[*PG185]SCIENCE AND “POST-DISCRIMINATORY” WTO LAW

Robert E. Hudec*

It seems to me that the word “kickoff” would describe my remarks today better than “keynote.” The word “keynote” makes its sound as if I am confident of what I am doing and that I will be guiding all of you. By contrast, “kickoff,” as you know, is an exertion of energy toward a football without very much idea exactly where it is going to land. That, I think, is a somewhat more appropriate description of my remarks today.

The prospectus for this conference asked us to focus on provisions of international trade law that call for decisions about the scientific validity or the scientific justification for certain regulatory provisions. The most visible of such provisions is, of course, the [World Trade Organization] [(WTO)] Agreement on the Application of Sanitary and Phytosanitary [SPS] Measures adopted in 1994 in the Uruguay Round, in which governments undertake an obligation to ensure that [SPS] measures must be based on scientific principles, and in certain cases, must be based on a proper risk assessment. Another agreement which seems to call for a similar kind of scientific investigation is the Agreement on Technical Barriers to Trade, the so-called Standards Code. In addition to those two, I would like to call attention to the fact that issues of scientific validity are sometimes central to the application of [General Agreement on Tariffs and Trade] [(GATT)] Article XX, an exemption provision for public policy measures, which states that if a measure is necessitated by some reason such as health or safety, it may be permitted even though its terms it would violate the agreement. And finally, under some formulations of GATT non-discrimination provisions, the character of a measure as discriminatory or non-discriminatory depends upon the question of whether its enactment can be justified by some valid non-trade purpose. It should be apparent that a measure’s legality by reference to that test often depends on the scientific validity of the claim of non-trade purpose.

[*PG186] Dean Garvey, in his introductory remarks, [reproduced in the present volume] anticipated a good deal of what I was going to say. One aspect of constitutional law in the United States, the so-called Dormant Commerce Clause, is very similar to what one sees in the GATT and WTO with regard to the interplay between the concept of discrimination and the scientific justification for regulatory measures that are being attacked. This is a central question about science-based legal criteria. Stated broadly, the issue is whether, and to what extent, science is actually capable of answering the questions posed by such provisions. Participants in this conference are asked to discuss the nature of the scientific issues that are likely to be confronted in the application of those trade law provisions from a variety of disciplinary perspectives and the relationship between those provisions, and of scientific analysis, to the policy goals which those provisions address. The goal is a better understanding of how such science-based provisions are likely to operate in practice and, ultimately, an understanding of the impact that such provisions are likely to have in the field of national regulations.

My role today is to provide some context for these questions and to point out some more particular questions that might usefully be addressed. My contribution will have nothing to do with the nature of science or scientific inquiry, a subject on which I am not qualified to speak, but rather will be framed from the perspective of a legal scholar familiar with the different types of trade law provisions at issue here and familiar with the adjudication process by which they are applied. Having served as a member of several GATT, WTO, and [North American Free Trade Agreement] [(NAFTA)] adjudicatory panels, I can perhaps offer some particular observations from the perspective of one who has actually had occasion to render legal judgments applying some of these provisions in practice.

One of the most useful things I might attempt to do is to clarify the various purposes of the legal provisions that are causing concern in this area. On the basis of that further explanation, I hope to be able to make some suggestions that will add greater precision to the questions asked in this conference and, accordingly, to the conclusions reached. I am sure that many of you will know a good deal more about what I am going to say, but I believe it can be quite useful from time to time to rearrange what one already knows about a subject.

The central theme of my remarks is that the approach one takes to the use of science in science-based legal criteria will often depend, consciously or not, on the assumptions that one makes about the purposes of these legal criteria. I think it would be useful if, at a mini[*PG187]mum, authors criticizing the science-based criteria in the various trade law provisions they are addressing were to ask themselves about the assumptions they have, the assumptions they make about the purposes of the provisions they are talking about. It would be even more helpful if they were actually to articulate the assumptions they make about the purposes of these provisions. Even the members of the trade policy community sometimes have difficulty in identifying the purposes of WTO rules.

