I. The Inherent Ambivalence of Subsidies as Reflected in the Original GATT/WTO System of Control

New WTO rules on industrial subsidies are advocated by Members and experts alike.Footnote ²⁰ Relatedly, over the past fifteen years, regulation of subsidies has played an increasing role in free trade agreements (FTAs) concluded outside the WTO and often expanding on the multilateral disciplines on the matter.Footnote ²¹ Obviously, the growing emphasis on the need for new international trade rules on subsidies does not solely reflect a sudden increase in awareness of their importance, including as a source of trade disputes.Footnote ²² Rather, mounting concerns of their use (and abuse) as a tool for strategic competition between countries and a careful consideration of the role of public aid to address key societal challenges are major determinants for the prospects of reforming existing WTO rules, particularly the Agreement on Subsidies and Countervailing Measures (ASCM).Footnote ²³

Yet, “no question in international trade law is as contentious, and as complicated, as the question of subsidies.”Footnote ²⁴ Part of the reasons are exquisitely technical: the definition itself and measurement of subsidies turn out to be extraordinarily complex. Beyond that, the question is intrinsically political: with subsidies, what is ultimately involved is a confrontation between different ideological, political, and social conceptions of the role of the state in the economy.Footnote ²⁵ The present analysis cannot thus disregard a certain teleology—that is, a discussion of the goals of (both) subsidies and international rules on subsidies control. Besides, a difference in focus regarding the goals can have a practical influence “on how rules actually work.”Footnote ²⁶ An investigation of the enduring shortcomings and possible reforms of the WTO rules on subsidies points to the very reasons for offering subsidies and, at the same time, to the goals of their regulation: “Why do governments in reality offer subsidies (positive question) and why should governments have an interest in preserving policy space to offer subsidies (normative question)?”Footnote ²⁷

In response, the literature does not yield a strong consensus.Footnote ²⁸ For the limited purposes of the present analysis, I only sketch the contours of the issue with general remarks.Footnote ²⁹ International regimes on subsidy control are inextricably linked to the consideration of their effects, which must be understood from at least two perspectives. First, from a trade perspective, subsidies are frequently seen as undermining the market access prerogatives that result from tariffs and other trade concessions: domestic subsidies may directly offset the effects of a reduced or eliminated tariff, producing the same protective result.Footnote ³⁰ Secondly, from a competition perspective, the deepest concern that subsidies raise is their possible negative impact on the competitive process.Footnote ³¹ Economic analysis shows that subsidies can distort market functioning and introduce inefficiencies at various levels.Footnote ³²

Still, subsidies are ambivalent, because they can produce positive effects and pursue legitimate goals too. Public expenditure seems in some way inevitable, and there is a robust economic and political case to maintain it.Footnote ³³ Subsidies are a key instrument of government policy in virtually all States in the world, an instrument that is by no means necessarily inefficient,Footnote ³⁴ and is “not viewed by economists as intrinsically trade-distorting or welfare-reducing.”Footnote ³⁵ As it is known, positive externalities and public goods, which are not internalized in market prices for particular products and services, may justify subsidization.Footnote ³⁶ Subsidies “may create negative international externalities and distort global resource allocation. But they may also represent sensible policy responses to a range of market failures or play a useful role in addressing income inequality.”Footnote ³⁷ Schwartz and Harper express this when they remark that subsidies can correct the market process at least as much as they may distort it. Footnote ³⁸ Also, in practice, subsidies often constitute a favored tool of industrial policy in several states.Footnote ³⁹ Their inherent ambivalent nature explains why it is proposed that any system of subsidy control that “pretends to be balanced should take into account (at least some of) the public policy objectives pursued and positive externalities produced by subsidies.”Footnote ⁴⁰

In essence, the first objective of any international system of subsidy control is discouraging subsidies that can harm foreign competitors either by impairing their market access expectations or by affecting their competitive position.Footnote ⁴¹ But this is only part of the story. A well-balanced system of subsidy control should go beyond the producers’ interests and also recognize the legitimacy of subsidies targeting market failures and pursuing public/societal goals. The framers of the General Agreement on Tariffs and Trade (GATT)Footnote ⁴² reflected subtly on the scope/justification structure.Footnote ⁴³ The resulting discipline on subsidies was ambivalent: “On the one hand, their legitimacy as tools of public policy was affirmed while their capacity to distort trade was also acknowledged. On the other hand, self-help against such subsidies was permitted in the form of countervailing duties, provided that the subsidies caused ‘material injury’ to domestic industry in the importing country.”Footnote ⁴⁴ In a similar vein, GATT revived in the initial recognition by the ASCM of the three categories of non-actionable subsidies (including certain subsidies for research and development, certain types of assistance to disadvantaged regions, and certain subsidies for compliance with environmental regulations)Footnote ⁴⁵ alongside with the rules to control industrial subsidies “as a potential non-tariff barrier to trade or a source of distortions in international trade that may undermine underlying comparative advantages or alternatively undermine tariff commitments.”Footnote ⁴⁶

Importantly, according to the terms of the ASCM, the permissive list of non-actionable subsidies expired in 2000 when the WTO members were unable to agree on extending them. Since then, it was never revised or reestablished. Therefore, “today no subsidy programs are explicitly protected as non-actionable.”Footnote ⁴⁷ Moreover, the WTO subsidies code does not include a general exception rule such as Article XX GATT that allow explicit trade measures, which would be otherwise inconsistent with the substantive obligations of that agreement, to pursue a range of public policy objectives, including the life and health of humans, animals or plants, the conservation of natural resources, and public morals. As a result, with the expiration of the short list of subsidies that were explicitly carved out from the discipline, current rules drastically confine WTO members’ flexibility to use subsidies as tools of public policy, the only limited concession being the more flexible rules Article 27 ASCM provides for developing country-members in recognition of the important role that subsidies can play in the economic development programs of these members.Footnote ⁴⁸

