Risk, science, and law in the WTO.

Position:International Law in a Time of Change - Proceedings of the 104th Annual Meeting of the American Society of International Law - Discussion

This panel was convened at 9:00 a.m., Thursday, March 25, by its moderator, John Jackson of Georgetown University, who introduced the panelists: Sungjoon Cho of Chicago-Kent College of Law; Gregory Shaffer of University of Minnesota Law School; Tracey Epps of the New Zealand Ministry of Foreign Affairs & Trade; and Annecoos Wiersema of Moritz College of Law, Ohio State University. *

* Sungjoon Cho did not submit remarks for the Proceedings.


It is sometimes observed that in our complex, technologically dynamic world, legal institutions attempt "to import science into law" and "export law's problems to science." (1) For legal institutions, it is sometimes thought that the expertise of science can help to resolve competing claims in a neutral manner, evade normative judgments with their distributional implications, and legitimize legal decisions through the authority of science. For many, however, this move to incorporate science into legal judgments is highly problematic because it permits judges, who do not understand scientific methods and their limitations, to evade judicial responsibilities. This concern lies at the heart of many critiques of decisions of panels in cases under the Agreement on Sanitary and Phytosanitary Measures (SPS Agreement) of the World Trade Organization (WTO). As you will see, in acknowledging this point, I take a complementary institutional tack. In my remarks, I will also extrapolate from what I have learned from my book-length study with Mark Pollack of the international law and politics of genetically modified foods. (2)

One starting point for a WTO panel in a case involving a measure that potentially falls within the scope of the SPS agreement is to ask why there is a reference to science and risk assessment in the agreement. In particular, Article 2.2 of the agreement requires WTO members to "ensure that any [SPS] measure ... is based on scientific principles and is not maintained without sufficient scientific evidence," regardless of whether it is applied equally to domestic and foreign products. Article 5.1, in turn, prescribes" "Members shall ensure that their sanitary or phytosanitary measures are based on an assessment, as appropriate to the circumstances, of the risks to human, animal, plant life or health." The only exception is "where relevant scientific evidence is insufficient," in which case, under Article 5.7, "a Member may provisionally adopt ... measures on the basis of available pertinent information," subject to certain conditions.

It seems clear from the negotiating history of the SPS Agreement that the reason why the references to science and risk assessment were made is because of the negotiators' concern with protectionism, and, in particular, discrimination against foreign agricultural products. Anyone who follows trade policy knows how politically sensitive agricultural trade is. This sensitivity is illustrated by (1) the lack of full incorporation of agriculture under the GATT system; (2) the creation of the Agricultural Agreement and the tariffication of non-tariff barriers, but with the result that there are much higher average tariffs and other trade distortions permitted for agricultural products than for non-agricultural products; (3) the creation of a distinct SPS agreement out of concern that SPS measures may be manipulated to evade the constraints of tariff bindings; and (4) the major concern of developing-country representatives with SPS measures confronting their exports to the United States and European Union, in particular, but also to other developing countries.

A separate starting point, however, is from the perspective of risk regulation itself in an interdependent world in which not only goods, but also harms, cross borders, whether they be harms to the environment or to human health. As social theorists Ulrich Bech and Anthony Giddens have written regarding the modern condition, through science we constantly manufacture new risks which we then attempt to address through science. The concept of risk refers to "the combination of the likelihood (probability) and the harm (adverse outcome, e.g. mortality, morbidity, ecological damage, or impaired quality of life) resulting from exposure to an activity (hazard)." (3) We mainly depend on our national governments and transnational regulatory cooperation to address these risks, complemented at times by specialized international institutions. However, national regulators often disagree. They often disagree based on factors other than environmental and health risks, and it can be difficult to discern the motivating factors.

Where there are disagreements among constituencies regarding the risk of a product, supporters of the product or technology tend to focus on the need for a neutral scientific risk assessment, as per Articles 2.2 and 5.1 of the SPS Agreement. In contrast, skeptics of a technology (such as of genetically modified (GM) products, for example) tend to focus on the uncertainty of its effects, insisting that uncertainty is a different concept than risk, and entails not just differences in degree but of kind. They point to the classic Knightian distinction that risk is something that one can calculate, while uncertainty is something that one cannot. Borrowing from sociologists such as Ulrich Beck, some argue that with such modern industrially produced risks, "the actual consequences ultimately become more and more incalculable." (4) They thus focus on the need for considerable precaution in risk management decisions, such as regarding the adoption of new agricultural production processes. In addition, they argue that risk assessments under uncertainty are inherently value-laden, since the perception of risk itself reflects cultural predispositions, including among scientists, so that risk assessments are not purely technical. Where value claims are raised, disputes gain political salience. In the case of the regulation of GMOs, for example, opponents of the technology emphasize the values of non-industrialized, non-corporate-dominated, non-gene-tampered, "slow food" agriculture. People committed to these views do not wish to engage simply in the discourse of risk assessment.

So what ought a WTO panel to do in light of the texts and conflicting claims? From a first order of analysis, we could view the texts as calling for a double delegation of authority to two non-majoritarian institutions for resolving disputes over risk regulation that affects trade: (1) delegate dispute resolution to a third-party decisionmaker, the panel and Appellate Body; (2) delegate (in part) assessment of the adequacy of a risk assessment to independent scientific experts selected by the panel. This latter delegation is not a complete one, but the scientific expert opinions can sway panels, cognitively shaping the way panelists view the issues.

From a second order of analysis, however, WTO panels and the Appellate Body face interpretive choices of how to weigh science and the application of a risk assessment that will have institutional implications. In terms of the assessment of the risks mentioned in Article 5.1, we can view a panel's choices in terms of three broad options, which are part of a continuum of showing greater to lesser deference to a national administrator:

(1) formal proceduralism. Under this approach, all that would be required is for the national administrator to reference a risk assessment that went through relatively easy procedural hoops, whether the risk assessment is conducted by the regulator or a third party.

(2) review process subject to circumscribed criteria. Here the panel would review the adequacy of the risk assessment along certain minimum substantive criteria defined by the Appellate Body.

(3) independent substantive judgment. This approach would make an independent determination on the weight of the scientific evidence, likely with a margin of maneuver for the national administrator.

Although the WTO panels and Appellate Body maintain that they are following the middle course, they can do so with different degrees of stringency regarding the requirements for a risk assessment. They often will be tempted to get a sense of the bottom line (the third option), even if they do not acknowledge they are doing so.

The panel's interpretive options can be viewed in broader institutional terms in terms of who decides, and how they decide, issues involving social risk. One can see the interpretive choices as implicating different allocations of institutional authority, which I categorize into five ideal types. Each of these choices involves different trade-offs, which likely will weigh differently in discrete cases.

First, a panel can show deference to the national decisionmaker, provided that the decisionmaker went through the formal procedural hoops of a risk assessment. This strategy is tempting for the reasons that Andrew Guzman has articulated. (5) The risk to the WTO system could be great if a panel gets the scientific judgment wrong, and these risks could exceed the benefits of stringent review of SPS measures. Science has gotten some matters dramatically wrong. For example, Francis Galton discovered the regression to the mean which is central to the empirical move in the study of international law (for which there is a panel taking place parallel...

To continue reading