The 103rd Congress, which concluded in November 1994 in a blaze of partisan bickering, will be forgotten for many reasons by those interested in environmental policy. With the exception of creating a new national park in the California desert, Congress failed to take action on a long list of environmental issues. However, the 103rd Congress will be memorable on at least one environmental count: it was the Congress that discovered risk analysis.
Congress has regulated risk for decades. For example, the national ambient air quality standards called for in the Clean Air Act of 1970 are required to protect against health risks to sensitive populations . The Toxic Substances Control Act, enacted in 1976, was probably the first law to explicitly use "unreasonable risk" as the criterion for government to take regulatory action. But Congress has never concerned itself with how risks were calculated or with comparing different risks. Risk as a general concept was of concern but, with a few notable exceptions, risk analysis was not. In 1993–1994, this situation changed dramatically.
Below I review some of the efforts in the 103rd Congress to deal with risk analysis; I then identify the major factors underlying lawmakers' interest in such analysis. I also outline what risk legislation can (and cannot) accomplish and distinguish among the uses of risk assessment, two issues about which Congress seems to be confused.
Legislative risk proposals
More than a dozen bills dealing with risk analysis were introduced in the 103rd Congress. Notable among these were bills introduced by Senator Daniel Patrick Moynihan (D–New York) and Representative Herbert C. Klein (D–New Jersey). Even more notable was an amendment to S.R. 171, a bill proposed by Senator John Glenn (D–Ohio) to make the U.S. Environmental Protection Agency (EPA) a cabinet department.
Senator Bennett Johnston (D–Louisiana) introduced the amendment, which would have required that EPA conduct a risk analysis for each of its regulations and compare the risk reduction to be achieved by the regulation with the cost of the legislation and with other types of risks. The Senate overwhelmingly passed it by a 95–3 vote, but later the content of the Johnston amendment was modified several times. (The original version required risk analysis of all final regulations; later versions made the requirement applicable only to major regulations and to proposed rather than final regulations.)
Legislators proposed adding this amendment to almost every pending environmental bill. The lack of action on environmental legislation during the 103rd Congress was due, to a great extent, to an inability to reach an acceptable compromise on the amendment's language. Junior members of the House surprised the leadership by defeating the rule under which the EPA cabinet bill would go to the House floor for a vote, in part because the rule would have precluded consideration of the Johnston amendment.
The basic requirements of the Johnston amendment were similar to the cost-benefit requirements already called for by a Clinton administration executive order (E.O. 12866). The Johnston amendment's one novel requirement was that the risks to be regulated be compared with other risks—a challenging requirement but not one that would bring to a halt all environmental regulatory efforts.
Senator Moynihan's bill (S.R. 110), the "Environmental Risk Reduction Act of 1993," would have required the EPA administrator to establish a Committee on Relative Risks to "identify and rank the greatest environmental risks to human health, welfare, and ecological resources," as well as a Committee on Environmental Benefits to provide expert advice on estimating the quantitative benefits of reducing risks. In addition, the bill would have required EPA to develop "guidelines to ensure consistency and technical quality in risk assessments." Finally, the bill would have required EPA to establish a research program on environmental risk assessment and to create an Interagency Panel on Risk Assessment and Reduction to coordinate federal efforts.
The Johnston amendment's one novel requirement was that EPA compare risks to be regulated with other risks—a challenging requirement but not one that would bring to a halt all environmental regulatory efforts.
Moynihan's bill, which was aimed at improving the quality and visibility of risk assessment, emphasized comparative risk analysis of the problems addressed by different EPA programs, rather than risk analysis of the problems addressed by individual regulations. A bill introduced by Representative Klein contained some of the same provisions as the Moynihan bill but focused on improving the quality of risk assessments done to support individual regulations. Klein's bill (H.R. 4306) would have established a Risk Assessment Program within EPA to develop, review, and update risk assessment guidelines. Other elements of the Klein bill included research and training in risk assessment and a pilot project on comparative risk analysis.
Terry Davies chaired a forum on risk assessment organized by the Center for Risk Management. Six high-level policymakers—including EPA Administrator Carol Browner, Senator Max Baucus, and Representative John Mica—attended the March 1994 event. Their discussion centered on legislation sponsored by Mica that would require EPA to conduct a risk analysis for all regulations it proposed
The Klein bill originally was supported by the Clinton administration. Environmentalists, who have generally opposed any efforts to promote risk analysis, stated that they would not oppose the bill. However, the House Committee on Science, Space, and Technology made a series of changes in the bill that caused both the administration and the environmentalists to oppose its passage.
The offending changes were put forward by congressional members and staff who believe that EPA risk assessments are generally biased in favor of regulation and exaggerate the degree of risk. The changes would have done two things. First, they would have made both risk assessment guidelines and EPA's risk assessments potentially subject to judicial review. In withdrawing support for the bill, EPA stated that the changes could make risk assessment "more a construct of the courts than of sound science." Second, the changes would have directed EPA to use "the most plausible" and "unbiased" assumptions to calculate "central estimates of risk" and to employ the "best information." Although these changes sound innocuous, they could have changed EPA's risk assessment methodology in fundamental ways, especially when combined with the threat of litigation.
