With the release of the latest US submission in United States-Origin Marking Requirement (DS597), it is worth thinking about one of the open questions of the GATT security exceptions in today’s climate – what does it mean to take action “in time of” an emergency in international relations?
Security exceptions can account for deviation from the reciprocal commitments that are the cornerstone of the GATT. The circumstances for invoking Article XXI of the GATT are broad, ranging from the traffic of goods for supplying military establishments to responses to severe disruption to the relationship between two countries (such as in cases of ongoing military conflict). Thus far, WTO panels have assessed measures taken in time of emergencies in international relations pursuant to Article XXI:b(iii) of the GATT.
As many know, the past few years have seen governments and scholars question the breadth of circumstances for invoking security exceptions. Does the language of Article 73 of the TRIPS Agreement (equivalent to Article XXI of the GATT) cover pandemics to override intellectual property rights to facilitate production and distribution of vaccines for COVID-19? Frederick M. Abbott argued that the existence of an “emergency” of international concern equates to an “emergency in international relations”, on the basis that allocation of scarce resources and disruptions to international trade and travel ‘cross the legal threshold of an emergency in international relations.’ (Abbott 2020, p.7). I question whether these are global emergencies for international relations rather than in international relations, as the former may impact more than a single trading relationship. But Abbott’s arguments in the context of a pandemic raises a more important question: the review of emergencies (whether in or for international relations) that exist in a void. On a related point, traders may not ever see a return to pre-pandemic trade – in fact, such an emergency may never end but instead set a new “normal.”
Timing of actions matter for the enforcement of all WTO commitments. In the case of security-related measures, the sole remedy remains prospective (withdrawal/modification of WTO-inconsistent measures). Without setting temporal parameters on “war” or “emergency in international relations” there could be never-ending disruption to WTO non-discrimination norms. When the Trump administration significantly expanded the interpretation of U.S. domestic laws to protect its domestic steel and aluminium industries on the ground of economic security – it opened the possibility for broad invocation of Article XXI. However, even economic security crises must, presumably, end.
Problematically, the Russia-Traffic in Transit (DS512) panel proposed a sliding scale for interpreting “emergency in international relations” in the context of the interests at stake (para. 7.134), but there was very little analysis of the phrase “taken in time of.” The DS512 panel observed that whether security-related measures are ‘taken in time of war or other emergency in international relations […] does not necessitate a prior determination that they would be WTO-inconsistent if they had been taken in normal times’ (para. 7.108). The intention here was to distinguish the exceptions from Article XX, such that the panel did not analyze the “necessity” of the measures taken. However, by only devoting a single paragraph to consider whether Russia’s measures were “taken in time of” the emergency in international relations (para. 7.124), the panel missed a key point of this circumstance – there needs to be an expectation that a time of war or other emergency in international relations ends. The DS512 panel report acknowledged the chronological issue, but only to affirm that whether action was ‘taken during war or other emergency in international relations ‘was an ‘objective fact’ (para. 7.70). The DS512 panel unintentionally forewent an opportunity to question Russia as to when exactly the “emergency in international relations” should end.
The European Union has highlighted the importance of reviewing the timing of measures when Article XXI(b)(iii) is invoked. The European Union observed here and here that the terms “in time” ‘require[s] a sufficient nexus between the action taken by the invoking Member and the situation of war or emergency in international relations, also in temporal terms.’ The European Union further stated, quite rightly, that ‘an emergency situation cannot cover a measure that is (supposedly) taken as a response to an action that occurred in the distant past.’ Thus, not only did time matter for understanding when emergencies must end – Members need to understand that actions in time of emergency actually happened in response to that emergency (thus distinguishing protectionist from non-protectionist actions). In sum, there are significant elements of scrutiny found within the temporal features of an emergency in international relations.
In contrast, the DS512 panel assessed the remoteness of the measures to the emergency, analyzing only the plausibility that measures taken were for protecting interests ‘arising out of that emergency’ (para. 7.145). For the DS512 dispute, the on-going “general instability” of relations had eliminated the need for periodization. In essence, the DS512 panel assumed the emergency occurred within a stable block of time. In seeking to avoid commenting on the ‘necessity’ of the measure, the DS512 panel avoided not only the consideration of event chronology but, more crucially, the imminence of the emergency in international relations, which could have invited a far stricter assessment than that undertaken in the DS512 dispute.[1]
So, what to do?
Unlike waivers with annual reviews, there is nothing within the text of GATT Article XXI to suggest Members cannot simply invoke Article XXI indefinitely, without any assessment as to when discrimination slides from justified to unjustified. Setting parameters on the zone of exceptions is necessary.[2] Perhaps a working Appellate Body would have sought to address this somehow. The DS512 panel could have done more to confirm when “taken in time of” meant there had to be evidence that a measure was versus was not ‘taken in time of emergency in international relations.’ Without this step in the analysis, there is risk of perpetual reinvention of what is a “normal” time of trading.
To process the temporal element of the interface of trade and security policies we must consider what kind of information sharing can occur. If Members will stretch the GATT security exceptions across the WTO regime, then it makes sense to institute information management like the existing systems for TBT and Sanitary and Phyto-Sanitary (SPS) measures. In other words, governments should be able to address national security outside of active dispute resolution.
Trade technocrats in Geneva need to understand when different countries regard an issue as becoming and ceasing to be a security interest. Governments could then discuss how plausible security issues plug into trade, and which agencies are responsible for these interconnected issues. The alternative is now laid bare – the United States reading of Article XXI in its submissions in United States-Origin Marking Requirement (2021) reveals a path whereby any Member may unilaterally act without having to provide any evidence for determining when (and how) an emergency stops. The issue here is that governments are then left in the dark as to when trade actions – taken in time of war or other emergency in international relations – are, in fact, persistently redrawing their trade relations on the basis of security interests.
* * *
[1] As a parallel, in other fields of international law, the “imminence” of an attack grounds an assessment of State actions in the light of threat or emergency. See, D. Bethlehem, “Self Defense Against an Imminent or Actual Armed Attack by Nonstate Actors” American Journal of International Law 106:769 (2012). My thanks to Greg Messenger for highlighting this point.
[2] On zones of exception, see N.C. Lazar’s chapter, ‘What is the State of Exception?’.
