Can’t buy me arbitrator love? How party-appointed arbitrators help ‘haves’ come out ahead in sports arbitration

Abstract

Party-appointed arbitrators are a common but controversial feature of many arbitral systems. The main point of contention is whether parties can and do use their power to appoint one of the arbitrators in a way that undermines impartial, equal, and fair arbitration. While the scholarly debate on party-appointed arbitrators is old and vast, claims made have been subject to limited empirical testing. Using the Court of Arbitration for Sport as a case study, this article empirically explores the effects of arbitrator appointments on arbitration outcomes and the relevance of arbitrator experience, arbitrator attitudes, and party capability on party success. In doing so, this study finds that arbitrator experience and attitudes affect outcomes and that this can be exploited by parties when selecting arbitrators. Importantly, the study also finds that the ability to make use of this strategic opportunity depends on parties’ capability. Whereas party-appointed arbitrators may be less problematic in arbitration between parties of equal capability, this finding calls into serious question the fairness and legitimacy of using party appointments in arbitration where there is a disparity in capability between the parties.

1 Introduction

1.1 Playing the arbitration game successfully

Access to objective, independent, and impartial judicial institutions is a fundamental element of any legal system based on the rule of law and respect for due process. To the uninitiated, it may therefore be surprising to learn that those who resolve their disputes through arbitration—as a private alternative to the primary or default jurisdiction of (state-based) courts—frequently select their own judges. As one experienced international arbitrator succinctly formulates the core issue, ‘[w]hy would any party have confidence in an arbitrator selected by its unloved opponent?’¹

So-called party-appointed arbitrators are a common feature of many arbitration systems, domestic as well as international, and have been around in various forms since ancient Greece.² Under the most common design, the parties appoint one arbitrator each. The party-appointed arbitrators then together, or some other actor such as the arbitration institution, appoint the panel president.³ This practice is so well established that Brower and Rosenberg claim that the ‘legal right’ of the parties to name an arbitrator is ‘beyond debate’.⁴

Arbitration can be thought of as a game. For both parties, the goal of the game is for at least two out of three arbitrators on the panel to vote for their preferred outcome.⁵ In order to achieve this goal, the parties and their counsel will seek to make the most persuasive arguments possible. A key strategic decision in order to achieve that goal is, however, who the parties appoint as ‘their’ arbitrator.⁶ As others have put it, ‘the selection is a key decision in winning or losing’⁷ and ‘the selection of the party-appointed arbitrator may be the most critical decision in an international arbitral proceeding’.⁸

In order to maximize the payoff of the selection pick, parties should obviously seek to appoint an arbitrator who is likely to vote in the direction of their desired outcome. This is however neither the only nor a sufficient criteria in selecting the ideal arbitrator. First, the arbitrator’s tendencies should not be so strong or obvious that it provides the opposing party with the basis for a challenge on the basis of conflict of interest.⁹ As one experienced arbitration counsel explains:¹⁰

when I am representing a client in an arbitration, what I am really looking for in a party nominated arbitrator is someone with maximum predisposition towards my client, but with the minimum appearance of bias.

Second, since a single favourable vote is insufficient to achieve a positive outcome, parties should appoint arbitrators who are likely to convince at least one of the other two arbitrators on the panel to vote in the direction that they prefer. Like in other contexts where a simple majority decides the outcome, we would, according to a theory developed for political processes¹¹ and extended to courts,¹² expect that arbitration outcomes are decided not by the outliers on the panel but by the median vote. Since both parties have the same goal and the same opportunity to select a positively disposed arbitrator, the member of the panel whose opinion is most likely to be swayed is that of the panel president.¹³ It is also, ceteris paribus, ‘the presiding arbitrators who are likely to be decisive decision makers as the median voter’ on the panel.¹⁴ One very important characteristic to look for in party-appointed arbitrators is, therefore, as one arbitration counsel phrased it, ‘the legal intellect to convince the president of the tribunal’.¹⁵ Like in the joke about two men outrunning a bear,¹⁶ a party only has to pick an arbitrator that will be more likely than the opposing party’s arbitrator to convince the panel president.

Whereas party appointment of arbitrators is in theory relatively straightforward and simple, and there is significant existing literature devoted to the topic, significant theoretical and empirical knowledge gaps exist. Can and does the arbitrator choice give the appointing party a litigation advantage, that is, to what extent is arbitrator appointment a litigation-strategic decision? If party-appointed arbitrators tend to vote in favour of their appointers, as some claim, why is this? Are party-appointed arbitrators simply ‘hired guns’ or ‘party representatives on the panel’ who vote for those who pay their fees? Or do they suffer from affiliation bias that cause them to push for and advance the interests of the appointing parties per se, that is, simply because they are their principals? Or can whatever litigation advantage a party may gain from appointing certain arbitrators be explained by the latter holding pre-appointment attitudes that align with the interests of the appointing party? Finally, are certain types of parties, such as those with more experience, money, and other resources, better than other parties at exploiting the strategic opportunity that arbitrator appointment provides?

By empirically exploring these questions, this paper seeks to enhance our understanding of (i) how parties’ capability affects their ability to make good strategic appointment decisions and (ii) how those decisions affect arbitration outcomes. To achieve this, I will use the Court of Arbitration for Sport (CAS) as a case study. As explained in Section 2.2, arbitration on the CAS has a number of features that provide an opportunity to study these questions in ways that other arbitration contexts, systems, and institutions do not.

1.2 Interests, impartiality, and influence of party-appointed arbitrators

The use of party-appointed arbitrators has both its critics and its defenders; critics have argued that the system is fundamentally problematic and ought to be altogether abolished.¹⁷ According to the critics, party-appointed arbitrators tend to act and vote in favour of the appointing party, and this introduces a ‘moral hazard’ in the arbitral system¹⁸ that undermines its independence and impartiality.¹⁹

The main point of contention between critics and defenders is not whether party-appointed arbitrators have a tendency to vote in favour of the appointing party—although the empirical support is quite limited, outright rejections of this claim are rare—but rather what the causes and effects of this phenomenon are. A straightforward explanation is economic incentives. As Branson puts it, ‘those who appoint want to win and those who want to be appointed want the income that appointment brings’.²⁰ In addition to the direct economic compensation that appointing parties provide appointed arbitrators, being appointed helps ‘build careers and markets’ for the future and brings more appointments.²¹

Others come to the defense of the party-appointment system by offering an alternative account of how parties and arbitrators behave. In response to the critics, the defenders argue that a biased arbitrator is more likely to harm the appointing party’s interests than advance them since a biased arbitrator will have no credibility and therefore no influence on the panel, including crucially on the panel president. Accordingly, it is in the interest of parties to select experienced arbitrators²² who have a reputation for being fair, impartial, unbiased, and independent.²³ Like the critics, the defenders use economic arguments to support their view: since arbitrators’ appointment opportunities depend on their reputation for being impartial and independent, they will be hesitant to risk their reputation by acting in favour of the appointing party.²⁴

