‘Unforeseen Developments’ Before and After US – Safeguard Measure on PV Products: High Standard or New Standard?

Abstract

This paper offers a comprehensive study of the jurisprudence on the ‘as a result of unforeseen developments’ test under the WTO's safeguards (SG) rules. It contributes to the existing scholarship by making three fresh arguments. First, the Appellate Body's decision to ‘revive’ this test as a prerequisite for the application of SG measures is not necessarily incompatible with the drafting record of the SG Agreement, even though this agreement does not make explicit reference to the test. Second, the test is not excessively difficult to satisfy under the standard of review established by case law, even though governments failed to pass it in almost all SG disputes to date. Third, in sharp contrast, the recent US–Safeguard Measure on PV Products decision took a strikingly more deferential approach which fell far short of the established standard of review, leading to the first and only decision in which the test was found to be satisfied. This decision has arguably created a new standard which could lead to abuse of SG measures and damage to the dispute settlement system and hence should be avoided in future disputes.

1. Introduction

The WTO's safeguards (SG) mechanism allows Member governments to impose trade measures otherwise prohibited under WTO rules. This mechanism was created under Article XIX of GATT 1947, titled ‘Emergency Action on Imports of Particular Products’, and was elaborated through the Agreement on Safeguards (SG Agreement) concluded in the Uruguay Round which created the WTO. As widely observed, this mechanism serves as a ‘safety valve’ necessary to incentivize governments to engage in further trade liberalization. The World Trade Report 2009 explains this key function of SG as follows:

countries need the flexibility to temporarily defect from their obligations under an international trade agreement in order to be ready to commit to a higher level of liberalization commitments. At the time that a trade agreement is concluded, countries are unable to foresee all future events that may lead to an intensification of competitive pressure from foreign imports. This may make contingent measures desirable for certain industries, be it as insurance against income loss, to facilitate industry adjustment to competition or for political reasons.¹

From a political economy perspective, the SG mechanism provides room for governments to respond to the demand for protection by import-competing industries, particularly when they have difficulty in competing with foreign counterparts.² In reality, protectionist interest does constitute a major driver of the application of SG measures.³ To avoid potential abuse, the WTO rules impose a range of requirements or obligations on governments both before and after SG measures are applied. The balance between maintaining sufficient flexibility for the use of SG measures and restraining their abuse hinges on how these requirements or obligations are interpreted and applied.

One of the most contentious prerequisites for the application of SG measures concerns the requirement that investigating authorities must show that import surges are a result of ‘unforeseen developments’. Article XIX:1(a) of the GATT states:

If, as a result of unforeseen developments and of the effect of the obligations incurred by a contracting party under this Agreement, including tariff concessions, any product is being imported into the territory of that contracting party in such increased quantities and under such conditions as to cause or threaten serious injury to domestic producers in that territory of like or directly competitive products, the contracting party shall be free, in respect of such product, and to the extent and for such time as may be necessary to prevent or remedy such injury, to suspend the obligation in whole or in part or to withdraw or modify the concession. (emphasis added)

The GATT provides no further guidance on the meaning of ‘unforeseen developments’. Since the ‘as a result of unforeseen developments’ test was rarely applied in the GATT era and the SG Agreement makes no reference to it, it is widely observed that this test should no longer be a prerequisite for the imposition of SG measures.⁴ Commentators have also questioned whether this test can be properly applied in practice due to the difficulties in developing a clear and coherent standard for determining whether an alleged development is ‘unforeseen’ and if so, whether a logical connection exists between the unforeseen development and import surges.⁵ As will be discussed below, these issues have been litigated intensively in a series of WTO disputes, and the tribunals have sought to establish standard approaches to the interpretation and application of the test. As a threshold issue, the Appellate Body (AB) maintained that despite a lack of reference to the test in the SG Agreement, it remains a prerequisite for the imposition of SG measures based on the operation of GATT Article XIX:1(a). In practice, this prerequisite has become a major hurdle to the use of SG measures. The fact that the governments imposing SG measures failed to satisfy the prerequisite in almost all SG disputes begs the question whether WTO tribunals have set the bar too high. In contrast, in the recent US–Safeguard Measure on PV Products case,⁶ the panel found, for the first time, that the findings of the United States International Trade Commission (USITC) satisfied the prerequisite. As of this writing, this remains the only decision in which the government using SG measures has passed the test. It raises the question of whether the US–Safeguard Measure on PV Products decision has somehow created a different standard making the test easier to pass.

While the ‘as a result of unforeseen developments’ test has been increasingly applied and elaborated in WTO disputes, current studies have paid insufficient attention to the development of the relevant jurisprudence.⁷ To fill this gap, this paper conducts a detailed review of the jurisprudence developed in all SG disputes between 1995 and 2022. It makes three main arguments. First, as will be elaborated in Section 2, the AB's interpretation that this test remains a prerequisite for the application of SG measures is not necessarily contrary to the drafting record in the Uruguay Round. Second, the relevant jurisprudence does not impose an excessively high standard but remains reasonably balanced. Third, as will be discussed in Section 3, in US–Safeguard Measure on PV Products the panel took a remarkably more deferential approach to the findings of the US authority, thereby creating a standard evidently lower than the one applied in other disputes. This section will discuss the implications of this new standard, particularly in the context of the prevailing scholarly views noted above and the ongoing AB crisis which largely arose out of US concerns about the AB's judicial activism⁸ including creating law on ‘unforeseen developments’. Section 4 then examines the two panel decisions after US–Safeguard Measure on PV Products and shows that the panels did not adopt the more deferential approach but have maintained the standard applied consistently in other cases. Section 5 sets forth the conclusion.

2. ‘As a Result of Unforeseen Developments’: A High Standard?

