Enter through the narrow gate. For wide is the gate and broad is the road that leads to destruction, and many enter through it. (Gospel according to Matthew, 7:13–7:14)
This piece is a comment to the Canada–Renewable Energy/FIT disputes. Its focus is, in particular, on the future of benchmarking after this important case. It draws on the already rich commentary, the latest example being Steve Charnovitz's and Carolyn Fischer's excellent article. Our analysis is largely complementary to theirs.
For the benefit of the reader, it is worth briefly summarizing what the case is about.
The WTO dispute settlement was called to decide whether Ontario's feed-in tariff (‘FIT’) system, which provides incentives to renewable energy producers and critically includes a local content requirement, was in line with WTO laws on national treatment and subsidies. After easily disposing of the local content requirement, which was patently in breach of the prohibition of non-discrimination, the Panel and the Appellate Body were both asked to establish whether the measure was also a prohibited subsidy. The determination of whether the FIT was a subsidy was the preliminary step of this analysis. In this respect, it was necessary to conclude that the incentive scheme was providing a benefit to its recipients. Both courts rejected the claims, arguments, and benchmarks put forward by the claimants, and concluded that they were not in a position to find for the existence of a benefit.
The Appellate Body's reasoning in this respect is particularly important, and has been the subject of much analysis and discussion. In a nutshell, the World Trade Court introduced two innovations to its long-standing jurisprudence on the benefit analysis under Article 1 of the SCM Agreement. It first required that a full market definition analysis be carried out.Footnote 1 This exercise's function is to identify the universe within which the appropriate market benchmarks can be found. Crucially, the Appellate Body found that this analysis should not stop at demand-side substitutability, but should extend to consider also the supply-side.Footnote 2 After concluding that the market at issue in the case was that of renewable energy only (as opposed to the broader energy market),Footnote 3 the Appellate Body introduced its second innovation. It distinguished two scenarios. Those where governments (simply) create markets and those where they intervene in already existing markets.Footnote 4 Importantly, the former is not ‘in and of itself’ a subsidy. It also transpires that, in the first scenario, the key requirement is that the government creative action should be proportionate and not excessive. In sum, if there is a new market and the creative act is simply conducive to bringing this market into existence, WTO subsidy laws do not apply. It is clear that, through this interpretation, the Appellate Body managed to take certain forms of subsidization in the clean energy sector outside of subsidy control.
My assessment of this decision is negative. In other writings, I have severely criticized both the Panel (majority) and the Appellate Body.Footnote 5 The burning question was: How is it possible that what is almost a textbook example of subsidy has turned out not to be a subsidy at all?Footnote 6 More specifically, we are of the view that their interpretation of the notion of benefit amounts to a fundamental misconstruction of the law. The result is a limited but unwarranted carve-out for a certain type of subsidies from subsidy disciplines. The potential implications of this exception are troublesome. If our reading is correct, it is desirable that future decisions limit the impact of this decision. Whether, and how, this is possible is the focus of this piece.
This article is structured as follows. After a brief overview of the rich commentary on the disputes in Section 2, Sections 3 and 4 briefly review the two main innovations (and mistakes) introduced by the Appellate Body. A brief interlude on the significance of this decision for the EU (Section 5) paves the way to the two interpretations of the Appellate Body's benefit analysis that are now possible. Several questions are left open and the WTO judicature could either follow an expansionist reading or a narrow path. These two alternatives are discussed in Sections 6 and 7. With Section 8, the analysis focuses on the serious challenges that a more desirable narrow reading of the ‘Canada–Renewable Energy/FIT exception’ may pose. Section 9 concludes.
2. Literature review
As testament to the true importance of the case, commentary on and analysis of these rulings followed soon after they were delivered. Most of the attention focused on the subsidy analysis, and in particular on how the Panel and especially the Appellate Body construed the benefit test. Despite their diversity, virtually all pieces share one common denominator. Their assessment is negative.
This section aims to highlight the key findings of this literature. Its goal is not to offer a comprehensive review, but rather to convey the gist and feel of the assessment of this decision so far.
The first reactions to the decision can be found on the web, and in particular in various posts in the International Economic Law blog.Footnote 7 Consultancies also acted quickly. Commenting on the Panel Report, economist Amar Breckenridge perceptively noted:
The reason governments choose to support the development of renewables is not primarily because of concerns regarding the adequacy and affordability of supply. Rather it is because lower emissions sources of energy, such as renewables, confer a positive benefit to society that is not normally captured by market arrangements absent some specific form of government intervention. In such circumstances, economists speak of market failure as a result of externalities. One way to address this externality is by subsidizing the production of energy from renewable sources. FIT schemes are one example of such subsidies. The benefit of the subsidy accrues to the producers of energy from renewable sources. In the absence of such a benefit, there would be no incentive for them to invest. The public benefit to society from lower emissions is entirely predicated on there being a private benefit to the producers of energy from renewable sources. These private benefits take the form of fixed prices over a lengthy period of time representing a significant mitigation of investment risk – more so than would likely be available on commercial terms, and more than made available under contracts between government and private investors.Footnote 8
This quote summarizes the economics of the case and how economists view FIT schemes. They are incentives that benefit private producers and subsidize the production of energy while tackling one or more market failures.Footnote 9
It is, however, the Appellate Body decision that attracted the large majority of published comments. The first ones were mainly in the form of short policy briefs. Casier and Moerenhout (Reference Casier and Moerenhout2013) observed that the Appellate Body ‘seemingly wanted to avoid ruling on whether an FIT constitutes a subsidy, thereby creating uncertainty for investors in renewables and governments designing their renewable energy policies'.Footnote 10 After exploring the findings on ‘market creation’ and ‘benchmark markets’, which we will delve into below, the main conclusion of the two authors is that the primary responsibility for clarifying the law is not that of the Appellate Body, but of the WTO members. The way forward for the regulation of renewable energy support in the WTO is law reform, ‘in the form of general exceptions or the revival and renegotiation of the expired non-actionable category of subsidies in Article 8’.Footnote 11 In a similar vein, Cosbey and Rubini (Reference Cosbey and Rubini2013) criticized the Appellate Body, highlighting in particular the implications of the decision for the future. If the Appellate Body gave more flexibility, this came with a hefty price, especially with respect to subsidy transparency and good governance. The uncertainties and dangers, with which we are left after this ruling, support, in the authors' view, the case for reform of subsidy rules. They note:
All this supports the case for reform of subsidy rules. As this case shows, a case-law solution has inherent limitations. Members – not dispute settlement – should take the lead and responsibility on identifying what is good policy and should therefore be permitted. Only reform, which may take various forms [interpretative understanding, authentic interpretation, temporary waiver, treaty amendment], all ultimately in the hands of Members, would enable to reach the objectives of desirable policy space, in respect of the integrity of the rules, and safeguard of transparency and good governance. Only reform can ensure the legitimacy of the fundamental decision of what type of government intervention should be permitted and what should not. Only reform can ensure the necessary legal certainty to both government and business action.Footnote 12
In a second client briefing, Amar Breckenridge focused on, and criticized, the Appellate Body's market definition analysis.Footnote 13
The first comprehensive academic commentaries appeared in the first quarter of 2014. In a lengthy, at times caustic, working paper, Rubini (Reference Rubini2014a) produced an extensive critique to both the Panel's and the Appellate Body's decisions. He surveyed the various economic and legal errors of these rulings, which, in his view, all boil down to an incorrect approach to legal analysis and are an excellent example of an unwarranted activist attitude, where what are perceived as desirable policy outcomes seem to count more than correct legal interpretation.
