The establishment of the Appellate Body in the World Trade Organisation (WTO) is widely regarded as ‘[o]ne of the most important innovations of the WTO dispute settlement system’Footnote 1 that is ‘almost revolutionary’.Footnote 2 It is innovative in two ways. First is the addition of an appellate phase, which is ‘rather creative, given that such a stage is not commonplace in State-to-State dispute settlement mechanisms’.Footnote 3 The second innovation is that it is a standing body, unlike the ad hoc dispute settlement panels under the General Agreement on Tariffs and Trade (GATT) and even the WTO. As noted by Lacarte, this is ‘one of the most marked departures from previous GATT practice’.Footnote 4
As a new institution, however, the Appellate Body could not benefit from the respect accorded to long-established institutions, and must instead ‘earn’ its reputation by ‘everything that the Appellate Body said and did’.Footnote 5 While it has ‘earned’ its reputation ‘quickly in a short span of time’,Footnote 6 it is not free of controversy. As the result, the Appellate Body has been subject to many criticisms over the years. Some of these are minor technical issues such as the tone of the Appellate Body in its reports ‘that was widely perceived as excessively critical, if not derisive, of panel reports’.Footnote 7 Others are more controversial issues, such as opening the hearing to the publicFootnote 8 and the acceptance of amicus curiae briefs.Footnote 9 However, nothing in the history of the Appellate Body has been more serious than the current reappointment debate, which has been widely regarded as a ‘legitimacy crisis’ in the WTO.Footnote 10
Most of the discussions so far have only focused on whether the US has the right to block Appellate Body reappointments, or if Appellate Body members may be chastised for their statements in Appellate Body reports. This article approaches the debate over the US blockage from a fresh perspective by arguing that the US allegation of obiter dicta is invalid as there is no basis for such a claim in the WTO. Through these discussions, the paper provides important insights into the nature of the WTO dispute settlement system, the effects of panel and Appellate Body reports, and their respective functions in WTO dispute settlement.
1. The Appellate Body reappointment saga
On 11 May 2016, the United States (US) shocked the world by announcing that they will block the reappointment of Appellate Body member Prof. Seung Wha Chang. In a joint statement issued by Deputy US Trade Representative (USTR) Michael Punke and USTR General Counsel Tim Reif, the US declared that
The United States is strongly opposed to Appellate Body members deviating from their appropriate role by restricting the rights or expanding the obligations of WTO members under the WTO agreements… The United States will not support any individual with a record of restricting trade agreement rights or expanding trade agreement obligations.Footnote 11
At the DSB meeting held on 23 May, the US further clarified their position by stating that ‘we do not consider that his service reflects the role assigned to the Appellate Body by WTO Members in the WTO agreements’.Footnote 12 In particular, the US referred to four reports in which Prof. Chang allegedly ‘add[ed] to or diminish[ed] the rights and obligations provided in the covered agreements’.Footnote 13 In three of the four reports, the US accused Prof. Chang of addressing issues which were moot (Argentina–Financial Services), not appealed (India–Agricultural Products), or not raised by parties (US–Countervailing Measures (China)).Footnote 14 According to the US, these amounted to obiter dicta as they are not related to ‘issues necessary to resolve the dispute’.Footnote 15 As to the fourth report (US–Countervailing and Anti-Dumping Measures (China)), the US claimed that the Appellate Body has adopted ‘a very problematic and erroneous approach to reviewing a Member's domestic law’ by ‘substitut[ing] the judgment of WTO adjudicators for that of a Member's domestic legal system as to what is lawful under that Member's domestic law’.Footnote 16
The US blockage led to widespread criticisms from the WTO membership.Footnote 17 Prof. Chang's home country South Korea, in particular, claimed that the US opposition ‘would seriously undermine the independence and integrity of the Appellate Body’Footnote 18 and reportedly declared its opposition to the reappointment of any Appellate Body member.Footnote 19 At the DSB meetings where the issue was discussed, the US position received no support from other WTO Members, which called the US blockage ‘extraordinary, exceptional’,Footnote 20 and ‘unprecedented’.Footnote 21 At the dedicated DSB session on the issue on 26 October, Korea, supported by Brazil, India, and Mexico, tried to solve the problem by tabling a non-paper that proposed to limit future appointments to the Appellate Body to a single term.Footnote 22 As not all Members supported the proposal, however, it was not adopted.Footnote 23
There were also strong reactions from both current and former Appellate Body members. On 18 May, 2016, the six remaining Appellate Body members sent a letter to DSB Chairman Ambassador Xavier Carim. In the letter, they stressed that the Appellate Body reports are reports of the Appellate Body as a whole and ‘no case is the result of a decision by one Appellate Body Member, nor should interpretations or outcomes be attributed to a single Member’.Footnote 24 Moreover, they argued that ‘the tying of an Appellate Body Member's reappointment to interpretations in specific cases’ could undermine the trust of WTO Members ‘in the independence and impartiality of Appellate Body Members’.Footnote 25
On 31 May 2016, the 13 living former Appellate Body members also wrote another open letter to Carim, in which they warned that the US action would not only undermine the ‘impartiality and independence of the … Appellate Body’, but ‘put the very future of the entire WTO trading system at risk’.Footnote 26 In particular, they pointed out that the US approach is flawed, as ‘[a] decision on the reappointment of a Member of the Appellate Body should not be made on the basis of the decisions in which that Member has participated as a part of the divisions in particular appeals … Nor should either appointment or reappointment to the Appellate Body be determined on the basis of doctrinal preference.’Footnote 27
They proposed two solutions instead. First, to the extent that ‘WTO Members ever conclude that the Appellate Body has erred when clarifying a WTO obligation in WTO dispute settlement’, they should try to adopt a legal interpretation according to Article IX:2 of the Marrakesh Agreement.Footnote 28 Second, the Members can also abolish the current system of one four-year term with the possibility of a second four-year term and replace it with ‘a single, longer term for all Members of the Appellate Body’.Footnote 29
The reappointment saga also generated considerable interest in academic circles, with most scholars condemning the US action and expressing their support to Prof. Chang.Footnote 30 Most of the discussions so far revolved around the following issues: whether it is appropriate to attribute the views in an Appellate Body report to one division member only; whether a WTO Member has the right to block the reappointment of an Appellate Body member; should the decision to reappoint an Appellate Body member be influenced by his position on certain WTO law issues or preference for a certain judicial style? There are sharply divided views on all of these issues, with the US government saying yes to all, while other WTO Members, former and current Appellate Body members, and the academia mostly saying no.
