Appendix A6

Differences Between the Draft and Final Reports

Including a Representative Sample of Comments Organized by Table of Contents Along with Commission Responses (in Italics)

Fundamental Difference in Organization of Final Report

• We recommend separating the Commission’s work products into two publications. Many of the Commission’s more detailed suggestions raise complicated issues that are not thoroughly addressed in the draft report. These sections detract from the Commission’s principal product: its risk management framework. The report would be strengthened if it focused on laying out why this framework is needed and what we need to do to implement it. This offers the best approach for ensuring the Commission’s recommendations become the "Red Book" for risk management in the 1990s.

The Commission created a two-volume final report, the first volume devoted exclusively to the Commission’s Risk Management Framework and the second to a more comprehensive revision of the 1996 Draft Report. We also moved the risk management chapter from chapter 5 in the Draft to chapter 3 in the Final, to put greater emphasis on risk management and to save some readers from an early immersion in more technical risk assessment issues.

1. Introduction

We added a box defining risk, risk assessment, and risk management.

2. Framework for Environmental Health Risk Management

• We strongly support the Commission’s call for a more systematic, consistent, and disciplined approach to risk management, but the recommendations fall considerably short of what is necessary to develop a coherent risk management program across federal agencies. The Commission needs to propose an adequate strategy for addressing statutory constraints that would limit implementation of its recommendations and establish a guiding set of risk management principles.

• The stakeholder involvement framework as set out in the draft report would be extremely resource-intensive for both agencies and stakeholders, and may not always be workable. While communication with all stakeholders throughout the regulatory development process is paramount, it is neither practical nor constructive to have the same level of involvement at all levels of the process by all.

• As a state agency, we see little new or unique guidance in this framework which will aid us in our day-to-day tasks. We have struggled with difficult risk management decisions for many years and would like practical guidance to help us coordinate our efforts. This chapter discusses only generalities with little practical guidance. It is always a good idea to put a problem into context relative to other sources of similar exposure or even background exposure. However, what is a state regulator to do if residual risks from sources outside of his control exceed a source within his control? Such comparisons often anger the public by belittling their local problem and saying it is insignificant compared to some larger problem beyond their control.

• While the Commission stresses that the appropriate contexts for a problem are likely to be situational, we are concerned that risks associated with individual sources or sites will be inappropriately compared with unrelated exposures. The Commission should recognize that many of the issues suited to analysis by comparative risk are best addressed at the program level, in the decision to regulate a process, release, or exposure, and not in individual risk assessments conducted as part of that program.

• We strongly support the comission’s recommendation to increase public participation in risk assessment and risk management processes. Recommending public involvement early in the risk assessment process is an important step forward from the guidance to separate risk assessment from risk management, as per the Red Book. However, the final report should address the increased funding required for meaningful stakeholder participation at the local level. For example, stakeholder participation from exposed citizens is extremely important in many situations, yet citizen groups do not usually have individuals with the appropriate technical background to educate stakeholders or to argue effectively for their position. These groups need funding to hire technical expertise.

• The scope and content of the proposed risk management framework are of limited value. The proposal breaks no new ground and does not assist risk managers, the professional community, or stakeholders in thinking about risk management in a different way. A more productive path would be the development of a companion set of principles, procedures, recommended practices, and case studies to complement those already in existence for risk assessment.

• Increased stakeholder involvement in both risk assessment and risk management is both necessary and inevitable, but the Commission provides no clear reasoning process or clarity about how the elements of stakeholder participation are to be effectively managed. The Commission could have presented a synthesis of best practices of incorporating stakeholder perspectives into both risk assessment and risk management decisions.

• The Commission should abandon its proposed risk management framework, which fails to address many of the major practical and scientific/analytical issues involved in identifying and implementing risk management options.

• We strongly support the Commission’s conclusions that an increased emphasis on explicit consideration of total environmental exposure is needed in the risk assessment process, that affected populations should be consulted about routes of exposure, that all stakeholders must be brought into the decision-making process from the beginning, and that there is a need for developent of better disease surveillance data.

• A clear strength of the report is the broad risk-management framework. The framework is compelling because it is flexible enough to allow consideration of a diversity of decision strategies, ethical perspectives, and types of technical guidance. It also calls for early stakeholder involvement, use of sound science, and good-faith efforts at consensus formation, features of decision making that are not always evident today.

• We are gratified to note that the Commission recommends enhanced stakeholder involvement and collaboration in its draft report. Our concern, however, is that the Commission does not give practical guidance on how to implement stakeholder involvement throughout the process of risk assessment and risk management. We urge the Commission to carefully expand upon its guidance on how to involve stakeholders and to clarify the nature of interactions between interested and affected parties. Most importantly, mechanisms of stakeholder involvement should be designed with the goal of preserving the critical role of science in risk assessment.

• The Commission report codifies the importance of stakeholder participation and explicitly anticipates a role for stakeholders in all phases of the risk assessment/risk management process. In light of the apparent increased importance of stakeholders there is a need to clarify their role, addressing: (1) criteria for determining both balance and legitimacy in the choice of representative stakeholders, and (2) power sharing, i.e., whether legislative changes are needed to give stakeholders advisory status or grant them the power to approve or disapprove regulatory decisions.

• The report falls short of providing a concrete scheme or evidence to support the conclusion that the projected process would actually provide time savings in the long run. Other issues not adequately addressed include the effective coordination among federal agencies, which requires changes in how government operates; the new directions in research, which require more funding support; and more stakeholder interactions, which require more time and resources. It is appropriate to lay down some alternatives leading to potential future implementations.

• We commend the Commission’s efforts to present a framework for risk management that incorporates a rigorous science-based approach to problem solving as well as full consideration of societal, economic, and cultural conditions and needs. With the goal of accounting for the connections between environmental health, human and economic well-being, and the processes by which our society’s actions create long-term changes, both beneficial and adverse, the Commission established an intriguing framework for making decisions on the reduction of risks to public health, safety, and the environment.

• We support the report’s recommendations on the inclusion of stakeholders in risk management processes conducted by any regulatory agency with jurisdiction over health, safety, and environmental issues. One of fundamental tenets of our organization’s Risk Principles is the need for an open public process with participation by stakeholders at every stage of the process. It is incumbent on stakeholders to clearly communicate their views and to provide information and analyses. We suggest that the Commission address the process for stakeholder selection and the means for participation.

• We compliment the Commission on recognizing that everything is connected. Tribal health and environmental equity are essentially synonymous, as our elders have said for thousands of years. They have also said that yes, everything is connected.

• We agree that an integrated approach to risk is needed. To do this, however, we should turn the entire paradigm around, and concentrate less on building a comprehensive set of micro-exposures (such as the TEAM approach, although this does, of course, provide useful information), and concentrate more on describing what we want to protect and all the ways it might be at risk.

• This report defines risk solely in terms of public health. This definition is too narrow. Our tribal risk model recognizes that contaminants pose risks to tribal rights, resources, health, and culture. Discussion should be added that recognizes that risk has broader definitions than simple mechanistic human exposure assessment.