The first issue of the 2014 Journal of International Economic Law (JIEL) features two articles on the case. Cosbey and Mavroidis (Reference Cosbey and Mavroidis2014) provided a complete review of the Appellate Body decision. Their assessment is also negative, and the approach followed by the Appellate Body, aimed at avoiding finding that a scheme promoting a public good is a subsidy, is tagged ‘legal acrobatics’. In particular, they noted:
There are many problems with the approach followed: it is hardly reconcilable with the text and spirit of the relevant WTO rules; it might be giving WTO Members the wrong incentives by opening the door wide to industrial policy unlimited; it places squarely before us the question whether the WTO courts have behaved as agents called to apply a law decided by their principals, or whether they re-invented themselves as principals and decided what the law should be.Footnote 14
Their conclusion is that the ‘rationale for subsidization should matter’ and that WTO subsidy laws should be re-thought accordingly.
Equally critical is the other article in the same issue of the JIEL. Pal (Reference Pal2014) complained about the Appellate Body's interpretation of the benefit. In his view, this ‘exhibits a bias on the part of the Appellate Body to exempt government support for renewable electricity from the disciplines of the SCM Agreement’.Footnote 15 This may result in giving the green light to trade-distorting subsidies for inefficient production technologies.Footnote 16
Hestermeyer and Nielsen (Reference Hestermeyer and Nielsen2014) concentrate on the legal status of local-content requirements under WTO law. The authors also summarize and briefly comment on the Appellate Body's subsidy analysis. If they do not openly criticize this jurisprudence, they however note the ‘liberty to establish “artificial” markets’ and the reference to policy imperatives which is now making it considerably harder to establish a benefit.Footnote 17
If virtually all commentary outlined so far is critical, one piece, that appeared in the Journal of World Investment and Trade, is more on the positive side in its evaluation. While conceding that the Panel's and Appellate Body's approach may well be regarded as ‘activist’, Kent and Jha (Reference Kent and Jha2014) believe these rulings are better classed as ‘evolutive’. They note in particular that the two adjudicating bodies ‘had reasonable grounds to develop the law and adapt it to today's circumstances, necessities, as well as to the overwhelming state practice’.Footnote 18 The two authors also noted important limitations to this approach, introduced by the Appellate Body itself, and in particular the strict prohibition of discrimination as well as the reference to the use of competitive bidding procedures.Footnote 19
Time has enabled scholars to provide more theoretical analysis on this case too, which is evidence of the inspirational potential of its issues and findings.Footnote 20 Importing concepts from sociology, Andrew Lang analyzes various subsidy cases, including Canada–Renewable Energy/FIT. Footnote 21 In his view:
[t]he point is crystal clear: what is happening in these and other examples is not the objective definition of a subsidy by reference to an idealized natural market; instead, there is the contingent claim that a certain kind of measure should (or should not) be treated as if it were a subsidy for present purposes, based on a particular understanding of the objectives of the Subsidies Agreement, and the institutional role of WTO dispute settlement. The notion of subsidy, in other words, is only ever defined for a specific purpose and relative to a particular context.Footnote 22
The (so far) final piece commenting on the Canada–Renewable Energy/FIT disputes is that of Charnovitz and Fischer. The authors highlight the apparent shift indicated by the Appellate Body's analysis from the ‘primacy of the market’ to the ‘primacy of government volitions’. They also note that, before this case:
the conventional wisdom was that Ontario's FIT program (like many FIT programs) purchases electricity at above-market prices. The Appellate Body, in effect, moved the goalpost by redefining the market to be electricity from renewable sources. As a result, the question of whether the prices paid in the FITs are above market no longer has an obvious answer.
Most crucially, while, courtesy of the special characteristics of renewable energy, Charnovitz and Fischer (Reference Charnovitz and Fischer2015) do not necessarily reject the ‘selection of a renewable energy electricity market as the relevant market for determining whether the benefit exists’, they are concerned for the ‘lack of cogency’ in the conclusion that supply-side factors outweigh demand factors, and, more generally, for the lack of methodology on how to compare the relative weights of these factors. Their main apprehension is formulated as follows:
in our view, it is disturbing to think that whatever government policies emerge from public choice distortions are automatically mapped into the baseline in determining whether a recipient of a financial contribution is better off than it would have been in the marketplace. Besides preferences about diversification in electrical generation, governments may have preferences about the merits of state-owned enterprises, the benefits of technological localization, the value of export-led manufacturing, the recalibrating of transatlantic income tax symmetry, or redistributing income from other countries.
