Hostname: page-component-745bb68f8f-s22k5 Total loading time: 0 Render date: 2025-01-14T23:26:07.786Z Has data issue: false hasContentIssue false

Second-Generation Flanking Policies: Addressing Extraterritorial and Non-Economic Costs of Trade Liberalization

Published online by Cambridge University Press:  13 January 2025

Timothy Meyer*
Affiliation:
School of Law, Duke University School of Law, 210 Science Drive, Durham, NC 27708-0187, USA
Rights & Permissions [Opens in a new window]

Abstract

Flanking policies – policies that aim to address potential negative effects of trade liberalization, and/or the concerns of domestic stakeholders regarding those negative effects, and that are either legally or factually linked to trade liberalization – have been a critical component of international trade policy since at least 1962. Over the years, however, flanking policies have changed. This Article argues that there is a heretofore unnoticed distinction between what I term first-generation flanking policies and second-generation flanking policies. Specifically, first-generation flanking policies target negative economic effects, or costs, of trade liberalization experienced within the enacting country. Trade adjustment assistance is the paradigmatic example. By contrast, second-generation flanking policies target non-economic costs that arise outside of the enacting country. Examples include the European Union's Carbon Border Adjustment Mechanism, Deforestation-free Products Regulation, and the United States' Uyghur Forced Labor Prevention Act. Because second-generation flanking policies directly target foreign activity, they often employ more trade-distorting policies – tariffs, imports bans, and associated administrative hurdles for imports – than first-generation flanking policies, which more often relied on domestic subsidies. Moreover, they reflect a significant reorientation of the limits of state authority in international trade law. Whereas authority to tax and regulate production in international economic law has historically been based primarily on a territorial link to productive activity, second-generation flanking policies target production but rely on a territorial nexus with consumption of goods and services.

Type
Original Article
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
Copyright © The Author(s), 2025. Published by Cambridge University Press on behalf of The Secretariat of the World Trade Organization

1. Introduction

In 1962, President Kennedy proposed the Trade Expansion Act to the US Congress. Prior to that time, US Presidents had negotiated tariff reductions with other nations pursuant to delegated authority under the Reciprocal Trade Agreements Act of 1934 and its successor statutes. The economic challenges and opportunities created by the emergence of the Common Market in Europe and the US security interests in deeper economic integration required, in President Kennedy's view, greater trade liberalization, which the act would facilitate. At the same time, though, the Kennedy Administration recognized that the reduction in tariffs since the end of the Second World War had disrupted some US industries, and deeper economic integration would likely cause further disruptions. He thus paired his proposal for expanded authority to negotiate trade liberalization with a proposal for a new program, Trade Adjustment Assistance (TAA), that would provide various kinds of financial assistance to firms and workers that had lost businesses or jobs due to competition with imports. President Kennedy described TAA's aim as ‘tak[ing] out an insurance policy’ for cases ‘where individual companies or groups of workers will face genuine hardships in trying to adjust to this changing world and market, and lack the resources to do so. … Instead of the dole of tariff protection, we are substituting an investment in better production’.Footnote 1

TAA is perhaps the most well-known of what I term first-generation flanking policies. Flanking policies, in general, aim to address the negative effects (or costs) of trade liberalization, or domestic stakeholders concerns with those effects, in a manner legally or factually connected to trade liberalization.Footnote 2 Although governments have deployed flanking policies for many years, the legal, political, and economic relationship between flanking policies and trade agreements – and the relationship between different kinds of flanking policies, the subject of this Article – pose a series of novel research questions that scholars have only recently begun to address.Footnote 3

As I explain in Section 2, first-generation flanking policies have two features that distinguish them. First, they target the domestic losses caused by trade liberalization, rather than losses that are most directly experienced in foreign countries. Second, first-generation trade policies focus on economic losses, as opposed to other kinds of costs that trade liberalization may create. In short, they try to make the ‘losers’ from trade liberalization better off.Footnote 4

Over the years, however, flanking policies have evolved to deal with a broader set of concerns about trade liberalization. As Section 3 explains, second-generation flanking policies target extraterritorial, non-economic costs of trade liberalization, rather than the domestic, economic costs targeted by first-generation flanking policies. Examples include the European Union's (EU) Carbon Border Adjustment Mechanism (CBAM), the EU's Deforestation-free Products Regulation, and the United States' Uyghur Forced Labor Prevention Act. To be sure, these policies target activities that may have effects within the enacting country, and the distinction between economic and non-economic targets is blurry in practice. CBAM, for instance, targets foreign carbon-intensive production – an economic activity – for the purposes of mitigating climate change, including within the EU. But the CBAM targets production primarily because of its environmental effects, rather than directly because of its economic features. Likewise, although the EU's ultimate purpose is to effect environmental conditions in the EU, CBAM achieves this purpose only indirectly, by trying to influence production and trade patterns outside of the EU. Having enacted domestic measures that make carbon-intensive production more costly within the EU, CBAM is necessary to prevent carbon leakage – the importing of cheap carbon-intensive products from abroad to replace more expensive but less carbon-intensive domestic products. In effect, CBAM helps ensure that EU consumers do not buy carbon-intensive products, no matter where in the world they are produced. TAA, by contrast, acts directly upon an economic loss – primarily job losses – suffered within the enacting country, the United States.

These second-generation flanking policies portend significant changes for international trade law, as I discuss in Section 4. First, because these flanking policies directly target extraterritorial activity, they often employ more trade-distorting policies – tariffs, imports bans, and associated administrative hurdles for imports – than first-generation flanking policies, which more often relied on subsidies. Second, they reflect a significant reorientation of the limits on state authority in international trade law. Whereas authority to tax and regulate production in international economic law has historically been based primarily on a territorial link to productive activity, policies are conditioned on production but rely on a territorial nexus with consumption of goods and services. WTO members adopting second-generation flanking policies that implicitly rely on consumption as a sufficient basis of authority to tax and regulate productive activities should offer interpretations of WTO rules that support their view of their own authority. I conclude with some thoughts on what such interpretations might look like.

2. First-Generation Flanking Policies

From an economic point of view, the overarching justification for trade liberalization is the promotion of economic growth through the more efficient allocation of resources. Reducing barriers to trade encourages specialization, allowing countries to produce what they are comparatively better at producing and trade (at lower prices) for everything else. In general, the result is lower prices for consumers, increased efficiency and innovation for producers, and increased profits for efficient and innovative firms. These economic effects produce other effects as well, which have served as justification for trade liberalization, such as increased political integration and the promotion of peace and security.Footnote 5

For all the good it does, though, scholars and policymakers have long recognized that trade liberalization also has distributional effects: it creates winners and losers.Footnote 6 Historically, and even today, the losers from trade are primarily characterized in economic terms. They are the firms that go out of business if they are forced to compete with cheaper imports, the farmers that have to sell their land for the same reason, and the workers that lose their jobs when either of these things happen or just when a multinational company decides to relocate its production facilities to a country with lower labor costs and fewer regulations. Economists and policymakers argue that many of these losses are merely short-term disruptions.Footnote 7 The efficient allocation of resources that trade liberalization promotes should lead to the creation of new jobs, ideally better paying ones. Coupled with lower consumer prices that result from trade liberalization (functionally a tax cut), the losers from trade should ultimately be beneficiaries as well.

In the early 1960s, though, the Kennedy Administration had already recognized two realities. The first was an economic one. Not all firms, farmers, and workers would be able to seamlessly move into new economic roles. There would be short-term transition costs, and some of those costs might be prolonged if subsets of the labor market failed to share in the more general economic growth.Footnote 8 The second reality was a political one. Import-competing firms, farmers, and their employees had an incentive to oppose trade liberalization if they thought they might bear the brunt of these transition costs. Moreover, the degree of political opposition could be disproportionate to the actual economic effects. Risk-averse workers might oppose trade liberalization out of the fear that they would lose out, even if trade liberalization might in fact benefit them. Community stakeholders who otherwise benefit from trade liberalization might nevertheless oppose trade liberalization out of a concern for others in their communities.