That arbitrators have incentives to be perceived as unbiased, and may even honestly and actively seek to set aside biases that they inevitably have,²⁵ does not however mean that they do not favour the party that appointed them. First, even where there are no actual monetary interests involved, and participants are explicitly asked to make unbiased decisions, a psychological effect may remain.²⁶ Second and most importantly, it is commonly claimed that parties seek to appoint arbitrators ‘who are likely to be favourable’²⁷ and whose views are similar to theirs and align with their interests.²⁸ As Puig explains, a ‘rational litigant will try to select a judge who has reliably shown the appropriate judicial philosophy towards a particular set of issues relevant to the litigant’s goals’.²⁹ Few, if any, who are experienced in arbitration seriously challenge this view. For example, the President of the Swiss Arbitration Association has characterized party-appointed arbitrators as ‘sympathetic friends’ of the appointing parties.³⁰

1.3 Effects of party capability

That some of the adjudicators are appointed by the parties on the basis that they might render a favourable ruling causes party-appointed arbitration to deviate markedly from how requirements of independence and impartiality are ensured in ordinary courts. In arguing why adjudication by institutions where the majority of the adjudicators are appointed by the litigants, although different from courts, complies with the same legal requirements of independence and impartiality, defenders of the system will point to the formal equality of opportunity: both parties have exactly the same opportunity to pick a favourable arbitrator. So what is the problem? While it is true that the use of party-appointment arbitrators is formally equitable, in the sense that the parties get the same number of picks, there are good reasons to believe that in practise not all parties have the same capability to make use of this opportunity and that the use of party-appointed arbitrators therefore systematically benefit certain classes of litigants and disadvantages others.

In his seminal article, Galanter provides a simple yet powerful framework for understanding why dispute-settlement systems tend to favour certain parties.³¹ The theory rests on the idea that parties can be distinguished based on party capability.³² As summarized by Szmer et al., ‘[p]arty capability theory assumes that litigants with certain characteristics (eg prior litigation experience and substantial financial resources) have several advantages when facing less capable litigants and these advantages may influence judicial behavior’.³³

Parties can broadly be divided into two categories based on party capability. The first is so-called ‘haves’ who tend to be well-established, resource-rich, repeat-players with access to high-quality representation and counsel by experienced lawyers. Examples of actors that commonly belong to this group are governments and their agencies, large corporations, and large non-governmental organizations. ‘Have-nots’, by comparison, are relatively inexperienced and resource-poor litigants. Their resulting litigation disadvantage could be partially offset by having lawyers that act on their behalf, but because they have fewer resources than other litigants it is also likely that they are represented by fewer and less experienced lawyers. They also tend to be ‘one-shotters’, that is to say actors that are unlikely to be involved in more than one dispute.³⁴

Galanter’s theory is one of the most influential in law.³⁵ There is an extensive literature employing the theory to study the litigation advantages of haves over have-nots across various jurisdictions. This includes, to name just a few, Australia,³⁶ Canada,³⁷ China,³⁸ the European Union,³⁹ the United Kingdom,⁴⁰ the United States,⁴¹ Taiwan,⁴² and Sweden.⁴³

Galanter emphasizes the distinction between state-based court litigation and more informal dispute resolution mechanisms, such as arbitration,⁴⁴ and existing research has applied and demonstrated the relevance of the theory in the context of arbitration. For example, Bingham was able to show that repeat players tend to have greater success in employment arbitration.⁴⁵ Franck and Wylie similarly found that parties with more experienced counsel fared better in investment arbitration.⁴⁶

2 Party capability and party-appointed arbitrators on the Court of Arbitration for Sport (CAS)

2.1 The CAS, its jurisdiction, and procedure: an overview

The CAS is an international arbitration institution seated in and governed by the laws of Switzerland⁴⁷ established by the International Olympic Committee (IOC) in 1984. In response to concerns about its independence from sports governing bodies (SGBs),⁴⁸ the CAS has since 1995 been governed by the International Council of Arbitration for Sport (ICAS). ICAS’s board members are directly or indirectly appointed by the IOC, international federations of Olympic sports, and the national Olympic Committees.⁴⁹ ICAS serves a number of functions,⁵⁰ including:

1. revising the code governing CAS and its procedures,

2. appointing the members of the CAS Court Office who run the day-to-day business, including the formation of arbitration panels: the Secretary-General and the Presidents of the Ordinary Arbitration Divisions, the Appeals Arbitration Division, and Anti-Doping Division, as well as...,

3. deciding which individuals are eligible to serve as arbitrators by adding and removing individuals from the list of approved CAS arbitrators (the ICAS list).

Upon its founding, IOC President Juan Antonio Samaranch had a vision for the CAS to become ‘a supreme court for world sports’.⁵¹ The CAS began its assent to this position in earnest in the mid-2000s as it was designated the final body of appeal for major international sports governing bodies (SGBs) and for international disputes under the World Anti-Doping Agency’s (WADA) World Anti-Doping Code.⁵² As illustrated in Fig. 1, this has led to a massive increase in the number of procedures. Most of this is caused by an increase in the number of appeals procedures which make up almost 80 per cent of all cases that end up before the CAS.

Like other arbitration panels, the CAS’s jurisdiction over a dispute—as well as the corresponding lack of jurisdiction for ordinary courts—rests on the mutual agreement and consent of the parties ‘to displace the courts’ competence with a mutually constructed alternative’.⁵³ For the CAS, that ‘mutual agreement’ in most cases, and particularly when it comes to appeals procedures, is found in the instruments authored by SGBs and that govern sports, and which may, for example, be entitled statutes, rules, and regulations. By becoming a member of an SGB or participating in a sporting event, like the Olympic Games, clubs, athletes, officials etc. become bound by those instruments, including clauses conferring jurisdiction over disputes on the CAS. For example, Article 61 of the Olympic Charter (2020) states:

The decisions of the IOC are final. Any dispute relating to their application or interpretation may be resolved solely by the IOC Executive Board and, in certain cases, by arbitration before the Court of Arbitration for Sport (CAS).

Any dispute arising on the occasion of, or in connection with, the Olympic Games shall be submitted exclusively to the Court of Arbitration for Sport, in accordance with the Code of Sports-Related Arbitration.

Many international SGBs include similar clauses in their statutes. For example, Article 56 of the FIFA Statutes (2022) provides:

FIFA recognises the independent Court of Arbitration for Sport (CAS) with headquarters in Lausanne (Switzerland) to resolve disputes between FIFA, member associations, confederations, leagues, clubs, players, officials, football agents and match agents.