The Appellate Body may not recognize the slippery slope it has created by introducing policy considerations into the determination of benefit (in seeming contradiction to its own case-law) and to equate politician preferences with consumer preferences. Perhaps the Appellate Body should have given more attention to the dissenting opinion which warned that the fact that a market does not meet the objectives that a government might have does not shield financial contributions from benefit analysis.
In conclusion, we consider it highly significant that (with a couple of exceptions) virtually all commentary so far is highly critical of the benefit analysis of the Panel (majority) and the Appellate Body. Vox populi, vox dei! – we would be tempted into adding.
After this literature survey, it may now be useful to briefly review the two innovations introduced by the Appellate Body – and explain why they amount to serious misconstruction of the benefit and of subsidy laws at large.
3. First mistake: the definition of the market
It is common ground in WTO subsidy laws that the benefit determination requires the identification of an appropriate market benchmark against which the measure at issue should be assessed.Footnote 23 In the Canada–Renewable Energy/FIT case, the Appellate Body immediately made it clear that, in order to identify an appropriate market benchmark, it is first necessary to define the relevant market.Footnote 24
Market definition is a prevailing tool of antitrust analysis, preliminary and functional to the assessment of market power, where its goal is to determine whether the firm under investigation is subject to competitive constraints. This is done by enquiring whether the products or services at issue are substitutable with each other. The Appellate Body's approach and jargon seem to directly refer to this antitrust world.
What is crucial is that, according to the Appellate Body, the definition of the relevant product market in the context of the benefit analysis cannot stop at the demand-side but should also consider the supply-side.Footnote 25 This way of proceedings has momentous consequences (which will be explained later on). For now, we concentrate on whether this methodological approach is correct or not.
The key criterion of assessment of the Appellate Body's methodology is whether the nature of subsidies and the specific goal of the benefit test allow this antitrust import.
We argue that the Appellate Body is wrong because it makes a too-easy transplant from market definition in antitrust analysis to subsidy analysis. Subsidies pose different problems as compared to collusive or anti-competitive unilateral conduct. As a direct consequence of this, product market definition for subsidies is different.Footnote 26
To explain this point, it may be useful to make reference to the EU experience where competition laws encompass typical antitrust laws (regulating collusion and abuses of dominance), merger control, as well as State aid control. Market definition is a very common and entrenched tool of analysis. The 1997 Notice on the definition of the relevant market for the purposes of Community competition law of the EU Commission makes an express distinction between market definition in antitrust and market definition in State aid.Footnote 27 We can read:
The focus of assessment in State aid cases is the aid recipient and the industry/sector concerned rather than identification of the competitive constraints faced by the aid recipient.
This concept is expressed even more clearly in a study carried out by two economists and ordered by the EU Commission in the mid-2000s. Nitsche and Heidhues (2005) note:
Market definition in the context of State aid should focus only on demand side substitutability. This is in contrast to the standard competition policy approach where both supply and demand substitutability are considered. In standard competition policy cases, the focus is on the competitive constraints that make prices increases unprofitable. For this question, it is important to include those potential competitors that render a price increase unprofitable because they have an incentive to start producing substitute products once the price is raised. In State aid cases, the relevant question is to trace the effects of a price decrease or an output expansion of the aid recipient's products. Thus, market definition should focus on how other existing producers are affected and on the potential effects on producers in markets that the aid recipient intends to enter.Footnote 28
It is worth noting that these observations are not system-specific, i.e. they are not valid only in the specific context of the EU and its State aid control. They are ‘economics 101’.Footnote 29 Subsidy control is not antitrust. Supply-side analysis is largely irrelevant.
If the above is correct, it is clear that, by considering supply-side in its product market definition, the Appellate Body has made a fundamental mistake of economic methodology, which, being imported into and relevant to legal analysis, turns into a fundamental mistake in the legal analysis.Footnote 30
We do not therefore share Charnovitz and Fischer's (Reference Charnovitz and Fischer2015) view that the Appellate Body may have been correct in concluding that renewable energy is a market on its own, and that the main issue with their decision is in the lack of reasoning – better: of methodology – justifying that conclusion. Surely, as they note, ‘government renewable energy policies do not view all electricity generation as the same’.Footnote 31 This does not mean, however, that they necessarily belong to a different market. The fact itself that support to clean energy may harm conventional energy producers is evidence of the fact that what we have is one single market.Footnote 32 What we are trying to say is that there is no (full) overlap between policies and markets.
What are now the momentous consequences of the Appellate Body's wrong approach?
In this case, the Appellate Body considered the different costs structures and operating costs and characteristics of renewable energy production, vis-à-vis other electricity, as relevant supply-side considerations. This led the Appellate Body to conclude that the relevant market is only that of electricity produced by wind and solar (as opposed to the broader energy market).Footnote 33
The result of the Appellate Body's narrow market definition was that the appropriate available benchmarks became those of the renewable energy sector and its characteristics. In a sense, the economic and legal standard against which Ontario's FIT had to be judged was essentially ‘tailored’ to that same policy. In other words, the Appellate Body's move – our set of reference is renewable energy only – made it easier for Ontario's policy to become the norm, or close to the norm, and thus reduced the possibility that it could be regarded as exceptional and hence advantageous. (And the reader should recall – this practically means a ‘no benefit’ and hence ‘no subsidy’ finding.)
Had the Appellate Body applied market definition correctly, they would have concluded that the market was broad, encompassing all electricity, irrespective of its green or less green origin. This would have paved the way to the easy conclusion that there was a benefit and a subsidy. In sum – an outcome opposite to the one reached by the World Trade court.