Traditionally, trade agreements have focused on limiting or eliminating discrimination against foreign trade by disciplining governmental measures that impose competitive disadvantages on foreign goods vis--vis domestic goods with which they compete. In the recent Uruguay Round trade agreements, however, it appears that the draftsmen of two key agreements added another goal, one that can be described as the prevention of unjustified regulation per se, whether or not such a regulation creates a competitive disadvantage for foreign goods vis--vis domestic goods. Thus, for example, a food safety measure that is not based on scientific principles would be a violation of Article 2 of the Agreement on the Application of [SPS] Measures, whether or not it discriminates against foreign goods. While other rules of the SPS Agreement are directed at traditional trade agreement concerns about discrimination, it is clear that violations of provisions like Article 2 do not require findings of discrimination.

Another example is the Appellate Body decision in the hormones case, which finds a violation of Article 5.1 of the SPS agreement on the basis of the European Community’s failure to base its hormone restrictions on a proper risk assessment. The Appellate Body in that case expressly rejected the panel’s finding that the measure was discriminatory. In reaching this conclusion, the Appellate Body said that the absence of a risk assessment is a violation of the agreement even though there is, it found, no trade discrimination. I am told by those who negotiated the Uruguay Round Standards Code that the same is true of the key obligation, Paragraph 2.2 of that agreement. It provides for the prohibition of product standards that could create “an unnecessary obstacle to international trade.” Although, to my knowledge, there have been no WTO decisions actually so holding, there seems to be quite broad agreement that this is the meaning of Article 2.2.

The emergence of this new type of legal standard, which I would propose to call “post-discriminatory”—going beyond discrimination—surely changes the purpose of the scientific inquiry called for by such legal standards. One can loosely describe such standards as still being [*PG188]devoted to removing barriers to international trade, in the sense that any unnecessary regulation impedes commerce of all kinds, including international trade. To the extent that these legal criteria do not require proof of discrimination, however, they become a rule simply by calling for an international tribunal to second-guess the rationality of a regulatory judgment at the national level.

Measures that have this very broad second-guessing purpose are, in my view, open to legitimate criticism on several grounds. These are reasons that do not fundamentally involve the nature of the scientific judgments that the rules call for, objections which you might say are prior to the question of the feasibility of the scientific inquiry. One can object that such a regulatory purpose of preventing unnecessary regulation is simply outside the scope of the jurisdiction or the purpose of the [WTO]. The [WTO] is supposed to be about promoting international trade, and this is not part of that original mission or indeed its stated mission.

One can also object that such a rule, which disciplines the rationality of national regulation even absent a finding of discrimination, actually gives foreign traders a higher set of legal rights, a greater set of legal rights, than is given to the domestic producers with whom they compete. To my knowledge, in no legal system are there provisions which give rights to object to the rationality of regulations which are as broad as those in the WTO Agreement on the Application of [SPS] Measures. There are, of course, regulatory laws under which scientific judgments are reviewed by courts as a matter of administrative law, by reference to a test of whether they comply with the standards set by the originating legislation. But the WTO provisions on discrimination are much broader than that. So one could object to a kind of super-status given to foreign traders that actually would be better than the standards given to domestic traders with whom they compete. In this sense, I think these provisions are a little like the investment provisions that you will be hearing about later in the day [and which are the subject of Todd Weiler’s contribution to the present volume]—the investment provisions which also seem to give foreign investors more rights against certain kinds of regulations than their domestic competitors would have.

This is not to say that questions of scientific validity, or the questions we are asking in this conference about the nature of scientific inquiry and scientific justification, are irrelevant. I think they add a third level of criticism to these kinds of measures. On that level one can say, in addition to the previous observations I have mentioned, that these measures are not really as objective as they appear. The sci[*PG189]ence under which they qualify is not, cannot be, as objective as one might pretend.

The reason I ask for a careful identification of purpose is that it is fairly obvious that it is important to be clear about the nature of objections one makes to these provisions. Clarity for its own sake is always good. In this kind of case, in which one has a highly objectionable provision—and I think they are highly objectionable—it is very tempting to let your objections get mixed in with what you are saying about the nature of scientific inquiry. If one looks at a provision that is thought to be genuinely wrong from the bottom up, it is possible, it is tempting, consciously or unconsciously, sometimes to exaggerate the particular problems that one has with the application of the scientific criteria to that measure. So my first point, again, is to ask for clarity about the purpose of the measure that one is talking about: to identify those of the post-discriminatory kind, to be clear about the criticisms one is making, and to identify which of the many criticisms one can make about them if one is in fact making a particular point.

