Environmental Risk Analysis: A Review of Public Policy Issues II

98-618 ENR

CONTENTS FOR THIS SECTION

Influence of Interagency Guidelines on EPA Risk Analysis
State of the Art of Risk Analysis at EPA

Risk Management at EPA

Comparative Risk Analysis at EPA

Influence of Interagency Guidelines on EPA Risk Analysis. Other federal agencies with responsibilities for protecting human health and safety, such as the Food and Drug Administration (FDA) and the Occupational Safety and Health Administration (OSHA), also were conducting risk analyses by the mid 1970s. Each agency independently developed analytic procedures suitable for its mission, authority, and budget; published results of risk assessments; and designed risk management activities with reference to scientists' risk estimates. However, because agencies often shared jurisdiction over an industry or chemical and sometimes came to different conclusions about the level of risk associated with chemicals or industries, as well as the level of risk that should be regulated (a risk management decision), some independent investigators and industry scientists asserted that certain federal scientific risk assessments were of poor scientific quality. They criticized the federal government for allowing inconsistent analyses of risks.4 Often criticisms targeted EPA which, it was argued, tended to produce higher risk estimates and to regulate more stringently than some other agencies.5

These criticisms of federal risk analyses focused on the diverse choices made by different agencies to cope with inherent uncertainties of risk analysis that arise from missing or ambiguous information on hazards and gaps in current scientific theory. Environmental risk analysis is especially beset by many uncertainties, because data usually are sparse, and scientific theories explaining hazards, exposures, and effects often have not been established. To conduct a risk analysis under these conditions requires choices among plausible alternative assumptions and competing theories to bridge the gaps. Because these choices cannot be based on science alone, they are subject to challenge. For example, if data are available for two animal species, which data should an analyst use or how should data be combined to estimate risk to a third species? The National Academy of Sciences(NAS) identified 50 choices beyond the realm of science (so-called "inference choices") that affect cancer risk analyses.6 For an excellent, though somewhat dated, discussion of inference choices in the context of cancer risk analysis, see chapter two in Making Cancer Policy by Mark Rushefsky.7

In 1977, the EPA, OSHA' the Consumer Product Safety Commission (CPSC), and FDA responded to the criticisms of agencies' risk analyses by establishing an Interagency Regulatory Liaison Group (IRLG) to coordinate procedures for analyzing cancer risks.8 The intent was to establish guidance for agencies on risk assessment procedures, including procedures for filling in gaps in data and scientific theory through the choice of inference options (i.e., "default assumptions"). The Food Safety and Quality Service of the U.S. Department of Agriculture (USDA) joined the IRLG soon thereafter. The IRLG proposed in 1979 a general "cancer policy" to coordinate risk analysis and risk management across agencies to the extent permitted by statute. According to the NAS, this was "the first evidence that all the [F]ederal regulatory agencies agreed on the inference options applicable to the identification of carcinogenic hazards and measurement of risks".9

However, the IRLG policy was controversial. Some scientists and industrial and environmental groups accused the Carter Administration of allowing policy prescriptions to influence scientific judgments. In response, Congress authorized a study by the NAS on institutional arrangements that might improve the agencies' use of risk analysis. 10

The NAS published the results of the study in a 1983 landmark report, Risk Assessment in the Federal Government: Managing the Process, also known as "The Red Book." It described the risk analysis policies and practices of all federal agencies and concluded that no change in institutional structure was necessary or desirable to improve risk assessments. In addition, the Academy provided a general framework for cancer risk assessment that still is used today; recommended that agencies separate risk assessment from risk management; and suggested development of uniform general risk assessment guidelines for the federal government.

In 1985, the NAS framework was adopted by the White House Office of Science and Technology Policy (OSTP), after a comprehensive review of the scientific basis of risk assessment by scientists from the regulatory agencies, the National Institutes of Health, and other federal agencies.11 Although the OSTP decision was not binding on the agencies' risk analysis practices, it provided a consistent basis for developing agency guidelines.

