The European Commission's Glass Fibre Fabrics Investigation and the Boundaries Between Investment and Trade

Abstract

In a 2020 anti-subsidy investigation concerning glass fibre fabric (GFF) products from Egypt, the European Commission (EC) attributed the Chinese government's conduct to the government of Egypt in a way that raised a systemic question about the boundary between trade and investment. This article argues that despite some overlap between the boundaries of these legal disciplines, the notions of trade and investment remain conceptually distinct. Customary rules of interpretation dictate that World Trade Organization (WTO) covered agreements are construed as facilitating trade relations and no further. Hence, an extension of WTO subsidy rules to cover outward investment promotion measures using the principles of state responsibility is untenable. Such a unilateral approach disproportionately affects the interests of developing countries, harming their efforts to draw green investments. This article recommends that new, balanced rules be designed to promote outward investments while limiting adverse trade impacts.

Home states frequently grant incentives to foreign investors independently or in the context of the growing complex network of international investment agreements (IIAs).¹ While the parties involved generally welcome this exercise of sovereign prerogative, the issue remains largely unregulated in international law. However, it has become evident that large-scale grants of incentives for outward investments can distort third-country market trade, especially when such incentives facilitate investments in export-oriented sectors of the host state. The problem is particularly acute regarding worldwide growing Chinese outward foreign direct investment (OFDI) promotion activities.²

The anti-dumping (AD) and countervailing duty (CVD) investigation conducted by the European Commission (EC)³ in 2020 against certain glass fibre fabric (GFF) producing firms in Egyptian and Chinese special economic zones brought this problem to the forefront. Taking it as a test case to tackle Chinese Belt and Road Initiative (BRI) projects,⁴ the EC argued that incentives granted by the Government of China (GoC) to its investors could be considered subsidies provided by the Government of Egypt (GoE). As a result, the Commission decided that those incentives were countervailable under the European Union (EU) regulation⁵ implementing the World Trade Organization (WTO) Agreement on Subsidies and Countervailing Measures (SCM).⁶ The imposition of countervailing duties on certain imports of Indonesian steel was partially based on the same argument in 2022.⁷

Scholarly opinions on the approach invented by the EC have shown considerable divergence. Framing OFDI promotes incentives as transnational production subsidies, Crochet and Hegde concluded that they fell within the SCM Agreement's scope.⁸ The authors also summarily negated the Commission's approach in reaching the same conclusion; they disagreed with the EC legal argument that Article 11 of the Articles on the Responsibility of States for Internationally Wrongful Acts (ARSIWA)⁹ could provide interpretative support.¹⁰ By contrast, Evenett and others succinctly argued that a subsidy, as defined by the SCM Agreement, is territorially confined, meaning that the beneficiary of the subsidy and the granting state should be in the same territory.¹¹ The authors also provided more arguments on the implausibility of interpretative recourse to Article 11 of the ARSIWA.¹²

Given that the EC Argument turns an OFDI incentive into a countervailable subsidy, the question that has remained unanswered so far is whether all OFDI promotion activities are subject to subsidy scrutiny under WTO rules. The EC approach contradicts the previously shared view of some scholars that, despite there being similarities, WTO rules on subsidies do not regulate home country measures promoting OFDI.¹³ Settlement of this question requires a comparative analysis of the normative boundaries between the notions of investment and trade and also calls for a deeper assessment of the merits of the EC argument from that perspective. That is the primary goal of this paper. Doing so will also highlight the need for new, consensus-based rules restraining aggressive OFDI incentives.

As a secondary goal, the paper sheds some light on the implications of the EC argument on developing countries’ ability to attract new and additional capital for domestic growth sectors. In particular, it contrasts the adverse impact of such an approach on developing countries access to green FDI with the growing green domestic subsidy race among global superpowers. This discussion adds weight to the narrative of public goods benefit arising from green OFDI promotion and also facilitatesadequately nuanced future approaches to OFDI regulation towards a balanced outcome.

The structure of this paper is as follows. After briefly looking back at the Commission's key new arguments in the GFF investigation and their implications, particularly regarding developing countries’ ability to attract green investments (Part I), this paper explores whether a clear boundary between trade and investment can be drawn based on existing legal and dispute settlement jurisprudence (Part II). Moving on, it assesses the merit of the EC argument by first looking into the legal plausibility of the claimed understanding of Article 11 of the ARSIWA and then the legal validity of a new interpretation of the term “government” under Article 1.1(a)(1) of the SCM Agreement (Part III). To conclude, the paper summarizes its findings and suggests charting a path that better preserves the interests of all involved (Part IV).

I. The EC's Argument and its Implication

A. The Commission's Novel Argument

In June 2020, the EC published its conclusions regarding the AD and CVD investigations of imported GFFs from Egypt and China.¹⁴ It resulted in, inter alia, the imposition of a 10.9% ad valorem CVD on some GFF products from Egypt.¹⁵ The reason was that the producers in question, subsidiaries of Chinese firms established in the Suez Economic and Trade Cooperation Zone (SETC),¹⁶ had received preferential financing in the forms of direct and indirect loans, capital investment support, and export credit insurance from Chinese state-owned entities (SOEs) in connection with their investments.¹⁷ Despite being linked to foreign investment and granted by a foreign government, those supporting measures appeared to the Commission as countervailable trade subsidies granted by the Egyptian government.

Since this argument was unprecedented, demonstrating how the EC reached its conclusion is worthwhile. At the outset, we should recall that under both EU and WTO rules, a subsidy requires a financial contribution granted by the government or a public body in the country of origin.¹⁸ According to the Commission, this requirement is satisfied in two steps. First, the Chinese SOEs in question are public bodies,¹⁹ or alternatively private bodies 'entrusted and directed' by the government.²⁰ Hence, their offered finance terms can be a 'financial contribution' by the GoC as direct transfer of funds.²¹ Second, with support from the ARSIWA, the actions of the GoC are attributable to the GoE.

