Before the U.S. Senate Committee on Governmental Affairs
On September 12, 1997
NRDC has worked on behalf of its members, which number more than 350,000 people throughout the United States, to carry out goals that the American public strongly supports: environmental quality that is free from manmade threats to human health and that preserves and enhances the wonderful diversity of natural resources with which America has been blessed. Like the American public NRDC also strongly supports a healthy economy and the opportunity for all to prosper. We believe that Americans want their government to help achieve these goals in concert, not to pursue policies that set one goal against the others.
We urge members of the Committee not to support S. 981. We do so because we believe, despite its principle authors’ intent, that the bill would prevent government from acting to protect Americans against serious environmental, health, and safety threats and would jeopardize existing legal safeguards that protect us today against such harms. In the second part of my testimony I will highlight why NRDC believes the bill would have these damaging consequences. First, however, I wish to discuss the broader issue of whether it is wise to enact another law to constrain the agencies that carry out today’s health, safety, and environmental laws.
America's environmental quality is among the best in the world. We have made enormous progress in reducing health risks and restoring our air and water over the last quarter century, due in very large part to landmark environmental laws passed by Congress since 1970. The same can be said about the quality of our food supply and the working conditions in many of our businesses. Much ground has been gained but without continued effective action by government to insist on compliance with existing safeguards and the ability to adopt new requirements to address remaining hazards, the benefits we now have easily could be lost. That is why we believe Congress should refrain from adopting any new laws that will make it harder for health, safety and environmental agencies to do their jobs.
Real people live or die due to actions by the many firms whose behavior is shaped by today’s health, safety, and environmental laws. Drinking water can be assured pure or not depending on the adequacy of the regulatory system that protects that resource. Tainted meat gets to our kitchen tables or is stopped (better yet kept pure from the outset) if our rule-making agencies are given adequate tools to do their work. Ecosystems like the Great Lakes or the Chesapeake Bay thrive or decline according to how long it takes agencies like EPA to determine the causes of environmental degradation, identify solutions and then carry out an already extended process of adopting and enforcing rules to remedy existing abuses. Kids play in smog or clean air based on rule-making actions by government and responding actions by regulated interests.
Yet the cry of "excessive regulatory costs" has been raised with increasing intensity as polluters, distributors of unsafe products and operators of hazardous workplaces all have been required to change their behavior. These interests have pressed Congress for new laws designed to force greater reliance on cost-benefit analysis and risk assessment at health, safety, and environmental agencies.
Of course, health, safety, and environmental rules have changed behavior in the regulated community. If they had not, we would suffer from far worse pollution, health, and worker safety threats than are with us today. Cleaning up pollution, protecting health, and making workplaces safer has shifted costs from the people who suffered such harms to the enterprises that inflicted them. But where does the truth lie with the claim that those costs are excessive and what are the appropriate steps to take to assure agencies are properly considering the impacts of their rule-making actions?
In the last Congress there was almost no inquiry into these complex factual issues before members introduced their legislative "solutions" to "the problem." It was taken as a given that regulatory costs are excessive and that the obvious remedy was to enact new laws forcing regulators to go through additional rounds of analysis and review before new rules can be adopted. Further, it was assumed that if a new rule could not pass new and rigorous economic tests the rule should be prohibited or at least treated as an example of bad policy.
Today we are once again considering regulatory reform legislation. Its lead authors have solid track records as thoughtful supporters of effective and responsible government. Yet you still do not have the benefit of a thorough and objective analysis of the nature of the alleged regulatory cost problem. Nor has the Congress adequately explored whether there are remedies more appropriate than requiring additional layers of analysis and review before more protective rules can be issued. Additional consideration of these issues could help Congress craft better legislation or decide whether the laws you have passed already are adequate to address the subject of better rule-making.
Consider S.981; it requires substantial economic analyses to accompany major rules but defines "rules" so that most significant economic regulation is exempt from the bill’s additional analytical burdens. This means that the bill applies primarily to health, safety, and environmental rules. Is the decision to ignore economic regulation and zero in on health, safety, and the environment the product of a considered judgment that economic rules are as a group economically efficient, while the rules that protect us are economic losers? This seems an unlikely explanation. According to the Office of Management and Budget’s recent Draft Report to Congress on the Costs and Benefits of Regulation, economic regulations are estimated as a group to impose costs billions of dollars greater than their estimated benefits, while health, safety, and environmental rules are estimated to produced benefits billions of dollars greater than the costs they impose. 62 Fed. Reg. 39351, 39364, July 22, 1997. If the purpose of S.981 is to produce a reduction in the costs of regulation without reducing the benefits they provide, it would seem the bill is aimed at the wrong target.
