International Economic Law and Policy Blog

Is TTIP-Light Enough?

At the outset of the TTIP talks, people were talking up regulatory issues as the most important part.  But perhaps it was just too much to deal with, at least in such a short period of time.  Can everyone accept a "TTIP Light"?  This is from Inside US Trade:

As the end of 2015 approaches, U.S. and EU trade officials are entering a new year without the "outline" of the Transatlantic Trade and Investment Partnership (TTIP) that political leaders had expressed hope for last summer. But what is becoming clear to both sides is that if a deal is to come together under the Obama administration, it will fall far short of the sweeping U.S.-EU regulatory alignment project that it was initially framed to be.

The plan of action for 2016, as laid out in private conversations by U.S. officials, is not focused on the regulatory cooperation side of the agenda. Instead, it envisions an exchange of modest government procurement market access offers by February. By summer of the new year, the aim is to be in the "middle game" of the negotiations, during which both sides envision having all the text of the deal essentially agreed and further advancement on sensitive areas. 

...

Progress in the TTIP's so-called "sectoral" regulatory talks -- those dealing with the weedy differences in how each side regulates cars, chemicals, pharmaceuticals and other industrial sectors -- has been slow. While there are some hints of progress on cars, sources say, regulators particularly on the U.S. side have not gotten the political direction to identify and pursue specific outcomes. As a result, many agree that any sectoral regulatory outcomes will be small if a TTIP deal is wrapped up in 2016.

Regulatory issues have been under discussion for many years prior to the TTIP. Was it unrealistic to expect much success in the TTIP?

And what about the future: Are there any U.S. Presidential candidates who might be interested in pursuing this agenda?

Will the EU Grant China Market Economy Status?

This is from the FT:

Washington has warned Brussels against granting China ‘market economy status’, saying the long-sought trade concession could hamper efforts to prevent Chinese companies flooding US and European markets with unfairly cheap goods.

... 

US officials have warned EU counterparts that granting Beijing MES would amount to “unilaterally disarming” Europe’s trade defences against China. In private, they are scathing about the move, which they see as the latest example of Europeans seeking to curry favour with Beijing to gain billions of euros in investment.

But the European Commission, the EU’s executive arm, is growing increasingly sympathetic to China’s pleas. The commission is expected to make its decision as early as February. 

...

The commission declined to comment on its decision, which officials say is still under consideration. However, diplomats and businessmen involved in the process say momentum is building for a positive decision on MES in the first quarter of 2016.

“Brussels’ attitude towards China is better than Washington’s,” said Tu Xinquan, a trade expert at the University of International Business and Economics in Beijing. “Of course, EU industries won’t welcome market economy status for China. But I think it’s quite likely to be granted.”

Under WTO rules, China’s non-market economy status gives the US and EU far greater latitude in determining the fair cost of production for Chinese companies when conducting anti-dumping investigations. As a consequence, say trade lawyers, that has made it easier for complainants to prove their cases against China.

The Obama administration, which is being cheered on by US industry, is advocating a policy of inaction, which would force China to bring a challenge in the WTO and thus put the onus on Beijing to prove that its state-heavy economic model has met all the criteria for MES.

The difference of opinion between the U.S. and EU is interesting. If we do up with litigation against the U.S., will the decision of other countries to treat China as a market economy be used as evidence that the U.S. is acting unreasonably?

How High Were Tariffs Before the GATT?

In various papers I've written, I'm sure I have made the point that prior to the GATT, world tariffs were around 40%, and the GATT/WTO brought them down from there to much lower levels.  I suspect others have done the same.

But Chad Bown and Doug Irwin have a new paper telling us that figure is wrong:

This paper has attempted to shed light on the height of tariff barriers on the eve of the 1947 Geneva conference that established the GATT and negotiated the first postwar tariff reductions. We find that the average tariff among the key GATT participants – the United States, Western Europe, and Japan – was about 22 percent at the time of the first Geneva conference, significantly lower than the oft-reported 40 percent figure (World Bank, 1987). This figure refers to the unweighted tariff average; the import-weighted tariff average would be much lower than this. ...

A shorter version of the argument, over at vox, is here.

