Law and Contemporary Problems: Judith L. Goldstein, Negotiate Or Litigate? Effects Of Wto Judicial Delegation On U.S. Trade Politics, 71 Law & Contemp. Probs. 257 (winter 2008)

Cited: 71 Law & Contemp. Probs. 257
[*pg 257]

NEGOTIATE OR LITIGATE? EFFECTS OF WTO JUDICIAL DELEGATION ON U.S. TRADE POLITICS

JUDITH L. GOLDSTEIN*

RICHARD H. STEINBERG^**

I. INTRODUCTION

II. DELEGATION IN THE GATT/WTO
   A. Delegation to the Secretariat
   B. Judicial Delegation: How Did It Occur and Is It Binding?
   C. Understanding Judicial Lawmaking -- and Its Limits
   D. The Expansiveness of Judicial Lawmaking: The GATT and WTO Compared
   E. Implications of WTO Judicial Action

III. UNDERSTANDING TRADE DELEGATION AND COMPLIANCE

IV. CONCLUSIONS

FOOTNOTES

I. INTRODUCTION

Since its creation in 1995, the World Trade Organization's (WTO) critics and supporters have agreed on two things: there has been a surprising amount of judicial lawmaking¹ and an equally surprising lack of success at advancing a broad trade "deal" among members. This article focuses on judicial lawmaking and argues that in some circumstances, decisions rendered by court-like bodies in the WTO are adhered to even when the same policy would not gain support in multilateral negotiations. In this sense, the two may be thought of as substitutes. Why? As U.S. expectations of reciprocity with developing countries have increased, and as the General Agreement on Tariffs and Trade (GATT)/WTO system has expanded in number and type of members, consensus-building to legislate new deals has become increasingly difficult, reflecting the often-conflicting constituent interests brought to the table.² As the prospects for broad legislative rulemaking have declined, judicial lawmaking has become more common, especially through interpretation of unclear rules and the filling of gaps in WTO agreements. Such lawmaking has been particularly evident in cases challenging subsidies or countervailing measures, anti-dumping duties, and safeguards measures, and has had considerable effects on the steel sector and agriculture. Adherence to such liberalizing judicial [*pg 258] action is surprising, since the same trade liberalization could face domestic resistance in the context of legislative trade negotiations.

In this article, we examine the question of judicial lawmaking and adherence to dispute-settlement decisions in the context of overall delegation to the GATT/WTO. The notion of delegation used here is drawn from Terry Moe, who argues that "[t]he principal-agent model is an analytic expression of the agency relationship, in which one party, the principal, considers entering into a contractual agreement with another, the agent, in the expectation that the agent will subsequently choose actions that produce outcomes desired by the principal."³ As noted by Bradley and Kelley in this issue, the institutional effect of such delegation is to change the incentives of all parties and, as they note, may lead to a "sovereignty cost."⁴ We consider the case of delegation to the secretariat, but focus on dispute resolution in the WTO Appellate Body; we assume that the relationship among the principals of the GATT/WTO, that is, its member nations or customs territories, always held some degree of a principal-agent relationship with the central administrative structures. Even though the membership exerted far more oversight over the secretariat than is found in most other international economic agencies, an agency relationship did develop with a nascent secretariat in the mid-1950s and after 1995 grew with the establishment of the contemporary dispute-settlement system. What we suggest, however, is that it is not necessary to think of the relationship as "binding." It has always been understood that the principals may, in practice, renege on agreements made by the collective, including informal opinions by the legal office of the secretariat as well as formal decisions of the dispute-settlement process.

As opposed to international organizations created in the same period, such as the World Bank or the International Monetary Fund (IMF), the founders of the early GATT never intended to delegate authority to a central agent. Rather, delegation was an unintended byproduct of the creation of an underspecified set of rules and procedures, first by the delegates who created the GATT in 1947, and later by those who participated in its reinvention as the WTO in 1995. There has been no de jure delegation to the regime in the sense that most countries have not given up legal authority under domestic law to maintain trade rules they deem appropriate. But there is "behavioral" delegation, meaning that countries have de facto begun to act as if they have given authority to the regime.⁵

This form of delegation has had lasting effects, and the functioning of the secretariat and the dispute-settlement system should be viewed in this context. [*pg 259] In particular, behavioral delegation may have enabled judicial activism as an unintended and imperfect substitute for liberalization through legislative action. Through judicial action, the dispute-settlement system (and the secretariat's influence on dispute-settlement panel reports) has been able to define and redefine trade rules so as to keep markets open. But whereas the actions of panels, the Appellate Body, and the secretariat appear to be consistent with their liberal world views, the absence of de jure delegation makes the high degree of national adherence to these judicial decisions a puzzle that begs for explanation.

This article focuses on this puzzle of adherence to judicial decisions by the most powerful member of the WTO -- the United States. We argue that, given the configuration of domestic interests in the United States, judicial compliance is not surprising. Rather, even when powerful domestic sectors seem to have an effective veto over liberalization in trade rounds, the rules and procedures brought to bear in complying with the WTO's judicial decisions reconfigure domestic coalitions in ways that favor liberalization long thought unimaginable. Liberalization is occurring, not because of delegated sovereignty, but because judicial actions are reconfiguring the relative power of domestic actors. Looking back, similarities are evident between the role of the WTO's judicial arm and the European Court of Justice's (ECJ) liberalizing role in the early days of the European Union (EU).

The article begins with a review of delegation in the GATT/WTO and the rise of judicial lawmaking. Next, it examines the court's actions and offers a one-dimensional preference space model, described below, that illustrates why U.S. compliance may be expected. Finally, it concludes with some general observations on the issue of sovereignty costs in the GATT/WTO regime.

II. DELEGATION IN THE GATT/WTO

When the United States invited fifteen nations to join in an initial round of trade talks in 1946, participants did not expect the meeting to yield the rules and procedures for commercial policy that would regulate trade for the subsequent century. Rather, participants were looking ahead to what was the draft of the forthcoming Havana Charter,⁶ which would have established the International Trade Organization (ITO), as the document that would specify the procedural rules and governance of the new regime. When agreeing to the interim agreement, the Protocol of Provisional Application of the GATT,⁷ participants assumed that the regime's rules and procedures were short-term and paid scant attention to the structure and procedures of the new organization. The GATT [*pg 260] structure did survive, however, and with minimal changes over the following decades. This absence of detailed procedural rules and the lack of a consensus on the scope of the trade organization influenced the trajectory of the trade regime in two fundamental ways.

First, because of this unusual history, the relationship between the principals (states and customs territories) and an agent (initially, the secretariat) has remained under-defined throughout GATT/WTO history. In fact, many would say that thinking in these terms is misleading because the organization was always "member-driven." The early GATT did not even have a secretariat per se since the creators envisioned the organization as a treaty and not as an international organization.⁸ But from the start, there needed to be at least a small central governance structure: what was called a "secretariat" for almost fifty years was formally a staff under the executive secretary of the Interim Commission for the ITO; the GATT contracting parties called this executive secretary their "Director-General." By the 1980s, that structure had grown into numerous divisions with a separate legal office. Yet it was not until the Agreement Establishing the WTO⁹ that the secretariat was formally created and the structure of rule adjudication, in the form of a judiciary, was more fully defined.¹⁰

Second, liberalization at the international level has occurred within a regime that never fully settled the basic procedural issues accompanying the setting of trade policy. From the beginning, the GATT liberalized episodically through trade rounds.¹¹ In each round and within each negotiating country, protectionists threatened with near-certain economic loss from liberalization joined together to make sure they were not on the chopping block. Each nation devised different mechanisms by which to circumvent protectionist interests at home, with the result that liberalization rounds were highly polarized and occurred under the shadow of politics at home. Deals were precarious, and politicians needed to worry about whether they could find support at home for agreements signed in Geneva.

These negotiating procedures became increasingly problematic in the 1980s, due first to the rapid increase in the number of nations participating in the regime and, second, to increasing efforts to discipline behind-the-border [*pg 261] measures. Enlarged membership and an expanded agenda made it more difficult to legislate liberalization. As evidenced by the Doha Round, it has become difficult to reach consensus on anything other than general principles.

This inability to specify detailed rules and procedures on trade opened the door for increased de facto delegation to the secretariat over the course of the GATT years and for considerable de facto delegation to the dispute-settlement system since the creation of the WTO.

A. Delegation to the Secretariat

In theory, the GATT/WTO secretariat is granted limited independence from the membership, and the Director-General, chosen by a consensus of the members, is given little real authority. The members have stated clearly that autonomy might lead the organization to promote policies inconsistent with political pressures at home, a problem of particular importance for the more powerful nations participating in the regime. As a result, the members of the secretariat rarely chair committees; and the growth in the number of professionals, as opposed to support staff, has been far less than expected, given the rapid increase in membership. But even with these constraints, delegation occurred; the inability of the membership to concur easily on comprehensive and detailed rules, both for trade policy and the specific scope of the secretariat's powers, led to more autonomy than most countries would have wanted. As with all agents, those in Geneva had an often differing position from the membership on trade matters: they consistently advocated for more open markets.

The growth in the secretariat's functions is not surprising. As the demands on the organization increased to both provide information on trade practices and support the process of trade liberalization, the secretariat was forced to perform more functions and become the legal memory of the regime. The secretariat performs at least three functions that have been delegated de facto by the members.