Curiously, however, one key issue seems to be ignored in the debate. Everyone, no matter where he/she stands in this debate, seems to accept the US accusations that Prof. Chang's views in these decisions amount to obiter dicta, or at least it is possible to have obiter dicta in WTO panel and Appellate Body reports. As I will demonstrate in this article, the entire theory of obiter dictum in WTO dispute settlement reports is flawed as none of its three underlying premises is valid. Thus, in the end, the whole US theory of the possibility of an Appellate Body member violating its role by giving obiter dicta in Appellate Body reports is nothing but a dictum on dicta.
2. The concept of obiter dictum
To understand the US position, we first need to understand what is ‘dictum (plural dicta)’. Traditionally, a dictum is defined as ‘an expression of opinion in regard to some point or rule of law, made by a judge in the course of a judicial opinion, but not necessary to the determination of the case before the court’.Footnote 31 Over the years, alternative definitions have been suggested. These include, for example, ‘a legal conclusion stated in the opinion but not applicable to the particular facts of the case’, or ‘overgeneralization in light of the particular facts [of the case]’.Footnote 32 However, neither of these two variations adds anything new. If a legal conclusion is divorced from the facts of the case, then it is not necessary to the decision of the case. Similarly, overgeneralizations would also not be necessary for the decision. Thus, both of them can be submerged into the traditional definition.
2.1 Categories of dicta
Depending on the manner that the statement in question is pronounced, dictum can be divided into several categories such as obiter dictum, judicial dictum, gratis dictum, dictum proprium, and simplex dictum.Footnote 33 Many of the distinctions are now obsolete, and the most widely recognized distinctions are between the first two.Footnote 34 The key here is whether the statement in question is fully argued by the parties and deliberated upon by the court. Thus, if a point is neither argued by the parties nor fully deliberated by the court, it is regarded as obiter dictum.Footnote 35 In contrast, judicial dicta ‘are the product of a more comprehensive discussion of legal issues, and usually involve points briefed and argued by the parties’,Footnote 36 but they are ‘not essential to the decision’.Footnote 37 As judicial dicta is a more thoughtful opinion than obiter dicta, it is often accorded more persuasive authority, but it is still non-binding.Footnote 38 Nowadays, however, even this distinction is losing its significance, as many courts in major Common Law jurisdictions such as the US no longer follow the traditional distinction.Footnote 39
2.2 The effects of dicta
The traditional view is that, unlike the holding or the ratio decidendi of the court, a dictum is not binding.Footnote 40 As explained by Black, because dicta ‘are not the judicial determinations of the court, they are never entitled to the force and effect of precedents, in the same or other courts, and do not preclude the rendering of a subsequent contrary decision’.Footnote 41 However, this does not mean that a dictum has no effect at all. As observed by Black, dicta ‘though not precedents, may possess considerable value as persuasive arguments’.Footnote 42 Thus, lower courts in particular shall treat a dictum ‘with respectful consideration, not only because it proceeds from the appellate court, but also as furnishing a suggestion of the decision which that court might be expected to make if the point should come fairly before it for determination’.Footnote 43 In particular, ‘long repetition of a dictum … may clothe it with the weight of a precedent’.Footnote 44 This is especially the case for the dicta of the Supreme Court, which has been regarded by lower courts as ‘very persuasive’ and followed ‘slavishly’.Footnote 45
As we can see from the discussions above, while in theory a dictum is supposed to have no effect upon the development of the jurisprudence, the ultimate fate of a dictum depends upon the treatment, or the ‘attitudes’ of subsequent courts.Footnote 46 In many cases, various dicta have been picked up by later courts and elevated to ‘a position hardly distinguishable from that of a direct adjudication’.Footnote 47 Thus, some scholars have cast doubts on the utility of the formalistic categorization between holdings and dicta.Footnote 48 Instead, they suggested that the effect of dicta should be approached with a more pragmatic method. The best attempt is the functional approach by McAllister, which divides dicta into three categories: vibrant dicta, where ‘the otherwise non-binding judicial pronouncement promptly flowers into law … either by the court that issues the dicta, or by the accumulation of rulings from other courts’;Footnote 49 ‘dead dicta’, i.e., dicta that die either ‘through the issuing court's explicit pronouncements’ or ‘implicitly through a series of unfavourable rulings’;Footnote 50 and ‘divergent dicta’, which, as its name suggests, receives divergent treatments from subsequent courts due to their disagreements over its persuasiveness and effect.Footnote 51
2.3 Why the concept?
The concept of dicta, along with the distinction with holdings, is unique and ‘essential’Footnote 52 to the Common Law system. According to the doctrine of stare decisis Footnote 53 or precedents, later courts are supposed to follow the holdings by earlier courts.Footnote 54 Without the concept of dicta, everything stated by the earlier courts would be binding on the later courts. This might not be a problem if the court always restricts itself to what is absolutely necessary for the resolution of the case at hand, but this is rarely the case. Instead, as has been observed by Llewellyn and Aldisert, judges tend to ‘over-state’ their case ‘in the heat of the argument’ and ‘overwrite opinions’,Footnote 55 with the result that ‘discussion outran the decision’.Footnote 56 Or worse still, judges may deliberately ‘plant’ dicta to steer the development of the law and ‘preempt colleagues who might later decide a further issue in a manner not to their liking’.Footnote 57 These concerns make it necessary to draw the distinction.