One interesting caveat they offer (in Table 1) is the following: "Data not available for 1947 GATT Contracting Parties Burma (Myanmar), Ceylon (Sri Lanka), China, Cuba, Czechoslovakia, Lebanon, Luxembourg, Pakistan, South Africa, Southern Rhodesia (Zimbabwe), and Syria."  Perhaps, then, there is more refinement to be done, but for now, I'm going with the 22 percent figure.

Philip Morris Loses its ISDS Plain Packaging Case against Australia

Investment Arbitration Reporter has the scoop:

A tribunal at the Permanent Court of Arbitration has issued a decision dismissing a high-profile case between cigarette multinational Philip Morris and the Commonwealth of Australia.

The PCA website indicates that a Decision on Jurisdiction and Admissibility was issued on December 17, 2015. A source involved in the case has confirmed to IAReporter that the tribunal has dismissed the investor’s claims of breach of the Australia-Hong Kong bilateral investment treaty.

...

The tribunal’s detailed reasoning for rejecting jurisdiction is not yet known. The award will not become public until the parties agree on the redaction of any confidential information contained in the award.

Jurisdictional grounds is a little frustrating, because I was hoping for some substance to talk about. How would an early 1990s version of an investment treaty apply to plain packaging laws? We may never know!

And we can't even talk about the jurisdictional ruling, because it's not public yet.  (I'm imagining this outraged blog post by a business group: "Secret international tribunal rules against job creator!").

The Philip Morris press release is here.

Appeals of Article 22.6 Arbitration Decisions

Last week, in a post about a systemic issue involving the calculation of the level of nullification or impairment in an Article 22.6 proceeding, Joost said:  "as systemic as this finding is, 22.6 arbitration reports cannot be appealed. Is this, however, not the kind of finding one would want the AB to weigh in on?"  In the comments, and kind of half-jokingly, I asked whether the DSU text could perhaps be manipulated interpreted creatively to support an argument that appeals are, in fact, permitted for these arbitrations.  Joost shot me down by pointing to Article 22.7's reference to "final" decisions, etc., and I conceded there was probably nowhere to go with this. 

But maybe I gave up too quickly. Here is the EU talking about a different Article 22.6 interpretive issue in the same case, related to whether requests for arbitration on the level of suspension are automatically referred to arbitration without DSB action, or whether the DSB needs to act on the request:

The European Union disagrees with the substantive analysis and conclusion of the Arbitration Panel on this point for the reasons set out in our prior communication. We note that it has not had the benefit of appellate review and we doubt that will occur in this particular case. We further take note of the fact that the Arbitration Panel bases its conclusion, in this particular case, on the absence of any disagreement between the parties. At the same time it recognises that this "is a contentious issue among Members"; that Article 22.6 "does not provide clear guidance"; and that "a resolution of this issue by Members would be desirable".

Pending such resolution, either by Members or by the Appellate Body, and given the substantial legal uncertainty thus persisting, for its part the European Union intends to continue the practice used in the past in almost all cases. ...

I don't want to get bogged down in the substance of the issue (about which I have no strong feelings).  Rather, the point here is that the EU thinks the Arbitrator got the answer wrong, and implies that the Appellate Body could, in theory, have a say ("it has not had the benefit of appellate review and we doubt that will occur in this particular case"; "Pending such resolution, either by Members or by the Appellate Body").  And note how the EU keeps referring to the "Arbitration Panel," presumably to emphasize that the Arbitrator is really just the original panel (and panel reports can be appealed).

So, does the EU think that Article 22.6 decisions could be appealed?  Would it pursue this thought in an Article 22.6 proceeding to which it was a party?

I Keep Hearing that the WTO Negotiating Function is Broken ...

... but then I see this:

More than 50 members of the World Trade Organisation on Wednesday concluded the biggest tariff-reduction deal in almost two decades, eliminating restrictions on the $1.3tn trade of 201 IT products.

Wednesday’s expansion of the 1996 Information Technology Agreement will increase global gross domestic product by $190bn a year, according to IT experts, as the costs are cut in trading goods from GPS devices and video game consoles to next-generation semiconductors.

...

Barriers will be eliminated in three stages under the expanded ITA. 65 per cent of the new tariff lines representing 88 per cent of imports will be cut immediately on implementation of the deal in July 2016.

