Robert Staiger and Alan Sykes have posted "International Trade and Domestic Regulation" on SSRN. From the abstract:
Employing the terms-of-trade framework for the modeling of trade agreements, we show how "large" nations may have an incentive to impose discriminatory product standards against imported goods once border instruments are constrained, and how inefficiently stringent standards may emerge under certain circumstances even if regulatory discrimination is prohibited. We then assess the WTO legal framework in light of our results, arguing that it does a reasonably thorough job of policing regulatory discrimination, but that it does relatively little to address excessive nondiscriminatory regulations.
The issue of how WTO rules should and do address non-discriminatory measures has long been an interest of mine and is something I have blogged about before (see, e.g., here). In my view, the extent to which trade agreements should go beyond discrimination is one of the most important issues facing the trade regime. Are there good reasons to regulate non-discriminatory measures? Is there a downside to doing so? The Staiger/Sykes paper provides a number of insights on these issues.
Early on in the paper, they state:
We then suppose that governments agree to a national treatment clause that applies to both domestic taxation and regulatory standards, and we ask: Will a tariff agreement that is protected by this broader nondiscrimination rule allow governments to reach internationally efficient policies? Again we show that the answer is "no," because governments have an incentive to distort their consumption taxes to inefficiently high levels even if these taxes cannot be set in a discriminatory fashion, and they may (and will, if product-level consumption taxes are unavailable) have an incentive to distort upward their nondiscriminatory product standards as well.
Our economic analysis thus leads to the following broad conclusion. To achieve internationally efficient policy levels in this environment, tariff agreements must include rules that prevent the use of discriminatory domestic tax and regulatory policies, while at the same time preventing governments from setting excessively high nondiscriminatory taxes and product standards in response to the ability to shift some of the costs of these policies onto foreign exporters.
Thus, they make the point that a non-discrimination rule is insufficient to achieve "internationally efficient policies." In this regard, they refer to "excessively high" taxes and product standards. They argue that additional rules might be useful in this regard:
The analysis suggests as well a possible need for additional legal disciplines on nondiscriminatory regulatory and tax policies to prevent international cost-shifting and over-regulation in appropriate cases.
However, they note, WTO rules do little to address this issue:
the explicit obligations contained in GATT Article III and the SPS and TBT Agreements probably do little to address this ... issue. We then evaluate the possible role for "nonviolation" claims in cases involving nondiscriminatory regulations, which would allow exporting nations to seek compensation for reduced market access, but suggest that as currently interpreted the nonviolation doctrine is also probably ineffective in providing the needed discipline.
They elaborate on these points as follows. With non-violation, they see it as effectively limited, in the case law, to discriminatory measures, and suggest that even if it could be used to address non-discriminatory measures, this would only be appropriate in exceptional cases. With regard to the SPS and TBT Agreements, they argue that:
it is at least questionable whether the obligations to use international standards, the scientific evidence requirements, and the consistency requirements will do much in practice to address instances of economically excessive nondiscriminatory regulation. Indeed, it is possible to interpret all of these provisions as falling within the nondiscrimination framework. If a nation declines to adopt international standards when they would achieve its objectives adequately, it might be presumed to be acting strategically to disadvantage foreign suppliers, whose goods will often meet the international standard but perhaps not the domestic alternative. The obligation to use international standards is then perhaps little more than a corollary of the least restrictive means principle. Similarly, a regulation that has no scientific justification might be considered a sham, again designed to raise the costs of foreign suppliers relative to their domestic competitors.
Their focus here was on international standards, scientific evidence requirements, and consistency requirements. Let me note here that I disagree a bit with the authors on whether provisions of the SPS and TBT Agreements go beyond non-discrimination. To me, all of the provisions they mention, as well as the "necessity" provisions, do, in fact, go beyond this, although exactly how far beyond is a little hard to say. There is clearly some overlap between these provisions and the non-discrimination requirement. However, there are at least some cases (although it is not clear how many) where such obligations will catch measures that would not violate the non-discrimination obligations. For example, while a regulation not based on science might be a sham designed to raise costs for foreigners (as suggested by the authors), it could also be that the legislators/regulators believed that the science was too uncertain to follow or that science is not the only consideration. It is also possible that while such a regulation raises costs for foreigners, it raises them for domestic entities even more. As a result, it could be argued that if these kinds of rules are a proxy for discrimination, they are not a very good one.
Given their emphasis on the importance of disciplining non-discriminatory measures, it is not surprising that the authors are skeptical of a view of trade agreements that is restricted to non-discrimination:
For Regan (2006), the goal of trade agreements is to eliminate protectionism. But why do governments care about protection imposed by other governments? In our view, the answer lies in the fact that their exporters are harmed, and earn less on their export sales than otherwise. This is precisely the injury that terms-of-trade theory captures. Thus, we view the “governments do not take account of the terms of trade” objection to the terms-of-trade theory as misguided. It is not necessary for governments to literally “take account” of terms-of-trade effects in their decision making processes for the issues that the model highlights to arise. It is enough that government policy is the result of a political process in which domestic interest groups are represented and foreign interests are not. The resulting political equilibrium will then naturally select policies that ignore the harm done to foreign interest groups, and in particular the harm due to the fact that tariffs, consumption taxes, regulatory standards and the like may force foreign exporters to reduce their prices to remain competitive. As long as governments ignore such harm to foreign interests, they will tend to behave “as if” they were consciously manipulating their terms of trade.
In response to this, I offer three points. First, it is no doubt true that there are non-discriminatory policies that affect trade. So, trade concerns clearly go beyond simple protectionism. However, when trying to build a trade regime, we need to balance out the goal of achieving efficient regulatory outcomes with that of creating a sustainable and legitimate legal and political structure for trade agreements.
Second, an agreement requiring "internationally efficient" domestic regulation would be extremely broad in scope. There are likely to be arguments that a wide range of domestic regulation is inefficient. Subjecting all such regulation to international discipline would lead to a situation where trade agreements could seriously hinder the adoption of much domestic regulation.
As a third point, the authors note that "governments care about protection imposed by other governments" because "their exporters are harmed, and earn less on their export sales than otherwise." While this is no doubt true, I think it is likely that they care more about lost export sales when they are the result of discrimination against their products. Most governments can probably live with a certain amount of incidental impact on their trade due to variations in regulatory policies; they are more likely to be upset when their goods are explicitly or intentionally targeted, or if discrimination can otherwise be shown.
With regard to all of these points, the following statement by the authors later in the paper seems relevant: "An international system that second-guessed the cost-benefit determinations of national regulators would also likely intrude heavily on notions of national sovereignty and meet considerable political resistance." Arguably, restricting trade agreements to the goal of eliminating protectionism is a way to achieve the appropriate balance between efficiency and regulatory autonomy.
Finally, it does seem clear that, all else equal, it would be a good thing if we had "internationally efficient policies" and could avoid "excessively high" taxes and regulatory standards. The question I would ask is, are binding international agreements the best way to achieve this? It may be that we could achieve just as much, if not more, through a cooperative and consultative system, as this could alleviate some of the sovereignty concerns noted above.