GATT Article XIX was modelled on SG provisions in the then existing trade treaties of the US.⁹ During the negotiations of the GATT, the term ‘unforeseen developments’ was barely discussed.¹⁰ However, not long after the conclusion of the GATT it became an issue in the US–Fur Felt Hats dispute in which Czechoslovakia challenged the US's imposition of SG measures on certain women's fur felt hats and hat bodies.¹¹ Amongst other issues, the two sides disagreed on whether a change of consumers’ preferences for certain styles or special finishes of the goods involved was an unforeseen development. The US argued that during the negotiations of the relevant tariff concession, it did not foresee ‘the degree of the future shift to special finishes or the effect which it, together with the concession, would have on imports’.¹² The Working Party held that the term ‘unforeseen developments’ means

developments occurring after the negotiation of the relevant tariff concession which it would not be reasonable to expect that the negotiators of the country making the concession could and should have foreseen at the time when the concession was negotiated.¹³

It then found in favour of the US, holding that the US negotiators ‘could not reasonably be expected to foresee this style change in favour of velours’ at such a large scale and for such a long period as to trigger a significant increase in imports of special finishes and considerably reduce the competitiveness of US manufacturers in this industry.¹⁴

After this dispute, there were no further decisions on ‘unforeseen developments’ in the GATT era.¹⁵ It was observed that this legal condition was largely ignored in national laws and the practices of many GATT Contracting Parties, particularly the US, leading to its absence in the SG Agreement.¹⁶

2.1 Reviving the ‘Unforeseen Developments’ Test

The issue of whether the criterion of ‘as a result of unforeseen developments’ needs to be established before an SG measure can be applied arose in the first two SG disputes under the WTO. In Korea–Dairy Safeguard and Argentina–Footwear Safeguard, the defending parties contended that this criterion was no longer such a prerequisite because it was omitted in the SG Agreement and the conditions set out in the agreement should prevail and be applied.¹⁷ In both disputes, the panels supported this contention holding that ‘unforeseen developments’ do not constitute an additional condition and that the satisfaction of the requirements of the SG Agreement is ‘sufficient for the application of safeguard measures within the meaning of Article XIX of GATT’.¹⁸ The AB disagreed and ruled that the imposition of SG measures must conform with both GATT Article XIX and the SG Agreement.¹⁹ In the AB's view, Articles 1 and 11.1(a) of the SG Agreement suggest clearly that the Uruguay Round negotiators did not intend that the agreement would entirely replace Article XIX.²⁰ Based on the interpretative principle of effectiveness, the AB felt obliged to give meaning to the phrase ‘as a result of unforeseen developments’.²¹ It held that while this phrase does not create ‘independent conditions’, it imposes an obligation on investigating authorities to demonstrate the existence of (1) unforeseen developments as a matter of fact and (2) a logical connection between the developments and the increase in imports causing injury to domestic producers.²² The AB believed that this interpretation is consistent with the approach taken by the GATT Working Party in US–Fur Felt Hats.²³ The AB also opined that the word ‘unforeseen’ means ‘unexpected’, although it did not offer further guidance on the exact scope of either of the words.²⁴ In Korea–Dairy Safeguard, the AB was unable to determine whether Korea had fulfilled its obligations to establish this prerequisite because the panel did not make any findings on whether the alleged increase in imports was a result of unforeseen developments.²⁵ In Argentina–Footwear Safeguard, the AB did not deliver any findings on whether Argentina fulfilled this prerequisite either because the contested measure was already found to be in breach of other requirements under the SG Agreement.²⁶

The AB's decision that the ‘as a result of unforeseen developments’ test remains a prerequisite for the use of SG measures has been consistently applied in subsequent cases. However, as noted above, one major criticism remains, that is, the view that this decision goes against the intention of the drafters of the SG Agreement in the Uruguay Round. Contrary to this criticism, we show that the drafting record of the SG Agreement does not provide a definitive answer to the question of whether the omission of ‘unforeseen developments’ was actually intended to remove this criterion.

The drafting record is sufficiently clear on the fact that the negotiators paid little attention to the criterion of ‘unforeseen developments’ since the outset of the Uruguay Round. The Ministerial Declaration on the Uruguay Round, which set forth the mandate for negotiations including the main elements/items for the negotiation of an SG agreement, made no reference to ‘unforeseen developments’.²⁷ To assist the negotiations, the GATT Secretariat prepared a summary of the work that had been undertaken on SG, which also showed a lack of consideration of ‘unforeseen developments’.²⁸ As a consequence, subsequent negotiations were focused on the items listed on the negotiating agenda whereas the criterion of ‘unforeseen developments’ was largely overlooked.²⁹ In June 1989, the Chairman of the Negotiating Group on Safeguards (Chairman) circulated a draft SG agreement which included the following condition:

there has been an unforeseen, sharp and substantial increase in the quantity of such product being imported.³⁰ (emphasis added)

However, one may argue that this reference to the term ‘unforeseen’ was not necessarily a reference to ‘unforeseen developments’ as it was concerned with an unforeseen increase in imports as opposed to an unforeseen circumstance causing such an increase. Subsequent negotiations, based on this draft text, continued to pay little attention to ‘unforeseen developments’.³¹ The next draft agreement changed the word ‘unforeseen’ to ‘unexpected’, although it was unclear why this change was made from the negotiating record.³²

Mexico seems to be the only Contracting Party which proposed that the negotiations should take into account the existing legal elements including ‘unforeseen developments’³³ and that this element should be added to the draft text as follows:

there has been an unforeseen, sharp and substantial increase in the quantity of such product being imported; as a result of unforeseen developments and of the effect of the obligations, including tariff concessions, incurred by a contracting party under the General Agreement. ³⁴ (original emphasis)

This proposal suggests that the reference to ‘an unforeseen … increase’ was considered to be something different from an ‘unforeseen development’, thereby lending support to our observation above. However, Mexico's proposal received no attention in subsequent negotiations³⁵ which led to the removal of the reference to ‘an unforeseen/unexpected … increase’ in another revised draft agreement.³⁶ This lack of reference to both ‘an unforeseen/unexpected … increase’ and ‘unforeseen developments’ remained unchanged in the next draft circulated by the Chairman on 26 October 1990. Article 2 of this draft set out the conditions for the application of SG measures as below:

A contracting party may apply a safeguard measure to a product only if the importing contracting party has determined, pursuant to the provisions set out below, that such product is being imported into its territory in such increased quantities, absolute or relative to domestic production, and under such conditions as to cause or threaten to cause serious injury to the domestic industry that produces like or directly competitive products.³⁷

