2. What the GATT Understood and the WTO Undervalued

The GATT began as a club of leaders in democratic or quasi-democratic nations who shared the desire to bring down substantial, and often discriminatory, trade barriers that had emerged during the 1930s.Footnote ² The membership was mindful of domestic constraints and wary of too much international delegation. Indicative of these concerns, decisions in practice were made by consensus, and obligations were qualified by numerous exceptions for measures that pursued legitimate domestic objectives. Politically unfortunate commitments could be withdrawn temporarily (through safeguard measures) or permanently (through Article XXVIII renegotiations), and countermeasures could be taken against ‘unfair’ trade practices by counterparties. The flexibility of the regime was viewed positively by current and potential members and facilitated GATT expansion.

Looking back, cooperation was supported as much by a set of shared norms and experiences as it was by explicit rules. Participants agreed that beggar-thy-neighbor policies had contributed to the Great Depression and ultimately to World War II. The founders also believed that the balance between market independence and government intervention had moved in the wrong direction in the interwar years, and, as a result, trade had subsided as nations had pursued nationalist goals. To avoid another economic and political disaster, they advocated for a stable trading system that would encourage and support international commerce. The method chosen to reach the goal of ‘peace through trade’ was to facilitate reciprocal trade deals based on the MFN principle, so that all members of the GATT would receive benefits and have a stake in the trading regime.Footnote ³

The founding members understood that they needed to guard against ex post opportunism, but they also recognized that the organization could not force members to act against their overall national interest. As such, limited power was delegated and the main role of the secretariat was to collect information on trade activities. GATT Article XXIII allowed members to seek redress for actions by other members that undermined the GATT bargain, and evolved into a process whereby aggrieved members could seek an arbitral panel if consultations did not resolve a dispute.Footnote ⁴ The parties to a dispute, however, had to agree to establish the panel for most of GATT's history, and the membership had to agree by consensus to ‘adopt’ the report of a panel to give it legal force and to authorize suspension of concessions against a recalcitrant violator.

Under this system, roughly three-quarters of the dispute reports were accepted (‘adopted’) by the membership and suspension of concessions was authorized only once in an early case.Footnote ⁵ The GATT held together for almost half a century, not because members were compelled to follow the rules but because fidelity to most of the rules was in their mutual interest – the members believed ongoing cooperation to be valuable, notwithstanding occasional unresolved disputes.

To be sure, the system was ‘power-based’ to a considerable degree. ‘Enforcement’ in thorny cases was through self-help, which inevitably advantaged powerful members. No country could be definitively adjudicated to be in violation of the law without its consent, and some disputes were left unsettled. However, despite its imperfections, the system functioned reasonably well, dramatically reducing average tariff rates over time and reducing some non-tariff barriers as well.

3. The Innovations of the WTO and Their Consequences

The motivation for moving on from the GATT is well documented. An increasing number of intransigent disputesFootnote ⁶ led the United States and eventually others to employ unilateral sanctions against measures that they unilaterally adjudicated to be GATT violations. Disputes over matters outside of the GATT – notably services and intellectual property (IP) – further contributed to this trend toward unilateralismFootnote ⁷ and led to a constituency for new rules in those areas. Increasing numbers of plurilateral agreements on matters going beyond the original GATT created a patchwork of obligations that bound some members but not others.Footnote ⁸

At the same time, following the breakup of the Soviet Union, those who designed the WTO believed that a new era of market liberalization had begun. Capitalism had triumphed, and, as a result, the WTO could and should expand its reach. This burst of enthusiasm for free markets helped to spur support for a broader scope of market-centered trade rules. What is now known as the ‘Washington Consensus’ formalized some of this shared world view in a set of what were seen as ‘optimal’ rules pertaining to the government-market relationship.