II. Enduring Problems and Emerging Deficiencies of the WTO Subsidies Rules

I. Trade Concerns over SOEs and the Prospect of WTO Reforms

State enterprises have long constituted, and are likely to remain, an important instrument in any government’s toolbox for a variety of economic and societal goals.Footnote ¹⁰⁶ In addition, as explained in the introductory piece to this Volume, governments around the world have lately instructed their infrastructure state enterprises to deliver public goods and services more widely and equitably among the population, not only to minimize the pandemic’s impact, but also to address the economic downturn. The significant extent of state ownership among the world’s top companies,Footnote ¹⁰⁷ SOEs’ quantitative and qualitative transformation, their growing presence in international trade and global value chains,Footnote ¹⁰⁸ and hybrid natureFootnote ¹⁰⁹ create an assortment of legal issues. In particular, the conflation of SOEs’ economic governance with the political and legal order of the state raises the issue of their impact on international trade flows and the competitive process, giving rise to an extensive debate about the appropriate approach to these entities in international trade law.Footnote ¹¹⁰ Different proposals on WTO reforms point at two main problems surrounding SOEs’ commercial and financial operations.

As is indicated, for example, in the Trilateral statement of January 14, 2020,Footnote ¹¹¹ one problem is the application of subsidies rules in the WTO to the state sector; a concern largely caused by a controversial ruling of the then WTO AB that Chinese state-owned banks do not necessarily qualify as public bodies to whom subsidies disciplines apply.Footnote ¹¹² This decision has been widely criticized and was in part qualified by the AB itself in a later pronouncement.Footnote ¹¹³ In related scholarship, several authors have explored its flaws in depth, and some have proposed different solutions, essentially in the (alternative) forms of a revision of the caselaw and additional treaty-making.Footnote ¹¹⁴

Beyond the application of WTO rules on subsidies to SOEs, there is a more fundamental concern (and increasingly demands) that new international trade rules, mainly of prophylactic nature, are needed to address the negative spill-overs from government influence on SOEs and affecting firms from other countries.Footnote ¹¹⁵ The argument goes on that these rules need to incorporate (or be based on) a concept of competitive neutrality that “would in some way requires states to establish a level playing field between state and private capitalism.”Footnote ¹¹⁶ The European Commission’s position is emblematic of such an approach. In its 2021 Communication on “An Open, Sustainable and Assertive Trade Policy”, the EU executive body asserts:

The following analysis examines these two issues separately. With respect to each matter, I first clarify the terms of the problem in relation to existing WTO rules and caselaw. Secondly, I consider whether, and to what extent, deeper FTAs—those that experts designate as models for WTO reforms on the matter—establish new rules that permit to adequately address the trade concerns raised by SOEs’ commercial and financial activities. Specifically, I critically discuss how these treaties operate in relation to SOEs, paying attention to how their drafters have defined such entities not least to how they establish new and specific rules for SOEs (vis-à-vis private enterprises), and defined exceptions.

II. The Application of the ASCM to the State Sector

The application of the ASCM provisions to SOEs has proved highly problematic. In determining the existence of a subsidy, a central issue is the meaning of the term “public body” in the ASCM agreement. The question of how to determine whether an entity is a public body becomes particularly challenging when the entity involved is a SOE. According to Article 1.1. ASCM, a subsidy could be conferred by a government, a public body, or a private body that has been “entrusted or directed” by the government to make a financial contribution. But the boundary between a government and a public body is not crystal clear. In particular, it is controversial whether the two terms are functionally equivalent, or the term “public body” covers something that is not covered by the term “government”.Footnote ¹¹⁸ Clearly, such characterization is key for the subsidy assessment under the ASCM: In many situations, the state confers a subsidy not through the government itself but through entities affiliated with, owned/controlled, or influenced by the government. Thus, SOEs may constitute a ready means for providing subsidies in an opaque way. As a matter of fact, the “SOE-as-provider” problem has raised intense conflict under the ASCM and attracted growing academic and policy debate. Within the limited scope of this article, recall that in the US-Countervailing Measures (China) the AB considered that a “public body” determination involves an assessment of the core features of the entity concerned and its relationship with government, particularly, whether the entity exercises authority on behalf of government.Footnote ¹¹⁹ Despite a certain ambiguity of the legal standard as formulated by the AB, as well as practical reasons for interpreting the rules on subsidies in a more adaptive way, the adjudicatory body of the WTO has subsequently reaffirmed the demarcations made by existing rules, explicitly rejecting the US argument that government-ownership may be a dispositive factor to making an entity a public body.Footnote ¹²⁰

The cautious approach taken by the AB was arguably a justifiable application of the law: the notion of “public body’” was understood as an entity that, although not institutionally a part of a government, still functions like one. As I argued in a previous work, “[t]his reading had an important corollary: the recognition that SOEs can be run as genuine commercial enterprises and operate on a level playing field with private enterprises.”Footnote ¹²¹ The upshot of this is that SOEs should not be treated differently simply because of majority-ownership by a government.Footnote ¹²² The AB’s narrow interpretation seems also to adhere to the original intent of the drafters, which presumably was not to extend the bound of government too widely. Yet, even if one assumes that this interpretation is correct, it leaves unaddressed the issues raised by SOEs as pass-through vehicles for subsidies. At the heart of the matter are the following questions: What is the standard of proof that a commercial entity is part of the state? Having refuted the formal ownership test, could the AB have clarified what else matters?