In conclusion, the first innovation of the Appellate Body has been to request a full market analysis at the level of the benefit analysis and, most importantly, to indulge in considering supply-side. This has led to the narrowing of the market and of the universe of reference for benchmark identification.
4. Second mistake: the creation of the market
After having defined the relevant market in a narrow way, the Appellate Body carried out its analysis and introduced another important innovation. They said it is necessary to distinguish new markets from existing markets. In particular, in an already oft-quoted passage, we can read:
a distinction should be drawn between … government interventions that create markets that would otherwise not exist and … other types of government interventions in support of certain players in markets that already exist, or to correct market distortions therein … While the creation of markets … does not in and of itself give rise to subsidies within the meaning of the SCM Agreement, government interventions in existing markets may amount to subsidies.Footnote 34
This distinction is puzzling.Footnote 35 Unfortunately, however, the distinction between new, indeed created markets, and existing markets has crucial practical significance. The Appellate Body is in particular saying that the acts that led to the creation of new markets do not ‘amount in and of itself to a subsidy’.Footnote 36 To paraphrase – they are not saying that market creation does never amount to a subsidy but that this does not automatically happen. You need to have other factors or circumstances present that may turn this act of creation into a subsidy problem. Absent them, the Appellate Body is crucially suggesting, you do not have a subsidy. As has been repeatedly said, this is a ‘carve-out’, ‘exception’ or ‘exclusion’ of certain types of action from subsidy laws.
The attentive reader may have started to ask questions: how does one identify a new market? What does creation of a market mean, from a prescriptive and legal perspective (which is the one that matter in the current context)? What can turn a ‘no-subsidy scenario’ into a ‘subsidy scenario’? Where, and how, do you draw the line?
Let's carry on with the analysis of the indicia left by the Appellate Body. They do indeed provide various indications of what the act of market creation may involve: governments decisions on the supply-mix of energy (which does depend on policy choices and objectives),Footnote 37 the regulation of the quantity and type of electricity supplied through the network and the timing of supply,Footnote 38 what are called the parameters of the system,Footnote 39 price setting, such as FITs (based on cost-recovery and a reasonable margin), and quantity mandates.Footnote 40 If this is what the market creation scenario includes, it is clear that this is a broad notion.
Does all this mean that all types of activity included in the previous list are sheltered from the wrath of subsidy disciplines? It seems not. If we focus on the measure at issue in the Canada–Renewable Energy/FIT case, it looks like a FIT may be protected if its key economic component – the tariff – is adequate and not excessive. But the identification of the dividing line between what is adequate and what is excessive may be difficult. A new set of questions naturally arises. In particular: How do we define the appropriate level of costs and profit? Is the Appellate Body providing a shelter for the compensation of all costs of renewable energy production? Or – rather – should the safe-harbor be recognized only for the extra or additional costs for producing clean energy? Shall we really – as the Appellate Body suggests – limit our search to the clean energy sector only? Or – rather – should our analysis consider the costs and profits of producing energy or electricity more generally? These are crucial questions which ultimately lead to the identification of the standard against which the measure at issue needs to be scrutinized. We will return to this below.
Let's now pause a bit and discuss the several innovations introduced by the Appellate Body. We explained elsewhere our serious misgivings about the Appellate Body's notion of ‘market creation’.Footnote 41
This is, in our view, faulty in three fundamental respects.
First, it is wrong inasmuch as it does not have any legal or economic justification. From a legal standpoint, it is, in essence, an example of unwarranted consideration of public policy objectives at the level of the definition.Footnote 42 This takes place by making governmental energy supply-mix decision relevant. From an economic view of point, a paradox comes out. To say that if you create a market from scratch you may be safe under subsidy laws – as opposed to less significant interventions in the market – can be rephrased in the following way. The bigger the market failure, the action of the government, and the possible distortions, the less likely you are found – legally speaking – to confer a subsidy. Can this be right?
Secondly, the boundaries of the ‘market creation’ scenario – which, it is important to remember, does not have a purely descriptive function but aims to represent a prescriptive legal notion, capable of distinguishing what does amount (or otherwise) to a subsidy, are ultimately undefined and unclear. To put it differently: can we simply rely on the adequacy or excessiveness of remuneration? We are not persuaded into believing that the benefit and subsidy question is just a matter of proportionality of public action. Surely, the question is different. Would market forces have acted as the government has acted?
Thirdly and finally, as it has been formulated, the language of market creation, is dangerously open-ended. Considering its important practical effect (i.e. carve-out of certain measures from subsidy control), more precision (and arguably strictness) is necessary.
As noted, the Appellate Body also gives some indications about what type of benchmarks could be of use in discriminating creations from interventions. As Charnovitz and Fischer (Reference Charnovitz and Fischer2015) note, the Appellate Body suggests a (not-so-rigid) sequence of benchmarks.Footnote 43 In particular, there seems to be two different kinds of yardsticks. The first category seems to a large extent to rehearse the acquis on the issue, referring to market prices, administered prices, and proxies. As it is also common practice, the possibility for construction or adjustment is expressly mentioned. At this point, there is, however, an interesting twist in the analysis, which may indeed constitute an important hint, and an indication of what lies ahead in the future of benchmarking. After outlining the various benchmarks outlined above, the report reads:
Alternatively, such benchmark may also be found in price-discovery mechanisms such as competitive bidding or negotiated prices, which ensure that the price paid by the government is the lowest possible price offered by a willing supply contractor.Footnote 44
What is the meaning of ‘alternatively’ (in particular if read together with the adverb ‘also’)?Footnote 45 Is it closer to ‘available as another choice’, essentially suggesting price-discovery mechanisms as one of the various alternatives? Or does it rather mean ‘in the absence thereof’, which would, by contrast, suggest a stricter subsidiarity of the use of these procedures to the more prevailing practice of discovering benchmarks from the market? These questions have fundamental practical importance.