The provisions of the CAS Code of Sports-related Arbitration shall apply to the proceedings. CAS shall primarily apply the various regulations of FIFA and, additionally, Swiss law.

Whether this approach constitutes sufficient consent to support the CAS’s jurisdiction has been called into question.⁵⁴ In Mutu and Pechstein v. Switzerland⁵⁵, the European Court of Human Rights had reason to consider this question in the context of determining whether the CAS meets the requirements of independence and impartiality under Article 6 of the European Convention on Human Rights since that requirement does not apply to ‘voluntary arbitration to which consent has been freely given’ (para. 96). The Court found that this is not the case when it comes to the CAS and athletes like Mrs. Pechstein who had no real choice as her only choice:⁵⁶

was between accepting the arbitration clause and thus earning her living by practising her sport professionally, or not accepting it and being obliged to refrain completely from earning a living from her sport at that level.

Having regard to the restriction that non-acceptance of the arbitration clause would have entailed for her professional life, it cannot be asserted that she had accepted that clause freely and unequivocally.

The Court thus concludes that, even though it had not been imposed by law but by the ISU regulations, the acceptance of CAS jurisdiction by the second applicant must be regarded as ‘compulsory’ arbitration within the meaning of its case-law.

Unless the parties agree or the CAS Appeals Division President decides to submit the case to a sole arbitrator, the following appointment procedure is used in CAS appeals procedures. The appellant begins by appointing an arbitrator, followed by the respondent nominating a second arbitrator. After consulting with these two party-appointed arbitrators, the CAS Appeals Division President appoints the panel president.⁵⁷ Only arbitrators that appear on the ICAS list are eligible for appointment in any of these positions.⁵⁸

The SGBs exert ‘real influence’ over arbitrator appointments on the CAS in several ways.⁵⁹ The ICAS members, who are directly and indirectly appointed by SGBs, decide on which arbitrators can and cannot be appointed. Moreover, ICAS also appoints the CAS officers that in turn form the panels. The CAS officers particularly exercise power over the appointment of sole arbitrators and panel presidents. In their separate opinion in Mutu and Pechstein, Judges Keller and Serghides of the European Court of Human Rights, who were of the opinion that the CAS does not live up to the requirements in Article 6 of the European Convention, held that:⁶⁰

the system of selection of arbitrators procures for the organisations (directly and indirectly through the ICAS) a disproportionate and unjustified ‘influence’ over the procedure for choosing the arbitrators who are responsible for settling disputes between the organisations and the athletes. In other words, this system ‘revealing the existence of a certain link between the ICAS and organisations that might be involved in disputes with athletes before the CAS, especially those of a disciplinary nature’ [...]. A link which [...] we regard as ‘worrying’.

2.2 Qualities of the CAS as a case study

The CAS is an excellent case for advancing our understanding of the appointment and effects of party-appointed arbitrators generally, even though every form of arbitration is to some extent unique. The CAS has at least three features that distinguish it from many other arbitration institutions but which provide discrete advantages when it comes to illuminating general features related to arbitration in general and arbitrator appointment in particular. First, the CAS is a relatively transparent arbitration institution and access to data on the CAS is exceptionally high (see further Section 5.1) for a type of dispute resolution institution that, due to a strong tradition of confidentially, generally suffers from a lack-of-data problem.⁶¹ Second, a general methodological problem in empirically studying arbitration is that it is difficult to measure outcomes in terms of litigation success in a reliable manner. For example, in commercial disputes, it is difficult to compare a monetary award actually granted to what it could or should have been. However, for almost 80 per cent of all cases that come before the CAS that concern an appeal of a decision by sporting bodies (see Fig. 1), there are clearly, easily, and reliably measurable outcomes as the appealed decision is either (partly or wholly) overturned or not (see further Section 5.2). Third, because only those arbitrators who appear on the ICAS list (see Section 2.1) are eligible to serve as party-appointed arbitrators and panel presidents on the CAS, unlike in arbitration institutions where anyone can be appointed, we have information about all appointable arbitrators. This allows us to study not only which arbitrators were appointed, but also which arbitrators were not appointed. This provides an exceptional opportunity to explore the appointment decision and its effects.

Just like investment arbitration, which directly involves states as parties, is different from commercial arbitration,⁶² so is sports arbitration different from other forms of arbitration. However, the nature and degree of those differences also vary between different forms of sports arbitration. For example, a dispute over contractual damages between two football clubs that have respectively sold and bought a football player is in many regards similar to many other commercial disputes. One of the more unique characteristics of the CAS as an arbitral institution is that a majority of its decisions involve a review of the rules and decisions of SGBs. In those cases, the CAS serves a function that is akin to an administrative or constitutional court.⁶³ This aspect of the CAS, which relates to the regulatory, administrative, and disciplinary functions of SGBs, makes it an excellent test case for theories about party capability. Although far from every case studied is of this type,⁶⁴ these same features may affect the generalizability of its findings.⁶⁵

3 Related work

This study sits squarely in the intersection of and primarily speaks to two strands of literature where the first strand focuses on party-appointed arbitrators and the second strand on the role of party capability in arbitration. These fields and some of the central related work were introduced above (Section 1), whereas I will here highlight some of the empirical studies that are most closely related and relevant to this study.

Although some empirical studies of arbitration, including studies related to arbitrator selection, were conducted in the 1980s⁶⁶ and the 1990s,⁶⁷ empirical research addressing the issues examined herein were for a long time limited.⁶⁸

The situation has to some extent improved in recent years as a result of an ‘empirical turn’ in arbitration studies. This turn is led by and heavily focused on investment arbitration,⁶⁹ and is to a large extent motivated by a criticism of it and crisis in its legitimacy.⁷⁰

To the extent that early critics of party-appointed arbitrators provided empirical evidence for their claims, they did so on the basis of small-sample-size studies of awards that contain dissenting opinions and which showed that dissenting opinions are almost exclusively issued by party-appointed arbitrators in favour of their appointing party.⁷¹ While interesting and weighing against the complete absence of appointing-party bias, one should be aware both that dissents are relatively rare and that party-appointed arbitrators thus frequently vote against the appointing party.⁷² Moreover, data on dissenting opinions do not speak to the relative strength of the affiliation and selection effects as explanations.

Puig has studied and made important contributions to our understanding of arbitrator bias. By studying ICSID awards, Puig showed the presence of pro-state and pro-investor bias and that litigants tend to select party-appointed arbitrators on this basis.⁷³ Moreover, using experimental surveys and tasking arbitrators with deciding a hypothetical case, Puig and Strezhnev showed that participants who were told they were party-appointed were more likely to reach an outcome that favoured the fictive appointing party compared to participants who were told they were appointed by ‘the Parties’, thereby demonstrating the existence of affiliation bias.⁷⁴

These quantitative studies were recently supplemented by qualitative insights by Traxler who conducted forty-eight interviews with investment arbitration counsel.⁷⁵ In addition to confirming the (perceived) importance of arbitrator selection on outcomes, the main contribution of his study when it comes to the questions explored herein lies in the depth, richness, and empirical support that it provides to some familiar claims made about arbitration in general and party-appointed arbitrators in particular.

