Justiciability of Security Exceptions in the US Steel (and other) Disputes: Some Middle-Ground Options and the Requirements of Article XXI lit. b (i)-(iii)

The US – Certain Measures on Steel and Aluminium Products case (US Steel Dispute) has aroused numerous comments in the blogosphere (see e.g. here, here, here, here, here and here) which already give a very good impression of the legal questions involved and of what is at stake at the WTO these days. One of the most controversial legal issues brought up by the case (and by two other recent cases: Russia – Measures Concerning Traffic in Transit (Russia – Transit) and United Arab Emirates – Measures Relating to Trade in Goods and Services, and Trade-Related Aspects of IP Rights; for comments see: here and here) is the justiciability of Article XXI GATT (security exceptions). The question of justiciability, however, has sometimes been portrayed as an either/or question by bloggers: Either justiciability or complete discretion for States. Moreover, commentators have scarcely elaborated on the further requirements of Article XXI para. b (i)-(iii) GATT with regard to the US steel dispute.

The following post shows that there are more options on the table than to allow States full discretion (option 1), or declaring security exceptions justiciable under a limited good faith standard of review (option 2), and that under all but the first option Panels are likely to declare Trump’s tariffs on steel and aluminium imports not to be covered by security exceptions. Still, finding some middle-ground position on justiciability could be useful (politically) to avoid the impression of judicial overreach.

Judicial Review: Several Options

Article XXI GATT (in the case at hand lit. b) seems to allow a Member State to self-judge what it “considers necessary for the protection of its essential security interests”. Security exceptions have scarcely been used in the GATT and earlier WTO era, and unfortunately the meaning of “considers necessary” in Article XXI GATT so far has not been authoritatively elucidated by a Panel or the Appellate Body (AB).

As has been highlighted by previous posts, Panels tasked today with interpreting Article XXI GATT are in a difficult situation. Judicial review that is too deferential could open the floodgates for further States “illegitimately” invoking security exceptions, thereby eroding the rule of law in international trade. Declaring the invocation of security exceptions judiciable and holding States accountable would uphold the rule of law formally but would also reinforce the US argument that Panels and the AB are overreaching, possibly leading to non-compliance by the US in the case at hand and a further weakening of the WTO’s Dispute Settlement Mechanism.

Although I generally agree with most commentators, the ICJ in the Case Concerning Certain Questions of Mutual Assistance in Criminal Matters (Djibouti v. France) (para. 145), and arbitrators in EC – Bananas III (Article 22.6) (para. 52), to name but a few, that so-called self-judging clauses are subject to limited (good faith) justiciability, I also think that Panels could adopt other middle ground positions in the context of Article XXI GATT. For example, a panel could find the travaux préparatoires argument by the US (elaborated in its third party submission in the Russia – Transit case; interestingly siding with Russia) against justiciability convincing with regard to the first part of lit. b, but still declare some parts of Article XXI lit. b, namely sub-paragraphs (i)-(iii), judiciable. Indeed the Panel in Russia – Transit appears to be aware of that possibility (see question 3 by the Panel here; the US for obvious reasons rejects this interpretation). States may for example enjoy unfettered discretion in determining what they consider “necessary” to protect essential security interests but not what objectively can be considered an “essential security interest” (option 3). Following this approach, as highlighted by the EU in its third party submission in Russia –Transit, Panels could also declare all the other requirements of lit. b (i)-(iii) e.g. “in times of war or other emergency” fully justiciable. A more limited approach (option 4) would be to allow States wide discretion (only subject to a good faith standard of review) in determining their essential security interests and in assessing what is necessary to protect these interests, but fully control the further requirements in sub-paragraph (i)-(iii).

Do Trump’s Tariffs Meet the Requirements of Article XXI lit. b (i)-(iii) GATT? 

The US so far has only generally referred to Article XXI GATT and national security and omitted justifying its measures under the further requirements of lit. b (i)-(iii) (see e.g. here). The US argument in that regard is rather simple: Steel and aluminium are needed in the defence industry and for the maintenance of US critical infrastructure. Cheap imports threaten US domestic production of these products. Therefore, protecting domestic producers is necessary to uphold national security (see in more detail the report of the US Department of Commerce).

