Thank you, Simon, for the link to the full GMO interim report.
One really has to read the full report to appreciate the Herculian job performed by this panel, and the WTO staff advising it. The report is amazingly well-written.
Given, in particular, our earlier discussions, I want to stress five findings that were not easily detectable in what was publicly available so far.
1. Relevance of non-WTO treaties
In response to EC claims that the panel look at the Biosafety Protocol and other non-WTO treaties to interpret the SPS agreement, the panel – for the first time ever -- defined head-on what rules are covered under Article 31.3(c) of the Vienna Convention on the law of treaties. For the panel, only treaties binding on all WTO members can be “relevant rules of international law in the relations between the parties” (para. 7.68). Treaties binding as between the disputing parties only do not qualify. Yet, and very importantly, other non-WTO treaties can still impart meaning, even if such treaties are not binding on all WTO members, not even on all disputing parties, if they – much like the Oxford dictionary -- shed light on the ordinary meaning of WTO terms pursuant to Article 31.1 (para. 7.92). (This is, by the way, a position I defended in my 2003 book on Conflict of Norms in Public International Law, p. 260, although I would thereafter open the door much wider when it comes to applicable law).
In passing, the panel also declined to find that the so-called precautionary principle is a principle of general or customary international law (para. 7.88).
2. A measure with multiple objectives (SPS and other) can in effect be two measures
If one and the same requirement has two objectives – say, first, the protection of human health and, second, consumer information or public morals – the panel is of the view that this can transform the requirement in effectively two measures. With reference to the first objective (health), the requirement can then be an SPS measure. With reference to the second objective (consumer information or public morals), the requirement can then be a TBT or GATT measure (para. 7.165).
This means that if a GMO restriction is enacted for health or certain environmental purposes, it needs justification under SPS. If, in contrast, it is enacted for other, non-SPS reasons (as well), it can be justified (independently) under TBT or GATT Art. XX.
As the panel noted, this may be important for how the EC can implement this ruling (para. 7.173). The EC could, in principle, reframe its entire GMO scheme on the basis of public morals or consumer information and claim that it thereby implemented the adverse DSB recommendations under SPS.
Later, the panel finds that one of the EC regulations (Regulation 258/97) was based on two non-SPS objectives (namely, avoid misleading consumers and avoid nutritional disadvantages). Yet, having found that there were SPS violations, the panel never reached the stage of examining whether these violations could be “cured” under TBT/GATT with reference to these other objectives.
3. Article 5.7 (provisional measures) is not an exception but a conditional right
As an important concession to the EC, the panel found that Article 5.7 is not an exception to either Article 5.1 (risk assessment) or Article 2.2 (sufficient scientific evidence). Rather, Article 5.7 is a conditional right that carves out the scope of application of both of these provisions (in the same way that 3.3 carves out 3.1).
This means that complainants -- here the US, Canada and Argentina -- have the burden of proving that the conditions under 5.7 are not met. It is not for the EC to prove that it does meet the four conditions in 5.7 for a provisional measure.
This raises the question, however, of whether, as a result, the complainants should have raised or claimed violation of 5.7 in their panel request. If 5.7 is a conditional right, not an exception, it is, in theory, up to the complainant to claim violation of it. The US, for example, did not mention 5.7 in its panel request. Remember, in this respect, that in EC – GSP the AB found that even if the Enabling Clause really is an exception (not a conditional right) it was still for India, the complainant, to raise it (whatever that means). When faced with a conditional right – as here in GMOs – this should, all the more, be a requirement. In the end, however, the panel finds a violation of 5.1 (not, strictly speaking, of 5.7) (see para. 8.23(a)), so this technical argument may be moot (surely, now that the panel found that 5.1 and 5.7 are self-standing alternatives, they cannot ever be both violated at the same time).
4. Once a 5.1 risk assessment can be made, provisional measures under 5.7 become unavailable
Heavily relying on the AB report in Apples (at para. 179), the panel concludes that once (the EC) makes a risk assessment, (EC member states) can no longer rely on 5.7 and enact provisional measures. To deviate from EC policy, EC member states must present their own (new) risk assessment in line with 5.1 (para. 7.3057 and 7.3224). Sufficiency of evidence under 5.7 is, in other words, linked only to the possibility to make a 5.1 risk assessment. Sufficiency of evidence has, according to the panel, no relation whatsoever to a country’s tolerance of risk or appropriate level of protection (para. 7.3234; a finding, in my view, in tension with the AB report in hormones, para. 124).
As noted in my earlier comment, this is, I think, the most vulnerable part of the report. Surely, if Austria were to find new scientific evidence tomorrow that GMOs are really deadly, it must have the right to provisionally ban GMOs, even in the face of a previous EC risk assessment and without yet having completed a full, new risk assessment of their own that incorporates the new evidence. Yet, according to the panel, Austria would, in that case, not ever be able to impose provisional measures under 5.7 as, before that, a risk assessment (by the EC) was possible which, in turn, means that there is, for ever after, “sufficient scientific evidence” and, hence, no longer any right to impose provisional measures. To do anything different from the EC, Austria must, according to the panel, complete a full risk assessment and only then would it be able to restrict the GMO. This would be absurd.
Instead, the panel could have found that even if there were insufficient evidence, EC member states did not complete a more objective risk assessment within a reasonable period of time (fourth condition under 5.7). To fault WTO members on such procedural requirements makes, in my view, much more sense than to fault them on substantive points.
In para. 7.1525, in respect of the “undue delay” in EC processing, the panel seems to admit that new evidence itself (not necessarily a full new risk assessment) can justify a decision “to suspend all final approvals pending an appropriate assessment of the new evidence”. In later parts of the report, however, a 5.1 consistent risk assessment – and not just new pieces of evidence – are required.
Finally, for the panel, the lack of a risk assessment automatically means insufficient scientific evidence (5.1 violation necessarily violates also the third condition in 2.2). It did so relying on the AB in Salmon (para. 137) even though there the AB never specified which of the three elements in 2.2 it was talking about. In my view, lack of a risk assessment implies a measure not based on scientific principles (which seem to include risk assessments). Yet, it may well be that a country does not have a procedurally correct risk assessment (in violation of 5.1), but still has sufficient scientific evidence (in some other form) in support.
5. Relevant Date for Evidence is Establishment of the Panel
Note, finally, that the panel refused to look at scientific evidence that came up after 29 August 2003, that is, the date of the panel’s establishment. In other words, even if some crucial new piece of evidence had turned up during the panel process, it would not have counted. Actually, when it comes to Austria’s safeguards, Austria did supply new studies that look quite important – on allergenicty, toxicity, gene transfer etc. – but since out of time, the panel refused to consider them. Unlike trade remedy cases, should not SPS panels look at all evidence submitted to them up to the date of its decision, or at least, up to the closure of debates?