Case law shows that the AB insisted on drawing a malleable line, providing a multi-factor test to reply to the questions above.Footnote ¹²³ Still, the additional evidence that may be required is not entirely clear at present. One of the last cases ruled by the AB before its demise in 2020 confirms this conclusion. In US—Countervailing Duty Measures on Certain Products from China,Footnote ¹²⁴ the AB ruled (by majority) that the absence of an express delegation of governmental authority does not necessarily preclude concluding that an entity is a “public body” under ASCM Article 1.1(a)(1), allowing even private entities, under certain conditions, to be considered public bodies.Footnote ¹²⁵ The AB further clarified that a public body inquiry under the same provision does no hinge on the conduct of the investigated entity, but rather on the core features of the entity itself and its relations with government, in light of the legal and economic environment in which the entity operates.Footnote ¹²⁶ Once it is established that an entity is a public body, all conduct of that entity shall be attributable to the Member concerned for purposes of Article 1.1(a)(1).Footnote ¹²⁷ For such an assessment a holistic approach is required. This should take into due account: (i) evidence that an entity is exercising governmental functions, especially if it is a sustained and systematic practice; (ii) evidence of the scope and content of the relevant government policies; (iii) evidence of the meaningful control over an entity by the government; and (iv) whether the conduct or functions of an entity are ordinarily classified as governmental in the legal order of the relevant Member. Accordingly, a public body determination must be conducted on a case-by-case basis.Footnote ¹²⁸

A final remark. “SOEs-as-providers” can be captured by ASCM Article 1.1(a)(1)(iv) that include private bodies which have been entrusted or directed to carry out the functions that constitute potential subsidies (such as transfer of funds).Footnote ¹²⁹ Yet, this application of the ASCM relevant provisions is no decisive solution to the problem. In fact, not all governmental measures vis-à-vis a private intermediary would necessarily amount to “entrustment” or “direction”, as both terms demand a significant degree of command-and-control authority on the side of the government.Footnote ¹³⁰ Moreover, the use of private vehicles poses more of an evidentiary challenge for Panels, as they have to examine how the seemingly private conduct can be ascribed to a government entity, which will not necessarily be eager to disclose that relationship to the rest of the world.Footnote ¹³¹

III. How Deeper FTAs Address the Problems of Subsidies Involving SOEs

How, and to what extent, is the problem of the application of international rules on subsidies to the state sector addressed by FTAs concluded outside the WTO?

Most FTAs have provisions on subsidies.Footnote ¹³² Deep(er) FTAs also have adopted approaches to deal with SOEs.Footnote ¹³³ Comparative treaty analyses suggest that, leaving aside the peculiar disciplines established by EU FTAs that incorporate for both SOEs and private enterprises the EU competition law rules on public aid and the language of Article 106 TFEU, specifying that entities charged with public tasks are subject to competition rules if this does not preclude them from performing their public service obligations, the most advanced formulation of international trade rules on SOEs are contained in two treaties, namely the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP)Footnote ¹³⁴ and the United States-Mexico-Canada Agreement (USMCA).Footnote ¹³⁵ Their approach on subsidies involving SOEs is based on two central elements, namely, the definition of SOEs and a general prohibition of providing state support on non-commercial terms to the commercial activities of SOEs.

As to the scope ratione personae, both agreements include a broad and, at the same time, precisely demarcated definition of SOEs, which details the relevant notion of ownership/control. Article 17(1)(3) CPTPP defines an SOE as an enterprise that is primarily engaged in commercial activities and has one or more of the following features: the contracting party (i) directly owns more than 50 per cent of the stocks of the enterprise; (ii) controls the enterprise through the ownership of more than 50 per cent of its voting rights; or (iii) has the right to appoint the majority of the members of the executive board or any other decision-making body. More comprehensive is, however, the notion of control as embodied in Article 22(1) USMCA, which expands the definition of the CPTPP “by including also indirect ownership of more than 50 per cent of share capital” and—inserting a sort of “decisive influence” test—“the power to control the enterprise through any other ownership interest, including indirect or minority ownership.”Footnote ¹³⁶ A footnote to the same provision further clarifies that this power exists if, through an ownership interest, the government can determine or direct important matters affecting the enterprise, excluding minority shareholder protections and that, in determining whether a party has it, all relevant legal and factual elements shall be considered on a case-by-case basis. Another important definitional aspect is that both trade agreements regulate for-profit institutions only, requiring that SOEs “must be” those “principally engaged in commercial activities.”Footnote ¹³⁷ Thus, the CPTPP and the USMCA adopt clear-cut rules based exclusively on quantifiable proxies, including SOE size and threshold revenues,Footnote ¹³⁸ and eliminate uncertainties by suggesting a binary means of assessing the commercial nature of SOEs activities for the application of the relevant international obligations upon governments. As a result of the broad definitions of SOEs they establish, these treaties address the two problems affecting the relevance and applicability of the pre-CPTPP rules to the current developments in international trade: the lack of unambiguous tests provided to determine ex-ante government control, and an implicit requirement of state action.

Yet, as I argued previously, clear-cut definitions of this sort “may be counterproductive and in effect weaken SOEs disciplines.”Footnote ¹³⁹ A problem with quantitative thresholds defining ownership/control in the CPTPP is that they fail to consider the many other, lower levels of ownership or board appointment that can actually be structured so as to retain governmental control, as well as situations of indirect control. Furthermore, there is no provision on interlocking directorates.Footnote ¹⁴⁰ Also, rigid rules provide actors with incentives to reframe a control/ownership pattern in order to dodge their application—not an implausible hypothesis, given contemporary corporate practices. Additionally, the threshold revenues that SOEs must reach are problematic, given the several inventive accounting techniques that could be used to avoid meeting these minimal requirements. Alongside transfer pricing manipulations, this can result in transfers between intra-group related entities that may not appear as revenue.

There is another element that impacts on the scope ratione personae and ratione materiae of the CPTPP and USMCA rules regarding SOEs, i.e., the exemptions, exceptions and carve-outs included in the relevant texts. Consider the text of the CPTPP. Exemptions regard specific categories of state entities. The extensive number of exceptions and carve-outs included in such treaties confirms that countries have different views on the legitimate role of the state, and which of those purposes should be subject to international discipline. Acceptance of such difference is embedded in the new trade regimes. However, while certain exclusions are quite sensible, (such as, the exemptions of smaller SOEs and government procurement), others are controversial, and may undermine the anticipated benefits of the new regimes. Notable examples are the sweeping exemption for SOEs that a sub-central level government owns or controls; the already mentioned state-owned domestic service providers; and Sovereign Wealth Funds (especially if one considers some Asian countries are actively discussing the potential transformation of their SOEs into SWFs).