The EC argument regarding attribution is based on the claim that the notion of government, as appearing in the definition of subsidy, can be interpreted in the light of Article 11 of the ARSIWA because the latter is a relevant rule of international law “applicable to relations between the parties”.²² Article 11 makes states responsible for conducts otherwise not attributable to it if the state “acknowledges and adopts” the conduct as its own.²³ The EC provided supporting evidence that the GoE acknowledged and adopted the Chinese conduct of preferential financial support as its own. The Commission pointed to a cooperation agreement signed in 2016 regarding the SETC zone (hereinafter, the Cooperation Agreement),²⁴ which included the fact that the GoE expected to benefit from the interaction.²⁵ The EC further indicated that successive Egyptian Presidents expected Chinese investments in the SETC zones and encouraged them.²⁶ Moreover, the preferential finance terms under the BRI being public knowledge, it was claimed that the President of Egypt must have been aware of the same.²⁷ The Commission also highlighted the wording of the 2016 Cooperation Agreement and its related consultation mechanism, which show a similar expectation and encouragement of Chinese investments, including the Egyptian government officials’ mandate to implement the incentives granted under Chinese and Egyptian laws.²⁸ As the EC concludes, it amounts to 'acknowledgement and adoption' by the GoE of Chinese financial contributions.

B. Implications for Developing Countries

The key implication of this line of argument is that it blurs the conceptual distinction between investment and trade transactions. As a result, not only is the line dividing these disciplines further obfuscated, but the investment policy freedom of a sovereign entity is also effectively limited due to the aggressive application of trade rules. This also goes against the Lotus principle of liberty: in the absence of specific legal prohibition, any restriction on the regulatory independence of a state cannot be presumed.²⁹

From the multilateral trade law perspective, there are two key implications. First, the understanding advanced by the EC, especially that of expanding the scope of the term “government” in the subsidy definition, significantly alters the balance of rights and obligations of the WTO members under the SCM agreement. Second, although such a radical reinterpretation of law can invite challenges from WTO members whose “benefits accruing […] under the covered agreements are being impaired”,³⁰ the framing makes it difficult. Regarding the GFF case, while the incentives are purported to be domestic measures of the GoE, the latter is arguably less inclined to pursue dispute settlement as neither its domestic producers nor its financial resources are threatened. By contrast, while Chinese producers in Egypt faced the challenge, the GoC was left without any forum to resolve the issue, as the measure was not directed against it.³¹

Evenett and others point to the systemic implication of the GFF decision in spawning similar investigations against other host states of foreign investment,³² many of which are developing countries. Unsurprisingly, in March 2022, the EC partially relied on the same argument in its anti-subsidy investigation of steel exports from Indonesia,³³ to bring Chinese measures supporting establishment of an Indonesian industrial park and providing preferential financing for investments therein under the SCM Agreement's coverage.³⁴ Referring back to the glass fibre investigation under discussion here, the EC held in paragraph 647:

Indeed, […] the terms ‘by a government’ in Article 3(1)(a) of the Basic AS [anti-subsidy] Regulation and in Article 1.1(a)(1) of the SCM Agreement, interpreted inter alia in light of Article 11 of the ILC [International Law Commission] Articles on State Responsibility for Internationally Wrongful Acts (‘ILC Articles’) permits the attribution to the GOID [Government of Indonesia] of the financial support granted by the GOC [Government of China] to Indonesian exporting producers in the Morowali Industrial Park in Indonesia.

This marks the start of a trend that may eventually institutionalize the EC approach as an entrenched policy position. Under the new EU regulation enacted in 2022 to tackle foreign subsidies, the Commission appears to have adopted a dual-track approach. The definition of “foreign subsidy” proposed in the law covers the contributions that benefit an entity economically active in the EU internal market.³⁵ Whereas the EC committed to trade defence instruments and state-to-state dispute settlements against “certain foreign subsidies granted by WTO Members and limited to goods”,³⁶ OFDI supporting measures to third countries will continue to be responded to by deploying the Commissions’ countervailing measures. This is a double standard, especially considering that EU public financial institutions are engaged in similar forms of OFDI promotion, which include supporting entities in the SETC zone in Egypt, but without encountering any legal repercussions.³⁷

Without prejudice to the generality of the EC approach affecting developing countries’ ability to attract FDI, particular challenges it poses to promotion of green investments deamand special attention. The indispensability of additional Foreign Direct Investment (FDI) to accelerate climate action, combat loss and damage, and gain access to low-carbon technologies by developing countries is well documented.³⁸ While post-COVID-19 economic stimulus has increased green FDI flows, most remain concentrated in developed countries.³⁹ This trend is further exacerbated by explicit recourse to large-scale, WTO-illegal, local content-based green domestic subsidy programmes adopted by the United States⁴⁰ and the EU.⁴¹ The combined effect of all these measures, including the EC anti-subsidy approach, is to heavily skew the landscape of green industrial supply chains to the advantage of European entrepreneurs.

It should also be noted that, while targeting Chinese OFDI incentives, both anti-subsidy investigations that follow the new EC approach affect green industrial value chains. The investigation against Egypt involves glass fibres, essential components of wind turbine blades, solar panels, and building insulation. During the investigation hearing, European wind energy equipment producers (Siemens and Gamesa) unsuccessfully argued that the imposition of CVD may lead to a loss of market share, higher prices for renewable energy (RE) equipment, and a lower uptake of RE solutions.⁴² Moving further, imposing CVDs on Indonesian steel exports surprisingly coincides with the EU's interest in ensuring continued access to nickel ore from that country. Nickel and nickel-steel alloys are indispensable materials for various climate change solutions, including high-capacity batteries for electric vehicles, solar energy, and carbon capture and storage.⁴³ Before introducing the CVDs, the EU brought a WTO complaint against Indonesia's export restriction of raw materials for stainless steel (nickel) in 2019,⁴⁴ which was purported to have been adopted for the growth of a domestic value-added industry for nickel. Subsequently, the CVD imposed in March 2022, during the pendency of the Panel proceedings, had, among others, the effect of restricting Indonesia's movement further up the nickel value chain.

