This is Andrew Lang's Typepad Profile.
Join Typepad and start following Andrew Lang's activity
Join Now!
Already a member? Sign In
Andrew Lang
Recent Activity
Steve, Simon, Brett, and all thanks for these very helpful thoughts. The conversation prompted me to go back to the Panel decision in Indonesia - Safeguards (DS490 and DS496), currently under appeal. It suggested that: ' ... not any measure suspending, withdrawing or modifying a GATT obligation or concession will fall within the scope of Article XIX:1(a). Rather, it is only measures suspending, withdrawing, or modifying a GATT obligation or concession that a Member finds it must be temporarily released from in order to pursue a course of action necessary to prevent or remedy serious injury that will constitute "safeguard measures".' (para 7.14). And, a little later: 'one of the defining features of a safeguard measure is the suspension, withdrawal, or modification of a GATT obligation or concession that precludes a Member from imposing a measure to the extent necessary to prevent or remedy serious injury, in a situation where all of the conditions for the imposition of a safeguard measure are satisfied.' (para 7.17). The case is somewhat different from the present one, because the measure in question was found not to be a safeguard measure in part because Indonesia did not have a tariff binding with respect to galvalume. But it seems pretty clear that on the basis of this test the announced US measures are not safeguards. (It’s worth noting that the Panel came to its conclusion even though the matter was initiated and conducted as a safeguards measure under Indonesian law.) We will have to see what the AB does with this, but it does raise the important point as to which of the conditions in GATT XIX and the Safeguards Agreement go to the question of the characterisation of the measure (is it a 'safeguard?') and which go to its legality (is it a legal safeguard?). The Panel seems to have suggested that 'serious injury' and GATT inconsistency are of the first kind. If that is what it is saying, then I have to admit that makes sense to me. If, taking into account all the evidence, the GATT-inconsistent measure is one which - based on its 'design, structure and revealing architecture' of course! - is genuinely designed to respond to a serious injury or perhaps even a perceived threat thereof, then it can reasonably be considered a safeguard. It's not clear where this leaves the other conditions for the WTO-consistent application of a safeguard measure: increased imports; unforeseen developments; causation; notification requirements; adequate investigation, etc. Some would seem probably to go to legality, and not characterisation, e.g. notification, some requirements concerning the investigation, causation. Others, like the existence of increased imports, are on the borderline. You can actually read the Panel to be saying that all conditions go to the first question ('in a situation where all of the conditions for the imposition of a safeguard measure are satisfied'). But if that is the correct interpretation of the Panel Report, I am less convinced. I take your point, Steve, to be that that interpretation can't be right, and I can see why you say that. But would you go so far as to say that a 'safeguard' need not - as a matter of definition and characterisation - be demonstrably a response to a 'serious injury', or at least a genuine perception of one? I can see the issue is an open one, and the stakes of the AB's decision in the Indonesian case just got higher!
1 reply
Professor Jackson will be remembered as much for the generosity of his spirit, as for the leadership and vision he showed in his work. So many of us on this list have been the beneficiaries of his warmth and kindness, and that of his wife Joan - and that never goes away. It is a hugely sad day, but also an occasion to celebrate the immense intellectual, pragmatic and personal contributions that Professor Jackson made in all areas of his life.
1 reply
Simon, I think this is a really interesting question. Specificity may be a problem (though may not), but I agree the benchmark question is a real one here. Very few publicly run water utilities around the world charge the full social cost of water (at least in the narrow sense that their income does not cover their long-term costs, including necessary infrastructure investment). And in any case water markets are very specific to localities, so comparisons will be of little use. Interesting that exactly this issue (ie how to deal with underpriced inputs) is raised in the AD context in DS459, 474 479 and 480 ...
1 reply
Lorand, can you explain a bit more what you mean when you equate institutional conditions with latent demand? Might this be a way out of the dilemma in some cases?