What did the Appellate Body mean to say? Are they hinting at something, indicating in particular a preference for these benchmarks which, as they seem to believe, ‘ensure that the price paid by the government is the lowest possible price offered by a willing supply contractor’? To put it another way: do governments now – in order to be on the safe side – have an incentive to use price-driving mechanisms? We do indeed believe this is the case. As the following section tries to show, there may other contextual indications to support this reading.
Now, although the use of price-driving mechanisms does not necessarily exclude the existence of a subsidy,Footnote 46 it may be a useful path to follow to ensure that (a) the remuneration is the minimum necessary (thus keeping distortions to the minimum, if not avoiding them), and (b) the awarding government acts in a transparent way (which would contribute to avoid abuse and circumvention). This concept is thoroughly analyzed in the following sections.
We conclude our analysis of the two innovations on the benefit analysis of the Appellate Body with one gloss. The two innovations of the Appellate Body – a narrow market definition and the idea that new markets are to some extent beyond subsidy laws scrutiny – are two sides of the same coin. They share the belief that renewable energy is ‘special’ and that for this reason action to support it cannot be considered as exceptional or derogating (which is the essence of the legal notion of subsidy).Footnote 47
5. The real victory of the EU
We have noted our impression that the Appellate Body seems to be hinting at a preference for ‘price-discovery mechanisms such as competitive bidding or negotiated price’.
But where does this type of benchmark come from? Who mentioned it or put it forward during the litigation? The reference to public procurement procedures as devices to exclude financial contributions that confer undue advantages will certainly ring a bell with any reader with some knowledge of EU State aid law. One of the most prominent debates in the State aid world in the beginning of the 2000s was the legal status of the financing of public services.Footnote 48 In particular, the question was whether the simple compensation of their costs can confer an economic advantage and hence qualify as State aid. After a fluctuation in the jurisprudence, which featured an interesting clash of views,Footnote 49 the European Court of Justice (‘ECJ’) passed the famous Altmark decision in 2003. The ECJ concluded that there is no advantage (and no State aid) if four conditions are satisfied, notably:
• first, the recipient undertaking is actually required to discharge public service obligations and those obligations have been clearly defined;
• second, the parameters on the basis of which the compensation is calculated have been established beforehand in an objective and transparent manner;
• third, the compensation does not exceed what is necessary to cover all or part of the costs incurred in discharging the public service obligations, taking into account the relevant receipts and a reasonable profit for discharging those obligations;
• fourth, where the undertaking which is to discharge public service obligations is not chosen in a public procurement procedure, the level of compensation needed has been determined on the basis of an analysis of the costs which a typical undertaking, well run and adequately provided with means of transport so as to be able to meet the necessary public service requirements, would have incurred in discharging those obligations, taking into account the relevant receipts and a reasonable profit for discharging the obligations.
To rephrase and distill the key legal issues, under the Altmark jurisprudence there is no aid from a legal standpointFootnote 50 if (i) there is a clear entrustment of public service, (ii) the compensation is defined beforehand and in a clear and transparent manner, (iii) the compensation is necessary, and (iv) the enterprise is chosen via public procurement procedures or, in the absence of these, the compensation is the one that a ‘typical undertaking’, similarly placed, would need to perform the public service. All these conditions need to be cumulatively present in order to make the measure of support safe under EU State aid law.Footnote 51 If any of them is not present, the ‘subsidy’ will qualify as ‘State aid’, and will be subject to notification requirements and the Commission's control. If the conditions of the relevant exceptions are there, the ‘subsidy’ may eventually be declared legal, that is ‘compatible with the internal market’ in the EU language.
This case and its test were clearly on the minds of EU litigators who certainly tailored some of their legal arguments to the Altmark jurisprudence.Footnote 52
This enables us to make a step further. The real winner of the Canada–Renewable Energy/FIT case is the EU – and not only in the superficial sense of having obtained its litigation goal, i.e. the condemnation of Ontario FIT programme's local content requirement. In this sense, the EU is a winner as much as Japan, the other complainant (and, certainly, all other countries that, altough often using them themselves, dislike localization requirements imposed by other countries).
The EU's success is much deeper and can have broader implications. It is based on canny strategic lawyering which secured that one interpretation of the benefit, which is prevailing and on the rise in domestic EU law, be accepted by the Appellate Body and, if followed and built upon in future cases, this may crucially render a broad range of State aid measures that are compatible with EU law are secure also under WTO law .Footnote 53
The requirement of price-discovery mechanisms is now pervasive in domestic EU State aid lawFootnote 54 and will soon become the only alternative for renewable energy support. For example, according to the new Energy and Environmental Aid Guidelines, introduced on 9 April 2014,Footnote 55 with the possible exemption of small producers (i.e. small installations or technologies),Footnote 56 feed-in tariffs are progressively replaced by ‘competitive bidding processes’, which, in the EU's view, ‘will increase cost effectiveness and limit distortions of competition’.Footnote 57 In other words, the use of competitive bidding is a legal requirement for Member States to have their renewable energy subsidies compatible with EU law.Footnote 58 In Europe, it will in principle become the norm in 2017,Footnote 59 subject to few exceptions.Footnote 60
We will go back to Altmark and its four conditions as a source of inspiration for techniques to clarify and narrow the ‘Canada–Renewable Energy exception’. We will also crucially outline the challenges faced in the EU in making use of public procurement procedures to exclude the existence of State aid.
6. The wide road
As noted, the innovative approach introduced by the Appellate Body probably raises more questions than answers. It may also be that the Appellate Body did not want to be too much precise in some of its rulings, perhaps wishing to see the reactions to them. Whatever it may be, generally speaking, two approaches to the ruling and its reasoning are now possible: one broad, the other narrow. In this section, we focus on the former, leaving the analysis of the latter to the next section.
Future WTO panels (and indeed the Appellate Body) could follow the ‘wide road’. The premise of this approach is that the Canada–Renewable Energy/FIT decision is good law, inasmuch as it offers a proper interpretation of the benefit. Crucial to this stance is that the ruling does in no way create a carve-out or exception. In a nutshell, the adoption of this approach is premised on the ‘normalization’ or ‘justification’ of the decision.