The Reagan and Bush Administrations convened numerous interagency meetings, conducted studies, and issued guidance in the hope that uniform policies and procedures might be established to guide risk analysis and management in the federal government. General guidelines for risk analysis were issued as part of the "Regulatory Impact Analysis Guidance" in Appendix V of the 1991-1992 Regulatory Program of the United States.12

The Clinton Administration continued to search for better interagency coordination of risk assessment. In August 1994, an interagency work group for the Clinton Administration released Draft Principles for Risk Assessment; Management; and Communication to serve as a "general policy framework" for implementing regulatory policy. In 1996, 0MB issued a report, Economic Analysis of Federal Regulations Under Executive Order 12866, which included general guidance for federal agencies conducting risk analysis, including risk characterization.

Within the agencies, risk managers also strove for consistency. In 1977, EPA became the first federal agency to propose interim guidelines for its cancer risk assessments. In 1986, it was the first agency to establish final guidelines for analyzing risks of cancer and other health effects, all of which were based on the 1983 NAS framework (51 Federal Register 33992-34054, Sept.24, 1986). In addition to cancer risks, the 1986 guidelines for analysis addressed: the risk that a chemical will cause mutations affecting future generations or damage to human development (developmental risks); human exposure to individual chemicals; and human health risks of chemical mixtures. The Agency revised its guidelines for developmental risks in 1991 and for exposure in 1992 (56 Federal Register 63798-63826, Dec. 5, 1991; 57 Federal Register 22888-22938, May 29, 1992). In recent years, EPA has continued to be "the main player in developing and revising risk assessment guidelines

[O]nly EPA has completed scientific reviews of some of its guidelines and formally modified them in response to new scientific information," according to the Congressional Office of Technology Assessment. 13 EPA's cancer risk assessment guidelines currently are being revised. Guidelines for assessing risks to reproduction were issued in 1996 (61 Federal Register 56274-56322, Oct.31, 1996), and EPA issued final guidelines for analyzing neurotoxicity risks and ecological risks in 1998 (63 Federal Register 26,846 and 26,926, respectively, May 14, 1998).

State of the Art of Risk Analysis at EPA. The most detailed and well established risk assessments at EPA probably are those used to assess human cancer risks of chemicals. These methods evaluate and model the results of animal experiments and human studies to estimate the risk that people will develop cancer following various levels of exposure to individual chemicals. Many of EPA's environmental standards, emission limits, and quality criteria are based on the results of cancer risk assessment. Other categories of risks, such as developmental or immunotoxicity, are rarely assessed.

Responding to a mandate in the Clean Air Act Amendments of 1990, the NAS National Research Council assessed the current state of cancer risk analyses for hazardous air pollutants at EPA. The 1994 NAS report Science and Judgment in Risk Assessment refrained from any pronouncement about the overall quality of analysis, but concluded:

· EPA should retain its conservative approach to risk assessment (in which the Agency makes judgments that err on the side of public safety) in the initial phase of setting standards, but EPA should more clearly state its principles for choosing and departing from inference guidelines. 14 The selection of principles cannot be established scientifically, because it ultimately depends on policy judgments about how best to respond to uncertainty.

· EPA should develop and use an iterative approach to risk assessment, beginning with relatively inexpensive screening techniques and moving on to more resource-intensive levels of data gathering, model construction, and model application as each situation warrants. At each level, risk should be reevaluated to produce a more precise estimate. Iteration should cease when no further refinement of the risk estimate is needed to inform risk managers.

· EPA should work to continually improve the models and data used in risk assessments and develop a standard procedure for deviating from its conservative approach to risk assessment when warranted by scientific considerations.

· In its reports to decision makers and the public, EPA should present information about the sources and magnitudes of uncertainty as well as point estimates of risk.

· "Risk assessment is a set of tools, not an end in itself The limited resources available should be spent to generate information that helps risk managers to choose the best possible course of action among the available options" (p.E14).

More specific NAS recommendations are referenced throughout this report.

The Clean Air Act Amendments of 1990 also established a Risk Assessment and Management Commission to consider the NAS report, methods for measuring and describing risks of chronic human health effects from exposure to hazardous substances, methods to reflect uncertainties, and risk management policy issues. The Commission also was directed to comment on the possibility of developing a consistent risk assessment methodology, or standard of acceptable risk, among various federal programs.

The Commission got a late start, holding its first meeting May 16, 1994.15 The final report was not issued until 1997. Members of the Commission agreed that detailed interagency guidelines cannot be established, because the different missions and objectives of federal agencies require them to adopt different approaches to analyzing and managing risks. Instead, it described a general framework for environmental health risk management discussed below.