First, the secretariat is the keeper of information. Not only does it collect and disseminate data provided by members to others in the organization, but it also collates and organizes -- and often creates -- the data sets on trade flows and trade restrictions used as bases for trade round negotiations.

Second, the secretariat plays a key role in dispute settlement. Even though members do not allow cases to be based formally on precedent (that is, while precedent is highly persuasive, there is no stare decisis), the secretariat helps assure the intellectual continuity of panel decisions. Even with no formal legal role, members of the secretariat influence decisions through advice to panel participants and recommendations on written panel reports. The effect has been generally consistent panel reports and far fewer decisions that would be politically difficult to administer. Particularly through the dispute-settlement panel stage, the secretariat is the collective memory of the organization: understanding the political pressures of members, it serves to mediate the [*pg 262] dispute process and increase the likelihood that decisions resonate with political realities.

Third, the secretariat provides support for small developing nations in both rounds and in dispute-settlement cases. The regime provides training for government officials on how to interpret rules and how to stay in compliance with these rules.¹² Training sessions both transfer information and socialize the trade officials of new members to support the underlying purposes of the regime. Past employees of the secretariat advise developing countries on a range of activities, from dispute-panel processes to the collection of appropriate information for trade talks to reform proposals on how to increase their voice in the organization. While the larger nations can rely on expertise at home, smaller and developing economies rely more extensively on the secretariat for support.

Further, whereas the first director-generals were selected because of their expertise as international civil servants, the most recent appointees have political, rather than technical, backgrounds. The lack of expertise at the top of the organization, either in economics or in international law, has created a widening division between the professional staff and the political leadership. Because the Office of the Director-General relies upon the permanent staff for support, its relationship with the civil service has become increasingly strained over time. This is not to suggest that power resides in the Director-General's office. Its personnel are as constrained by the membership as is the secretariat. Still, the relationship between the professional staff and the Director-General's office has evolved with changes in the organization. Whereas the early secretariat saw the Director-General as its representative to the contracting parties, the more politicized relationship of recent years has led the secretariat to view the Director-General's office as pursuing interests often at odds with the more liberal policies being promoted by the professional staff.

The mantra among the staff and the members of the WTO is that the organization is "member-driven," but the reality of a complex organization supported by an educated and sophisticated staff is not entirely consistent with this image. Institutional change has had unanticipated effects on the organization; and given that the rules of the WTO became more complex and inclusive, some agent had to be entrusted with the job of maintaining the institutional memory important to interpreting rules and monitoring the evolution of members' trade policy. Unintentionally, the role of the secretariat expanded, even as the membership attempted to keep the bureaucracy in a subservient role and as the leadership became more a political position and less one entrusted to an individual with trade expertise.

[*pg 263]

B. Judicial Delegation: How Did It Occur and Is It Binding?

The WTO's dispute-settlement rules and processes were an attempt to reform problems in the GATT's dispute-settlement rules and processes. The GATT had been created to facilitate bargaining among participants over the regulation and liberalization of trade policy. Although created with the concern that contracting parties live up to their promises, the mechanism for oversight of rule adherence depended upon a contracting party, in the name of a producer, complaining about a violation. The secretariat had neither oversight nor judicial power. Although the secretariat compiled data on trade practices starting in 1989 by means of the Trade Policy Review Mechanism (TPRM), the data in these reports constituted neither formal grounds for a dispute nor legal evidence of a country's trade practices.¹³ Monitoring occurred through oversight by individual contracting parties; when a complaint was filed, the secretariat assisted with forming a panel. If the panel found in favor of the complainant, and the respondent did not block a consensus to adopt the panel report, then the contravening party was required to change its behavior.¹⁴ In the absence of a change in practice, little could be done other than sanction retaliation -- though that, too, required a consensus.¹⁵

The weakness of the GATT dispute-settlement procedure, the "panel procedure," became increasingly apparent in the 1980s. In the early years of GATT, from 1948 through 1959, contracting parties brought relatively few (fifty-three) legal complaints against each other.¹⁶ The panel procedure was developed in these years,¹⁷ and it was used in over half of these cases.¹⁸ As the number of contracting parties grew, the number of conflicts increased. Perhaps reflecting dissatisfaction with the settlement procedures, the number of formal complaints did not rise and, in fact, fell after 1963. Whereas almost sixty cases were dealt with by the dispute-settlement process through 1963, only one new case was brought forward through 1970.¹⁹ Hudec argued the issue was legitimacy -- the process was viewed as unfair.²⁰ Legitimate or not, the caseload [*pg 264] increased in the 1970s to a total of thirty-two,²¹ and in the 1980s the panel process began to be invoked regularly.²² Of the 115 complaints filed in the 1980s, forty-seven led to panel reports.²³ However, only about two-fifths of rulings for the complainant resulted in full compliance by the respondent.²⁴ Nonetheless, the increased caseload forced the secretariat to create a separate legal division, which had the effect of encouraging even more legal complaints.²⁵ As the number, visibility, and importance of cases increased, so too did the number of cases in which consensus was blocked.

During the 1970s and '80s, in response to a growing trade deficit, perceptions of unfair trade practices abroad, and frustration with the sclerotic GATT dispute-settlement system, the United States turned increasingly to domestic law to deal with its trade disputes. Specifically, a "unilateral" approach to addressing trade disputes was enacted by the U.S. Congress in the form of Section 301 of the Trade Act of 1974.²⁶ Section 301 permits (and in some cases requires) the President to impose retaliatory trade sanctions on countries engaging in any practice that "burdens or restricts United States commerce" and is "unjustifiable," "discriminatory," or "unreasonable"²⁷ -- as determined by the Office of the United States Trade Representative (USTR).²⁸ Thus, when a foreign government blocked the GATT dispute-settlement process, the U.S. government often found itself in a position of threatening unilateral trade retaliation against that government unless it agreed to change its trade practices in accordance with Washington's demands. This approach to the settlement of trade disputes was not viewed favorably by the rest of the world.

Perhaps as part of an attempt to credibly commit to the regime in the post-Cold War days or in response to growing criticism of American intransigence in a number of foreign policy arenas, the United States championed a change in dispute-settlement procedures.²⁹ In 1989, the United States supported a new dispute-settlement understanding adopted in the Midterm Review of the Uruguay Round;³⁰ in the next year the United States heralded an even more radical change in the rules. The underlying theme was that it was time for member nations to agree to give up their right to block a consensus in the establishment of panels, the adoption of panel reports, and authorization of retaliation. Initially, the U.S. position aligned closely with Canada's but not with [*pg 265] those of the other Quad members -- Japan and the European Union. The U.S. proposal not only included the automatic adoption of panel reports, but it also created a right to appeal to a new WTO Appellate Body whose purpose was to oversee the work of panels on questions of law.³¹ The time limits established in this process would be modeled on the United States' own Section 301 statute. For the United States, this seemingly radical position was contingent on a crucial proviso -- that the substantive rules adopted in the Uruguay Round had to be adequately specific and reflect U.S. policy objectives.³²

At home, increased delegation under these terms was argued to be consistent with American interests: the United States would remain more often in compliance with WTO rules that reflected its interests and policy objectives than would its trading partners. If the WTO's substantive rules were to the United States' liking, and the dispute-settlement procedures were both consistent with the timeline for action under Section 301 and could ultimately authorize retaliation for noncompliance, then a more legalized WTO dispute-settlement system would legitimize U.S. use of its market power to pressure other countries to comply with its trade-policy objectives.³³ Tellingly, the United States shifted to this position on dispute-settlement reform at the end of 1990, which is the same time that it reached agreement with the European Communities (EC) to impose the results of the Uruguay Round on developing countries via the "single undertaking."³⁴ Few nations agreed with the U.S. analysis of who would end up in front of panels. But curbing U.S. unilateralism was one of the most salient elements in both Japan's and the EC's publicly stated interest in the Round, and both endorsed the reforms.

In the end, the new procedures occupied twenty-four pages of text, elaborating a process that is more complex and precise than in the past and that covers all areas of WTO agreements and state behavior subject thereto, which would include U.S.-imposed trade sanctions pursuant to Section 301. The Dispute Settlement Understanding (DSU)³⁵ is far more obligatory, automatic, and apolitical than the GATT rules. The effect of the change -- a vast increase in use of the DSU -- was far broader than anticipated in any commentary of the time. Although 452 dispute-settlement complaints were filed in the forty-six [*pg 266] year period of the GATT system,³⁶ 279 complaints were filed in the first eight years alone of the WTO system.³⁷

Reform meant that two substantive changes occurred in the WTO judicial system. First, judicial action became more automatic. A consensus is now required to block the establishment of a panel,³⁸ the adoption of a report,³⁹ or an authorization of retaliation for continued noncompliance⁴⁰ -- a reversal of the former rule that required a consensus to move through each of these stages. Of course, complainants would not agree to block the establishment of a panel they are demanding, and prevailing parties would not block the adoption of favorable panel reports.

Second, the reform led to the creation of a judicial body to which nations could appeal panel reports -- the WTO Appellate Body. This body's mandate is formally limited to the review of legal findings made by panels, given the facts established by the panel.⁴¹ The Appellate Body has seven members, chosen by the WTO's Members at large, and appeals are heard by a subset of three of the seven members.