Now the other side of the coin, of course, is that we should also think carefully about the traditional tests, the anti-discrimination provisions. The place to start is to recognize that the word “discrimination” that we use in all these cases is a normative term expressing a judgment of disapproval. When we see a regulation that has the effect of putting foreign goods at a competitive disadvantage, the neutral descriptive term for that situation is that the regulation has a “differential impact.” If we think there is nothing wrong with that differential impact, we continue to call it a differential impact. If we think there is something wrong about the differential impact, then we call it discrimination.

The normative judgment that all differential impacts are wrong, are discriminatory, usually requires something more in a particular situation than the mere fact that a measure or regulation has a differential impact. The most common additional element to a charge of discrimination is the element of purpose or intention, the notion that the measure was intended to harm the disfavored class. GATT and WTO law may use many other more mechanistic kinds of legal criteria and I do not propose to trace all of these in these prefatory remarks. I am going to simply suggest that if you scrape off all that surface legal doctrine, you will find that most of these criteria, most of the cases, are really aimed at establishing whether or not there has been a discriminatory purpose to the measure in question.

I find it interesting on this question of intention or purpose that this particular standard seems to play on both sides of the debate. [*PG190]Whenever I see a paper arguing that a particular measure should not be regulated by GATT or WTO, I almost always see that measure described by a statement about its intention; it is intended to protect human health; it is intended to protect the environment. This is another way of saying that this is a legitimate, good measure that we should therefore support.

There is more to this business of intention than simply the normative mechanics of how we look at these situations. There is something of an unstated agreement among governments as to how much of this kind of regulation they can really tolerate. A regulation itself often entails impacts that are difficult to control. One cannot tell when one adopts a regulation whom it is going to affect and how it is going to affect them. There are many cases of unintended differential impact; this is part of the game, part of the consequences of regulation. Governments looking at this situation have come to the conclusion that they cannot stand the level of intrusion that would occur if regulations, all regulations, were subject to prohibition merely based on the fact of discriminatory impact, whether accidental or purposeful. So this is an attempt to regulate the intrusiveness of GATT/WTO supervision of these issues by saying in effect that one has a justified claim to regulators in cases in which they are deliberately discriminating. This intention test is also kind of a pragmatic standard, which attempts to draw a line delineating how much regulation is too much.

Now I would like to turn to the application of these discriminatory rules. There are two kinds of WTO cases in which the issue of discriminatory purpose comes up. One is a case in which a government explicitly, deliberately, adopts a different and less favorable rule for foreign products, but in which the government claims it has a valid public policy reason for imposing the disadvantage. The issue of discriminatory intent turns on the validity of the claim of public policy purpose. Here, I am happy to accept what Dean Garvey told you a few moments ago as a matter of internal U.S. constitutional law under the Dormant Commerce Clause.

The second type of discrimination case is probably the more common and also the more difficult case, in which the government adopts an ostensibly neutral regulation, such as the avocado regulation that was mentioned by Dean Garvey in his introductory remarks, a distinction between products on ostensibly neutral grounds. The question is whether that distinction, which places foreign goods at a disadvantage, is based on some valid neutral and non-discriminatory purpose or whether, alternatively, the purpose is in fact to discriminate against California avocadoes, for instance. Although there are [*PG191]significant differences between the two types of cases, I am going to concentrate today on the first case because I think I can make the points I want to make by talking about that kind of case.

The first category of discrimination cases is illustrated by the WTO’s first environmental case—the reformulated gasoline case. The regulation was a product standard for gasoline that imposed a separate, different, and more rigorous standard for foreign gasoline. The U.S. government claimed that the different standard was necessitated by the environmental, the clean air goals of the regulation. It said that applying the same product standard for domestic gasoline to foreign gasoline would not in fact achieve equally clean gasoline because the government could not enforce that standard effectively against producers located in other countries. This is not a scientific judgment in the sense that this word is being used in this conference. That was a judgment about regulatory feasibility—the likelihood that the objectives of the regulation could be enforced. The primary issue confronting the WTO decision-maker was whether the decision was in fact made on that basis. The facts of the case at least raised the issue about this central question. To be sure, enforcement would have been costly in the case of foreign gasoline and there was actually a plausible argument that it might not work as well.