Three years later, additional cuts will mean 89 per cent of tariff lines representing 95 per cent of imports will have been cut. The remaining tariffs will be eliminated by January 2024.

...

The 53 countries participating account for 90 per cent of the trade in IT products. All WTO countries will be able to benefit from the deal, however, because the cuts will be included in the ITA participants’ WTO so-called schedules of concessions.

People probably have varying expectations and hopes for what the WTO can and should accomplish, but to me, MFN-based tariff liberalization of this kind is a pretty big success.

 

Justice Florentino Feliciano

I was saddened to hear of the passing of Justice Florentino Feliciano, one of the members of the original WTO Appellate and a distinguished international jurist.  Toy brought his brilliant and razor-sharp mind to whatever task he had and was at his best in harmonizing differing bodies of international law as he did in early WTO caselaw as presiding appellator in the Shrimp and Gasoline cases.  With an early book in 1961 on "Law and Minimum World Public Order" coauthored with Yale Law Professor Myers McDougal, Justice Feliciano, over the next several decades, did as much or more than any international jurist to promote a world of optimal public order including sustainable development. A collection of essays in his honor was published in 2005 by Cambridge University Press under the title "Law in the Service of Human Dignity: Essays in Honour of Florentino Feliciano" (edited by Steve Charnovitz, Debra P. Steger, and Peter Van den Bossche). The essay by Professor Michael Reisman, titled "A judge's judge" captures the respect for Justice Feliciano shared throughout the international legal community.  The photo below taken in April 2009 at Yale Law School shows Justice Feliciano with Carolyn Lamm and Steve Charnovitz.

 

 

A

Senator Cruz on Promoting Imports of Medical Products That Have Been Approved Abroad

Here's something from an op-ed by Senator, and Presidential hopeful, Ted Cruz, making the case for legislation he is co-sponsoring:

... as scientists and investors abandon the U.S., the FDA continues to prevent life-saving drugs from coming into the country. Take, for example, the multiple-sclerosis drug Lemtrada, which has been approved and used in Europe, Canada, Mexico, and Australia. Research has demonstrated that patients who are prescribed Lemtrada are only half as likely to relapse as those prescribed a leading drug in the U.S., but the FDA has refused to approve Lemtrada because the trials did not fit within the agency’s placebo testing requirements. Patients are getting sicker, and the FDA’s outmoded scheme is keeping them that way. 

The bill I am introducing takes the first step to reverse this trend. It provides for reciprocal drug approval, so that cures and medical devices that are already approved in other countries can more expeditiously come to the U.S. It requires the FDA to respond to the treatment sponsor within 30 days of its application. ...

This sounds like one of those common sense, no-brainer, clearly beneficial ideas that will be opposed by some interest group and go nowhere.  Or perhaps someone will make the argument that we can't liberalize unilaterally like this, or else we'll lose negotiating leverage on regulatory issues in the TTIP.

But we'll see.

More details here.

Trump-State Dispute Settlement

This is from the Washington Post, in the context of an article about how Donald Trump's recent statements are affecting his business activities:

In the longer term, however, Trump could face problems. There have already been several worrisome harbingers for the Republican presidential candidate. This week, a major retailer in the Middle East cut ties with him. Canadian politicians have warned developers they could lose business if they don’t terminate a real estate deal with the mogul.

I'm particularly interested in that last sentence.  I've tried to track down more details, but haven't been able to find anything.  Is Trump's real estate deal an "investment" in Canada?  Seems likely.  Do politicians' warnings not to do business with a particular foreign investor meet the "fair and equitable" treatment standard?  I'm not sure they do.

So does Canada need worry about a NAFTA Chapter 11 lawsuit?  I guess that depends in part on how litigious Trump is.

The Paris Climate Change Draft Treaty: Where Are the Dispute Settlement Provisions?

The global climate change conference in Paris has adopted the draft of an international treaty aimed at transforming the world's fossil fuel-driven economy within decades and slowing the pace of global warming to well below 2 degrees Celsius. Many consider this text as “historic” and  according to Mr Fabius, French Minister of Foreign Affairs, the text is "legally binding".