II. Is There an Objective Boundary between Investment and Trade?

Against this backdrop, we explore whether there is any objectively discernible boundary between the concepts of investment and trade in the law and the practice of respective disciplines. A positive finding to that effect would be a strong reason to preclude one category from manipulating the rules meant for the other. The flip side would be that trade rules can be fluidly applied to investments and vice versa. Even if a precise answer cannot be found, an exploration will clarify the legal reception of GFF incentives.

However, looking at the past is important to gain much-needed context for the study before embarking upon a detailed inspection of the status quo. Historically, the practices of trade and investment may have existed in parallel. Still, the two were distinct – not only because of their separate evolutionary pathways but also due to the intrinsically different forms of legal protection required by each. In Europe, the earliest forms of investments that received legal protection were properties and capital assets brought in by merchants and traders from one country to another throughout the seventeenth and eighteenth centuries.⁴⁵ Countries relatively equal in status increasingly made agreements that offered protection and provided equal treatment to the properties of aliens as part of improving trade relations.⁴⁶ Trade, the older economic affair of the two, received legal protections, including lower customs duties and freedom from the jurisdiction of local courts.⁴⁷ In parallel, European colonial powers engaged in one-sided investment and trade activities in large parts of the globe.⁴⁸ Foreign corporations of merchants received monopoly concessions to invest abroad and engaged in competition, exploitation, oppression, and war to obtain market and territorial dominance in those lands.⁴⁹ Apart from exclusive trade competence from their home countries, these corporations' acquisition of ruling authority resulted in special tax rates and selected monopolies in internal trade.⁵⁰ With the waning of colonialism and the arrival of the United States in the economic arena, trade and investment interests co-existed in the evolving bilateral Friendship Commerce and Navigation (FCN) treaties.⁵¹ After the stillbirth of the International Trade Organization (ITO),⁵² the bilateral trend continued for foreign investments, whereas trade multilateralism began with the GATT rules,⁵³ later culminating in the World Trade Organization (WTO). The divergence of these two legal disciplines is so well settled that some suspect whether their confluence can add any value at all.⁵⁴

A. Lessons from Modern Investment Law and Practice

FDI, like the Chinese OFDI in the glass fibre sector in Egypt, embodies the core hallmark of an investment: foreign commercial presence. The public international law discipline on investment protection has emerged from the necessity to strengthen the protection of foreign assets and establishments.⁵⁵ To further identify the characteristics that may be exclusively applied to distinguish between investment and trade, the ideal place to start is the convention establishing the International Centre for Settlement of Investment Disputes (ICSID),⁵⁶ particularly Article 25,⁵⁷ given its near-universal acceptance by states. Key dispute settlement jurisprudence and practice in the IIAs context can then be used to further support this.

It is well known that attempts to find a shared definition of investment during the Convention's early life were unsuccessful.⁵⁸ Despite that, those attempts, combined with subsequent dispute settlement practice, have successfully established the outer bounds of the notion. It has been generally accepted that what passes as an investment under Article 25 is not entirely left to the discretion of the disputing parties. Schreuer holds: “while it is clear that the parties have much freedom in describing their transaction as an investment, they cannot designate an activity as an investment that is squarely outside the objective meaning of that concept”.⁵⁹ Over time, typical recurrent characteristics of an investment, such as durability, regularity of return, substantial nature, contribution to the host state's development, etc., have also taken shape.⁶⁰ These, while not being treaty-mandated criteria or a sine qua non for establishing the existence of an investment, play a significant role in determining the nature of a transaction in question.⁶¹ This objective approach towards defining investment has also influenced a trend in IIA drafting that additionally requires an asset to display the “characteristics of an investment” for it to receive protection of the generally open-ended definition of the term.⁶² Based on these criteria, the disputed Chinese support to its firms established in Egypt falls within the objective boundary of the term “investment”.

Keeping the objective criteria aside, most asset-specific definitions of investment in the IIAs retain financial support measures facilitating investment within its coverage.⁶³ For example, the definition of investment in the EU-Vietnam IIA explicitly includes “bonds, debentures, and loans and other debt instruments, including rights derived therefrom”.⁶⁴ In Fedax v. Venezuela, the ICSID tribunal found loans and credit support to be within its jurisdiction.⁶⁵ Similarly, in Sempra v. Argentina, loans that were part of an investment's financial arrangement were considered to be within the scope of the term.⁶⁶

To establish the scope of “investment”, some past tribunals have distinguished the term from “trade”. These distinctions are useful to trace the difference between the characteristics of these two regimes. In this regard, the majority position appears to reject simple, straightforward, and immediate commercial transactions from the scope of investment. For example, in the annulment hearing of the Malaysian Historical Salvors v. Malaysia, the ad-hoc committee, while exploring the drafters’ intent behind Article 25 of the ICSID convention, held that “[i]t appears to have been assumed by the Convention's drafters that use of the term ‘investment’ excluded a simple sale and like transient commercial transactions”.⁶⁷ Similarly, in Global Trading v. Ukraine, ⁶⁸ the tribunal refused to exercise jurisdiction because money owed by the claimants regarding the performance of standard CIF contracts were “pure commercial transactions and therefore cannot qualify as an investment”.⁶⁹ It was further mentioned that the contract may have benefitted Ukraine's government, given the particular circumstances. Still, such a benefit was nothing more than what all legitimate trade brings.⁷⁰

To sum up, current investment law and practice suggest the existence of clear boundaries distinguishing investment and trade transactions. The distinction has been repeatedly upheld in several investment arbitration panels where jurisdiction over trade measures was declined. In addition, language, particularly that of the EU IIAs, indicates that the type OFDI incentives that are in issue in the GFF case, fall squarely within the scope of the term “investment”. As a result, it can be argued that any questions regarding those incentives should be settled between the home and host states, or before an investment arbitration tribunal.