How to draw the distinction then? There are two possible approaches. The first approach is looking at the substantive merits of the respective statements. The most trustworthy statements are those made by the courts when ‘they are aware of the relevant facts, and the possible competing legal positions have been argued at length by lawyers’.Footnote 58 In contrast, the courts’ ‘passing comments on peripherally related legal subjects’ are much less reliable as they ‘are expressions of opinions … which may not have been argued at the bar or duly brought to the attention of the court, or that they do not embody the mature and deliberate opinion of the judges’.Footnote 59 While this approach provides a satisfactory solution to most cases, it would not be able to distinguish judicial dictum from holding as the former also benefits from arguments in court and the full consideration of the court. This is why we need a second approach that also takes into account the raison d'être for the judicial law-making power. The core function of the court is to resolve disputes, thus the law-making power is only a by-product of such a function. Therefore, ‘the authority of a particular court should extend only to what is needed to resolve the dispute’, lest the court, as the ‘occasional legislators’,Footnote 60 usurp the legislative function of the ‘fulltime’ legislature, which is the legislative branch proper.Footnote 61 This means that the law-making power only extends to those statements necessary to resolve the dispute at hand, but not the unnecessary sound bites.
3. Obiter dictum in WTO law?
As discussed above, dictum is a concept unique to Common Law. Thus, the argument that the concept also exists in WTO law presumes that the WTO law system is modelled after the Common Law system. Or, short of that, that there is a system of precedents in WTO law. Or alternatively, that the WTO has rules against dictum. However, as I will elaborate below, none of these three assumptions is valid.
3.1 WTO follows common law
As I have demonstrated in the last part, the concept of dicta is used to distinguish the unnecessary parts of a decision from the key holding, or the ratio decidendi of the case. Ratio, in turn, forms the main body of precedents, or judge-made law, which is the hallmark of the Common Law system.
In contrast, neither judicial precedent nor judicial law-making is recognized under the Civil Law system. In France, for example, ‘[t]he central conviction … is that judges cannot be lawmakers’.Footnote 62 Instead, ‘[l]aw-making … was entirely reserved for the legislature’.Footnote 63 As pointed out by many comparative law scholars, such aversion to judicial law-making is a legacy of the French revolution, which viewed the judiciary as an enemy due to their excessive power under the ancien regime and sought to reign in the courts with the monopoly of legislative power by the legislature.Footnote 64 Such deep distrust of the judiciary even leads to explicit codifications of the prohibition on judicial law-making. The most well-known examples are ‘[t]he courts may not directly or indirectly take any part in the exercise of the legislative power’ under Article 10 of the Code de l'organisation judiciare (Code of Judicial Organization)Footnote 65 and ‘[i]t is forbidden for judges to make pronouncements by means of general and regulatory provisions on the cases submitted to them’ under Articles 5 of the Code Civil.Footnote 66 While there have been references to the jurisprudence or the jurisprudence constante of the Cour de cassation (Court of Cassation), the fact remains that ‘the Cour de cassation's jurisprudence is not officially binding on the lower Courts’ and ‘French lower courts are always free to depart from the Cour de cassation's jurisprudence’.Footnote 67 Indeed, as noted by de S.-O.-I'E. Lasser, the French Cour de cassation is rather annoyed with the lower court's reference to its ‘so-called jurisprudence’, and jurisprudence of the Cour cannot be cited by lower courts as the legal basis for overturning a decision.Footnote 68
Similarly, in Germany, there has been a long tradition of following Justinian's maxim ‘non exemplis sed legibus iudicandum est’ (decisions should be based on legislations, not on precedents).Footnote 69 While there has been increasing reference to court decisions after the Second World War,Footnote 70 there is still ‘no more inclination than before to urge a strict theory of precedent on German courts’.Footnote 71 Indeed, attempts to make the decisions of the Constitutional Court binding were explicitly rejected by the Great Senate for Civil Matters, which noted that making the decisions of the court binding on all organs of the government ‘would be to elevate a court into a supreme law-making body not chosen by the people, to eliminate the safeguarding formalities essential to law-making in a Rechtstaat, to destroy the separation of governmental powers and the federal system, and to borrow from the Anglo-Saxon legal world conceptions of judicial power that are alien to Germany’.Footnote 72
As I will demonstrate in the next two sub-sections, the WTO rules do not recognize either a binding precedent system or judicial law-making power. This clearly rules out the possibility that the WTO is based on the Common Law model.Footnote 73 Instead, if one has to choose one between the two systems, the Civil Law system has a much stronger claim for influencing the WTO legal system. There are three reasons for this.
First, the main source of law in Common Law is judicial decisions or judge-made law, while legislations reign supreme in Civil Law countries.Footnote 74 Similarly, the paramount role of formal legislations in the WTO has been repeated ad nauseam. In the WTO parlance, these legislations are known as ‘covered agreements’, i.e., agreements listed in Appendix 1 of the Dispute Settlement Understanding (DSU).Footnote 75 In the DSU, the phrase ‘covered agreement’, in either singular or plural forms, appeared 72 times in the main text and seven times in the Appendixes and footnotes. The key provision is Art. 3.2, which emphasizes that the purpose of the dispute settlement system is ‘to preserve the rights and obligations of Members under the covered agreements’. It also warns that the DSB ‘cannot add to or diminish the rights and obligations provided in the covered agreements’ in its recommendations and rulings, which is repeated verbatim in Art. 19.1. Similarly, according to Arts. 7.1 and 11, the function of the panel is confined to examine the consistency of the challenged trade measure with the relevant provisions in the ‘covered agreements’. Such slavish reliance on legislations can only be found in the Civil Law system.