As someone familiar with two previous historic multilateral treaties, i.e. the Law of the Sea treaty (UNCLOS) and the WTO Agreement, I immediately took a look at the Paris Draft Treaty to see if, similarly to the previous mentioned ones, it includes dispute settlement provisions giving teeth to this treaty. Unfortunately, I did not find such provisions. This is definitely a Soft Law accord despite that it would be adopted formally as a treaty. The central compliance mechanism is described in Article 15:

A mechanism to facilitate implementation of and promote compliance with the provisions of this Agreement is hereby established. The mechanism… shall consist of a committee that shall be expert-based and facilitative in nature and function in a manner that is transparent, non-adversarial and non-punitive. The committee shall pay particular attention to the respective national capabilities and circumstances of Parties. 

In fact, the only way to pressure parties is an obligation on them (under Article 13.7) to provide the following information: (a) A national inventory report of anthropogenic emissions by sources and removals by sinks of greenhouse gases (b) Information necessary to track progress made in implementing and achieving its nationally determined contribution.This Information will undergo a “technical expert review”:

The technical expert review under this paragraph shall consist of a consideration of the Party’s support provided, as relevant, and its implementation and achievement of its nationally determined contribution. The review shall also identify areas of improvement for the Party, and include a review of the consistency of the information with the modalities, procedures and guidelines referred to in paragraph 13 of this Article, taking into account the flexibility accorded to the Party under paragraph 2 of this Article. The review shall pay particular attention to the respective national capabilities and circumstances of developing country Parties

People specializing in International Environmental Law are familiar with this kind of soft law taking the form of a  treaty. Should we be concerned, considering the stakes, or is it the usual way things work in International Environmental Law?

Cool Arbitrator Finds That Only Lost Trade Is “Recoverable” Not Broader Economic Harm Suffered By Canada/Mexico

The December 7 retaliation decision in US - COOL includes a finding of great systemic importance to the trading system. When a country is the victim of a WTO violation, what “damages” are “recoverable”?

We all know that the WTO never awards cash. The closest it gets is to put a number on the “nullification or impairment” caused by the violating measure, here the US country-of-origin label (COOL) on meat. This number is then used as the maximum amount of retaliation the victim, here Canada and Mexico, can impose.

In this case, Canada and Mexico wanted to retaliate not only for the export revenue losses they suffered because of COOL (e.g. lost exports / lower export prices of Mexican/Canadian livestock sold in the US). They also wanted to include revenue losses suffered by Mexican/Canadian ranchers at home (for Canada alone, calculated at over 1 billion$): because of COOL restrictions, prices in Mexico/Canada fell, making ranchers worse off (less exports to the US because of COOL meant more domestic supply, hence lower domestic prices).

In the end, the question was whether “nullification or impairment” in WTO lingo is limited to “lost trade”, or whether it covers more broadly the “economic harm” or all “adverse effects” suffered because of a WTO violation. Put differently, are the “benefits” that Mexico and Canada can legitimately expect from a US commitment not to discriminate (TBT national treatment, the WTO violation found in this case) limited to the trade volumes that come with non-discrimination or do they include the “benefit” of not suffering any adverse effects as a result of discrimination?

In earlier 22.6 arbitration cases this question had so far been avoided: when victims claimed broader economic losses, arbitrators kicked those out as speculative or for lack of causal effect; they never defined what kind of “benefits” the WTO protects from nullification or impairment in the first place.

The Arbitrator in COOL found that “benefit”, at least “benefit” linked to TBT national treatment, is limited to trade losses; it does not include other adverse effects such as lost revenues on domestic sales (note, however, that DSU Art. 3.8 equates “nullification or impairment” with “adverse impact” more broadly so the Arbitrator's finding is not completely obvious).

This has major consequences for what the WTO and trade agreements more broadly are about.

First, the general rule for beach of international law – including investment treaties -- is that the “responsible State … makes full reparation for the injury caused by the internationally wrongful act” (Article 31 ASR). “All the consequences” of an illegal act must be “wiped out” so that the victim is put in a situation “which would, in all probability, have existed if that act had not been committed” (Chorzow Factory). Also “[c]ountermeasures must be commensurate with the injury suffered” (Article 51 ASR).