To what extent are the decisions of the WTO Appellate Body indicative of binding delegation? As a matter of international law, there is a legal obligation to comply with their decisions.⁴² But as a matter of domestic law, few if any national legal systems give direct effect to Appellate Body decisions.⁴³ As a matter of domestic law, sovereignty has not been "delegated" to the Appellate Body. Therefore, behaviorally, compliance is a domestic political decision, analyzed in greater detail below.

The preceding analysis of the negotiating history behind the move to legalization of GATT/WTO dispute resolution suggests that it was not intended to lead to expansive judicial lawmaking. The switch to automatic dispute-resolution and the establishment of the Appellate Body were seen by the United States as an opportunity to foster implementation of and compliance [*pg 267] with the deals struck in the legislative process -- even if those deals were not optimally efficient, and even if they were not considered equitable. The dispute-settlement process was intended to fulfill that purpose by offering a neutral judicial process to enforce the WTO agreements. The United States was willing, and remains willing, to delegate to WTO dispute settlement the authority to enforce the WTO "contract." The U.S. government may win a few cases, and it may lose a few, but WTO dispute-settlement losses are of little consequence if the overall effect of the system is to help enforce substantive agreements supported by the United States.⁴⁴ Hence, to the extent that it is performed effectively, the judicial function helps reinforce political support for the WTO by powerful countries. Furthermore, from the U.S. perspective, the DSU is a vehicle by which the U.S. government may legitimately challenge WTO-inconsistent practices by foreign governments, and it has simultaneously helped solve America's credibility problem arising from unilateralism. From the U.S. government's perspective, the radical judicial reforms of the Uruguay Round represented not a multilateralization of U.S. unilateralism, but an Americanization of the GATT/WTO dispute-settlement process. The WTO dispute-settlement system was not intended as some new form of delegation (in the sense of delegating sovereignty de facto), and certainly it was not construed as binding (in the sense of delegating sovereignty de jure). The United States could always refuse to comply with an Appellate Body report, if inconvenient.

C. Understanding Judicial Lawmaking -- and Its Limits⁴⁵

Few, if any, architects of increased legalization at the WTO foresaw the institutional development that would follow. Many scholars have suggested that judges may behave strategically and favor increasing their own authority,⁴⁶ yet few Uruguay Round negotiators anticipated or intended the Appellate Body to engage in lawmaking.⁴⁷ As suggested above, U.S. policymakers expected that [*pg 268] the new dispute-settlement system panelists and Appellate Body members would apply WTO rules to cases presented before them -- not make law.

As in domestic legal systems, rules and principles guiding the interpretation of public international law permitted the Appellate Body to take a range of interpretive stances: at one extreme, a restrained interpretive stance that avoids opportunities to make law and is highly deferential to the express consent of states; at the other extreme, an expansive interpretive stance that is less deferential to state consent, favors dynamic interpretation of treaty provisions, and expands upon terms and gaps. Largely in the interests of completeness, coherence, and internal consistency of WTO law, the Appellate Body chose a more expansive stance on both the question whether to interpret and on the method used for interpretation. The resulting judicial decisions have created an expansive body of new law.

WTO judicial lawmaking has two dimensions: filling gaps and clarifying ambiguities. Gap-filling refers to judicial lawmaking on a question for which there is no legal text directly on point, whereas ambiguity clarification refers to judicial lawmaking on a question for which there is legal text that needs clarification.⁴⁸

First, the DSU's silence on many procedural questions has been seen by some as an invitation to the Appellate Body to make procedural rules. In some cases, the Appellate Body has created law that fills procedural gaps in WTO agreements, even though the gaps resulted from sharp disagreement among WTO members about how to fill them. For example, the Appellate Body decided -- without clear guidance from WTO agreements -- that dispute-settlement panels could consider amicus curiae briefs submitted by nonstate actors.⁴⁹ In so ruling, the Appellate Body relied on general language in DSU Article 13.1, which provides that "[e]ach panel shall have the right to seek information and technical advice from any individual or body which it deems appropriate."⁵⁰ Regardless of the merits on the question, the Appellate Body's interpretation of Article 13 was made in the context of several years of north-south deadlock on the question whether to permit amicus briefs; few developing countries would have consented to an agreement with that outcome, yet the Appellate Body chose to interpret the DSU as supporting it.

Similarly, the Appellate Body established that private lawyers may represent WTO Members in oral proceedings, despite European Community [*pg 269] (EC) and U.S. opposition on grounds that the practice from the earliest years of the GATT was to permit presentations in dispute-settlement proceedings "exclusively by government lawyers or government trade experts."⁵¹ The Appellate Body acted at odds with nearly fifty years of GATT practice, reasoning that "nothing in the Marrakesh Agreement [ -- the Final Act Embodying the Results of the Uruguay Round of Multilateral Trade Negotiations -- ] . . . nor in customary international law or the prevailing practice of international tribunals . . . prevents a WTO Member from determining the composition of its delegation in Appellate Body proceedings."⁵² At the panel stage, this practice of permitting participation by nongovernment lawyers was subsequently adopted.⁵³

Second, the WTO Appellate Body has engaged repeatedly in a form of lawmaking by which it has given specific meaning to ambiguous treaty language. Such clarifications may cause a negative political reaction by WTO Members or nongovernmental stakeholders that engaged in behavior within a range of possible meanings. For example, in U.S. -- Shrimp-Turtle, the Appellate Body decided whether the United States could rely on GATT Article XX(g) to ban the importation of certain shrimp and shrimp products from WTO Members that did not maintain laws guaranteeing particular methods of protecting endangered sea turtles in the process of shrimp fishing.⁵⁴ GATT Article XX(g) excepts certain measures from the GATT's affirmative obligations that are necessary for the "conservation of exhaustible natural resources,"⁵⁵ but the provision is ambiguously silent on the question of whether such exhaustible natural resources must be located within the jurisdiction of the country invoking the exception. Earlier decisions, which suggested that they must, catalyzed enormous debate between the members. The Appellate Body offered a dynamic interpretation of the conditions under which the GATT Article XX(g) exception for conservation of exhaustible natural resources could be invoked, stating that it must be read "in the light of contemporary concerns of the community of nations about the protection and conservation of the environment."⁵⁶ After concluding that the measures in question fell within the meaning of Article XX(g), the Appellate Body interpreted the chapeau to Article XX and established at least five specific factors that would apply in considering whether a measure contravenes the terms of the chapeau.⁵⁷ Some of the factors had no textual lineage (for example, whether the respondent's [*pg 270] actions have an "intended and actual coercive effect on the specific policy decisions made by . . . Members of the WTO").⁵⁸ In short, the Appellate Body ruling provided an approach to balancing trade-environment issues, despite WTO Members' decade-long deadlock about how to achieve balance on the question.

Third, in a number of instances, the Appellate Body has given precise and narrow meaning to language intentionally left vague by negotiators, either because they could not agree on more specific language, or in order to permit a range of alternative behaviors or national practices. For example, in three decisions, the Appellate Body fleshed out the causation analysis to be used in safeguards cases,⁵⁹ which Uruguay Round negotiators had intentionally left ambiguous.⁶⁰ In the U.S. -- Lamb Meat case, for example, the Appellate Body established an affirmative requirement that national authorities analyze not only the nature but also the "extent" of causes other than those prompting the safeguards investigation so as to not attribute injury from other causes to imports subject to the investigation.⁶¹ The Appellate Body took a similar approach in the anti-dumping context.⁶² In a more recent cause celebre in Washington, the Appellate Body overturned a longstanding domestic practice of "zeroing"⁶³ -- a practice by which the U.S. Commerce Department, when calculating how much product had been dumped, would ignore (or set at zero) examples in which the imported product had a negative dumping margin (that is, the product was sold for a higher price in the United States than in the domestic market).⁶⁴ Despite an EU-U.S. informal understanding to permit the continuation of this practice and the consequent exclusion of this topic from the WTO Anti-Dumping Agreement, the Appellate Body ruled that zeroing was [*pg 271] illegal.⁶⁵ The U.S. government and commentators have identified several other cases in which the Appellate Body or the dispute-settlement panels have given a specific and narrow interpretation of language in WTO agreements that was intended, by at least some of its negotiators, to be ambiguous and to permit a range of national practices.⁶⁶ Decisions like these might enhance efficiency,⁶⁷ but they are certain to engender negative political reactions in countries that intended to consent to broader interpretations.