There are indeed two possible variations to this view. On the one hand, one could argue that the decision is about creating a shelter for the renewable energy sector, in the absence of specific exceptions. Additionally, it could be observed that it is the specific factual and economic circumstances of the case that make this law very case-specific and the possibility of exporting it extremely difficult. If this is the case, the argument goes, there would simply be no need, or indeed possibility, to refer to this ruling in future cases. This does not exclude, however, that, in practice, it is highly likely that future complainants, if necessary, will base their claims and arguments on this ruling and its reasoning, forcing adjudicating bodies to difficult exercises of distinguishing . Hence, the findings of the Canada–Renewable Energy/FIT case could have – de facto – broader application, beyond this case and this sector.
On the other hand, even more dangerous than a simple normalization of the ruling is its express ‘generalization’ or ‘expansion’. In other words, the main concern and risk is that this law is intentionally interpreted and applied broadly. The logical fabric of the reasoning of the Appellate Body would be considered as a good prototype to replicate and use in other cases.
These scenarios are not mere speculation. Prominent players have already taken positions very close to the ones just depicted.Footnote 61 The arguments are various: the ruling could hardly apply beyond the specific circumstances of the case; the Appellate Body approach would simply reflect a ‘widespread technique for concluding that there is no aid or subsidy’, one reflecting a quid pro quo analysis whereby the alleged benefit of the measure is simply compensating a cost, disadvantage, or obligation on part of the recipient; there is no issue (and ‘no benefit’) if the measure is not discriminatory and pursues a legitimate objective in a proportional way; the use of ‘price-discovery mechanisms’ does ensure that there is no over-compensation; ultimately, the decision, and in particular the distinction between market creation and market interventions, simply reflects the distinction between regulation and subsidies; the lack of application of subsidy laws does not imply a legislative vacuum but more simply that other provisions, such as national treatment ones, would find space.
We are not persuaded by these arguments. First, the innovative findings of the Appellate Body, and in particular their ‘market creation carve-out’, are sufficiently broad, and unclear, to find application in many other cases, beyond the clean energy sector, and beyond this case. The variety of real-life and litigators' ingenuity will offer context and incentives. Secondly, the quid pro quo approach, imported from EU State aid law, is not widespread at all, quite the contrary. We will return to this point. Most importantly, irrespective of its general or otherwise currency, the adoption of a broad ‘compensation logic’ when it comes to define a measure as subsidy brings with it the risk that a (net) benefit will rarely be found to exist. Thirdly, to simply rely on whether the measure does (not) discriminateFootnote 62 or on whether there is a proportion in relation to a legitimate objective leads to the same type of inconvenience. It is conceptually vague and practically dangerous. Fourthly, the use of price-discovery mechanisms does not necessarily ensure that there is no over-compensation. It all depends on whether what the public body wants and is prepared to pay is in line with what the market wants and would be prepared to pay. In other words, the use of price-discovery mechanisms cannot replace the validity of the indications coming from market benchmarks. We will come back to this crucial aspect below. Finally, the distinction between regulation and subsidies cannot be found in the law. It is a rationalization that is probably more useful to describe an outcome rather than crucially offering indications on what is or is not covered by the definition of subsidy. Furthermore, even assuming its correctness, there is a very large grey area in the middle. The measures subject to analysis in cases like Canada–Renewable Energy/FIT or, in the EU, PreussenElektra are excellent examples to the point.
In conclusion, an expansionist reading of the Appellate Body's analysis can give a huge blow to subsidy disciplines, lessening transparency and control of potentially trade-distorting measures.
As Matthew warns, ‘broad is the road that leads to destruction, and [unfortunately, we add] many enter through it’Footnote 63. This can be aptly applied to our case too.
7. The narrow road
If there is a wide road, there is also a narrow one. The premise of the narrow reading stands in contrast with the assumption of the alternative approach. Stakeholders, litigators, and especially adjudicators, must at least be aware (if not even recognize) that Canada–Renewable Energy/FIT has created a carve-out for certain subsidies and that its reasoning and conclusions are in various respects ‘exceptional’ (or, more bluntly, incorrect).
With this premise, clearly the first-best scenario would be to simply ignore this ruling. It was a case-specific solution to a case-specific problem. The argument, mentioned above, whereby the decision would be justified by various peculiar facts and circumstances, leads in the same direction.
We believe this is unlikely. As suggested, it is rather highly likely that this decision, which has become one of the key rulings in WTO subsidy law, will be relied on. The ruling and its analysis will provide ammunition to those that will consider them favorable to their litigation and strategic position.
It must mean something that virtually all commentators that have criticized the Appellate Body decision have also called for its clarification and narrowing. After noting that the ruling may open the way to trade-distorting subsidies for inefficient production technologies, Pal (Reference Pal2014) concludes with an important gloss: ‘[t]he Appellate Body may therefore find itself in a position in the future where it needs to rein in its holdings in these disputes'.Footnote 64 Cosbey and Mavroidis (Reference Cosbey and Mavroidis2014) equally highlight the risk that the rulings ‘open the door to industrial policy unlimited’ and accordingly call for the need to impose some limits to the Appellate Body's findings, especially those concerning the creation of the market scenario.Footnote 65 In the same vein, Cosbey and Rubini (Reference Cosbey and Rubini2013) hint at the need for the Appellate Body to impose ‘rigorous conditions’ on the shelter it has created.Footnote 66 Rubini (Reference Rubini2014a, Reference Rubini2014b) notes that it is important that the flexibility granted by the Appellate Body is not abused. Hence, ‘[t]o avoid undue distortions, we also need to keep compensation at the minimum level necessary to achieve the goal’.Footnote 67 In practical terms, this would mean imposing conditions of transparency, proportionality, and also the use of bidding procedures.Footnote 68
If a clarification and narrowing seems to be highly desirable, we can now elaborate on what this could mean. The reader needs to keep in mind the various uncertainties left open by the Appellate Body (as discussed in Sections 3 and 4).