1 reply
Simon, a quick reply in response to this comment of yours: 'my view is that regardless of whether the subsidy is market-correcting or market-distorting, the SCM Agreement does a pretty good job identifying a narrow category of subsidies that are designed to favor domestic producers over foreign competitors. There is never a good reason for these, and it's fine for them to be disciplined under the SCM Agreement.' I think there is the same core problem here as there is under the GATT non-discrimination jurisprudence. Under GATT III one of the most difficult questions has always been how to treat measures which are not designed to favour domestic producers, but which do so in practice anyway (and of course you know this well). I think this same fundamental problem is at the heart of much of the developing discussion around the SCM agreement. How does the SCM Agreement in its present form deal with these measures? Of course it strikes a delicate balance between discipline and regulatory autonomy, just as the GATT does. I agree with you that, eg, the combination of the specificity requirement, as well as the need for adverse effects in non-prohibited subsidy cases, are an important part of that balance, and can do a lot to protect this sort of measure. But I also think that the term 'benefit' is also a crucial part of the balance that the drafters struck, and that benefit analysis is one of the important points of interpretation where panels and the AB must uphold that balance. On the question of 'hard cases', the cases I listed above are hard for different reasons. (b) and (c) are hard because of practical difficulties of application. To determine whether supply or demand has been changed, you need to determine what supply or demand would have been in the absence of the measure, and the difficulty is that you don't have good evidence of that which is not heavily speculative. (d) is hard because there are by hypothesis many different markets to choose between, and the SCM Agreement provides no clear steer in its text, context or object and purpose as to the basis on which to choose between them. And (a) is hard because there are strong normative reasons - including those based on a particular interpretation of the object and purpose of the Agreement - for discarding the market-as-is benchmark, and no strong textual reasons for adopting it. There's another way of describing the problem as I see it, which might be better. Imagine we use the 'market-as-is-minus-the-impugned-measure' benchmark. The prices in this imagined market are still (in principle at least, it would depend on the facts) massively influenced by a huge range of background governmental measures, including tax law, health and safety laws, consumer protection law, as well as more basic arrangements regarding the details of property rights regimes, corporate and insolvency laws etc. These prices are not purely the result of forces of supply and demand - or, better, the forces of supply and demand are themselves in part a function of these background laws. As Alan Sykes has said, we can't control for all of these laws in our analysis - apart from any of the problems of conceptualising a market in the absence of background laws, it is a practical impossibility. So, in a benchmark analysis we necessarily have to treating some laws as being part of the background institutional context of the market, and other laws as interventions into it (namely, the impugned measure). We HAVE to do that, otherwise the analysis is impossible. The point is how do we draw the line? By what non-arbitrary criteria do we treat some laws in one way, and others in the other? This for me is the really difficult problem. Economists often try to think through this problem with the help of categories such as market-correcting and market-distorting measures (and, for institutional economists, market-creating action). I agree that they have their own problems, and it is true that the SCM Agreement does not use those categories. The problem is that the SCM is almost entirely silent on this, and the text provides precious little explicit guidance on how to think it through. That's why the AB is forced to resolve the problem in ways which in their view best fit the text, context and object and purpose of the agreement, in view of their own institutional mandate.
1 reply
Hi Marc, thanks for this. I meant from a legal point of view, and I guess I should have made that clear. The difficulty is that the text of the law does not provide the answer to this problem. 'Benefit' is not defined, and we have to turn to context etc (esp Article 14) to help us to interpret it. There is nothing in the text which explicitly or conclusively says precisely what 'market' is to be used to determine benefit.
1 reply
Lorand, yes, I think they can be, and of course they were ultimately in the FIT case. But the point is, if that is the best result, then (as Marc says above) they need somehow to fall outside the scope of the SCM Agreement. Just because of the principle of cumulative application. so we can't avoid this difficult issue of defining what is and is not a subsidy/benefit.