Second, as the legislature monopolizes the law-making power in Civil Law countries, the role of the judiciary is reduced to that of a technicalFootnote 76 or even grammaticalFootnote 77 interpreter within rigid parameters.Footnote 78 This approach is grounded in the idea that the Codes provide a complete and perfect set of legal text that can encompass ‘all cases that life could possibly offer’Footnote 79 and judges are ‘merely applying pre-existing rules – the rules laid down in the code’.Footnote 80 Again, such an approach is adopted by the WTO legal system, which does not recognize any source of law other than the sacred ‘covered agreements’. The role of a WTO panel, according to Art. 11.1 of the DSU, is to make ‘an objective assessment of the facts of the case and the applicability of and conformity with the relevant covered agreements’. The wordings suggest that all that a WTO panel need to do is to mechanically apply the covered agreements and then determine the conformity of the challenged measure accordingly. Indeed, it could even be argued that, strictly speaking, the WTO panel and Appellate Body do not even have the power to ‘interpret’ the covered agreements. Instead, according to Art. 3.2 of the DSU, they only have the power to ‘clarify the existing provisions of those agreements in accordance with customary rules of interpretation of public international law’. On the other hand, one could argue that such approach is naive and unworkable and may even point to the reference to ‘legal interpretations developed by the panel’ under Art. 17.6 as an implicit acquiescence of the interpretive power of the panel. But any faith one might place on such implicit interpretive power must be severely shattered in the face of the explicit warning under Art. 3.9 that ‘[t]he provisions of [the DSU] are without prejudice to the rights of Members to seek authoritative interpretation of provisions of a covered agreement through decision-making under the WTO Agreement’. In other words, here the WTO legal system is again taking the traditional Civil Law approach. One could even say that it is much stricter than that of most modern Civil Law countries, and only falls an inch short of the explicit ban on the interpretive power of judges by the Roman emperor Iustinianus in his Codex.Footnote 81
Third, another key difference between Common and Civil law systems is whether the judicial decisions are made on a collective or individual basis. In Civil Law jurisdictions, ‘judicial decision is rendered by the entire court as a unit’ with judges remain ‘anonymous’.Footnote 82 Individual opinions such as ‘[d]issenting and concurring opinions are forbidden’Footnote 83 or at least discouraged.Footnote 84 In contrast, a defining feature of Common Law courts decisions is the ‘personalisation’ of views,Footnote 85 with each judge given the freedom to expound on his own point of view.Footnote 86 Indeed, it could even be said that a Common Law judgment is ‘the sum of the decisions of the individual judges’.Footnote 87 Again, between the two, the WTO legal system bears more resemblance to the Civil Law approach for the following reasons:
1. Under the DSU, strictly speaking, it is not the panel or the Appellate Body which decides individual disputes. Their role is limited to making recommendations to the DSB,Footnote 88 which is the WTO Membership acting on a collective basis.Footnote 89 As argued by Debra Steger, ‘[t]he Appellate Body only has jurisdiction for a particular appeal once a notice of appeal has been submitted to the DSB, and that jurisdiction is lost once its report has been circulated and adopted by the DSB. It does not have any continuing jurisdiction outside of these periods during particular appeals. There is no true separation of powers in the WTO. The DSB (a political body) governs the dispute settlement system: it decides to establish panels, adopt panel and Appellate Body reports (which have no legal status until they are ‘blessed’ by the DSB) and authorize suspension of concessions’.Footnote 90 In other words, ‘[i]t is the DSB that makes decisions, and the role of the Appellate Body is to advise the DSB on what to do’.Footnote 91 Thus, ‘their status is clearly subsidiary to that of the Dispute Settlement Body’.Footnote 92 Such approach could not be further from the individualist approach in the Common Law system.
2. In the DSU, reference to the panel or Appellate Body decision in a given case always refers to ‘the report’, implying that it is the decision by the panel or Appellate Body as a whole rather than the sum of individual opinions. The Working Procedure of the Appellate Body made this explicit, by stating that the Appellate Body shall ‘make every effort to take their decisions by consensus’Footnote 93 as the appellate process is a collegial process that is designed to ‘ensure consistency and coherence in decision-making, and to draw on the individual and collective expertise of the Members’.Footnote 94 One may argue that this is not the case, as Art. 17.1 of the DSU states that only three out of seven Appellate Body members shall ‘serve on any one case’ as a Division. However, one should not mistake this to mean that the other four Members play no role as there is the practice of ‘Exchange of Views’,Footnote 95 whereby ‘the division responsible for deciding each appeal shall exchange views with the other Members before the division finalizes the appellate report for circulation to the WTO Members’.Footnote 96 To ensure that even members who are not part of a Division make meaningful contributions, the Working Procedures also explicitly state that ‘each Member shall receive all documents filed in an appeal’. As explained by former Appellate Body Secretariat Director Valerie Hughes and former Appellate Body Chairman Claus-Dieter Ehlerman, such exchange of views is not merely rubberstamping the decision of the Division but has been ‘of enormous benefit to the work of the Appellate Body’ by allowing Divisions ‘to draw on the individual and collective expertise of all members’.Footnote 97 This is confirmed by the first Appellate Body Secretariat Director Debra Steger, who noted that ‘[i]n one particular early appeal, the exchange of views took five days, including two days during which the Appellate Body members listened with tremendous respect to a member who was not part of the division for that particular case as he tried several different ways to convince the division of his point of view’.Footnote 98
3. Under the DSU, there is no explicit prohibition of dissenting or individual opinions as in the European Court of Justice.Footnote 99 The only implicit reference to dissent can be found in Rule 3(2) of the Working Procedure, which states that the Appellate Body shall ‘make every effort to take their decisions by consensus’.Footnote 100 While there have been calls to allow dissenting opinions by some scholars,Footnote 101 they remain extremely rare in both panel and Appellate Body reports.Footnote 102 In practice, as the DSB always adopts the panel or Appellate Body report as a whole, the Common Law approach of allowing individual and sometimes conflicting opinions could create difficulties for WTO Members. If a report with dissent is adopted, does it mean that the dissenting opinion is accepted by the WTO Membership as well? Thus, it seems safer to follow the Civil Law tradition of not allowing dissents, or at least not to encourage them. Steger provided some hint into the origin of the Appellate Body's aversion of dissent when she noted that one of the reasons for the lack of dissents is because ‘some of [the Appellate Body members] emphasises that in their legal systems dissents were not common’. Apparently, these members must be from Civil Law countries.