There is a broader issue presented by S.981’s focus on health, safety, and environmental rules. By requiring such rules to judged by whether they maximize economic efficiency, the bill departs from the fundamental reasons these laws were enacted. Health, safety, and environmental laws are a reflection of the fact that Americans define their lives, their hopes, their expectations in terms that are not governed exclusively by an economic calculus. Of course, economic considerations play a major part in each person’s life but we pursue other aims as well. In philosopher Mark Sagoff’s terms, yes we are consumers but we are also citizens.1 When we think and act as citizens we promote behavior that is compatible with our notions of a civil society, one that reflects shared values shaped by families, friends, neighbors, community.
Because we are no longer just a village where all actors can be induced to adhere to community values through peer pressure, citizens insist that our government be given the authority and tools to promote behavior that is desirable but is not a natural outcome of economic transactions. Hunger for increased economic efficiency did not move the public to ask the Congress to pass laws to prevent air and water pollution, to protect worker health, to ensure a safe food supply. These laws are a product of the public’s view that the government is an essential instrument for delivering the benefits of clean air, clean water, safe food and safe workplaces—things we want because they comport with our sense of how society should function, not because of the hypothetical prices that academic theorists attempt to attach to these benefits.
Does the motivation for these laws mean that Congress should delegate to agencies the irrevocable power to do whatever they want to pursue health, safety, and environmental goals? Of course not, but it does mean that laws designed to improve rule-making in this area should not be predicated on maximizing economic efficiency. It also means new laws that increase the difficulty of issuing new safeguards should take great care that they do not impair the government’s ability to provide the public with the benefits that it expects from the laws that pledge a clean environment, safe food and decent working conditions.
Turning to the specifics of S. 981, while it may be well-intentioned, we believe it would harm programs to protect health, safety, and the environment and for that reason we oppose it. In this testimony I will focus on these critical impacts of S. 981 on the government’s ability to protect the public:
- weakening of health, safety and environmental protections due to new cost-based criteria for acceptable rules that would restrict current authority through the courts and the political process;
- lengthy delays in issuing rules due to analytical demands and review procedures that would sharply increase the time, difficulty, and costs of developing new safeguards;
- major reductions in resources available to enforce safeguards and respond to new threats due to requirements to review existing rules through processes that could dominate agency workloads and jeopardize retention of important safeguards;
- understatement of risks to the public due to imbalanced risk assessment requirements;
- domination of agency decision-making processes by firms with a direct interest in the content of new and existing rules.
Each of these impacts considered separately poses a risk of unwarranted damage to agencies’ ability to carry out protective laws; taken together they would have a cumulative impact that has the potential for lasting damage to these important programs. I firmly believe that you do not intend these results and you are therefore skeptical that the bill could have these effects. Accordingly, I would like to discuss these areas in greater detail.
Overriding Existing Authority to Protect the Public
The sponsors of S.981 have clearly stated that their intent is not to override existing health, safety, and environmental laws with this bill. We agree wholeheartedly that this is the correct policy. Unfortunately, the text of the bill leaves enormous latitude for regulated interests to argue that the bill, if enacted, does override and limit pre-existing authority to adopt protective rules. Beyond the legal jeopardy to protective rules, the bill establishes a political standard of acceptability for public protections that we believe would drive agencies to adopt less protective rules.
The bill contains several interlocking provisions that will provide new grounds for overturning health, safety, and environmental rules. First, there are the "cost benefit" determinations required by section 623(c)(3). Second, there is the absence of a savings clause that would expressly preserve existing authority to issue protective rules. Third, there is the judicial review provision of section 627, requiring courts to hear claims that that rules should be overturned on various grounds created by S.981’s requirements, including flaws in the analyses required by the bill. Fourth, there is the provision for judicial review of existing rules in section 633(l) that would allow any previously legal, existing safeguard to be repealed simply because it was not the best possible rule. Finally, there are the mandatory regulatory analyses, risk assessment, and peer review procedures discussed in detail in the next section of this testimony.
The Bill’s Mandated Cost Benefit Tests Would Prevent Protective Rules
We have two reasons for opposing the cost-benefit tests established by the bill. First, the tests are biased against the adoption of more protective rules. Second, the bill as written may well be construed to override most existing protective laws as a legal matter, making it impossible for agencies to uphold protective safeguards in court.
Section 623(c)(3)(A) of the bill establishes three cost-benefit tests (the second pair framed as an alternative) : do the rule’s likely benefits justify its likely costs; and is the rule more cost-effective or does it result in greater net benefits than other regulatory alternatives considered by the agency. While an agency that does a competent job of articulating the benefits of a chosen rule is likely to be able to satisfy the first test (benefits justify costs), the second set of tests require an unworkable quantification and monetization of benefits that will operate against selection of more protective rules.
Let us consider the tests in turn. The "more cost effective" test will normally be passed only by the least protective regulatory option because the typical pattern for rules that abate an existing harmful exposure is that the cost-effectiveness ratio (dollars of costs per unit of benefits) goes up as the level of protection increases. If an agency were required to select the option that is "more cost-effective" than any other legally permitted alternative, it would be barred from choosing an option that provided far greater protection simply because of a slight increase in the cost-effectiveness ratio.