It was this provision that eventually became Article 2.1 of the SG Agreement without any substantive change.³⁸

This brief review of the drafting record of the SG Agreement shows that there was almost no discussion of the criterion of ‘unforeseen developments’. The drafting record, therefore, leaves it uncertain as to whether this criterion should remain a prerequisite for the use of SG measures. One may argue that the removal of the reference to ‘an unforeseen/unexpected … increase’ and the disregard of Mexico's proposal to add ‘unforeseen developments’ to the draft text indicate the drafters’ intention to remove ‘unforeseen developments’ as a prerequisite.³⁹ However, this is not the only plausible interpretation of the drafting record. As suggested above, the reference to ‘an unforeseen/unexpected … increase’ was not necessarily the same as ‘unforeseen developments’ which cause an increase in imports.⁴⁰ Thus, the removal of the former does not show definitively the negotiators’ position on the latter.

An alternative interpretation is that the drafters simply overlooked the possibility that GATT Article XIX may continue operating cumulatively with the SG Agreement. Here, it is worth noting the Chairman's remark on the last draft text, which stated that the draft agreement ‘would not amend Article XIX or any other Article of the General Agreement’.⁴¹ This statement suggests an understanding emerged from the negotiations that SG measures may be governed by both GATT Article XIX and the SG Agreement. The failure of the negotiators to pay sufficient attention to this possibility or to explicitly remove the criterion of ‘unforeseen developments’ provided scope for the AB to ‘revive’ this criterion.⁴² The AB's decision, thus, is not necessarily incompatible with the drafting record.

2.2 Setting the General Standard

Since GATT negotiators failed to consider ‘unforeseen developments’ in the Uruguay Round, WTO tribunals have been tasked with clarifying the meaning of this term and developing a proper standard for its application. They started to do so in Korea–Dairy Safeguard and since then have developed some general requirements on domestic authorities which conduct SG investigations. As shown above, authorities must demonstrate the existence of (1) ‘unforeseen’ circumstances as a matter of fact and (2) a logical connection between the circumstances and the increase in imports causing injury to domestic producers. In US–Lamb, the AB held that authorities are required to set forth ‘reasoned and adequate explanations’ and findings on both issues in their published report(s) that decide to impose SG measures.⁴³ Such explanations and findings ‘cannot be supplemented by the Member concerned during the course of WTO dispute settlement procedures’ because doing so would constitute an ‘ex post facto explanation’ for SG measures already imposed.⁴⁴ Overall, whether an explanation of unforeseen developments or a logical connection between such developments and import surges is reasoned and adequate depends on the factual circumstances of each case.⁴⁵ This general standard of review does not provide a clear answer as to whether the standard would be difficult to satisfy, although it has left some flexibility for domestic authorities and for WTO tribunals themselves to assess whether the authorities have complied with their obligations. To determine whether the case law has set the standard too high, one needs to explore the reasons behind WTO tribunals’ dismissal of the relevant explanations and findings of the authorities in individual disputes.

2.3 Applying the General Standard

Prior to the US–Safeguard Measure on PV Products dispute, the failure of the governments imposing SG measures to satisfy the criterion of ‘as a result of unforeseen developments’ can be attributed to three main reasons, as categorized below.

The first reason is simply that the authorities did not provide any assessment of this criterion, especially in the earlier disputes. In US–Lamb and US–Line Pipe, the US did identify some unforeseen circumstances in the panel proceedings such as changes in the type of lamb meat imported into the US in the former case and the collapse in oil and gas prices and the East Asian financial crisis in the latter.⁴⁶ However, this was an ex post facto explanation as these circumstances were merely considered in sections of the USITC reports that dealt with other issues. The lack of consideration of ‘unforeseen developments’ by the USITC was understandable as the US law did not require the authority to do so at that time.⁴⁷

The second category involves cases in which the authorities did consider ‘unforeseen developments’ but a reasoned explanation was largely lacking. Argentina–Preserved Peaches was the first dispute in which the panel examined claims on the existence of ‘unforeseen developments’. Argentina identified three circumstances: ‘(a) increased production as a result of an exceptional Greek harvest; (b) substantial increase in world stocks; and (c) a downward price trend’.⁴⁸ The panel found that the report of the Argentine authority did not consider the increase in world stocks as an unforeseen development.⁴⁹ In assessing the other two circumstances, the panel ruled that the authority did not offer any explanation as to why they were unforeseen. What the authority did was merely to find that ‘the increase in imports, or the way in which they were being imported, was unforeseen’ which in the panel's view, ‘does not constitute a demonstration as a matter of fact of the existence of unforeseen developments’ (original emphasis).⁵⁰ Accordingly, the panel drew a distinction between an unforeseen increase in imports and unforeseen circumstances leading to such an increase.⁵¹ In Ukraine–Passenger Cars, Ukraine failed to establish ‘unforeseen developments’ for a similar reason. The panel found that a demonstration of an unforeseen increase in imports was different from and insufficient for the required demonstration of unforeseen circumstances causing the import surges.⁵² The panel also found that Ukraine's allegation relating to the global financial economic crisis as being unforeseen was merely an ex post facto explanation as the authority did not explain why this crisis constituted an unforeseen development in its report.⁵³ In Dominican Republic–Safeguard Measures, the panel considered two alleged circumstances: (1) ‘the entry of China into the WTO and the effect that this had on international trade’ and (2) tariff reduction made by Dominican Republic as a result of the entry into force of two free trade agreements (FTAs).⁵⁴ The panel found that the authority provided no explanation or reasoned conclusion as to why these events were unforeseen at the time of the accession of Dominican Republic to the WTO in 1995 and how they resulted in the increase in the products concerned.⁵⁵ In Indonesia–Iron or Steel Products, the Indonesian authority identified two circumstances: ‘(a) the 2008 global financial crisis; and (b) a change in Indonesian raw material preferences from wood to light steel’.⁵⁶ However, the authority did not provide any further explanation to support the allegation that these circumstances constituted unforeseen developments.⁵⁷