• We strongly endorse the emphasis on stakeholder involvement at each step in the risk assessment and risk management process, although this was not carried through the report. Most of the document, while raising many important points, tends to remain within the conventional risk paradigm (hazard id, dose response, exposure assessment, and risk characterization). We would prefer to step a little further out of the "box" and add a component of cultural toxicity to the paradigm along with human and ecological toxicity.

• We strongly support the key observation that environmental problems must be addressed comprehensively rather than on a chemical-by-chemical, media-specific basis. The draft report almost completely fails to address solutions to the problems it identifies. Implementation deserves sustained attention in the report. The proposed risk management framework should be modified to (1) incorporate a more comprehensive approach to hazard identification, (2) incorporate effective incentives for generating the scientific information required for sensible decision-making, (3) address the resource constraints that bias and limit stakeholder participation in regulatory decisions, and (4) recognize and counteract the risk of increased transaction cost and "paralysis by analysis" that the framework poses in its current form.

• The framework is an important step in the direction of providing the components and philosophy of a decision-making approach, but a greater degree of specification would prevent unintended outcomes from occuring where the wrong process was taken.

• Advocating public involvement without establishing a process for involvement is politically and managerially unsound. The discussion of public involvement should be expanded to include elements of or alternatives for such a process or at least how such a process would be developed.

• The report underscores multimedia approaches and the provision of contexts for exposures. Without further development, however, it is too easy for this to be interpreted as the larger the problem, the less the marginal impact of any incremental additional contamination or exposure. This can obscure situations in which the mountain of contamination is increasing and any given addition has to be continually larger to be acknowledged.

• Ever since the environmental movement began, integration to avoid the pitfalls of categorical thinking and fragmentation has been a constant battle cry. No one denies that in many circumstances integration is absolutely critical to effective management. Most importantly, identifying how the institutional barriers to integration can be overcome is more significant than acknowledging that integration should occur.

• I would like to commend the Commission for its bold proposal of a new risk management framework. Given [our federal agency’s] environment, safety, and health challenges, complex web of stakeholders, and technical, budgetary and regulatory constraints, this risk management framework will be very useful.

• There is utility to the collaborative framework concept outlined in the report. If inflexible and prescriptive federal rules remain, collaboration at the local level will be meaningless. New statutory and regulatory flexibility for state and local decision making will be necessary to implement any such framework.

• The draft report suggests making stakeholders partners in risk assessment and risk management. This should not mean that risk assessment becomes a political process. Risk assesssment must be science based. Stakeholders can contribute by providing scientific information.

The Framework is reaffirmed and supported considerably better than in the draft report with principles, guidance regarding implementation, and examples. For example, Principles for Risk Management Decision-Making have been added, as have Guidelines for Stakeholder Involvement. Each of the six stages of the framework is described in detail, including questions to ask and considerations to address. The role of stakeholders in each stage of the Framework is explained. The connection of each stage to the Principles for Risk Management Decision-Making is clarified. The different contexts that should be considered (multisource, multimedia, multichemical, and multirisk) are clearly described. Examples are given that illustrate different contexts and indicate how creative, integrated strategies to risk reduction can be implemented.

We recognize that regulatory agencies operate under the constraints of their enabling statutes and regulatory policies and we recommend that to address environmental problems more comprehensively and in context, Congress should initiate joint oversight hearings and agencies should fully use their discretionary authority to expand stakeholder involvement and to address the most significant sources of risk. We also recognize that local stakeholders may need technical and sometimes financial assistance to be effective participants in risk management decisions. To avoid having risk assessments become too politicized, however, we note that stakeholders can contribute valuable information (about exposures, for example) to risk assessors but should not participate directly in the assessment itself.

We acknowledge the important role that cultural considerations play in risk management decision-making in addition to considerations of risk, feasibility, cost-effectiveness, etc., but we have chosen to restrict our definition of risk to that of human and ecological health in the mechanistic sense. We emphasize, however, that cultural considerations should be included in risk assessments, especially in terms of their impacts on exposure and susceptibility.

3. Risk Management and Regulatory Decision-Making

Communicating and Comparing Risks

• Identifying Risk Communication Needs

• There is little or no discussion of the role of the media in developing and shaping of perceptions of risk and of the need to bring reporters and others into the process in a way that recognizes their role but that reflects accurate and thoughtful information.

• We agree that better risk communication is needed. However, it is far harder for decision makers to understand other perspectives (once they’re convinced that their computer has provided them with the "right" answer) than it is for tribal technical staff to understand the details of risk assessment. As you are all too well aware, any disagreement with the computer’s answer, or with the assumptions of the modeler who designed the code, is labeled junk science. From a tribal perspective, over-reliance on numerical results is irrational and might be labeled junk ethics. Science should be done in service to values. Data are subservient to wisdom.

• Risk communication is a two-way process, and just as much emphasis needs to be placed on communicating tribal and community risks to managers and assessors as risk managers place on communicating probabilities to the public. Please do not overestimate the ability of risk managers to understand either the details of risk assessment or of tribal/community concerns, and do not underestimate the expertise of tribal technical staff. This applies not just to explaining risk results, but also (and even more importantly) to the improvement of the risk assessment methods themselves. We don’t need to be communicated at; we need to be able to use risk assessment as equal partners and peers.

• We would prefer to see less emphasis on funding risk communication and more on funding tribes to develop in-house technical expertise.

We strongly support two-way communication and a sense of respectful partnership. We have tried to avoid recommending greater quantitation than is needed for understanding risks and examining options for action. We added recognition of the important role that the media play in risk communication.

• Communicating About Risk by Comparing Different Kinds of Risk

• We fully support the report’s recommendations regarding the value of using comparative assessments of risk to convey information about the nature and magnitude of risks. We agree with the types of comparisons that the report recommends but suggests some additional information. Information about the benefits associated with the risk should also be provided along with information about "substitution" risks. That information is necessary to place the risk fully in context and to identify the potential tradeoffs inherent in any risk management decisions.

• When used appropriately, comparisons of unlike risks can be illustrative in certain situations and can convey a sense of magnitude in terms familiar to the recipients of that information. The Commission should acknowledge the potential usefulness of such risk comparisons and suggest appropriate scenarios for their use.

The Commission acknowledges the useful role that risk comparisons can make in risk communication, but we believe that comparing unlike risks is generally inappropriate. There are other useful methods of communicating a sense of magnitude, which we discuss. We also recognize the importance of understanding "substitution" risks, but have chosen to emphasize their consideration when we address examining options for risk management actions, not as a tool for communicating about the nature and magnitude of a particular risk.