The second key element of rules on subsidies involving state enterprises is the inclusion of a general ban on provision of state support on non-commercial terms to the commercial activities of SOEs. Thus, the CPTPP and USMCA are the only treaties that incorporate new rules on non-commercial assistance (NCA), the main aim of which is to prevent adverse effects or injury to the interest of other parties as a result of advantages that SOEs obtain because of their proximity to the government. While rules on NCA are not covered in earlier trade agreements,Footnote ¹⁴¹ these provisions should be essentially considered as an expansion of the WTO-based approach. This is because the new rules espouse the basic principles of the ASCM and, at the same time, “extend those disciplines to cases not covered by the same treaty, such as NCA provided with respect to trade in services.”Footnote ¹⁴² Indeed, many overt and covert advantages SOEs may enjoy over private enterprises boil down to a direct or indirect subsidization by the government. Hence, the new rules prohibit a contracting party from causing an adverse effect or injury to other contracting parties through NCAFootnote ¹⁴³—generally defined as any assistance to an SOE by virtue of its government ownership or control—in purchase, and sale of goods and services and investments.

The CPTPP and USMCA lay out very specific instances where NCA is deemed to adversely affect the interests of another party or cause injury to its domestic industry, making investigations of the violation easier. Indeed, the two treaties define these conditions more narrowly than the ASCM.Footnote ¹⁴⁴ Despite the fairly lengthy character of these provisions, the bottom line is that the substantive discipline does not generally inhibit NCA as such. Instead, the new rules aim at controlling certain negative effects that may arise from such assistance.Footnote ¹⁴⁵ If a given form of NCA is causing adverse effect or injury to another party, the providing country needs to take appropriate action to remove those effects or withdraw the assistance concerned. Otherwise, it will be confronted with the countermeasures permitted under the two treaties’ dispute settlement procedures.Footnote ¹⁴⁶ Central in the new regime, is the use of the word “indirectly” to qualify the provision of NCA to SOEs.Footnote ¹⁴⁷ As is explained in a footnote to Article 22(6) USMCA, which echoes Article 1.1(a)(1)(iv) of the ASCM, “indirect provision includes the situation when a Party entrusts or directs an enterprise that is not a state-owned enterprise to provide non-commercial assistance.”

Even more meaningfully, both treaties clarify that SOEs, just like states themselves, are prohibited from providing NCA to other SOEs when it causes adverse effect or injury to another party.Footnote ¹⁴⁸ Therefore, SOEs cannot serve as pass-through vehicles for direct and indirect subsidization to other SOEs, even when they would not match the definition of “public body”under the current interpretation of ASCM by the AB. Thus, the prime benefit of the new NCA rules is their application to cases where governmental assistance slips through the net of ASCM. Another such case is when SOEs engage in trade of services: in the absence of WTO comprehensive subsidy disciplines connected to trade in services, the impact of the CPTPP and USMCA rules in this area is, for the respective parties, potentially vast. However, the NCA rules do not apply to a service supplied by a SOE within the territory of the subsidizing party.Footnote ¹⁴⁹ A result of intense negotiations, this compromise muddles all domestic services, irrespective of whether they have public goals or not, and seems to flatly contradict free market principles and the objective of the international contestability of markets.

Importantly, as in the GATS, the NCA rule does not apply to services supplied in the exercise of governmental authority.Footnote ¹⁵⁰ Clearly, the very definition of services supplied in the exercise of governmental authority, which supposedly safeguards the sovereign power of states to carry out public functions, reveals the neo-liberal vision of a laissez-faire economy, (minimalist) state, and society. Moreover, the prohibition of NCA to SOEs does not apply to services that are supplied neither on a commercial basis, nor in competition with one or more services suppliers. Similar exemptions have attracted intense criticism for its limited scope in that many predominantly government-provided services, for example health care, education, and water distribution, are provided in actual or potential competition with private providers, and so may fall outside its scope. Not surprisingly, the proposed text for the Trade in Services Agreement (TiSA),Footnote ¹⁵¹ which incorporates the same exemption in Article I-1:3, has been criticized as a tool to further reduce the area of public services. For these reasons, Uruguay eventually withdrew from the TiSA negotiations.Footnote ¹⁵²

As for the procedural provisions, the CPTPP and USMCA rules do not contain a mechanism of countervailing duties. Also, they do not clearly indicate an alternative approach. For example, they do not incorporate a mechanism somewhat akin to the undertakings system envisaged under ASCM Article 18. In essence, the complaining party is given wide discretion and control over all the counteracting processes.Footnote ¹⁵³

To conclude, NCA rules are not without problems. Even leaving aside the issue of the nature and relevance of exceptions and carve-outs, which has been widely explored in the literature,Footnote ¹⁵⁴ some significant lacunae remain under the NCA rules. While the CPTPP rules seek to cover NCA that may remain uncaptured by the ASCM (e.g., assistance provided by a state enterprise not vested with governmental authority), they do not contemplate the situation where the assisted exports adversely affect the industry of another party operating within that party’s domestic market, i.e., the case that would be subject to countervailing duties under the ASCM.Footnote ¹⁵⁵ Thus, if assistance is provided in a form that does not fulfil the definition of subsidy under WTO law, the importing party may have no means to address the adverse effects upon its domestic industry caused by assisted exports.Footnote ¹⁵⁶ Moreover, the CPTPP and USMCA rules do not cover situations where assistance is provided by SOEs to private enterprises. Hence, the “SOE-as-provider” problem is only partially addressed. Under the two agreements, these entities could still represent means for providing subsidies in an opaque way where the final recipients are private enterprises and the ASCM does not apply. Finally, some of the NCA provisions in the two FTAs are quite complex and introduce for the first-time central concepts such as the “adverse effect” requirement and the “cause injury” standard. Besides, there are further provisions that provide exceptions to those injury standards.Footnote ¹⁵⁷ This reduces the utility of reference to past WTO practice. As history of the ASCM demonstrates many central terms and methodologies were clarified and developed through the interpretation of the Panels and the AB. The same is likely to be true for the CPTTP and USMCA. For this reason, the new NCA rules require the concurrent development of a functioning dispute settlement mechanism. If this is not achieved the SOE regulation under such treaties will result in a lex imperfecta,Footnote ¹⁵⁸ “with individual parties interpreting more obscure terms in accordance with their preferences and interests.”Footnote ¹⁵⁹

