home page

Law Firms   ::   Tax Consultants   ::   Intellectual Property Law Firms   ::   Immigration Law Firms   ::   Property Law Firms


The Immunity of Arbitrators.
© Anastasia Tsakatoura, 20 June 2002. All Rights Reserved.


Ronald Dworkin, one of the most important legal philosophers of our times, in the preface of his book “Law’s Empire”[1] stated: “we live in and by the law. It makes as what we are: citizens and employees and doctors and spouses and people who own things… we are subjects of law’s empire, liegemen to its methods and ideals bound in spirit while we debate what we must therefore do.” This has long been one of my favourite phrases, justifying why legal studies are so important. If we attempt to adjust it to the context of international commercial arbitration there is one question that crosses ones mind. Since we are all “subject’s of law’s empire”, how come and arbitrators are immune of legal liability in England, the United States and other common law countries or by the regulations of some arbitral institutions such as the ICC?

The key concern of this essay would be to investigate the underlying principle of such immunity as arisen by case law and supported by legislation and various commentators. We will attempt to evaluate the arguments for and against the immunity and conclude on whether it is desirable to sustain such a principle. However, due to the restricted size of this essay we will not scrutinize in details the legal relationship between the arbitrator and the parties, neither shall we analyze the duties owed by the arbitrator to the parties in depth. Furthermore, since this essay does not refer simply to English domestic arbitration but is extended to the field of international commercial dispute resolution it seems to be essential to refer briefly and without a very deep analysis to some other jurisdictions so as to get a more integral idea of the subject. Also, although there will be some reference to theories concerning the nature of the arbitration as a general concept, this analysis shall not be concentrated on the theoretical evaluation or re-evaluation of those theories and will only cite them as the jurisprudential background of the existing attitude towards the immunity of arbitrators in various common law systems or the opposite view taken by civil law systems.

To begin with, it appears to be useful, to provide with some information for the historical background of arbitration and the evolution of the concept in the contemporary reality.

Historical Background

In the past, when arbitration was not as commonly used a dispute resolution regime, as it is today, the appointment of a professional to act as an arbitrator by one’s peers had been regarded to be an honour for that person. It was an indication of his standing in the community, a sign of his success in the field of his profession. The remuneration of such arbitrators had been more of a nominal compensation of the time they devoted to the case at issue, rather than a fee for the services they provided. Arbitrators were not professionals; their livelihood did not depend on the fees they received for stating their expertise opinion on the subject matter of the dispute. “Arbitrations were conducted mainly during lunch hours and in the evening after office hours” as a practicing arbitrator Manfred A. stated in one of his article[2]. Back then, the issue of whether an arbitrator should be accountable of his conduct was hardly a matter for consideration. The existence of immunity was somewhat axiomatic and “the question of enforcing remedies against an unsatisfactory arbitrator would rarely arise.”[3]

Conversely, this position has changed during the years. In the complex contemporary society with the arbitration assuming a role of enormous importance in dispute resolution and unprecedented power, the situation is hardly self-evident. Still, it is considered to be an honour for an arbitrator to be selected to resolve a particular dispute, but in most cases arbitrators are full-time professionals. A number of institutions have been established so as to regulate commercial arbitrations more effectively and the fees charged to do so are barely nominal. In the contrary, the cost of such a procedure is rather substantial and the expectations of the parties for the skill and experience of the arbitrator or the arbitral tribunal are considerably high. In many occasions the parties expect that the arbitrators owe them a duty of diligence and that they will act accordingly. Merely, that an incompetent arbitrator can be removed form his position or have his award nulled does not compensate for the loss they might have suffered until then[4]. The question arising seems to be rather crucial: why then should arbitrators be excluded from liability towards those parties?

Rationale of the immunity

By using the word international commercial arbitration one might expect to find a uniform international law regulating this field. However, such a uniform law does not exist with the sole exception of the UNCITRAL Model Law, which has only been adopted by a limited number of nations and in general does not deal with the matter. Therefore, since international instruments are not coherent or sometimes are even silent on the question of the arbitrator’s liability, the legal rules regulating the immunity of arbitrators are mainly depended upon the different legal systems and much still remains to be determined by the area of law designated as ‘conflict of laws’ and of private international law. It is not the purpose of this essay, however to inquire into questions raised by private international law but merely to attempt and provide for a basic investigation of the approach taken by common law and civil law countries in general towards the issue of immunity of arbitrators.