4 Hypotheses

4.1 Arbitrator experience

It is well established that parties prefer to appoint arbitrators that are experienced, in general⁷⁶ and/or when it comes to deciding cases that are similar to the present dispute.⁷⁷ There is also clear evidence for this preference in empirical data. For example, in his study of appointments in ICSID arbitrations, Puig demonstrated preferential attachment under which ‘appointments are distributed among arbitrators according to their existing share of prior appointments’.⁷⁸ It is less clear exactly why parties prefer experienced arbitrators. Some of those reasons are unrelated to the puzzle explored here, for example, that picking an experienced arbitrator helps shield the appointer from criticism in case of a loss,⁷⁹ or that experienced arbitrators might manage the arbitration better, thereby reducing process costs. For the purpose of playing a better arbitration game, at least two reasons for parties’ preference for experience can be advanced.

First, as elaborated on in Section 1, we expect that parties research potential appointees for the purpose of gauging their attitudes. It is nearly impossible for parties to predict how individuals who have never previously sat as arbitrators will act if appointed. However, for each panel they sit on and each award that they participate in rendering, that uncertainty decreases.⁸⁰ Inexperienced arbitrators are, in other words, seen as ‘jurisprudential wildcards’ that parties should avoid.⁸¹ Both general experience in arbitration and more domain-specific expertise are therefore valuable to parties and the arbitrators’ experience is a proxy for the amount of information that parties can have about them.

Second, there are reasons to expect that more experienced arbitrators may exert stronger influence on the other members of the arbitration panel and, most importantly, on the median-vote president. In their seminal study of international arbitrators, Dezalay and Garth found that so-called ‘Grand Old Men’ or, as one commentator calls them, ‘super arbitrators’,⁸² have an out-sized influence on individual panels, as well as more systemically on arbitration institutions.⁸³ Similarly, Puig found that ICSID suffers from a ‘small world phenomenon’ where a small, elite group of ‘power-brokers’ dominate the field and exert an out-sized influence on its development.⁸⁴ Arbitrators with greater experience can both be expected to exert greater influence over other arbitrators and to be less likely to be influenced by others, by merit of superior knowledge, greater ‘symbolic capital’,⁸⁵ or some combination of the two. Appointing an experienced arbitrator can thus increase a party’s chance for success, assuming that the party makes a correct assessment of the arbitrators’ tendencies and attitudes.

If these expectations are correct, it is clear why previous experience is an attractive trait in a potential arbitrator to appointing parties that seek to play the arbitration game well. Appointing an experienced arbitrator increases the likelihood (i) that the appointed arbitrator will vote in the direction desired by the party and (ii) that they will convince other panelists to vote in the same direction. Thus:

Hypothesis 1. Appointing an experienced arbitrator increases the likelihood of a successful outcome for the appointing party.

In testing this hypothesis, the reasoning outlined above dictates that the effect on outcome is not determined solely by a single party’s appointment choice, but also by the opposing party’s choice. To use a simple analogy, we can imagine the two party-appointed arbitrators playing tug-of-war with the expected median-voting president and that whichever party-appointed arbitrator is stronger will be successful in moving the median point to their side. However, we would expect the president’s experience to affect this in the sense that the party-appointed arbitrators' influence on the president will decrease if the latter is more experienced. In other words, it can be expected that it is more difficult for party-appointed arbitrators to move a more experienced president.

4.2 Arbitrator attitudes

According to the so-called attitudinal model, judges—like other humans—favour certain outcomes and these attitudes will—intentionally or unintentionally—affect how they act and vote on the bench.⁸⁶ There are no reasons to expect that arbitrators are different from judges in this regard. That is to say, we should expect that arbitrators have subjective views, preferences, and biases that vary between individual arbitrators. Moreover, it is likely that the attitudes with which they come into arbitration may, at least under certain conditions, affect how they think and act as arbitrators.⁸⁷

This is largely uncontroversial and undisputed. Based on this line of thinking, it is claimed that parties to an arbitration research potential appointees in order to gauge how they might decide the dispute in question. Experienced arbitration counsel have also publicly and explicitly acknowledged that this is indeed a common practice in field.⁸⁸ Thus, as explained in Section 1, much of the existing literature assumes that arbitrators have certain attitudes and that parties make their appointment picks based on those attitudes.⁸⁹

One possibly relevant factor in this regard is the arbitrators’ professional background. This is informed by the existing literature which expects that arbitrators’ professional experience and background may influence how they think and act as arbitrators.⁹⁰ For example, in their study of international commercial arbitration, Dezalay and Garth found a perceived strong difference in behaviour between arbitrators who were practitioners and those who were academics.⁹¹ In the context of international investment law, Waibel and Wu note:⁹²

Especially in more controversial areas of international investment law where the views of arbitrators may diverge, it is a possible that their background, life experience and ideology may influence how they decide cases. How an arbitrator balances competing considerations may hinge in part on a given arbitrator’s view of the world, on their mindset based on their professional experience and on the incentive the arbitrator faces when making the decision...

Arbitrators’ individual tendencies may however also more directly reflect what can best be described as ‘ideological’ leanings.⁹³ In an empirical study of international investment arbitrators, Waibel and Wu found that arbitrators’ policy preferences affected case outcomes, even more so than incentives.⁹⁴ For the purpose of this study, we are not primarily concerned with arbitrators’ attitudes per se, but rather to what extent parties’ selecting arbitrators based on those attitudes can result in a litigation advantage.

Hypothesis 2. Appointing an arbitrator whose attitudes align with the appointer’s interests increases the likelihood of a successful outcome for the appointing party.

4.3 Party capability

It is intuitive that the amount and quality of experience, knowledge, resources, and contacts that parties have—collectively here referred to as ‘party capability’—have an impact on their ability to make an effective appointment choice. Accordingly, we would expect that high-capability parties or, to speak with Galanter’s nomenclature, ‘haves’ are better than ‘have-nots’ when it comes to selecting arbitrators. First, with experience, resources, and connections come an informational advantage. So-called repeat players ‘have access to prior awards in arbitrations in which they were a party, which for large law firms can number into the hundreds’.⁹⁵ Also, as Franck explains:⁹⁶

For parties who are not repeat players or otherwise lack the ability to hire elite international law firms with unique expertise about arbitrator reputation, there may be informational gaps.