IV. Incorporating "Competitive Neutrality” in the Legal Order of International Trade

To ensure that the decision-making of state trading enterprises (STEs)—state enterprises or private enterprises operating under state-conferred monopolies or privileges—would be subject to non-discrimination norms, the drafters of the GATT included a provision, Article XVII, that frames non-discrimination in terms of such enterprises basing their purchase and sales decisions on “commercial considerations.”Footnote ¹⁶⁰ In addition to Article XVII, as one GATT panel stated, “the note to Articles XI, XII, XIII, XIV and XVIII provided that throughout these Articles the terms “import restrictions” and “export restrictions” include restrictions made effective through state-trading operations.”Footnote ¹⁶¹

The GATT acquis is arguably of limited use in regulating the magnitude of today’s state capitalism.Footnote ¹⁶² First, STEs are only a subset of SOEs.Footnote ¹⁶³ Secondly, not only is its application restricted to the above-referred commercial transactions of subset of SOEs, but it seems also marked by some uncertainty with respect to the principle of national treatment.Footnote ¹⁶⁴ Thirdly, case law has been weakened by the finding that it suffices for STEs to act in a non-discriminatory manner to comply with the provision.Footnote ¹⁶⁵ Thus, whatever the precise meaning of “commercial” here, Article XVII “has been interpreted by the WTO Appellate Body as establishing non-discrimination as an obligation of result in respect of “state trading enterprises.”Footnote ¹⁶⁶ While it is questionable whether in so doing, ipso facto, STEs act also in accordance with commercial considerations and afford competitors adequate opportunity to compete for participation—and economic logic would not support this view of the ABFootnote ¹⁶⁷—this interpretation is by now water over the dam, as “there is not one single deviation from this case law”.Footnote ¹⁶⁸ Finally, according to the then WTO AB, GATT Article XVII does not apply to an STE’s transactions when no foreign enterprise is directly involved.Footnote ¹⁶⁹

GATT Art. XVII marked the beginning of international trade regulation on state enterprises which has lately evolved into complexed free trade agreements that contain far-reaching obligations on non-discrimination and non-commercial considerations. The latter kind of obligations incorporate the “competitive neutrality principle” advocated, among others, by the European Commission. I discussed this concept as a normative foundation of new international trade rules on SOEs in an earlier work. In that context, I argued that competitive neutrality, is “a highly condensed form of rhetorical material” that “has come to seem real and part of legal routine in current trade negotiations.”Footnote ¹⁷⁰ Moving from a critical approach to the debate surrounding this notion, I eventually concluded that in order to fully capture the approaches of contemporary FTAs to the regulation of SOEs “[p]articularly clarifying is the recognition that there are two goals at work […] simultaneously. On the one hand, negotiators seek the creation of the conditions for internationally contestable markets, while simultaneously permitting domestic constituencies to maintain sufficient policy space for using SOEs to contribute to important public goals.”Footnote ¹⁷¹ Here, it is important to add that, as evidenced by Howse’s systematic engagement with relevant GATT/WTO caselaw, “under existing GATT rules on non-subsidies measures claimants have often been successful in obtaining rulings against protectionist discrimination in the policies or practices of state enterprises or government treatment of them.”Footnote ¹⁷²

Moving from these different orders of considerations, I now turn to assessing how and to what extent the introduction of “competitive neutrality” into the legal order of international trade in the form of new rules on non-subsidies measures would be an adequate alternative to existing multilateral rules in addressing ways in which government measures concerning SOEs create barriers to international trade of a protectionist nature that are not justified by legitimate non-protectionist public policy objectives.

The obligations of non-discrimination and commercial considerations are present in the CPTPP and USMCA. In subjecting SOEs to such requirements, the new rules are plainly rooted in the GATT rule on STEs. However, they significantly expand on such discipline. First, in both the CPTPP and USCMA, obligations extend to trade in services and investments from other parties.Footnote ¹⁷³ Second, with a view to favoring the international contestability of markets, they also cover SOEs’ activities within domestic territory. Third, the new rules dissolve any uncertainty with respect to the principle of national treatment, which is now explicitly incorporated, avoiding the confusion that exists concerning GATT Article XVII. Fourth, the same provisions, as they read, reverse the AB’s interpretation of the same rule. As observed supra, the AB had long interpreted the duty of non-discrimination treatment and the commercial considerations rules as illustrative of each other. In contrast, like other deep FTAs, the CPTPP and USCMA put such obligations in two distinct provisions with no logical link between them.Footnote ¹⁷⁴ A literal and contextual reading of these provisions would command that an SOE cumulatively satisfy both requirements in order for the party to comply. However, here one may notice a first ideological bias against SOEs: “Why should trade law be concerned with whether state enterprises act commercially as a general matter, apart from situations where the non-commercial behavior reflects an alteration in the conditions of competition that disadvantages imports?”Footnote ¹⁷⁵