Jurisprudential theories as a feedback to immunity

There is a vast literature in analyzing, contrasting, developing and contemplating those theories in regard to arbitration and its nature. However, as we have already mentioned is not possible to get into much a detailed analysis in this paper. It seems though to be crucial to quote them so as to roughly get an idea on wherefrom different jurisdictions derive their attitude towards the arbitrator’s liability.

During the years three individual analyses became apparent vis-à-vis the juridical nature of arbitration, namely, the contractual, the jurisdictional and the mixed or hybrid theory. Subsequently, the autonomous theory had been added to those three.

Originally, the contractual theory evolved from Merlin’s[5] perception on the subject, as early as in the 19th century, where he supported that the arbitration agreement has the character of a contract. In few words, Adam Samuel in his book[6] summarizes the classical contractual theory as follows: “it is the agreement to arbitrate that alone gives the arbitrators the authority to make the award. They, in turn, in resolving the dispute, are acting as the agents or ‘mandataires’ of the parties.” According to that theory an arbitrator cannot be regarded as a judge since his function is not of a public character. The parties’ will is of paramount importance and his powers do not emanate from the state’s authority. This theory has been supported or criticized a lot in the subsequent years and many arguments were presented for and against it. However, the theory left unanswered many questions and it was rather vague in many aspects. Therefore even those who followed it modified it so as to embrace some of the elements left out by the traditional theory. Bernard, attempted to analyse the relationship between the arbitrator and the parties, and it is his analysis that more or less has been followed by civil law jurisdictions. According to him, the arbitration agreement is still a contract but which is determined by special rules. The arbitrator’s authority stems from the agreement of the parties, which have the right to limit it in time. Nevertheless, the arbitrator is not considered to be an agent of the parties, since the duty determine the mutual obligations of the parties cannot be fitted to that of the agent’s duties. He went on by defining the nature of this ‘special contract’ as “a contract sui generis, governed by the rules appropriate to it and which must be dealt with by taking into account both the principles of the contract and the particular nature of the function exercised by the arbitrator[7]

The jurisdictional theory has been divided into two different schools of thought as well. On the one hand are those who regarded the arbitrator similar to a judge and the arbitral award ‘an act of jurisdiction’ (judgment theory)[8] while, on the other hand where the arbitrator is performing a public function as a temporary judge (delegation theory). The jurisdictional theory has been the one adopted by jurisdictions that support the immunity of the arbitrators and we will further examine it later in this essay when looking into those jurisdictions.

The mixed or hybrid theory encompasses a mixture of the above two. The commentators who support it believe that the reality lies somewhere in the middle of the contractual and jurisdictional theory, namely, that neither the arbitrator performs a legal function nor that the award is a contract. “The parties, by their agreement, created and fixed the limits of their private jurisdiction[9] The arbitrator’s duty is to judge but the power to do so is conferred to him by the agreement of the parties.

Lastly, the most recent theory is the autonomous one, whereby the main idea was roughly that arbitration should be detached from all the above theories and acquire an autonomous character. As described by Rubellin-Devinchi: “in order to allow arbitration to enjoy the expansion it deserves, while all along keeping it within its appropriate limits, one must accept that its nature is neither contractual, nor jurisdictional, nor hybrid but autonomous.[10] Nevertheless, this theory does not seem to have any practical application by any jurisdiction; therefore, even if its content seems to be appealing, technically, there are many problems in its appliance, because it implies a partial surrender of legal sovereignty.

Although these are in basic lines the main theories for the nature of arbitration that influenced consequently the attitude towards the arbitrators, other theories, variants of these, have emerged during the years and many jurists spent a lot of time and paper to analyse them. However at this point it seems vital not to insist on analyzing them but go on further to attempt to investigate what is the present reality technically in various different nations concerning the liability or immunity of the arbitrators.

