(Cross-posted from the new Kluwer Arbitration Blog).
Two weeks ago, an LCIA tribunal issued its Award on Remedies in a dispute filed by the U.S. against Canada under the 2006 Softwood Lumber Agreement (SLA). This dispute is interesting in many respects. Most obviously, it is a state-to-state dispute adjudicated under the auspices of the LCIA, more commonly used for commercial arbitration. This is in part a result of the SLA’s genesis as a “mutually agreed solution” under Article 3.6 of the WTO Dispute Settlement Understanding (DSU), attempting to settle six disputes in the WTO (and some more under NAFTA). Although the WTO context of the SLA is clear, and the parties’ notification to the WTO Dispute Settlement Body (DSB) of the mutually agreed solution included the full text of the SLA, under conventional wisdom, “mutually agreed solutions” in the WTO cannot be enforced by the WTO’s own dispute settlement system. Hence the need for outside arbitration.
Here I would like to focus on the implications of this dispute for one of the taboos of international trade law: retrospective remedies.
With all its substantive complexity, the Softwood Lumber dispute boils down to a clear-cut trade standoff. The US claims Canadian measures unfairly support Canadian industry and employs unilateral trade remedies (antidumping and countervailing duties) to block Canadian imports. Canada challenges these duties as inconsistent with trade rules. On one hand Canada is allegedly culpable of distorting bilateral trade; on the other hand the US is allegedly guilty of protectionism. For present purposes, it is immaterial which of these allegations is true. Indeed, it might even be the case that both of them are. What matters is that the SLA is an effort to resolve the ongoing dispute in a manageable way. Under the SLA, instead of Canada granting questionable benefits to its producers and the US negating them through questionable duties, the US has agreed to refrain from unilateral protection, and Canada has agreed to restrict exports to the US, in accordance with elaborate price and quantity formulas. Like any such voluntary export restraint (or ‘VER’), the SLA is not a liberalizing agreement. If the SLA were an agreement between private corporations, it would likely be a violation of antitrust laws. Yet when consummated by governments, it is kosher; but it is still an anti-competitive arrangement.
In this first of SLA disputes, the US argued that the Canada had not applied the export restrictions it had committed to, between January and July, 2007, leading to “overshipping” of lumber from Canada to the US. The award on liability upheld some of the US claims, leading to the second stage of arbitration on remedies. The central question at the remedies stages was whether Canada had to compensate for this overshipping during the period prior to the initiation of the arbitration (e.g., by asserting stricter price and quantity restrictions subsequent to the award), or whether rectification of its calculations of restrictions from July, 2007 onwards was in itself sufficient remedy.
With respect to this question, Article XIV SLA is somewhat confusing. Under Article XIV:22 SLA, if the LCIA tribunal finds a breach of the SLA, it shall identify a reasonable period of time for curing the breach, up to 30 days from the award, and determine “appropriate adjustments” to the export restrictions to compensate for the breach, if the party “fails to cure the breach within a reasonable period of time.” Under Article XIV:24, these adjustments apply from the end of the reasonable period of time until the breach has been cured. Now, by one reading of this, the respondent bears no liability for a breach prior to a finding by a tribunal that there has in fact been a breach and a remedy specified. Only from then on, must the respondent cure the breach as ordered, within a reasonable period of time specified by the tribunal. If it fails to do so, it will bear the cost, not of the original breach, but of not curing the breach as required. The second reading is that the remedy must address the damage done by the entire breach since its inception, until the end of the reasonable period of time, or until the breach has been cured and compensated for, whichever is later.
The first reading (prospective remedies only) is a WTO-like reading; the second reading (prospective and retrospective remedies) is a non-WTO-like reading. This is because, de lege lata, retroactive remedies for a violation of trade commitments are not possible in the WTO (Article 19.1 DSU).