B. Trade and Investment Interplay at the WTO

This section explores whether, like the investment discipline, the multilateral trade regime excludes investment issues from being entertained or addressed. As the issue is not comprehensively settled, there is room for reasoned opinion. The narrative below highlights the key determinants of trade and investment interplay within the framework of WTO rules, considering the regime's legal structure and dispute settlement practice. It highlights the importance of treaty interpretation, which can arguably create or obviate such a distinction.

The Marrakesh Agreement Establishing the WTO ascribes to the organization, inter alia, the goal of facilitating “trade relation among its Members”.⁷¹ The Preamble to the agreement also aspires to conduct “relations in the field of trade and economic endeavour” to, inter alia, raise living standards, increase employment, and expand trade in goods and services.⁷² However, the scope of this expression is not explicitly clarified anywhere in the rules. Moreover, unlike investment tribunals, a trade dispute settlement panel is not required to determine what trade is before exercising its jurisdiction. According to the Dispute Settlement Understanding (DSU), breaches of covered agreements,⁷³ including violations of obligations thereunder, nullification or impairment of the benefits arising therefrom, and impediments to attaining their objectives automatically result in the compulsory jurisdiction of a WTO Dispute Settlement panel.⁷⁴ “Any act or omission attributable to a WTO Member”⁷⁵ can come under a panel's scrutiny – it is in no way constrained to being exclusively trade-related.⁷⁶ This is further supported by the rich practice of challenging protectionist trade distortions arising from non-trade motivated measures.⁷⁷

Defining the exclusive trade domain, apart from investment, becomes more complicated when the areas of potential overlap are considered. For example, the Agreement on Trade-Related Investment Measures (the TRIMS Agreement) prohibits WTO member host countries from adopting investment policies that result in discrimination (for example, local content requirements).⁷⁸ Also, trade and investment disputes occasionally arise from the same facts. In the Canada–Renewable Energy dispute, feed-in-tariff (FIT) contracts deployed as investment promotion measures were found to breach the respondent's commitment under the TRIMS agreement because the tariffs were attached to local sourcing preconditions.⁷⁹ Renewable energy FIT schemes were also the frequent subject of investor-state arbitrations.⁸⁰ Similarly, the Australian government's Tobacco Plain Packaging regulation has been challenged multilaterally at the WTO and in investor-state arbitration.⁸¹

Even though these overlaps are often taken as illustrations of inextricable linkages between trade and investment disciplines,⁸² a closer look reveals that the issues addressed remain different, even in areas where trade and investment rules come together. The aforementioned plain packaging case is a good example of it. While several WTO members challenged Australia's alleged breach of commitments under the TRIPS agreement, the investor-state arbitration dealt with the alleged failure to uphold Australia's fair and equitable treatment commitment to the investor and possible indirect expropriation resulting from the tobacco plain packaging measure.⁸³ Wu correctly observed that these two disciplines were set on unique and non-convergent paths.⁸⁴

Furthermore, despite the absence of explicit clarification of the substantive scope of trade relations, as mentioned earlier, the structure of the covered agreements and the nature of transactions regulated therein supply significant contextual guidance in deciphering the term “trade”. A broad survey of the landscape of WTO rules reveals that trade principally involves the supply and reception of products and services between independent operators of different nationalities who interact at arm's length, dictated by prevalent market conditions. Generally speaking, trade regulation should include substantive and procedural commitments made by sovereign entities in areas such as reducing trade barriers, promoting transparency and predictability, creating fair conditions of competition, and protecting intangible property rights. Most of these issues fall within the scope of the expression “transient commercial transactions” observed by the tribunal in the Malaysian Historical Salvors dispute.⁸⁵ This shows that the conceptual understanding of the distinction between the notions of trade and investment is largely shared between the respective regimes.

In addition, some past Dispute Settlement panels have explored the meaning of trade in varying contexts. For example, to interpret the phrase “in the course of trade” in Article 10 of the TRIPS agreement, the panel expressed the ordinary meaning of trade as “[t]he action of buying and selling goods and services”.⁸⁶ A similar expression, “sale in the ordinary course of trade”, appears in Article 2 of the WTO Anti-Dumping Agreement and clarifies the identification of benchmarks to determine the dumping margin. Although the provision has been interpreted,⁸⁷ the discussion did not lead to any elucidation of the term “trade”. The language of these provisions establishes that sales and similar transboundary commercial arrangements are the prominent trade domain.

Putting it all together, it is submitted that the WTO's explicit institutional mandate of facilitating trade relations, coupled with the general framework and nature of the covered agreements regulating commercial transactions between states, serves as an important guiding function. They inform the outermost boundaries of the WTO rules. When clarifying any provision of the covered agreements in line with the customary rules of interpretation, a WTO panel should take account of these factors as the necessary context⁸⁸ and expression of shared intent of the members. Therefore, despite not being required to distinguish between trade and investment upfront, the panel is nonetheless responsible for construing the applicable WTO rules in a way that does not venture beyond the outer boundaries of trade. Presumably, the explicit discharging of such a responsibility is not called for in most cases. It will, nonetheless, exist and require specific attention in a rare situation where a WTO member advances interpretation potential that expands the scope of the covered agreements beyond the boundaries of trade relations. Such is the situation regarding the EC approach to redefining the scope of a subsidy in Article 1.1 of the SCM Agreement.⁸⁹ After summarizing this section below, the matter will be addressed in greater detail.