First-generation flanking policies aim to address these two problems. Specifically, as I use the term here, first-generation flanking policies target (1) negative economic effects (2) experience within the enacting country as a result of trade liberalization. They are, in this sense, inward looking.Footnote 9 These programs, by and large, are domestic programs that are neither mentioned in, nor required by, the trade agreements that codify the trade liberalization provisions in international law. In the United States, for instance, TAA is commonly authorized in the same legislation that authorizes either the negotiation or implementation of new trade agreements, but no reference appears to TAA in the trade agreements themselves.Footnote 10 In this sense, first-generation policies such as the TAA are factually linked to trade agreements, and may be legally linked under domestic law to the extent that a single piece of legislation approves the trade agreement and the flanking policy, but they are not usually linked as a matter of international law.Footnote 11

First-generation flanking programs vary in the extent to which they address costs explicitly and purposefully, as opposed to implicitly and/or incidentally. TAA in the United States explicitly and purposefully targets workers, firms, and farmers that have been laid off due to new competition with imports. Although the criteria have changed over time, claimants for benefits under TAA have generally had to show that they lost their job (or business in the case of firms and farmers) as a result of new trade agreements.Footnote 12 This explicit requirement serves two functions. First, it allows the US government to claim that it is doing something specifically to combat harms caused by trade liberalization. As Kim and Pelc have found, TAA reduces political pressure for protectionism,Footnote 13 making TAA politically useful as part of a program of trade liberalization. Second, it limits the program's costs. Causality is difficult to prove in practice, even when a worker may have in fact lost their job in part due to competition with imports. Limited costs, in turn, make it easier for members of Congress to support the program, although over time the difficulty in actually accessing TAA has led to a high degree of cynicism about the program's effectiveness, with labor leaders referring to the program as ‘burial insurance’.Footnote 14

Other countries have programs similar to TAA. In 2006, the EU created the European Globalization Adjustment Fund, which, like TAA, provides funding for training, job searches, and business startups for workers who have lost their jobs as a result of globalization.Footnote 15 Like TAA, the Globalization Adjustment Fund contains eligibility criteria requiring an explicit showing that the job losses stemmed from globalization, although the definition of globalization is broader than equivalent eligibility criteria used in TAA's various iterations. As with TAA, however, the Globalization Adjustment Fund has not been as effective as advocates might have wished. A study conducted by the European Parliament found that, in part, applications for the program were lower than expected because the eligibility criteria were not broad enough, both in relation to the definition of globalization and the number of workers at a firm that had to be laid off to trigger eligibility (500 workers), a criterion that limited the program's use for small and medium-size enterprises.Footnote 16

Other examples abound. Canada introduced a TAA program in 1965 in response to the United States–Canada Automotive Products Agreements (or Auto Pact) and deployed several versions of the program over the ensuing decades.Footnote 17 Those programs largely ceased operation at the end of the 1980s, in part due to a government-appointed commission arguing that it was impracticable to try to trace specific job losses to specific trade liberalization policies.Footnote 18 South Korea introduced its first TAA program in 2007 in order to gain domestic support for the pursuit of new preferential trade agreements.Footnote 19 Australia introduced its Special Adjustment Assistance and Special Assistance for Non-Metropolitan Areas in 1973, although it canceled those programs in 1977 due to significant demand that created budgetary constraints. Since then, it has deployed more targeted sectoral or regional policies.Footnote 20 In 2017, even China began experimenting with a TAA-like program as part of its Shanghai Pilot Free Trade Zone.Footnote 21

These programs provide subsidies that are explicitly linked to costs of trade liberalization. They can be compared with at least two other kinds of programs. The first are measures that are not factually or legally linked to trade agreements but nevertheless mitigate the negative effects of trade liberalization.Footnote 22 Mitigating measures that are not flanking policies include general social safety net programs, such as unemployment insurance and benefits, government-provided healthcare, housing policies that prevent eviction in the event of economic hardship, and basic income support programs. These programs generally do not require a showing that economic hardship or job loss resulted from trade liberalization, but job losses resulting from trade liberalization would be covered.

Policymakers may face a tradeoff when choosing between general mitigating measures and flanking policies. On the one hand, general mitigating measures are often more available, and hence more effective on a societal scale, at addressing the negative effects of trade liberalization precisely because they are not linked to trade agreements. Indeed, a common refrain regarding TAA-like programs is that the eligibility criteria, while politically useful at least in the short-run, ultimately harm the program's effectiveness. Research has found a correlation between countries that are more open to trade and countries that have more generous social safety nets.Footnote 23 This correlation could suggest that general social safety net policies are in fact linked to trade liberalization, at least in citizens' political consciousness. On the other hand, general mitigating measures are arguably less politically useful in generating support for trade liberalization, since recipients do not experience the receipt of benefits as part of a political quid pro quo.Footnote 24

The second type of measure that can be compared to TAA-like programs are trade remedies, such as antidumping duties, countervailing duties, and safeguards. Unlike both TAA and general social safety net programs, trade remedies are not subsidies. They are, rather, trade restrictions. But trade remedies nevertheless can be classified as flanking policies.Footnote 25 They aim to limit the harms caused by cheap imports and, since they are explicitly authorized in trade agreements, they are clearly linked to trade liberalization.Footnote 26 Trade remedies also fit the definition of a first-generation flanking policy. All three kinds of trade remedies require a demonstration that some activity (which varies depending on the kind of trade remedy) has caused injury to domestic industries within the enacting country, thereby satisfying the domestic component of the definition of first-generation flanking policies.Footnote 27

The relevant activity for each kind of trade remedy varies in the extent to which it is domestically focused. Safeguards are the most domestically focused, requiring a sudden surge in imports and not requiring any demonstration of any kind of unfair trade practice on the part of exporters. Countervailing duties are at the other extreme, as they target subsidies to foreign producers and thus do not involve any conduct within the enacting country. Antidumping duties occupy a middle ground, focusing on a comparison of pricing decisions in foreign markets and the domestic market of the enacting nation.

Whether subsidies or trade remedies, first-generation flanking policies sought to generate political support for trade liberalization by either compensating and helping those harmed adjust to the new economy or providing selective protection in certain instances. The hope, which some research supports, is that subsidies might reduce the pressures to use protection to address trade liberalization's economic harms.Footnote 28 But in either case, these first-generation policies remained focus on economic dislocations within the enacting country.

3. Second-Generation Flanking Policies

Second-generation flanking policies, as I use that term, differ from first-generation policies in two respects. First, their primary focus is on the non-economic costs of trade liberalization, such as environmental damage or labor and human rights. Second, second-generation policies are extraterritorial in the sense that their direct targets are activities that occur in foreign countries, rather than the enacting country. As noted above, examples of second-generation flanking policies include the EU CBAM, the EU Deforestation-Free Products Regulation, and the US Uyghur Forced Labor Prevention Act.

Second-generation flanking policies are the product of three trends. One is the growing awareness of the ways in which trade liberalization can affect values other than core economic ones. The second is the realization that in a globalized economy in which trade barriers have already been substantially reduced, national legislation to address these broader problems is insufficient if it is not paired with an effective international trade policy. The third is a political dynamic in which import-competing interests support trade restrictions motivated by non-economic objectives in an effort to obtain protection, but they might struggle to muster the political support needed if framed as straightforward protection. I discuss these factors in turn.

3.1 Non-Economic Consequences of Trade Liberalization

The growing awareness of how trade liberalization affects values other than purely economic ones is an oft-told tale. During the early 1990s, the first Bush Administration in the United States negotiated the North American Free Trade Agreement (NAFTA).Footnote 29 After losing the 1992 presidential election to Bill Clinton, Bush signed the agreement but left it to his successor to obtain congressional approval. To build support for the agreement, President Clinton negotiated the so-called NAFTA Side Agreements on the Environment and Labor.Footnote 30 The aim of these provisions was to ensure that NAFTA members did not lower their environmental and labor standards in order to induce the movement of jobs across borders. From the point of view of the United States, the proponent of the provisions, these were outwardly facing provisions aimed primarily at Mexico (as the United States and Canada already had a free trade agreement). Although critiqued at the time as not going far enough, and modest by today's standards, the NAFTA Side Agreements inaugurated a process through which ‘trade and … ’ provisions have become increasingly prevalent and strict in terms of both substantive and procedural terms.Footnote 31

The inclusion of labor and environment provisions in trade agreements in the mid-1990s (as opposed to earlier) might seem odd. After all, the Democratic Clinton Administration was centrist, reflecting the general rightward shift of US politics, and the labor movement had lost some of its political clout by then.Footnote 32 The modern environmental movement was born in the 1970s, and yielded a series of landmark treaties in the 1980s and early 1990s, including the Montreal Protocol, the UN Framework Convention on Climate Change, and the UN Biodiversity Convention. Trade agreements only belated began to take notice of this growth in international environmental law.

The end of the Cold War, however, opened up political space for the labor and environmental movements to be more influential within trade policy. From the end of World War II forward, politicians, again especially in the United States, justified trade liberalization in terms of foreign policy objectives.Footnote 33 President Kennedy's speech introducing the Trade Expansion Act, for instance, cited the way in which trade liberalization promotes the ‘common defense’ and indicated that the Act sought to allay fears ‘that the United States may someday abandon its commitment to European security’.Footnote 34

With the Cold War over, however, general foreign policy arguments, especially those focused on security in a polarized world, declined in political salience. The justification for trade agreements thus began to depend more on their economic benefits. That, in turn, has invited more scrutiny of both the distributional implications of trade policies, typically those addressed via first-generation flanking policies, as well as the non-economic costs that trade liberalization creates.Footnote 35 Although trade institutions, such as the WTO and trade officials at the national level, rarely addressed these costs as quickly as advocates would have liked, evidence of the growing need to engage with trade's non-economic costs are reflected both in institutional roles – such as the WTO's Committee on Trade and the Environment or the Assistant US Trade Representative for the Environment and Natural Resources – as well as in an increased receptiveness by WTO dispute panels to environmental defenses.Footnote 36

3.2 How Extraterritorial Activities and Foreign Policies Undermine Domestic Political Commitments

If the end of the Cold War empowered political groups to push for trade policy that addressed the non-economic costs of trade liberalization, it also highlighted the way in which trade liberalization could undermine government policies pursuing broader social objectives. While this had long been clear in the case of job losses caused by import competition,Footnote 37 increased interest in addressing non-economic harms domestically started to put pressure on trade policy.