Moreover, in defining “commercial considerations” as equivalent to those factors that a privately-owned enterprise in the same business or industry would also consider,Footnote ¹⁷⁶ the CPTPP and USMCA incorporate a standard, absent in WTO law, that pretends to determine in abstracto whether a firm’s buying or selling behavior is commercially sound. Here again, a seemingly ideological bias against SOEs shines through. “Why would the decision-making considerations of private firms be the baseline against which state enterprises should conform themselves?”Footnote ¹⁷⁷ As Joseph Stiglitz argues, “the distinction between competitive and non-competitive (including bureaucratic) forms of economic organization may be more important than the distinction between private and public ownership…”.Footnote ¹⁷⁸ Accordingly, both public and private enterprises depend on managers/agents to achieve the goals of their principals, the owners. Different types of public and private enterprises exhibit different incentive structures for managers, some of these addressing agency costs better than others,Footnote ¹⁷⁹ but, ‘”in any case, there is no systematic evidence of distortion of global competition from the way in which agency costs are managed in the incentive and monitoring structures of public firms as opposed to private ones.”Footnote ¹⁸⁰

There are however elements of the new disciplines that appear to relax the standards of non-discrimination expected of SOEs. For example, insofar as the obligation to act in accordance with commercial considerations is concerned, the CPTPP seems to allow SOEs the right to sell at different prices and even allows them to refuse to sell, which at first glance seems to authorize violation of the MFN and national treatment principles. However, contrary to what has been claimed elsewhere,Footnote ¹⁸¹ my take is that these provisions do not introduce a systemic loophole in SOE rules. Instead, they seem to reduce the bias against SOEs: privately-owned enterprises, by definition, are not subject to a duty of non-discrimination. In fact, as for ownership-neutrality in the GATT/WTO, it is only the motivation and behavior in the market that matters. If a state-owned firm behaves in a commercial manner, the argument might go, its ownership structure would be irrelevant. Hence, like its privately-owned competitors, it should enjoy contractual freedom.

In my view, the main technical problem with the obligations of commercial considerations remains that there is no clear-cut economic test to determine whether a firm’s behavior is commercially rational, despite any effort to provide clarifications. Actions such as selling at very low prices to hook customers can be practiced by commercially motivated firms and those with further motives alike.Footnote ¹⁸² As I have pointed out elsewhere, “[i]t would be easier to test between conduct and motivations if there were a clear alternative hypothesis: what, precisely, is it that SOEs would do if they did not act commercially?”Footnote ¹⁸³ The answer is likely to vary with the relevant market and business. Given the silence of the legal texts and the likely complexity of relevant cases, the examination of single instances of conduct would greatly benefit from the progressive stratification of interpretative practices (read: case law) at either the international or domestic level. However, at this time, neither agreement is endowed with a functioning dispute settlement mechanism, and to divine whether they could ever effectively work is impossible. Further, there is another obstacle for new rules on state enterprises to go further than the WTO acquis. Commercial rationality in a state-centered economy that needs to shoulder various transitions and crises cannot arguably exist in a uniform way. It is almost by necessity a matter of perspective.

I. Tensions Regarding the Future Design of Subsidies and SOEs Treaty Regulation

The tensions that now exist regarding the future design of subsidies and SOE treaty regulation are usually portrayed as part of a political and ideological contest. And, indeed, deciding on the international trade rules that should govern these prominent forms of state intervention in the economy raise “fundamental questions concerning the nature and degree of government involvement in commercial affairs and the right of other governments to inquire into that involvement.”Footnote ¹⁸⁴ Central tenets of a state economic system and basic beliefs concerning its governance are radically involved under the cover of technical notions like allocative efficiency, contestable markets, market failures, or regulatory neutrality. Against this backdrop, it is hardly surprising that decades of negotiation to reform the GATT/WTO system to address subsidiesFootnote ¹⁸⁵—with the recent supplement of growing demands to deal with trade-distorting effects of SOEs—“have never produced the sort of uniform, scheduled rules that apply to tariffs or even services.”Footnote ¹⁸⁶ Instead, WTO members developed rules to demarcate between tolerable and intolerable subsidization, and regulate self-help against such subsidies that are presumed or proved to be impermissible.

In this paper, I argue that existing WTO rules on subsidies and state enterprises—which represent a necessary piece of the interface mechanism designed to allow different economic systems to trade together without excessive tensions—are not suitable to ease, to say nothing of solving, recent controversies caused by the resilience of the state as an economic actor and the related increasing politicization of trade relations. Moreover, I submit (and demonstrate) that this system of rules is poorly equipped to assist states in addressing the substantial challenges they face in the post-pandemic world. In fact, extant rules on state intervention in the economy may make it more difficult for WTO members to engage in industrial policies that are needed to deal with the economic consequences of the COVID-19 pandemic, support desired changes in the real economy, and address the climate change shock. My approach is openly analytical: I explain and discuss existing multilateral rules and relevant caselaw on subsidies and state enterprises in light of the present geopolitical shifts, emerging changes in global economy, and environmental concerns. The ultimate goal is to dissect the enduring problems and emerging deficiencies of this system of rules, and, hence, lay the ground to investigate the direction taken by future possible reforms. However, if we now venture beyond the analytical account to the normative, some concluding considerations can be drawn from the foregoing work in relation to the main problematic issues explored and related prospects of reform.

