TSCA Implementation is Risky Business
With the recently amended Toxic Substances Control Act (TSCA), Congress left many of the most critical policy decisions to EPA. At stake are tens if not hundreds of millions of dollars in compliance costs and potentially more valuable improvements to human health. Depending on how the Agency decides to implement the new law, regulatory outcomes could be fast, cost-effective, and clear or they could be slow, costly, and confusing.
Two weeks ago, the public comment period ended on two issues of critical import: (1) prioritization of chemicals and (2) risk evaluation for high priority substances. Both involve the use of risk analysis to evaluate chemicals against the new safety standard.
Many good ideas can be found in the public comments. But along with the good comes the bad. Here are the five worst ideas from those public comments—ones that, if adopted, would impose the greatest opportunity cost—and why the Agency should resist them.
The mere presence of a “safer” alternative equates to unreasonable risk, and subsequent regulation should restrict consumer choice. It is not reasonable to continue to expose humans or the environment to an avoidable hazard.
This recommendation flies against the history of environmental law, where the magnitude of the risk distinguishes acceptable risk from unacceptable risk. If EPA adopted this recommendation, any risk analysis showing one chemical to be less risky than another—even if both chemicals pose relatively low risk—would presumably lead to regulation banning the higher risk substance. In theory, no two chemicals pose the exact same risk, and very few substances in commerce are indispensible. Therefore, EPA would be forced to restrict choice to one substance for any particular use, even if risk reduction were minimal, and even if the regulation led to inferior products and higher prices for consumers. The result would be an explosion of regulation, much of which would not pass a cost-benefit test.
Determination of “unreasonable risk” can be based on qualitative findings (i.e., hazard data) and for which there is also potential for exposure. EPA guidance should reflect that risk, by definition, is the possibility of harm, not the certainty of harm. Numerical estimates of risk should be addressed in the risk management phase after a chemical is deemed to be of unreasonable risk.
This recommendation ignores the fact that Congress, in amending TSCA, requires EPA to employ the risk paradigm to chemicals in commerce. The risk paradigm is based on the premise that it is hazard coupled with exposure that determines risk, that all chemicals pose some degree of risk, and it is only when the risk is unacceptable (or in the case of TSCA, unreasonable) that regulation warranted.
If EPA adopted this as policy, it would effectively regulate on the basis of hazard, even if the risk were de minimis (i.e., negligible health benefits). It is an axiom of good public policy that regulation ought to be cost-effective (an axiom explicitly recognized in TSCA), so this recommendation fails when measured against this well-accepted criterion.
Default assumptions used in a risk assessment should be health protective in all cases and under all conditions. EPA should determine the maximum total risk for the most exposed or most vulnerable people from potential future maximum use.
One of the age-old arguments in risk assessment is how to address scientific unknowns, such as how to extrapolate toxicological results in animals exposed to large doses of a chemical to humans that are exposed to smaller doses of the chemical. A typical risk assessment employs dozens of assumptions to address gaps in scientific understanding or lack of empirical data. By employing worst-case (also known as conservative, or health-protective) assumptions, the analyst is ensuring a much higher risk estimate than would otherwise be the case.
EPA has long acknowledged that it employs “plausible conservatism” in risk assessment. This policy recommendation would be the equivalent of plausible conservatism on steroids (i.e., implausible conservatism). The Agency would be much more likely to determine unreasonable risk when the true risk was reasonable.
Adherents to this policy recommendation argue that to choose assumptions that underestimate risk could allow an unreasonable risk to escape notice. Better to regulate even if there is a risk of a false positive than not to regulate and allow a false negative.
The solution here lies in risk characterization—the Agency should provide a range of plausible risk estimates along with a probability distribution showing the likelihood where the true risk lies within that range. For example, the Agency might find that the highest estimate of risk has a very low probability and that the median estimate of risk is the most probable. Such a presentation of risk is objectively correct and would not provide a false sense of precision to EPA risk managers when making a risk evaluation under the Act.
Low priority does not mean low risk. A full risk evaluation is needed to determine if a substance is low priority.
This perspective is puzzling. Besides raising an obvious question (What does “low priority” mean if it does not mean low risk?), it implies that significant resources should be expended to prove a substance is safe. A better use of resources would be to focus on chemicals that might be unsafe, as those substances are the ones that need to be controlled to improve public health.
The amended TSCA is worded carefully. A low-priority substance is one that is not high priority, and since a high priority substance must undergo an extensive risk evaluation, one can argue that only through an extensive risk evaluation can a substance be designated low priority.
This, however, is just one way a substance can earn a low-priority designation. It is quite possible that there are other pathways to a low-priority designation, and indeed, EPA has identified chemicals as safe or safer under various voluntary programs, and the methodology employed is not as extensive as a risk evaluation for a high priority chemical. The Agency can and should allow for multiple pathways to a low priority designation. The saved resources can best be used to identify and evaluate high-priority substances.
EPA should never rely solely or principally on an external draft risk evaluation, even if it conforms to EPA guidance and practice.
This recommendation ignores the considerable expertise in risk analysis that lies outside EPA. If the Agency follows this recommendation, it will be squandering a real opportunity to leverage external analysis toward realizing the objectives of the Act. Risk evaluations will be slower and more expensive as a result. The Agency should most certainly do the opposite—set standards for risk evaluation that it will follow and make these available for external use, and accept draft risk evaluations that follow its own standards.
Suffice to say, if EPA adopts these particular recommendations during implementation of the amended TSCA, it will have a significant, long-lasting, and detrimental impact because implementation decisions have staying power. Ironically, these recommendations, voiced by public interest groups, will not serve the public interest because they will make the new law slower, more costly, and less clear than it otherwise would be.
Instead, the Agency should make implementation decisions with an aim toward directing resources where it could do the most good. This would entail a focus on high priority substances, being transparent about its risk evaluation practices, characterizing risk objectively, and leveraging external expertise. The end result will be a regulatory program where decisions are faster, more predictable, and cost-effective.