Weihuan Zhou and Victor Crochet have a new working paper called "Attribution to the State: A Critique of Cross-Border Subsidies," which combines two of my favorite issues: Cross-border subsidies and non-violation complaints. Here's the abstract:
While industrial policy and subsidies proliferate worldwide, the emergence of cross-border subsidies is intensifying legal and policy debate on how the rules of the World Trade Organization (WTO) can be applied to address the spillover effects of such subsidies. Amid this debate, this paper tackles a relatively more fundamental, yet under-explored, issue relating to attribution of state responsibility. We argue that the European Commission’s approach to attributing subsidies provided by one government to another government has overstretched the attribution rules under public international law and the WTO subsidy rules, setting a harmful precedent for similar abuses beyond subsidies. We propose non-violation claims as a viable and more balanced response to the challenges posed by cross-border subsidies. We call upon governments to retreat from unilateralism and resolve these challenges via cooperation.
After arguing that the EU approach to cross-border subsidies is not permitted, they say that non-violation claims can be useful here:
2. Developing non-violation claims
A more, and perhaps the only, sensible option is to bring non-violation claims (NVCs) pursuant to Article XXIII:1(b) of the GATT. This provision came out of the original GATT negotiations in the 1940s. It was applied and clarified by several GATT panels and was further negotiated in the Uruguay Round before it was also codified in Article 26.1 of the WTO’s Dispute Settlement Understanding (DSU).72 Article XXIII:1(b) and Article 26.1 maintain the avenue for governments to challenge measures which are not in breach of any WTO rules but which nullify or impair benefits under, or impede the attainment of objectives of the relevant covered agreement(s). It is clear from the historical development of these provisions that governments saw the importance of NVCs in addressing the issue of an incomplete GATT/WTO contract where the outcomes of tariff negotiations may be undermined by acts which were not foreseen at the time of the negotiations and hence were not captured by the GATT/WTO rulebook.73 At the same time, however, governments have resorted to NVCs with caution and generally accepted the interpretative approach taken by GATT/WTO panels to confine the application of NVCs to limited circumstances. Thus, governments brought and won NVCs only in a few disputes throughout the life of the GATT/WTO. These disputes were predominantly concerned about action taken by an importing Member subsequent to the negotiation of its tariff concessions with an exporting Member leading to the nullification or impairment of benefits (i.e. market access opportunities) derived from the concessions for the latter. 74 In Japan – Films (1998), the WTO panel summarised the legal conditions for an NVC: (1) the existence of a governmental measure, (2) a benefit accruing to the complaining Member from the relevant tariff concession or agreement, and (3) the benefit is nullified or impaired as the result of the application of the measure. 75 The panel also iterated that “the non-violation remedy should be approached with caution and should remain an exceptional remedy.”76
While GATT/WTO panels have developed restrictive interpretation of NVCs,77 we argue that their interpretation still provides room for establishing an NVC against cross-border subsidies. At a general level, there are at least three reasons for that optimistic observation. First, the legal conditions for NVCs remain unsettled. In the Uruguay Round negotiations, governments deliberately left out the conditions developed by GATT panels from the text of Article 26.1 of the DSU,78 arguably leaving the flexibility for WTO tribunals to further develop or refine these conditions. In a handful of disputes where NVCs were raised, WTO panels seem to have maintained GATT panels’ restrictive approach. 79 However, WTO panels also recognised that NVCs must be assessed on a case-by-case basis and can be applied to benefits unrelated to a tariff concession but stemmed from a WTO agreement.80 Second, a majority of the past NVCs, including the few successful ones, involved subsidy schemes, hence providing a solid precedent for applying NVCs to cross-border subsidies which as shown in Section III, are typical forms of financial contributions contemplated in the ASCM. Third, in adjudicating NVCs, GATT/WTO panels have focused on assessing whether a measure frustrated the competitive opportunities generated by tariff concessions rather than the impact on actual trade flows.81 This means that an NVC may be brought even before cross-border subsidies have caused an increase of subsidised imports into adversely affected countries.