• Need for a Common Metric

• We agree with the finding that the dichotomy in methods for assessing cancer and non-cancer risks causes inconsistencies in risk management actions and makes comparisons of risk difficult. A margin-of-exposure approach for carcinogens may be useful in addressing this dichotomy but potential confusion may arise when a risk manager is presented with somewhat inconsistent measures of risk for linear and non-linear carcinogenic agents. In addition, such a method must be clear about whether methods or assumptions are based on science or policy (i.e., defining an acceptable margin of exposure).

• A margin of exposure approach is a major departure from the standard risk assessment/risk management practice used in the evaluation and regulation of carcinogens. The draft document does not adequately describe the disadvantages of such a practice. The discussion overlooks the limitations of cancer bioassays and fails to put into perspective the purpose and utility of probabilistic cancer risk estimates. This [probabilistic] approach is very useful for risk management and should not be written off as "speculative." The non-probabilistic margin of exposure approach is useful in pointing out the dosage gap between animal testing and human exposures, but it does little to clarify whether there is a plausible risk for humans and how large this risk might be.

• There is no scientific evidence for potential risk and it should not be used in public policy. Evidence for health effects at low dose, by traditional scientific interpretation, is negative. We believe public policy in the United States and Europe would have followed a different course if the public had been told about the successful experience regulating the natural carcinogens arsenic, radon, and aflatoxin. Restoring public trust after years of misleading science advice and failing policy is the most urgent issue. The U.S. National Institutes of Health should lead the science advice. EPA should not be involved in risk assessment or in science advice in the future.

• The draft report strongly supports margin of exposure as a cancer risk characterization tool but fails to provide any guidance on how to use such a comparison in risk management.

• When providing guidance to regulatory agencies, the document contradicts itself by relying solely upon probabilistic estimates of cancer risk for decision making, rather than also using the margin of exposure approach.

• Its use should be limited to helping to provide perspective on probabilistic risk estimates for non-genotoxic carcinogens whose mechanistic database is sufficiently developed to demonstrate a strong likelihood for a threshold dose for cancer.

• We strongly support the Commission’s recommendations to develop compatible protocols for evaluating cancer and noncancer health effects, so that the two types of risks may be compared. However, the Commission should stress the importance of being able to use all available scientific information and, in the absence of data to the contrary, to incorporate conservative toxicity and exposure assumptions to protect human health.

• The recommendation to adopt a "margin of exposure" type analysis for carcinogens essentially gives up on producing quantitative estimates (uncertain though they are) of the consequences of interest for policy how many cases of cancer do we expect may be caused by a specific exposure, and how may cases might we be able to prevent by various measures to control that exposure? What we need to do for nongenetically acting carcinogens and for the array of noncancer effects is to do the fundamental research to eventually convert the old NOAEL/uncertainty factor system into one that makes quantitative predictions of the likely incidence of harm.

• The draft report leaves readers with the impression that margin of exposure is essentially interchangeable with the margin of protection. Using a margin of exposure for comparative risk purposes presumes that exposure is a perfect surrogate for response; thus, it assumes, albeit implicitly, that all dose-response relationships are linear. Although the margin of exposure approach would give both cancer and noncancer endpoints a common metric, it provides us with an essentially linear metric, the worst possible metric to be using for noncancer endoints, thereby trivializing all of toxicology.

This section was changed to emphasize that we recommend common metrics as risk communication tools and not necessarily as a substitute for current regulatory practices. We note that the U.S. Environmental Protection Agency recommends and uses a margin of exposure approach. We also point out that a margin of protection approach is used everywhere (except in the U.S.) to evaluate both carcinogens and noncarcinogens. Probabilistic estimates of risk are, after all, upper bounds on expected risk, not estimates of actual risk or of the number of cases of cancer we expect to occur. We strive to indicate that the margin of exposure is not the same as the margin of protection, and that basing risk management decisions on such metrics must include evaluations of the available scientific information.

Comparative Risk Analysis for Risk Management Priority-Setting

• Risk-Based Priorities and Resource Allocation

• We share the Commission’s support for the use of comparative risk assessment by federal agencies for prioritization purposes but disagree with the Commission that comparative risk should be conducted only on a demonstration or experimental basis. We believe that there has been enough experience with comparative risk to begin using it now to set priorities among agency activities.

• The procedure for comparative risk assessment for priority-setting, while intuitively appealing, is extremely problematic. As an example, consider the problem of ranking the relative importance of controlling endocrine disruption and particulate air pollution. The former is poorly undertood, but has the potential to impact the entire human race and all living animals. The latter problem is fairly clear, and we have quite precise estimates of the number of deaths annually attributable to particulate pollution. Which poses the greater risk? That question is currently impossible to answer. Because of the reliance on data availability, a precautionary approach is generally not supported by comparative risk assessment.

This section now acknowledges the disparate views on the utility of comparative risk analysis, but retains the recommendation that comparative risk analysis should be conducted on a demonstration basis, which seems appropriate in light of those disparate views.

Strategies for Managing Risk

• Risk Management Options: Alternatives to Command and Control

• We strongly support the Commission’s recommendation that there should be increased reliance on market-based alternatives to command and control regulation.

• We agree with the need to explore alternatives to command-and-control regulations. The tools identified in the report as alternatives to command-and-control systems should be risk-based or used to identify and respond to risk reduction opportunities. The report did not adequately emphasize the goal of risk reduction as a key component in each of these tools. The usefulness of a tool should be measured by its ability to identify, compare, and reduce risks in a way that permits the optimal use of resources.

This section now acknowledges the important role that command-and-control regulation has played in environmental protection. The importance of evaluating alternative tools for reliability, feasibility, effectiveness, and efficiency is emphasized. Material on the roles that research, monitoring, and surveillance play in managing risks was added.

• Bright Lines and Regulatory Standards for Risk Management

• The recommendation on bright lines is equivocal and needs further clarification. Bright lines may be successfully used only as a criterion to differentiate those problems, issues, sites, etc., that do not require further inquiry from those that require further evaluation by a more exhaustive and extensive analytic process to determine if action is necessary. They should be used only for screening, and not for risk management decision-making.

• We do not oppose the use of bright lines when they are used only for screening purposes and are based on scientifically sound methods. The final report should include language warning against the inappropriate use of bright lines for non-screening purposes.

• The Commission should endorse the use of bright lines only in screening analyses to determine whether a more detailed risk assessment is required, and clarify its discussion on the use of bright lines as screening tools and specify how to use these tools flexibly.

• We commend the Commission for recognizing both the benefits and limitations of bright lines and concur that bright lines are necessary as guidelines in exposure limits and that agencies need to be flexible in their interpretation of bright lines. However, the Commission should recognize that qualified staff are needed to interpret and qualify the range of uncertainty and variability that underlies the bright line. The less we depend upon bright lines, the greater our costs will be.

• The Commission should give more thought to its recommendation about multiple bright lines and ranges of bright lines. It is our experience that given a range of "acceptable risks," the upper end of a risk range becomes the de facto bright line. This practically eliminates the flexibility that in thoery results from the use of a range. Further, the use of a risk range may result in inconsistent application of the range and questions of fairness, both real and perceived.