For example, to limit a contaminant in drinking water EPA might identify the following options: level 1 (the least protective option) would prevent 1 million cases of illness at $20 per case; level 2 prevents 10 million cases of illness at $22 per case; level 3 prevents 10.1 million cases of illness at $1,000 per case. In this example only level 1 would pass the "more cost-effective" test in S.981. If this test governed, EPA would not be able to select level 2 even though it prevented ten times as many illnesses with only a slightly higher cost per case avoided than the least protective option.
Since the more "cost-effective" test generally would force selection of the least protective option, to adopt a more protective rule under the bill’s tests an agency would have to demonstrate that its preferred option provided "greater net benefits" than any other reasonable alternative. While the bill elsewhere defines costs and benefits to include non-quantifiable elements, this sensible proviso is undone by the "greater net benefits" test in §623. The net benefits test forces precise quantification and monetization of benefits because regulated interests will argue an agency cannot show that a more protective rule has greater net benefits than another option except by expressing the benefits and costs of both alternatives in common units so that the differences in costs and benefits (net benefits) between the options can be compared. The common units that regulatory analysts typically insist on is dollars. As discussed below, the difficulty in quantifying and monetizing benefits could result in a rule being overturned because of a court’s view that an agency did not have adequate support for a conclusion that the rule selected had greater net benefits than other less protective options.
Even if monetization were easily done, there is another objection to the bill’s greater net benefits test: as formulated the test could require rejection of a much more protective option which had benefits exceeding costs, simply because the option did not maximize net benefits compared to other regulatory alternatives. These determinations could turn on small differences in assumed monetized values for important benefit categories, such as the value of a life saved.
Consider a hypothetical EPA rule to limit the amount of a cancer-causing additive to gasoline. Alternative A would require only modest reductions in the additive and would save 100 lives per year at a predicted annual cost of $100 million. Alternative B would more sharply restrict levels of the additive and would save 1000 lives per year at a predicted cost $2.4 billion per year. (See Figure 1). Alternative A is more cost-effective ($1 million per life saved compared to $2.4 million per life saved under Alternative B) and would be chosen under the section 623 tests unless the agency can show Alternative B has "greater net benefits."
While Alternative B saves ten times as many lives, that alone does not prove it has greater net benefits. To pass this test the agency must assign an assumed value to each life saved and then calculate the difference between costs and benefits under each alternative. Unfortunately, this can result in a huge difference in benefits being driven by a small change in the assumed value of a life. For example, if EPA assumed a life saved was "worth" $2.6 million, the more protective Alternative B would have greater net benefits than Alternative A ($200 million versus $160 million) and thus could be adopted. 2
However, if EPA assumed only a slightly smaller value per life saved, $2.5 million, the outcome would be the opposite. At the lower assumed value of life, saving 100 lives would have greater net benefits than saving 1000 lives ($150 million versus $100 million) and the more protective option would not meet the section’s cost-benefit test. 3
Note that both alternatives in the example above are calculated to have benefits greater than their costs but due a difference in the assumed value of a human life as small as $100 thousand, the alternative that saves ten times as many lives would flunk the §623 tests. We believe that a test which could produce this result is not a sound test for evaluating health, safety, and environmental rules and hope you will agree.
Below I discuss the risks of judicial rejection of protective rules due to the §623 tests but there is another threat to such rules – the political threat. If §623 were enacted, its cost benefit tests would become the presumptive criteria for evaluating whether rules "made sense" on policy grounds. The likely result would be that agencies would avoid adopting rules that could not clearly pass these tests, either out of concern for congressional disapproval of the rule or because the agency would want to avoid the stigma of adopting a rule that did not "make sense" judged by the lights of the bill’s cost benefit tests. Accordingly, poorly designed tests could do great harm, even if they were not judicially enforceable, by pushing agencies to adopt less protective rules.
It is important to point out that the second pair of cost benefit tests in §623 also conflicts with the stated objectives of the bill. Section 2 of S.981 includes the following findings:
(2) Cost-benefit analysis and risk assessment are useful tools to better inform agencies in developing regulations, although they do not replace the need for good judgment and consideration of values.
(3) Cost and risk need to be considered in evaluating regulatory proposals which address health, safety, or the environment. Other factors such as social values, distributional effects, and equity, must also be considered.
(emphasis added).
Yet the second pair of cost tests in §623 rule out, as a practical matter, the agency’s ability to consider factors other than its quantifiable estimates of the rule’s benefits and costs.