In the closing days of the session, Congress enacted a U.S. Department of Agriculture reorganization bill with a version of the Johnston amendment attached to it. However, the amendment applies only to environmental and health regulations promulgated by the Department of Agriculture. No other risk legislation passed, but the issues raised in the debate over the Klein bill will be high on the agenda of the 104th Congress, many of whose Republican members have promised reform of federal regulation as part of their "Contract with America." The reasons for interest in risk have become, if anything, more pressing, and the Republicans have generally been more supportive of risk legislation than the Democrats.
Factors underlying Congress's interest in risk
Why the sudden passion for risk analysis and comparative risk assessment? Several interrelated factors account for Congress's newfound interest.
The first factor is a shift in the public's view of environmental problems. Whether because of the increasing costs of environmental remedies, the rightward shift of the nation's politics, growing cynicism toward all groups and institutions, or other reasons, many people no longer believe that all environmental problems are urgently pressing. The notion of priorities—of some problems being more important than others—has entered the environmental debate.
State and local governments have seized upon comparative risk assessment as a potent weapon for fighting expensive and often unwanted federal environmental requirements.
The second factor is the squeeze being put on some state and local governments by unfunded environmental mandates. These governments have seized upon comparative risk assessment as a potent weapon for fighting expensive and often unwanted federal requirements. In many cases, states and localities believe they can show that they are being required to expend funds on problems that either pose smaller risks than those arising from other problems on which the money could be spent or that pose trivial or nonexistent risks. This "grass roots" dimension of the push for comparative risk analysis is politically of great significance.
In Congress, risk analysis also has been linked with the issue of takings, uncompensated restrictions on private land use. Environmentalists have dubbed risk analysis, unfunded mandates, and takings as "the unholy trinity," although risk and takings do not have the direct, substantive connection that risk and unfunded mandates often do. The three have become linked because each potentially could slow or halt federal environmental regulation.
A third factor contributing to the interest in comparative risk is the shortage of public funds at all governmental levels. The shortage emphasizes the need to set priorities and to make hard choices. Not coincidentally, the congressional committees responsible for appropriating money to EPA have been strong supporters of applying comparative risk analysis to different EPA programs (as opposed to different proposed regulations). For these committees, risk analysis holds the promise of providing a rationale and a defense for difficult budgetary choices. At the same time, the results of risk analysis are sufficiently broad and uncertain that the committees do not have to worry about losing control over budgetary decisions.
What risk legislation can accomplish
No other congressional issue is marked more by confusion and misinformation than the current debate over risk assessment. One reason is that legislators seem confused (perhaps in some cases deliberately) about what risk assessment legislation can accomplish.
Members of Congress have an understandable tendency to blame EPA for problems that local constituents have with pollution-control requirements. Since risk assessment supposedly guides EPA decisions, they believe that changing the way risk assessment is done can alleviate the problem of unwanted or unreasonable requirements imposed on local governments and corporations. However, for Congress, in many cases both Shakespeare and the comic strip character Pogo are apt. The fault is not in the stars—Congress has met the enemy and it is them.
The unfunded mandates that have caused the most problems for local governments are those related to drinking water. Communities complain that EPA is requiring them to monitor for chemicals that pose no risk and that the agency is demanding expensive capital investments to deal with nonexistent threats. But most of these difficulties arise from the 1986 amendments to the Safe Drinking Water Act—amendments that required EPA to set standards for forty water contaminants within two years of the act's passage and to keep issuing standards for additional contaminants at an equally rapid pace. Congress directed that the standards be set "as close to the maximum contaminant level goal as is feasible." In turn, the maximum contaminant goal is to be set "at the level at which no known or anticipated adverse effects on the health of persons occur and which allows an adequate margin of safety."
To put it bluntly, Congress should not pass laws that require absolute protection for the public and then complain when EPA promulgates standards that provide such protection. It should not pass laws that require EPA to move rapidly to promulgate numerous regulations and then complain when the agency moves rapidly to promulgate numerous regulations. Implementing the law should not be considered a political crime.
Congress should not pass laws that require absolute protection for the public and then complain when EPA sets standards that provide such protection. Implementing the law should not be considered a political crime.
Another "confusion" in Congress is that risk drives all environmental decisions. In fact, many environmental regulatory requirements are statutorily determined by technology and thus relatively unaffected by risk findings. For example, the initial standards for controlling hazardous air pollutants under the Clean Air Act Amendments of 1990 are to be based on the best technologies employed by each type of polluting facility, not on risk. Similarly, many of the regulatory requirements under the Clean Water Act are based on "best available technology," a determination of which is unrelated to risk. EPA actions under these provisions will not be influenced by any changes in risk assessment methods.
Uses of risk assessment
A more general source of confusion in the current debate over risk assessment arises from a failure to distinguish among different uses of risk assessment. At least four different policy uses of risk assessment exist. Each involves different methodologies and raises different problems.