This section was changed to clarify that the Commission opposes the use of inflexible risk-based bright lines because risk estimates are so uncertain and variable, but that we believe measurable, concentration-based bright lines (standards) can be useful. We note, however, that bright lines or ranges of bright lines tied to specific exposure concentrations should be used only as guideposts or goals for decision-making, and should not be applied inflexibly. We also note that risk-based bright lines can be useful for screening, but should not be the sole basis of risk management decision-making.

Judicial Review of Regulatory Decisions to Manage Risk

• This section reflects the personal views of the Commission members on legislative initiatives considered by the 104th Congress. The legislative mandate creating the Commission did not envision comments on pending legislation as an appropriate issue for Commission consideration. It would be useful to revise the section on judicial review to assess the impact of the process on risk assessment/risk management practices historically. It would also be of value to assess judicial review in the context of the internal debate within EPA between those advocating inclusion of alternative interpretations and uncertainties in a regulatory document and those who believe such inclusions weaken the agency’s position during judicial review.

• We find the report’s section on judicial review to be very good and agree with most of its recommendations, for example, that final agency action must be final, that a Congressional mandate to all agencies to follow detailed cost-benefit and risk assessment requirements that would then be judicially reviewable is unwise and inappropriate, and that the standard of judicial review should not be expanded.

• This section should be refocussed away from issues raised in specific past legislative proposals toward articulating the general principles that should govern judicial review in the event Congress adopts a new framework for risk assessment and risk management decisions. The Commission’s generalizations simply do not represent an accurate description of the proposals that have received most serious consideration. This issue lies outside the scope of the Commission’s mandate and fails to recognize the extensive oversight role that the courts now play under existing health, safety, and environmental statutes.

• The judicial review chapter of the draft report is a great disappointment and should be deleted or revised substantially. The chapter is extremely one-sided and simply adopts the administration’s position on the regulatory reform debate without providing a useful analysis to illuminate the issues.

The judicial review section was changed to de-emphasize previously proposed regulatory reform legislation, but the Commission believes that the issues raised in those earlier proposals may very well surface again in future proposals and, more importantly, that those issues go well beyond any particular legislation. Those issues served to focus the debate on the appropriate role of judicial review of agency action in the regulatory process. The section does not intend to overlook or impugn the important oversight role that the courts play at present, and we believe that failing to address judicial review-related issues in our report would constitute failing to acknowledge that important role.

The section has been changed to clarify that the Commission does not support strict decisional criteria. The section addressing the impact of increased litigation on agencies, parties, and the courts has been deleted.

• Premature Interruption of the Administrative Process

• The regulatory reform bills did not call for the interruption of the administrative process by premature judicial review. Instead, the bills would have required that risk assessments and benefit-cost analyses be made part of the rulemaking record, to be considered in connection with judicial review of final rules.

This statement is not true. Early proposals would have redefined final agency action. Final agency action must be final, and should not be redefined to permit review of agency action and discrete issues until after agency action is complete.

• The Nature and Extent of Judicial Review

• All environmental laws, including recent safe drinking water and food safety legislation, employ risk assessment and scientific terminology that are ultimately reviewable in courts. The Commission’s recommendation is one-sided and fails to consider the need for accountability of federal programs. Arguably, smarter programs may result in less litigation. Nonetheless, any statutory change is likely to increase litigation at first.

• If the well-established principles for judicial review developed under the APA and individual laws are reflected in regulatory reform legislation, the courts would continue to perform their long-standing responsibility to assure principled decision-making by agencies without creating new litigation opportunities or subjecting rules to unproductive judicial oversight.

• If an environmental rule were based on more sound science and greater consideration of costs and benefits, industry may well find that the final rule is reasonable enough that it should not be challenged. Thus, litigation actually may decrease.

New litigation opportunities would not result from the appropriate application of the APA. They could result from the establishment of detailed decisional criteria and procedural requirements that would supplement all existing enabling statutes, which were included in versions of the draft regulatory reform legislation.

• Standard for Judicial Review

• The general principles that govern judicial review should be based upon current standards of judicial review of administrative action.

• The attack on the "substantial evidence" test is impassioned but misguided. There is no evidence to support that under the substantial evidence test, courts would replace administrative agencies as the ultimate decision-maker on highly technical issues.

The Commission believes that proposals to require substantial support and in the rulemaking file for agency findings and conclusions at the expense of the arbitrary and capricious standard could be interpreted to mean that the substantial evidence standard would be expanded beyong formal hearings to all rulemakings, reducing a court’s ability to defer to agency decisions.

• Consensual Approaches as Alternatives to Increased Judicial Review

• Our agency makes extensive use of consensual approaches as regulatory alternatives. For example, through meetings and cooperation with industry, we have addressed a number of safety hazards. This has allowed us to address these hazards quickly and without resorting to rulemaking. We also work extensively with consensus voluntary standards groups to develop effective voluntary standards.

Consensual approaches as alternatives to increased judicial review would involve stakeholders in decision-making, as recommended by the Commission’s Risk Management Framework.

4. Uses and Limitations of Risk Assessment for Risk Management Decision-Making

Toxicity Assessment

• Using Rodent Tests to Predict Human Cancer Risk

• There are many important chemicals that are candidates for a future version of the table listing potentially irrelevant mechanisms, tumors, and chemicals. Further research may show that tumors observed in rodents or even in humans at extremely high doses are not relevant to human cancer risk at low doses by obtaining the understanding that different biological mechanisms govern the development of these tumors and that these mechanisms are not operative for humans at low doses.

• If animal models are wrong, deficient, limited, etc., then addressing these concerns, i.e., what criteria or research is needed to identify when the models are wrong, should be a major focus of the report, not that the results of animal models should be discarded when they are not relevant to humans. Can the report provide criteria or guidelines as to when results can or cannot be used? Most, importantly, what are the alternatives to animal models?

• We concur with the recommendation to classify as irrelevant to human cancer risk assessment a limited set of rodent cancer responses where the physiological mechanisms for tumor development in the rodent do not have a corresponding human mechanism. Some education and risk communication may be required on this point, however, so that the public does not lose confidence in other rodent bioassay data. We strongly support the Commission’s recommendations to focus on testing

• We agree with the recommendation that certain rodent responses should be classified as irrelevant to human cancer risk assessment. This recommendation is particularly applicable in cases where there are tumors that result from mechanisms that are unlikely to occur in humans or that occur at very high doses that are irrelevant to human exposures.

• The Commission’s recommendation is critically important to ensure that risk assessments are realistic and to help policymakers understand true human cancer risks and set priorities among their efforts accordingly.

This section was changed to emphasize the importance of rodent bioassay results and the role that they can play in identifying carcinogens, along with knowledge of mechanisms of action. A table listing known human carcinogens was added. Material was added that describes weight-of-evidence considerations for evaluating a chemical’s or a tumor’s human relevance.