Any call by Congress for agencies to compare the benefits and the costs of their actions must recognize and prevent the potential for abuse of such comparisons. We cannot shut our eyes to the fact that powerful interests will use cost benefit comparisons as a weapon to attack rules that cause them to change their behavior. Members of Congress need to lead in criticizing those who would seize on numerical estimates of costs and benefits, while ignoring broader factors, to buttress their claims that a health, safety, or environmental rule is unjustified. It is one thing to say that unquantifiable benefits and equitable issues must be considered in sound decision-making. To make this principle real, agencies who attempt to follow this guidance must be defended when they are attacked by those who trumpet numerical estimates and ignore the additional considerations that accompany the estimates.
It is worth mentioning a real-world example of the misuse of cost benefit comparisons in the policy debate. As you know, EPA recently adopted revisions to the National Ambient Air Quality Standards for Ozone and Particulate Matter to improve public health protection against the effects of these pervasive air pollutants. As required by executive order, EPA prepared a Regulatory Impact Analysis (RIA) to describe the costs and benefits of the final rules. Not surprisingly, some of the interests who attacked the standards are now distributing broadsides claiming the RIA proves the rules will have more costs than benefits. For example, C. Boyden Gray’s group, the Citizens for a Sound Economy Foundation, recently released a charge that "based on the new RIA, the new standards may no longer be worth the expense." Their charge concluded, "The EPA has now weighed the costs and the benefits, and found that the new standards may actually harm the public, potentially producing net negative benefits of $26 billion." (emphasis in original)4 While many may read Mr. Gray’s two-page release, few will read EPA’s several hundred page RIA and that is a shame because it may allow this misuse of EPA’s analysis to continue.
What are the facts? EPA faced a significant challenge in estimating costs to meet the revised standards because it needed to forecast the amount of pollution reductions needed to achieve the standards in many different communities and then to forecast what methods might be employed to achieve those reductions as far in the future as the year 2010. The RIA states that improved and less costly methods of pollution reduction almost certainly will be developed over this period and points to substantial evidence of past examples of dramatic breakthroughs in technology prompted by tighter standards. However, rather than attempting to assume the degree of cost reductions that would occur, the agency used for purposes of the RIA a conservative assumption of costs several times higher than current costs to achieve the reductions needed for full attainment of the standards. The RIA makes it clear the agency does not believe anything like these assumed costs will actually occur and provides evidence to support its view.
Regarding the RIA benefits estimates, the agency pointed out that it was able to assign dollar value estimates to only a fraction of the health and environmental benefits of the revised standards and that the monetized benefits substantially understated the actual benefits of the standards. However, by ignoring this information and by highlighting the low end of EPA’s partial benefit estimates and the high end of cost estimates the agency shows are almost certain not to occur, Mr. Gray’s group has confected a claim that the RIA proves the revised standards are bad policy. This is the type of abuse Congress can expect to proliferate, poisoning debate and perhaps the public, unless you are diligent in insisting on an honest discourse regarding the use of benefit and cost considerations.
More Protective Rules Would be Subject to Judicial Rejection
As discussed above, § 623(c)(3)(A) of the bill establishes three cost benefit tests: do the rule’s likely benefits justify its likely costs; and is the rule more cost-effective or have greater net benefits than other reasonable alternatives? While §623(c)(3)(A) of the bill is framed as simply requiring a statement "whether" the agency’s selected rule passes these specific cost-benefit tests, the next subparagraph (B) contains language that courts are likely to construe as sharply limiting an agency’s authority to adopt rules that do not pass the bill’s cost-benefit tests.
Subparagraph (c)(3)(B) imposes several requirements on an agency that selects a regulatory alternative that the agency "cannot reasonably determine" passes the subsection’s cost-benefit tests. First, the agency must "explain why such determinations cannot be made." (emphasis added). Second, the agency must "identify any statutory provision or other factor that prevents such determinations." (emphasis added). Third, the agency must describe a regulatory alternative that would pass the required cost-benefit tests.
While the bill’s authors describe this section as allowing an agency to select any currently lawful regulatory alternative, the bill’s language suggests that more than a policy choice would be required to sustain a rule that does not pass the tests. Parties challenging a rule will argue that the words "cannot be made" and "factor that prevents such determinations" mean that an agency must identify an impediment or barrier that makes it not possible for the agency to select a rule that passes the cost-benefit tests.
The argument of an industry lawyer seeking to overturn a protective rule can be concisely summarized as follows:
Subparagraph 623(c)(3)(A) requires the agency to determine whether the rule meets the cost benefit tests in that subparagraph. Subparagraph(B) sets forth the conditions under which the court can sustain a rule if the agency selects an alternative that does not pass the tests in (A). To sustain such a rule the agency must explain why it "cannot" determine that the cost benefit tests are met and "identify any statutory or provision or other factor that prevents such determinations." These requirements indicate Congress intended to allow agencies to adopt rules that could not pass the cost benefit tests only when it was impossible to adopt a complying rule; for example, when an existing statute required a specific regulatory outcome and that outcome could not pass the cost benefit tests.