The basic problem that trade liberalization creates for the pursuit of broader social policies is often, and unfortunately, referred to as ‘leakage’. If a government taxes or regulates certain methods of production, domestic production becomes more expensive relative to foreign production. If trade barriers are sufficiently low, it becomes cheaper for domestic consumers to purchase imported products as opposed to domestic products.Footnote 38 The term ‘leakage’ is unfortunate because it suggests that the mechanism for this substitution effect is domestic firms shifting (or leaking) their productive activities overseas. While that does happen, the substitution effect does not depend on specific firms relocating production facilities and jobs. Rather, it is sufficient that domestic consumption shifts from domestic to imported goods and services after the imposition of domestic rules on production make domestic goods and services relatively more expensive.

This substitution effect can undermine domestic policies and priorities in several ways. First, it can contribute to the same kind of domestic economic harms to which first-generation flanking policies respond.Footnote 39 Second, it can prevent domestic policies from having their desired effect by giving a cost advantage to imported goods and services from countries that do not regulate the activity regulated in the domestic market. Because trade liberalization makes it less costly to produce goods in one place and sell them in another, trade liberalization means that a nation's consumers continue to fund activities that their government is trying to limit or prohibit.

Climate change offers a case in point. Since 2005, the EU has operated its Emissions Trading System (ETS), its primary policy to induce its domestic industries to decarbonize. In essence, the ETS requires producers in covered sectors to purchase permits, or allowances, for their emissions. By limiting the number of available allowances, the government can increase the cost of carbon-intensive production policies, thereby inducing producers to decarbonize their production.

A central problem with the ETS has always been the fear of carbon leakage. Higher production costs in the EU will lead to increased consumption of carbon-intensive imports. This effect both undermines EU competitiveness and also means that the ETS will not be as effective in reducing global carbon emissions. Because carbon emissions contribute to global warming no matter where they occur, the substitution effect would defeat the point of the ETS.

The EU's initial response to this problem was to give industries free allowances, but because the price of carbon in a cap-and-trade system, like the ETS, is determined in part by the supply of available allowances, increasing the supply at no cost to producers drove down the price of carbon within the EU.Footnote 40 To address this problem, as of October 2023 the EU began to implement its CBAM, which will charge importers a duty based on the carbon emitted during production.Footnote 41 The amount of the duty is based on the price for the same level of carbon emissions for the same covered product under the ETS.Footnote 42

The direct target of the EU CBAM is thus foreign production processes. The EU has long been a leader in trying to combat climate change, but its domestic efforts in that vein are undermined if carbon-intensive imports can come into the EU without having to pay a cost to account for their carbon emissions. The CBAM aims to ensure that the EU can regulate its own producers' carbon emissions and ensure that any domestic reductions achieved actually contribute to reducing climate change. Put differently, if it works as intended, the EU CBAM is a necessary component of an overall climate mitigation strategy aimed at ensuring that European consumption does not fund carbon intensive production practices. Without the CBAM, the EU would struggle to reduce the extent to which its consumers fund climate change.

Other second-generation policies share this feature. Deforestation, for example, exacerbates climate change and contributes to biodiversity loss, regardless of where in the world it occurs. Frans Timmermans, the European Commission's vice president, justified the Deforestation-free Products Regulation explicitly in consumption terms, saying that ‘EU demand for commodities like palm oil, soy, wood, beef, cocoa, and coffee are strong drivers of deforestation’.Footnote 43 Similarly, the European Commission has written that ‘[a]s a major economy and consumer of these commodities linked to deforestation and forest degradation, the EU is partly responsible for [deforestation] … This initiative will provide a guarantee to EU citizens that the products they consume on the EU market do not contribute to global deforestation and forest degradation.’Footnote 44 Of course, not all of these products from deforested land are themselves produced in the EU. Soy and palm oil, for example, are primarily produced in the Americas and Asia. But they compete with an EU crop, rapeseed, that – like soy and palm oil – is used to make the biodiesel fuel that has been critical to EU plans to decarbonize its transportation sector.Footnote 45 Importing either soy and palm oil into the EU for purposes of making biofuels, or importing biofuels made with those products, shifts EU consumption to products that are made through deforestation abroad, rather than domestic deforestation-free products.Footnote 46

Efforts to enforce labor standards overseas also share this justification. Poor labor practices overseas are not a public good (or bad) in the same sense that climate change or biodiversity protection are. They do not, for instance, have physical effects in the enacting country. But the labor movement has long had a global component, with labor leaders embracing an increase in global labor standards as a moral imperative, in addition to a way to ensure that production does not shift to countries with low labor standards. Poor labor practices in foreign countries, in other words, have long been viewed as a reasonable object of international concern.Footnote 47 Also, like environmental concerns, if trade barriers on products made with such practices are not imposed, countries cannot ensure that their consumption does not fund practices that are illegal or viewed as immoral domestically. A commitment to ending forced labor or providing opportunities for unionization, for instance, requires both domestic laws implementing those commitments at home and trade laws, such as the Uyghur Forced Labor Prevention Act, the EU's new Forced Labor Regulation, or the USMCA Rapid Response Mechanism, that can trigger trade barriers to prevent domestic consumption from supporting those practices abroad.

3.3 Mixed Motive Policies

The increased acceptance of policies that target extraterritorial, non-economic costs of trade liberalization has also shifted the dynamics of how interest groups build coalitions to enact such policies.Footnote 48 First-generation flanking policies focused on specific products or sectors of the economy. TAA, as noted above, has historically had restrictive eligibility criteria that mean the program is not widely available. Trade remedies are product specific. Even early second-generation flanking policies, such as import bans on tuna or shrimp caught in ways that endanger dolphins or sea turtles, are limited to narrow sectors of the economy. By contrast, most contemporary second-generation flanking policies, like the EU CBAM or the US and EU prohibitions on forced labor, apply either economy-wide or to foundational sectors of the economy. Second-generation policies thus make it possible to justify trade restrictions that are potentially much broader in scope than those available under first-generation policies.

Second-generation flanking policies are thus highly likely to be mixed motive policies.Footnote 49 They seek to ameliorate costs of trade liberalization, but they are also supported – and their design often influenced – by domestic import-competing groups that benefit from the reduction in competition that accompanies trade restrictions. For protection-seeking interest groups, supporting second-generation flanking policies is often the most realistic way to obtain broad protection from imports. The result is Baptist-Bootlegger coalitions in which second-generation flanking policies, as they become more acceptable, will attract more political support, making them more prevalent.Footnote 50

Elsewhere, I have referred to this dynamic as ‘channeling’.Footnote 51 The basic idea is that government policies require a degree of political support in order to be enacted. To gain that support, proponents must craft a policy that appeals to a sufficiently wide coalition of interests. However, many policies that pursue non-economic values, especially the kinds of policies protected by the exceptions in GATT article XX, create benefits that are diffuse, while they may create costs, such as foregone economic investments or trade, that are relatively concentrated. The resulting collective action problem may mean that these policies are not pursued because they lack sufficient political support.

Protectionist policies have the opposite problem. They create concentrated benefits for the protected economic interests, but they create costs for consumers that can be diffuse (in the case of end-use consumers) or concentrated (in the case of the purchasers of intermediate products). Moreover, the possibility of retaliation following dispute resolution can lead to concentrated costs for exporters. These costs create disincentives for using protectionist policies.

Policies designed to appeal both to proponents of non-economic policies and protectionist interests can solve both groups' problems. The former gain political support for a policy from a group that stands to benefit in a specific way from the policy. US shrimpers, for instance, benefitted directly from US efforts to prohibit foreign shrimp on the grounds of risks to sea turtles.Footnote 52 At the same time, protectionist interests are likely to encounter less resistance to their policies if the policies stand a chance of being justified under WTO exceptions, even if dispute panels ultimately reject the exception's application. The mere possibility of legal justification gives exporters that might be the target of retaliation less reason to object. A plausible argument about legality also may reduce internal resistance from government lawyers.Footnote 53 Put differently, the availability of plausible legal arguments – even ones that are not likely to succeed under existing interpretations of WTO rules – can smooth the way for the adoption of mixed-motive, second-generation flanking policies.Footnote 54

A case in point involves the development of the EU's deforestation policy prior to the Deforestation-free Products Regulation. As part of its Renewable Energy Directive, the EU had sought to create incentives for the increased use of biodiesel fuel in the transport sector. These incentives stimulated development of the EU's own biodiesel sector, but it also led to a flood of new imports from, among other places, Indonesia and Argentina. The EU first employed trade remedies, a first-generation flanking policy, in an attempt to protect its domestic biodiesel industry.