4. The fourth hint for the Civil Law influence can be found under Arts. 14.3 and 17.11, which mandate that opinions expressed in the panel or Appellate Body report by individual panellists or Appellate Body members ‘shall be anonymous’. Instead, the reports are issued by the ‘faceless foreign judges’Footnote 103 and nobody is supposed to know who authored particular parts of the decision. This again is a hallmark feature of the Civil Law system.
3.2 WTO follows stare decisis
As I have demonstrated above, the view that the WTO follows the Common Law model is a fallacy. However, is it possible that the WTO, while not adopting the Common Law system on a wholesale basis, still follows the rule of stare decisis or binding precedents? Again, the answer has to be no.
At the outset, we should recall, as John Jackson has pointed out, ‘the international legal system does not embrace the common law jurisprudence … which calls for courts to operate under a stricter ‘precedent’ or ‘stare decisis’ rule’.Footnote 104 Thus, it is no surprise that most international tribunals do not follow the rule of stare decisis.Footnote 105 Some courts explicitly reject the idea. For example, the Statute of the International Court of Justice made it very clear that ‘[t]he decision of the Court has no binding force except between the parties and in respect of that particular case’.Footnote 106 While other tribunals do not have such explicit language in their constituting documents, they usually do not recognize the binding authority of previous decisions.Footnote 107 There have been some suggestions that the International Criminal Court is different in this regard, as Art. 21.2 of the Rome Statute states that ‘[t]he court may apply principles and rules of law as interpreted in its previous decisions’.Footnote 108 In my view, however, this is far from acceptance of the doctrine of stare decisis, as it merely uses the permissive language ‘may’, which still falls far short of granting binding force to precedents.
Similarly, the concept of precedent is also far from uncontroversial in the multilateral trading system. During the GATT era, the Contracting Parties took differing views on the issue. The European Economic Community (EEC), for example, argued that panel findings shall be ‘limited to the specific measures under examination’ and should not have precedential effect.Footnote 109 The US, on the other hand, argued that ‘when the Council adopted a report, those interpretations became GATT law’.Footnote 110 Moreover, even GATT panels themselves have not recognized the precedential effect of previous panel reports. For example, in the 1989 case of EEC – Restrictions on Imports of Dessert Apples – Complaint by Chile, the panel refused to follow the 1980 panel report on EEC – Restrictions on Imports of Apples from Chile,Footnote 111 even though it involved ‘the same product and the same parties as the present matter and a similar set of GATT issues’.Footnote 112
When the WTO came into being, the Ministerial Conference and the General Council was bestowed exclusive authority to adopt interpretations of the covered agreements.Footnote 113 With such explicit grant of interpretive power, it is not unreasonable to infer that such authorities cannot be exercised by other institutions.Footnote 114 This in turn means that, in principle, the legal interpretations adopted by the panel and Appellate Body do not have precedential power. Notwithstanding this, many commentators have argued that stare decisis does existFootnote 115 and WTO Appellate Body reports do have ‘precedential value’.Footnote 116 In the paragraphs that follow, I will investigate the validity of this claim with a detailed survey of the key WTO cases.