In WTO cases like COOL, however, the only injury considered is trade loss. The WTO model is not one of compensation of harm, but tit-for-tat trade concessions and suspension. In investor-state disputes, in contrast, all losses are covered. In Cargill v. Mexico, for example, Mexico (found to violate national treatment under NAFTA Chapter 11) had to pay not only for investment returns lost on Cargill de Mexico’s distribution center (no HFCS was produced within Mexico), but also for sales/trade losses suffered “up-stream” by Cargill USA from where the HFCS was exported. In response, TPP Article 9.28.2 provides that an investor “may recover only for loss or damage that it has incurred in its capacity as an investor of a Party”. But it is unclear that this will avoid situations like Cargill, which came to its broad damages conclusion based on an equally broad definition of investment as including “business income … associated with a physical asset” (the distribution center) in Mexico (para. 522).    

Second, if WTO retaliation covers only trade loss, not all harm suffered by the victim, it is hard to see how continuing a WTO violation but “paying” retaliation for it, amounts to, as some have argued, “efficient breach”. For breach to be “efficient”, the victim must be fully “compensated”, for all harm, not just some harm (e.g. lost export revenues). In this sense, WTO violation-cum-retaliation is “under-compensating”. It tolerates breach where it is not “efficient”. In TPP and an increasing number of FTAs, the amount of retaliation is also used to set the amount of damages that the violator can, as an alternative to retaliation, pay in cash (“monetary assessment”), namely 50% of the amount of retaliation. Also in the WTO, the amount of retaliation set by arbitration is increasingly used to determine compensation payments (think of US – Cotton). So how one thinks about retaliation indirectly determines also levels of compensation.

Third, strangely enough there is one set of WTO “violations” that do open the door to “full compensation”, namely: actionable subsidy cases where, as found in the US – Cotton retaliation, retaliation does include all adverse effects caused by a subsidy (e.g. reduced sales value and production effects suffered by Brazilian cotton farmers, including on the home market, as a result of US subsidies lowering the word price for cotton by 9.38%). So a subsidy that is not even prohibited or a violation, can offer more retaliation rights than a discriminatory regulation or prohibited subsidy …

Finally, as systemic as this finding is, 22.6 arbitration reports cannot be appealed. Is this, however, not the kind of finding one would want the AB to weigh in on?

The COOL Retaliation Decision

Recall that in the U.S. - COOL arbitration proceeding over the amounts of suspension of concessions:

Canada asked for $2.41 billion in retaliation; the U.S. has countered with $43.22 million. ... Mexico wants $713 million; the U.S. countered with $47.55 million

The Arbitrator's decision is out today, and the amounts are, not surprisingly, in between what the parties were arguing for:

6.81. Based on our calculation, the level of nullification or impairment for Canada amounts to CAD 1,054.729 million annually.

6.82. Based on our calculation, the level of nullification or impairment for Mexico amounts to USD 227.758 million annually.

No doubt there is some interesting reasoning to get to in the decision, but turning to more practical concerns, what happens now?  Does the U.S. repeal the measure completely and come into compliance?  Does it pass a replacement measure, using "voluntary" labelling, that some believe will bring the U.S. into compliance, but others think will lead to continued non-compliance?  Does it do nothing and accept the retaliation? 

Burden of Proof for Less Favorable Treatment under TBT Agreement Article 2.1

This is from the recent AB 21.5 report in Tuna:

7.32 ... Under Article 2.1, ... a complainant must show that, under the technical regulation at issue, the treatment accorded to imported products is less favourable than that accorded to like domestic products or like products originating in any other country. Provided that it has shown detrimental impact, a complainant may, therefore, make a prima facie showing of less favourable treatment by, for example, adducing evidence and arguments showing that the measure is not even-handed, which would suggest that the measure is inconsistent with Article 2.1.  If, however, the respondent shows that the detrimental impact on imported products stems exclusively from a legitimate regulatory distinction, it follows that the challenged measure is not inconsistent with Article 2.1.

I've very puzzled by this.

Let's just assume that detrimental impact has been shown.  What happens next?  According to the AB, the complainant "may" show that less favorable treatment exists by proving the measure is "not even-handed."  In addition, if the respondent can show that any detrimental impact "stems exclusively from a legitimate regulatory distinction," there is no violation.

So who has the burden here?  Does the complainant have the burden to show the measure is not even-handed?  If so, why did the AB say "may"?