After WTO disputes went against the EU in cases brought by Argentina and Indonesia, the EU shifted tactics.Footnote 55 As part of its revised Renewable Energy Directive, it limited biofuel incentives to biofuels not presenting a high risk of indirect land use changes, i.e., deforestation. Following a study, the EU concluded that only palm-based biodiesel from Indonesia and Malaysia presented a sufficiently high risk of deforestation. The EU thus achieved an environmental objective, as well as a protectionist objective that it had sought to achieve directly via trade remedies, in a single deforestation measure.Footnote 56

4. Addressing Extraterritorial and Non-economic Costs

Second-generation flanking policies reflect a significant reorientation of international trade policy, especially at the domestic level. This reorientation, in turn, is putting significant stress on the international trading system. Efforts to respond to extraterritorial, non-economic costs of trade liberalization bring flanking policies into conflict with international trade rules.

I discuss two related types of conflict here. First, second-generation flanking policies employ different governmental tools than first-generation flanking policies. The latter often relied on subsidies, which are feasible for first-generation policies because the subsidy received is entirely within the enacting jurisdiction. Because second-generation flanking policies target extraterritorial production practices, though, governments primarily use sticks instead of carrots. These sticks take the form of import restrictions, as in the case of the EU Deforestation-Free Products Regulation and the US Uyghur Forced Labor Prevention Act, and duties on imports conditioned on the production process, as in the case of the EU CBAM and the proposed Global Arrangement on Sustainable Steel and Aluminum. These tools, in turn, are much more likely to violate established understandings of international trade rules and thus pose a threat to the existing trade law system.

Second, I argue that the inconsistency of second-generation flanking policies with existing interpretations of trade law is part of a larger shift in international economic law. Because these policies are primarily unilateral, rather than embedded in international trade agreements, they implicitly rest on claims of authority to tax and regulate that differ significantly from the theories underlying most modern international trade law. Specifically, jurisdiction in international economic law to tax and regulate production has historically rested on a territorial nexus with production. If production occurred within a nation's territory, that nation had primary and usually exclusive jurisdiction to tax and regulate production. First-generation flanking policies fit comfortably within this norm. They targeted dislocations to productive resources (primarily labor) within the enacting country.

Second-generation flanking policies, by contrast, often target production processes abroad, such as carbon-intensive production or the use of forced labor, and thus are incompatible with limits on authority that rest on a territorial link to production. Instead, second-generation flanking policies rest on a theory that consuming a good or service provides a sufficient basis for imposing taxes or regulations that are conditioned on extraterritorial production practices. I refer to this shift in norms as a shift from production jurisdiction to consumption jurisdiction. WTO members that have adopted second-generation flanking policies that implicitly rest on consumption jurisdiction have not, as yet, articulated clearly how their claims of authority should influence the interpretation of established trade law doctrines. Section 4.2 thus concludes with a brief outline of how members might formulate such interpretations.

4.1 From Subsidies to Trade Restrictions

The political economy and legality of first-generation flanking policies rests in large part on the fact that they primarily address economic harms within the enacting country. This creates political and legal dynamics that permit the use of flanking policies, especially subsidies, that are relatively transparent, are somewhat self-disciplining insofar as they impose costs on the enacting government, and are relatively less disruptive to international trade patterns. However, as the nature of the costs of trade liberalization that governments wish to address become extraterritorial and non-economic, the political dynamics that shape the specific kinds of policies change, as does the legality of those policies. Specifically, policies that try to confront non-economic, extraterritorial harms from trade liberalization impose fewer domestic costs that discipline their use and are more likely to take the form of trade restrictions that both disrupt trade to a greater extent and violate international trade rules more frequently than first-generation flanking policies do.

4.1.1 The Types of Measures Used in First-Generation Flanking Policies

Putting aside trade remedies for the moment, first-generation flanking policies mostly involve various kinds of labor market subsidies. For instance, as described in Section 2, TAA offers job training and relocation assistance to workers who have lost their jobs due to US trade agreements. More general social safety net programs, such as unemployment insurance or government-provided healthcare, offer the same kinds of fiscal subsidies.

The subsidies that make up the bulk of first-generation flanking policies are both relatively less trade distorting and relatively more likely to be self-disciplining. Subsidies, for example, are a relatively transparent form of redistribution. The government raises revenue via taxation and then transfers that revenue to subsidy recipients via government spending. As such, subsidies can be politically controversial within a domestic polity. TAA in the US is chronically underfunded relative to its stated objectives in part due to objections to government spending for subsidies to labor groups. Subsidies, in other words, show up as spending in government accounts and can be characterized as handouts, making them subject to political attack, especially in countries in which a ‘small government’ ethos is prevalent.

However controversial (or not) subsidies may be within a domestic polity, they would be considerably more controversial if given to foreign groups.Footnote 57 Put differently, the large-scale use of subsidies as a form of flanking policy – that is, to address negative effects of trade liberalization, as opposed to other foreign policy objectives that may be tied to foreign aid – depends as a political matter on the bulk of those subsidies going to recipients in the granting country. Spending money to directly benefit citizens of foreign countries is considerably more difficult politically than spending money to directly benefit one's own citizens.Footnote 58

To be sure, there are examples of trade subsidies that go to foreign companies. Famously, rather than remove its subsidies for its own cotton farmers, the US paid Brazil what in effect amounted to subsidies for Brazilian cotton farmers.Footnote 59 However, because the US payments were linked to the continued ability to subsidize US farmers in the form of a settlement of a WTO dispute with Brazil, they do not obviously show a willingness to subsidize economic activity in other countries. Governments also frequently subsidize businesses that operate in foreign countries.Footnote 60 Oftentimes these subsidies stem from a national connection to the company involved, even if the economic activity occurs overseas. A contemporary example that would clearly qualify as a flanking policy would be Japan's ‘China exit’ subsidies for Japanese firms that relocate their supply chains to countries other than China.Footnote 61 These subsidies aim to address the risks created by supply chain concentration that trade liberalization, coupled with Chinese government policies and comparative advantages, have created. Most such subsidies, however, would not qualify as flanking policies. They are generally the product of rent-seeking by firms, rather than a government's intention to address externalities created by trade liberalization by inducing changes in behavior in foreign countries.

From a legal point of view, international trade rules generally permit subsidies. This is perhaps in part due to the fact that the transparent cost of subsidies to the granting jurisdiction may discipline their use politically, and partially due to the fact that subsidies, especially production subsidies, are among the least distorting types of government interventions in the market.Footnote 62 Indeed, the kinds of subsidies that violate WTO rules are those that seek to shift some of the costs of the subsidies to foreigners unrepresented within the domestic political process. Subsidies that contain domestic content requirements, for example, violate both the GATT's national treatment rule and the Subsidies and Countervailing Measures (SCM) Agreement.Footnote 63 Other subsidies, if they are specific to an industry or enterprise, can be actionable under the SCM Agreement if they cause adverse effects. In practice, though, the need to demonstrate the existence of a specific subsidy that causes adverse effects is extremely difficult, which has left subsidies largely unregulated at the international level.Footnote 64

As noted above, first-generation flanking policies also involve some trade restrictions in the form of trade remedies. The use of trade remedies predated the GATT in 1947 and thus states incorporated rules permitting their continued use into the GATT.Footnote 65 Trade remedies are thus in general legal, although in practice WTO panels and the Appellate Body significantly tightened the interpretation of rules governing states' use of trade remedies.Footnote 66 States have also used trade remedies in a tit-for-tat fashion to deter their expansive use.Footnote 67

Trade remedies mostly come in the form of duties imposed either to halt a sudden surge in imports or to respond to unfair trade practices that cause injury within the enacting country, including imposing countervailing duties on foreign subsidies. The effect of these flanking policies is relatively less transparent than subsidies (although more transparent in terms of its price effects than regulations) for the familiar reason that they alter the prices consumers pay, but also because they show up as revenue rather than spending on the government's ledger even though they impose economic costs on the enacting country. To the extent that consumers are unaware of the existence of these duties, there is little to discipline their use from a political economy point of view. In practice, end-use consumers are likely to have relatively little information about trade remedies.

The consumers of intermediate goods (themselves producers) are likely to be aware and to push back against the use of these tools. The Biden Administration's decision to lift safeguard duties on solar panels provides one illustration.Footnote 68 Solar panel installers in the United States represent a very informed and significant portion of the US solar industry, one consistently opposed to the US safeguard duties on solar panels. Although lobbying by organizations such as the Solar Energy Industries Association failed to prevent the initial adoption of the solar duties by the Trump administration, they consistently supported lifting them and eventually prevailed.Footnote 69 More generally, in a world of complex supply chains, duties may be subject to political pressures more similar to subsidies than might have been true in prior generations.

