COMMENT BY: MICHAEL GOUGH, Cato Institute
SUBJECT: Response to Dr. Milman's comment on EPA Cancer Risk Assessment Guidelines
DATE: August 30, 1999

Dr. Harry A. Milman's email compliments EPA for its efforts to review its draft Cancer Risk Assessment Guidelines and suggests making the guidelines more consumer friendly. My take on the guidelines and on EPA is quite different.

EPA's big change in the draft guidelines is a commitment to "consider all the evidence" and to abandon the linear, no-threshold model (LNT) as the default procedure for estimating cancer risks. As an alternative, EPA proposes using a safety factor approach, but it provides few details about the magnitude of the safety factors.

David Gaylor of the National Center for Toxicological Research (NCTR in Arkansas; I may not have the name exactly right) and, independently, Steve Milloy and I in a Cato Policy Analysis, (https://www.cato.org/pubs/pas/pa-263.html) point out that the practical effect of using a method different from the LNT may be negligible. EPA lists 5 factors that might be multiplied together to calculate the appropriate safety factor. If each of those five factors is assigned a value of 10, then the safety factor becomes 100,000. Use of a safety factor that large produces estimates of acceptable exposure levels that differ by a factor of about two from the acceptable level calculated by use of the LNT.

Granted, using a safety factor of 100,000 is the extreme, and EPA does not stipulate that it always be so large. But, in fact, arguing about risk estimation models is besides the point. EPA is guided by policy; not by guidelines, let alone by science.

EPA has carried out two high visibility exercises with the draft guidelines in mind. The "dioxin re-assessment" completed in 1994, was reviewed and trashed by the EPA's Science Advisory Board in 1995. All the evidence considered leads almost every observer to conclude that the LNT is not applicable to dioxin, and the SAB refused to accept EPA's reliance on it to estimate risks from dioxin. (EPA told the SAB in May 1995 that a revised re-assessment would be available in 6 months. The SAB and everyone else is still waiting.) There was little justification for EPA's use of the LNT and none for its out-of-hand rejection of other methods to estimate the cancer risk from dioxin.

In last December's rule making about the risks from chloroform in drinking water, EPA accepted the judgment of its scientists that those risks can be estimated using a safety factor approach. Despite that acceptance, the agency decided to apply the LNT to chloroform risks. Why? Because EPA recognizes that using anything else than the LNT would be a big change, and, evidently, it can't make such a change based on science. It has to consult with stake holders.

There's no reason to think that new guidelines, if they are ever made final, are going to change EPA policy.

But far beyond that, everyone knows that environmental pollutants contribute little to the cancer burden. What have we, as a nation, purchased for the $100,000,000 we have spent for dioxin cleanup? What do we gain from a policy that insists that any level of exposure to chloroform in water increases cancer risk? In both cases, very little, and, perhaps, nothing.

New guidelines aren't the answer. The answer, if it ever comes, will be from politicians courageous enough to apply science to the assessment of environmental health risks and to tell the public that its resources are being squandered on trivial or nonexistent health risks.