Putting that aside, how does the "even-handed" test relate to the "stems exclusively from a legitimate regulatory distinction" test?  Are these two completely separate standards?  After detrimental impact has been shown, who has the burden to show what exactly?

More on Free Trade in Canadian Beer

What is it with Canadian provincial restrictions on internal trade in beer? I blogged about one issue here. Now there's this:

On Tuesday, the Alberta budget included an overall five per cent liquor tax increase on top of the 10 per cent increase introduced by the former Tory government last spring. The latest increase is actually more for breweries outside of Alberta, Saskatchewan and British Columbia.

“It is actually a 20 to 30 per cent increase depending on where you’re from,” said Vern Raincock, the Alberta account manager for Garrison Brewing based out of Halifax.

...

Before Tuesday, breweries outside the western provinces were charged 20 cents a litre. Today, they’re paying $1.25 a litre.

“We’re supposed to have fair trade and free trade between each province and unfortunately what I see is preferential treatment provided currently to British Columbia, Alberta and Saskatchewan-based brewers to the tune of roughly $1.05 or $1.10 per litre,” said Raincock.

For specific details, you can see the liquor mark-ups on the Alberta government's web site.

There are so many possible violations here, it's hard to know where to begin!

First off, perhaps there is a constitutional claim that could be brought, in the hopes of finally getting the free trade part of Canada's constitution working property.  Recall that there is already one such claim underway.

Then there is Canada's Agreement on Internal Trade, which allows for various claims against provincial trade barriers.  There are concerns that this process doesn't work all that well, but I think it has been successful in a few cases.

And of course, there is the WTO.  Recall that there were some successful GATT claims against Canadian provincial measures of this sort, brought by the EEC and the U.S.  Presumably the measures here differ somewhat from the mark-ups at issue in those cases, but at least on the face of it, the issues appear to be similar.

Finally, perhaps there is an ISDS claim somewhere in here as well.  Let's say, for example, there are some U.S.-owned companies who produce beer in one of the disfavored Canadian provinces.  I think they probably have some good arguments under NAFTA Chapter 11 related to national treatment/fair and equitable treatment violations.

Thanks to Derek James From of the Canadian Constitution Foundation for the tip.

Discriminatory Measures and Perfect Measures

This is from the U.S. statement at yesterday's DSB meeting, commenting on the Tuna 21.5 panel and Appellate Body reports:

Panels and the Appellate Body should not make their conception of the “perfect” measure the enemy of all the possible good ones. In pursuing legitimate objectives, Members should not be held to the impossible standard of designing and applying a measure that corresponds exactly to the one that a panel or the Appellate Body would have designed to achieve the legitimate objective at issue.

Do WTO obligations ever require perfect measures?  What about the requirements that SPS measures be science-based?  Or that TBT measures not be more trade-restrictive than necessary?  And now the AB has applied a high standard of regulatory competence under the "stems exclusively from a legitimate regulatory distinction" element of less favorable treatment.  

From ISDS to ICS

From the European Commission:

The European Union and Vietnam today finalised talks for a free trade agreement (FTA), marking the end of the negotiating process. After the agreement in principle reached in August, today’s announcement follows the successful resolution of the few remaining issues on the table.

...

Vietnam has also agreed to accept the EU's new approach on investment protection – in particular a permanent investment dispute resolution system with an appeal mechanism.

The FTA text has not been released yet, as far as I can see.  But if the ISDS provisions in the EU-Vietnam FTA really do reflect the EU's new Investment Court System (ICS), that would be an intriguing development.  The battle over what ISDS should look like has led to a wide range of varying provisions in the thousands of investment treaties around the world.  As always, complexity means lawyers are the big winners!

Of course, it may be a while before the EU-Vietnam FTA actually takes effect.  We are still waiting on the CETA and the EU-Singapore FTA.  Vietnam might have to stand in line a bit.

Note that the Commission doesn't refer to ISDS at all, whereas I keep using the term.  Is the old ISDS really all that different from the new ICS?  Can't we use the same term to describe it?  Even under ICS, there are still investors and states settling disputes, aren't there?

And then there's the question of the yet to be ratified pre-ICS FTAs that still have old ISDS rules in them, such as the CETA and the EU-Singapore FTA. Can those really be ratified as currently written, when the new and improved version of ISDS is out there, being sold as a better alternative?