C. Summing Up

Analysis in this section has revealed that investment and trade are both spheres of transactions and regulation that have historically remained adjacent yet displaying unique characteristics that have influenced respective institutional evolution. This is manifested in the way both these regimes define their outer boundaries. On the one hand, ad hoc investment tribunals frequently fall back on the objective or treaty-defined scope of investment, occasionally distinguishing it from trade to affirm its jurisdiction. On the other hand, jurisdiction for a WTO dispute settlement panel is automatic and covers all breaches of covered agreements. This difference in approach, as explained above, is owed to the difference in the legal framework of the two regimes. Although a WTO Panel may seem to be allowed to entertain non-trade measures, issues devoid of trade impact (for example, exclusive investment measures) do not come within the WTO domain.

The preceding analysis also highlighted the important role of treaty interpretation by the panel with regard to defining the outer boundaries of the WTO rules. While interpreting the covered agreements to address members’ claims, the panel is guided by the customary rule of ascertaining the meaning of the law by looking at its ordinary meaning in the light of its context, object, and purpose. Especially when a claimed interpretation of a WTO rule appears to go beyond its traditional ambit, the panel must assess the same, inter alia, against the backdrop of the purpose of the WTO as an institution facilitating trade relations. This is the situation with the new EC interpretation that renders Chinese OFDI promotion measures as countervailable trade subsidies. As this is made possible by the novel recourse to Article 11 of the ARSIWA and, with its aid, extending the scope of the term “government” in the WTO definition of a subsidy, our next task is to examine the legal plausibility of the whole proposition.

III. Use of the ARSIWA to Reimagine Subsidy Scope at the WTO

This section explores the relevance and utility of the ARSIWA attribution provisions regarding alleged overlaps between OFDI incentives and the SCM Agreement. Following the earlier detailed EC line of argument, OFDI incentives are attributable to the recipient government and, hence, are considered a subsidy; two questions come into focus. First, whether such an attribution between governments is supported by Article 11 of the ARSIWA in the factual context of the GFF case. Second, whether this attribution can reinterpret the understanding of the term 'government' in the SCM Agreement. These are discussed below.

A. Attribution Between Governments Under Article 11 of the ARSIWA

The EC argument is premised on the claim that, in the given circumstances, Article 11 of the ARSIWA dictates the attribution of a measure taken by the GoC to the GoE. Here we explore the context and plausibility of such an outcome.

Article 11 of the ARSIWA is a supplementary provision introduced originally as Article 15 bis by the then special rapporteur James Crawford during the second reading of the Draft Articles.⁹⁰ The reason was that, at that stage of the draft, all of the provisions concerning attribution of conduct (Articles 5–14), except one, required prior state mandate.⁹¹ As state endorsement was not considered a prerequisite for responsibility attribution, the proposed Article 15 bis (now Article 11) made ex post facto adoption of conduct a valid basis for attribution.⁹² Examples of state practice in support of the proposition were drawn from the United States Diplomatic and Consular Staff in Tehran case and the Lighthouses arbitration.⁹³ Upon being generally accepted by all the members of the International Law Commission,⁹⁴ the provision made its way into the final draft in the following language:

Conduct which is not attributable to a State under the preceding articles shall nevertheless be considered an act of that State under international law if and to the extent that the State acknowledges and adopts the conduct in question as its own.⁹⁵

The language of Article 11 allows attribution of wrongful conduct between states, arguably within certain parameters. Although it does not specify whether a state or a non-state entity was the original actor, Crawford explained in a subsequent commentary that it was not intended to limit the ambit of the provision to the conduct of non-state entities only, and attribution between two states is equally possible.⁹⁶ However, state practice supporting the latter scenario is restricted to situations where the state authoring the conduct has ceased to exist, and another has taken its place; for example, state succession⁹⁷ or situations where another has subsumed a former sovereign entity or its parts.⁹⁸ This favours the argument that attribution of responsibility between states under Article 11 of the ARSIWA should apply when the state originally authoring the conduct cannot meaningfully assume or discharge their responsibility. The genesis and objective of Article 11, as recounted above, support such a position. It is also corroborated by the fact that the ARSIWA contains a separate chapter covering situations of state responsibility for another existing state's conduct, which Article 11 is not a part of.⁹⁹ The alternative is to open the possibility of consecutive attribution of responsibility upon two or more states for the same action. This is not illegal, although it was not foreseen by the drafters either. Keeping in line with state practice, attribution of responsibility between states should ideally be limited to situations comparable to prior instances.

However, the EC's application of Article 11 to the GFF case contradicts the above rationale. According to the EC itself, the acts in question are those of Chinese public bodies, which are attributable to the GoC.¹⁰⁰ The GoC is in no way unable to discharge any responsibility that may arise out of the conduct of its SOEs. Therefore, the underlying justification behind having a provision like Article 11 of the ARSIWA attributing conducts between governments, that being such conducts will otherwise remain unattributed and unresolved, does not apply in this case. The EC provides no convincing argument for why Article 11 may apply, departing from the prior state practice mentioned above. Such consecutive attribution to the GoC and the GoE also raises questions of shared responsibility, an under-explored and uncertain area of international law.

One pivotal question is the evidentiary standard one should meet to prove that a state has made a “subsequent acknowledgement and adoption” of conduct as its own.¹⁰¹ It was explained in the ILC commentary that, in the absence of explicit confirmation by a state, the act of adoption and acknowledgement can be inferred from the conduct of that state.¹⁰² As the provision has not seen any animation in dispute settlement over the years, we know very little as to what standard such factual analysis should fulfil, except that it is guided by the term “adoption”, itself meaning that “the conduct is acknowledged by the state, in effect, as its own conduct”.¹⁰³ It is submitted that an "acknowledgement, in effect" should require that the entity claimed as responsible has complete knowledge of the particular conduct and its effect regarding which responsibility is claimed – simply because one cannot acknowledge what one does not know.