Only two cases, in the third category, involved more detailed claims and assessment by WTO tribunals of the ‘as a result of unforeseen developments’ criterion. The main reason behind the failure to satisfy this criterion was that the authorities’ explanations and supporting evidence were found to be inadequate. In US–Steel Safeguards, the panel, for the first time, found that the existence of ‘unforeseen developments’ was established and then examined whether the authority had provided a proper assessment of a logical connection between such developments and an increase in imports of the steel products involved. The USITC identified four factors – i.e. ‘the Russian crisis, the Asian crisis and the continued strength of the United States’ market together with the persistent appreciation of the US dollar’ – arguing that they each and in combination constituted ‘unforeseen developments’.⁵⁸ The panel observed that an unforeseen development may arise from either individual events or a confluence of events.⁵⁹ It held that the Asian financial crisis, the dissolution of the former Soviet Union and their impacts on the global steel markets were circumstances that the US negotiators could not foresee at the conclusion of the Uruguay Round in 1994.⁶⁰ As regards the other two factors, the panel held that they constituted unforeseen developments only in the context of or in combination with the two crises, but not on their own.⁶¹ With respect to the relationship between the unforeseen developments and the import surges, the US claimed that (1) the Asian financial crisis caused a decrease in steel consumption in five major economies in the region and a significant depreciation of their currencies; (2) the post-USSR economy led to a sharp increase in steel production which flooded into the global markets; and (3) coupled with an appreciation of the US dollar and a robust US economy, these circumstances resulted in surges of steel imports into the US market.⁶² The US failed to substantiate its claims largely due to the lack of a reasoned explanation to show the existence of such a relationship ‘in respect of the specific steel products at issue’ (original emphasis).⁶³ On appeal, the AB upheld the panel's findings. It clarified that the ‘reasoned and adequate explanation’ standard requires the USITC to use the data included in its report to explain how unforeseen developments resulted in increased imports of each product subject to the SG measures.⁶⁴

India–Iron and Steel Products was the last SG case before US–Safeguard Measure on PV Products. The Indian authority considered two circumstances: (1) the fast-developing steel production capacity in a number of major exporting countries in 2014 and 2015 leading to a global overcapacity, and the decline in demand for steel imports in the US, the EU, and China, and (2) the depreciation of the Russian currency and the restricted access of Russian steel products to its traditional export markets. These circumstances, coupled with the growing demand in India, were alleged to be unforeseen resulting in surges of steel imports into India.⁶⁵ The authority provided some data to show the excessive global steel production capacity.⁶⁶ The panel ruled that ‘negotiators could not reasonably have expected this confluence of events when India negotiated its tariff concessions’.⁶⁷ Interestingly, the panel did not question whether the authority had adequately explained why these circumstances were unforeseen. As far as the relationship between the unforeseen developments and the import surges is concerned, the panel observed that Article XIX:1(a) does not mandate how the relationship should be examined, thereby leaving some room for authorities to choose the appropriate method in each case.⁶⁸ India failed to show this relationship due to several shortcomings in its authority's report. First, the authority did not provide data on steel production and exports to India by its two FTA partners, Japan and Korea. The authority did not explain ‘why the alleged increase in imports, with a predominant share from Japan and Korea, occurred due to the unforeseen developments of different origins’.⁶⁹ Second, while the authority considered the increase in steel imports in general, it did not provide evidence to show that the unforeseen developments resulted in an increase in the products subject to the SG investigation.⁷⁰ Third, the panel questioned the lack of evidence showing a link between the excess steel supply and the change in demand in several markets, on the one hand, and the alleged increase of imports into India, on the other.⁷¹ Finally, the authority did not provide an adequate explanation for the timing of some of the unforeseen events to show a clear temporal connection between the events and the import surges.⁷²

2.4 Summary

In summary, WTO tribunals have provided further guidance for how the general standard of review would be applied, that is, what authorities are expected to do to provide a reasoned and adequate explanation under the ‘as a result of unforeseen developments’ test. There are three major steps. The first is to identify the circumstances or events alleged to be unforeseen. This is a relatively easier task as the tribunal does not preclude any circumstance from being a potential unforeseen development.⁷³ The decisions above suggest that this step merely requires authorities to identify the relevant circumstances as part of the assessment of ‘unforeseen developments’ in their reports.

The second step is to explain why the circumstances identified were unforeseen when the relevant GATT/WTO concessions or obligations were made. In most of the decisions above, the panels took the position that a mere allegation is insufficient and some explanation based on evidence is required. The flexibility, though, is that in cases where an event is conspicuously unforeseen, less explanation/evidence would be needed. The circumstances relating to the Asian and Russian crises in US–Steel Safeguards are arguably an example of such cases. However, in India–Iron and Steel Products, the panel took a more deferential approach without examining whether the Indian authority had provided adequate explanations as well as supporting evidence to substantiate the claim that the circumstances concerning excessive steel production, currency depreciation, and changes in domestic demand were unforeseen. Compared to the Asian and Russian crises, it is less evident that these circumstances could have not been foreseen. The Indian authority did not provide any explanation for why these circumstances were unforeseen at the time when the relevant concessions were made. Such an unsubstantiated allegation was consistently rejected in the other decisions. This panel's approach, therefore, lowered the established standard of review and may tilt the balance embedded in the requirement of ‘unforeseen’ in favour of the application of SG measures. As a similar issue also arose in the US–Safeguard Measure on PV Products decision, we will elaborate our arguments in Section 3.

The third step requires authorities to provide reasoned and adequate explanations for the existence of a logical connection between unforeseen developments and an increase in imports. It is insufficient to record such developments and increases separately. Instead, there needs to be some explanation and supporting evidence to show how the developments have resulted in the import surges. Moreover, where an SG measure is applied to a subset of goods that falls within a broad category of products (such as steel products), authorities must show how unforeseen developments have led to an increase in the importation of the goods subject to the measure.