High-capability parties and their counsel are also both more likely to have better connections with experienced arbitrators. Finally, it is possible that experienced arbitrators are more likely to accept appointments by high-capability parties.⁹⁷ It is therefore expected that high-capability parties are better than low-capability parties at playing the arbitration game in general and appointing arbitrators that improve their chances of preferable outcomes in particular. By relating this expectation to the hypotheses previously defined, this leads to our final two hypotheses:

Hypothesis 3. Parties with greater capability are more likely to appoint experienced arbitrators.

Hypothesis 4. Parties with greater capability are more likely to appoint arbitrators whose attitudes are aligned with their own interests.

5 Method and data

5.1 Data

I depart from 1,471 final awards⁹⁸ issued by the CAS on the merits⁹⁹ in an equal number of unique cases between 1986 and 2022.¹⁰⁰ These are cases where a decision by a sporting body was appealed to the CAS under the ordinary appeals procedure and decided by a three-party panel that included two party-appointed arbitrators.¹⁰¹ Since there are three appointed arbitrators for each case—one appointed by the applicant, one appointed by the respondent, and one president appointed by the CAS office—the data includes information on 4,413 appointed arbitrators, two-thirds of which (n = 2,942) were appointed by parties. Based on a larger set consisting of all published CAS decisions (n = 2,465), I generate a number of variables, presented below, on the arbitrators, parties, and representatives that are involved in those ordinary appeals decisions.

Due to limited statistical information provided by the CAS, there is some uncertainty regarding the size and representativity of the sample. However, we expect that it constitutes slightly more than half of all such awards issued during the studied period.¹⁰² Although it is difficult to know with certainty, it is likely that cases that were not included in the sample are largely due to the parties jointly agreeing to keep them confidential. According to Rule R59 of the CAS code, ‘[t]he original award, a summary and/or a press release setting forth the results of the proceedings shall be made public by CAS, unless both parties agree that they should remain confidential’. One of the main purposes of the CAS, which is also reflected in its comparatively high degree of transparency, is that it should contribute to uniform application of sports law by setting out authoritative interpretations and general principle though its jurisprudence. SGBs are thus generally positive towards the publication of CAS awards and can accordingly be expected to be reluctant to agreeing to confidentiality, except for when the particular circumstances call for discretion, for example, in cases concerning minors. It is more likely that confidential awards derive from cases that do not involve SGBs on either side of dispute. While this study is thus not based on a random sample of cases, there is no reason to expect that the data selection will bias the results. It can in this context also be noted that this study cannot and does not make any claims about unseen cases, such as those not submitted to the CAS or those which were settled before an award was rendered.¹⁰³

Only those lower decisions that were appealed to the CAS are included in the data and we have no information about the decisions that were not appealed or how the former might differ from the latter. A potential problem in this regard is that high-capability parties, for the same reasons that they might be more successful before the CAS, are better at identifying decisions that have a high likelihood of being overturned by the CAS. In other words, decisions appealed by high- and low-capability parties may be qualitatively different.¹⁰⁴ To adjust for this, we include as a control whether the case was appealed by an SGB.¹⁰⁵

5.2 Variables

Outcomes

As outlined above, this study seeks to study and explain differences in case outcomes. Following inter alia McCormick¹⁰⁶ and Songer and Sheehan,¹⁰⁷ case outcome is approached by comparing the outcome of arbitration to the decisions that were appealed, using whether the appealed decision was overturned or not as the outcome variable. This is a simple but intuitive and reliable outcome variable. There is no doubt that the applicant is in a better situation if the CAS overturns the appealed decision, even if the appeal is only partially granted.¹⁰⁸ While one could argue that an ‘only’ partially granted appeal can be categorized as a partial win for the respondent, I think that in the particular context of appeals cases at the CAS, where the respondent side almost always includes a party that made or was involved in making the appealed decision, it is more appropriate to code it as a loss. We can safely assume that the respondent generally (i) would like to see its decision completely confirmed by the CAS and (ii) will perceive the CAS overturning its decision as a loss.

I capture the outcome in the dummy variable success which indicates whether the outcome in terms of the final award was successful for the party. The outcome of a case is coded as a success for the applicant if the CAS overturned, entirely or partially, the appealed award and as a success for the respondent if the CAS confirmed the appealed decision.

It can in this context be noted that the CAS with regard to outcomes exhibits one of the basic traits of an adjudication institution for which party capability plays a significant role. Previous studies of the advantages of repeat-players over one-shotters have found that one-shotters tend to have a higher litigation rate but lower win rate than the former.¹⁰⁹ This is also true for the CAS. Of the cases in the data with a clear case outcome, 88 per cent were brought by applicants that did not include an international SGB. However, those litigants were successful in only 47 per cent of the cases compared to international SGB applicants that were successful in 83 per cent of the cases that they brought.¹¹⁰

Arbitrator experience

Although the ICAS’s list of appointable arbitrators at any given time contains 300–400 names, a small group of individuals receive most of the appointments. As illustrated in Fig. 2, arbitrator appointments on the CAS follow a power law distribution. Among all 688 unique individuals who have appeared on the ICAS list, the 24 most highly appointed arbitrators (3.5 per cent) have received 50 per cent of all appointments (those above the dotted line in the figure).

Because of how arbitrator appointments on the CAS are distributed and inspired by the findings of previous research (see Section 4.1), I seek to develop a variable that captures whether arbitrators are super-arbitrators. These super-arbitrators serve a role on CAS that is comparable to what other arbitration researchers have called ‘Grand Old Men’¹¹¹ and ‘power-brokers’¹¹². They are the repeat appointees that make up the institutional backbone of the CAS and who have an out-sized influence over how its case law develops.

A methodological challenge when capturing whether an individual arbitrator belongs to this elite class of arbitrators is that the network is continuously changing as new appointments are continuously made, as well as arbitrators added and removed from the ICAS list. Accordingly, a variable that designates an arbitrator as having the status of super-arbitrator (or not) must take change over time into account. I therefore here define as a super-arbitrator an arbitrator that at the time of appointment belongs to (i) the group of appointable arbitrators and (ii) the group of most-appointed arbitrators that as a collective have received at least 50 per cent of the total number of previous appointments of all appointable arbitrators at that time. To use a cumulative appointment share of 50 per cent as the community criteria is to some extent arbitrary. It is however natural given the power law distribution of the appointments. This information about arbitrators is subsequently used to generate case-level dummy variables describing whether the parties’ appointees were super-arbitrator or not.

