EPA's New Framework for Human Health Risk Assessment: Clean Science Versus Dirty Costs and Benefits

Last month, EPA published for public comment a draft “framework” for human health risk assessment.  It is the culmination of years of work done by EPA staff who are part of the Risk Assessment Forum, a select team of experts from various offices throughout the agency whose efforts were overseen by the Office of the Science Advisor.  Billed as a response to the National Research Council’s Science and Decisions:  Advancing Risk Assessment (a.k.a. the “Silver Book”), the Framework really only addresses one of NRC’s recommendations.  And it proposes policy changes that threaten scientific integrity.

The Silver Book’s Chapter 8 looks innocuous at first glance, suggesting that improvements in EPA’s risk assessment process should include reforms that “make risk assessments more useful for informing risk-management decisions.”  But the simple utilitarian façade masks a proposal that would induce a revolutionary shift in the way EPA risk assessments are conducted.  To make risk assessments more useful, NRC proposed a process wherein risk managers (the agency policymakers with the closest ties to political appointees) would sit down with risk assessors (the agency scientists who read the scientific studies and interpret their meaning in terms of potential risks to the public) before the risk assessment begins, with an objective of defining the scope of the risk assessment.  The concept of risk assessors and risk managers working together is not unreasonable, but as with any practical effort to put in place a general recommendation, the devil is in the details. 

How best to maintain protective distance between risk assessors and risk managers is a controversial subject.  Without independence from policymakers and their agendas, risk assessors are extraordinarily vulnerable to manipulation.  The high profile case of the Fish and Wildlife Service’s Julie MacDonald, a Bush Administration political appointee who tampered with scientific evidence behind endangered species listings shows just how far these problems can go.

Of course, it is also fair to recognize that EPA operates under time and resource constraints that establish an undeniable need to manage the scope of risk assessments.  When BP spilled Macondo crude all over the Gulf of Mexico and wanted to use the dispersant Corexit to help with the cleanup, a full-blown, multiple-pathway, human health and ecological risk assessment just was not possible.

In publishing the draft Framework for Human Health Risk Assessment to Inform Decision Making, EPA’s Risk Assessment Forum has proposed opening up sizeable holes in the firewall between risk assessors and risk managers.  The most striking example is the Framework’s discussion of the link between human health risk assessment and cost-benefit analysis.  It begins by making the argument,

In some regulatory contexts, a risk assessment (or some quantitative aspect of it) may be a key input into benefit-cost analyses of alternative regulatory options; these analyses may impose different or additional requirements by comparison to regulatory contexts wherein costs are not considered.

As a description of how certain regulatory programs work, that passage is mostly unobjectionable.  But the document goes on to argue that “early communication between the risk assessment and benefits analysis teams” is advisable and that the leading consideration in the discussions should be:

Identifying a set of human health endpoints that are economically meaningful; that is, endpoints that can be linked to human well-being and for which the risks can be monetized using economic valuation methods.

CPR Scholars and staff have published more than a few documents that highlight the fallacies inherent in cost-benefit analysis, including the misconception that regulatory decisions should be driven by the benefits that can be predicted and monetized.  It is bad enough when the ultimate choice between different regulatory options turns on the ability to monetize benefits.  Frank Ackerman and Lisa Heinzerling make the point eloquently:

When the question is whether to allow one person to hurt another, or to destroy a natural resources; when a life or a landscape cannot be replaced; when harms stretch out over decades or even generations; when outcomes are uncertain; when risks are shared or resources are used in common; when the people “buying” harms have no relationship with the people actually harmed—then we are in the realm of the priceless, where market values tell us little about the social values at stake.

The Risk Assessment Forum’s draft Framework sets the stage for scientists to ignore some human health risks posed by chemicals simply because economic advisors expect it will be difficult to monetize the costs.  As we learn more about chemicals’ epigenetic and intergenerational effects, the pricelessness of risk assessors’ work becomes more evident; so too, as we discover toxic chemicals’ ability to spread across the globe.  (Phthalates in the arctic, anyone?)

When the draft Framework is finalized, EPA should eliminate the passages that suggest risk assessment should be subsumed by cost-benefit analysis.  More importantly, the Framework needs to identify who makes the ultimate decisions about the scope of agency risk assessments.  As it stands, the draft is vague on that point.  It encourages consultation between risk assessors, risk managers, and others, but fails to say who will lead those discussions, how conflicts should be resolved, and who will have the final say.  Again, these are controversial issues and they must be resolved before this document is finalized.  If risk assessors’ autonomy and control over their work is not firmly established in the final Framework, it will undermine the scientific integrity of EPA’s work.