• Evaluating Chemical Mixtures

• We strongly support the Commission’s recommendations to focus on testing chemical mixtures. This type of research will provide better information on potential interactions among chemicals and help reduce uncertainty. However, testing mixtures adds costs and the results are hard to generalize to slightly different mixtures in different settings.

• We agree with the recommendations to test mixtures and that, in some cases, adding together risks from individual chemicals is generally appropriate and is unlikely to result in an underestimate of risk.

• It is neither appropriate nor workable to test environmental mixtures that exist at low concentrations. As the Commission itself notes, for "environmental" mixtures (as distinct from consumer or occupational exposure), effects of mixtures seen at high doses are probably inappropriate for conducting risk assessments for much lower levels of actual exposure because the interactions that may occur experimentally at high doses would not occur at low doses. Furthermore, mixture permutations are infinite and their variability huge.

• We agree with the finding that dose response data are needed for mixtures. However, we urge caution about the recommendation to consider risks separately if mechanisms are different. Chemicals do not usually act through single mechnisms, and secondary mechanisms may exist. The same is true of primary and secondary target tissues and primary and secondary effects.

• The draft report is a bit too strongly slanted toward the contentions that: (1) at low doses (i.e., environmentally relevant exposure levels), toxicologic interactions are unlikely, and (2) at low doses, interactions, if present, are mostly likely additive in nature. These contentions, in my view, are not necessarily true.

The Commission disagrees that testing environmental mixtures is of no utility; we recognize that testing at high doses may not be relevant in a dose-response sense to low environmental doses but it would certaintly be useful as a screening tool, to help identify exposures that should be examined or controlled more carefully. We also recognize that mixtures are highly variable and suggest that methods such as Monte Carlo analyses may prove useful in attempts to generalize testing results. We added material suggesting how a coordinated research program could facilitate enlarging the mixture toxicity data base.

The Commission continues to support the view that low-do0se interactions, if they occur, are unlikely to be detectable and that assuming response additivity is precautionary. The data submitted by commenters supported our conclusions.

• Accounting for Differences in Susceptibility

• We agree with the recommendation that risk assessments should include considerations of differences in susceptibility. Population subgroups must be considered and we agree that recognition of subgroup susceptibility should not "result in more stringent regulations." However, we object to the reference to additional "bright lines." Instead of using the most susceptible subpopulations to justify the most stringent bright lines, we believe that knowledge about differences in susceptibility should be used to identify where more stringent restrictions may be needed. Such situations should be dealt with on a case-by-case basis and considering all aspects of the situation. Thus we believe that where appropriate, knowledge of differences in susceptibility should be used to tailor risk management responses for identified susceptible subpopulations.

• Environmental equity needs further interpretation. Suggesting separate bright lines for protecting susceptible subgroups should not inadvertently impede their economic development or career opportunities. Further deliberation is needed to avoid unintended consequences.

• Susceptible subpopulations should not become a vehicle for maintaining the conservatism of the maximally exposed individual. The Commission should make it clear that risk assessments that include susceptible subpopulations should not be used as the basis for risk management actions that treat all exposed individuals as members of the susceptible population. Actions should be specifically targeted towards the susceptible subpopulation.

• This finding and recommendation imply that risk assessments do not take into consideration host susceptibility factors. Standard risk assessment practice for noncancer endpoints is to include a 10-fold safety factor for inter-individual differences in susceptibility to the chemical. This approach should be pointed out and its merits/faults discussed.

We now indicate in our finding that current regulatory approaches for reducing risks associated with chemical exposures generally do not include information on differences in individual susceptibility or encourage getting evidence to identify them, and note that in the absence of specific information about differences in susceptibility, risk assessments rely on assumptions and safety factors that are presumed to be protective of sensitive individuals. We concur that susceptible subpopulations should not be used as the basis of risk management actions that treat all members of an exposed population as especially sensitive, and that risk management responses should be tailored for specific identified subpopulations.

Exposure Assessment

• Design of Exposure Assessments to Meet Risk Management Goals

• We agree that exposure assessments should be designed to be commensurate with the needs of the risk management decisions at issue and that the use of a tiered approach in exposure assessments is a good strategy for effective resource allocation.

• Using Realistic Exposure Scenarios

• We agree that exposure assessments should not be based on a maximally exposed individual. Using high-end exposure estimates for screening assessments and distributions of a population’s exposures for more refined assessments is reasonable.

• We commend the Commission on its recommendation regarding population-based exposure. Another terminology might be receptor-based as opposed to source-based exposure assessment. The TEAM studies have clearly demonstrated that source-based exposure assessment can be seriously misleading when compared to personal measurement results.

We clarified our recommendation to indicate that exposure assessments should rely on population exposure data where possible instead of assumptions about exposure derived from sources and models. We include the terms receptor-based and source-based and refer to the TEAM study as a good example of receptor-based exposure assessment. We emphasize that considering the size of an exposed population is important in addition to considering the distributions of its exposures.

• Identifying Highly Exposed Populations

A reference to the Food Quality and Protection Act of 1996 and its role in addressing susceptible populations was added. Our reference to education as a risk management tool was clarified to indicate that there are other alternatives and that education might not be considered appropriate by stakeholders in all cases.

Ecological Risk Assessment

• Framework for Evaluating Ecological Risk

• We agree that EPA and other agencies and interested parties should continue to work together to refine ecological risk assessment and risk management approaches. Guidance on problem formulation, methods, characterization of uncertainty, and the appropriate role of stakeholder participation in the process is necessary and should incorporate the views and expertise of all practitioners of ecological risk assessment/risk management.

• We applaud the Commission for including a discussion of ecological risk asssessment. However, ecological receptors will continue to be underrepresented in environmental regulations as long as there is a lack of focus on this topic. This section should be expanded and revised in keeping with the level of discussion on direct human health issues. Formal recommendations on support for basic research tools and data should be included in the final report.

• This report does not engage in a robust and comprehensive discussion of ecological risk assessments, alluding to only a single purpose for conducting an ecological risk assessment protecting resources in terms of their direct human uses. There are multiple and legitimate purposes that various constituencies may have in conducting ecological risk assessments. This subject needs more evaluation and discussion.

• The limited scope of the Commission’s review is illustrated by the sole identification of the index of biotic integrity as an "important diagnostic tool," giving the impression that Karr’s index is a preferred choice. There are numerous other diagnostic tools available that may be equally or more applicable and useful.

• The link between the discussion phase and the risk management phase in the framework should be illustrated with only a downward arrow. An arrow going in both directions communicates that a risk management decision is never final or complete.

The Commission has not expanded this section, although we continue to refer the reader to other excellent sources of information. We were not mandated to address ecological risk assessment, but we would have been remiss not to include it. An acknowledgment and analysis of the U.S. Environmental Protection Agency’s new ecological risk assessment guidance was added and the arrow on the framework was modified. We acknowledge that there are other purposes for conducting ecological risk assessments besides protecting resources in terms of their direct human uses.