Under this interpretation of §623, a statute that expressly specifies an outcome that does not include considerations of costs would likely suffice to justify a rule that does not pass cost-benefit tests. However, most statutes do not fit this pattern. The typical health, safety, or environmental law grants an agency authority to adopt rules as needed to accomplish a broadly stated purpose, such as protection of health, safety, or the environment.
For example, §211(c) of the Clean Air Act authorizes EPA to control or prohibit a motor fuel additive if the Administrator determines that the additive causes or contributes to air pollution that may "endanger the public health or welfare." EPA used this authority to require lead additives to be phased out of gasoline for health purposes. Under S. 981 EPA would be required to determine whether its selected lead phase-out rule passed the §623 cost-benefit tests. Under §623(c)(3)(B) of the bill, if EPA selected an alternative that was not more cost-effective or did not have greater net benefits than other alternatives, EPA would have to explain why it "cannot" make the determinations that the selected option passed the section’s cost benefit tests and it would have to identify a statutory or other factor that "prevents" it from making the favored determinations. Under a statute like CAA §211 that gives EPA substantial discretion, there is no statutory mandate preventing the agency from choosing a less protective option that would clearly pass the section’s cost-benefit tests. So the question would be on what grounds could EPA assert that it "cannot" choose an option that would pass those tests.
While the agency could attempt to argue that uncertainties in quantifying and monetizing benefits counseled adoption of a more protective rule, S. 981 casts doubt on the success of this rationale. First, the language "cannot make" and "prevent" could support an interpretation that the agency has a high burden to show that the uncertainties are so great that it is effectively impossible to tell on which side of the cost benefit line its selected option falls. This route is not without peril, however. Such a claim invites a response that it is arbitrary and capricious to select a more expensive option with "uncertain benefits" when the agency has identified other permissible options which clearly pass the cost benefit tests. As discussed below, the bill’s judicial review provisions mandate courts to consider the §623 determination in judging whether the rule is arbitrary and capricious.
Thus, there is a significant potential that under §623 as written, existing statutory authority to adopt precautionary health, safety, and environmental safeguards would be sharply curtailed by judicial decisions allowing agencies to adopt rules only if they pass the section’s cost benefit tests except in the rare instance where consideration of costs was clearly prohibited by the underlying statute.
Moreover, attempts by agencies to sustain rules by relying on significant dollar estimates of the value of "non-market" benefits, such as lives saved, also will be closely scrutinized in the courts. Under the bill, the agency’s selection of assumptions is subject to judicial review and the regulated community has the means and the incentive to create huge records attacking the agency’s assumed values for lives saved and other benefits as too high. This presents the real prospect of judge-made law forcing agencies to use unreasonably low values for such benefits, which in turn would make it impossible to select more protective rules.
New Grounds to Attack Rules in Courts
The judicial review provisions of S.981 will be seized upon by industry lawyers to mount attacks on rules on grounds that the analytical, procedural, and determination requirements of the bill have not been satisfied. Section 627 of S.981 requires courts to consider these claims when reviewing a challenge that a rule is arbitrary and capricious. The only exception to this mandate for review is when the agency can prove the analysis or assessment "would not be material to the outcome of the rule." This exception will not protect sound rules from being overturned. When faced with a claim of a technical flaw in an analysis, the only way an agency could meet its nonmateriality burden would be to redo the analysis without the flaw and show it made no difference. But how is the agency to conduct this reanalysis unless the court remands the rule to the agency for further consideration? Thus, the rule must be overturned to make use of the exception intended to protect the rule.
Another method of proving nonmateriality would be to show the authorizing statute prohibited consideration of the challenged analyses in making decisions. But, as discussed above, such statutes are the exception, not the rule in the health, safety, and environmental area.
Section 627 is likely to be interpreted to require overturning of rules based on a court’s review of the adequacy of the bill’s required analyses because of the absence of language stating that such review is precluded. In S.1001, a regulatory reform bill sponsored by Senators Glenn, Levin and many others in the 104th Congress, express language was included to preclude review of the contents of the analyses.5 In addition, S.1001 contained a critical savings provision, making it clear that the bill did not alter the criteria for adopting rules under preexisting law.6 Neither of these crucial clarifying provisions is contained in S.981.
As a result of these provisions we believe S.981 as written would prevent agencies from adopting more protective health, safety, and environmental rules, undermining existing legal authority both as a legal and a practical matter.
Existing rules also would be threatened with judicial repeal under the bill. Section 633(l) of the bill allows courts to overturn existing rules that have been reviewed if the challenger shows the agency "could have adopted a reasonable alternative that would substantially increase net benefits" and meet statutory objectives. This remarkable provision raises the prospect that settled rules that are legal and working well, could be overturned years after adoption simply because a court is persuaded by a litigant that the agency could have adopted a "better" rule, using the criterion of net benefits. As I have argued above, the net benefits test is basically flawed and should not be used to judge new rules. It would be even more misguided to allow the test to be used to overturn sound existing rules.