There is also the issue of proving intent. It ought to be established that the state alleged to be adopting the conduct was aware of its potentially wrongful nature and intended to assume responsibility for it. During the initial presentation of the Draft Articles and later in the ILC commentary, it was cautioned that a factual determination should not be made lightly, especially because states often make statements or endorsements of support without ever intending to be bound by legal responsibility.¹⁰⁴ The ILC commentary also provides that a state may be held responsible for acts it did not initially approve “provided the State's intention to accept responsibility for otherwise non-attributable conduct is clearly indicated”.¹⁰⁵ Proof of intent is even more important when an act that is not wrongful in the hands of the initial actor may become so after such acknowledgement and adoption.

Regarding the GFF case, the Commission was correct in attempting to establish “acknowledgement and adoption” through the acts of the state in question.¹⁰⁶ However, it is doubted whether the facts presented were enough to reach that conclusion. The EC did not produce any information to show that the GoE had comprehensive information about the existence, amount, scope, or the rate of support provided by the GoC. While the Commission presented the 2016 Cooperation Agreement between the GoC and the GoE as supporting evidence, the quoted excerpts fall short of demonstrating government knowledge of a wrongful trade subsidy or endorsement of the same.¹⁰⁷ More importantly, as the agreement took place before the grant of incentives, its provisions play no role in proving subsequent acknowledgement or adoption, as required by law. Furthermore, the Commission showed that the GoC and the GoE had different motives.¹⁰⁸ While the GoE, like many other developing countries, expected and welcomed foreign investments, none of the acts on record show intent of a prohibited trade subsidy.¹⁰⁹ By construing the host state's (GoE) intent to receive foreign investment, including incentives facilitating the same as acknowledgement and adoption of a prohibited trade subsidy, the EC may have committed the exact error that Crawford warned about.¹¹⁰

So, despite the attribution of responsibility between governments being possible under Article 11 of the ARSIWA, it is highly likely that this will not be the result in the GFF case due to the extreme difficulty, if not altogether impossibility, of concluding with reasonable certainty that the GoE intended to adopt the conduct of the GoC as effectively its own.

B. Reinterpreting the Term “Government” under the SCM Agreement

Even if the Commission's proposed understanding of Article 11 of the ARSIWA is accepted, the question of whether that will suffice to dress up an OFDI support measure as a breach of the SCM Agreement will nonetheless remain. To find an answer to that, it is necessary to take a critical look at what the Commission argues concerning the SCM Agreement: the term “government” appearing in the definition of subsidy is wide enough to construe measures taken by another government as that of the government under investigation.

The use of ARSIWA provisions to support the interpretation of WTO law is not new. In the US – Gambling dispute,¹¹¹ a WTO panel took recourse to the ARSIWA as a supplementary means of interpretation under Article 32 of the Vienna Convention on the Law of Treaties (VCLT).¹¹² The Panel looked particularly at Article 4 of the ARSIWA to reinforce its finding that a regulatory measure taken by a state organ is attributable to the United States government.¹¹³ The Appellate Body (AB) once again came close to seeking recourse to the Articles in the US Anti-Dumping and Countervailing Duties dispute (hereinafter US-AD/CVD)¹¹⁴ to determine the meaning of the term “public body” under Article 1.1(a)(1) of the SCM Agreement.¹¹⁵ In that dispute, the parties extensively debated the relevance of the provisions on responsibility attribution in the ARSIWA to find characteristic similarities between “government” and “public body” – as aids to interpret the latter term.¹¹⁶ Based on textual meaning and contextual analysis, the AB concluded that a public body is an “entity vested with certain governmental responsibilities, or exercising certain governmental authority”.¹¹⁷ The AB maintained that the rules of attribution would be a valid recourse under Article 31(3)(c) of the VCLT, but only when they were established as customary international law or general principles and not solely because they are part of the ARSIWA.¹¹⁸ The AB later found that the language of Article 5 of the ARSIWA corresponded with its findings.¹¹⁹ As the provisions are supportive and conclusions that have already been reached will not be altered in light thereof; it was not considered necessary to determine ex-ante whether Article 5 of the ARSIWA was a valid source of law.¹²⁰

It is important to remember that Article 31(3)(c) of the VCLT is not a standalone tool to elicit the meaning of any treaty provision. In any dispute, the shared role of the customary rules of interpretation, as manifested in Articles 31 to 33 of the VCLT, is to reinforce or clarify the ordinary meaning of the treaty text in the context of its intended use.¹²¹ This is also the essence of a harmonious approach to interpretation. In the present case of interpreting the term “government”, appearing in Article 1.1(a)(1) of the SCM Agreement, the ordinary and established meaning of the term, as well as the situation of the provision in one of the multilateral agreements dealing with trade in goods annexed to the Marrakesh Agreement purposed to facilitate trade relations, are all important and unavoidable considerations.¹²²

In opposition to the view that Article 11 of the ARSIWA can be considered a rule of international law that is relevant and applicable in the relation between parties¹²³ to reconstruct the understanding of the term “government” and thereby expand the coverage of a trade subsidy, several arguments can be presented. First and foremost, it has already been mentioned that no state practice supports the EC's intended use of Article 11 of the ARSIWA as a vehicle of attribution between two states. This weakens any claim regarding this provision’s attainment of a customary status and thereby being a “rule of international law” in line with the AB decision in the US – AD/CVD dispute. Next, Article 11, as it stands, is much younger than the SCM Agreement.¹²⁴ Naturally, the GATT signatories negotiating the SCM Agreement in the late-1980s and early-1990s did not foresee such a possibility. Combined with the lack of supportive state practice, it is difficult to see how Article 11 can be considered “applicable” to interpret the notion of “government” within the definition of a subsidy. Furthermore, the SCM Agreement is a specialized legal instrument whose scope for attribution of conduct to a government is intentionally constrained.¹²⁵ During the Uruguay Round negotiations, the European Community highlighted the necessity of subsidies for domestic policy. It sided with the view that the conferment of benefits was not the sole determinant of subsidy.¹²⁶ It further expressed that “the concept of subsidy be limited to actions which imply expenditure of public funds, or otherwise a cost for the government”.¹²⁷ This position, now a part of the definition of subsidy in the SCM Agreement, contradicts the EC position in the GFF case.