Immunity in England and the USA

Common law jurisdictions that support the exclusion of liability for the arbitrators have traditionally based the justification for it, on the ground that arbitrators should be treated akin to judges.

In England as well as in the USA there is a long tradition of that reasoning hence, granting them immunity from suit in the absence of bad faith in England and; full immunity for all actions undertaken in fulfilling their duties as arbitrators in the USA. The new Arbitration Act 1996 in England by sec. 29 illustrates and sustains that tradition. The question raised is whether this presumed analogy with the judiciary is sufficient to justify the exclusion of liability.

In Bremer Schiffbau v. South India Shipping Corp. Ltd[11]. Donaldson J asserted that “courts and arbitrators are in the same business, namely the administration of justice”. Even though a number of commentators’ articles and statutory provisions relied on that assumption, there are many and not trivial differences between the two categories.

A starting point for comparison lies in the source of the power of the judges as compared to that of the arbitrators. A judge’s power derives directly from the State (or the Crown in the UK) and the general law of the nation. On the contrary, an arbitrator’s jurisdiction derives directly from the agreement of the parties. Secondly a judge is neither nominated nor remunerated by the parties, while an arbitrator is. As Lord Kilbrandon in a descending opinion in the seminal case of Arenson v. Arenson &Casson Beckman and Ruttey & Co.[12]asserted “the citizen does not select the judges in this system, nor does he remunerate them …the judge has no bargain with the parties…he pledges them no skill…his duties are to the state.” Taking into consideration all these disparities outlined, on what ground could such correlation be made? By applying the assumption that an arbitrator should be impartial and independent like a judge, and the judge in so doing should be immune of liability, commentators assimilated the two, and concluded that, thus, an arbitrator should be immune as well. In the leading case Sutcliffe v. Thakrah[13] this contemplation was underlined by Lord Salmon: “it is well settled that judges, barristers, solicitors, enjoy an absolute immunity from any form of civil action being brought against them in respect of anything they say or do in court during the course of a trial” and this presumption has been equally applied to arbitrator’s on the grounds of public policy.

As it has been argued an arbitrator “should be able to perform his task without constantly looking over its shoulder in the fear of being challenged through legal process.”[14] Nevertheless, it is rather self-evident that the reasons for the existence of the judicial immunity are significantly different than those supporting arbitral immunity. A judge shall not be personally accountable for the decision he reached in a particular case to the litigants, because, to put it straightforwardly he is supposed to be only accountable to the State. However, his decisions could be reversed or rectified upon appeal and therefore his duty to apply the law of the state should not be disrupted by a fear of personal liability. A judge’s independence should be shield against the chance of personal implication to litigation based merely to the dissatisfaction of a loosing party, so that the judiciary could effectively enforce the law of the state.

Pierre Lalive[15], an experienced arbitrator, in an attempt to rebut the presumption of the ‘judicial character of the arbitrator’ he distinguished between ‘public’ and ‘private’ justice. He defended that the term justice does not have an identical meaning to the two domains, and he went further by citing the words of Rene David[16], a French comparatist who believed that “when parties resort to international commercial arbitration, it is in general because they wish to have ‘another kind of justice’ “.

To sum up, it seems to be true what Mustill and Boyd argued in their book, namely that “the analogy between an arbitrator and a judge is tempting, but if pressed to far can lead to false conclusions”[17].

Before going further to assess and contrast the law regarding arbitral immunity in other jurisdictions, it seems to be essential to point out the provisions of professional liability in those jurisdictions and summarize the duties owed by the arbitrator to the parties.