Finally, conflict between GATT/WTO texts (or between text and GATT practice) may create ambiguity. In a handful of cases, the Appellate Body has read language across agreements cumulatively in a way that has generated an expansive set of legal obligations. Perhaps most controversially, the Appellate Body ruled that national authorities imposing a safeguards measure must demonstrate the existence of "unforeseen developments."⁶⁸ In a 1951 case, a GATT Working Party agreed that the application of Article XIX safeguards measures could be based on an argument that an unexpected degree of change in consumer tastes that increased imports constituted demonstration of "unforeseen developments."⁶⁹ Given this implicitly broad interpretation of the phrase, which would seem to allow almost any increase in imports to constitute "unforeseen developments," subsequent GATT panels did not require national authorities to demonstrate "unforeseen developments" prior to imposing safeguards measures. Moreover, the WTO Safeguards Agreement makes no reference to a requirement to demonstrate "unforeseen developments,"⁷⁰ and the negotiators of the Agreement expressly considered and rejected inclusion of any such requirement.⁷¹ The cumulation of GATT practice, relevant texts, and negotiating history created an ambiguity over whether "unforeseen developments" must be demonstrated in safeguards cases. Focusing on GATT [*pg 272] Article XIX:1(a), the Appellate Body read all of the relevant GATT/WTO law and practice cumulatively so as to conclude that a demonstration of "unforeseen developments" must be shown if a safeguards measure is to be applied.⁷²

D. The Expansiveness of Judicial Lawmaking: The GATT and WTO Compared

Several indicators suggest that judicial lawmaking is more expansive in the WTO than in the GATT. Compare, for example, discussions about judicial lawmaking among DSU negotiators in the GATT Uruguay Round to discussions among DSU negotiators in the WTO's current Doha Round. Analysis of publicly available notes by the secretariat concerning the Uruguay Round Negotiating Group on Dispute Settlement indicate no instance in which a negotiator suggested that GATT dispute-settlement panels were activist or engaged in expansive lawmaking.⁷³ Although Uruguay Round preparatory materials are incomplete, the leading secondary history of the Uruguay Round DSU negotiations nowhere mentions any discussion of activism or lawmaking by GATT panels.⁷⁴ Finally, interviews with lawyers from the USTR identify only four GATT panel reports criticized by contracting parties as instances of overly broad lawmaking or inappropriate interpretation.⁷⁵

In contrast, analysis of publicly available official documents of the current Doha Round Negotiating Group on Dispute Settlement, which convenes in Special Sessions of the DSB, indicates that in the first eighteen months of negotiations -- by June 2003 -- concern about instances of or proposed solutions [*pg 273] to judicial lawmaking by WTO panels or the Appellate Body had been raised seventy-seven times by representatives of fifty-five WTO Members (including the EC and the United States), focusing on at least ten dispositive WTO dispute-settlement reports adopted in the WTO's first eight years.⁷⁶ Moreover, confidential notes on meetings of the DSU Reform Group, in which ambassadors representing approximately ten WTO Members have met regularly to informally discuss concerns about operation of the DSU, reveal that judicial lawmaking was discussed at almost every meeting during the period for which the notes are available.⁷⁷

Consider journal articles on the subject. Of 110 selected U.S. and Canadian law-journal articles on GATT dispute settlement published from 1982 through 1994 (the last year of the GATT system),⁷⁸ only two suggest cases in which controversial, expansive, or activist judicial lawmaking might have taken place.⁷⁹ In contrast, at least fifty-one articles published in similarly selected U.S. and Canadian law journals in the first eight years of the WTO discuss controversial, expansive, or activist judicial lawmaking or cases in the WTO.⁸⁰

This expansive interpretive stance by the Appellate Body is not without limits. It is constrained by politics. For example, powerful members, particularly the EC and the United States, have had a de facto veto over the appointment of [*pg 274] Appellate Body members: in the WTO's early years, these powerful members engaged in a comparatively cursory review of Appellate Body nominees; in more recent years, as the Appellate Body's capacity to make law became apparent, the United States began engaging in a thorough review and interview of Appellate Body nominees, blocking the appointment of some nominees who were seen as too activist.⁸¹ Similarly, members are not shy about complaining when the Appellate Body engages in lawmaking they dislike, and proposals by powerful members to rewrite parts of the DSU in the Doha Round may have had a sobering effect on the Appellate Body. Loud public claims that the Appellate Body is engaged in "activism" raise the possibility that WTO judicial lawmaking could become a mass politics issue. To some extent, agent slack is limited.

E. Implications of WTO Judicial Action

Three key points can be derived from WTO case law. First, judicial action is a substitute for legislative action. Under the guise of interpretation, the Appellate Body has legislated rules for regime members. Of course, as shown below, judicial lawmaking differs in form, rate, and substance from negotiated lawmaking, so they are imperfect substitutes.

Second, in just about all cases, Appellate Body interpretations have favored more openness, which is its preference.⁸² The observation that the dispute-settlement system favors openness is intuitive: in all cases, complainants advance interpretations of WTO agreements that challenge a respondent's trade barrier, and respondents argue for interpretations that would permit the maintenance of the barrier. If the complainant wins even some of the cases, the result is more open trade. For WTO cases initiated before 2001, eighty-nine percent of the 152 dispositive reports held that at least one of the national measures at issue was WTO-inconsistent.⁸³ Qualitative assessments of Appellate Body decision, such as those by Daniel Tarullo, have also shown a liberalizing bias.⁸⁴ This liberal bias is particularly apparent in cases that have increasingly reigned in the use of anti-dumping duties, countervailing measures, and safeguards measures, and -- more recently -- in successful challenges to the maintenance of agricultural protection and subsidization measures.⁸⁵

[*pg 275]

Third, judicial activism occurs in the shadow of power politics, both between nations and within nations. The Appellate Body is not ignoring the interests of the membership: it often acts to induce behaviors that are consistent with the interests of members but are politically difficult to implement at home.

III. UNDERSTANDING TRADE DELEGATION AND COMPLIANCE

Why do nations comply with a policy that moves trade in a liberal direction, even when inconvenient, given the absence of de jure delegation? In only about one out of twenty dispositive WTO cases has a losing party been held to have failed to comply within a "reasonable period."⁸⁶ This ninety-five percent approximate rate of compliance with judicial decisions would be respectable in most national legal systems. That it persists in an international legal context, which many describe as anarchic, may be surprising. This compliance rate is particularly impressive, given that compliance with WTO dispute-settlement decisions typically involves political defeat of historically powerful and intransigent protectionist sectors. Why does the WTO dispute-settlement system enjoy such a high compliance rate?

The essence of the mechanics by which the dispute-settlement system enjoys such a high compliance rate is rooted in decentralized enforcement -- a legitimized retaliatory threat that reconfigures domestic politics in the contravening country. At the interstate level of analysis, when ruled against, a WTO Member is expected to comply within a "reasonable period of time" (which may be determined by WTO binding arbitration).⁸⁷ If the member does not, the adversely affected complainant can retaliate. Retaliation takes the form of raising tariffs on goods originating in the territory of the contravening country to a level intended to have the effect of eliminating demand for the imports in proportion to the adverse effect of the contravening measures.

At the domestic politics level, the potency of this mechanism becomes clear. The dispute-settlement system provides national leaders with a way to get [*pg 276] around domestic pressures to keep the market closed. The potential of retaliation is an important means to motivate exporters who would be hurt by the threatened sanction. Since countries publish a list of products that will be affected by the sanction,⁸⁸ this has the effect of pitting those politically powerful, export-oriented producers against the industry that champions the contravening measures. If targeted smartly, the proposed retaliation list mobilizes sufficient political muscle within the contravening country to result in WTO-consistent reform. In this way, the DSU works in conjunction with domestic laws, regulations, and politics to foster compliance with the legislative outcomes that are codified in the WTO agreements.⁸⁹

Illustrating the range of political preferences within the United States suggests why compliance should be expected with most, but not all, WTO decisions.⁹⁰ Figure 1 below provides the traditional organization of trade-policy preferences in the United States. A couple of assumptions underlie this distribution of preferences. First, although both exporters and import-competing groups -- firms and labor -- influence Washington policy, the latter groups, which face a certain and substantial loss from liberalization, are more likely to get around collective action costs and to voice their preferences to Congress than the former. Thus, the preferences of Congress are aligned more closely with those of the latter group. Second, the size of the constituency has a significant effect on the trade-policy preferences for elected officials. Export-oriented producers are more likely than import-competing producers to be geographically dispersed; a high proportion of local production is consumed locally. By that logic, which has been demonstrated empirically, import-competing producers wield more influence over smaller electoral districts than do larger producers.⁹¹ Hence, the House is less free-trade oriented than is the Senate, and they are both more prone to trade closure than is the President. Finally, since all dispute-settlement bodies include non-Americans, the assumption is that, across cases in which the United States is a respondent, dispute-settlement panelist and Appellate Body preferences will be more pro-liberalization than those of any U.S. domestic actor, except export interests.

[*pg 277]

Figure 1

In the diagram above, the continuum ranges from no liberalization on the left to liberalization on the right. I represents the ideal point of import-competing groups, and SQ represents the status quo. The ideal positions on liberalization (whether through negotiation or other means) of a majority of the House, a majority of the Senate, and the President are represented by H, S and P, respectively. AB is the preference of the Appellate Body, and E represents export interests. AB and E have general preferences favoring trade liberalization, not preferences in any particular case or with respect to any particular industry. The win set for each actor is established as an arc, with each actor's ideal point equidistant from the status quo (on the left) and an increase of liberalization (on the right) beyond which the actor would prefer reversion to the status quo.

Given contemporary procedures for negotiated trade liberalization, such as those established under current Trade Promotion Authority, the President proposes a trade deal initialed in Geneva. That negotiated deal, however, must be acceptable to both the House and the Senate. A new policy will be acceptable to all if it falls within that small circle centered on the House. In this scenario, the President proposes a position at his ideal point and Congress agrees. The dot on the line above P represents this outcome.