Crafting a new organization was not easy. The GATT membership had expanded greatly over time, and gaining consensus had become increasingly difficult. The major players seized on the idea of a ‘single undertaking’ that would solve the problems of GATT and bring all members on board. New rules would be negotiated on services and IP, and formerly plurilateral codes would be further negotiated and refined. Even agriculture could be tackled in a new agreement. Pursuant to a new Dispute Settlement Understanding (DSU), members complaining of legal violations could secure adjudication as a matter of right, its results could no longer be ‘blocked’ by a losing disputant, and the loser could not prevent eventual authorization for sanctions following its refusal to comply with a ruling. Errors by dispute panels would be corrected by the AB before they became legally binding. To ensure buy-in from all members, the major players would withdraw from GATT (while preserving its legal provisions) and enter the WTO – all countries would be forced to go along lest they lose the benefits of GATT membership.Footnote ⁹

This ambitious move seemed to solve many of the perceived deficiencies of the GATT, but it did not change the de facto ‘voting rule’ for most decisions, which remained wedded to the norm of consensus – and given the growth of the membership, achieving such consensus was an enormous task. Moreover, the treaty text for the WTO was roughly seven times longer than that of the GATT, not counting the many decisions and understandings that have followed. The sheer volume of legal obligations, and the attendant potential for interpretive disputes, not unexpectedly exploded. Until the United States disabled the AB by blocking judicial appointments, members accused of violating an obligation and unwilling to settle could not avoid going through the full adjudicative process and could not avoid designation as a scofflaw by blocking the adoption of rulings. When compliance with a ruling was politically unacceptable, complainants could seek authority for retaliatory measures, and the violator would be harangued about its plans for eventual compliance at regular meetings of the membership.

Important aspects of the single undertaking were in tension with the central insight of the GATT's founders. The GATT allowed members to bend the rules to placate powerful domestic groups without endangering a general commitment to the regime. Many of its obligations were vague and open to interpretation, and it tolerated certain ‘extra-legal’ measures that deviated from the letter of the agreement (such as grey area measures in response to import surges) as a necessary compromise between market discipline and political necessity. GATT also enabled members to block formal dispute settlement when litigation was unlikely to facilitate a mutually acceptable outcome. The WTO approach curtailed much of this political wiggle room in favor of more precise and detailed rules and a guarantee that disputes could proceed to a definitive ruling. In retrospect, however, the perceived weaknesses of GATT may have been more of a virtue than a vice.

China's entrance into the WTO in 2001 revealed another problem in the trading regime. Existing members, mostly democracies, worried about organized interest groups. Their internal political dynamic focused on building domestic coalitions to pursue the collective and shared interest of opening foreign markets. China may have been expected to evolve toward Western-style market orientation but in fact it has not, with its ongoing emphasis on central planning and industrial policy aimed at dominating key industries with little regard for market discipline or global consequences. WTO rules were not designed to address the scope and magnitude of the attendant political externalities, highlighting the need for new rules.

4. Institutional Rigidity and the Declining Support for Liberal Trade

The economic and political implications of the ‘China Shock’ have been thoroughly investigated.Footnote ¹⁰ Its effects in the United States were concentrated in manufacturing industries, many of which were geographically concentrated in electoral swing states. The United States was not alone in suffering economic dislocation from surging Chinese imports, and the WTO had little capacity to craft a politically sufficient response. Its architecture is designed to constrain the policies of member governments so that market forces can determine competitive outcomes. It was never designed to deal with centrally controlled economies where market forces do not operate. China's export-led economy is fueled by a centralized industrial policy, often carried out by state-owned enterprises (SOEs) that either engage in exporting directly or supply capital and inputs to exporters. The activities of these nominally independent entities do not fit easily into WTO subsidy rules. China's export growth was also, at times, attributed to currency manipulation, another practice that WTO rules barely address.Footnote ¹¹

The obvious solution from the perspective of China's trading partners was the negotiation of new rules. But in the consensus-based WTO system, new rules would require the assent of China and all its political allies. Even a plurilateral negotiation among a subset of members requires authorization by consensus in current practice. The consensus-based system for decision making thus makes it exceedingly difficult for new challenges, such as those posed by China, to be addressed legislatively. Indeed, it has led to paralysis in WTO negotiations on almost all issues of major importance (although we do not mean to suggest that China is the only culprit). Even issues that seem completely straightforward, such as curtailing the subsidization of fishing in overfished fisheries, have run up against a lack of consensus.