4.1.2 The Types of Measures Used in Second-Generation Flanking Policies

If first-generation flanking policies consist of subsidies with some trade restrictions that are generally legal, second-generation policies tilt more toward trade restrictions or duties that involve complicated new administrative processes that themselves create new barriers to trade. These restrictions are thus both less likely to generate the political dynamics that might discipline their use and also less likely to be consistent with international trade rules.

Regulations that restrict trade are among the least transparent forms of trade restrictions.Footnote 70 Because they do not operate directly on prices, their effects on prices can be difficult to observe, especially for uninformed consumers. Moreover, as compared to subsidies, the observable costs of outright trade restrictions are externalized on to foreign producers. An import restriction, for instance, obviously denies market access to foreign producers. In so doing, it creates a benefit for import-competing domestic producers. These costs and benefits are easily observable, while the costs to domestic consumers can be diffuse and thus may not prompt a reaction from consumers. The result is that trade restrictions may not engage the same kind of self-disciplining political dynamics that come with subsidies, where the cost to the enacting jurisdiction (in democracies, at least) is transparent because it shows up as government spending. In the absence of the self-discipline that politics can bring, legal conflict is more likely.Footnote 71

Imposing new duties at the border also creates new kinds of non-transparent trade restrictions. Take the EU CBAM as an example. Much of the literature on the CBAM has focused on whether it complies with the WTO's primary rules and whether it can be justified under GATT article XX.Footnote 72 But the EU CBAM specifically, and CBAMs generally, create significant trade barriers in the form of administrative hurdles for importers. If importers do not wish to simply pay rates based on default carbon-intensity values, they will have to demonstrate that their production processes are less carbon-intensive than default values – an administrative burden.Footnote 73 Moreover, if they wish to claim credit for carbon prices paid in their home markets, as the EU CBAM allows, they will have to demonstrate that they have actually made such payments.Footnote 74 Finally, as CBAMs spread to other countries, exporters operating in multiple markets will have to master multiple new sets of import requirements. While agreements that streamline and harmonize CBAMs are possible, in the short run it seems unlikely that governments will use the same kinds of paperwork, or even that they will necessarily calculate carbon emissions in the same way. In short, the administrative hurdles associated with trying to calculate duties in a technically accurate way create new trade barriers, which are not transparent and thus may not generate the same kinds of political pressure to correct.

Legally, the specific policy tools associated with second-generation flanking policies are also on a less firm footing. Unlike subsidies or trade remedies, which are in general permitted under the WTO's primary rules, import bans (such as contained in the Uyghur Forced Labor Prevention Act) and duties above one's tariff bindings (such as carbon tariffs) violate either GATT article XI or GATT article II. Moreover, WTO case law has interpreted the non-discrimination rules in Articles I and III in a manner that focuses on whether a measure modifies the equality of competitive opportunities, and without regard to regulatory purpose.Footnote 75 As discussed in further detail in Section 3, under these interpretations, many second-generation flanking policies likely violate non-discrimination rules.

The result is that the legality of second-generation flanking policies likely hinges on whether they can be justified under exceptions to WTO rules. But the reliance on exceptions to justify the pursuit of entirely unexceptional policy goals has come under fire.Footnote 76 Perhaps most importantly, WTO panels and the Appellate Body virtually never simply uphold the application of the WTO's general exceptions. Instead, dispute reports have tended to recognize that challenged measures fall within the scope of the exceptions but nevertheless identify discriminatory aspects of the measure that the respondent must amend to bring it into compliance.Footnote 77 For members where trade policy is formulated by an executive-like entity, this system works reasonably well. The European Commission, for example, can often make changes to its programs. For members that must look to their legislatures to amend challenged laws, a situation the United States often finds itself in, this process of revision in response to dispute reports is unwieldy. The unwillingness to uphold measures that fall within the scope of the exceptions without further amendment may render the exceptions to justify second-generation flanking policies practically unavailable.

4.2 Consumption as a Basis of Extraterritorial Authority over Production in International Economic Law

The second source of conflict is related to the first but is more fundamental. Second-generation flanking policies implicitly rest on novel claims about the authority of nations to prescribe rules of conduct. These novel assertions of authority lead to a fundamentally different understanding of how trade law should be interpreted. The developed countries that have adopted second-generation flanking policies most aggressively, however, have not yet fully articulated how the claims of authority on which they rest should lead to a reinterpretation of trade law. This section outlines both the conflict between the old and new paradigm, as well as how WTO members adopting second-generation flanking policies that rest on the new paradigm might more forcefully articulate the legal basis for their authority in established doctrinal categories.

4.2.1 Changing Norms Regarding Authority to Tax and Regulate

Traditionally, jurisdiction to tax and regulate production in international trade law rests with the country in which production occurs.Footnote 78 Although other countries are free to tax and regulate products and services within their territory, those taxes and regulations were not supposed to be conditioned on aspects of the production process that occurred within the territory of another state. In other words, international trade law historically limited nations' otherwise plenary authority to tax and regulate goods within their own territory by excluding foreign production from the permissible bases of taxation or regulation. From the 1990s forward, the debate over so-called process and production methods (PPMs) sought to clarify the exact contours of this rule, with some scholars in particular advocating for a more permissive understanding of the rule.Footnote 79

The increasingly widespread adoption of second-generation flanking policies, especially by developed nations, marks a shift away from this production theory of jurisdiction in international trade law – the rule that nations must have a territorial nexus with production to impose taxes, regulations, or market access restrictions that are conditioned on the manner in which a good or service is produced. Instead, second-generation flanking policies rest on an implicit theory of consumption jurisdiction, under which any nation that consumes a product or service may condition taxes and regulations applied within its own territory on the manner of production that occurs extraterritorially. In effect, a territorial nexus with consumption has become, at least in the eyes of developed nations, a sufficient basis to tax or regulate production.

A bit of background on jurisdictional principles and their relationship to international trade law is useful to unpack this claim. Under general international law, a number of bases exist for a nation's jurisdiction to prescribe.Footnote 80 The chief among these is territorial jurisdiction, the notion that a nation has plenary authority to prescribe rules governing activities that occur within its territory and, absent another basis for jurisdiction, lacks the authority to prescribe rules governing conduct in other nations' territories. A little over a century ago, the US Supreme Court laid down the strictest version of this rule, stating that ‘the general and almost universal rule is that the character of an act as lawful or unlawful must be determined wholly by the law of the country where the act is done’.Footnote 81

Since then, however, the notion that territorial jurisdiction creates an exclusive right to prescribe rules has been under stress.Footnote 82 Although other bases for jurisdiction can limit the exclusivity associated with territorial jurisdiction, the most significant new limitation on exclusive jurisdiction effects jurisdiction. The Fourth Restatement of Foreign Relations Law provides that ‘[i]nternational law recognizes a state's jurisdiction to prescribe law with respect to conduct that has a substantial effect within its territory’.Footnote 83 In essence, a territorial nexus with the effects of an act can provide a basis for prescribing rules governing the conduct that causes the effect, even if that conduct is extraterritorial.Footnote 84

Second-generation flanking policies, I argue, reflect a further extension of territorial jurisdiction beyond only effects. At their extreme, these policies rest on the implicit assertion that a nation can prescribe rules for foreign conduct (production) even in the absence of sufficient effects caused by those activities. In asserting the right to impose duties on products based on the amount of carbon emitted during production, or to prohibit the import of products made with labor standards the importing nation finds objectionable, nations are implicitly claiming – and sometimes explicitly claiming, as in the case of the EU Deforestation-free Products Regulation discussed above – that the mere fact of consumption provides a sufficient territorial nexus to permit them to prescribe rules for the production of goods entering their territory.

Put differently, the demand for flanking policies created by trade liberalization is spurring changes in the allocation of authority in international trade law (and international economic law more broadly).Footnote 85 In order to adopt policies that address non-economic, extraterritorial costs of trade liberalization, countries have to rely on newly expansive theories of state authority. Those theories are disruptive for existing interpretations of legal rules, leading to disputes and misunderstandings.

What has started to emerge are two distinct visions for how international trade law should be interpreted. One, advocated by developing countries such as India, Indonesia, and Malaysia, relies on existing and well-established interpretations of GATT/WTO law. As a specialized area of the law, international trade law has not used either the traditional territorial/effects language of general international law, or the functional language of production and consumption that I urge. But this view is very much grounded in a theory of state authority that holds that only states with a territorial nexus to production may impose taxes and regulations based on production.

Doctrinally, this traditional production-based view has three main components. The first is an interpretation of the esoteric concept of ‘border adjustability’.Footnote 86 Over the course of the twentieth century, the distinction between taxes and regulations on production (not border adjustable) and those on the product (border adjustable) gave trade what amounted to a jurisdictional theory under which the nation in which production occurred had exclusive jurisdiction to tax and to regulate production.Footnote 87 The traditional understanding of border adjustability is thus the trade law doctrine that embodies the production theory of jurisdiction.