Thus, more specifically, affected Members like the EU can argue that the cross-border subsidies provided by the Chinese government to its investors overseas have the effect of upsetting the competitive opportunity that EU producers should have enjoyed in their home market. The benefit, which must be interpreted broadly,82 comes out of the EU’s right to seek multilateral remedies or impose countervailing measures pursuant to GATT Article VI, as elaborated by the ASCM. Two complications can arise in this claim of “benefit”. One pertains to the fact that the benefit is related to the EU’s own tariff concessions. This presents an atypical situation compared to past NVCs, which consistently involved a tariff concession of a responding Member that created more favourable market access for the relevant exports of a complaining Member.83 The other complication involves the EU’s evidentiary burden to show that the Chinese subsidies were not reasonably anticipated at the time when the concessions were made,84 essentially during China’s WTO accession negotiations.
These hurdles, however, are not so difficult to overcome. A plausible argument would be that when the EU offered reciprocal tariff concessions, there was a legitimate expectation that it would have access to WTO-permissible remedies to offset the impact of subsidised imports to maintain certain degree of market opportunities for domestic industries. That China now provides harmful subsidies not captured by GATT Article VI or the ASCM nullifies or impairs the benefit that the EU should have enjoyed via these remedies. Here, it should be noted that the “adverse effects” claims under Article 5 of the ASCM would not apply as it requires the existence of a subsidy within the meaning of the ASCM in the first place.85
On the question of whether the subsidies could have been reasonably anticipated during China’s accession negotiations, the Working Party Report on the Accession of China (WPR) suggests that Members were unaware of issues related to cross-border subsidies. The discussions recorded in the WPR focussed on how the relevant economic policies, regulations, conduct of state entities, industrial subsidies may affect imports or investments into China, or how subsidised exports from China may impact trading partners.86 It follows that Chinese cross-border subsidies, which only became prominent after the launch of the BRI twelve years after China’s WTO accession, were not reasonably foreseeable by WTO Members.
Normally, I'm a non-violation skeptic, often on the basis that whatever measures were taken could have been anticipated. But here, as the authors point out, cross-border subsidies were not well known at the time of the Chinese accession negotiations, so it is plausible for a WTO Member to argue that it did not anticipate the Chinese government taking such measures in the way it did. Fleshing out the history of cross-border subsidies, in China and elsewhere, would be of great important to this claim.
Zhou and Crochet also make two other points that I want to highlight:
Not only is an NVC against cross-border subsidies viable, but it also provides a more balanced approach to tackle such subsidies for three reasons. First, a successful NVC does not require the removal of the contested measure but merely requires the disputing parties to reach a mutually satisfactory adjustment (Article 26.1(b) of the DSU). This provides the flexibility for the subsidising government to maintain the subsidies in pursuit of its policy goals and for the affected government to take action to rebalance the relevant rights and obligations under the WTO framework. Such action may involve withdrawal of tariff concessions (which would lead to an outcome similar to what a countervailing measure can achieve), compensation, or even cooperation between the parties in other areas of trade and investment. NVCs therefore provide a better avenue for the subsidising and aggrieved governments to negotiate a compromise that would work for both without causing a decision on whether the subsidies are WTO-unlawful.
... Third, compared to a violation claim, an NVC can be adjudicated more efficiently. As some have proposed in the context of using NVCs to address security-based measures, a “shortened and simplified dispute settlement procedure” can be developed to determine “nullification or impairment of benefits”, and the actual level of benefits can be agreed upon by the parties or decided by an arbitrator under Article 21.3(c) of the DSU.87 An NVC therefore can facilitate the settlement of disputes and the provision of remedies for the affected Member in a more timely fashion.
As to the resolution of successful non-violation complaints, we may not all agree on how DSU Article 26.1 is likely to work in practice, but I suspect that the first such complaint is going to lead to a good deal of confusion and controversy over how exactly things are supposed to proceed after a finding of nullification or impairment.
Relatedly, on their efficiency of adjudication point, maybe I am still scarred from my experience on the Japan - Film case, but I have doubts about whether non-violation complaints can be made significantly shorter and simpler than violation complaints. Among other factors here is that the responding party is unlikely to agree to something shorter and simpler. For obvious reasons, responding parties tend to try to drag the proceedings out, and I'm not sure why this would be any different in the non-violation context.