Now consider the situation when a matter is litigated. Eventually, the Appellate Body makes a decision. As long as it rules within the preference space represented by the arrow, the United States should comply with the ruling. Why would the President not prefer a position at his ideal point and so reject the Appellate Body's ruling? In practice, if the President were to respond by proposing his ideal point, Congress would bargain with him. The outcome, which would be in the absence of agenda control by the Appellate Body, would likely be closer to the Congress's ideal than to the President's. In such a world, the President has an interest in giving the Appellate Body agenda control, and Congress will not object, as long as the outcome is within the win-set that centers on the House.

[*pg 278]

This model is suggested only as a simple illustration. Its central point is that it is not counterintuitive to have the U.S. President go along with liberalization efforts by the Appellate Body, and similarly, for Congress not to object to this judicial liberalization. Given this very rudimentary model of domestic preferences, such liberalization may be an outcome that everyone sees as superior to the status quo.

Five additional observations, none of which should be surprising, derive from this simple diagram. First, the President may be able to move outside the House's and Senate's respective win-sets and liberalize further than Congress would like, if compliance can be achieved by a regulatory change or sole executive action (such as an Executive Order). In those cases in which compliance can be achieved by liberalizing without an affirmative act of Congress, the President has considerable freedom to do what the Appellate Body prescribes. However, the President cannot accommodate the Appellate Body by liberalizing so much that he catalyzes veto-proof action by both the House and the Senate. Empirically, in many cases, the United States has complied (partly or fully) with an Appellate Body decision through sole executive action that arguably went further than the House (or Congress) would have preferred, including U.S. -- Shrimp-Turtle,⁹² U.S. -- Lamb Meat,⁹³ and United States -- Subsidies on Upland Cotton.⁹⁴

Second, the Appellate Body must engage in some compromise and accommodation to enjoy compliance, particularly when U.S. compliance requires an act of Congress. If Congressional action is required for compliance, then the United States will comply only if the prescribed action is within the House's win-set. Assuming that the Appellate Body prefers substantially more liberalization than the House, such that its ideal point resides outside the House's win-set, the Appellate Body will need to offer compromise opinions. Empirically, the Appellate Body has offered such politically sensitive reports in several cases. For example, in U.S. -- Shrimp-Turtle,⁹⁵ the Appellate Body could have limited the jurisdictional scope of GATT Article XX(g), which might have created conflicts between GATT and various multilateral environmental agreements, and would certainly have angered environmentalists and members of Congress, but it evaded the question. In several cases the Appellate Body has embraced the doctrine of judicial economy,⁹⁶ thereby limiting the scope of judicial lawmaking and enhancing the prospects for compliance.

Third, the model may be used to suggest longitudinal change. The dispute-settlement process can move policy toward liberalization by influencing the [*pg 279] number and type of groups on the left side of the continuum. Once liberalized and faced with competition, import-competing firms and jobs disappear. Once gone, the political landscape shifts. House, Senate, and Presidential preferences may shift to the right. In this sense, Appellate Body rulings can have direct effects on U.S. politics.

The existence of international delegation, however, does not imply an inevitable move toward openness. Competition may lead to a changing number of actors on either side of the continuum. With the growth of competition at home, more firms are likely to find themselves aligning on the left, stopping any attempt to open the U.S. market; likewise, the growth of export opportunities, or the threat of retaliation, will mobilize exporters on the right-hand side and could move policy in that direction. In the end, it is the preferences of U.S. industries that will move policy in one direction or the other, no matter the degree of delegation to the WTO dispute-settlement system.

Fourth, the model shows how some contemporary U.S. statutes have moved U.S. trade policy in a more open direction by shifting authority over trade matters from Congress to the executive branch. Prior to 1974, the implementing legislation for comprehensive trade agreements originated in the House (as required by the Constitution), then went to the Senate, then to the President for signature. As Figure 1 suggests, with the House acting as the agenda-setter, the Senate and then the President were presented with legislation that was at the more closed end of the spectrum. "Fast-track" negotiating authority in the Trade Act of 1974 and subsequent statutes, as well as the current Trade Promotion Authority, have provided the President with an opportunity to propose a trade agreement and implementing legislation to Congress, which could not amend the agreement or the legislation, requiring Congress to accept or reject the package by an up-or-down vote. By shifting agenda-setting authority to the President, these mechanisms have moved U.S. trade policy in a more open direction.⁹⁷

Finally, the model may roughly suggest the extent to which the Appellate Body may successfully push for trade openness when considering measures maintained by non-U.S. domestic political systems. For example, the model suggests that the Appellate Body may successfully push an authoritarian system further than the United States toward the trade-openness end of the spectrum because authoritarian rulers do not need to deal with an effective legislative branch with a more closed orientation. Similarly, the model may indicate that the Appellate Body faces the opposite situation when addressing EU trade measures: the Commission is up against multiple veto points, giving it and the Appellate Body relatively little leeway to push Europe toward greater openness. In the end, the success of this and all international courts may well be [*pg 280] determined as much by their willingness to act strategically as by the wisdom of their rulings in any particular case.

IV. CONCLUSIONS

Today, more than at any time in the last sixty years, liberalization is difficult and export interests do not necessarily hold sway over powerful private groups unwilling to face international competition. Partly because of powerful and recalcitrant protectionist sectors, the most optimistic expectation for the Doha Round is a minimalist package, modest by contrast to the results of the Uruguay Round. As of this writing, the Round has delivered even less. Whereas big bundled deals gained majority support in the United States and among other power members of the regime during the GATT years, judicial action in the WTO since then has come to be an increasingly salient and effective means of defeating recalcitrant protectionism. In trade rounds, U.S. steel, sugar, cotton, apparel, and other inefficient producers have succeeded for a half century in assuring their continued protection: facing certain, devastating losses from liberalization, they have remained united in successfully keeping their protection off the bargaining table, even if including their liberalization in U.S. offers could have helped U.S. export-oriented producers. Liberalization of these sectors has been stuck in the mud.

This outcome is no surprise. From the start, delegation of authority in the United States, first from Congress to the President and then from the President to a multilateral trade regime, has been a means to get around domestic resistance to trade reform. America's tolerance for de facto delegation to the WTO and judicial lawmaking continues to perform that function.

From the perspective of large democracies, the WTO dispute-settlement process is now a highly salient method to keep markets opening, given domestic resistance and the difficulties that now accompany multilateral negotiations. In broad terms, granting agenda-setting power to the Appellate Body forces leaders in member nations to compare the liberalizing stance of the Appellate Body in its reports with the status quo; if the Appellate Body has been strategic, its overall stance on liberalization will be closer to those of national leaders than the status quo. Mechanically, at the end of each individual dispute, the threat of retaliation usually pits one of the offending protectionist subsectors against a large number of export-oriented interests. If retaliation occurs, certain loss will befall the exporters, motivating them to support ending the protection under challenge. In the dispute-settlement context, the export-oriented producers, who in the legislative context merely had a possible gain, are now threatened with a clear and credible loss, motivating them to act decisively against protection. This asymmetry -- many exporters against a single import-competing group -- creates the political space that pushes liberalization forward. As long as the Appellate Body does not deviate too far from underlying interests in member countries, they can rule, countries will comply, and liberalization [*pg 281] moves forward -- albeit with a Greek chorus of politicians and the public complaining about judicial activism and a loss of sovereignty.

Furthermore, liberalization itself can be self-reinforcing: as groups are unable to maintain protection and cannot compete, they "peel" off and become less powerful domestic interests.⁹⁸ To be sure, the Doha Round negotiations have increasingly operated under the shadow of the law and politics of these liberalizing judicial decisions; in this way, litigation may be feeding back onto legislation. In any case, the "judicial liberalization" of WTO judicial action is currently a crucial piece of the contemporary WTO liberalization story.

This important role for the WTO dispute-settlement system is reminiscent of the crucial liberalizing role played by the ECJ in the 1960s through the late-1980s -- until the Single European Act.⁹⁹ In that period, the Council was paralyzed by the Luxemburg compromise,¹⁰⁰ which effectively required unanimity for any important action. The ECJ's exercise in "negative liberalization," striking down national protectionist measures in such famous cases as Rheinheitsgebot¹⁰¹ and Cassis de Dijon,¹⁰² is credited with being the main engine of internal market liberalization in the period. Moreover, as we have argued here about compliance with WTO Appellate Body decisions, domestic politics have been crucial to compliance with ECJ decisions.¹⁰³ Of course, the ECJ's role is distinguishable from that of the WTO Appellate Body in the contemporary period: perhaps most significantly, the ECJ established direct effect and unqualified supremacy of its decisions in member-state legal systems, whereas the Appellate Body has to rely exclusively on the political mechanisms described above for compliance.

As long as this WTO agent supports open trade, the agency relationship will lead nations to continue to open up markets. Powerful WTO members like the United States will abide by these liberal decisions to the extent that the decisions catalyze decisive action by efficient, export-oriented producers and [*pg 282] are favored by elites and leaders who still deem openness to be in the national interest.