The doctrine works as follow. GATT/WTO law imposes severe restrictions on border measures. The GATT generally prohibits regulations that restrict imports at the border.Footnote 88 The GATT permits tariffs (taxes on imports) so long as the tariff charged is ‘no less favourable than that provided for in’ a nation's individual schedule of tariff limits.Footnote 89 In practice, repeated rounds of negotiations on reducing tariffs over the last 75 years have resulted in significant limits on tariffs, especially for developed countries. As a result, measures that are on production and hence are not border adjustable likely violate WTO rules.

In contrast, measures that tax or regulate the product as such are deemed border adjustable and hence are treated as domestic measures. Whether framed as taxes or regulations, domestic measures are subject to the principle of national treatment.Footnote 90 Although the precise formulation of the tests differ, the general rule is that imports must be treated no less favorably than ‘like’ domestic products.Footnote 91 In other words, domestically GATT/WTO rules permit nations to tax and regulate products however they see fit, so long as they do so in a non-discriminatory fashion.Footnote 92

Much thus hinges on whether a measure is classified as a border measure or a domestic measure. For instance, a country might have a regulation that bans a certain type of product, such as tuna caught in a manner deemed risky for dolphins, from being imported or produced domestically.Footnote 93 If the ban on imports is treated as a border measure, it violates GATT rules even if there is a similar ban in place for domestic production. If it is treated as part of a domestic prohibition on the sale of dolphin-unsafe tuna, it may be GATT-consistent because it applies to both domestic and imported products.

The second component of the traditional view is the manner in which non-discrimination rules are interpreted. To say that a measure is lawful so long as it is non-discriminatory sounds benign in theory. In practice, as noted above, the traditional view has adopted a restrictive understanding of non-discrimination that focuses on the economic relationship between two products (the ‘like products’ test) and how the measure impacts the equality of competitive opportunities (the ‘treatment no less favourable’ test).Footnote 94 As they have been interpreted, these components have minimized, if not eliminated, the ability of a WTO member from justifying a difference in treatment between two products on the grounds that the products are different in light of the government's regulatory purpose.Footnote 95 The result is that under the traditional and still prevailing view, de facto discrimination is often not that difficult to demonstrate when products that are substitutes in the marketplace are treated differently.

Doctrinally, the place to explicitly consider regulatory purpose under the traditional view is the exceptions – the third component of the traditional, production-based view. From the time the WTO was created, dispute panels became more open finding challenged measures ‘provisionally justified’ under the exceptions, meaning that measure was related to or necessary to fulfill a permissible purpose.Footnote 96 However, as discussed in section 4.1.2 above, panels have generally rejected an exception's application, even after finding a measure related or necessary to fulfill a legitimate objective, unless any discriminatory aspects of the measure can be justified in light of that objective.Footnote 97 In other words, panels have imposed a high standard of purity of intent and application in order to find an exception applicable.

Many second-generation flanking policies likely fail under one or more of the components of this traditional view. As discussed above, the DSB recently adopted a decision concluding that the deforestation aspects of the EU's Renewable Energy Directive are unjustifiably discriminatory. Although no challenge has yet been brought to CBAM, it is also likely discriminatory in ways that are unjustifiable under traditional interpretations of GATT article XX.Footnote 98 The Uyghur Forced Labor Act, which imposes different procedures for assessing whether products from Xinjian province in China are made with forced labor than the procedures used for goods from other parts of the world, could be viewed by some as an unlawful trade embargo in violation of GATT article XI (because it regulates production and hence is not border adjustable), and unjustifiably discriminatory in singling out China's treatment of the Uyghur population.

4.2.2 Articulating Trade Law Doctrines Based on Consumption Jurisdiction

The result is that second-generation flanking policies are likely to create difficult conflicts over international trade rules. This difficulty is made worse by the fact that WTO members embracing second-generation flanking policies have not articulated clear interpretations of WTO rules that take the implicit consumption-based theory of state authority, on which second-generation policies rest, into account. I offer here a brief outline of what such a doctrinal reformation might look like.

First, a better approach is to bring the theory of border adjustability in line with more general theories of the jurisdiction to prescribe in international economic law. In particular, recognizing that nations have an interest in the way goods they consume are produced should lead to the conclusion that taxes and regulations on extraterritorial production are border adjustable and legal so long as they are non-discriminatory. Such a theory better harmonizes the rules with recent state practice by, for example, the United States and the EU. It also supports an open global trading system by allowing countries to assert their interest in ameliorating the extraterritorial costs of trade liberalization while also maintaining their good standing in a rules-based trading system. To do otherwise forces states to an either/or choice: either seek to mitigate the costs of trade liberalization or maintain one's role in the rules-based trading system. Recent events suggest that the rules-based system will not come out the better for putting nations to that choice.

Second, non-discrimination rules themselves need to be interpreted in light of a nation's legitimate regulatory purposes. Failing to take regulatory purpose into account when assessing discrimination forces respondent nations to use exceptions to justify ordinary government actions. Indeed, norm-exception frameworks have a number of drawbacks as applied to unexceptional policies such as second-generation flanking policies. For one, these policies are not likely to be temporary. Nor are they narrow in scope. Early WTO disputes about the scope of exceptions, such as Shrimp–Turtle, dealt with discrete products. By contrast, many second-generation flanking policies deal with huge sectors of the economy (e.g., CBAM) or even apply economy-wide (labor rules). Assessing the permissibility of these broad measures through the application of exceptions puts treaty interpreters in a bind. Either they interpret the exception capaciously enough to incorporate the broad policy, which risks letting the exception swallow the rule, or they reject the exception's application, leaving second-generation flanking policies impermissible under WTO rules. Finally, the need to justify environmental or labor policies as exceptions puts those policies in a subordinate position to rules on trade liberalization, making them clear second-class citizens within international trade law.Footnote 99

If WTO rules were being explicitly renegotiated now, we might expect second-generation flanking policies to receive the same kind of legal grace that the tools associated with first-generation flanking policies – subsidies and trade remedies – received. Indeed, in negotiations outside the WTO, we see this happening. The negotiations over the Rapid Response Mechanism in the USMCA, efforts to negotiate a Global Arrangement on Sustainable Steel and Aluminum, and the Indo-Pacific Economic Framework have all offered opportunities to think about how international rules might be designed to create space for second-generation flanking policies.

But second-generation flanking policies remain subject to WTO rules, meaning that there is only so far nations can go in rewriting the rules while preserving the WTO. WTO parties and future dispute panels thus should embrace regulatory purpose as relevant in assessing the existence of discrimination. Although the Appellate Body rejected such a role early on,Footnote 100 second-generation flanking policies and the political and legal dynamics that produce them and conflict over them suggest that the time for reconsideration is nigh. Although in principle regulatory purpose could be relevant to either a “like” products analysis or a less favorable treatment analysis,Footnote 101 the former might be a more natural fit. High-carbon steel, for instance, should not be deemed “like” low-carbon steel for purposes of administering carbon taxes, even if they are “like” for purposes of other kinds of measures, such as sales taxes.

Finally, WTO members should call for, and panels should embrace, a greater margin of appreciation in assessing whether discrimination can be justified under GATT article XX and its chapeau. The reality that domestic politics produce mixed motive policies, combined with the difficulty that legislatively led governments will have changing details of a policy, counsel strongly in favor of a ‘predominant motive’ test, under which discrimination can be justified so long as a panel concludes that the primary motive for the measure is a legitimate one.Footnote 102

Although reforming the application of WTO exceptions, and especially the chapeau of article XX, is necessary, it is in an important sense also the least important of these three reforms. The governments that are adopting second-generation flanking policies are not adopting them as exceptions to general policies, nor are such policies small in scope. Second-generation flanking policies aim to rebalance global trade rules by conditioning market access on extraterritorial, non-economic costs associated with production. WTO members would be well served to adopt interpretations of legal rules that fit the scope of that ambition.

5. Conclusion

Trade liberalization has been one of the most significant and positive international policy developments in the last century. But it has not been an unalloyed good. In recent decades, policymakers have become aware that trade liberalization has non-economic, as well as economic, costs. Unfortunately, though, the very different dynamics of addressing non-economic, extraterritorial harms in an already liberalized trading system create additional stresses on the trading system above and beyond the externalities that second-generation flanking policies are designed to address. As nations' appetites for addressing these externalities seem to be increasing, the trading system must adjust to the challenges these new flanking policies pose.

References

1 J.F. Kennedy (1962) ‘Address before the Conference on Trade Policy’, 17 May. Online by Gerhard Peters and John T. Woolley, The American Presidency Project, https://www.presidency.ucsb.edu/documents/address-before-the-conference-trade-policy.