Most of – if not all – the attention about clarifying and narrowing in the commentary so far has focused on the market creation carve-out. There is a good reason for this. The consideration of the supply-side in the definition of market subsidy is simply wrong, and, as such, it looks extremely difficult to advance adjustments that could make this option ‘less wrong’. The only, minor, piece of advice we could give is: ‘if you want to do it, then do it properly!’Footnote 69 This inevitably means that, for the reasons outlined above, supply-side considerations are irrelevant when it comes to defining the relevant market in a subsidy scenario.
It is indeed with respect to the creation of the market carve-out that there seems to be more scope for clarification and narrowing. On the one hand, we should seek a clear and firm definition of the circumstances and conditions that would allow governments, businesses, and adjudicators to identify a scenario of market creation. On the other hand, clear conditions ensuring the proportionality and transparency of the transfer of economic resources should be introduced.
We delve into the details, and challenges, of possible improvements in the law in the next section.
8. ‘Many will try to enter and will not be able to’: will the Appellate Body find the narrow gate?
Matthew told us that many follow the wide road. In the corresponding passage of Luke's gospel, we find a complementary statement: ‘Make every effort to enter through the narrow door, because many, I tell you, will try to enter and will not be able to.’Footnote 70 For centuries, religious and moral wisdom has taught that taking the narrow road is more difficult. A narrow gate is harder to pass through, a narrow road is less pleasant to follow.
The key question for us (and for this section) is whether the Panels and, especially, the Appellate Body will find the ‘narrow gate’ and thus take the narrow road. Translating the moral discussion above to the legal plane, the wide gate (that is a broad reading of the Canada–Renewable Energy/FIT ‘exception’), by offering more ‘policy space’ to subsidizing governments, does not require unveiling the weaknesses of the law, and would also presumably meet Members’ acceptance.Footnote 71 To step through the narrow gate would, by contrast, require adjudicators to shamelessly expose the deficiencies of the system (that is certain, desirable measures may be caught and even objected by subsidy disciplines), and force Members to recognize them and react.
But the difficulties of the narrow road may simply lie at more technical level. It is interesting to speculate whether the formulation of the Altmark conditions, and the EU experience in their application, can offer some guidance. Let's repeat them again. Under EU law, there is no aid if: (i) there is a clear entrustment of public service, (ii) the compensation is defined beforehand and in a clear and transparent manner, (iii) the compensation is necessary, and (iv) the enterprise is chosen via public procurement procedures or, in the absence of this, the compensation is the one that a ‘typical undertaking’, similarly placed, would need to perform the public service.
We now review each of these conditions and consider what lessons they could give to the WTO.
It is worth immediately highlighting to the non-EU lawyer that the Altmark decision, passed in 2003 after much controversy and fluctuations in the previous case-law, has generated much debate and two reform processes with extensive public consultations. The Commission, in an effort to clarify this important decision, which drew a new line between aid and non-aid, embarked in a first reform in 2005Footnote 72 and, more recently, a second one in 2011.Footnote 73 It adopted various soft-law instruments which inter alia attempt to clarify the impact of Altmark. One of the crucial areas of contention has been the fourth condition, and in particular the criteria for the selection of the provider.Footnote 74
This very brief account of the complexity of the debate on the relationship between state aid control and the use of public procurement attempts to pinpoint one simple fact. Far from being a panacea, the Appellate Body's reference to ‘price-driving mechanisms’ may – inevitably – have opened Pandora's box. The goal of this article is merely to highlight some of the issues that may come out of, and not to provide a comprehensive account of, what in Europe has become a significant body of the law.Footnote 75 The big question now is whether the necessary clarification can be left to future judicial decision-making and to its inherently piecemeal and erratic nature, or whether more comprehensive and legitimate law-making processes should further define and clarify the law in this respect.Footnote 76
The first Altmark condition refers to a clearly defined scenario: the recipient firm must have been entrusted with a public service obligation which justifies the compensation of the relative costs out of public purse. This entrustment does not need to be formal but it certainly needs to be expressed, what matters is the content of the act and not its form.Footnote 77 This is the first and main condition which makes the relevance of Altmark and its quid pro quo approach unique, and, to a large extent, exceptional. In other words, Altmark is not a general test for any type of payment of public money to firms. It only applies for the compensation of public service obligation costs to firms that are specifically entrusted. In other words, Altmark has always been perceived as an exception, not the norm, in EU State aid law. Since it results in removing measures from State aid control, it has been subject to careful and restrictive interpretation.
Most importantly for our analysis, the condition of entrustment is not generally subject to much contention – or, in any case, any question arising out of it can be clearly and satisfactorily answered in the ordinary process of clarification of the law. If we now move to the WTO, it is clear that the definition of the ‘market creation’ scenario, which should provide a clear and limited scope to the Appellate Body's carve-out, is far more difficult. Admittedly, we are at a loss if we try to think at how this scenario could be better defined and formulated having in mind the need to guide the behavior of governments and businesses (and adjudicators).
The second condition, which codifies the requirements of good administration and transparency, should arguably be inherent in the reference to price-discovery mechanisms, and, more generally, in any test that wants to ensure that no benefit is given. It is also arguably of paramount importance also for another, crucial reason.
All the tests we are discussing are about drawing the line between what is and what is not a subsidy. As noted, there are important consequences in terms of transparency and governance depending on the subsidy qualification. It is only if you have a benefit – and hence a subsidy – that you need to notify the measure to the WTO. If there is no benefit, there cannot be any subsidy under the SCM Agreement, and consequently there is no notification duty. It is therefore clear that to require that the procedure that may lead to exclude the existence of an economic advantage be transparent would also partly serve – although in a different way – the pervasive need for transparency in subsidy control.Footnote 78
The third condition (necessity) corresponds to the main thrust of the Appellate Body decision in Canada–Renewable Energy/FIT. It, however, refers to a very ambiguous concept.