The first WTO case to address the precedential effect of panel reports is the 1996 case Japan–Alcoholic Beverages II. In its report, the panel stated that ‘panel reports adopted by the GATT CONTRACTING PARTIES and the WTO Dispute Settlement Body constitute subsequent practice in a specific case by virtue of the decision to adopt them. Article 1(b)(iv) of GATT 1994 provides institutional recognition that adopted panel reports constitute subsequent practice. Such reports are an integral part of GATT 1994, since they constitute “other decisions of the CONTRACTING PARTIES to GATT 1947”.Footnote 117 This view was rejected by the Appellate Body, which noted that, first, under GATT 1947, adopted panel reports only bind ‘the parties to the dispute in that particular case’, but not subsequent panels; second, only ‘the conclusions and recommendations in an adopted panel report’ are binding, but not the ‘legal reasonings’ in the report.Footnote 118 Citing the grant of exclusive authority to adopt interpretations by the Ministerial Conference and General Council under Article IX:2 of the WTO Agreement, the Appellate Body also held that panel reports would not ‘constitute a definitive interpretation of the relevant provisions of [covered agreements]’.Footnote 119 At the same time, the Appellate Body also noted that ‘[a]dopted panel reports are an important part of the GATT acquis … often considered by subsequent panels’, which ‘create legitimate expectations among WTO Members, and, therefore, should be taken into account where they are relevant to any dispute’.Footnote 120 Even unadopted panel reports could provide ‘useful guidance’ to future panels.Footnote 121 However, to prevent any illusion on the binding effect of panel reports, the Appellate Body also made it explicit that ‘they are not binding, except with respect to resolving the particular dispute between the parties to that dispute’.Footnote 122
In a way, it is not surprising that the Appellate Body took a cautious approach on the precedential value of panel reports in this case. The Appellate Body was barely one year old when the case was decided, thus it was better to avoid controversial statements so as not to undermine its own legitimacy as a new institution. Moreover, the Appellate Body did not address the precedential effects of its own reports, a question that is only answered in the subsequent case of US–Shrimp (Article 21.5 Malaysia). In that case, the Appellate Body expanded its approach in Japan–Alcoholic Beverages II in two very important ways. First, the Appellate Body confirmed that Appellate Body reports, just like panel reports, ‘provided interpretative guidance’ for panels.Footnote 123 This is not very surprising, because it is only natural that the Appellate Body, as the institution reviewing panel decisions, would have at least the same power as the panel. Second, in addition to confirming that ‘[t]he Panel was correct in using [the Appellate Body's] findings as a tool for its own reasoning’,Footnote 124 the Appellate Body went one step further by stating that the panel ‘was right to use’ and ‘rely on’ the ‘reasoning’ of the Appellate Body report in US–Shrimp.Footnote 125 This is one big step towards recognizing the precedential effect of Appellate Body reports, as the key in a precedent is its ratio decidendi or reasoning. Furthermore, to dispel any speculation that the reasoning in the Appellate Body report in US–Shrimp applied to the current case only because the two cases concern the same dispute on the same measure between the same parties, the Appellate Body also made clear that such reasoning shall be relied on by not only ‘the Panel in this case’, but also all ‘future panels’.Footnote 126
Some ‘future panels’, however, chose to ignore the Appellate Body's edict, resulting in a tug of war between the panel and the Appellate Body. The most contentious battle is fought over the legality of ‘zeroing’ practices by the US, where some WTO panels persistently refused to follow settled Appellate Body jurisprudence on the issue. In the US–Stainless Steel (Mexico) case, for example, the panel refused to follow previous Appellate Body decisions even though it was aware that its reasoning is very similar to those of the two panel decisions that have been overruled by the Appellate Body.Footnote 127 According to the panel, such an approach is mandated by Article 11 of the DSU, which requires panels to carry out an objective examination of the matter at issue.Footnote 128 The EU, one of the third parties in the case, became so frustrated that it asked the Appellate Body ‘to unambiguously re-confirm that all panels are expected and therefore also obliged, to follow its rulings on these issues’ (emphasis original).Footnote 129 Their frustration is shared by the Appellate Body, which stated that they ‘are deeply concerned about the panel's decision to depart from well-established Appellate Body jurisprudence clarifying the interpretation of the same legal issues’.Footnote 130 Citing the need to ensure ‘security and predictability’ in the dispute settlement system in Article 3.2 of the DSU, the Appellate Body held that ‘[w]hile the application of a provision may be regarded as confined to the context in which it takes place, the relevance of clarification contained in adopted Appellate Body reports is not limited to the application of a particular provision in a specific case’.Footnote 131 Thus, the Appellate Body concluded, ‘absent cogent reasons, an adjudicatory body will resolve the same legal question in the same way in a subsequent case’.Footnote 132
Does this statement mean that Appellate Body reports now shall be treated as binding precedents? While WTO Members differ widely on this issue,Footnote 133 the strong wordings of the Appellate Body certainly provided plenty of ammunition for the claim that the WTO now has a system of binding precedents, or stare decisis.Footnote 134 However, I think that such exuberance about the existence of a precedent system in the WTO dispute settlement system is not only irrational but also premature, as the Appellate Body itself explicitly stated, at the beginning of its discussion on the issue, that ‘[i]t is well settled that Appellate Body reports are not binding, except with respect to resolving the particular dispute between the parties’.Footnote 135 Compared to the highly cautionary language used earlier, the Appellate Body has made it very clear, in a straightforward and unequivocal manner, that there is no formal or de jure system of precedents in WTO dispute settlement.Footnote 136
Therefore, at most, one can only claim the existence of a de facto precedent system in the WTO, but ‘it is certainly not stare decisis’,Footnote 137 as pointed out by John Jackson, who argued that
[the] precedent effect in the jurisprudence of the WTO …is not so powerful as to require panels or the Appellate Body considering new cases to follow prior cases, with the possible exception that once prior cases have been numerous regarding a particular issue and approach, and apparently accepted by all members of the WTO, then the language of the Vienna Convention about ‘practice under the agreement’, may suggest a stronger impact. But short of that situation, it appears that the ‘flavor’ of the precedent effect in the WTO is still somewhat fluid, and possibly will remain somewhat fluid for the time being.
To paraphrase Jackson, much of the confusion regarding the precedential effect of the panel and Appellate Body reports arose because the word ‘precedent’ is a ‘complex concept’ with ‘many flavors’.Footnote 138 To avoid this, Jackson proposed to view the word as ‘a multi-layered concept, or at least as having a number of different approaches of different flavors’.Footnote 139 However, as we can see from the foregoing discussion, such an approach could still lead to confusions. Instead, I would suggest ceasing to refer to the previous decisions of the Appellate Body as precedents, but to call them as ‘jurisprudence’ instead.