FOOTNOTES

*	Professor of Political Science, Stanford University. ** Professor of Law, University of California at Los Angeles. The authors thank Karen Alter, Curtis Bradley, Andrew Guzman, Larry Helfer, Judith Kelley, Douglas Rivers, and participants in the Duke workshop on International Delegation for their many helpful comments on earlier drafts.
1.	"Judicial lawmaking" is defined as action by a judicial body that establishes new legal rules. Such action is usually taken pursuant to a court's perceived authority to interpret legal text. A court may take a restrained or expansive stance on interpretation, leading to variance in the extent of judicial lawmaking. As elaborated in Part II, below, judicial lawmaking at the WTO is done by the Appellate Body. WTO judicial lawmaking is treated here as having two dimensions: filling gaps in legal text and clarifying ambiguity.
2.	For a more detailed explanation of increasing stalemate in WTO trade rounds, see Judith L. Goldstein & Richard H. Steinberg, Regulatory Shift: The Rise of Judicial Liberalization at the WTO, in THE POLITICS OF GLOBAL REGULATION (Walter Mattli & Ngaire Woods eds., forthcoming 2008).
3.	Terry M. Moe, The New Economics of Organization, 28 AM. J. POL. SCI. 739, 756 (1984).
4.	Curtis A. Bradley & Judith Kelley, The Concept of International Delegation, 71 LAW & CONTEMP. PROBS. 1, 27 (Winter 2008) (quoting Kenneth W. Abbott & Duncan Snidal, Hard and Soft Law in International Governance, 54 INT'L ORG. 421, 436-37 (2000)).
5.	For more on the distinction between formal or de jure sovereignty and informal or de facto sovereignty, see Richard H. Steinberg, Who Is Sovereign?, 40 STAN. J. INT'L L. 329, 340 (2004).
6.	Final Act of the U.N. Conference on Trade and Employment, Havana Charter for an International Trade Organization, chs. VII-IX, U.N. Doc. E/CONF.2/78 (Mar. 24, 1948) [hereinafter Havana Charter].
7.	Protocol of Provisional Application of the General Agreement on Tariffs and Trade, Oct. 30, 1947, 61 Stat. A-2051, 55 U.N.T.S. 308 [hereinafter Protocol of Provisional Application].
8.	See generally Richard Blackhurst, The Capacity of the WTO to Fulfill Its Mandate, in THE WTO AS AN INTERNATIONAL ORGANIZATION 31 (Anne O. Krueger ed., 1998). In a comparative analysis of international organizations, Blackhurst concludes that the GATT/WTO has had a small and underfinanced secretariat. Id. at 41. He notes that the only agency with a smaller staff and budget among the seventeen most active international agencies is the United Nations Conference on Trade and Development. Id. at 40, tbl.1.1.
9.	Final Act Embodying the Results of the Uruguay Round of Multilateral Trade Negotiations, Apr. 15, 1994, 33 I.L.M. 1125 (1994) [hereinafter WTO Agreement].
10.	The secretariat remained under formal UN control until the WTO's creation. Thus, through the mid-1990s, the UN was the legal entity that could appoint the Director-General and set compensation.
11.	See generally JOHN H. JACKSON, WORLD TRADE AND THE LAW OF GATT 201-48 (1969) (discussing "The Legal Framework of Tariff Concessions and Negotiations in GATT").
12.	See generally World Trade Organization, Aid For Trade Global Review, http://www.wto.org/english/tratop_e/devel_e/a4t_e/global_review_e.htm (last visited Jan. 25, 2008) (describing the WTO's goals for monitoring and evaluation of its "Global Aid-for-Trade Review" program).
13.	See generally WTO Agreement, supra note 9, Annex 3 (describing the Trade Policy Review Mechanism).
14.	See John H. Jackson, The WTO Dispute Settlement Understanding-Misunderstandings on the Nature of Legal Obligation, 91 AM. J. INT'L L. 60, 62 (1997). See generally Improvements to the GATT Dispute Settlement Rules and Procedures, Apr. 13, 1989, GATT B.I.S.D. (36th Supp.), at 61 (1990) [hereinafter Midterm Dispute-Settlement Rules], available at http://www.wto.org/gatt_docs/English/SULPDF/91420188.pdf.
15.	See General Agreement on Tariffs and Trade arts. XXIII(2), XXV, Oct. 30, 1947, 61 Stat. A-11, 55 U.N.T.S. 194 [hereinafter GATT]; Richard H. Steinberg, In the Shadow of Law or Power? Consensus-Based Bargaining and Outcomes in the GATT/WTO, 56 INT'L ORG. 339, 344 (2002).
16.	ROBERT E. HUDEC, ENFORCING INTERNATIONAL TRADE LAW: THE EVOLUTION OF THE MODERN GATT LEGAL SYSTEM 11 (1993).
17.	Robert E. Hudec, The New WTO Dispute Settlement Procedure: An Overview of the First Three Years, 8 MINN. J. GLOBAL TRADE 1, 4-5 (1999).
18.	HUDEC, supra note 16, at 12.
19.	See id. at 13.
20.	See Hudec, supra note 17, at 21.
21.	HUDEC, supra note 16, at 13.
22.	See id. at 8.
23.	Id. at 14.
24.	Marc L. Busch & Eric Reinhardt, Testing International Trade Law: Empirical Studies of GATT/WTO Dispute Settlement, in THE POLITICAL ECONOMY OF INTERNATIONAL TRADE LAW 457, 473 (Daniel L. M. Kennedy & James D. Southwick eds., 2002).
25.	HUDEC, supra note 16, at 13.
26.	19 U.S.C. § 2411 (2000).
27.	Id. § 2411(a),(b).
28.	Id. § 2411(a)(1).
29.	For more details on this argument, see Judith Goldstein & Joanne Gowa, U.S. National Power and the Post-War Trading Regime, 1 WORLD TRADE REV. 153, 158-64 (2002).
30.	See generally Midterm Dispute-Settlement Rules, supra note 14.
31.	This description of the U.S. position in 1990 is documented in USTR, Dispute Settlement, in THE U.S. DELEGATION BRIEFING BOOK FOR THE BRUSSELS MINISTERIAL (Dec. 1990) (on file with authors).
32.	Id.
33.	"Legalization" is a particular form of institutionalization characterized by precision, obligation, and delegation. See generally Kenneth W. Abbott, Robert O. Keohane, Andrew Moravcsik, Anne-Marie Slaughter & Duncan Snidal, The Concept of Legalization, 54 INT'L. ORG. 401 (2000).
34.	See Steinberg, supra note 15, at 359-60.
35.	Understanding on Rules and Procedures Governing the Settlement of Disputes, Apr. 15, 1994, Marrakesh Agreement Establishing the World Trade Organization, Annex 2, Legal Instruments-Results of the Uruguay Round, 33 I.L.M. 1125, 1226 (1994) [hereinafter DSU], available at http://www.wto.org/english/tratop_e/dispu_e/dsu_e.htm (last visited Feb. 21, 2008).
36.	Busch and Reinhardt count 654 disputes filed from 1948 through June 2002. Busch & Reinhardt, supra note 24, at 462. From the date of the WTO's establishment through June 2000, 202 disputes were initiated. See World Trade Organization, Chronological List Of Disputes Cases, http://www.wto.org/english/tratop_e/dispu_e/dispu_status_e.htm (last visited Jan. 27, 2008). See also Kara Leitner & Simon Lester, WTO Dispute Settlement 1995-2003: A Statistical Analysis, 7 J. INT'L ECON. L. 169 (2004).
37.	Each of these 279 disputes is listed chronologically in World Trade Organization, supra note 36.
38.	DSU, supra note 35, art. 6(1).
39.	Id. arts. 16(4), 2(4).
40.	Id. art. 22(7).
41.	Id. art. 17(6).
42.	See Jackson, supra note 14, at 61-64.
43.	See Inge Govaere, The Reception of the WTO Agreement in the European Union: The Legacy of GATT, in REGIONALISM AND MULTILATERALISM AFTER THE URUGUAY ROUND 703, 707 (Paul Demaret, Jean-Francois Bellis & Gonzalo García-Jiménez eds., 1997) (suggesting that GATT/WTO law has almost no direct effect in the law of the European Union and its Member States); Richard H. Steinberg, Direct Application of Multilateral Trade Agreements in the United States, in REGIONALISM AND MULTILATERALISM AFTER THE URUGUAY ROUND 715, 723 (Paul Demaret, Jean-Francois Bellis & Gonzalo García-Jiménez eds., 1997) ("The GATT 1947 and the Uruguay Round Agreements, including the GATT 1994, have had only limited direct effects in US law.").
44.	See Richard H. Steinberg, Judicial Lawmaking at the WTO: Discursive, Constitutional, and Political Constraints, 98 AM. J. INT'L L. 247, 250 (2004) (arguing that politics is the tightest constraint on judicial lawmaking at the WTO, and that political constraint is unlikely to operate unless such lawmaking changes the fundamental balance of WTO rights and responsibilities favored by powerful states).
45.	See id. for a more detailed statement of the arguments in this section.
46.	See, e.g., Anne-Marie Burley & Walter Mattli, Europe Before the Court: A Political Theory of Legal Integration, 47 INT'L ORG. 41, 73-76 (1993); John Ferejohn & Barry Weingast, Limitation of Statutes: Strategic Statutory Interpretation, 80 GEO. L.J. 565, 581-82 (1992); Erik Voeten, The Politics of International Judicial Appointments: Evidence from the European Court of Human Rights, INT'L ORG. (forthcoming 2008).