Thus, the satisfaction of the ‘as a result of unforeseen developments’ test largely depends on whether a reasoned and adequate explanation has been provided in the last two steps. Whether an explanation was adequate would in turn depend on the facts in each case. While the WTO tribunal has imposed some requirements on investigating authorities, these requirements were not unreasonably difficult to fulfill. It is true that the governments imposing SG measures failed to pass the test in all the decisions discussed above. However, in most of the decisions the reason for this failure was simply because the authorities either ignored this criterion completely or addressed it too briefly. The US–Steel Safeguards and India–Iron and Steel Products decisions did suggest that WTO panels would examine whether there are notable gaps in the authorities’ reports which would render the evidence and explanation provided inadequate. However, the gaps identified in both decisions can be fixed by providing more data/evidence to show that the increase in the imports concerned was in fact a result of the unforeseen developments.⁷⁴ Overall, while the WTO tribunal will need some evidence and explanation by authorities under the ‘as a result of unforeseen developments’ test, the standard of review, as applied in the aforesaid decisions, was reasonably balanced and did not set the bar too high.⁷⁵

3. The US–Safeguard Measure on PV Products Decision: A New Standard?

The US–Safeguard Measure on PV Products case concerned the US's SG measures on imports of certain crystalline silicon photovoltaic cells (CSPV Products). In the SG investigation, the USITC prepared a ten-page supplemental report⁷⁶ to specifically address the issue of ‘as a result of unforeseen developments’. This report set out four circumstances as being unforeseen at the time that the US acceded to GATT 1947 or later to the WTO, or at the time that it agreed to China's accession to the WTO, including that

1. 1. China would implement ‘industrial policies, five-year plans, and other government support programs favoring renewable energy product manufacturing, including CSPV products’ which directly contradicted its accession commitments;⁷⁷

2. 2. ‘These industrial policies, plans, and support programs would lead to the development and expansion of capacity to manufacture CSPV products in China to levels that substantially exceeded the level of internal consumption’;⁷⁸

3. 3. This overcapacity ‘would largely be directed to export markets such as’ the US;⁷⁹

4. 4. A series of antidumping and countervailing duties the US imposed on CSPV products imported from China since 2011 would be of limited effectiveness while the ‘rapid changes in the global supply chains and manufacturing processes’ would immediately follow the imposition of these duties. The latter development concerned the expansion of investments by major Chinese CSPV producers in other markets so that they continued to increase their production capacity causing an increase in exports to the US from these foreign markets.⁸⁰

The report provided some data to show the expansion of China's CSPV production capacity both in China and the third countries involved and the increase in CSPV exports to the US from these markets during the period of investigation.⁸¹ It then concluded that these unforeseen developments resulted in an increase in the importation of CSPV products into the US.⁸²

3.1 Creating a New Standard

In reviewing the US's claims, the panel took a strikingly more deferential approach. The panel accepted all US claims on ‘unforeseen developments’ without asking for explanations and supporting evidence to substantiate these claims. One of China's major arguments was that the USITC failed to identify the relevant Chinese industrial policies, five-year plans, and other government support programs and to explain why they were unforeseen.⁸³ The panel dismissed China's argument holding that it is sufficient for the authority to refer to their previous countervailing decisions in which these policies, plans, and programs were considered.⁸⁴ Contrary to this ruling, however, previous panels had consistently required authorities to set out the relevant evidence and explanations in the section of their reports that is specifically dedicated to the issue of unforeseen developments.

Moreover, the panel ignored the requirement for a reasoned and adequate explanation for why these Chinese policies and measures were unforeseen and simply accepted the US's allegation. This issue is not so evident or self-explanatory as to permit a discharge of the authority's obligations. For instance, the panel should have asked the USITC to explain why, during China's WTO accession negotiations, the US negotiators could not foresee that China was to foster its renewable energy industries including the solar power sector through industrial policies and support programs. To the extent that the US's arguments concerned China's WTO obligations, especially those tailored to China, more evidence and explanation should have also been required to support the allegation that Chinese industrial policies and subsidies were ‘in direct contradiction of these commitments’⁸⁵ and hence were unforeseen.

In the panel proceedings, the US did claim that ‘what was unforeseen was the scale of [China's] effort, the speed with which it boosted Chinese production, the overcapacity that it created, and the degree to which these effects spilled into other countries where Chinese producers expanded their operations’ (original emphasis).⁸⁶ However, this claim which emphasized the scale of China's industrial policies and support and the resultant fast-growing production capacity did not appear in the supplemental report. It therefore constituted an ex post facto explanation which was consistently rejected by WTO tribunals in previous cases.

As regards the alleged developments associated with the ineffectiveness of antidumping and countervailing duties and rapid changes in the global supply chains and manufacturing processes for the products concerned, the panel accepted the US's allegation, again without asking for an explanation based on evidence to show that these developments were actually unforeseen.⁸⁷ The panel tried to justify its ruling by relying on the lack of dispute over these developments by China. However, the panel did not question whether the US's unsubstantiated allegation fell short of the established standard. The lack of dispute by China did not relieve the US from the obligation to demonstrate ‘unforeseen’ developments in a reasoned and adequate manner before the SG measures can be applied.⁸⁸

Regarding the connection between the developments identified and the import surges, the supplemental report did provide some data as noted above. However, compared to the panels’ approaches in other disputes particularly US–Steel Safeguards and India–Iron and Steel Products, the US–Safeguard Measure on PV Products decision was evidently more deferential. The panel held that it is sufficient for the USITC to simply record the evidence relating to China's industrial policies and support programs, the expansion of China's CSPV production capacity, and the increase in Chinese CSPV imports into the US separately in different reports, including previous countervailing decisions and other reports which were not dedicated to addressing the ‘as a result of unforeseen developments’ criterion.⁸⁹ Similarly, the panel accepted the US's claim that the expansion of major Chinese CSPV producers in foreign markets (i.e. Korea, Malaysia, Thailand, and Viet Nam) resulted in a significant share of imports into the US even though the supplemental report provided almost no evidence or explanation to show that the increased imports were necessarily linked to the expanded Chinese production capacity in these markets.⁹⁰