Environmental Hazards Other Than Chemicals

• Risks from Radiation Hazards

• The report focuses on the differences between radiation and chemical standards and the level of cancer risks at the exposure limits. There is no discussion of differences in methodology, data, and assumptions from which standards are derived. Harmonization of the fields of radiation and chemical risk assessments will remain an illusive goal without these basic differences being articulated and discussed.

• The Commission did not recognized standards recommended by authoritative bodies in radiation health protection such as the ICRP and the NCRP.

• The report downplays the ALARA principle that underpins all radiation standards.

A discussion of the differences in approaches to deriving radiation and chemical standards, a reference to the ICRP and NCRP standards, and an emphasis on ALARA were added to the section.

• Risks from Microorganisms

Material has been added that describes ongoing efforts intended to strengthen microbiological risk assessment.

Risk Characterization

• Effective Risk Characterization to Support Decision-Making

• Risk managers and the public need to understand, and have the right to know, what the weight of the scientific evidence says about a given health or environmental risk. Unfortunately, current agency practice and EPA’s recent cancer guidelines appear to ignore the weight-of-the-evidence approach; as a result, decision makers and the public will continue to be misinformed and billions of dollars will be wasted on excessively hypothetical and exaggerated statements of cancer risks.

• Failure to achieve a clear separation of scientific evidence from policy considerations has negative implications beyond impaired credibility: it inevitably obscures the overall level of uncertainty in the assessment and will influence consideration of societal tradeoffs and ultimate decision-making. The Commission recommends "weighing the evidence," but weighing the evidence cannot be done unless it is kept separate from assumptions. Clear understanding of uncertainties is a key component of considering tradeoffs because the public may often opt for deferring measures to address very uncertain gains in order to devote scarce resources to achieve more pressing and certain gains.

This section was changed to emphasize the role of the precautionary principle and the importance of considering the weight of the scientific evidence. Material on uncertainty was moved to the uncertainty section (next) and material on resolving the need for decision-making versus collecting more data was moved to the new value-of-information section.

• Characterizing the Uncertainty Associated with Risk Estimates

• The emhasis of your major statement on uncertainty that most risk assessments do not need quantitative uncertainty evaluation is discouraging to those of us who believe that making our best efforts to fairly and honestly express uncertainties is a prime duty of any technical analyst. It is probably quite literally true that most risk control decisions would not be changed by a quantitative uncertainty analysis. But we should be encouraging risk managers to face the facts of the imprecision of our estimates as a fundamental part of appropriately communicating the limitations of our available information.

• The report is a step backward in regulatory decision-making because of a serious inconsistency. The Commission recommends against performing full analyses of uncertainties because it would be difficult and confusing to the decision-maker, but at the same time they encourage risk comparison, risk ranking, and cost-effectiveness or cost-benefit analyses that are meaningless without explicit treatment of uncertainties. The issue is not so much the treatement of variability, which is sometimes currently done in exposure assessment, but of fundamental epistemic uncertainties, for example, about dose-response relationships.

• We urge the Commission to reconsider its preference for providing qualititative descriptions of the range and distribution of risks over quantitative ones. Risk assessors should provide such quantitative estimates where appropriate. Qualitative information, including descriptions of the major assumptions, uncertainties, and policy judgments embodied in a risk assessment, are always necessary to more clearly describe a risk and place it in context. However, a point estimate accompanied by qualitative information describing uncertainties does not sufficiently inform the risk manager or the public. The apprehension of risk managers regarding quantitative estimates of the range is understandable. Point estimates provide a sense of certainty and the appearance of consensus on the estimate, but that sense is unreal. The Commission should suggest a paradigm for how risk management decisions can be made from ranges and distributions of risk. At a minimum, the Commission should recommend that if a full range and distribution of estimates are not going to be provided, more than one point should be identified, including a central or most plausible estimate along with a high-end estimate.

• Formal uncertainty analyses have much to add to the risk assessment and risk management process. If risk managers’ eyes start to glaze over when they are presented with such information, the fault is ours, not theirs, and the obligation is ours to improve communication and raise their level of awareness and appreciation for the real value contributed by such analyses.

• The Commission should recommend that uncertainties be quantified for both exposure and dose response when comparing risks. If uncertainty is not quantified, then it is not possible to know whether one risk is truly higher or lower than the other. When uncertainty is quantified, it is possible to determine whether one risk is likely to be higher or lower, or whether the risks are indistinguishable from each other given scientific uncertainties. This information is useful to policy-makers. In the former case, the risk assessment may be influential in the decision process. In the second case, other factors, such as equity issues, economics, etc., may provide a stronger basis for making a decision.

• As long as cancer risk calculations are being made by risk assessors and used by risk managers, these calculations need to be made by appropriate methods and uncertainties need to be carefully explained by the risk assessors to the risk managers and to stakeholders. I do not believe progress will be made by retreating to qualitative discussions of uncertainty, especially if the use of a quantitative point estimate of cancer risk is retained.

• While it is appropriate to avoid excessive use of complex probabilistic methods, quantitative methods for uncertainty analysis can enable improved communication between scientists/risk assessors and the decision-makers and stakeholders. The draft recognizes this potential for economic analysis and exposure analysis. The draft should be revised so that this point is also made for uncertainties about the dose-response relationships for cancer and other health effects.

• Quantitative uncertainty analysis may not be necessary in all instances, but should be conducted for risk assessments underlying major regulatory decisions that include parameters with significant uncertainty.

This section was changed to clarify that the Commission supports using probability distributions of the variability in a population’s exposures as appropriate to enhance characterization of exposures and communication of risks. We continue to recommend against routine use of formal quantitative analysis of uncertainties in risk estimation, particularly that related to evaluating toxicity, relying on narratives instead.

• Value of Obtaining Additional Information

• The major analytic weakness of the report is its silence on the promise of formal "value of information" methods to enhance the quality of risk management decisions when the stakes are high and the quality of available information is low.

This section was added to clarify the conflict between the need to make decisions in the face of uncertainty and the need to gather more information on which to base decisions.

5. Uses and Limitations of Economic Analysis for Risk Management Decision-Making

Benefit-Cost and Cost-Effectiveness Analysis

Only minor additions and clarifications have been made.

• Useful Roles in Regulatory Decision-Making

• We strongly support the report’s conclusions regarding the importance of benefit-cost and cost-effectiveness analysis in making regulatory decisions. We also agree that not all costs and benefits can be assigned monetary values; nonmonetary values should be included in decision-making nonetheless. When making regulatory decisions, the decision-maker should be able to determine whether the benefits of a rule including non-quantifiable factors justify the costs.

• Protecting people’s health and safety and the environment should always be the primary goal of risk management regulation but economic analysis should always be a factor in deciding when and how to take regulatory action. The role of economic analysis in regulatory decisions should be clearly set out by the law.

• The report does not take a strong enough position on the appropriate role that cost-benefit analysis can play in regulatory decision-making, and does not confront existing statutory barriers to the use of economic analysis in policy-making. The final report should call for expansion of the use of cost-benefit analysis in risk management decision-making.

