My friend Tim Meyer has a new paper entitled "A Political Theory of Legal Exceptions." Back before the pandemic, Tim and I and a few others submitted a proposal for a SIEL 2020 panel entitled "Making Space for Public Policy in International Economic Law Agreements: The Role of Obligations and Exceptions." That panel never happened, of course, and this paper of Tim's predates that proposal, but nevertheless it would have been an opportunity to discuss these things. With any luck, we'll all have that chance soon!
Talking about trade policy with Tim is always interesting because we seem to either strongly agree or strongly disagree on issues. This paper has a little of both. I'm not going to give a comprehensive summary of the paper in this post, so read the whole thing to get the full flavor of his argument. Instead, I'm going to pick out a few aspects to comment on.
First off, where we agree is that existing exceptions language in trade agreements, and the judicial standards derived from them, are sometimes too narrow. Here's how Tim describes a flexible approach he advocates:
An alternative approach within new trade agreements would borrow from the experience of investment treaties. In response to decisions that they perceived as overreaching, a number of states began including clarifications in the investment chapters of their trade agreements and standalone investment treaties. This language provides that “[n]on-discriminatory regulatory actions by a Party that are designed and applied to protect legitimate public welfare objectives, such as public health, safety and the environment, do not constitute indirect expropriations . . .”
This language was adopted to clarify the meaning of rules prohibiting expropriation, but other countries have adopted the means-ends relationship – “designed and applied to”—to their general exceptions clauses. Specifically, they have modeled the scope of their general exceptions clauses on GATT article XX, but replaced “necessity” with “designed and applied to.” They have also retained the chapeau of Article XX and its nondiscriminatory requirement.
This approach dramatically improves the workability of general exceptions clauses. Unlike the necessity test, the “design and applied to” standard does not preclude the possibility that a measure had more than one motive. It merely requires that one purpose—as evidenced by the measure’s design and application—is a permissible regulatory objective. The original context of this language’s development further underlines this point. The “designed and applied to” standard was created to provide defendant states greater discretion in how they pursue legitimate regulatory objectives. This negotiating history further makes clear to treaty interpreters that multiple objectives should not be fatal to invoking an exception. By contrast, new exceptions clauses that mirror or incorporate by reference existing WTO exceptions risks codifying, in the eyes of future treaty interpreters, existing WTO case law on exceptions.
To be sure, states have adopted the non-discrimination requirement from the chapeau of GATT article XX even in those treaties that have discarded the necessity test. In recent cases like EU—Seals and Brazil—Tyres, the non-discrimination requirement has been the stumbling block to successfully invoking the exception. Modifying the language of GATT Article XX in new agreements could make clear that a Predominant Motive test applies to the non-discrimination test in the chapeau as well. Currently, that language provides that:
Subject to the requirement that such measures are not applied in a manner which would constitute a means of arbitrary or unjustifiable discrimination between countries where the same conditions prevail, or a disguised restriction on international trade . . .
The chapeau could be rewritten to codify a Predominant Motive test in the following way:
Subject to the requirement that arbitrary or unjustifiable discrimination between countries where the same conditions prevail is not the predominant objective of such measures . . .
This language makes two changes. First, it deletes the prohibition on “a disguised restriction on trade.” This language suggests that tribunals should look for hidden motives, a requirement that should be eliminated in favor of asking directly whether a legitimate objective constitutes the challenged measure’s primary motive. As the Channeling Paradigm suggests, multiple objectives is more likely the norm in policy making. Giving tribunals license to root out motives they deem “disguised” seems particularly unwise in the current context of fragility in international dispute resolution.
Second, this language makes clear that arbitrary or unjustifiable discrimination is only problematic if it is the predominant objective of discrimination among countries where like conditions prevail. This language preserves the fundamental nondiscrimination requirement of the chapeau. At the same time, though, it expands the discretion states have to choose among measures that might incidentally discriminate among countries.
In so doing, it will expand the margin of appreciation member states enjoy from second-guessing by dispute panels. Trade panels will still be able to strike down measures that are predominantly protectionist or discriminatory. But mixed motive measures that predominantly pursue permitted objectives will survive. That rebalancing between the right to regulate and the objectives of trade liberalization will bolster the legitimacy of trade law and its dispute settlement function at a time when such legitimacy is sorely needed.
I think Tim and I would end up in about the same place on these issues, although I might apply a slightly different standard. First of all, I would emphasize the relevant obligations a bit more, and perhaps distinguish between the exceptions being applied to certain obligations versus other obligations. For example, Article XX might look very different when applied to Article III than Article XI. In fact, you might not even need Article XX when Article III is involved, if you interpreted Article III in a certain way. But regardless, for each exception/obligation combination, what I would want to look at is a combination of the degree of discriminatory effect and the degree of discriminatory intent (mostly objective intent, but I wouldn't rule out subjective intent as a supporting element). In terms of the intent, I like the "predominant motive" framing. I just want to make sure that effect is not ignored.