3.2 Major Shortcomings of the New Standard

Thus, the US–Safeguard Measure on PV Products decision has arguably created a new standard which is remarkably lower than the standard of review established and applied in all other SG cases (including two subsequent decisions to be discussed below). While the panel followed the AB's position that investigating authorities must comply with the ‘as a result of unforeseen developments’ test, it softened the relevant requirements to a significant degree making the test considerably easier for authorities to satisfy. As such, the panel's approach provided a way to address the longstanding criticisms about the rigidity and applicability of the test. On rigidity, however, we have shown above that the standard of review developed in the other SG cases is not unreasonably high. On applicability, we have seen a positive development of judicial guidance for the level of evidence and explanation required, as also discussed above. Here, major concerns in the existing scholarship point to the lack of clarity in determining the relevant period for assessing whether a future development is unforeseen, and the difficulties in showing such developments could have been foreseen in a long-lived agreement (such as GATT 1947) and a causal link between unforeseen developments and import surges.⁹¹

As regards the relevant period, the case law has clarified that it should be when the relevant GATT/WTO concessions or obligations were made. One may further argue that where the concessions or obligations were reviewed or modified in subsequent negotiations, the relevant period should become the latest round of negotiations.⁹² While more clarity on this issue would be desirable, WTO tribunals have had no difficulty determining the relevant period for assessment of unforeseen developments in the SG cases to date.

Furthermore, with the passage of time, it may well become increasingly difficult to argue that a future development could have been foreseen during the relevant negotiations. However, it would not be too difficult or impossible to advance this argument. To illustrate, let's consider the US claims that China's industrial policies and support programs were unforeseen in US–Safeguard Measure on PV Products. There is no shortage of evidence to suggest that these policies and programs could have been foreseen despite China's accession commitments. As early as 1986, in its 7th Five-Year Plan, China had maintained an overarching industry policy to grow the energy sector.⁹³ This was elaborated and reinforced in subsequent policy documents including the State Council's decisions which reiterated the essential role of industry policies for China's economic development, named the energy sector as one of the key sectors subject to industry policies, and encouraged the development of high value added and high-tech products including renewable energy technologies.⁹⁴ Later during the 9th Five-Year Plan period (1996–2000), the central government identified new energy as a key sector for technological development and industrialization to be supported by subsidy and other programs, with the development and application of solar power being one of the priorities.⁹⁵ This brief review of China's national policies shows that China already had and was developing robust industrial policies in the solar power sector during the period of its WTO accession negotiations (i.e. 1986–2001). Moreover, there was evidence to suggest that China's industrial policies could lead to massive support programs which in turn could boost the development of the covered sectors. For example, government initiatives introduced in the 1990s provided considerable support to the renewable energy industry. The so-called ‘Bright Project’ alone injected approximately 10 billion RMB into the solar industry, contributing to a remarkable growth of China's solar manufacturing capacity.⁹⁶ Thus, even with the passage of over three decades since the inception of China's accession negotiations, it is not impossible for China to demonstrate that the relevant developments could have been foreseen. In addition, China's industrial policies, subsidies, and other forms of government intervention in the market and their potential impact attracted considerable concerns during the negotiations.⁹⁷ These concerns led China to agree to a wide range of WTO-plus obligations including some broad commitments to restrain the conduct of state-owned enterprises (SOEs) and to ensure domestic prices for goods and services were determined by market forces.⁹⁸ However, there is no commitment under China's accession package which requires an elimination or reduction of such policies and subsidies. Rather, while the negotiators were fully aware of such policies and subsidies, they agreed to confine China's obligations to certain areas such as notification of subsidies and elimination of prohibited subsidies.⁹⁹ Like other WTO Members, Chinese industrial subsidies, other than the prohibited ones, are actionable only. The fact that the negotiators of other Members insisted on the imposition of certain China-specific rules on industrial subsidies is also strongly indicative of the understanding that China may continue to use subsidies in pursuit of industrial policies and hence the intention to facilitate the application of countervailing measures to offset the impact of Chinese subsidies.¹⁰⁰ The analysis above is not meant to re-litigate the case. Rather, it reinforces our observation that the complexity in establishing an unforeseen development can vary from case to case depending on the nature of the development, the negotiators’ knowledge about it at the relevant period and other relevant evidence. The passage of time does not make every modern event an unforeseen development such that the requirement that authorities offer sufficient evidence and explanation to facilitate an objective assessment by WTO tribunals remains necessary.

Likewise, the complexity in showing a logical link between an alleged unforeseen development and import surges also varies among individual cases. As discussed above, the panel in India–Iron and Steel Products already allowed some flexibility as to how authorities may demonstrate such a link. The panel also identified the deficiencies in the Indian authority's report, thereby providing guidance for how these deficiencies can be rectified. In EU–Safeguard Measures on Steel (Turkey), the latest SG dispute as of this writing, the panel provided further guidance and flexibility for the ‘logical link/connection’ test (which will be considered later). Such continued effort to develop enhanced clarity and flexibility has served to maintain an overall balance in the application of the ‘as a result of unforeseen developments’ test.

In contrast with the balanced approach, the US–Safeguard Measure on PV Products decision seems to have reduced the standard of review to the minimum so that allegations about ‘as a result of unforeseen developments’ can be accepted in the absence of adequate evidence and explanations. Such an excessively deferential approach may well render this prerequisite practically meaningless.

3.3 Systemic Implications

The reasons behind the panel's approach cannot be ascertained. One explanation is that the panel sought to advance a systemic change of the existing case law to be more deferential in trade remedy cases in response to US concerns about the AB's judicial activism – a major cause of the US's blocking of the resumption of the appellate review mechanism.¹⁰¹ Dating back to the early 2000s, the US already expressed the concern that WTO rulings on ‘unforeseen developments’ were ‘confusing and difficult to follow’.¹⁰² More recently, the United States Trade Representative (USTR) detailed these concerns in a report criticizing the AB's decision on ‘unforeseen developments’ as ‘erroneous’, creating obligations that Members did not agree to, and making it nearly impossible for governments to impose SG measures.¹⁰³ The more deferential approach adopted by the panel in US–Safeguard Measure on PV Products offered a solution to the US's dissatisfaction with the existing case law. In the DSB meeting following the panel report, the US welcomed the panel's approach and ultimate findings.¹⁰⁴ Indeed, one may view the panel's approach as a positive development towards alleviating the US hostility to the AB. However, for the US to change position, such a deferential approach may have to be taken on the other major issues which the US considers to be obstacles to its imposition of trade remedies.¹⁰⁵ Moreover, an effort to introduce such systemic changes to the established case law by WTO panels would cause more problems than it may resolve. It would undermine not only the predictability and certainty of the rules-based trading system but also the credibility of the appellate review mechanism. If all panels start to deviate from past AB decisions by adopting more deferential approaches to trade remedies, would it not disincentivize the US from agreeing to restore the AB?