Most scholars appear to think that this is a silly argument. Joseph Weiler has called the argument a black lie, Simon Lester and Steve Charnovitz have called it a sham, and I suppose a panel would basically agree with them if applying option 2 (good faith standard of review). Even US Defense Secretary James Mattis did not fully buy the argument and highlighted his department’s ability to meet national defence requirements with available steel and aluminium from US producers. According to him, the “U.S. military requirements for steel and aluminium each only represent about three percent of U.S. production”. Moreover, most imports of steel and aluminium come from traditional allies of the US (see here) and are unlikely to become unavailable in the near future. Therefore, commercial reasons rather than purely security related ones appear to be the driver of US steel and aluminium tariffs.

Albeit Commerce Secretary Wilbur Ross expressed the opinion that “national security is broadly defined to include the economy, to include the impact on employment, to include a very big variety of things”, this is obviously not the understanding of essential security interests in Article XXI GATT. Otherwise, a whole number of other exceptions (most notably Article XIX) would be superfluous and the rule of law in international trade meaningless. The argumentum ad absurdum proves that point. GATT would be robbed of all legal meaning and capacity to uphold a liberal trading regime, if essential security interests would include commercial concerns. The controversial invocation of “the spirit of Article XXI” by Sweden in 1975 to protect its footwear industry, as well as the US argument that imports of automobiles impair US national security, are cases in point. The wording of lit. b requires an essential security interest, not any commercial interest somehow linked to security (soldiers need US cars to drive to work will not do it). Interestingly, US representatives had also rejected the invocation of commercial interests as “essential security interests” in the early stages of GATTs drafting history (see here).

But what if a panel found the US assertion that economic security is an “essential security interest” to be unreviewable? Under most options presented above the further requirements of lit. b are justiciable. Moreover, in my opinion it appears hardly convincing to apply the good faith standard of review to those rather detailed requirements in subparagraphs (i)-(iii) which, in addition, are textually not connected to the wording of “considers” (see also: Hestermeyer, Article XXI, in: Wolfrum/Stoll/Hestermeyer). These further requirements are unlikely to be met by US measures in question for the following reasons.

US actions on steel and aluminium do not relate to products in (i) “fissionable materials or the materials from which they are derived” or (ii) “arms, ammunition and implements of war”. “[O]ther goods and materials” can only be made subject to restrictions if “carried on directly or indirectly for the purpose of supplying a military establishment”. While this is a broad exception (potentially covering numerous goods from toothbrushes to tanks, including metals) I wonder whether it was really conceived of by its drafters to allow for anything else than export restraints on named goods (see in that direction also the examples provided by the Analytical Index of GATT, p. 602). So far I found no clear opinion in scholarship on the issue but it appears to me rather clear that the whole sub-paragraph (ii) is aimed primarily at allowing States to restrict exports of such goods that could be used against them (or allies) in military conflicts. In the case at hand, it is the other way round. Only, under some cases of “modern” security threats such as terrorism one may consider import restrictions on “other goods and materials” justified by subparagraph (ii), but only if they aim at curtailing the support for “military establishments” of opposing terrorists/troops (e.g. on own territory), which is clearly not the case for US steel and aluminium imports. In any case, indicated by the statement by James Mattis above, only a small proportion of US steel and aluminium imports are used in the defence industry.

Accordingly, measures could only fall under sub-paragraph (iii). This would require the US to be in “a time of war or other emergency in international relations”. Although war is to some extend a rather vague term (it could for example also mean a “trade war”) it appears rather clear that the clause only addresses armed conflicts. The wording “other” indicates that the “other emergency” must be of similar severity than outright war but potentially covers situations beyond military conflicts, such as natural or economic crisis. However, the wording “emergency” further indicates that there must be some form of suddenness or unexpectedness. The US has made reference to possible national emergencies in the future, in which US steel producers must be capable of meeting additional demands, but as far as I can see has not argued to be in a time of “other emergency in international relations” or even “war” related to imposed measures right now. In particular, it would be circular to argue that a trade war started with reference to Article XXI is an “other emergency” under the same article. Systematically, serious injury to domestic producers of steel and aluminium caused by imports of particular products is no emergency in the sense of Article XXI but under Article XIX (Emergency Action on Imports of Particular Products; see also: Balan, p. 15). Whether Trump’s tariffs would be legal under Article XIX (if the clause had been invoked), and whether the EU in imposing countermeasures has rightfully treated US measures as safeguards (a white lie according to Weiler), is another story hopefully to be explored by a WTO Panel soon.