As a final point, the use of FTAs by major economies such as the U.S. and EU to promote their policy agendas -- here, a revised ISDS -- is interesting.  Have people been writing about this?  If not, they should!

Detrimental Impact in the Less Favorable Treatment Analysis

With the AB's recent Tuna 21.5 report in mind, I think it's fair to say that the focus when analyzing less favorable treatment tends to be on the "stems exclusively from a legitimate regulatory distinction" element.  I agree that this element is crucial, but I'm worried we aren't doing enough with the other element, detrimental impact.  Here's a quote from the tuna report:

7.234. Beginning with the first step, we consider whether the labelling conditions under the amended tuna measure, taken together, modify the conditions of competition to the detriment of Mexican tuna products in the US market. In so doing, we must take "due cognizance" of the DSB's recommendations and rulings, including the findings of detrimental impact in the original proceedings, which constitute relevant background for our assessment. Accordingly, we begin by recalling the Appellate Body's finding in the original proceedings that the detrimental impact of the original tuna measure was "caused by the fact that most Mexican tuna products contain tuna caught by setting on dolphins in the ETP and are therefore not eligible for a 'dolphin-safe' label, whereas most tuna products from the United States and other countries that are sold in the US market contain tuna caught by other fishing methods outside the ETP and are therefore eligible for a 'dolphin-safe' label." ...

Here's my concern: Is it really enough to say that "most" Mexican products get one kind of treatment, and "most" U.S. products get a different kind of treatment?  "Most" could mean 99% for each; or it could mean 51% for each.  The exact percentages matter to me.

If you go to the original panel report, you get some more details on this:

7.314 Based on the above, we note that it is undisputed that at least two thirds of Mexico's purse seine tuna fleet fishes in the ETP by setting on dolphins (therefore fishing for tuna that would not be eligible to be contained in a "dolphin-safe" tuna product under the US dolphin-safe labelling provisions). We also take note of Mexico's indication that the portion of its fleet that the United States identifies as catching tuna eligible to be included in dolphin-safe tuna products (i.e. vessels under 363 metric tons) represents only a limited share of the total catch of the fleet, although we also note that the exact figures are disputed.

7.315 With respect to the US fleet, the United States has indicated that US vessels used to set on dolphins to catch tuna at the time the US dolphin-safe labelling provisions were enacted. The United States explains that there were, at the time, 46 US purse seine vessels along with 52 Mexican vessels that fished for tuna in the ETP, most of which were authorized to set on dolphins to catch tuna. The United States further indicates that US vessels did not fully discontinue the practice until years later, in the mid-1990s.

7.316 From these undisputed elements, it appears that the US fleet currently does not practice setting on dolphins in the ETP. We note also that two US full time purse seine vessels are currently registered to fish in the ETP for 2010 but neither has sought or obtained a Dolphin Mortality Limit (or DML) under the AIDCP, which implies that they are not allowed to set on dolphins. Mexico has also observed that these two vessels are also registered in the WCPFC, which suggests that they are not operating exclusively in the ETP.

7.317 From the above, it can be inferred that, as the practices of the US and Mexican tuna fleets currently stand, most tuna caught by Mexican vessels, being caught in the ETP by setting on dolphins, would not be eligible for inclusion in a dolphin-safe product under the US dolphin-safe labelling provisions. However, most tuna caught by US vessels is potentially eligible for the label, provided that it otherwise complies with the requirements of the measures.  

So the ultimate conclusion relies on "most," but there are some additional facts leading up to that conclusion.

I have two main points here.  First, panels looking at these issues should put a lot of effort into ascertaining the facts.  They should press the parties, and come to factual conclusions related to detrimental impact that are as precise as possible.

And second, when panels and the Appellate Body make findings on less favorable treatment, the degree of detrimental impact should matter.  For example, in the context of TBT Agreement Article 2.1, when considering detrimental impact and the purported legitimate regulatory distinction, the degree of detrimental impact should play a role in reaching an ultimate conclusion on less favorable treatment. (Under GATT Article III:4, where the current state of affairs looks only at detrimental impact, I suppose any degree of detrimental impact would lead to a violation, although I'm not sure how sustainable this approach is in general.)