Furthermore, an overly deferential approach would likely be treated by other WTO Members as judicial activism, leading to the same problem currently faced by the AB. In the same DSB meeting in which the US praised the panel decision, China criticized the decision as being ‘systemically harmful’ as it ‘substantially lowered the threshold’ and ‘would lead to abuse of safeguard measures and thus seriously undermine the rules-based multilateral trading system’.¹⁰⁶ These concerns led to the first and only case to date in which China appealed a panel's decision into the void.¹⁰⁷ China did so, contrary to its support for a stable and effective dispute settlement system, apparently to minimize the potential influence of the panel's approach on future disputes and SG jurisprudence. It also sent a strong message that if this or a similar approach is not avoided, it would provoke China to appeal into the void in more disputes and over time undermine China's faith in the dispute settlement system.¹⁰⁸

In addition, the panel's deferential approach in US–Safeguard Measure on PV Products would adversely affect the balance that WTO SG rules are designed to achieve, particularly the balance embedded in the ‘as a result of unforeseen developments’ test. As reiterated by the AB, SG measures are designed to allow governments to address matters of urgency, particularly where ‘an importing Member finds itself confronted with developments it had not “foreseen” or “expected” when it incurred [the relevant] obligation’.¹⁰⁹ Drawing a distinction between SG measures and the measures designed to address ‘unfair trade’ actions (i.e. antidumping and countervailing measures), the AB has also stressed that it is important to ensure that the former ‘are not applied against “fair trade” beyond what is necessary to provide extraordinary and temporary relief’.¹¹⁰ The lowered standard applied in US–Safeguard Measure on PV Products tilted the desired balance in favour of the application of SG measures by effectively removing the substantive requirements under the ‘as a result of unforeseen developments’ test. Since this case was predominantly concerned with Chinese industrial subsidies, the panel's overly deferential approach also led to the use of SG measures to address unfair trade actions enabled by subsidies, thereby confusing the distinct functions of SG measures and countervailing measures.¹¹¹ Overall, if this approach becomes the new standard, it would likely lead to tit-for-tat (ab)use of SG measures,¹¹² which will cause damage to the integrity and credibility of the WTO.

4. Decisions after US–Safeguard Measure on PV Products

Two panel decisions – i.e. US–Safeguard Measure on Washers and EU–Safeguard Measures on Steel (Turkey) – came several months after the US–Safeguard Measure on PV Products decision. In contrast with the US–Safeguard Measure on PV Products decision, however, the two panels have continued to apply the standard of review established in the previous decisions.

In US–Safeguard Measure on Washers, the US sought to challenge the established case law by contending that authorities do not need to set out reasoning and findings on ‘unforeseen developments’ in their report(s) and that claims relating to this matter can be raised exclusively in WTO proceedings.¹¹³ The panel quickly rejected the US contention and endorsed the previous decisions.¹¹⁴ The panel went on to find that the USITC's report contained no explanations or findings on ‘unforeseen developments’ and it was insufficient for the US to refer to materials in sections of the report that did not deal with this issue.¹¹⁵ Accordingly, the US's failure to pass the test was simply due to the lack of assessment of ‘unforeseen developments’, just like the failures in the category 1 cases discussed in Section 2.3.

In EU–Safeguard Measures on Steel (Turkey), the panel added more clarity to the existing jurisprudence and offered a thorough examination under the ‘as a result of unforeseen developments’ test. It observed that ‘unforeseen developments’ refer to ‘an event, fact, or circumstance that emerges or comes to light, including a new stage in an evolving situation, that was not anticipated or expected’.¹¹⁶ It confirmed that ‘the point in time at which a development must have been “unforeseen” is when the relevant obligations were incurred’.¹¹⁷ Based on the US–Steel Safeguards and India–Iron and Steel Products decisions, it clarified that developments such as changes in market conditions and financial circumstances as well as economic crises can constitute ‘unforeseen developments’.¹¹⁸ It also pointed out that to show the existence of ‘a logical connection’ between unforeseen developments and import surges, authorities do not ‘need to provide the same quantum of reasoning and evidence’ as required under the ‘causation’ test.¹¹⁹ These clarifications are consistent with our understanding of the standard of review established and applied in previous decisions, although more detailed guidance – e.g. the exact scope of unforeseen developments, the evidentiary requirements under the ‘logical connection’ test – will need to be developed in future disputes.

The panel then went on to examine whether the EU authority adequately demonstrated the ‘as a result of unforeseen developments’ criterion in its reports through the three-step test discussed in Section 2.4. First, the panel agreed that the authority satisfactorily identified three circumstances including: (1) the significant expansion of global steel production capacity; (2) the growing use of trade restrictive measures on steel products in markets of major steel importers particularly the US; and (3) the US's Section 232 measures which imposed extra tariffs on steel imports since March 2018.¹²⁰

Under the second step, the panel accepted the authority's explanation of why these circumstances were unforeseen. In relation to the global overcapacity, what was unforeseen was the expansion of the overcapacity to unprecedented levels even though overcapacity per se was known at the time of the Uruguay Round.¹²¹ As regards the rise of trade restrictive measures, the panel agreed with the EU that such measures can constitute unforeseen developments even though they are already contemplated under WTO agreements.¹²² In this respect, what was unforeseen was ‘the unprecedented and increased number of [antidumping and countervailing] measures taken by third countries’.¹²³ Regarding other types of trade restrictive measures, the authority's report set out these measures and the countries which applied them.¹²⁴ Finally, despite the longstanding US legislation which authorizes the use of Section 232 measures, the steel tariffs imposed under the legislation were still unforeseen because they covered almost all countries and had the effect of reducing around 13 million tonnes of steel imports into the US.¹²⁵ Altogether, the panel upheld the authority's findings that these measures, introduced in the context of the persistent global overcapacity, amounted to unforeseen developments. This decision suggested that the unforeseen developments arose from a confluence of events and not necessarily from each of these trade restrictive measures.