The traditional approach in England has long established a requirement of reasonable skill and care to be exercised by professionals as regard their field of expertise. That could be materialized either by the law of obligations and contract or by tort proceedings. As Lord Kimbrandon stated in Arenson: “…the result, whether in tort or in contract, is the same[18]. However, in contract law the duties of the arbitrator towards the parties could be taken as an implied term to the, in-between them contract, while in tort a claim can merely be brought for breach of duty of care. In particular in tort the notion of professional liability has been affirmed and expanded after the two seminal rulings of the House of Lords in Donogue v. Stevenson[19] and Hedley Byrne &Co. Ltd v. Heller & Partners Ltd[20] where, by the former, a general cause for action has been established and by the latter, a cause for negligent misstatement has been founded. In the contemporary society where the laws of negligence have been expanded to embrace an ever-growing legislation on medical negligence it seems to be rather awkward why arbitrators should remain immune. As Tindal CJ asserted in Lamphier v. Phippos as early as the 19th century: “every person who enters into a learned profession undertakes to bring to the exercise of it a reasonable degree of care and skill[21]and indisputably practicing arbitration is a full-time profession which is also remunerated very generously.

There are, some commentators, therefore, such as Veeder J.[22] who supported that there seems to be no reason why arbitrators should be immune from liability in negligence, whereas other professionals are not, since, according to his opinion, ‘arbitrators provided what was surely a service as defined by the Supply of Goods and Services Act 1982’. Furthermore, arbitrators undeniably, have a range of obligations towards the parties to the arbitration. Those obligations include, the duty to take care, to act impartially and to proceed diligently. Nevertheless, in contrast with other professionals, those duties are unenforceable due to their immunity and consequently a suffering party might be left with no remedy.

The same line of reasoning on the issue of professional liability has been adhered to, in the USA. There, negligence is more objectively measured against the yardstick of the ‘reasonable competent member of that profession’[23] and professional liability has gained increasing recognition. Notwithstanding, arbitrators are granted full immunity even in cases that bad faith can be founded. This is obviously because the American courts do not perceive arbitrators as professionals when exercising their skills. They have strongly emphasized on the analogy of their functions and that of a judge’s. The justifications for that attitude seem to have the purpose of encouraging the continuing use of arbitration and expansion of its application.[24] The court in the USA indicated in a line of cases a growing interest in arbitration as a “desirable method of dispute resolution, particularly in international transactions[25] and therefore by granting immunity, they wished to induce individuals to “volunteer to arbitrate disputes” without having to be “caught up in the struggle between the litigants and saddled with the burdens of defending a lawsuit[26]. It seems to be rather unconvincing, though, why will liability deter arbitrators, who are paid large sums of money for a full time job and did not have the same effect on i.e. doctors and architects.

It is noteworthy, however that, while in the USA arbitrators have full immunity, in contrast, the Arbitration Act 1996 has limited that immunity to acts and decisions taken in bad faith.[27] Therefore, if judges are immune from suit for fraud, and English jurisprudence supports the analogy of judges and arbitrators, how come and arbitrators do not experience the same immunity? [28]

Other jurisdictions

In other jurisdictions, the situation is somewhat different than that adopted in the UK and the USA, especially in civil law countries, where the relationship between the arbitrator and the parties is mostly based on the contractual theory. In such countries, there is not any presumed analogy among the arbitrators and judges. Judges are public officials whereas arbitrators are solely professionals that may be held liable in damages where their wrongful conduct caused loss to the parties.

In France, there is not any legal provision concerning the liability of the arbitrators in particular. Judges and arbitrators are clearly distinguished. Hence, the arbitrator is fully liable for his acts. However, there is a precondition as to the admissibility of a claim for liability against an arbitrator: there should be no other remedy against the award.[29] This full liability is based on the contractual nature of the relationship between the parties and the arbitrator. The contract makes the arbitrator liable to the parties, if he fails to fulfill the terms of his reference. Furthermore, in tort, an arbitrator can be found liable under art. 1382 of the Civil Law, as against claims of third parties. Nevertheless, complexities arise regarding a bad decision included in the award.