Anyway, I basically agree with Tim here, although what I said in the previous paragraph may obscure that a bit. What I think we agree on is that the fundamental inquiry that should be undertaken -- either in the obligations or the exceptions -- is whether the measure is protectionist. Under the current standards, it's not always clear to me that's the inquiry being carried out, and I worry that non-protectionist measures will be found in violation as a result.
Shifting to the bigger picture, Tim sets out three theories of trade law exceptions. The first two are the conventional ones, and the third is something he came up with:
Despite the variety of exceptions within trade law, this Part argues that there is a conventional wisdom regarding exceptions. I distill that conventional wisdom into two approaches to exceptions: what I refer to as the Policy Space Theory and the Safety Valve Theory. Under the Policy Space Theory, governments invoke exceptions to accommodate objectives from policy domains other than trade policy, such as to fulfill domestic social or environmental objectives. An unlawful trade measures might be justified, for instance, as necessary in order to take aggressive measures to combat climate change.
The Safety Valve Theory, by contrast, holds that governments invoke exceptions to defuse political pressure that trade liberalization itself creates. This pressure typically comes from domestic groups that are harmed by competition from imports. Exceptions for measures that target unfair trade practices, for instance, allow states to reintroduce some protection into the trading system in order to reduce attacks on the trading system as a whole from these interests.
...
I ... introduce the Channeling Paradigm. The intuition behind the Channeling Paradigm is that a certain amount of domestic political pressure for protection from foreign economic activity will remain in any political system. That pressure will seek an outlet, which exceptions provide. Those interested in increasing barriers to foreign economic activity will therefore seek to form political coalitions in favor of policies that both create protection and serve some other objective authorized by the exception. The availability of international legal exceptions, in other words, channels domestic political pressure for otherwise impermissible objectives into policies that plausibly fit within the exceptions. As a result, “exceptional” trade cases are very likely to involve government policies that have mixed motives. A trade jurisprudence that does not have greater tolerance for governments’ mixed motives is one for which governments themselves are likely to have little tolerance.
His explanations of the first two theories make sense to me. But with the third theory, the disagreements may be starting. Tim argues the following:
The basic idea behind the Channeling Paradigm is that exceptions incentivize explicit or implicit bargaining among domestic interest groups. Groups with different potentially trade restrictive policy objectives, such as an environmental group and a domestic import-competing industry, will unite around a single policy that plausibly fits within an exception. Such policies will be more likely to be enacted domestically and stand a better chance of an international tribunal finding that the measure falls within an exception. ...
Here are some questions I have: Do trade exceptions really play much of a role in how domestic politics is carried out? Are the exceptions doing much "channeling"? Wouldn't all this be happening in domestic politics even if trade agreements didn't exist? I can certainly imagine that domestic actors would cite to the trade agreement exceptions as they formulate their positions, as people will always cite to things that support their argument. But I'm not sure the political results would be any different if those exceptions did not exist. It's a question of cause and effect, and I'm just not sure the exceptions are causing anything.
Also, when this does happen in domestic politics, couldn't it lead to a tainting of the non-trade policies? A measure's objectives seem a bit less pure when special interests lobbying for protection from competitors join the team.
Next up, there is the classic disagreement I have with Tim: Doesn't this bargaining among interest groups end up leading to less effective implementation of the non-trade policy at issue? It may be true that local content requirements help with the political task of generating support for clean energy subsidies, but they also put in place policies that undermine the impact of the subsidies in pushing people towards clean energy.
And finally, there is a disagreement over the economic literature about the impact of trade liberalization. Tim says:
Even the reduction in tariffs achieved through free trade agreements like NAFTA and the dramatic expansion of the WTO in the decades after the end of the Cold War had significant domestic economic effects. These agreements allowed multinational companies to disaggregate their supply chains across countries, moving jobs offshore. While the globalization of supply chains had beneficial effects for companies, and led to job growth in some communities within the United States, it also led to significant and long-lasting economic turmoil in communities in which jobs were shipped offshore.
I agree that trade liberalization is disruptive because it changes the economic status quo. But don't we all agree that we as a (both national and global) society are better off overall with NAFTA and the modest liberalization under the WTO due to that disruption? And that we would be worse off overall if the higher tariffs were still in place? Even taking into account the job losses in some regions, on balance we as Americans (and not insignificantly, poor people outside the United States) are better off. And when I say "we," I don't mean companies, some of whom were helped and some of whom were hurt by all this. I mean ordinary Americans, including those with the least economic security. Yes, we could have kept some tariffs in place and perhaps particular jobs would have been preserved, at least for a little while. But when you take into account the economic harm caused by those tariffs (job losses in other places, higher prices for basic goods), joining these trade agreements was still the right policy. Keeping the tariffs would have been economically harmful to a lot of vulnerable people.
I feel like what we need is more analysis of the disruptive impact of long-standing and newly imposed tariffs, including an examination of who in particular is harmed by them. That would be a useful way to reframe the discussion. People are always putting trade liberalization on trial, and I think we need to put tariffs and other forms of protectionism on trial a bit more. The Trump administration's various unconventional tariffs have been helpful in providing opportunities to measure the impact of tariffs, but there is probably more work to do here.