In the absence of new rules, members face unattractive options. First, they can seek to negotiate outside the WTO while respecting WTO rules, but they have no leverage to the extent that a counterparty, such as China, is comfortable with the status quo, and WTO rules would have to be violated to put pressure on China to come to the table.

A second option is to rely on the ‘flexibility’ mechanisms built into the WTO treaty system, but these tools accomplish little in relation to most of the problems posed by China. Renegotiation of tariff commitments on an MFN basis under GATT Article XXVIII does not permit the concerns regarding China to be targeted. Temporary safeguard measures are of little value when the problems emanate from China's long-term industrial policies, and safeguards too require non-discrimination in most instances. The general exceptions of GATT do not speak to most concerns with China, as the complaints about China rarely relate to health and safety, resource conservation, and the like. The national security exception might offer useful flexibility in some situations but, as noted below, measures taken against China on national security grounds have been held to be inconsistent with GATT by an arbitral panel. Finally, the remedies for ‘unfair’ trade practices involving dumping and subsidies have been brought to bear in cases involving China, with some regularity, but they often prove unsatisfactory. China has challenged many of these measures with significant success as noted below. Further, China can often relocate production to other countries to avoid countermeasures. Lastly, as these cases require extensive investigations industry-by-industry or even company-by-company, a trade remedy approach involves exceedingly high transaction costs.

The final option is to ignore WTO rules and proceed with unilateral measures to stimulate negotiation or restore levels of trade protection that exceed WTO-permissible levels. The United States has largely converged on this third approach after many years of being stymied by existing WTO law.

US frustration with the WTO, however, was not all about China. Article 3(2) of the WTO Dispute Settlement Understanding recites that dispute settlement rulings ‘cannot add to or diminish the rights and obligations provided in the covered agreements’. Yet, they may nevertheless ‘clarify the existing provisions of those agreements in accordance with customary rules of interpretation of public international law.’ In the view of the United States, the AB proved overly aggressive at ‘clarifying’ obligations to the point that it often ignored the admonition not to ‘add to or diminish’ them. Moreover, although the DSU does not provide for stare decisis in relation to AB rulings, the AB in practice insisted that its rulings be followed in future cases, chastising panels that deviated from them.Footnote ¹² A series of dubious AB rulings against the United States in the trade remedy area, on matters of high political salience in the United States, were thereby locked in place for future cases raising comparable legal questions.

The key decisions in this regard are well known. One group involved safeguard measures.Footnote ¹³ in which the AB insisted that import surges be linked to ‘unforeseen developments’ even though this requirement found in the original GATT was omitted from the WTO Safeguards Agreement, had been omitted from national statutes authorizing safeguards for many years, and was effectively a dead letter in modern GATT practice.Footnote ¹⁴ The understandable failure of national governments to attend to the matter under national law thus immediately rendered their safeguard measures impermissible. To make matters worse, the AB insisted on an analysis of causation or ‘non-attribution’ that was logically incoherent and impossible to satisfy,Footnote ¹⁵ and sowed enormous confusion over the acceptable way to determine whether imports had increased.Footnote ¹⁶