II. Expanding, Reinforcing and Refining Extant WTO Rules on Industrial Subsidies

The first order of considerations is about the expansion, strengthening and refinement of existing multilateral trade rules on industrial subsidies. Some scholars view disciplines beyond rules preventing the nullification and impairment of tariff commitments as both needed and feasible. Exemplars of this assessment of the future position of subsidies disciplines are two works by distinguished international trade experts: Chad Bown and Jennifer Hillman,Footnote ¹⁸⁷ on the one hand, and Gary Horlick and Peggy Clarke,Footnote ¹⁸⁸ on the other. At the risk of obscuring different gradations of their analysis, key “common themes emerge in their proposed reconfiguration”Footnote ¹⁸⁹ of extant WTO rules on subsidies, ranging from expanding the category of prohibited subsidies by including, inter alia, subsidies to the fossil fuel industry that aggravate risks of climate change; extending the scope of public bodies in the ASCM so as to include SOEs; reinstalling a green light category of non-actionable benign subsidies; and finding new institutional arrangements for enforcing subsidies rules.Footnote ¹⁹⁰ These studies offer very sensible suggestions, some of which are clearly mirrored in the recent EU proposal.Footnote ¹⁹¹ To be fair, one cannot disregard that, although recent, the two articles at issue were published in a period where the crisis of the WTO did not appear as severe as it is today. As widely discussed in different contributions to this Volume, the situation has dramatically changed over the last few years. At present, in few other areas, divergence between the major trade powers as to the priorities accorded to the subject of reform is as profound as in the area of subsidies.Footnote ¹⁹² Moreover, in the aftermath of the financial crisis and the Covid-19 pandemic shock, the struggle for what counts as normal relations between state and market has taken new colors and intensity—with the United States “steadily moving from a commitment to economic and political liberalization in its foreign policy”,Footnote ¹⁹³ and new paradigm of industrial protection spreading among other WTO members.Footnote ¹⁹⁴ Therefore, in the current geopolitical setting one cannot be but skeptical about the prospects of any consensus within the WTO on new multilateral rules on subsidies in the foreseeable future. As I observe above, any modification of the current rules necessarily entails the re-definition of the confines between illegitimate state intervention in the market and legitimate, hence “normal”/normalized, level of state actions.

This is why authors like Hoekman and Nelson view(ed) the launch of a broader international work program on subsidies in the 12^th WTO Ministerial Conference as a necessary preliminary step to—at least this is the second step I can safely infer from/add to their analysis—subsequent plurilateral negotiations, which, in turn, could respond to the consensus problem. In that perspective, the WTO would provide a platform “to members willing to invest resources into a work program to compile information and analyze existing subsidy programs in systemically important economies,”Footnote ¹⁹⁵ including in sectors where multilateral rules do not exist (e.g., services). This program would build a more solid evidence base on the magnitude, purpose and effects of subsidy policies and remedy to a significant (and widely acknowledged) information gap about the use of these tools by WTO members. With that evidence at their disposal, members would be able to identify shared objectives and mutual gains in relation to possible reforms,Footnote ¹⁹⁶ which, most likely, could take place through plurilateral agreements. Regrettably, such an approach was not considered during the last Ministerial Conference, the result being the remaining focus on piecemeal efforts like unilateral action (e.g., the EU regulation on foreign subsidies) and FTAs subsidies rules (some elements of which are portable at the multilateral level, while the overall disciplines are too narrow in scope and focus).

Still, I argue that an international form of cooperation promoted by and within the WTO to gather and analyze information on subsidies in systemically important (and cooperative) economies could effectively assist also other responses to the consensus problem on multilateral reforms. This is the case, for instance, of the proposal advanced by Michael Trebilcock who posits that, within the multilateral system, “there is a strong case for abandoning one-size-fits-all subsidy disciplines for all member countries across all economic sectors.”Footnote ¹⁹⁷ Rather, he favors a more flexible system, viz. an approach “that draws on the negotiating modalities adopted in the General Agreement on Trade in Services (GATS).”Footnote ¹⁹⁸ Such a system, transposed to international trade in goods, would entail member countries of the GATT/WTO having the latitude of opting-in to different subsidies disciplines on an à la carte basis. Commitments on the use of subsidies “(akin to opening offers in tariff negotiations) would form the basis for future international negotiations on subsidies disciplines, along with appropriate tariff adjustments.”Footnote ¹⁹⁹ In case “of agreement (which, like tariffs, may not necessarily be symmetric), countries would be committed to not subsequently adopting or expanding their subsidy policies at variance with negotiated commitments, causing or threatening injury to producers in another country in domestic or foreign country.”’Footnote ²⁰⁰ Clearly, an approach of this kind would more closely align subsidies rules with tariff commitments, and complement rather than undermine them. Further, it would ultimately determine a graduation of WTO members’ rights and obligations regarding subsidies through progressive regulation.

III. Reforming the ASCM to Address the Climate Change Shock and Green Industrial Transition

The second concluding remark concerns the climate change shock and the role of public spending to favor industrial green transition. Extant WTO rules that divide all subsidies into either prohibited or actionable categories are no longer fit for purpose. Particularly for the purpose of supporting WTO members in levelling the competition between fossil fuels and renewable fuels and make the needed transition to a decarbonized economy. Unless a close connection between climate change policy and the public aids employed is established, climate change subsidies may conflict with both GATT Articles VI and XVI, and the ASCM. Technically, modernization of rules in this area are not hard to conceive. There is wide agreement among experts that WTO rules need to be re-examined because of their inadequate flexibility for public policy.Footnote ²⁰¹ In this article, I maintain that international trade rules on subsidies should be conditioned on what governments are planning to do, implying determining what goal they pursue, and differentiating economic from non-economic objectives. Theoretically, a sensible way to allow this sort of assessment would be the introduction in the ASCM of a general exceptions provision, akin to GATT Article XX, which relates to a range of legitimate public policy goals, including human, animal, and plant life and health, and the conservation of natural resources.Footnote ²⁰² As observed above, an alternative option is the reinstallation of the category of benign permitted subsidies, including subsidies to renewable energies. Similar considerations about the technical feasibility of reforms in this area can be readily extended to fossil fuel subsidies with negative climate change externalities. But, again, the core problem is to reach consensus at the multilateral level.Footnote ²⁰³ This is accentuated by the refusal of some WTO members to separate reforms on subsidies to ease green transition and other changes in multilateral trade rules that may be equally or even more difficult to achieve.Footnote ²⁰⁴