Party capability

Several different factors affect party capability, and this study includes four measurements. First, party capability may come from the party having gained valuable experience by appearing in previous cases before the CAS. The number of cases that a party has previously had before the CAS constitutes a reasonable proxy for the party’s knowledge of the applicable rules and case law, the process before the CAS, as well as the tendencies and preferences of arbitrators. In the data, this is captured in terms of the number of previous appearances that the party has in the CAS (party n appearances) as ascertained from the larger dataset.¹¹³

As we can see in Fig. 3, most parties (76 per cent) that have come before the CAS are true one-shotters that are ever only involved in a single case. This is particularly true for individuals who are on average involved in 1.15 cases. On the opposite side of the distribution are international SGBs who appear on average in 13.2 cases. Only international SGBs have more than 30 appearances in the data and some international SGBs appear in more than 100 cases,¹¹⁴ making them clear repeat-players.

Party capability may also result from the amount of financial and personnel resources that the party can devote to the litigation.¹¹⁵ In cases before the CAS there is frequently clear inequity of resources between different classes of parties. In particular, the resources of the individuals who come before the CAS—most of whom are athletes—are generally much more limited than those of the SGBs and, in particular, the large international SGBs, like FIFA, IOC, WADA, and UEFA¹¹⁶, that frequently appear on the other side. International SGBs also hold a unique position among sport litigation ‘haves’ that most closely resemble those of governments. International SGBs are monopolists within their respective domains, and are, like governments, immune from opposing parties withdrawing and refusing to continue relations.¹¹⁷ Moreover, like governments, international SGBs created and control the institutional and procedural framework of the CAS as well as the substantive rules that the CAS enforces. They are accordingly capable of structuring the system to their benefit.¹¹⁸ The ability of CAS arbitrators to interpret and apply sport rules is ultimately subordinate to the international SGBs' power to set and modify those rules.¹¹⁹ Finally, akin to how governments influence judicial appointments, international SGBs through the ICAS exert significant influence over appointments. All of these factors compound to make whether a party is an international SGB an important proxy for party capability in the context of the CAS.

A particular type of resource is the lawyers who assist, advice, and represent the parties. The number of representatives who appear in the CAS on behalf of a party is a simple but reasonable proxy for the financial and personnel resources that a party has available and can devote to litigation and can therefore be a valuable proxy for party capability.¹²⁰

An alternative measurement of the quality of representation is the amount of prior litigation experience that a party’s lawyers have. Having more experience should reasonably improve a representative’s ability to argue persuasively, but being frequently hired by parties can also reflect the competence and quality of lawyers, at least according to the appointing parties. Counsel experience is therefore frequently used as a measure of party capability.¹²¹ As with arbitrator and party experience, counsel experience in the CAS follows a power law distribution, as displayed in Fig. 4. It is reasonable to expect that where on this distribution a party’s counsel is positioned can affect the likelilhood for a successful outcome as more experienced counsel can be expected to have more experience with and expertise in the applicable substantive and procedural rules, CAS case law, the process before the CAS, as well as the tendencies and preferences of arbitrators. In the data, this is capture in the variable representatives n appearances.¹²² It can be noted that there is a clear difference between international SGBs and other classes of litigants before the CAS regarding the experience of their representatives. Lawyers who appear on behalf of international SGBs appear on average in more than twice as many cases (44.2) than lawyers who appear on behalf of national SGBs (18.5), clubs (20.6), and individuals (19.7).

Arbitrator attitudes

What policy dimensions and related arbitrator policy preferences are relevant varies depending on context. For example, there are no compelling theoretical reasons for applying the progressive-conservative dimension used by Waibel and Wu in the context of sports arbitration.¹²³ I therefore instead define two theoretically relevant attitudinal dimensions for the particular context of the CAS and sports dispute resolution more generally.

The first of these I call Tough-on-Discipline (ToD). A majority of the cases that come before the CAS concern doping violations or some other disciplinary action. Those cases almost invariably began with an SGB taking a disciplinary decision, most often against an individual, such as an athlete. That action was appealed either directly to the CAS or via one or multiple lower sport-internal institutions entrusted with examining such decisions. Under what conditions—for example, whether someone should be punished for mistakenly and inadvertently consuming a prohibited substance—and how severely—for example, whether the length of suspension should be two or four years—is a common and obvious policy dimension that CAS arbitrators face. Tough-on-Discipline can in this regard be thought of as a sport-analogous version of a ‘tough-on-crime’ tendency in national criminal courts.

To capture this dimension, I first assign values to awards in disciplinary cases based on outcome where decisions that uphold or increase the disciplinary liability and/or disciplinary consequences are coded as advancing tough-on-discipline (ToD = 1), whereas decisions that reduce liability and/or consequences are coded as the opposite (ToD = 0). Second, for each appointment opportunity, I calculate for all appointable arbitrators the mean ToD-score (as a value between 0 and 1) for all previous awards they have participated in rendering and which has a value on the dimension. Third and finally, I compare, on the one hand, the appointing party’s presumed preference on ToD¹²⁴ to the appointee’s mean ToD. The resulting measurement, party-appointee ToD alignment, has a value between 0 and 1 where a higher value represents greater attitudinal alignment between the appointing party and the appointed arbitrator.

The second attitudinal dimension is institutional and seeks to capture whether arbitrators favour certain classes of parties. This dimension is inspired by investment arbitration research where arbitrators are commonly seen and characterized as having either a pro-investor or pro-state bias.¹²⁵ A comparable type of bias in sports arbitration would be a bias for or against the interests of governing bodies (SGBs) in relation to the interests of those who are governed (individuals and clubs). On a theoretical level, this dimension can be understood as reflecting the arbitrator’s view on the autonomy of SGBs and the balancing of collective interests in sports—as interpreted by the SGBs in their norm- and decision-making capacity—against individual interests—such as the interests of individual athletes. I produce a measurement for this attitudinal dimension in the same way as for ToD, tracking to what extent cases that arbitrators have previously been involved in have resulted in outcomes that favor  or disfavor SGBs and then comparing this to the appointing party’s class to generate a value for party-appointee class alignment.¹²⁶

It should be acknowledged that measuring arbitrators’ attitudes, which are not directly observable, is methodologically challenging.¹²⁷ One way to gauge arbitrator attitudes is to study the awards they have previously delivered, and more specifically the outcome. This is an imperfect but relevant proxy; if cases that an arbitrator partook in deciding consistently go in a particular direction, this can be taken as an indication of their attitudes.¹²⁸ It is also a likely that this approach to some extent overlaps with how parties and their representatives go about researching prospective appointees before making their appointment decision.