2 Pauwelyn, J. and Sieber-Gasser, C. (2024) ‘Addressing Negative Effects of Trade Liberalization: Unilateral and Mutually Agreed Flanking Policies’, World Trade Review, this issueGoogle Scholar.

3 Laurens, N., Winkler, C. and Dupont, C. (2024) ‘Sweetening the Liberalization Pill: Flanking Measures to Free Trade Agreements’, Review of International Political Economy 1-18Google Scholar.

4 E.g., Shaffer, G. (2019) ‘Retooling Trade Agreements for Social Inclusion’, University of Illinois Law Review 1Google Scholar; Roberts, A. and Lamp, N. (2021) Six Faces of Globalization. Harvard University PressGoogle Scholar.

5 Irwin, D. et al. (2008) The Genesis of the GATT. Cambridge University Press, 197200CrossRefGoogle Scholar.

6 International Monetary Fund, World Bank, and World Trade Organization (2017) ‘Making Trade an Engine of Growth for All: The Case for Trade and For Policies to Facilitate Adjustment’; Shaffer, supra n. 4, 3.

7 E.g., Kennedy, supra n. 1 (‘For most affected firms will find that the adjustment involved is no more than the adjustment they face every year or few years as the result of changes in the economy, consumer taste or domestic competition.’); B.S. Bernanke, Chairman, Fed. Reserve (2007) ‘Speech to the Greater Omaha Chamber of Commerce: The Level and Distribution of Economic Well-Being’, 6 February 2007 (‘Because labor markets are adaptable, outsourcing abroad does not ultimately affect aggregate employment, but it may affect the distribution of wages.’).

8 Kennedy, supra n. 1.

9 Pauwelyn and Sieber-Gasser, supra n. 2.

10 T. Meyer (2020) ‘Misaligned Lawmaking’, Vanderbilt Law Review 73, 151, 189–191.

11 Trade remedies are an exception, which I discuss further below. However, they were a feature of domestic law prior to being uploaded to trade agreements.

12 C. Michael Aho and T.O. Bayard (1984) ‘Costs and Benefits of Trade Adjustment Assistance’, The Structure and Evolution of Recent US Trade Policy. University of Chicago Press, 153.

13 S.E. Kim and K.J. Pelc (2021) ‘The Politics of Trade Adjustment versus Trade Protection’, Comparative Political Studies 54, 2354.

14 E. Kapstein (1998) ‘Trade Liberalization and the Politics of Trade Adjustment Assistance’, International Labour Review 137, 501, 509.

15 Regulation (EC) No. 1927/2006, as amended.

16 L. Puccio (2017) ‘Policy Measures to Respond to Trade Adjustment Costs’, European Parliamentary Research Service, 7; Roberts and Lamp, supra n. 4.

17 D. Lysenko et al. (2017) ‘Does Canada Need Trade Adjustment Assistance?’, International Journal 72, 91, 94–97.

18 Ibid, 97.

19 W. Cutler and J. Bell (2018) ‘Adjusting to Trade: Asia-Pacific Approaches to Assisting Displaced Workers’ (Asia-Pacific Policy Institute 30 March), 4.

20 Ibid, 5.

21 Ibid.

22 Pauwelyn and Sieber-Gasser, supra n. 2.

23 D.R. Cameron (1978) ‘The Expansion of the Public Economy: A Comparative Analysis’, American Political Science Review 72, 1243; D. Rodrik (1998) ‘Why Do More Open Economies Have Bigger Governments?’, Journal of Political Economy 106, 997.

24 Cf. Kim and Pelc, supra n. 13.

25 G. Shaffer (2024) ‘Package Treaties: Addressing the Negative Effects of Trade’, World Trade Review, this issue.

26 For the general definition of a flanking policy, see Pauwelyn and Sieber-Gasser, supra n. 2.

27 ‘General Agreement on Tariffs and Trade’ (first published 1947), vol. 55 U.N.T.S., art. VI.6.

28 Kim and Pelc, supra n. 13.

29 North American Free Trade Agreement, Can.–Mex.–US, 17 December 1992, 32 I.L.M. 289 (1993).

30 North American Agreement on Environmental Cooperation, opened for signature 8 September 1993, 32 I.L.M. 1480; North American Agreement on Labor Cooperation, 14 September 1993, 32 I.L.M. 1499.

31 E.g., O.E. Herrnstadt (2018) ‘Renegotiating NAFTA is an Opportunity to Get Trade Policy Right’, Economic Policy Institute, 10 January, www.epi.org/blog/renegotiating-nafta-is-an-opportunity-to-get-trade-policy-right/.

32 M. Wallerstein and B. Western (2000) ‘Unions in Decline? What Has Changed and Why’, Annual Review of Political Science 3, 355.

33 T. Meyer and G. Sitaraman (2019) ‘Trade and the Separation of Powers’, California Law Review 107, 583 (describing the ‘foreign affairs paradigm’ that prevailed in US trade policy during the latter half of the twentieth century).

34 Kennedy, supra n. 1.

35 Meyer and Sitaraman, supra n. 32, 612–617.

36 Compare Panel Report, United States – Restrictions on Imports of Tuna, DS21/R-39S/155 (3 September 1991)(unadopted) [hereinafter ‘United States–Tuna’] with Appellate Body Report, United States – Import Prohibition of Certain Shrimp and Shrimp Products, WT/DS58/AB/R (adopted 6 November 1998)[hereinafter United States–Shrimp]; Appellate Body Report, European. Communities Measures Prohibiting the Importation and Marketing of Seal Products, WT/DS400/AB/R, WT/DS401/AB/R (adopted 18 June 2014)[hereinafter ‘EC–Seals’].

37 Shaffer, supra n. 4, 3.

38 S. Barrett (2005) Environment and Statecraft: The Strategy of Environmental Treaty-Making. Oxford University Press, 307–334.

39 To combat these kinds of effects, Shaffer has called for social dumping duties. Shaffer, supra n. 4, 33.

40 K. Appunn and J. Wettengel (2023) ‘Understanding the European Union's Emissions Trading Systems (EU ETS)’, Clean Energy Wire, 26 January.

41 This is an oversimplification, as determining this amount itself involves the use of default values subject to appeal.

42 The EU also reduces the amount owed or completely exempts the producer if the producer has paid a carbon tax at home or is from a non-EU country that has either joined the ETS or linked a carbon trading system to the ETS.

43 Press conference by Executive Vice-President Frans Timmermans and Commissioner Sinkevičius on a package of proposals on soil, waste and deforestation (17 November 2021).

44 European Commission (2021) Questions and Answers on New Rules for Deforestation-Free Products (17 November 2021), https://ec.europa.eu/commission/presscorner/detail/en/qanda_21_5919.

45 C. Fischer and T. Meyer (2020) ‘Baptists and Bootleggers in the Biodiesel Trade: EU-Biodiesel (Indonesia)’, World Trade Review 19, 297, 309–313.

46 A WTO panel in case brought by Malaysia recently concluded that this distinction was discriminatory. Panel Report, European Union – Certain Measures Concerning Palm Oil and Oil Palm Crop-based Biofuels, WT/DS600/R (adopted 26 April 2024).

47 D. LeClerq (2021) ‘The Disparate Treatment of Rights in US Trade’, Fordham Law Review 90, 1, 23–26 (discussing the history of ILO rights and efforts to incorporate them in US trade law).

48 Laurens et al., supra n. 3, provide an insightful analysis of the political motivations underlying flanking policies.

49 T. Meyer (2022) ‘The Political Economy of WTO Exceptions’, Washington University Law Review 99, 1299.

50 Cf. B. Yandle (1983) ‘Baptists and Bootleggers: The Education of a Regulatory Economist’, Regulation 7, 12.

51 Meyer, ‘The Political Economy of WTO Exceptions’, supra n. 47.

52 United States–Shrimp, supra n. 35.

53 Meyer, ‘The Political Economy of WTO Exceptions’, supra n. 49.

54 For a discussion of how communities of legal practice can shape interpretation and meaning, see H.G. Cohen (2015) ‘International Precedent and the Practice of International Law’, in M.A. Helfand, ed., Negotiating State and Non-State Law: The Challenge of Global and Local Legal Pluralism. Cambridge University Press.

55 See Panel Report, European Union-Anti-Dumping Measures on Biodiesel from Indonesia, WTO Doc. WT/DS480/R (adopted 28 February 2018) [hereinafter EU–Biodiesel (Indonesia); Appellate Body Report, European Union-Anti-Dumping Measures on Biodiesel from Argentina, WTO Doc. WT/DS473/AB/R (adopted 26 October 2016); Fischer and Meyer, supra n. 45, 309–313.

56 A dispute panel did, however, find that the measure was unjustifiably discriminatory. European Union – Certain Measures Concerning Palm Oil and Oil Palm Crop-based Biofuels, supra n. 45.

57 H.V. Milner and D. Tingley (2013) ‘Public Opinion and Foreign Aid: A Review Essay’, International Interactions 39, 389, 392 (‘relative to domestic programs foreign aid is unpopular.’).