In Switzerland similarly, the relationship is of a contractual nature, although it has been argued that in practice arbitrators are conferred a large degree of immunity.[30] In Austria, damages could be claimed by the injured party for any loss caused by intent, negligence or gross-negligence, where as in Germany, arbitrators are accountable for willful conduct and negligence. In Greece, by the very recent law in International Commercial Arbitration[31] which implements UNCITRAL Model Law, and articles 12 and 13 an “arbitrator may be challenged only if circumstances exist that give rise to justifiable doubts as to his impartiality or independence, or if he does not possess qualifications agreed on by the parties to the arbitration agreement”.[32]

Nevertheless, as Lew argued there seems to be no jurisdiction in which arbitrators are fully liable for any error of view or judgment in the decisions they reach. “Between, the state of absolute immunity and absolute liability, the state of qualified immunity arises in respect of acts performed by arbitrators in the exercise of their functions and of which they will or should exercise judgment and discretion within the terms of the law and the applicable arbitration rules. [33] Various, countries, such as Austria, Germany and Norway, have adopted that qualified middle way, whereas others, as France, Spain Sweden, shift towards greater liability.


Having presented briefly the main arguments concerning the legal relationship of arbitrators and judges in various jurisdictions and having examined the views of various commentators, many questions still remain unanswered. Should the arbitrators be completely immune from liability or should they be fully liable as any other professional? Is it preferable to follow the purely contractual and quasi-contractual approach or the judicial and quasi-judicial one?

As P.Lalive argues in his article, when the parties consciously preferred to resolve their dispute in arbitration rather than through litigation, they have knowingly accepted the underlying risks of such a procedure. Such risks as the finality of the award, the choice of arbitrators, however, do not seem to cover cases of fraud, corruption and gross-negligence. “Human nature being what it is, the irresponsibility of arbitrators, especially if absolute immunity applies, is likely to encourage the carelessness or negligence of arbitrators[34]. Is then preferable in the name of immunity to leave injured parties without a remedy in such cases? Public policy does not normally tolerate exclusion of liability in fraud. Why should this be the case in arbitration? A mere setting aside of the award or a removal of the arbitrator does not compensate for the time and money lost since “justice delayed is justice denied”. On the other hand, would it be desirable in the name of justice for the wronged, to harass arbitrators with unjustifiable claims that will undermine the finality of the award by having the case eventually re-litigated and which discourage people volunteering to arbitrate?

Many commentators have presented of several possible solutions to those hurdles. Mr. Li, an Australian jurist, proposed a model in which arbitrators would be obliged to join one professional institution. Arbitrators, in that model, could be immune from liability, whereas the institution would take the responsibility for the claims against them. In a case where they would be found liable, the institution shall take disciplinary procedures against them. The question is whether this ‘institutionalization’ undermines consequently, the ad hoc arbitration. Another objection to this proposal might be found in the contemporary tendency of arbitral institutions to exclude liability, not only on behalf of their arbitrators but also their employees (an example is presented by art. 34 of the ICC rules) [35]. Will it be possible to change the practice of the institutions so as to embrace the model proposed by Mr. Li?

Furthermore, other commentators have proposed the creation of an international court that could challenge the liability of the arbitrators among other matters relevant to arbitration. Whether this could be feasible in an international context and after taking into consideration the significant differences of various nations on the subject-matter, still remains to be investigated. It is essential though, in order for an international court to exist, a uniform law on arbitration to be adopted beforehand.

In my opinion, consequently, the most attractive solution seems to lie neither in excluding immunity totally, nor in restricting liability in full. The middle way of ‘qualified immunity’, whereas arbitrators are immune from unjustified claims (as judges) and could only be held liable in cases of fraud and negligence (as professionals), appears as a more of an ideal answer to the questions aroused. Yet, which of the four jurisprudential theories will back up such a model? Not the contractual, and jurisdictional, nor the autonomous seem to be adequate. It looks as if the mixed theory is more of application here. However, a mixed theory of a slightly different nature: having as a basis the jurisdictional theory and completing the inadequacies by the contractual theory. Hopefully then, neither the arbitrators nor the parties would feel insecure in practicing or preferring correspondingly, commercial arbitration as a dispute resolution.

back to Lex e-Scripta© Anastasia Tsakatoura, 20 June 2002. All Rights Reserved.

Add to Favourites  Publish an article  |  Free Listing  |  Advertise  |  Add URL  |  Currency Converter  |  Search

About Us  |  Contact Us  Links  |  Become an Editor  |  Terms & Conditions  |  Privacy Policy

Content & Design © 2000-11 INTER-LAWYER.com