47.	Interview with A. Jane Bradley, former chief U.S. dispute-settlement negotiator and Assistant USTR for Enforcement, in Washington, D.C. (Mar. 2007); Interview with Kenneth Freiberg, USTR Deputy Gen. Counsel, in Washington, D.C. (Mar. 2007). According to Bradley and Freiberg, a handful of lawyers in the USTR's General Counsel's office were concerned about judicial lawmaking, but those at the political level in both Washington and Brussels were persuaded by the clarity of the WTO agreements and by the mandate of DSU Articles 3.2 and 19.2 that neither panels nor the Appellate Body could add to nor diminish the rights and obligations provided in the covered agreements. Interviews with A. Jane Bradley & Kenneth Freiberg, supra. U.S. Senator Bob Dole was concerned enough about judicial lawmaking that he proposed establishing a commission to review the decisions and behavior of the Appellate Body, but only twelve co-sponsors joined him in support of the proposal, and it never passed the Senate. See generally WTO Dispute Settlement Review Commission Act, S. 16, 104th Cong. (1995) (establishing a WTO Dispute Settlement Review Commission to review dispute-settlement reports of the WTO's Appellate Body).
48.	Ultimately, the distinction between gap-filling and ambiguity clarification may be fragile, but the distinction is respected here out of convention. See generally H.L.A. HART, THE CONCEPT OF LAW (1961).
49.	Appellate Body Report, United States-Import Prohibition of Certain Shrimp and Shrimp Products, paras. 79-91, WT/DS58/AB/R (Oct. 12, 1998) [hereinafter U.S.-Shrimp-Turtle], available at http://docsonline.wto.org/DDFDocuments/t/WT/DS/58ABR.doc.
50.	DSU, supra note 35, art. 13(1).
51.	Appellate Body Report, European Communities-Regime for the Importation, Sale and Distribution of Bananas, paras. 9-10, WT/DS27/AB/R (Sept. 9, 1997).
52.	Id. at para. 10.
53.	See Panel Report, Indonesia-Certain Measures Affecting the Automobile Industry, para. 14.9, WT/DS54/R, WT/DS55/R, WT/DS59/R, WT/DS64/R (July 2, 1998), available at http://www.wto.org/English/tratop_e/dispu_e/54r00.pdf.
54.	See U.S.-Shrimp-Turtle, supra note 49, at paras. 20-28.
55.	Id. at para. 126 (quoting GATT, supra note 15, art. XX(g)).
56.	Id. at para. 129.
57.	Id. at paras. 150-86.
58.	Id. at para. 161.
59.	See Appellate Body Report, United States-Definitive Safeguard Measures on Imports of Circular Welded Carbon Quality Line Pipe from Korea, paras. 208-17, WT/DS202/AB/R (Feb. 15, 2002), available at http://www.wto.org/english/tratop_e/dispu_e/202abr_e.pdf; Appellate Body Report, United States-Safeguard Measures on Imports of Fresh, Chilled or Frozen Lamb Meat from New Zealand and Australia, paras. 166-69, WT/DS177/AB/R, WT/DS178/AB/R (May 1, 2001) [hereinafter U.S.-Lamb Meat], available at http://docsonline.wto.org/DDFDocuments/t/WT/DS/178ABR.doc; Appellate Body Report, United States-Definitive Safeguard Measures on Imports of Wheat Gluten from the European Communities, paras. 76-78, WT/DS166/AB/R (Dec. 22, 2000), available at http://www.wto.org/english/tratop_e/dispu_e/166abr_e.pdf.
60.	Interview with Tim Reif, Democratic Chief Trade Counsel, U.S. House Comm. on Ways and Means, in Washington, D.C. (Apr. 2002). In the Uruguay Round negotiations, U.S. negotiators refused to agree to a test that would require national authorities to quantify the relative effects of imports and other factors on domestic industry. In so refusing, the U.S. negotiators intended to enable the International Trade Commission to continue using its qualitative approach to analyze the "substantial cause" question in safeguards cases.
61.	U.S.-Lamb Meat, supra note 59, at paras. 105-06.
62.	See, e.g., Appellate Body Report, United States-Anti-Dumping Measures on Certain Hot-Rolled Steel Products from Japan, paras. 226-28, WT/DS184/AB/R (July 24, 2001), available at http://www.wto.org/english/tratop_e/dispu_e/184abr_e.pdf.
63.	Appellate Body Report, United States-Laws, Regulations and Methodology for Calculating Dumping Margins ("Zeroing"), para. 263, WT/DS294/AB/R (Apr. 18, 2006) [hereinafter U.S.-Zeroing], available at http://www.wto.org/english/tratop_e/dispu_e/294abr_e.pdf.
64.	Id. at para. 173 n.286.
65.	Id. at para. 263.
66.	See U.S. SECRETARY OF COMMERCE, EXECUTIVE BRANCH STRATEGY REGARDING WTO DISPUTE SETTLEMENT PANELS AND THE APPELLATE BODY: REPORT TO THE CONGRESS TRANSMITTED BY SECRETARY OF COMMERCE 7-10 (2002), available at http://www.ita.doc.gov/ReporttoCongress.pdf. See generally CLAUDE E. BARFIELD, FREE TRADE, SOVEREIGNTY, DEMOCRACY: THE FUTURE OF THE WORLD TRADE ORGANIZATION (2001); Daniel K. Tarullo, The Hidden Costs of International Dispute Settlement: WTO Review of Domestic Anti-Dumping Decisions, 34 LAW & POL'Y INT'L BUS. 109 (2002).
67.	See generally Alan O. Sykes, Protectionism as a "Safeguard": A Positive Analysis of the GATT "Escape Clause" with Normative Speculations, 58 U. CHI. L. REV. 255 (1991).
68.	U.S.-Lamb Meat, supra note 59, at para. 197. See also Appellate Body Report, Argentina-Safeguard Measures on Imports of Footwear, para. 151, WT/DS121/AB/R (Dec. 14, 1999), available at http://docsonline.wto.org/DDFDocuments/t/WT/DS/121ABR.doc.
69.	Report of the Intersessional Working Party on the Complaint of Czechoslovakia Concerning the Withdrawal by the United States of a Concession under the Terms of Article XIX, paras. 11-12, GATT/CP/106 (Mar. 27, 1951) (adopted Oct. 22, 1951), available at http://sul-derivatives.stanford.edu/derivative?CSNID=90310003&mediaType=application/pdf.
70.	Agreement on Safeguards, Apr. 15, 1994, Marrakesh Agreement Establishing the World Trade Organization, Annex 1A, 1869 U.N.T.S. 154 (1994), available at http://www.wto.org/english/docs_e/legal_e/25-safeg.pdf.
71.	Interview with Tim Reif, former USTR Associate General Counsel and member of the U.S. delegation that negotiated the Agreement on Safeguards, in D.C. (May 2005).
72.	U.S.-Lamb Meat, supra note 59, at paras. 69-72 (citing Appellate Body Report, Korea-Definitive Safeguard Measure on Imports of Certain Dairy Products, para. 85, WT/DS98/AB/R (Dec. 14, 1999), available at http://docsonline.wto.org/DDFDocuments/t/WT/DS/98ABR.doc.
73.	Judicial lawmaking was raised several times in the Uruguay Round dispute-settlement negotiations, but in all of those instances, participants were expressing a preference that prospective changes to the dispute-settlement system should not create, by constructive interpretation, obligations that were not established in the texts of GATT/WTO agreements. See, e.g., GATT Secretariat, Summary and Comparative Analysis of Proposals for Negotiations, MTN.GNG/NG13/W/14/Rev.2 (June 22, 1988), available at http://www.wto.org/gatt_docs/English/SULPDF/92050053.pdf; GATT Secretariat, Meeting of 25 June 1987, MTN.GNG/NG13/2 (July 15, 1987), available at http://www.wto.org/gatt_docs/English/SULPDF/92020218.pdf.
74.	See, e.g., JOHN CROOME, RESHAPING THE WORLD TRADING SYSTEM: A HISTORY OF THE URUGUAY ROUND (1995).
75.	Interviews with A. Jane Bradley & Kenneth Freiberg, supra note 47 (identifying the following reports: Report of the Panel, United States-Restrictions on Imports of Tuna, WT/DS29/R (June 16, 1994), reprinted in 33 I.L.M. 839 (1994), available at http://www.wto.org/gatt_docs/English/SULPDF/91790155.pdf; Report of the Panel, United States-Restrictions on Imports of Tuna, WT/DS21/R (Sept. 3, 1991), reprinted in GATT B.I.S.D. (39th Supp.) at 155 (1993), available at http://www.wto.org/gatt_docs/English/SULPDF/91530924.pdf; Report of the Panel, Canada-Measures Affecting the Sale of Gold Coins, L/5863, 1985 WL 291500 (Sept. 17, 1985) (unadopted) (Canada agreed with certain parts of the report and rescinded the tax but blocked adoption of the report because it disagreed with findings relating to the MFN principle); Report of the Panel, United States Tax Legislation (DISC), Nov. 12, 1976, GATT B.I.S.D. (23rd Supp.) at 98 (1977), available at http://www.wto.org/gatt_docs/English/SULPDF/90930121.pdf (the United States permitting its adoption only in conjunction with a Contracting Parties decision, the substance of which was negotiated between the EC and the United States, interpreting various points of GATT law)).