• The report inappropriately lumps benefit-cost analysis and cost-effectiveness analysis under the single heading of "economic analysis" and often refers to them jointly. In fact, these two policy-making tools are very different and the report should clearly distinguish between them. The final report should make clear the primary role benefit-cost analysis can play in formulating policies that maximize net benefits to society, unlike cost-effectiveness analysis.

• We recognize the necessity to convert all of our risk concerns into dollars if they are to play any real role in cost-benefit analysis. We also recognize, however, that risks and lives are not linear functions. There are miminum units of aggregation that must be evaluated (entire habitats, for instance, rather than cubic meters of soil). The inevitable consequence of setting a monetary value on a human life is the reductionist conclusion that a value per minute of life lost is a valid application.

• Distributions of Costs and Benefits

• Weighting benefits and costs quantitatively based on equity would be highly subjective and inappropriate. The better approach is to inform the decision-maker about who receives benefits and who pays the costs in a more appropriate way, to be used explicitly in making the regulatory decision.

Uncertainty and Inconsistency in Economic Analysis

• Characterizing the Uncertainty Associated with Cost and Benefit Estimates

• We agree that the preference for transparency in risk assessments applies to economic analyses as well as risk assessment and that there is a need to develop consistent methods for conducting such analyses for use in regulatory decisions. The value of risks should be stated explicitly and valued using best estimates or ranges of estimates and using consistent procedures and basic assumptions.

• Inconsistencies in Monetary Valuation of Benefits

• We agree in part with the recommendation to state explicitly the value of mortality risks but we would also recommend more emphasis on the use of illness risks and ecological risks. The costs of medical care and workdays lost to illness are probably large by comparison to mortality costs, and these are worth considering despite some added controversy. We hope to see a caution in the final report against too-rigid protocols, which reduce economists’ flexibility to choose the data and analytical approach that best fit the problem.

• Linking Risk Assessment and Economic Analysis

• Willingness to pay should be applied to the precautionary principle, so that regulation for protectiveness reasons (beyond a statistical NOEL level) should include a monetary penalty for data gaps when used in CBA.

• The problem with the cost-per-life-saved approach is that protecting small groups of people would be cost-ineffective. The population risk might be a more relevant measure in some situations. Other measures might be total environmental contaminant burden, total mutation burden on a small tribal gene pool, etc.

• There is a need to more closely coordinate the analyses of the risk assessors and the economists. There is a need to find ways to present risk information and to craft economic analyses to use that information in a way that will best enable a risk manager to determine which risk management alternative is the most cost-effective or provides the best balance of benefits and costs.

6. The Role of Peer Review in Regulatory Decision-Making

This section was changed to three findings and recommendations instead of one, targeting and clarifying the specific issues raised by commenters.

Improving the Quality of Regulatory Decisions

• We have serious concerns regarding the recommendation on the effect of potential conflicts of interest on the eligibility of potential peer reviewers. The recommendation to exclude potential peer reviewers with financial conflicts of interest is wrong. Full disclosure of all financial or organizational interests best serves as a criterion for selection of peer reviewers. Peer reviewers are not judges or juries, they are expert witnesses.

• We disagree with the recommendation that calls for the disqualification of any potential peer reviewer with "clear conflicts of financial interest." It is important to recognize that federal agencies use scientists with expertise in specific technical disciplines who are employed by industry, consulting firms, advocacy groups, and academia. All of these affiliations could be construed to constitute a potential conflict of financial interest.

• We strongly endorse the report’s recommendation to expand the use of peer review to economic analyses, social science information, and the use of scientific and economic data in decision-making.

This section was changed to clarify which kinds of financial conflicts we believe should disqualify peer reviewers and which kinds do not.

The Conduct and Effectiveness of Peer Review

This section was not substantially changed.

Evaluating the Use of Peer Review and of Scientific and Economic Analysis in Regulatory Decision-Making

This section was changed to clarify that we think advisory groups should review the use of peer review information in regulatory decisions, not review the decisions themselves.

7. Recommendations for Specific Regulatory Agencies and Programs

Environmental Protection Agency

• Office of Air and Radiation

• We agree with the Commission’s support of the use of screening assessments to determine which facilities within source categories or subcategories need to take additional risk assessment or risk management steps. We believe MACT standards will greatly reduce emissions of hazardous air pollutants, and that further reductions should be based on site-specific considerations of remaining risk. We agree with the proposal of a tiered scheme to determine and manage residual risk after implementation of MACT.

• The report should acknowledge that although many questions still remain about indoor air quality, that significant progress has been made. Because consumer products used in the home may release chemical and biological pollutants, and the design of the home itself may contribute to indoor air quality problems, CPSC has taken an active role in protecting consumers from illnesses and deaths associated with poor indoor air quality.

The tiered scheme for assessing residual risks proposed by the Commission was revised, clarified, and streamlined. A section addressing critical data gaps and needs was added. The recommendation regarding risks from indoor air was expanded to include other regulatory agencies besides EPA and to acknowledge the work that CPSC has done in this area.

• Superfund

• It is inaccurate to say that it has been difficult to revise remedies. Revisiting remedies has always been possible, and is in fact being encouraged at sites with ground-water contamination where new science or remediation technologies hold promise for cost savings.

• Preparing an overall annual budget estimate, presumably including actions at sites by other federal agencies, would obscure the Commission’s emphasis on risk-based planning and actions in a fog of numbers of widely differing scales. EPA is committed to risk-based priority-setting, planning, and budgeting. At the same time, our experience to date in the risk-based allocation of cleanup funds, which the Commission commends, has demonstrated that there is great difficulty using one risk algorithm in considering emergency responses and long-term cleanup actions.

• We support Superfund’s goal of protecting human health and the environment, and agree with the Commission that Superfund should be amended to require site-specific and risk-based remediation standards. The remedy process should provide protection of human health and the environment through methods that are practical and achievable in a cost-effective fashion.

• We agree that the interpretation of ARARs has caused problems and led to needless cleanup. ARARs dramatically affect the final cost of remediation. They can result in remedy selection that is overly costly and technically infeasible. We believe that the "relevant and appropriate" language should be deleted from the law and that "applicable" requirements should be subject to a specific set of remedy-selection balancing criteria.

This section now acknowledges EPA’s new policies on revising remedies. We clearly distinguish between emergency responses, applauding their goal of expeditiously removing obvious sources of exposure, and long-term cleanup. The costs and disputes about remedy selection apply almost entirely to the long-term cleanup. The recommendation about brownfields was caveated to note that purchasers of brownfields cannot add to or exacerbate contamination and that access must be provided to authorities to ensure that no migration or increase in contamination occurs. The recommendation regarding an overall annual budget for Superfund has been deleted.

