A comparison between the standard of proof applicable in arbitration and formal adjudication

Introduction

Among the different mechanisms of alternative dispute resolution (ADR), arbitration, which refers to an impartial referee’s hearing and settling of a dispute between the parties concerned, holds a privileged place. The appearance of ADRs started, and became an alternative to formal justice in the US in the 1970s, and is currently being successfully implemented throughout almost every known jurisdiction with a relatively modern justice system (Twining, 1993). Long before the ADR idea was installed in legal culture as an alternative to formal justice, arbitration was considered both as part of the system of conflict resolution in civil and commercial matters, and as a complement to formal justice. Its first proper regulation by the British Parliament was the Arbitration Act 1889. Since then, it has gained general acceptance as a reliable alternative to civil litigation.

Nowadays, recourse to arbitration has grown exponentially in national and international disputes, when they incorporate an arbitral tribunal or an arbitrator as their dispute resolution mechanism. It should also be noted that arbitration is currently the most widely used conflict resolution mechanism in the world (Born, 2009). Many relevant cases are based in the United Kingdom and utilise English regulations as their applicable law, especially in international commercial disputes (Born, 2009: 1542). The reasons for the popularity of arbitration are its intrinsic procedural qualities, including freedom, de-legalisation and demystification, together with advantages over national jurisdiction, including speed, confidentiality, specialisation, flexibility, immediacy and greater involvement of the relevant parties. Considering the breadth of these features, arbitration is often described as the best form of dispute resolution (Andrews, 2013).

Domestic arbitration in England is regulated by The Arbitration Act 1996 (hereinafter ‘AA1996’). Furthermore, prior regulation was designed for the same purpose, which stipulated most aspects of arbitration activity, such as jurisdiction, applicable laws, initiation of the procedure, requirements of the award and its adoption before the court. In the quasi-judicial activity performed by the arbitrator (Emerson, 1970), there is a lack of regulations regarding evidence and proof, which is a shortcoming, given that these are some of the most relevant aspects of any adjudicative instance and decision-making process. In general, there are neither specific rules governing the approach to evidence in arbitration, nor particular norms telling the arbitrator how to proceed in these matters. In contrast to formal civil justice, where evidence is a well-regulated topic, both in statutory and common law, the regulation on arbitration in AA1996, s. 34(2)f, assigns the arbitrator the responsibility of ruling the proceedings under which evidence and proof should be reviewed. Yet there is no mention made of evidence or fact-finding as a specific topic.

It is important to remember the steps, also described as stages, of evidence activity. First, during the inclusion or admissibility step, the judge must examine the relevance of the proof material offered by the parties and exclude irrelevant and illicit evidence (Lawson, 2017: 21). Secondly, once this depuration process has been undertaken, and during the trial, the evidence is presented before the judge (or eventually the jury), who will assess it. Thirdly, they will be called upon to apply a given standard of proof in order to make a decision on the case (Taruffo, 2008).

Throughout this essay, we will review one of the most significant topics in evidence law, namely, the standard of proof. Standard of proof can be defined as a probabilistic judgment, whereby the adjudicator concludes that an allegation of fact made by a party is proven (Allen and Stein, 2013) and, accordingly, applies the proper law to resolve the conflict. The applicable standard of proof for civil and commercial cases in formal litigation is clearly established in England. Due to the increasing use of arbitration for the resolution of similar types of disputes, it seems necessary to address the question of whether it is consistent that the arbitrator applies the same standard to arbitration taking place in England and ruled by the AA1996. Furthermore, this query might lead to a deeper issue, i.e. whether the rules concerning standard and burden of proof are part of the public policy of English law and, if so, mandatory even for informal methods of adjudication, or, on the contrary, whether arbitration, as a private process, is not bound to those rules. Within this context, the arbitrator, within his/her autonomy, would be allowed to use different standards, according to the subject matter of the dispute.

This essay will argue that the standard and the burden of proof might be considered a procedural guarantee for the defendant, as rules under which the law distributes the risk of error among the parties, by settling: a) which party must bear the obligation to provide enough evidence, and b) what threshold is required to consider a certain fact proved (Roberts and Zuckerman, 2010). We will argue that these guarantees can be interpreted as part of a wider set of procedural rights and principles enshrined constitutionally as due process or procedural fairness. Due process undoubtedly applies to formal litigation, and some of its basic aspects have been recognised as necessary in arbitration (Paulsson, 2010). These aspects, according to Park, would be ensuring that both sides are heard: impartiality and independence (Park, 2011: 26–27).

Although some scholars assert that the function of each of these processes is determined by the type of conflict, and the fact that arbitration—due to its private nature and the lack of a constitutional function in the vindication of rights of individuals—should not be considered a mirror for civil litigation, we argue that the whole arbitration process must respect at least some aspects of the aforementioned procedural fairness. Within this context, the standard of proof should be guaranteed for the relevant parties to know which arguments are accepted and/or rejected in the claim.

As a methodological explanation, this article has undertaken a comprehensive review and a critical analysis of the relevant literature regarding the topic of evidence and arbitration, in order to provide a justified proposal of theoretical framework as well as an eventual change to the AA1996 as a lege ferenda solution.