76.	See generally Negotiations on the Dispute Settlement Understanding, Proposal of the Africa Group in the WTO, para. 2 (Oct. 2002) (on file with authors); Special Session of the Dispute Settlement Body, Minutes of Meeting-Held in the Centre William Rappard on 17-18 February 2003, TN/DS/M/9 (July 1, 2003); Special Session of the Dispute Settlement Body, Minutes of Meeting-Held in the Centre William Rappard on 28-30 January 2003, TN/DS/M/8 (June 30, 2003); Special Session of the Dispute Settlement Body, Minutes of Meeting-Held in the Centre William Rappard on 16-18 December 2002, TN/DS/M/7 (June 26, 2003); Dispute Settlement Body-Special Session, Special Session of the Dispute Settlement Body-Report by the Chairman, Ambassador Péter Balás, to the Trade Negotiations Committee, TN/DS/9 (June 6, 2003); Special Session of the Dispute Settlement Body, Minutes of Meeting-Held in the Centre William Rappard on 13 - 15 November 2002, TN/DS/M/6 (Mar. 31 2003); Special Session of the Dispute Settlement Body, Minutes of Meeting-Held in the Centre William Rappard on 14 October 2002, TN/DS/M/5 (Feb. 27, 2003); Special Session of the Dispute Settlement Body, Minutes of Meeting-Held in the Centre William Rappard on 10 September 2002, TN/DS/M/4 (Nov. 6, 2002); Settlement Body-Special Session, Negotiations on the Dispute Settlement Understanding-Proposals on DSU by Cuba, Honduras, India, Malaysia, Pakistan, Sri Lanka, Tanzania and Zimbabwe, TN/DS/W/18 (Oct. 7, 2002); Special Session of the Dispute Settlement Body, Minutes of Meeting-Held in the Centre William Rappard on 15 July 2002, TN/DS/M/3 (Sept. 9, 2002); Dispute Settlement Body-Special Session, Contribution of the United States to the Improvement of the Dispute Settlement Understanding of the WTO Related to Transparency-Communication from the United States, TN/DS/W/13 (Aug. 22, 2002).
77.	Confidential notes on meetings of the informal DSU Reform Group from Oct. 22, 1998-Sept. 24, 1999, provided to the author by a senior trade official (Apr. 2002) (on file with authors).
78.	These articles were selected by searching the Lexis-Nexis database of "US & Canadian Law Reviews, Combined" on Feb. 7, 2004, for articles that mentioned "GATT" at least ten times and "dispute settlement" at least five times.
79.	Steve Charnovitz, Environmental Trade Sanctions and the GATT: An Analysis of the Pelly Amendment on Foreign Environmental Practices, 9 AM. U. J. INT'L. L. & POL'Y 751, 788 (1994); Daniel A. Farber & Robert E. Hudec, Free Trade and the Regulatory States: A GATT's-Eye View of the Dormant Commerce Clause, 47 VAND. L. REV. 1401, 1408-09 (1994).
80.	These figures are based on the author's review of articles selected by searching the Lexis-Nexis database of "US & Canadian Law Reviews, Combined" on Feb. 7, 2004, for articles that mentioned "WTO" at least ten times and "dispute settlement" at least five times.
81.	Interview with A. Jane Bradley, supra note 47; Interview with Dorothy Dwoskin, Assistant USTR for Multilateral Affairs, in Washington, D.C. (May 2005).
82.	See Tarullo, supra note 66, at 168. Complainants win almost ninety percent of all dispute-settlement cases. This win rate may be explained by a combination of asymmetric information and the cost of bringing a case, rather than by the political leanings of the tribunal. Nevertheless, this high win rate it is consistent with other evidence, presented above, suggesting a bias toward liberalization.
83.	See Busch & Reinhardt, supra note 24.
84.	See generally Tarullo, supra note 66.
85.	For example, a liberal bias has been apparent in WTO rulings against the EC's banana regime, beef hormones directive, subsidization of sugar, and restrictions on GMOs and against U.S. trade-remedy laws (particularly those used to protect steel) and cotton subsidies. See Panel Report, European Communities-Measures Affecting the Approval and Marketing of Biotech Products, WT/DS291/R, WT/DS292/R, WT/DS293/R (Sept. 29, 2006), available at http://www.wto.org/english/news_e/news06_e/291r_e.htm; Appellate Body Report, European Communities-Export Subsidies on Sugar, WT/DS265/AB/R, WT/DS266/AB/R, WT/DS283/AB/R (Apr. 28, 2005), available at http://docsonline.wto.org/DDFDocuments/t/WT/DS/283ABR.doc; Appellate Body Report, United States-Subsidies on Upland Cotton, WT/DS267/AB/R (Mar. 3, 2005), available at http://www.wto.org/english/tratop_e/dispu_e/267abr_e.pdf; Appellate Body Report, European Communities-Measures Concerning Meat and Meat Products (Hormones), WT/DS26/AB/R, WT/DS48/AB/R (Jan. 16, 1998); Appellate Body Report, European Communities-Regime for the Importation, Sale and Distribution of Bananas, WT/DS27/AB/R (Sept. 9, 1997).
86.	This figure was approximated as of Aug. 31, 2007, by dividing the total number of dispositive WTO dispute-settlement rulings holding a Member in contravention of a WTO obligation by the number of cases in which a WTO Member has been found in noncompliance with such a ruling pursuant to DSU Article 21. This calculation is an approximation because it is possible (although rare) that a Member has failed to comply but that the successful complainant was compensated by the contravening Member or otherwise decided not to pursue the matter under DSU Articles 21 and 22. For the list of information upon which this calculation rests, see World Trade Organization, supra note 36.
87.	DSU, supra note 35, art. 21(3).
88.	In the United States, the government publishes a proposed retaliation list in the Federal Register thirty days prior to the effective date of retaliation.
89.	See generally Judith Goldstein & Lisa L. Martin, Legalization, Trade Liberalization, and Domestic Politics: A Cautionary Note, 54 INT'L ORG. 603 (2000).
90.	This analysis is consistent with that presented by Epstein and O'Halloran in this volume. See David Epstein & Sharyn O'Halloran, Sovereignty and Delegation in International Organizations, 71 LAW & CONTEMP. PROBS. 77 (Winter 2008). However, we consider the interaction as only a two-party game-the United States interacting with the WTO. The interests of the WTO are assumed to be fixed and for free trade the preferences of the United States are as suggested in Figure 1, which could lead to a range of possible policy outcomes.
91.	See Michael Bailey, Judith Goldstein & Barry R. Weingast, The Institutional Roots of American Trade Policy: Politics, Coalitions, and International Trade, 49 WORLD POL. 309, 328-29 (1997); Michael Hiscox, The Magic Bullet? The RTAA, Institutional Reform, and Trade Liberalization, 53 INT'L ORG. 669, 687-89 (1999).
92.	U.S.-Shrimp-Turtle, supra note 49.
93.	U.S.-Lamb Meat, supra note 59.
94.	Appellate Body Report, United States-Subsidies on Upland Cotton, WT/DS267/AB/R (Mar. 3, 2005), available at http://www.wto.org/english/tratop_e/dispu_e/267abr_e.pdf.
95.	U.S.-Shrimp-Turtle, supra note 49.
96.	See, e.g., Appellate Body Report, United States-Measure Affecting Imports of Woven Wool Shirts and Blouses from India, § VI, WT/DS33/AB/R (Apr. 25, 1997), available at http://www.worldtradelaw.net/reports/wtoab/us-woolshirts(ab).pdf.
97.	See, e.g., Bailey, Goldstein & Weingast, supra note 91, at 314; Judith Goldstein, International Law and Domestic Institutions: Reconciling North American "Unfair" Trade Laws, 50 INT'L ORG. 541, 559-60 (1996).
98.	See Oona A. Hathaway, Positive Feedback: The Impact of Trade Liberalization on Industry Demands for Protection, 52 INT'L ORG. 575, 584-88 (1998), for a similar argument.
99.	See Single European Act, 1987 O.J. (L 169) 1.
100.	The Luxembourg Compromise, signed on Jan. 31, 1966, provides, In the event of decisions that can be adopted by majority on the proposal of the Commission, when very important interests of one or several partners are at stake, the members of the Council will attempt, within a reasonable period of time, to arrive at solutions that could be adopted by all members of the Council in respect of their mutual interests and those of the Community. This is considered a "political declaration" by Foreign Ministers and did not amend the Treaty of Rome. See Arrangements Made in Luxembourg Between the Foreign Affairs Ministers of the Six, Jan. 31, 1966, 5 I.L.M. 316, 317 [hereinafter Luxembourg Compromise].
101.	Case 178/84, Comm'n v. Germany, 1987 E.C.R. 1227, available at http://eur-lex.europa.eu/smartapi/cgi/sga_doc?smartapi!celexplus!prod!CELEXnumdoc&numdoc=61984J0178&lg=en (last visited Feb. 21, 2008)
102.	Case 120/78, Rewe-Zentral AG v. Bundesmonopolverwaltung für Branntwein, 1979 E.C.R. 649, available at http://eur-lex.europa.eu/smartapi/cgi/sga_doc?smartapi!celexplus!prod!CELEXnumdoc&lg=en&numdoc=61978J0120 (last visited Feb. 21, 2008).
103.	See Karen J. Alter, The European Union's Legal System and Domestic Policy: Spillover or Backlash?, 54 INT'L ORG. 489, 493 (2000); Burley & Mattli, supra note 46, at 62-65.