58 Of course, countries have different attitudes on foreign aid, just as they have different attitudes on social safety net programs. The point here is a relative one – – a given country is likely to find it easier to give aid to its own citizens than to the citizens of other nations.

59 A. Soto and K. Hughes, ‘US to Pay $300 Million to End Brazil Cotton Trade Dispute’, Reuters (30 September 2014).

60 The EU General Court recently confirmed the European Commission's view that such subsidies are in principle countervailable. Case T-540/20, Jushi Egypt for Fiberglass Industry v Commission (1 March 2023).

61 Nikkei staff writers, ‘Japan Reveals 87 Projects Eligible for “China Exit” Subsidies’, Nikkei Asia (17 July 2020).

62 A. Dunkel and F. Roessler (2021) ‘The Ranking of Trade Policy Instruments Under the GATT Legal System’, in A.T. Guzman and J. Pauwelyn (eds.), International Trade Law, Second Edition, Aspen Casebook Series. Aspen.

63 H.P Hestermeyer and L. Nielsen (2014) ‘The Legality of Local Content Measures Under WTO Law’, Journal of World Trade 48, 553.

64 G. Shaffer et al. (2015) ‘Can Informal Law Discipline Subsidies?’, Journal of International Economy Law 18, 711, 718.

65 See GATT art. VI & XIX.

66 D.K Tarullo (2002) ‘The Hidden Costs of International Dispute Settlement: WTO Review of Domestic Anti-Dumping Decisions’, Law and Policy in International Business 34, 109.

67 B.A. Blonigen and C.P. Bown (2003) ‘Antidumping and Retaliation Threats’, Journal of International Economics 60, 249.

68 J. Mason (2022) ‘Biden to Waive Tariffs for 24 Months on Solar Panels Hit by Probe’, Reuters (5 June 2022).

69 E.g., Solar Energy Industries Association, Solar Energy Industries Association, ‘The Adverse Impact of Section 201 Tariffs: Lost Jobs, Lost Deployment and Lost Investments’ (December 2019).

70 This discussion draws on Dunkel and Roessler, supra n. 62, 223–224.

71 From a welfare perspective, this dynamic may actually be beneficial where the trade restrictions in question promote global public goods, such as in the case of climate change. Externalizing costs to foreign actors while providing benefits to domestic import-competing producers can mobilize political coalitions to supply policies for global public goods that otherwise might go unprovided. See T. Meyer (2015) ‘How Local Discrimination Can Promote Global Public Goods’, Boston University Law Review 95, 1937.

72 See, e.g., M. Mehling, H. van Asselt, S. Droege, and K. Das (2022) ‘Symposium on Carbon Border Adjustments,’ American Journal of International Law Unbound 116, 191.

73 Regulation (EU) 2023/956 of the European Parliament and of the Council of 10 May 2023 establishing a carbon border adjustment mechanism art. 7.2.

74 Russia, for example, suggested that it might impose a carbon tax on its producers in order to allow them to claim credit under the EU CBAM. Presumably, they might not actually pay the tax in Russia. To prevent these kinds of evasion tactics, greater administrative care will be required in ensuring that payments to foreign governments have actually been made.

75 E.g., EC–Seals, supra n. 35, para. 5.101.

76 J. Arato, K. Claussen, and J. Benton Heath (2020) ‘The Perils of Pandemic Exceptionalism’, American Journal of International Law 114, 627.

77 E.g., United States–Shrimp, supra n. 35); Appellate Body Report, Brazil – Measures Affecting Imports of Retreaded Tyres, WTO Doc. WT/DS332/AB/R (adopted Dec. 17, 2007); EC–Seals, supra n. 35); European Union – Certain Measures Concerning Palm Oil and Oil Palm Crop-based Biofuels, supra n. 45.

78 This also true in other areas of international economic law, such as international tax and competition law. I explore the broader trend toward consumption jurisdiction in T. Meyer (forthcoming 2024) ‘Consumption Governance: The Role of Production and Consumption in International Economic Law’, Brigham Young University Law Review 49.

79 D. A. Kysar (2004) ‘Preferences for Processes: The Process/Product Distinction and the Regulation of Consumer Choice’, Harvard Law Review 118, 525, 631; S. Charnovtiz (2002) ‘The Law of Environmental “PPMs” in the WTO: Debunking the Myth of Illegality’, Yale Journal of International Law 27, 59.

80 Restatement (Fourth) of Foreign Relations Law § 407–13.

81 Am. Banana Co. v. United Fruit Co., 213 US 347, 356 (1909).

82 N. Krisch (2022) ‘(Extra)Territorial Regulation as Global Governance’, European Journal of International Law 33, 481.

83 Restatement (Fourth) of Foreign Relations Law § 409.

84 As a matter of doctrine, commentators and policymakers do not agree on whether effects jurisdiction is an expansion of the territorial theory or an independent, and possibly extraterritorial, basis for asserting jurisdiction. Courts, for instance, have often treated it as an extension of territorial jurisdiction. See, e.g., United States v. Neil, 312 F.3d 419, 422 (9th Cir. 2002)(‘Under the territorial principle, the United States may assert jurisdiction when acts performed outside of its borders have detrimental effects within the United States.’); Restatement (Second) of Foreign Relations Law, § 18; Restatement (Third) of Foreign Relations Law § 402. Commentators, however, have often treated it as a distinct, extraterritorial basis for jurisdiction. E.M. Fox (2019) ‘Extraterritorial Jurisdiction, Antitrust, and the EU Intel Case: Implementation, Qualified Effects, and the Third Kind’, Fordham International Law Journal 42, 981. The Fourth Restatement takes the view that effects jurisdiction is separate, but still derived, from territorial jurisdiction, ‘reflect[ing] the evolution of the effects principle into a distinct basis for jurisdiction to prescribe under customary international law’. Restatement (Fourth) of Foreign Relations Law § 409 Rptr's n. 5.

85 See Meyer, ‘Consumption Governance: The Role of Production and Consumption in International Economic Law’, supra n. 78.

86 Ibid; C. Adkins and D.S. Grewal (2016) ‘Two Views of International Trade in the Constitutional Order’, Texas Law Review 94, 1495.

87 Meyer, ‘Consumption Governance: The Role of Production and Consumption in International Economic Law’, supra n. 78.

88 GATT art. XI.1 (‘No prohibitions or restrictions other than duties, taxes or other charges, whether made effective through quotas, import or export licences or other measures, shall be instituted or maintained by any contracting party on the importation of any product … ’).

89 GATT art. II.1(a).

90 GATT art. III.

91 GATT art. III.2 (taxes) & III.4 (regulations); see also AB Report – European CommunitiesMeasures Affecting Asbestos and Asbestos-Containing Products, ¶ 99 WT/DS135/AB/R (adopted 5 April 2001) (establishing that the scope of ‘like’ products under art. III.4, governing domestic regulations, is similar to the scope of ‘like’ products under Art. III.2 governing domestic taxes).

92 P.C. Mavroidis (2005) The General Agreement on Tariffs and Trade: A Commentary. Oxford University Press.

93 United States–Tuna, supra n. 35.

94 E.g., EC–Seals, supra n. 35, para 5.101.

95 A.D. Mitchell, D. Heaton, and C. Henckels (2016) Non-Discrimination and the Role of Regulatory Purpose in International Trade and Investment Law. Edward Elgar, 2 (‘A pressing issue facing both international trade law and international investment law is whether, if at all – and, if so, how – the so-called “regulatory purpose” of a measure is relevant to the question whether the measure infringes national treatment or MFN treatment obligations.’).

96 See cases cited, supra n. 35.

97 See cases cited, supra n. 75.

98 Espa, I. (2022) ‘Reconciling the Climate/Industrial Interplay of CBAMs: What Role for the WTO?’, American Journal of International Law Unbound 116Google Scholar, 208, 212 (‘Origin-based discrimination is also virtually certain since those countries that are either integrated or linked to the EU ETS are exempted from the CBAM in violation of the most-favored nation clause (Article I of GATT’); Vidigal, G. and Venzke, I. (2022) ‘Of False Conflicts and Real Challenges: Trade Agreements, Climate Clubs, and Border Adjustments’, American Journal of International Law Unbound 116Google Scholar, 202, 205 (‘One challenge with the EU's climate club exception is that it requires other countries to adopt systems with fundamentally the same design as the EU's.’).

99 Ibid.

100 Appellate Body Report, Japan–Taxes on Alcoholic Beverages, WT/DS8/AB/R (4 October 1996) 18–23; Hudec, R.E. (1998) ‘GATT/WTO Constraints on National Regulation: Requiem for an Aim and Effects Test’, International Lawyer 32, 619, 626–33Google Scholar.

101 Mitchell, Heaton, and Henckels, supra n. 93 provide a nuanced analysis of how it is relevant to both.

102 Meyer, ‘The Political Economy of WTO Exceptions’, supra n. 49, 1360.