Turning to the third step, the panel found some deficiencies in the authority's explanation of how the unforeseen developments resulted in an increase in steel imports into the EU. Essentially, the EU claimed that it was sufficient for its authority to show that the unforeseen developments and the import surges occurred in the same period of time.¹²⁶ The panel disagreed and held that more explanation was required. In relation to the global overcapacity, the authority's explanation was insufficient because it merely asserted that a connection existed between the overcapacity and import surges without offering any supporting evidence.¹²⁷ Likewise, the panel also found that the authority did not provide any evidence or explanation to show that the rise of the trade restrictive measures actually led to increased steel imports into the EU.¹²⁸ It observed that ‘a more detailed analysis’ was required in this respect because each of these measures ‘typically concerns only certain steel products imported into and exported from certain countries’.¹²⁹ It further held that a more detailed analysis was required on the connection between the Section 232 tariffs and the import surges because the measures were ‘adopted by a single Member’.¹³⁰ In addition, the panel suggested that more evidence or analysis is needed to substantiate the authority's claims about the attractiveness of the EU market for exporters excluded from the US market because of the tariffs.¹³¹ While rejecting the authority's assessment of whether a logical connection existed, the panel stressed that it only needed some more evidence and analysis which should not be difficult for the authority to provide.¹³²

Thus, the panel applied the established standard of review by requiring reasoned and adequate explanations for why a development was unforeseen and how the unforeseen development resulted in an increase in the imports subject to the SG measures. Its rulings were reasonably balanced by requiring some explanations and supporting evidence but leaving some flexibility for authorities to decide how to structure such explanations and what evidence to use. Like the findings of deficiencies in the authorities’ reports in US–Steel Safeguards and India–Iron and Steel Products, the panel did not set the bar too high but merely asked the EU authority to fix some notable gaps in its reports by offering some more explanation based on evidence that was apparently available to it. At the same time, however, the EU–Safeguard Measures on Steel (Turkey) decision has left some ambiguities in the application of the standard of review, particularly in relation to the evidence and explanations required to demonstrate the link between unforeseen developments and surged imports. Future panels should seek to provide more clarity on these matters while maintaining some flexibility for them to apply the standard of review based on the facts of individual cases.

Toward this end, Table 1 provides a summary of the findings by WTO tribunals under the ‘as a result of unforeseen developments’ test in all SG cases between 1995 and 2022, as discussed above.

Table 1. A Summary of Findings on ‘As a Result of Unforeseen Developments’ in Safeguards Disputes (1995–2022)

5. Conclusion

Since the birth of the multilateral trading system, the SG mechanism has been perceived, widely and consistently, as a ‘safety valve’ necessary to allow governments to temporarily ‘escape’ GATT/WTO obligations in response to protectionist pressures at home. This flexibility, however, is constrained by GATT/WTO rules designed to discipline and dissuade the abuse of SG measures. Like in many other areas of international trade regulation, the systemic challenge has been striking a delicate balance so as to ensure the rules are neither overly restrictive nor too flexible. To a large extent, this balance hinges on how the SG rules are interpreted and applied in disputes.¹³³

Among other conditions, the criterion of ‘as a result of unforeseen developments’ has emerged as a major constraint on the application of SG measures. WTO tribunals have developed a growing body of case law to elaborate the relevant requirements and have found that governments imposing SG measures failed to fulfill some of these requirements in almost all SG disputes to date. Through a comprehensive and detailed review of the jurisprudence and the application of the relevant standard of review by WTO judges, this paper fills a gap in the literature and presents three fresh arguments. First, contrary to the criticism about the AB's decision to ‘revive’ this criterion as a prerequisite for the application of SG measures, we show that the AB's decision is a plausible interpretation of the drafting record of the SG Agreement. While this criterion received little attention during the Uruguay Round negotiations, the record does suggest that it was possible that the negotiators simply overlooked the possibility that GATT Article XIX may continue operating cumulatively with the SG Agreement. Second, the standard of review developed and applied by WTO tribunals in most of the cases has maintained a proper balance. While requiring authorities to provide some explanations and evidence to support their claims, the tribunals have left sufficient room for authorities to choose their own approaches to do so and were not unreasonably demanding in terms of the degree of explanation and evidence required. The standard of ‘reasoned and adequate explanations’ is essentially subject to a case-by-case analysis which leaves the flexibility for WTO tribunals themselves to assess whether authorities have complied with their obligations. Third, in supporting the US's SG measures on CSPV imports from China, the panel on US–Safeguard Measure on PV Products applied a remarkably lower standard of review, thereby tilting the balance towards the use of SG measures. The panel's approach may lead to growing abuse of SG measures and adversely affect the balance that WTO SG rules are designed to achieve, particularly the balance embedded in the ‘as a result of unforeseen developments’ test. It also amounted to a disruption of the development of jurisprudence and hence of the predictability and certainty of the dispute settlement system. If this or a similar approach is not avoided in future disputes, it would provoke China to appeal into the void in more disputes and frustrate China's faith in the system over time.

Amid the ongoing wave of economic nationalism and unilateralism in the world economy, SG measures, which used to be an unpopular trade defence instrument,¹³⁴ appear to be on the rise and may well become increasingly utilized.¹³⁵ This constitutes one of the (potential) challenges that the WTO must effectively deal with to show its continued competence and credibility in promoting trade liberalization and cooperation. As far as trade disputes are concerned, this challenge calls for a more cautious and sophisticated approach to pursue not only a proper balance in maintaining the underlying function of the SG mechanism but also the predictability and certainty of the dispute settlement system.