But in the other direction, the Environmental Protection Agency[(EPA)] in drafting the regulation had been prepared to apply the same product standard to foreign gasoline and it had testified that it believed it could enforce that standard effectively against foreign suppliers. The U.S. Congress overruled the EPA and ordered the EPA not to use the same standard to regulate foreign gasoline. Now at this point the question is the following: Was Congress itself making an independent judgment about the effectiveness of the regulatory question involved, or was it engaging in some kind of deliberate protectionism? Further evidence gave some reason to suggest that the second might have been its purpose. The issue had been the subject of intense lobbying in Congress by the U.S. producers that competed with the principal source of foreign gasoline that was at issue here, Venezuelan gasoline being sold by, I believe, Citgo in the northeastern United States. Now the explanation suggested by this information was that the more rigorous product standard had been adopted simply to provide trade protection for a local constituent. That was at least a plausible possibility.

Put yourself in the position of the decision-maker dealing with this case. Here is a potential case of purposeful discrimination against foreign producers. How is the WTO decision-maker to decide [*PG192]whether or not this regulatory measure is in fact purposeful discrimination against foreign trade? A common answer at this point is that the WTO should give deference to national government decisions on such regulatory matters, but deference to what? What kind of deference? Deference to a technical judgment made by the government? Deference to a judgment of regulatory policy made by the government? Or should the WTO also show deference to a possibly questionable claim that the regulatory judgment was in fact made? This is a claim that every government will make in this case. If the last kind of deference is due—in other words, deference to the good faith of the government—the concept of deference becomes an ironclad defense. There is no way to challenge it. If the WTO decision-maker has to defer to good faith in what the government did, there is no way to get at what may be a deliberate trade restriction.

Note two things about the issue before the WTO decision-maker. First, officials who work in this area are convinced that governments make deliberately protectionist decisions all the time. They have internalized the teachings of experience by looking at the hundreds of previous cases brought under GATT or WTO rules or under the Dormant Commerce Clause in the United States, the cases referred to by Dean Garvey, or under Article 30 of the Treaty of Rome, the EU equivalent of our Dormant Commerce Clause. This is a collection of cases in which one can see hundreds upon hundreds of government decisions that discriminate in this fashion. Officials can also cite the simple common-sense observation that the political settings for these decisions, namely a contest between a local economic interest represented in the government and a foreign economic interest that is not represented in the government, makes protectionist decisions clearly the more probable outcome. And finally, they can cite the large regulatory institutions that have been constructed in the GATT/WTO, in the [United States] in the form of commerce clause litigation, and in the EU, to deal with protectionist measures like this one as evidence that generations of officials have in fact recognized this possibility and felt that it was serious enough to require addressing. Consequently, if the rule of decision is to avoid second-guessing the bona fides or good faith of government decisions in such a case, then a large amount of trade discrimination will go unchecked.

The second point illustrated by this case is that in most legal settings where tribunals actually do inquire into issues of discriminatory purpose like this one, it is a standard canon that tribunals cannot judge the actual motivation of the persons who made the decision—because, in modern governments, we cannot determine who actually [*PG193]made the decision in the first place, and because even if we could, we do not have the tools to know what motivated the decision of those actors, in any event not to any kind of satisfying degree of certainty.

The standard way of deciding such purpose questions in GATT law, U.S. Constitutional law, and European Community law is by reference to what I usually refer to as “objective indicators.” The objective indicators, when applied, look a great deal like second-guessing the judgments of governments. In general, one looks to whether the measure was really necessary to obtain the declared goal. For example, one examines whether less restrictive alternatives were available and not used, or whether the decision in areas affecting foreign goods is in fact consistent with the kinds of decisions that are taken in analogous areas which affect sellers of domestic goods.