The next line of problematic AB decisions concerned the practice of ‘zeroing’ in antidumping cases. Zeroing is a method of computing average dumping margins that treats ‘negative’ dumping as zero dumping, thus raising the average amount of dumping and inflating antidumping duties. It has long been criticized as an artifice for generating increasing trade protection, but it was routinely used in the United States and elsewhere for many years. During the negotiations for the Antidumping Agreement, US negotiators successfully avoided any reference to zeroing or any specific restriction on it in the text. The United States also secured a deferential standard of review (Article 17.6(ii) of the Antidumping Agreement) that permits members to implement national antidumping measures that comport with permissible interpretations of ambiguous text (akin to the Chevron deference standard of US administrative law at the time). The AB nevertheless held that zeroing was impermissible in a series of decisions that extended the prohibition to all manner of antidumping investigations including administrative reviews, new shipper reviews, and so on. In the US view, these decisions completely undermined the bargain that had been struck during negotiations.Footnote ¹⁷

Finally, the AB ruled against the United States in a case involving Chinese SOEs. The United States had determined in several cases that the pricing behavior of SOEs conferred a countervailable subsidy (such as a below market interest loan from a government-owned bank). But the finding of subsidization was consistent with WTO rules only if the SOE could be considered a ‘public body’ under the Agreement on Subsidies and Countervailing Measures. The AB held that China's ownership or control of an SOE is not enough to make it a ‘public body,’ and that it must also be performing a ‘governmental function’.Footnote ¹⁸ This ruling had no textual basis and did considerable violence to one of the few tools available to the United States to deal with Chinese industrial policy.

This series of AB decisions, along with subsequent panel rulings holding that the invocation of the GATT national security exception is reviewable under the DSU and that US steel and aluminum tariffs fall outside the GATT national security exception,Footnote ¹⁹ have led the United States to conclude that a functioning dispute settlement system is not in the national interest. A bipartisan willingness to deviate from core WTO obligations on tariffs, subsidies, and non-discrimination has only grown since the AB was disabled and the prospect of definitive adverse legal rulings has evaporated.

5. Can the Patient be Saved?

A variety of factors suggest a rocky road ahead. The China problem remains central. Chinese trade, Chinese industrial policies, and China's failure to move toward market orientation as was anticipated at the time of accession all continue to undermine cooperation. Members do not agree on how to respond, nor do they have the same economic ties with the United States and China. The long tail of the dislocation from the ‘China Shock’, increasing frustration due to the lack of trade concessions by the developing nations, and the history of overreach by the AB have all caused a backlash in the United States against the WTO and trade agreements in general.

Cracks in the political support for liberal trade first appeared during the Obama administration, which blocked the reappointment of certain AB judges who had issued adverse rulings. Things then accelerated in the first Trump administration with a complete blockage of AB appointments, steel and aluminum tariffs based on dubious ‘national security’ claims, and punitive tariffs against China tied to a range of Chinese measures. The Biden administration might have reversed the Trump policies but did not. Instead, it continued the blockage of AB appointments, maintained many of the Trump tariffs (even doubling down with 100% tariffs on Chinese electric vehicles and increased tariffs on Chinese EV batteries and solar cells), and introduced tens of billions of dollars in industrial subsidies, including prohibited import substitution subsidies.

For now, John Jackson's ‘rules-based’ system has run its course. Perhaps the source of the problem was too many lawyers and too few economists and political scientists. We will never know for sure. But what we did learn from the experiment is that the WTO's formal rules, and the manner of their adjudication, paid insufficient attention to the political pressures on national leaders. The WTO is too legalistic and inflexible. The hope of reining in the powerful with a rules-based system ran up against the reality that the powerful can strike out on their own when that offers a better outcome for them.

Today, the trading system is plagued by rising nationalism, nativism, and populist governments. Given America's unilateralism, there is a growing fear that global trade will decline, and cooperation may be largely bilateral or plurilateral rather than multilateral. Given that the GATT was created to combat a similar political environment, there is a lesson to be learned from the GATT period. The GATT was a thin organization with politically savvy rules. The members did not share all policy goals, but did share the view that liberalized trade is welfare enhancing. If new rules are to be written, and they are needed, they not only need to seek a stable and transparent system for the trading world, but they also need to be sensitive to the political exigencies of member states. The strength of the GATT may well have been its weakness.