Case variables

The CAS hears cases on various matters, including doping cases, cases concerning other forms of disciplinary matters, and eligibility cases. Depending on the subject matter, a case will involve distinctly different types of facts, applicable substantive rules, as well as evidentiary questions and rules. We can therefore expect that the probability of the applicant winning the appeal may vary significantly depending on the nature of the matter. Accordingly, where I seek to capture the effects of other variables on outcomes, models include controls for type of matter.¹²⁹ We also need to include a control for time since several of the variables included in the models vary over time, for example, the number of previous appearances by parties and their counsel. I therefore include a case-level control for the case year, that is to say the year that the case was brought before the CAS. Finally, since we, as elaborated above, have reasons to believe that cases appealed by SGBs might be qualitatively different from cases appealed by other classes of litigants, we include whether the case was appealed by an SGB as a control. Descriptive statistics for these variables are presented in Table 1.

6 Results

6.1 Reaching the desired outcome

Before considering the effects of party-appointed arbitrators on outcome, we shall briefly explore the effects of party capability on outcome per se, that is to say without considering the characteristics of the arbitrators involved. Table 2 presents four models testing the effects of the different measurements of party capability considered on the basis of applicant success (see Section 5.2). When considering each of the four party-capability factors in separate models (models 1–4), we find that the party being an international SGB, the number of previous appearances on the CAS that the party has, and the number of previous appearances before the CAS that the party’s counsel has previously made are all postively and significantly correlated with an increased likelihood of a beneficial arbitration outcome. However, when we include all of the factors in a single model (model 5), only the party being an international SGB has a significant and positive effect on litigation outcome. That whether a party is an international SGB has a significant effect on the predicted directions for both applicants and respondents is indicative of party capability playing a significant role at the CAS. It is arguably the best measurement of party capability when it comes to the CAS, and these models suggest that it is difficult for non-international SGBs to overcome this advantage by hiring more or more qualified representatives.

We turn now to consider the effects of the arbitrators’ characteristics on outcomes and begin by testing the effects of arbitrator experience (Hypothesis 1). To do so, we consider what if any is the effect of a party—here split by applicants and respondents—appointing a super-arbitrator on the applicant wining the case, as well as the effect of the panel president being a super-arbitrator.

To begin we can observe that successful outcomes for appellants appear to be connected to the presence of super-arbitrators on the panel. Fig. 5 illustrates the percent of cases in which the appellant has been  successful in the CAS under different panel configurations. Overall, appellants have been successful in 51 per cent of the cases, but this share increases to 55 per cent in cases where the appellant appoints a super-arbitrator, and decreases to 48 per cent where the appellant appoints a non-super-arbitrator. If the appellant appoints a super-arbitrator, the respondent’s choice has a comparable effect on appellant success. If the respondent also appoints a super-arbitrator, the appellant’s win rate returns to 51 per cent, whereas it increases to 59 per cent if the respondent appoints someone that is not a super-arbitrator. Interestingly, if the applicant appoints a non-super-arbitrator the respondent’s pick is not associated with a shift in the applicant’s win rate.

The effect of the panel president being a super-arbitrator is interesting, particularly when the panel includes a mix of super-arbitrators and non-super-arbitrators. One possible expectation would be that super-arbitrators, by virtue of constituting a close-knit network of individuals that serve repeatedly on panels, would develop a similar way of thinking and reasoning.¹³⁰ It would therefore not be surprising if where only one party appoints a super-arbitrator, super-arbitrator presidents tended to reason in a similar way as that appointee and to vote in favour of the appointing party. If so, such tendencies could explain the variance in outcome when respondents appoint super-arbitrators, but applicants do not.¹³¹ This is however not the case if the parties’ picks are reversed.¹³² It therefore seems that the effects of the president being a super-arbitrator or not is more complex. It is possible that super-arbitrators, for one reason or another, tend to favour respondents,¹³³ or that the CAS office picks panel presidents based, inter alia, on features of the case or which arbitrators the parties appointed. These are questions for future research to explore.

This discussion reveals that win percentages based on panel configuration alone fail to take into consideration the various factors that we can expect to affect outcomes. To consider these, we construct a number of regression models. As illustrated in Fig. 6, we find that when parties appoint super-arbitrators as their party-appointed arbitrators the litigation outcome is significantly more likely to go in the direction that they desire (Table A1).¹³⁴ This is true for applicants as well as respondents and the size of the effect is comparable. That is to say, applicants appointing super-arbitrators is significantly positively correlated with an applicant win and respondents appointing super-arbitrators is significantly negatively correlated with an applicant win. These findings allign with our hypothesis and support our expectations. In practical terms, this means that the applicant’s success chance increases by circa 30 per cent, all else being equal, if appointing a super-arbitrator and decreases by a similar degree if the respondent does the same, a sizable effect on arbitration outcomes. However, contrary to what was expected, whether the president is a super-arbitrator has no discernible effect on the impact of the party-appointed arbitrators on the outcome. It would thus seem that super-arbitrator presidents are not more difficult to sway than other presidents.

Turning our attention to arbitrator tendencies, we test whether a party appointing arbitrators who have exhibited tendencies that are aligned with the party’s preferred outcome increases the chance of an outcome that favours the party (Hypothesis 2). As displayed in Fig. 7, there is a clear and significant correlation between party-appointee tendency alignment and a successful outcome for the appointing party (Table A2).¹³⁵ This is true for both measurements of arbitrator tendencies, tough-on-discipline (ToD) alignment (model 1) and party class alignment (model 2). Thus, making an appointment based on the outcome of previous cases is positively correlated with positive litigation outcomes and an advantageous strategy for parties that seek to play the litigation game well.¹³⁶

6.2 Picking the right arbitrator

Having explored the effects of arbitrators’ characteristics on outcomes and found the predicted effects, we turn now to the effect of party capability on making the strategically best decision when appointing arbitrators, and we begin with appointing experienced arbitrators (Hypothesis 3). Table 3 presents models on the effects of party capability on the likelihood that parties appoint super-arbitrators as arbitrators. Like before, we consider the effects of each of the party-capability factors by themselves (models 1–4), as well as together. In line with our expectations, we find that international SGBs are significantly more likely than non-international SGB parties to appoint super-arbitrators. We cannot say whether this is because international SGBs are more well connected than other parties, that arbitrators prefer to be appointed by international SGBs, or if there is some other explanation. We do however know, as established immediately above, that this significantly improves international SGBs chances of winning. Non-international SGB parties do however appear to be able to mitigate the advantage that international SGBs have in this regard through increased representation quality: having a greater number of representatives and more experienced representatives are both positively correlated with a party appointing a super-arbitrator, even if we control for whether they are international SGBs.