In form, a decision that the purpose was discriminatory, but which does not actually mention purpose but instead is framed in terms of necessity or less restrictive alternatives, comes out looking like second-guessing. In the gasoline case, for example, the decision came out that the regulatory conclusion was wrong—that, contrary to the government’s claim of impossibility of regulating equally, in the view of the Appellate Body with a little work it would have been possible to achieve the same result by applying the domestic standard to foreign gasoline.

To be accurate, one should admit that there is more than a kind of objective second-guessing that goes on in such decisions. In my experience decision-makers who are familiar with government policy making in this area will usually take account of the entire context of the measure as they approach it and will probably come to an early opinion, tentative though it may be, about whether the measure “smells” discriminatory. The same smell test will be found in the community at large. Other government delegations will form an opinion, members of the Secretariat will form an opinion. A good decision-maker will try to base his or her decision on objective factors. I believe it would be dishonest to deny the influence of these intuitive judgments. Actually, in this whole business the best informed participants about government motivation are of course those who are representing the defendant government, who, more likely than not, actually participated in the decision-making process back home. Although such representatives will steadfastly defend the measure during the litigation, they often provide validation for the smell test after it is over. This is one of the ways in which everyone’s sense of smell begins to sharpen in one case after another. If you have been in this business for a while you have not only seen cases that look wrong, [*PG194]but you also have a strong number of confessions telling you that in fact they were as wrong as they looked.

The second-guessing of government regulatory decisions that goes on in a trade discrimination case occurs in a much different setting from, and for a much different purpose than, the second-guessing that seems so unwarranted in the post-discriminatory part of the SPS agreement that I discussed earlier in these remarks. First, the typical discrimination case concerns a situation in which foreign goods are in fact being placed at a disadvantage vis--vis domestic goods. There is a differential impact, the central problem at which the international trading system is directed. Second, a discrimination case concerns a question of wrongful purpose, not just a mistake in judgment. Third, discrimination cases involve a situation in which government wrongdoing is a highly possible element—not probable, but possible.

The next point is whether the need to regulate measures with discriminatory purposes may not be important enough to justify a workable, rather than a perfect, standard of decision-making. Domestic law often decides important questions with a degree of certainty of less than one hundred percent. In simple tort and contract suits, for example, courts decide which of two conflicting versions of the facts has occurred by asking which of the two versions is simply the more probable, even though we know that there is no way to know for certain what the truth is. In criminal cases we allow courts and juries to decide whether criminal defendants had certain states of mind. Although the standard of proof applied in those cases is the much higher “beyond a reasonable doubt” test, we know that we cannot really know for certain what state of mind the defendant had. We decide these issues nonetheless because the alternative of not deciding them is chaos.

Now the other question is this. In the gasoline case, how certain must a panel be that the same standard can be applied effectively to foreign gasoline before it can properly issue a decision saying that? Would it be good enough if [the panel was] convinced that the quality of foreign gasoline would be almost as good? What does this long exposition of the gasoline case have to do with the questions about science raised in this conference? Very briefly, it is my way of asking whether the questions being asked about science are the right ones, or, rather, whether the very good questions that are asked do not need to be supplemented by some other quite relevant questions.

The questions we are asking today are introduced by saying that the SPS provisions are based on a premise that the scientific commu[*PG195]nity, that the scientific issues that they pose, can be answered fully and in a completely objective manner. To the extent that is true, and to the extent that is the purpose of these scientific criteria, it is certainly highly relevant to demonstrate the ways in which science cannot perform the tasks assigned. I am trying to show that the WTO has many other provisions in which imperfect, best-that-we-can-do answers may be good enough, because a demand for perfection will block all regulatory activity.

If one thinks about what science can contribute to the operation of such legal provisions, one has to start adding some other questions. For example, instead of asking whether science can contribute an objective answer to the question whether the measure is based on sound science, should we also not be asking whether science can provide any relevant evidence on the question whether the purpose of that particular measure is protectionist? Or, to ask almost the same thing, whether science can provide any relevant evidence on the question whether a particular measure has any other credible regulatory purpose other than that of protectionism?

Thank you very much, and I look forward to hearing the thoughts of others on these important issues.

[*PG196]BLANK PAGE

?? ??