Late developments regarding fossil fuel subsidies are telling of existing challenges. At the 11th WTO Ministerial Conference in Buenos Aires in December 2017, twelve of the 164 members of the WTO launched an important initiative that could help commitment to fossil fuel subsidy reform under the Sustainable Development Goal 12 (c) of the 2030 Agenda. On this occasion, they issued a declaration expressing their desire to add specific disciplines to the existing WTO subsidies rules to rationalize and phase out inefficient fossil fuel subsidies. Their initiative is “aimed at achieving ambitious and effective disciplines on inefficient fossil fuel subsidies that encourage wasteful consumption.”Footnote ²⁰⁵ These dozen members have since embarked on a diplomatic campaign to enlist the remaining 152 WTO members in new negotiations on how the trade rules should be amended to address the climate change shock. In December 2021, 19 WTO members issued a similar statement that reiterated the urgency of accelerating fossil fuels subsidies reform and promised to “elaborate concrete options” in the Ministerial Conference ahead.Footnote ²⁰⁶ Yet, no real step forward emerges from the official reports of the 12^th Ministerial Conference.Footnote ²⁰⁷ And the whole membership of the WTO has yet to place this issue on the negotiating agenda. Adding it to the agenda is indeed highly problematic.Footnote ²⁰⁸

IV. New Treaty Rules on Subsidies to the State Sector?

The third order of considerations is about appropriate reforms to address subsidies to the state sector. As I explain above, some recent FTAs contain more detailed rules relating to SOEs as either originator or recipient of government subsidies (rules on NCA). The CPTPP and USMCA, often regarded by international trade scholars as models for WTO reforms on the matter, exemplify such rules. I only recall here that these rules are not without problems, starting with the very definitions of SOEs they establish and the way exclusions and exceptions are made. Moreover, I doubt that transposing the CPTPP or USMCA rules on NCAs—and the related definitions of SOEs—at the multilateral level would adequately address the problem of state enterprises as providers of government subsidies. In the CPTPP these rules were inserted with China in mind. But, for one thing, as Mark Wu and Ming Du show, while China resorts to SOEs as tools of policy, its economic system is vastly more complex. There, the state has the capacity to model market behavior, even that of firms that are entirely private, through complex governmental and Communist Party networks hard to detect from outside.Footnote ²⁰⁹ For another, the exceptionalism of China “should not drive a general (…) tainting of state ownership or control of business activity”Footnote ²¹⁰ in an organization whose membership count more than 160 States, especially when, in the aftermath of the last financial crisis, and in the wake of the Covid-19 global pandemic, the state has taken an even greater role as an investor in several countries.

Rather, I subscribe the view that a more adequate approach to the concept of “public body” in the ASCM is possible and opportune.Footnote ²¹¹ Importantly, in US—Countervailing Duty Measures on Certain Products from China, the then AB eventually qualified its previous caselaw and demonstrated openness in this respect.Footnote ²¹² In my view, the concept of public body is able to capture public enterprises where government preserves means of control or influence (for example, through appointment of board members), but the company is a private enterprise in other respects. In all such cases, “a presumption of governmental involvement or influence is warranted where the enterprise is acting in a manner that deviates from the normal market conduct of a competing private firm.”Footnote ²¹³ Whereas Mavrodis and Sapir argues that such a presumption should be included in new version of the ASCM, I fall in with Howse that, in leaving the concept open (also to evolutionary interpretations) and, hence, broad, the drafters of the original GATT and later those of the ASCM display wisdom. They prevented excessive rigidity in a concept the meaning of which can vary over time, and eluded the arguably futile search for a new formula that would more appropriately capture the use of state enterprises as a tool of government policy.Footnote ²¹⁴ It is sensible for the WTO to “defer to the decision of domestic agencies” that may presume, in economic systems like that of China, or where the peculiar context surrounding the activities of a given state enterprise suggests so, “that certain conduct has been influenced or guided by the state, even in the case of enterprises that are not formally charged with exercising governmental authority.”Footnote ²¹⁵

V. Ensuring the Equality of Competitive Conditions

Finally, as to the additional obligations of commercial considerations included in the CPTPP and USMCA to ensure competitive neutrality between domestic and foreign firms—advocated as broader reforms for the multilateral trade system too—I only add a couple of remarks to my previous analysis. First, and noted here only briefly, the institutional designs of transparency rules in the SOEs chapter of the CPTPP and USMCA do not offer a useful corrective to the scarce information we have about the extent to which state action—the only thing subject to international trade rules—alters the terms of competition in relation to the operation of SOEs.Footnote ²¹⁶ In the absence of reliable data, concerns with the potential for negative spill-overs on trade as a result of SOEs’ commercial operations may appear to be ideological rather than empirical.Footnote ²¹⁷ Secondly, the directives incorporated in the obligations of commercial considerations are designed to ensure that SOEs do not alter the terms of competition in domestic markets at the expense of foreign firms.Footnote ²¹⁸ In the absence of compulsory rules regarding property ownership for parties to comply with, rules on commercial considerations reduce the incentives for state intervention in the economy by imposing on SOEs obligations to which private enterprises are not subject. Hence, as I explain above, they reveal a clear ideological bias against SOEs. In fact, the resulting difference in treatment is not substantiated by systematic evidence of distortion of global competition from the way in which agency costs are managed in the incentive and monitoring structures of state enterprises as opposed to private ones. Even leaving this issue aside, if it is the terms of competition the main concern, it remains controversial whether rules of this sort are the best option for future reforms. As Mitsuo Matshusita argues in his article for this Special Issue, this kind of problems should be addressed by competition law (rectius: by an interaction between international trade law and competition rules still to be shaped at the multilateral level).Footnote ²¹⁹ However, in the enduring absence of a WTO antitrust code, the question remains why “competitive neutrality” should be introduced into the trade order through new prophylactic rules focused on state enterprises as opposed to private enterprises instead of relying on a sound application of existing disciplines (particularly, WTO rules on non-discrimination) in cases where government policies toward state enterprises or their conducts effectively distort international trade.Footnote ²²⁰