• Office of Prevention, Pesticides and Toxic Substances

• We do not support the Comission’s recommendation that TSCA be reopened and updated. Although we agree that TSCA should be administered to reflect advances in toxicology over the last 20 years, legislative action is not necessary to effect these changes. The current statute provides sufficient flexibility to update the TSCA requirements. Sufficient flexibility exists under section 4 to update tests that are being mandated. EPA is already updating requirements for reporting under sections 8(a), (d), and (e), and the agency is considering several changes to the section 4 requirements.

This section now acknowledges the changes in the Delaney clause that were made with the Food Quality and Protection Act of 1996, noting their responsiveness to the recommendations in our Draft Report. Instead of asking Congress to update TSCA, the Commission recommends a focussed stakeholder process to review the Act and its implementation.

• Office of Water

This section acknowledges the Safe Drinking Water Act amendments passed in 1996 after our Draft Report was released. As recommended in the Commission’s testimony to the Senate, which was cited in the report accompanying the 1996 amendments, the act recognizes that cost and risk are not the only factors that need to be considered in evaluating environmental programs and that other factors, including values and equity, must also be considered.

Occupational Safety and Health Administration

• We support the recommendation that OSHA and NIOSH facilitate effective collaboration to help guide OSHA’s regulatory needs and NIOSH’s research efforts. A cooperative and integrated approach to the nation’s occupational issues would result in more cost-effective actions.

• The allegation in the report that there is substantial underreporting of job-related injuries is unfounded. The Commission offers no evidence to support this statement.

• We support the Commission’s recommendation that OSHA develop a guideline that lays out scientific and policy defaults. Currently, we are supporting a larger industry intitiative to develop a process to update the outdated permissible exposure limits. One part of the industry-identified process is the development of guidance to assist the agency in risk assessment and cost-benefit analysis.

• The report makes three valuable recommendations for improving the administration of the Occupational Safety and Health Act, calling for better surveillance and intervention-effectiveness research, better coordination between NIOSH and OSHA, and creation of health assessment guidelines. We support these recommendations.

• Recommendations for additional OSHA and NIOSH actions, such as increased surveillance, intervention effectivess research, evaluations, and guidelines for decision rules should be accompanied by recommendations to Congress for additional resources to carry these steps out.

• The report should be amended to clearly state that OSHA needs to reduce the permissible exposure limits to many common chemicals. Risks to workers from prevailing chemical exposure levels are thousands to a million times greater than those addressed by controversial actions in the environmental health arena. Efforts to update these limits have been stalled for decades, although public health actions addressing those few substances for which OSHA has set standards have been successful. The OSHA PEL update project of 1984 had considerable industry support for adopting tighter consensus exposure limits for several hundred common substances. Nevertheless, this modest improvement in protection was blocked by the courts because risk assessment and feasibility analysis were judged to be insufficiently elaborate.

This section was modified to include some minor additions and modifications to the text supplied by OSHA and NIOSH staff. The statement that job-related injuries are under-reported was made by OSHA staff, so we changed our statement to read that such injuries are "considered" to be under-reported. We acknowledge that additional OSHA and NIOSH actions such as increased surveillance will require additional resources. A new finding and recommendation to design a streamlined PEL update process, involving stakeholders and possible Congressionial action, was added.

Food and Drug Administration

• Given the continuing vagaries and ambiguities of risk assessment, fully recognized in this draft report, including the rather systemic inability to address cumulative exposure to multiple carcinogens, there are no policy jusfications for interfering with the applications of the Delaney clause [besides those covered by the Food Quality and Protection Act of 1996].

• The Food Quality and Protection Act of 1996 leaves the text of the Delaney clause unchanged and thus leaves unaddressed the issues surrounding its application to substances that fall within the statutory definition of "food additive." My concern is that passage of legislation freeing EPA from the constraints of the Delaney clause will be cited by defenders of the clause as confirmation that action by Congress is necessary to permit FDA to take into account any of the scientific advances that are discussed in the toxicity assessment section of the Commission’s draft report. This could leave FDA even more constrained than it has been in the past. The Commission’s recommendation may be cited in support of this result. It will be more difficult to muster congressional support for any change now, precisely because the coverage of the Delaney clause has shrunk.

• By limiting the Delaney clause to additives "found" to "induce cancer," Congress clearly contemplated that the responsible agency would conduct a meaningful, rigorous evaluation of evidence concerning carcinogenicity. The statute does not prescribe how to conduct this evaluation and this comports with a clear Congressional expectation that the nature of the inquiry will evolve as scientific understanding of carcinogenesis advances. For this reason, although one may most certainly contest the scientific underpinnings of the Delaney clause and call for its modification, such criticism should not overlook the reality that decisions leading to the implementation of the clause were never meant to be simplistic, legalistic, or non-scientific. Rather, decisions were intended to be based on the most convincing scientific evidence and the very best exercise of scientific judgment. The Commission could enhance federal food safety decisionmaking by, in addition to endorsing reform or modification of the Delaney clause, underscore the value of and need for sound science in the implementation of the Clause in the absence of such reform.

• It would be a mistake to modify the Delaney clause to permit consideration of quantitative risk. This would make matters worse. The Delaney clause is not appropriate in light of today’s knowledge and it should simply be repealed. In carcinogen risk assessment, as in risk assessments focused on noncancer endpoints, we should call for the best scientific judgment as to whether harm is likely to occur to people under realistic conditions of exposure.

The Delaney clause is quite complex. We strongly support the use of the best available science and believe that the Delaney clause illustrates what can happen when Congress legislates scientific judgments, however well intentioned, in a manner that cannot evolve with advances in scientific knowledge. We understand that "found" to "induce cancer" should be open to rigorous evaluation of scientific evidence concerning likely human carcinogenicity, but must point out that the courts have interpreted it otherwise. Minor additions and corrections were made to the text at the suggestion of FDA staff.

Department of Agriculture

• It appears that information about risk assessment and risk management in USDA came principally from ORACBA. Risk assessment activities are currently underway in many USDA agencies, including FSIS, the Animal and Plant Health Inspection Service, the Agricultural Research Service, and the Economic Research Service.

• Half of USDA’s 100,000+ employees are engaged in service with the Forest Service, Natural Resources Conservation Service, or Farm Services Agency. That means that we have a very large share of the resource conservation activities in the federal government. Because of that, some mention must be included in the Commission’s report.

This section was modified to include minor corrections and additional information submitted by USDA staff, including acknowledgment of risk assessment activities taking place in offices in addition to those at ORACBA. A recommendation was added to develop and implement methods for monitoring and evaluating benefits of USDA’s conservation practices. We are gratified that the Administration has instituted substantial funding for improvements in testing for microbial contamination, as we urged in 1996.

Department of Energy

This section was modified to include information about DOE’s work planning procedures for managing worker health and safety risks.

Department of Defense

The section was modified to include minor corrections and additions suggested by the Office of the Undersecretary of Defense.

• Go to previous section.
• Go to next section.
• Go to report's Table of Contents.
• Go to Commission's homepage.
• Go to RiskWorld's Departments Page.