1 reply
Simon, thanks for the responses. On the question of 'hard cases', I suppose I would list a few: (a) where the market is distorted (b) where the governmental measure in question so heavily shapes the market that it is difficult to know what the market would look like in its absence (Softwood Lumber) (c) where one has to conjecture what (if anything) might be put in place instead if the impugned measure absent (d) sometimes, when the specific product market in question takes many radically different institutional forms (liberalised energy markets are a classic example, though I say 'sometimes' here because often it will be possible simply to take the institutional form as it exists in the respondent state. On the question of the distinctions to be drawn, I can see that many people are uncomfortable with the idea that state action is required to create functional markets (though most economists would agree with that proposition in some form). But what is your view on the distinction between market-correcting and market-distorting measures?
1 reply
Hi all, I am coming in late to this - no doubt too late. But just a few comments in direct response to Simon's post. (1) I agree that this is a hugely important debate, for at least two reasons. First, because I am continually struck by the extent to which international trade lawyers (probably necessarily) draw on their beliefs about the purpose of the regime in the interpretive choices they make. And second, because of the now quite common argument that part of the challenges facing the trade regime is to re-orient itself towards 21st century issues, which may well mean providing a modified or more developed sense of its own purpose. (2) But, one of the difficult things about the existing literature is that it tends to treat the 'purpose of trade agreements' as something to be discovered - ie something out there, already fixed and discoverable. Of it is in in part that, but actually it is not only that - there are of course multiple conceptions of the purpose available at any one point in time, and periodically they are revised, updated or modified implicitly or explicitly, as the regime re-orients itself in relation to real world events (3) THere are a number of useful methodologies for examining this question of the purpose of the trade agreements. Some methods seek to derive the 'function' of trade agreements from observed behaviours combined with theoretical models. Others try to infer something like a 'subjectively shared purpose' by looking to verbal or textual expressions about the perceived purpose of the regime from key actors. Legal methodologies do something different, by seeking to excavate the 'objective purpose' of the regime from the content, structure and architecture of the regime's law. All have obvious flaws, but all also have important strengths vis a vis the others. (All also have a different concept of 'purpose' in mind, which can sometimes be lost.) (4) One difficulty with the existing debate, such as it is, is that it treats the question of the 'purpose of trade agreements' as in some sense independent question from the broader issue of the purpose of other structures and regimes of global governance. This unduly narrows the horizon of possibilities, and in my view is unrealistic. The function/purpose of trade agreements almost certainly ought to be - and in my experience almost always is in reality - imagined as part of, and in the context of, a much broader vision of the purpose of institutions of global governance more generally.
1 reply
(apologies for the double post) ... another way of responding to your question Simon would be to say that there is no reason in principle to exclude the 'application' of this exception in cases of direct expropriation, even if most of us might agree that it might prove extremely hard to successfully apply it in such cases. Ie, it may be very difficult in practice to show that direct expropriation actually meets all the requirements of the exception.
1 reply
I guess I had in mind regulatory/indirect expropriation where I think the potential application is clear. Having this apply to the expropriation obligation would reduce the need creatively to read exceptions for legitimate regulation into the prohibition itself - instead you would have some (reasonably) firm text to guide the arbitrators in their development of this side of the jurisprudence,
1 reply
Hi Simon The standard of review point was a different one. I am not talking about the standard of review to be applied when assessing the measure for compliance with 2.1. I'm talking about the subsidiary question of the standard of review to be applied to the more specific assessment of whether the EU applied its own criteria in an objective way. I don't think this latter question has been addressed in the jurisprudence. On the intent question, I think the jurisprudence is a little muddled on this point, and that is part of the reason the Panel decision seems to me to fall into difficulty. In the end, I don't think the Panel is saying that, if the EU criteria were applied objectively, then the Greenland hunt would not have satisfied them (though I think that is one possible interpretation). After all, one of the problems that the panel had with the measure was that, properly applied, it seemed to single out Greenland. Nor do I think that the criteria were in themselves not well designed (see the para I reference in the original post). So we have (i) a legitimate regulatory objective; (ii) a legitimate regulatory distinction; (iii) legitimate criteria to operationalize that distinction; and (iv) objective application of those criteria. I don't yet see how we can then turn around and say the measure nevertheless fails on the ground of its objective intention. If there is a problem with the objective intention of the measure, then it should show up in one of those four elements.
1 reply