The facts giving rise to the need to interpret Article 1.1 of the SCM Agreement with support from provisions of the ARSIWA in the US-AD/CVD dispute are distinguished from those compelling the EC argument in the GFF case. As previously mentioned, the facts in that dispute called for interpreting the term “public body” to clarify its relationship with the government. Article 5 of the ARSIWA, by explaining that a government is responsible for the conduct of entities exercising elements of governmental authority, illustrated the nature of relationships between the government and such entities.¹²⁸ Unlike that situation, clarifying the meaning of “government” in Article 1.1 of the SCM Agreement does not necessarily require considering situations where its action can be attributed to another government. Yet, that is what the Commission set out to establish. Also worthy of reflection, following the EC's argument, is the distributional complication that will arise when the responsibility for any conduct is attributable to more than one government. As WTO members and duty-bearers in this predictable, rules-based multilateral trade regime, states will undoubtedly seek to avoid such a situation.

In the SCM Agreement, the government is primarily regarded as the provider of a subsidy solely because of its authority and control over financial contributions that may result in such forms of support. The Commission's approach is a significant deviation from it. Going back once more to the US-AD/CVD dispute, we find that the AB explored the meaning of the term “government” therein. After noting that the textual meaning of the term indicates “continuous exercise of authority over subjects; authoritative direction or regulation and control”, the AB concurred with its earlier decision made in the Canada-Dairy Dispute that “government” meant effective enjoyment of the “power to regulate, control, or supervise individuals, or otherwise restrain their conduct, through the exercise of lawful authority”.¹²⁹ As mentioned, this understanding is centred on governments having certain functions and powers and the authority to exercise the same. Using the acknowledgement and adoption standard will far outstretch a government's legal responsibility for financial contributions to areas and affairs beyond its regulatory domain or ability to control. The GFF case is but one example of this. This expanded view of government responsibility for subsidies can bring support from private entities and international organizations into its fold, provided that the government encourages and welcomes such activities.

Overall, the reinterpretation of the term “government” in the subsidy definition as proposed by the EC, solely using the VCLT Article 31(3)(c), is untenable. Article 11 of the ARSIWA, although an important part of general international law, cannot be deployed in the suggested manner to expand the scope of attribution under the SCM Agreement. This also contradicts the principle of harmonious interpretation and ends up clashing with the ordinary meaning of the term “government” in the context of the SCM Agreement in light of the parties’ shared understanding.

C. Summing Up

The previous section highlighted the importance of treaty interpretation tools to limit the outer boundaries of trade rules. Building upon that, further analysis clearly shows that the Commission's attempted conversion of OFDI promotion measures to a trade subsidy in the hands of the host government will face significant legal hurdles when challenged. This section illustrated that de facto attribution of responsibility between two existing governments is uncharted territory in international law. Recognized state practice supports such attribution methods in cases where the government authoring the conduct is non-existent or is incapable of discharging its responsibility. New legal claims beyond that scope cannot be claimed as customary and hence “applicable in the relations between the parties”. It is also shown that indirectly expanding the definitional scope of a subsidy by reinterpreting the term “government” in the SCM Agreement is challenging on many fronts. Although WTO law is not clinically isolated from the body of public international law,¹³⁰ it cannot sustain a situation where emerging rules of later origin are used to stretch the meaning of trade rules beyond their intended purport. Such a step will not only amount to a unilateral alteration of the carefully struck balance of rights and obligations among WTO members in the SCM Agreement, but also further endanger the institution's negotiating arm and erode the shared trust of the members in the system.

IV. Regulation of OFDI: Need for Better and More Balanced Approaches

Based on the analysis of this paper, the problem of the GFF anti-subsidy investigation can be framed as a set of contrasting interests waiting to be resolved. The re-imagination of Chinese OFDI incentives as subsidies enables the EC to deploy CVDs to discourage the practice by restricting market access for third-country (for example, Egypt and Indonesia) exports benefitting from such incentives. Not only does this provide a means to counteract Chinese BRI and other OFDI activities,¹³¹ it also accords greater protection to domestic economic interests from foreign competition. However, the extent to which the EC approach succeeds is also the extent to which capital inflows to developing countries are curtailed. This compounds the already manifested challenges of gathering adequate climate finance as well as post-pandemic economic recovery in developing regions. In addition, this approach does not provide the home country of foreign investment (China) any forum to present its concerns.

Against this backdrop, one of the conclusions of this paper, although arrived at differently, corroborates the earlier finding of Evenett and others that the Commission's imposition of CVD regarding outward investment incentives is illegal under the SCM Agreement. Although the settlement of trade disputes does not require isolating trade from other economic transactions, customary rules of interpretation prevent the scope of multilateral trade rules from being stretched to apply beyond trade relations. Hence, an interpretative extension of the SCM Agreement's definition of “subsidy” to cover investment promotion measures is not legally possible. Even if that were so, no state practice supports the meaning of Article 11 of the ARSIWA as purported by the EC, which will render it as applicable law.

It is submitted that, within the existing framework of WTO rules, filing a non-violation complaint against an OFDI home state is the multilateral course of action available to the EC to protect its internal market from exports that are allegedly under-priced due to having received investment incentives. By dint of Article XXIII(1)(b) of the GATT, read in conjunction with Article 30 of the SCM Agreement, the EC can claim that Chinese OFDI promotion measures in Egypt, despite not being in breach of WTO rules, have: (i) impeded the objective of the SCM Agreement, or (ii) caused nullification and impairment of benefits accruing to the EC under the per se prohibition of export subsidies under that Agreement. Evidence supporting the anti-subsidy investigation also supports such a claim.