Burdens of Increased Analysis and Review Procedures
S.981 would create detailed statutory analysis and review procedures in three overlapping areas: specific mandates for analysis of costs and benefits of rules and alternatives; risk assessment requirements; peer review procedures. Taken together these requirements will present agencies with a much more difficult task in adopting new safeguards.
Section 623 of the bill requires an analysis of benefits and costs that cannot be quantified. While well motivated, the requirement to analyze unquantifiable costs and benefits is not likely to be easy to carry out. Further the agency must analyze the costs and benefits of a range of regulatory alternatives that reflect the range of options that would achieve the statutory objectives addressed by the rule. It is not clear how many alternatives will be required to reflect the full range of options relevant to the statutory objectives but we can anticipate vigorous claims by regulated interests that agencies are not analyzing enough alternatives.
In addition, section 623 requires analysis of alternatives that require no government action and alternatives that accommodate differences among regions and among persons with different resources. The practical implications of these added analytical requirements are substantial. In an ideal world perhaps it would be appropriate to ask agencies to custom tailor rules to fit perfectly the characteristics and preferences of each regulated person. But in the real world this is a prescription either for rules designed for the lowest common denominator or a rule-making process so weighed down by considerations of source-specific differences that the agency is unable to complete its work. In the real world we know that the price of a custom-cut and fitted suit from a bespoke tailor or the price of a hand-built automobile is so high that few are made and few can afford them. A bill that would allow lawyers to force agency rules to emulate this model is not a recipe for prompt action or reducing the cost of government.
The agency’s estimates of a rule’s benefits must be conducted by means of a risk assessment process prescribed by section 624 of the bill for health, safety, and environmental rules. In general the bill requires two rounds of risk assessment; one for the proposed rule and one for the final rule unless the agency wants to take the legal risk of arguing that a new assessment at the time of final rule-making is not needed to respond to comments on the proposal. (§624(a)(1) and (a)(2)). Whenever a risk assessment involves the use of assumptions (nearly always the case), the agency in effect must carry out multiple risk assessments – one for each reasonable alternative assumption or combination of assumptions. (§624(c)(2)).
A regulated entity seeking to delay adoption of a rule it opposes, be it tobacco restrictions, meat safety or some other reform, need only propound alternative assumptions to those identified by the agency. This will put the agency to the difficult and risky choice of attempting to prove the alternatives are "unreasonable" and therefore not required to be analyzed, or to extend and expand the rule-making process by conducting alternative risk assessments that incorporate the alternative assumptions.
The agency must provide opportunity for public comment and participation "during the development" of a risk assessment. (§624(d)). Since a proposed rule must be accompanied by a completed risk assessment, the language of §624(d) suggests that the agency must provide for public comment prior to the publication of the proposed rule (during the risk assessment’s "development") and again after the publication of the proposal. To these process requirements the bill adds a mandate for peer review of risk assessments with a directive that the peer review written comments and the agency’s written responses be available to the public. (§625(b)(2)). The bill is not clear when these peer review materials must be made available but we can expect lawyers to argue they must be made available prior to each public comment period to permit informed comment. Of course, public comment opportunities are not something to discourage but the effect of multiple comment opportunities interlocked with independent peer review comment and response processes would be to make many agency decision-making processes impossible to complete in any reasonable time-frame.
The agency’s risk assessments also must examine ranges and probability distributions of risk and of corresponding exposure scenarios ((§624(f)); they must do analyses comparing the risks to other risks encountered by the general public (§624(g)); and they must address substitution risks when information is reasonably available (§624(h)). Thus, if the tobacco industry submits claims alleging the risk of a black market emerging in response to a proposed tobacco rule, the FDA must address the risk thoroughly or risk having its rule thrown out in court. The cumulative impact of all these risk assessment analytical requirements would be to add indefinite delay to the issuance of major new rules.
As mentioned above, the bill also requires risk assessments and cost benefit analyses to be peer reviewed. Later in this testimony I address the problem of conflicts of interests during peer review; here I wish to note that it adds to the large analytical process burden described above. When does the peer review of risk assessments and cost benefit analyses occur: during their development; when they are in draft just prior to proposal; after proposal; when they are in draft final form just prior to final rule-making; or at all four stages?
If the Committee goes forward with legislation we urge reconsideration of the basic decision in S. 981 to require peer review of the analytical components of each major rule. While it is important that agencies have reliable quality assurance procedures for their analyses, it is overkill to demand peer review for each covered rule. Congress doubtless would reject as unreasonable a bill requiring auto manufacturers to road-test each vehicle coming off the assembly line before sale; why impose a similarly burdensome requirement on health, safety, and environmental rules?