The US has argued the non-WTO-like reading consistently, since the original Request for Arbitration of August 13, 2007, saying that remedies should be awarded for both the original period of the breach (explicitly “retrospective and prospective”) and for any additional period of breach beyond the “reasonable period of time” to be determined (para. 61 of the request). In its Statement of the Case on Remedy from May 29, 2008, the US essentially argued that to cure the breach, Canada should be ordered to enforce export restrictions additional to its ongoing SLA commitments, during the “reasonable period of time” for compliance, “with the ultimate goal of placing the United States in the position it would have occupied absent the breach” (para. 5). In short, the US was asking for a full retrospective remedy. Canada, by contrast (and not surprisingly), argued for the WTO-like approach, relying especially on the fact that it had started applying the full export restraints before the tribunal set to work. Much of Canada’s arguments in this regard relate to the similarity between the language of Article XIV SLA to the DSU, claiming that they should be interpreted similarly, etc.
In the current SLA remedies award, the tribunal was less than sympathetic to these WTO arguments, approaching the SLA as a stand-alone agreement in international law, and not as an agreement done under the WTO DSU. Starting with the ILC Draft Articles and the Chorzow Factory, it established that there is a presumption under international law in favor of retrospective remedies, and found nothing to the contrary in the SLA, referring to the DSU only for comparative purposes. Hypothetically, had this been adjudicated in the WTO, the outcome would have been diametrically the opposite, with the default being prospective remedies only, and nothing to the contrary in the SLA.
The US has taken the argument for retrospective remedies under the SLA one step further in a separate Request for Arbitration regarding other Canadian measures allegedly circumventing the SLA (currently still pending on the merits). In this second arbitration, the US has requested that the award cure the alleged breaches not only by terminating the provincial programs in dispute, but by ordering Canada to recover “the benefits conferred upon the softwood lumber sector as a result of these programs.” This would appear to be a request that, if awarded, would require Canada to extract funds from private corporations, tantamount to a repayment of a government subsidy. To WTO lawyers, this will strike a discordant note with the history of application of Article 4.7 of the WTO Subsidies and Countervailing Measures Agreement, under which a WTO Panel that has found a measure to be a prohibited subsidy “shall recommend that the subsidizing Member withdraw the subsidy without delay”. In Australia -Automotive Leather (US Recourse to Article 21.5), a WTO compliance panel found that despite Article 19.1 DSU, “withdraw[wing] the subsidy” under Article 4.7 SCM “is not limited to purely prospective action, but may encompass repayment of prohibited subsidies” (para. 6.42) and furthermore, that in some circumstances (as in the case itself), repayment might be actually necessary in order to satisfy the requirement to “withdraw” the subsidy (para. 6.48). This introduction of retrospective remedies, albeit in a narrow slice of WTO law (subsidies) was opposed both by the parties to the dispute (including the US - see para. 6.12 of the Report) as well as the EC (as third party), and was subsequently criticized by Members in the Dispute Settlement Body. Such was the abhorrence of the idea of retrospective remedies in the WTO, even in the special area of subsidies, that the US and Australia subsequently reached a mutually agreed solution covering essentially the prospective part of the subsidy; and this aspect of the Panel’s findings has generally been ignored by later panels (see here and here for discussion).
With no intention of addressing either Canada’s arguments in the case or the LCIA tribunal’s analysis (and there is a connection between the two, of course), the outcome in this remedies case is very interesting, for at least the following reasons:
First, this is another case where it appears that the same treaty text would have received very different interpretation and application in a different forum (namely, the WTO, had it held jurisdiction), even though both tribunals are in the same economic sphere.
Second, we see here a division of legal culture that is not unfamiliar from other fragmentational contexts; but in this case it is the WTO that is left out to dry by the non-WTO tribunal, rather than the other way around.
Third, if we return to the understanding that the SLA is an anti-competitive agreement, the substantive paradox that emerges is striking. Whereas trade-liberalizing commitments in the WTO (and indeed, in regional trade agreements), including subsidies, are not enforced through retrospective remedies, a trade-restrictive agreement has been enforced retrospectively. This is bad news in these economic times.
Finally, perhaps most interestingly, the SLA does work both ways; if the US were to restrict lumber imports, the logic of the LCIA tribunal would mean that Canada could apply retrospective remedies reducing export restraints. So, despite WTO law’s resistance, aren’t retrospective trade remedies possible after all?