A panel of experts convened by the National Academy of Public Administration (NAPA) concurred in 1995 with the NAS assessment that EPA's approach to risk assessment is essentially sound, but needs refinement. 16 The panel suggested two general changes to make EPA's risk analyses more useful to decision makers. EPA risk analyses should-

· Be more transparent, so that "anyone reading a summary report should be able to evaluate the assumptions behind the conclusions and the value-laden components of the analysis"; and

· Cover the full range of impacts about which the public is likely to care, including risks to ecosystems, future generations, and economic welfare.17

Risk Management at EPA

Risk analysis is a tool for risk management, and EPA risk managers increasingly have relied on it, to the extent permitted by law. In setting priorities across or within program offices, EPA sometimes has compared risks or risk reduction potential of different regulatory targets. In developing regulations, EPA has used the results of risk analysis to set standards, to compare the effectiveness of various control measures, and to evaluate risks relative to costs. Recent reports by the Commission on Risk Assessment and Risk Management and NAPA recommend an expanded role for risk analysis in risk management.

Comparative Risk Analysis at EPA. In the mid 1980s, the EPA Administrator commissioned a special task force to compare the risks associated with major environmental problems that remained to be controlled, given the level of federal risk management that existed at the time. The purpose was to help the Administrator determine where available EPA resources could be applied to greatest effect.18 Senior EPA career managers and technical experts assigned to the task force ranked 31 environmental problems in a 1987 report Unfinished Business: A Comparative Assessment of Environmental Problems.

EPA scientists based their ranking on available data, but reported that data gaps and uncertainty about risks plagued their efforts. Ranks did not take into account the feasibility of controlling risks, the economic benefits of activities posing risks, the limits of EPA's statutory authority, or the distribution of risks and benefits geographically, over time, or among people. Problems were ranked based on relative risks within four categories: human cancer risks, other risks to human health, ecological effects, and human welfare (including such effects as visibility impairment and damage to building materials). Scientists grouped environmental problems within categories as relatively high, moderate, or low risks.

The exercise revealed that no environmental problem ranked relatively high or relatively low in all four categories of risk. Problems ranked relatively high or moderate in three or more categories included: criteria air pollutants (that is, lead, sulphur dioxide, nitrogen oxides, particulates, carbon monoxide, and tropospheric ozone), stratospheric ozone depletion, pesticide residues on food, and other pesticide risks. other relatively high risks to human health included: hazardous air pollutants, indoor air pollution, indoor radon, pesticide application, exposure to hazardous substances in consumer products, and worker exposures to chemicals. Additional problems posing high risks to ecology or human welfare included: global warming; surface water pollution; physical alteration of wetlands, estuaries, and other aquatic habitats; and mining wastes.

Interpretation of these results requires caution. For example, the low relative risk of hazardous waste sites (as indicated by data available in 1987) was due, in part, to the existence of regulations and availability of funds to treat the problem. Problems such as indoor air pollution were characterized as relatively risky, at least in part, because they were not regulated by any federal agency.

EPA scientists next compared the relative risk of each problem with its budget allocation and the results of national polls of public concerns. The polls indicated: high concern about chemical waste disposal, water pollution, chemical plant accidents, and air pollution; moderate concern about oil spills, worker exposure, pesticides, and drinking water; and low concern about indoor air pollution, consumer products, radiation (other than nuclear power), and global warming.

EPA concluded that its budget correlated better with the priorities of the public than with the scientists' evaluations of residual risks. However, there are several reasons why this conclusion may be suspect. First, EPA did not ask the public to rank environmental hazards, and it did not use the results of scientific studies of how people rank hazards based on risk. (The results of one such study are shown in the figure titled "How People Evaluate Hazards.")19 Rather, EPA staff compiled public responses to questions asked in national opinion polls in 1985 and 1986 about 19 environmental problems which roughly coincided with the 31 hazards ranked by EPA scientists. Thus, the scientists' rankings were assigned after hours of careful deliberation, while the public was responding to a few questions in an opinion poll. In addition, the scientists and public responded to different questions. Scientists addressed the question, "Of the environmental hazards that are recognized, which pose the highest risks and remain to be controlled?" The public was asked, "Which of these nine problems are most serious?" or in some cases, "How serious is the problem of [chemical waste disposal, for example]?" Finally, the public responses were compiled and interpreted by EPA staff