Beyond a non-violation complaint, regulating the trade repercussions of OFDI under the auspices of the WTO will call for developing new rules, ideally balancing the conflicting interests mentioned above. Members interested in regulating the promotion of OFDI to preclude unforeseen knock-on trade distortions can call for participatory and inclusive rulemaking. At the same time, as this paper has shown, it is of great importance to ensure that support for transboundary capital flows benefitting public goods is not only allowed but also actively encouraged.

One innovative multilateral solution to incorporate the above arrangement into the WTO rules is by including it in the TRIMS Agreement. The TRIMS Agreement regulates investment measures that are inconsistent with specific GATT rules. For example, a new Article 2 bis can be laid down as: “No Member shall allow a TRIM to be granted or maintained within its territory by another Member, which, if granted or maintained by the former, would violate its commitments under the Agreement on Subsidies and Countervailing Measures.” This can be further accompanied by applying several illustrative measures appended to the existing Annex of the TRIMS Agreement. The benefits of such a construction are twofold. First, it will use the already existing multilateral agreement on the subject. Second, Article 3 of the TRIMS Agreement makes general exception provisions of the GATT applicable to it. As a result, OFDI incentives that are public policy motivated and implemented fairly will remain outside the scope of the new restriction. The key limitation will be that the arrangement will only apply to investment measures that impact trade in goods.

Aside from a multilateral approach, the upcoming agreement on Investment Facilitation for Development (IFD) at the WTO is a potential plurilateral avenue to implement a set of rules to limit the practice of trade-distorting OFDI and promote investments for public goods. While slated to establish a transparent, efficient, sustainability-oriented, and development-friendly business environment,¹³² the exact contents of the IFD remain unknown to the public. One earlier consolidated text from March 2022 shows the convergence of members’ interest around issues that include, inter alia, information transparency and notification,¹³³ special and differential treatment for developing countries,¹³⁴ responsible business conduct,¹³⁵ and linkage with general as well as security exception provisions,¹³⁶ all of which can play a facilitative role to regulate OFDI. However, as far as the latter is concerned, very little agreement is visible in the text. While one proposal suggested that changes be made to incorporate home state obligations on encouraging OFDI through incentives, it was highlighted that these required further discussion.¹³⁷ By contrast, the text also documents a Chinese proposal to keep such OFDI incentives (that is, grants, insurance, and loans) outside the scope of the IFD,¹³⁸ showing a lack of willingness to cooperate on the issue. It is hoped that the members put in greater effort to find areas of agreement on OFDI regulation and promotion instead of dropping the issue altogether.

Outside the WTO perimeter, bilateral, or multi-party agreement among the largest capital exporting countries, including China, can also be a solution. However, amid the escalating geopolitical and security tensions, and the US actions to decouple from China in strategic sectors,¹³⁹ it is not a viable path in the short term. If an environment of cooperation is somehow restored in the future, the erstwhile Comprehensive Agreement on Investment (CAI)¹⁴⁰ will provide useful examples to channel OFDI into a more sustainable route. In this frozen agreement, both the EU and China committed to ‘facilitate and encourage’ investment in environmental goods and services.¹⁴¹ Using obligatory language, the CAI holds that “each party shall […] promote and facilitate the investment of relevance for climate change mitigation and adaptation”.¹⁴² Although the parties have committed to ensuring that their SOEs (“covered entities”) operate transparently, without discrimination and under commercial considerations,¹⁴³ exceptions are also built to maintain inconsistent incentives promoting public goods. The parties preserved their right to regulate in determining the desired level of environmental protection,¹⁴⁴ including the freedom to deploy measures necessary to protect human, animal, and plant life, and health.¹⁴⁵

Lastly, countries that are the biggest sources of OFDI can also self-regulate to enhance the positive impact of capital flows and limit the adverse consequences. The activities of the Development Finance Corporation (DFC) of the United States are one such example. With a limited budget of $60 billion (US), the DFC is mandated to support investment projects with development benefits in low-income countries.¹⁴⁶ Similar self-regulation also takes place in China. The Chinese government has introduced successive guidelines to divert investment resources under the BRI umbrella to climate-friendly sectors (the so-called “green BRI”).¹⁴⁷ While such efforts can benefit the recipient countries, the obvious limitation of self-regulation lies in the unilateral setting of priorities and lack of any reflection of other states’ concerns. Nevertheless, one could hope that self-regulation will generate positive and shared practices on OFDI over time, which can be generalized across economies worldwide.

V. Conclusion

Triggered by the EC's novel argument to consider OFDI support measures as trade subsidies in the hands of the host government of foreign investment, this paper searched for a tangible boundary between trade and investment. It reveals a structural difference between regimes in designating their outer boundaries. While an investment regime often makes a conceptual distinction between the notions, the effect-oriented principles that dictate trade jurisdiction mean that the necessity to make such distinctions is often non-existent under WTO rules. Nevertheless, this paper advances the argument that customary rules of interpretation will dictate that all WTO commitments be construed in the light of the regime's structure, operation, and, most importantly, the shared objective of facilitating trade relations among its members. It, therefore, agrees with prior scholarly findings that the Commission's purported expansion of the SCM Agreement's scope to indirectly include OFDI promotion activities is not legally possible. This paper concludes that the EC should either resort to non-violation complaints or develop new substantive rules to address the trade impact of OFDI promotion. The key general takeaway is that, instead of straining the multilateral trade system by stretching its rules beyond their intended limit, WTO members should ensure that emerging legitimate concerns are addressed based on inclusivity and partnership. Several avenues have been highlighted to that effect in the penultimate section of this paper.

Connected to the above, a secondary goal of this paper was to articulate the plight of developing countries that will arise from the standardization and further proliferation of the new EC anti-subsidy practice. It has shown that such an approach significantly impedes developing countries’ efforts to access new financial resources necessary to maintain, inter alia, sustainable economic development. This approach appears even more unfair when the growing race for green industrial subsidization is taking place among some developed countries. As the paper suggests, the takeaway is that any future regulation restraining the practice of OFDI must also retain necessary policy space to effectively engage in and promote outward investments that produce public goods.