To say that something needs to be necessary does not say much and leaves a lot –too much – unspecified. In particular, the necessity determination turns on the two terms of the relationship. What costs should be covered? Using the example of energy: any cost of energy production? Or – rather – only the additional costs to generate renewable energy (i.e. those costs that are extra as compared to conventional energy production)? Can the compensation cover every cost that is indicated in the bids, or only the costs of the most efficient firms in that sector? More fundamentally, what should the compensation be necessary for? To acquire energy or electricity? Or to buy green energy or electricity? Or even to purchase green energy by using a certain green technology? These questions touch the fundamental problem. From a subsidy perspective, price-driving mechanisms can really exclude any benefit only if public policy considerations are expunged from the procurement process to completely align the contracting body's action with that of any other private agent in the market. We will go back again to this fundamental assertion.
The fourth condition, and in particular the reference/incentive to public procurement procedures, has been discussed extensively, and is arguably the most significant area of overlap between EU and WTO law in this respect.
In principle, what the Appellate Body could do in this respect is to be clear and ensure that the design and execution of the procedure should be such as to achieve its result of cost-efficiency and low distortions.
There are two types of issues here.
The first concerns the possibility that ‘price-discovery mechanisms’ achieve the desired results. A tender may lead to very different results depending on its design and, most crucially, it does not necessarily lead to the provision of a service at the least possible cost.Footnote 79 One does not need to think of possible collusive behavior to reach this conclusion.Footnote 80
The second type of challenge focuses on the various issues that need to be conceptually and practically clarified to ensure that the act of selection of the provider does not confer a benefit – which is the concern of subsidy control.
What does ‘lowest cost’ mean for the Appellate Body? Does it refer to the ‘lowest price’ which in the EU is deemed to ‘obviously satisfy the fourth Altmark condition’? Or does it rather refer to the ‘most economically advantageous tender’ where also non-economic (say social or environmental) criteria may be considered? This distinction was very sensitive in the EU and caused much debate during the genesis of the European Commission Communication (European Commission, 2012a).Footnote 81 The issue, however, is that, strictly speaking, only a criterion that purely and exclusively refers to the lowest cost at which the market can offer the good or the service can ensure, in a subsidy context, a no-benefit and no-subsidy determination.Footnote 82
In other words, while public procurement may come to the aid of subsidy control, the focus is slightly different. The former's goal is efficiency, in the sense of public-expenditure reduction. The latter's objective is the exclusion of competitive advantages. In other words, in the context of the benefit analysis, the use of price-discovery mechanisms is of help only if it produces an outcome that is, at least, as competitive as that of normal market transactions.
That is why there is a fallacy in saying that the ‘mere compensation’ of costs does not give any benefit. This is far too indeterminate. The key economic issues – underlined in the analysis of ‘necessity’ above – are left unresolved. Sykes and Neven (2013) rightly note that these procedures, if properly designed, may reduce the rents of the bidders. But, when a subsidy determination is at issue, what matters is that the rents are not simply ‘reduced’ but rather that they are equal to those that the market would consider justified.
Consequently, the Appellate Body was right when it started its analysis by underlining the importance of market definition. This finding holds true even when public authorities resort to price-discovery mechanisms to source their needs. One thing is to define the market, quite another to recognize that a procedure may ensure that the price paid is the best (i.e. corresponds to the benchmark) in that market. Market definition and benchmarking are different and complementary steps. One cannot replace the other.
It is worth repeating this important point. It is market definition that defines the universe, the ‘search box’ of where to find the benchmark. It is only within the four corners of this universe that price-discovery mechanisms can be used to ‘create a competition in the market’ and benefit from the results of this contest. If properly designed, these procedures can genuinely lead to a situation where ‘the price paid by the government is the lowest possible price offered by a willing supply contractor’.Footnote 83
If one really wanted to remove any room for an economic advantage, the specifications should be formulated without referring to the source of energy, or process used for its generation, or indeed any other public policy objective, and just plainly refer to what a private operator would search for in the market, in our case: energy or electricity.
We go back to square one, and to the Appellate Body's original sin of having defined the market too much narrowly.
There is, however, some hope. In the Canada–Renewable Energy/FIT case, the first innovation (the definition of the relevant market) is conducive to the second one (benchmarking). The essence of the problem is that, in applying its new approach, the Appellate Body's definition of the market was too narrow and the ensuing benchmarking too vague, and potentially broad. The best service to clarification and narrowing of its law that the Appellate Body could now do is to take the inherent link between the definition of the market and benchmarking seriously.Footnote 84 This would firstly and immediately require performing market definition analysis correctly, that is according to the specific features of subsidies. By contrast, as noted by several commentators to the case, the best service to WTO subsidy control, and the green energy revolution, Members could do is to seriously start talking about law reform.
The Canada–Renewable Energy/FIT decision left us with bad law. This is not good news for governments and businesses alike. This piece kicked off from the premise that it will be difficult to have any momentous revirement, especially in the current regulatory framework that does not provide for any express justification for good subsidies. After outlining the risks that a normalization and expansion of this ruling may pose, we have responded to the call for clarification and narrowing of this case and speculated on how this could be done. The EU Altmark decision of the European Court of Justice, which was certainly on the minds of the EU litigators and whose ethos the Appellate Body embraced by referring to the use of ‘price-discovery mechanisms’, has inspired our speculation about the future of benchmarking. Our exercise has, however, exposed many challenges and difficulties, many of them having already occurred in EU law. The amount of helpful clarification the WTO judicature could offer is limited, and would probably be restricted to taking the link between market definition and benchmarking seriously. This unsatisfactory situation leads us to suggest, once again, that the only solution to the current ‘turquoise mess’ is law reform.Footnote 85