HARM CAUSED BY THE BILL’S PROCEDURES TO REVIEW RULES
Section 633 of S.981 requires each agency to schedule review of existing rules based on the recommendations of an appointed advisory committee. The authors of the bill have indicated that because the agency is not mandated to accept all the advisory committee’s recommendations and because the selection of rules is not subject to judicial review, this review requirement should have no negative impact on agency’s ability to perform their work. Despite the authors’ intent, we believe this provision would have substantial harmful consequences. The fundamental problem with this provision is that the makeup and authority of the advisory committee assures that agencies will be forced as a practical matter to review far more rules than warrant review, with the result that agency resources will be diverted from protecting the public to attempting to meet the statutory review schedules.
The bill requires the advisory committees to include interests "affected by" the agency’s rules. (§632(b)). Even without this mandate, it is a political certainty that such committees will be awash with regulated interests. Every large firm that has a significant disagreement with an existing rule will seek membership on these committees and your offices can expect many requests for assistance in landing them an appointment. While the bill requires the advisory committees to be balanced and to include "public interest groups," assuring balanced participation in fact cannot be achieved by statutory fiat. To achieve real balance, the bill would have to address the huge disparities in resources available to different interests. Regulatory experts are a scarce commodity in the public interest community and most groups would find it impossible to fill in behind a staff member who was tapped to serve on an advisory committee, even if the group had the resources to hire added staff.
Once constituted, the committees can operate with essentially no constraints. The bill does not require the committee to explain its choice of rules recommended for review; it imposes no limits on the number of rules that can be recommended; it does not require a finding by the committee that the agency’s resources are adequate to review the recommended rules while continuing to meet the agency’s mission to protect the public; it does not require the committee to propose a list before making final recommendations, so that the public could comment on the committee’s list. Nothing in the bill or in the dynamics of the committees will operate to limit the number of rules they recommend to the agencies. Each regulated interest on the committee will have an incentive to engage in log-rolling with its industry colleagues: "I’ll vote to review your rules if you’ll vote for mine."
The predictable result is that the agency will receive a list of recommendations far longer than the agency can possibly tackle without completely abandoning its responsibility to protect the public. The bill allows the agency to cite resource constraints as a reason for not reviewing a rule but agencies will quickly learn there is a high political price to be paid for relying on this provision. In Congress and on many editorial pages the advisory committees will be given the mantle of independent fresh-thinking advisors determined to sweep away the dead wood of the regulatory state. Agency heads that receive a list of 100 rules for review and respond they have resources only to review 20 in a five-year period can expect to be pilloried for falling back on the last stale cry of the bureaucratic hack, "we don’t have the money." As a consequence, agency heads who wish to remain in good graces with Congress and others will be tempted to accept far more rules for review than their staffs tell them can be reviewed on time. Some may even reflect that they are likely to be gone by the time that it is apparent the agency cannot meet the review schedule.
The next stage in this scenario of overload is that the combination of congressional pressure and court orders will force agencies complete the excessive number of reviews they agreed to and resources will be stripped away from public protection programs to make these reviews possible.
If large numbers of rules are proposed for weakening changes or repeal after review, Congress cannot depend on the public comment period as an adequate safeguard against bad policy choices. As noted above, the resources of the regulated community are far greater than those environmental, health, and safety groups. While critics of safeguards will be able to fill the record with submissions urging repeal or relaxation, advocates for public protection will be overwhelmed by any significant increase in actions to drop or change rules. We do not argue that all rules should be preserved forever. Rather, we ask that Congress consider the ability of different interests to participate when you design proposals to review those rules.
RISK ASSESSMENT BIAS
Section 624 of the bill establishes principles for risk assessment. Earlier I described the procedural burden of the multiple assessments required by this section. Another major issue is the direction in which the bill’s principles would push risk estimates. We believe that some of these principles would inappropriately characterize risks as smaller and less important than they actually are. In particular, the requirement to describe the "most plausible estimates" of risks may result in an improper assessment. (§624(e)(2). Recall the collapse of a hotel walkway in Kansas City some years back during a dance. Is the "most plausible" estimate of the risk of a collapse one based on normal usage of the walkway? Do we want an agency to design construction standards based on such an estimate? How would the "most plausible" directive apply in the field of nuclear safety? The accidents at the Fermi plant, at Three-Mile Island, and at Chernobyl certainly were not the most plausible.
Another troubling requirement calls for agencies to compare the risks they are addressing to other "reasonably comparable" risks routinely encountered by the general public. (§624(g)). This requirement has great mischief potential and is likely to be used to argue against correcting avoidable risks simply because they appear smaller than other risks society has a difficult time reducing. For example, analysts often compare the risk of harm from an environmental hazard to the risk of being killed or maimed in an automobile accident. It is a strange position that uses other risks as an argument against reducing the risk an agency is authorized to address. Carnage on the highways is not an index of what is good policy; it is an example of policy and practical failure.
