the government’s assessments—the oil might never have been found. Such uncertainties prompted William A. Vogely, head of the Department of Mineral Economics at Pennsylvania State University, to observe, “Their [resource assessment] results are so imprecise that a policy based on them may be worse than a policy that recognizes our complete ignorance of potential reserves.” The same can be said for government policies anticipating environmental risks.
A second reason for government’s inability to be our protector against future harms is that initial risks must be taken in order to reduce future harms. Only in rare areas, such as national defense, are politicians willing to tell their constituents they must suffer in order for future benefits to happen. Congress is currently incapable of controlling budget deficits in large part because legislators fear causing their constituents immediate, visible harm.
A telling example comes from the Clean Air Act amendments of 1977, which were designed to reduce sulfur dioxide from power plants and other sources. An effective approach would have been to mandate the amount of emissions of sulfur dioxide that would be allowed, and let the power plants choose the best means of reaching them. But that concept ran into political trouble. It would have allowed power plants to meet the standards by buying low-sulfur coal from the west and would have eliminated jobs in eastern coal fields, which produce high-sulfur coal.
To protect jobs, Congress mandated that scrubbers be installed in all new construction, regardless of the sulfur content of the coal. But scrubbers are less effective than coal switching, and companies kept using their old plants to avoid the costs of the scrubbers. Serving constituents was placed ahead of the environment. The result was probably dirtier air. While this case is dramatic and documented more extensively than most, politics frequently overrides environmental goals.
A third reason that governments are not suited for protecting against risk is that they have trouble carrying out the time- and place-specific experiments and trials that develop the ability to deal with surprises. Whether overseeing grazing, timber harvest, air quality, or water policy, governments are not able to discriminate by location or conditions. One size fits all.
Former EPA administrator William Ruckelshaus has pointed out that the Clean Air Act requires the Environmental Protection Agency to impose costly automobile inspection and maintenance programs in communities where certain pollution criteria have been exceeded only twice in a year. This rule applies even if the violations occurred because of the placement of the air-quality monitoring devices, not because the air was discernibly unhealthy. “The law does not allow the federal government to distinguish between (for example) Los Angeles and Spokane, Washington, in this regard—a restriction that defies common sense,” says Ruckelshaus. “In the same way, we cannot distinguish between a plant discharging pollutants into a highly stressed river in Connecticut and one discharging into Alaskan waters that bear no other pollutant burden. In other words, the law does not permit us to act sensibly.”
A fourth reason for government’s inability to anticipate errors and prevent them is that for such anticipation to work, it must be possible to make mistakes and learn from them. Private entrepreneurs receive feedback about their successes and failures through profit and loss information. But politicians’ and bureaucrats’ success does not depend on learning from mistakes. Instead, it depends on increasing budgets, responding to organized interests, and maximizing votes. Bureaucrats are insulated from the effects of their good and bad choices—someone else benefits or loses. Politicians are rewarded by voting for policies popular with their constituents, even if the policies’ cost to the nation are greater than the benefits to their constituents. These are hardly strategies for developing the ability to anticipate harm or to be resilient in the face of it.
Fifth, the public is ill-informed and easily excited about new risks. Its hysteria can cause public agencies and politicians to act in exceedingly harmful fashion, essentially through witch hunts. As the article by Robert Nelson in this issue discusses in some detail, it is a situation not unlike that of the sixteenth and seventeenth centuries, when witches were blamed for many harms. One reason why this hysteria was supported by the authorities was the fact that when witches were executed, their property was confiscated by the authorities. Something parallel happens today, write Richard Stroup and John Goodman. “Businesses [that are] politically `convicted’ are assessed billions of dollars in Superfund taxes and cleanup costs,” they write, “and thus help to fund the agencies which prosecute them.”
In sum, governments are poorly suited for achieving safety because safety is a process of discovery, undertaken without being able to see the end, rather than an object that is freely available to public or private decision-makers. As Aaron Wildavsky has pointed out, markets provide a better route for achieving safety. “Markets overcome defects to enhance overall safety not because `they’ know the answer, but precisely because they don’t; convinced that better bargains can always be struck, markets are based on the principle of incessant search. The more decentralized, dispersed, variegated, and, need I add, competitive markets become, the more likely it is there will be more different kinds of search and, therefore, more safety, especially against the unforeseen. . . .
“Attempting to short-circuit this competitive, evolutionary, trial and error process by wishing the end—safety—without providing the means—decentralized search—is bound to be self-defeating. Conceiving of safety without risk is like seeking love without courting the danger of rejection.”
The only meaningful alternative is to learn once again to rely on markets to help us cope with environmental and other risks. They sample the unknown, they have reliable feedback, and they allow trials, errors, and corrections. 
3. See Bruce A. Ackerman and William T. Hassler, Clean Coal/Dirty Air or How the Clean Air Act Became a Multi-billion Dollar Bail-Out for High Sulfur Coal Producers and What Should be Done About It (New Haven and London: Yale University Press, 1981).
4. William D. Ruckelshaus, “Risk, Science, and Democracy,” Issues in Science and Technology, vol. 1, no. 3 (Spring, 1985), p. 33. Cited in Aaron Wildavsky, Searching for Safety (New Brunswick: Transaction Books, 1988), p. 224.
5. See also William C. Clark, “Witches, Floods, and Wonder Drugs: Historical Perspectives on Risk Management,” in Richard C. Schwing and Walter Albers, Jr. (eds.), Societal Risk Assessment: How Safe is Safe Enough? (New York: Plenum Press, 1980), pp. 287-313.
6. Richard L. Stroup and John C. Goodman, “Making the World Less Safe: The Unhealthy Trend in Health Safety and Environmental Regulation,” National Center for Policy Analysis (NCPA) Policy Report #137, April 1989.