September 6, 2000 / Biotechnology Viewpoints
Several factors support a precautionary approach for genetically modified
(GM) foods, including the irreversibility of introducing living modified
organisms into the environment, the large uncertainties about the risks of
such introductions, and the widespread human and ecological exposure to GM
organisms. Yet, at the same time, excessive precaution can suppress
important benefits offered by biotechnology products, impose unnecessary
costs on society, and perhaps even increase net risks as a consequence of
risk-risk tradeoffs.
While some precaution may be warranted, application of the precautionary
principle (PP) to GM foods is problematic for at least two reasons. First,
the indeterminacy of the PP makes it an inappropriate and ineffective
regulatory decision-making tool. The PP provides no guidance on any of the
fundamental questions that are faced in making any risk decision. The PP is
ambiguous, for example, as to what level of risk is acceptable, what role
costs should play in risk decisions, what quantum of scientific evidence is
sufficient for making decisions, and how potential risk-risk tradeoffs
should be addressed. Proponents of the PP not only disagree on these
important questions, but also on whether the PP should apply in the risk
assessment process, the risk management process, in both risk assessment and
risk management, or as a substitute for the current risk assessment/risk
management paradigm.
If its proponents cannot even agree on its practical meaning, the PP surely
cannot provide meaningful decision-making guidance for governmental and
industry risk managers who must make defensible real-world risk decisions.
Especially in the U.S., regulatory agencies must follow or articulate
"intelligible principles" to cabin their regulatory discretion. As presently
formulated, the PP offers no intelligible decision-making principle. At most
then, the PP may serve as a general aspiration or goal for a regulatory
system, perhaps appropriate for the preambles of international treaties and
domestic statutes, but certainly not as a legally binding regulatory
stricture.
The second problem with the PP is that it represents a major leap backward
from a focus on risk to a focus on hazard. While used interchangeably in
common parlance, "risk" and "hazard" have distinct meanings in the risk
analysis literature. "Hazard" is the intrinsic potential of an agent to
cause an adverse effect, whereas "risk" is the likelihood and magnitude of
the adverse effect occurring under real-world exposure scenarios. While many
earlier regulatory decisions were based on hazard identification, there has
been increasing realization that a fuller characterization of risk usually
provides a better basis for making regulatory decisions. For example, for
carcinogenic chemicals, agencies such as the National Toxicology Program
(NTP), the International Agency for Research on Cancer (IARC), and the U.S.
Environmental Protection Agency (EPA) originally classified carcinogens
based only on findings of hazard (usually the results of chronic rodent
bioassays). However, all of these agencies have recently recognized that
exposure and mechanistic considerations can produce risk-based evaluations
that are more meaningful than the original classifications based on hazard
identification alone.
The PP, at least as defended by some of its strongest proponents, would
appear to be directed at hazard rather than risk by calling for
precautionary measures once some indicia of hazard exist. Yet, we know that
every substance or product has the intrinsic potential for some hazard,
which may or may not translate into real-world risks of concern. Because
hazard potential is ubiquitous, basing regulatory decisions on hazard alone
creates the potential for arbitrary, unfair and inefficient regulations.
With respect to GM foods, many potential hazards can and have been
hypothesized, ranging from interspecies gene transfer, toxicity to
non-target species, and new allergens or toxins in foods. While the limited
evidence supportive of such hazards may be sufficient to trigger the PP, and
thus to block the introduction of GM foods, there is no evidence to date
demonstrating that GM foods present an actual significant risk to human
health or the environment.
The finding, and particularly the hypothesis, of a potential hazard should
not automatically necessitate a ban or other prohibition, but rather should
be the impetus for additional scrutiny including data gathering, pre-market
test requirements, post-market surveillance, and risk assessment. To be
sure, sometimes the finding of a hazard alone will be sufficient to justify
interim precautionary steps, especially when the general type of evidence
produced is known to be predictive of actual risks. But the ultimate goal
should be the assessment of real-world risks using sound science and expert
judgment, critical inputs that the PP treats as dispensable.
Professor Gary E. Marchant
Arizona State University College of Law
Tempe, Arizona, U.S.A.