EPA's Science Advisory Board (SAB), an advisory group of independent scientists, reviewed EPA's efforts in its 1990 report Reducing Risk: Setting Priorities and Strategies. SAB praised EPA for considering "the long-term public policy importance of understanding relative risks," but criticized the accuracy and methods of ranking and the omission of important environmental problems. It devised its own method and ranked a different, though overlapping, set of environmental issues. The results were largely consistent with those of the EPA scientists, but the SAB expressed more concern about ecological risks because of "the vital links between human life and natural ecosystems." The SAB identified the following hazards that remained to be controlled in 1990 as the highest human health risks:

· ambient air pollutants (both toxic and criteria);
· occupational chemical exposures;
· indoor air (including radon); and
· pollutants in drinking water.

Relatively high-risk problems affecting ecology and human welfare included:

· habitat alteration;
· loss of biological diversity;
· stratospheric ozone depletion; and
· global climate change.

ENDNOTES

4 Environmentalists also criticized EPA's risk assessments, but their complaint was that EPA used risk assessment to justify less rigorous enforcement of environmental statutes.

5 For example, see Nichols ,Albert L.. and Richard J. Zeckhauser. The Perils of Prudence: How Conservative Risk Assessments Distort Regulation. Regulation, v.10, n. 2 (Nov/Dec)1986. p.13.

6 National Academy of Sciences. Risk Assessment in the Federal Government: Managing the Process. Washington, D.C., National Academy Press. 1983. p.29-33.

7 Rushefsky, Mark E. Making Cancer Pohcy. New York, State University of New York Press, 1986, p.37-54

8 The IRLG was disbanded in 1981.

9 NAS,ibid. p.61.

10 The study was authorized in the Act making appropriations for Agriculture, Rural Development, and Related Agencies programs for the fiscal year ending September 30, 1981 (P. L. 96-528). The study was carried out by the National Research Council with support from the Food and Drug Administration.

11 NAS, National Research Council. Science and Judgment in Risk Assessment. Washington, DC, National Academy Press. 1994. p.2-10.

12 U.S. Office of Management and Budget, Executive Office of the President. Regulatory Program of the U.S. Government, 1991-1992. Appendix V. Washington, U.S. Govt. Print. Off., 1992.

13 U.S. Congress Office of Technology' Assessment. Researching Health Risks. Washington, U.S. Govt. Print. Off. 1993. p.120.

14 According to the National Research Council, these "are generic approaches, based on general scientific knowledge and policy judgment, that are applied to various elements of the risk-assessment process when the correct scientific model in unknown or uncertain" (Science and Judgment in Risk Assessment, Washington, DC: National Academy Press, 1994, p.6-i.)

15 The ten members of the Commission are: Sheila McGuire(appointed by President Clinton), Peter Y. Chiu and Alan Craig Kessler(appointed by President Clinton), David P. Rall and Norman Anderson (appointed by the the Majority Leader of the Senate). Gilbert Omenn and Joshua Lederberg (appointed by the speaker of the House), Virginia Weldon (appointed by the Minority Leader of the House), John Doull(appointed by the Minority Leader of the Senate), and Benard Goldstein (appointed by the President of the National Academy of Sciences).

16 National Academy of Public Administration. Setting Priorities, Getting Results: A New Direction for the Environmental Protection Agency. Washington, DC: National Academy of Public Administration. 1995. p. 35.

17 Ibid

18 EPA, Office of Policy Analysis. Unfinished Business: A Comparative Assessment of Environmental Problems. Washington, U.S. Environmental Protection Agency, February 1987. p. xiii.

19 Scientists, engineers, and other experts in the evaluation of hazards tend to use and interpret the term "risk" in a narrow actuarial sense (e.g., as average, annual mortality rates for a population), whereas non-experts may employ or interpret any of several common meanings of the term, depending on the context. Often, the public interprets "risk" in a very personal way, depending on whether they or their families are exposed. For example, in Figure 1, "Active Club" members, who hunt and fish, rank handgun risks more highly than college students, risk experts, or members of the League of Women Voters, a civic group. In scientific studies that asked the public to estimate the annual mortality rate or hazards, public and scientific hazard rankings are more similar. For example, see Fischhoff B., S. Watson, and C. Hope. "Defining Risk." Policy Sciences, v.17, n. 2, (1984). p.123-139.