Another ignored factor in risk comparisons is that often a large component of the reference risk is caused by unlawful behavior (for example, speeding and drunk driving in the highway case). Yet the comparisons are used to argue that current risk-creating behavior (exposing persons to pollution, for example) should be allowed to remain legal (by not adopting the rule). This argument confuses the issue of allocating resources to enforce compliance, with the issue of deciding what the law should be. For example, the risk of death from homicide is probably substantially smaller in Pinewood, Tennessee than in Chicago. This comparison may justify allocating fewer policing resources to Pinewood but it is not an argument for making homicide legal in Pinewood. Yet when interests use comparative risks to argue against adoption of a new rule prohibiting certain harmful behavior, they are arguing that the smaller risk is a rational basis for making the behavior legal.
DOMINATION OF AGENCY PROCESSES BY CONFLICTED INTERESTS
As mentioned above, the bill requires the advisory committees on review of rules to include interests directly affected by the rules they will discuss. In addition, the bill’s peer review provisions allow a person to be member of a peer review panel even if the person has a financial interest in the outcome of the review, requiring only that the interest be disclosed. (§625(b)(1)(B)). In contrast, a person who has any connection with the agency program is absolutely barred from participating in the peer review. (§625(b)(1)(A)). It is difficult to understand this disparate treatment of persons who might fairly be assumed to have difficulty in approaching a review with full objectivity. The bill would appear to allow an industry representative to attack an agency expert’s research findings, then participate in a peer review of that expert’s work while the expert herself was excluded from the review panel. When one considers that the purpose of the peer review requirement is to increase confidence in the quality of the work relied on by agency decisionmakers, allowing affected industry representatives to "peer review" work that may impose millions of dollars in compliance costs on their employers seems the wrong way to go. How would public confidence in the food supply be enhanced by having research on E. coli contamination be peer reviewed by a panel that includes representatives of meat packers?
Apart from being destructive of confidence, allowing peer review panels to be populated by regulated interests simply increases the likelihood that the peer review process will simply be another tool in the regulated community’s arsenal to use against more protective safeguards. Simply by raising issues that take time and money to respond to, a regulated interest member of a peer review panel can delay the issuance of rules for months or years. It is simply inappropriate to design a process where this can happen.
Can it be that outside experts are in such short supply that regulated interests must be allowed on the panels to make rule-by-rule peer review practical? I have argued above that peer review should be a more selective process. This may be another reason for not requiring such rule-by-rule review.
CONCLUSION
Mr. Chairman, members of the Committee, we believe that before adopting new laws that require additional analyses for health, safety, and environmental rules that Congress should examine objectively how agencies are doing under the analysis requirements of the laws already enacted. As you know, a number of new analytical requirements were enacted in the 103rd and 104th Congresses. Doesn’t it make sense to study how these laws are helping to address the concerns that motivate S.981 before imposing an added set of requirements? We honestly believe S.981 would do significant harm and little good and that is why we oppose the bill even though we count among its authors and sponsors Senators whom we regard as friends of strong health, safety, and environmental protections. Thank you for this opportunity to testify. I’ll be happy to respond to your questions.
------------------------------
1 Sagoff, Mark, "The Economy of the Earth, at 51, Cambridge University Press, 1988.
2 With an assumed value per life saved of $2.6 million, the calculated benefits of Alternative A would be $2.6 million times 100 lives saved or $260 million; the net benefits of A would be $260 million -$100 million, or $160 million in net benefits. Under this assumption, Alternative B would have calculated benefits of $2.6 million times 1000 lives saved or $2.6 billion; the net benefits of B would be $2.6 billion-$2.4 billion, or $200 million in net benefits.
3 At an assumed $2.5 million value per life saved, Alternative A would have $250 million of benefits and $150 million in net benefits. Alternative B would have $2.5 billion in benefits but "only" $100 million in net benefits and thus would flunk the greater net benefits test.
4 "Capitol Commnet," August 26, 1997, Citizens for a Sound Economy Foundation.
5 Section 623(d) of S.1001 provided in part, "If an analysis or assessment has been performed, the court shall not review to determine whether the analysis or assessment conformed to the particular requirements of this chapter."
6 Section 622(h) of S.1001 provided as follows: "The requirements of this subchapter shall not alter the criteria for rulemaking otherwise applicable under other statutes."
Related Links
RiskWorld news article "Risk Science and Law Panel Weighs Regulatory Reform" published September 16, 1997
Full text of Paul Portney's testimony before the U.S. Senate Committee on Governmental Affairs hearing on the Regulatory Improvement Act of 1997 (bill S.981)
RiskWorld news article "Regulatory Reform Bill Requires Risk Assessment, Cost-Benefit Analysis" published June 27, 1997
Summary of the Levin-Thompson Regulatory Improvement Act
Statement of U.S. Senator Fred Thompson, R-Tenn., on June 27, 1997
Joint Press Release from U.S. Senator Carl Levin, D-Mich., and U.S. Senator Fred Thompson, R-Tenn.
Posted September 16, 1997.
Go to:
Copyright © 1997 by Tec-Com Inc.