Moreover, as Beshkar and Chilton have argued, conferring binding force on Appellate Body reports could raise substantive systematic costs.Footnote 140 For example, WTO Members might rush to bring cases, or at least participate as third parties, in a bid to shape the jurisprudence through litigation.Footnote 141 Wrong judicial precedents might perpetuate over time as the consensus requirement makes it difficult for the legislative branch to correct them.Footnote 142 All these will be unfair for the small and poorer countries as they are less likely to participate in WTO disputes.Footnote 143
In addition to these practical reasons, I would add another very important constitutional reason. The Appellate Body was set up as a ‘safety valve’Footnote 144 to check against ‘rogue’ panelsFootnote 145 which might render ‘bad reports’.Footnote 146 It was never meant to be a judicial branch that is on par with the legislative branch to safeguard the so-called separation of powers as under some domestic legal systems. Elevating Appellate Body reports to the status of binding precedents could seriously undermine the nature of the WTO as a ‘Member-driven’ organization.Footnote 147
3.3 WTO has rules against dicta
Even if the WTO does not have a system of precedent, could it still have rules against dicta? This is more than pure academic speculation, as many WTO Members, especially the US, has repeatedly referred to certain parts of panel reports as ‘dicta’ in their written submissions. This point, however, is highly contestable as what a WTO Member might view as ‘obiter dictum’ may often be a necessary link in the panel's overall analysis leading to the final findings. For example, in a case involving the non-discrimination obligations, the panel would have to first determine if the two products are alike before deciding whether the measure at issue is indeed discriminatory. If the panel makes a negative finding on likeness, this does not mean that the panel should stop its analysis there, because such a finding might be overturned on appeal. Thus, a more prudent course of action for the panel would be to continue making findings on the discrimination issue, lest the Appellate Body do not have sufficient facts to ‘complete the analysis’ when the likeness finding is reversed.Footnote 148
Compared to the allegations from WTO Members, what is even more worrying is the usage of the concept of ‘dicta’ by the Appellate Body itself in its own reports. What does the Appellate Body mean by ‘dicta’ then? Again this can only be found out from the Appellate Body's own words.
The very first case where the Appellate Body mentioned dicta is the Canada–Periodicals case, in which the Appellate Body held that, the statement by the panel in EEC–Oilseeds Footnote 149 that ‘it can reasonably be assumed that a payment not made directly to producers is not made “exclusively” to them’ is ‘obiter dicta’ because the panel already found that subsidies paid to oilseeds processors were not made ‘exclusively to domestic producers’.Footnote 150 However, the Appellate Body does not explain further why such statement is considered dicta. We can only surmise that the statement is regarded as dicta because it is about a moot issue.
In the US–Shrimp (Article 21.5 – Malaysia) case, the Appellate Body told us what is not dicta by stating that ‘[t]he reasoning in our Report in United States–Shrimp on which the panel relied was not dicta; it was essential to our ruling’.Footnote 151 This suggests that dicta is something that is not essential to the ruling of the Appellate Body.
In the US–Gambling case, the Appellate Body visited the issue again, when the parties debated whether the panel's statement on whether ‘practice’ as such may be challenged as a ‘measure’.Footnote 152 The Appellate Body ruled that, as Antigua, the Complainant, was not challenging a practice as such, the panel's statement did not constitute a ‘“finding” of the Panel’. Thus, the Appellate Body concluded, ‘the Panel's statement on “practice”, in our view, was a mere obiter dictum, and we need not rule on it’.Footnote 153 Ironically, however, the Appellate Body followed this statement with yet another dictum on dicta, by stating that
We nevertheless express our disagreement with the Panel's understanding of previous Appellate Body decisions. The Appellate Body has not, to date, pronounced upon the issue of whether ‘practice’ may be challenged, as such, as a ‘measure’ in WTO dispute settlement.Footnote 154
From these three cases, we can see that the Appellate Body's main criteria for distinguishing dicta from holding are whether the statement at issue is relevant or essential to the decision. However, as with any other legal issue, the Appellate Body's position here must be supported by provisions in the covered agreements. Unfortunately, the covered agreements do not include any explicit prohibition of dicta. Instead, the claim that dicta are not allowed can only be inferred from WTO provisions, as the US alleged in their statement at the DSB meeting on 23 May 2016. According to the US, ‘more than two-thirds of the Appellate Body's analysis [in Argentina – Measures Relating to Trade in Goods and Services] – 46 pages – is in the nature of obiter dicta’ because:
The Appellate Body reversed the panel's findings on likeness and said that this reversal rendered moot all the panel's findings on all other issues, including treatment no less favorable, an affirmative defense, and the prudential exception under the GATS. Yet, the Appellate Body report then went on at great length to set out interpretations of various provisions of the GATS. These interpretations served no purpose in resolving the dispute – they were appeals of moot panel findings. Thus, more than two-thirds of the Appellate Body's analysis is comprised simply of advisory opinions on legal issues.Footnote 155
As mentioned earlier, such a position is premised on the Common Law view that the law-making power of the court arises from its function to solve disputes, thus the rulings which are necessary to resolve the disputes become the holdings, while those which are unnecessary become dicta. Therefore, the hidden assumption of such argument is that the roles of the panel and Appellate Body are limited to resolving trade disputes. A closer reading of the DSU reveals, however, not only that there is no support for such a view in the text of the DSU, but also the DSU explicitly requires the panel and Appellate Body to go beyond merely solving disputes.
First, according to Art. 3.2 of the DSU, the WTO dispute settlement system serves not only to ‘preserve the rights and obligations of Members under the covered agreements’, but also to ‘clarify the existing provisions of those agreements in accordance with customary rules of interpretation of public international law’. As explained by Art. 3.4, the first function is achieved through ‘prompt settlement’ of disputes. But this apparently does not apply to the second function, as clarifications of treaty provisions often have to be conducted beyond the narrow confines of individual disputes. Indeed, it could be argued that the use of the term ‘clarify’ here widens the general roles of panels and the Appellate Body and enables them to provide ‘guidance’ to the Members’ future conducts under the covered agreements.Footnote 156
Second, the panel is under an explicit obligation to ‘address the relevant provisions in any covered agreement or agreements cited by the parties to the dispute’.Footnote 157 This means that, even if a provision cited by a party turns out to be inapplicable because the issue is moot, the panel still has to address it. Otherwise, the panel could well be accused of failing to fulfil its obligation under Arts. 7.2 and 11, especially if the Appellate Body decides to overturn the panel's finding that the specific provision is inapplicable.