Finally, we will explore whether party capability has an effect on parties appointing attitudinally aligned arbitrators (Hypothesis 4). We do this by generating two linear models, predicting party-arbitrator ToD alignment and class alignment respectively on the basis of our four party-capability factors. In line with our expectations, we find that high-capability parties are more likely to take advantage of the strategic opportunities that the party-appointment system offers when it comes to appointing attitudinally aligned arbitrators. However, this only holds true for some party-capability factors. As we can see in Table 4, a party being an international SGB is positively correlated with the party appointing an arbitrator that is attitudinally aligned under both dimensions. The same is true for parties that have a greater number of previous appearances, but not for parties with a greater number of representatives or more experienced representatives. Thus, unlike when it comes to appointing a super-arbitrator, it seems that non-international SGBs cannot compensate for the advantage of international SGBs when it comes to appointing attitudinally aligned arbitrators by hiring additional or more experienced lawyers. Finally, it is interesting to note that this is the only context where we find that a party themselves being more experienced, that is, having appeared in more previous cases before the CAS, has a significant effect. A possible explanation for this finding, which was also advanced above, is that such parties have access to more data points in the forms of awards and observed behaviour on which to base their pick.

Having made these findings, it is now possible to consider to what extent the litigation advantage of international SGBs (see Table 1) can be attributed to their increased ability to select attitudinally aligned arbitrators (see Table 4). The interesting question to consider is essentially this: would international SGBs litigation advantage over non-international SGBs exist even if there was no party-appointment system that they were able to strategically exploit? The models presented in Fig. 8 show that even when we control for party-arbitrator attitude alignment, international SGBs have a litigation advantage on the CAS. However, they also show that the size of that effect is reduced. This suggests that preventing international SGBs from appointing attitudinally aligned arbitrators by abolishing party-appointed arbitrators would help reduce their litigation advantage over other parties, albeit not eliminate it.

7 Conclusions

This study provides several important insights about party-appointed arbitrators. First, appointing an experienced party-appointed arbitrator significantly increases a party’s chance of a desired outcome of the arbitration. Moreover, the outcomes in previous cases that arbitrators have participated in can be used as predictors for the outcome in future cases that they participate in. This study thereby thus shows that appointing experienced arbitrators whose track records suggest that they are ‘harmoniously aligned’ with a party’s interests appears to be a successful litigation strategy. While these findings were expected and in line with the existing literature, being able to empirically demonstrate this improves our understanding of the ‘arbitrator appointment game’. The study does not support or dismiss the propositions that arbitrators exhibit appointer biases or tend to act in ways that benefit the appointing party. It does however show that it is not necessary for party-appointed arbitrators to be ‘hired guns’ in order to give the appointer a litigation advantage and provides empirical evidence in support of the appointed-arbitrator-as-sympathetic-friend thesis.

Second, this study has provided novel and important insights regarding the effects of party capability on the outcome of arbitration. The findings of this study suggest that high-capability parties—the ‘haves’—are generally better than low-capability parties—the ‘have-nots’—at exploiting the strategic opportunity that party-appointed arbitrators offer parties by both appointing more experienced arbitrators and arbitrators whose attitudes are more closely aligned with their own interests. Moreover, the study shows that low-capability parties cannot reduce the high-capability parties’ advantage simply by hiring more or better lawyers, but that eliminating party-appointed arbitrators would do so.

In addition to lending credence to the theoretical framework of the ‘arbitration game’ that this study builds on, these findings have some clear policy implications. Specifically, the findings raise serious concerns about the fairness and legitimacy of using party-appointed arbitrators in arbitration where there is a high degree of disparity in party capability or, to speak in Galanter's terms, in arbitration between ‘haves’ and ‘have-nots’.¹³⁷ While eliminating party-appointed arbitrators would not entirely eliminate the advantages that high-capability parties have in arbitration, this study has shown that doing so would place high- and low-capability parties on a more even playing field.

This study can speak most directly to the specific case study and the use of party-appointed arbitrators in the CAS. In the late 1990s, Menkel–Meadow observed that strong repeat players were increasingly establishing their own alternative dispute resolution (ADR) systems, such as arbitration, that by making arbitration mandatory they coerced low-capability parties into using these fora, and that this has given them ‘the ability to manipulate rules, fora, and decisionmakers’.¹³⁸ The CAS fits her observation, at least to some extent. As discussed above, there is frequently a large capability discrepancy between the parties on the CAS. Moreover, the substantive rules as well as the arbitration system itself are in the case of the CAS developed and governed by the highest-capability parties. Finally, as the European Court of Human Rights has established, the low-capability parties have not ‘freely and unequivocally’ accepted arbitration by the CAS whose jurisdiction must be regarded as compulsory.¹³⁹ The importance of lack of consent cannot be overemphasized.¹⁴⁰ Frequently repeated arguments in defense of party-appointed arbitrators focus on parties freely choosing such fora,¹⁴¹ and argue that the right to appoint arbitrators is ‘part and parcel of the parties’ autonomy’.¹⁴² These arguments weigh little when it comes to defending the use of party-appointed arbitrators in a compulsory arbitration institution, especially one designed by the same high-capability parties that compel low-capability parties to submit to its jurisdiction.

I fail to see how the litigation advantage that party-appointed arbitrators provide international SGBs (and other high-capability parties) in the CAS can be resolved by tweaking the existing system, which leads me to the conclusion that it ought to be abolished. I want to make clear that I am not arguing that the CAS itself—or sports arbitration more broadly—needs or ought to be abolished. It should however be significantly reformed. In this regard, I share the opinion of Giorgetti in that ‘[c]riticisms will allow the growth and betterment of the system’ and that ‘[a]s academics and practitioners, we should ensure the evolution of the species, not its extinction’.¹⁴³

The European Union has proposed that investor–state dispute settlement (ISDS) arbitration should be reformed so that it works more like national and international courts. This includes, in particular, replacing party-appointed arbitrators with ‘a single standing institution with permanent professional staff and with an appeal instance [which] is far more likely to rule more consistently and independently from the interests of disputing parties’.¹⁴⁴ The EU has also provided a concrete model for how such an institution can be setup through its Investment Court System (ICS).¹⁴⁵ I have elsewhere proposed that the present, appointment-based system on the CAS should be replaced with a system where a number of arbitrator–judges are appointed to serve with the institution itself for fixed periods of time and that these individuals be randomly assigned to serve on panels and resolve disputes.¹⁴⁶ We know from national and international courts that this design is superior for the purpose of ensuring the independence of the adjudicators. The findings of this study allign with and provide significant support for that recommendation.

I should finally address some of the limitations of this study. First, sports arbitration is a discrete field in which, as has been elaborated on above, party capability plays a special role and must be empirically operationalized in a field-specific manner. Second, the empirical study considers only appeals cases, which has methodological advantages but also leaves room for considering to what extent its findings hold true for other, more common procedural forms of arbitration. Both of these limitations should be borne in mind if one seeks to apply the study’s findings to other contexts, and future research should study whether and to what exent they apply to other arbitral institutions. Another interesting avenue for future research identified in this paper is the role of panel presidents.¹⁴⁷

Footnotes

Appendix

Author notes