The most common use of risk assessment in policymaking is in regulatory decisionmaking. For all significant regulations, E.O. 12866 requires the agency proposing the regulation to conduct a cost-benefit analysis. From the perspective of EPA and the other health and safety regulatory agencies, the benefit side of the cost-benefit equation generally is the amount of risk reduced by the regulation as calculated by some type of risk assessment. Within EPA, risk assessment is often used to gauge where to set a standard (although, as noted above, statutory requirements frequently preclude risk considerations), because it is the only way to determine how much (if any) danger a given substance, product, or activity poses.
A second use of risk assessment occurs in Congress's statutory definition of "acceptable risk." Probably the best example of this use is the Clean Air Act, which requires the EPA administrator to promulgate more stringent standards for emissions of hazardous pollutants when the technology-based standards for the emissions "do not reduce lifetime excess cancer risks to the individual most exposed... to less than one in one million."
These "bright line" provisions have been based on quantitative assessment of cancer risk, but cancer may not be the risk that is of most concern. Ecological threats, birth defects, liver damage, hormonal or immune deficiencies, or any of a thousand other problems may be the reason for regulating risk. Because the cancer risk may be irrelevant, gearing the risk standard to cancer may set the standard too high or too low. Risk assessment takes many different forms. Quantitative cancer risk assessment is only one of them and often not the most appropriate one to use.
The crudeness of risk estimates may make it impossible to establish clearly that one risk is greater than another. Moreover, such comparisons do not take into account the many dimensions of risk other than the amount of damage to health and the environment.
Another problem is that the bright line, acceptable risk approach assumes a precision that most risk assessments cannot achieve. Risk assessment is still a relatively crude science, and, depending on which methodological assumptions are used, its results may vary a hundredfold or more. Thus, placing great legal weight on one point estimate of risk is an open invitation to shade the assumptions in a certain direction in order to achieve the desired outcome.
A third use of risk assessment is priority setting for individual risks or regulations, which involves comparing one specific risk to another. Such comparisons can be useful in putting any particular risk into perspective; but two caveats, neither of which has received much attention in Congress, are important to note. The first concerns the crudeness of risk estimates. If the uncertainty range around any point estimate of risk is several orders of magnitude, it frequently will be impossible to establish clearly that one risk is greater than another. The second caveat relates to the many dimensions of risk other than the amount of damage to health and the environment. These dimensions include whether the risk is undertaken voluntarily, whether the victims can be identified, and whether the nature of the risk is catastrophic—that is, whether great damage occurs at one time, as in a plane crash, or whether less damage occurs and is spread over time, as in car accidents. These dimensions of risk are important politically, psychologically, and even ethically. They need to be taken into account when comparing risks.
The fourth use of risk assessment is priority setting for government programs and budgets. This use was pioneered by EPA in 1987 when it published its report Unfinished Business. Senator Moynihan has introduced legislation requiring this type of priority setting to be instituted within EPA. Both the House and Senate appropriations committees for EPA have expressed interest in this approach in the belief that it might provide a "scientific" way of making (or justifying) difficult budget choices.
Comparisons of risks regulated by different programs are a useful way to consider priorities, and they hold long-term promise of bringing greater rationality to government budgeting and goal setting. However, we do not have (and may never have) good methods for comparing different types of risks. Comparing health risks with ecological risks, for example, is clearly a value-laden process. Moreover, acting on the results of broad risk comparisons is almost always impeded by individual statutory mandates. Each environmental program has its statutory support, which is designed (in part) to give each program high priority and prevent its being compared to other programs.
The road ahead
Risk assessment can be a powerful tool for improving environmental policy and decisionmaking. Like all powerful tools, however, it can be abused and employed for nefarious purposes.
Most of the risk legislation that has been proposed would have little short-term effect on environmental policy. However, I believe some of the proposals could do major harm to the quality of the science behind regulatory initiatives by making risk guidelines judicially enforceable. Doing so would transform risk analysis from a scientific undertaking to a legal one, would preclude the exercise of scientific judgment on how to conduct risk assessments of individual chemicals, and would be a major obstacle to incorporating scientific advances into risk assessment. In addition, some proposals would make risk assessment information useless to decisionmakers by dictating which risk assessment methodologies are used. Some of these proposals can be interpreted to mean that risk assessments should determine risk to the average person rather than to the most vulnerable people.
However, the discovery of risk analysis by the 103rd Congress means that the new Republican Congress has an opportunity to forge legislation that will improve the long-term quality of regulatory decisions and environmental policy. If the varied interests with a stake in environmental policy can reduce the ideological and partisan coloration that has characterized the risk debate so far, and if they can accept both the uses and limitations of risk assessment, the risk debate could lead to a new era of more effective, efficient, and equitable environmental programs.
Terry Davies is director of RFF's Center for Risk Management. Portions of this article appeared previously in Inside EPA's Risk Policy Report (vol. 1, no. 2, October 14, 1994).
A version of this article appeared in print in the January 1995 issue of Resources magazine.
A version of this article appeared in print in the January 1995 issue of Resources magazine.