Third, Art. 17.12 imposes similar obligation on the Appellate Body with even more explicit language by requiring the Appellate Body to ‘address each of the issues raised in accordance with paragraph 6 during the appellate proceeding’. Again, failure to comply with the obligation could expose the Appellate Body to allegations of violations of its duties under the DSU.
Fourth, more importantly, even for issues or provisions not raised by the parties, neither Art. 7 nor Art. 17 prohibits the panel or the Appellate Body from considering or ruling on such issues. To the contrary, as every lawyer knows, they often need to consider the unnamed provisions in order to assess the contexts of the provisions at issue in the litigation. One might argue that such a restriction can be found under Art. 17.6, which states that ‘[a]n appeal shall be limited to issues of law covered in the panel report and legal interpretations developed by the panel’. However, this provision at best only delineates what may be appealed by the parties to the dispute, but it does not impose restrictions on what the Appellate Body may rule upon. Even though Art. 17.12 refers to Art. 17.6, one cannot conclude that the Appellate Body is thus subject to the same restriction as it only states that ‘[t]he Appellate Body shall address each of the issues raised in accordance with [Art. 17.6]’. To the contrary, had the Members intended to also limit the power of the Appellate Body, they would have used the same language as Art. 17.6 here by stating that ‘the rulings of the Appellate Body shall be limited to issues raised in accordance with paragraph 6 during the appellate proceeding’.
To sum up, as the discussions above have illustrated, the covered agreements do not really distinguish between holdings or dicta in a decision. Thus, if anything, the Appellate Body's announcement on so-called ‘dicta’ in panel and Appellate Body reports is nothing but dictum on dicta. This approach is dangerous not only because it lacks legal basis in the covered agreements, but also because it could backfire when Members in turn borrow the term and accuse the Appellate Body itself of rolling out dicta, which is exactly what the US has done in the reappointment saga.
4. Conclusion
As Hersch Lauterpacht wisely warned 60 years ago, ‘[i]t is not conducive to clarity to apply to the work of the Court the supposedly rigid delimitation between obiter dicta and ratio decidendi applicable to a legal system based on the strict doctrine of precedent’.Footnote 158 The reason for this, as explained by former International Court of Justice President Jennings, is that when international lawyers talk about an obiter dictum, they imply ‘the existence in international law decisions of something that the common lawyer calls the ratio decidendi’.Footnote 159 Similarly, ‘whenever judges or publicists talk about obiter dicta the point of the distinction from the ratio decidendi is conceded by implication’.Footnote 160
However, as I have argued in this article, the Appellate Body has consistently rejected the idea that there is a system of binding precedents or stare decisis in the WTO. To be consistent, the Appellate Body should then stop using such loaded words like ratio and dicta, both of which are unique to the Common Law system of binding precedents. If the Appellate Body agrees with the previous rulings of the panel or itself, they shall just refer to it as ‘persuasive jurisprudence’, or ‘jurisprudence constante’ if they prefer to use a more established term. If they do not like the ruling on a particular issue, the Appellate Body should just state that it is wrong, or that it is not relevant to the issues raised in the case, not necessary for the resolution of the dispute at hand, or concerns a moot issue that is not argued by the parties, rather than bury it under the ambiguous and unhelpful tombstone of ‘obiter dicta’.
This is not just a petty issue of mere semantics, because words could influence people's thinking. As shown by my previous discussions, with such injudicious usage of terms, the Appellate Body has brought the destiny upon itself when Members started to fling accusations of dicta towards the Appellate Body's own reports.
Does this mean that the WTO Members have no way of controlling unwanted or unpopular emanations from the AB? The answer is no. Instead, I think the WTO Members could make use of any or all of the following suggestions.
First, the DSB could choose to adopt only the findings and recommendations of the panel and Appellate Body. As stated in Arts. 11 and 19.2 of the DSU, what really matters in the panel and Appellate Body reports are their ‘findings and recommendations’. This is confirmed by Arts. 21 and 22, which state that the losing party only need to comply with ‘recommendations or rulings of the DSB’. The findings are simply the legal conclusions without the detailed reasoning. As the reasoning is not adopted, there is no need to debate whether it falls under ratio or dicta. However, amendment of the DSU might be needed as one could argue that, under Arts. 16.4 and 17.14, the entire panel or Appellate Body report shall be adopted.
Second, if the Appellate Body's interpretation on a particular provision is so universally endorsed by the Members that the Members really want it to become part of WTO law, it shall be adopted only with a three-fourths majority of the Members. This is how an authoritative interpretation by the General Council or Ministerial Conference may be adopted, and it should also be how the interpretations of the Appellate Body are accorded authoritative status. It would create perverse incentives if such interpretation could be sneaked in through the backdoor via the operation of the negative consensus rule for the adoption of Appellate Body reports, while a proper authoritative interpretation has to go through the painful process of garnering the requisite majority among all WTO Members.
Third, on the other hand, if the Members are so appalled by what they perceive to be a blatantly wrongful interpretation by the Appellate Body in its report, they could always prevent further contamination of the WTO jurisprudence by trying to adopt an authoritative interpretation of the relevant provision and explicitly reject the approach taken by the Appellate Body, just as a parliament would do in a Common Law jurisdiction. Of course, as some WTO Members might stand to benefit from the interpretation at issue, it might be difficult to reach the three-fourths majority required to adopt the authoritative interpretation, unless such benefiting Members are also willing to join the consensus for systemic reasons.
In view of the current paralysis of the decision-making mechanism of the WTO, it might take some time before such suggestions may be accepted in practice. In the meantime, the Appellate Body should at least rid its reports off words such as ‘dicta’, lest the self-fulfilling prophecy ultimately becomes the Appellate Body's own destiny.