Background on standard of proof and civil process

The standard of proof has been defined as a rule whose function is to indicate the party bearing the burden of proof. In other words, it pertains to ‘how much cogent or convincing the evidence must be to conclude that a fact actually happened’ (Murphy, 2008: 101). The Supreme Court of The United States has held that the living motive of a standard of proof is ‘to instruct the fact-finder about the degree of confidence our society thinks he should have in the correctness of factual conclusions for a particular type of adjudication’. ¹ This statement highlights that the standard should fulfil an extra-process function, which is to point out the height of the threshold for rightfulness in the fact-finding process required within the justice system. In summary, the standard of proof relates to the level of proof that society requires to consider that a practice actually occurred (Haack, 2014).

The standard of proof establishes whether a hypothesis can be declared proven according to the available evidence. The election regarding the standard height in different kind of proceedings is said to be both a political and judicial decision. In England and Wales, this has been conceptualised by Lord Hoffman in Re B (Children) (Sexual Abuse) as follows:

If a legal rule requires a fact to be proved (a fact in issue), a judge or jury must decide whether or not it happened. There is no room for finding that it might have happened. The law operates a binary system in which the only values are 0 and 1. The fact either happened or did not. If the tribunal is left in doubt, the doubt is resolved by a rule that one party or the other carries the burden of proof. If the party who bears the burden of proof fails to discharge it, a value of 0 is returned and the fact is treated as not having happened. If he does discharge, a value of 1 is returned and the fact is treated as having happened. ²

The evidentiary procedure that culminates in the application of a standard of proof to decide in a legal situation is no different from the decision-making process in any other field of knowledge (Laudan, 2006). In this order of things, the evidentiary standard will be used, in simple words, to establish when a factual statement is justified or should be considered true according to the evidence provided during the process. At this point, and after a detailed analysis of evidence with the truth and falsity of an affirmation, it should be determined only in accordance to the latter whether the background is sufficient to accept or reject the claim, that is, independent of the possibly applicable legal rules. Hence the standard must be applied regardless of whether or not the person entitled to decide has any legal training

The standard division in English law is twofold: the balance of probabilities and beyond any reasonable doubt (Zuckerman, 2013). The latter has been established for criminal cases, where the scale of persuasion is nearly 1, whereas the former is utilised in civil justice as the general rule, where the threshold is lower and is met at any value over 0.5. The difference between both standards is justified in the distinction of the values and public concern regarding the legal rights at stake. The standard is understood as a rule for the allocation of errors, of false positives and false negatives, whereby the parties bear the risk in the case where the barrier established by the standard is exceeded (Laudan, 2006). In criminal process, the risk for the accused is his/her liberty and in some cases his/her life—the ultimate possible values. As the criminal process concerns a guarantee for the correct and proportionate exercising of those rights, a higher standard as part of those guarantees is needed, considering that the risk of a false positive is the imprisonment of an innocent person (Laudan, 2006: 14). Likewise, in civil and commercial cases, litigation engenders economic consequences, where the risk borne by the losing party is the payment of damages or the fulfilment of specific acts, and in a cost-shifting system such as the English system, the costs for litigation.

The general rule and exceptions in civil and commercial matters

As argued above, the standard in civil cases is the balance of probabilities. This was established in LJ Denning’s Miller v Minister of Pensions ³ (which highlighted that the standard is met when ‘we think it is more probable than not’ (Hannibal and Mountford, 2002: 457). The aforementioned has been reaffirmed by the common law over the years, even in relatively recent judgments such as Re H (Minors) (Sexual Abuse: Standard of Proof), ⁴ where Lord Nichols completed the definition by declaring that: ‘The balance of probabilities standard means that a court is satisfied an event occurred if the court considers that, on the evidence, the occurrence of the event was more likely than not’. That is to say, is the suppletory standard in the civil cases.

Although this threshold has been established as the default standard, it can be criticised, on the basis that the level of probability that is posed is too low. In other words, to come to a decision we should not only abide, in some cases, by one more piece of evidence from the opposing party or a lack of evidence from the latter. In these cases, the most probable version is not necessarily the truest one (Taruffo, 2008). In addition, each type of proceeding provides a different protection framework depending on the values at stake—the more important the value, the higher the standard, and vice versa.

In accordance with the aforementioned, exceptions are warranted. The first exception occurs when—within civil proceedings—committal proceedings for civil contempt of court are undertaken (Hannibal and Mountford, 2002). Hence, due to the quasi-criminal nature and criminal-like consequences of the contempt—liberty privation—the standard must be raised to the criminal parameter, beyond any reasonable doubt, as stated by the Court of Appeal in Dean v Dean. ⁵

Additionally, there are always explicit exceptions in statutory law. The Royal Warrant Act 1949 may be mentioned as one noteworthy case. In Judd v Minister of Pensions and National Insurance [1966] 2 QB 580, it was explicitly mentioned that, in the aforementioned Act, Parliament introduced the necessity of a standard beyond any reasonable doubt to prove that the evidence was insufficient to effectively support the applicant’s claim (Hannibal and Mountford, 2002).

In the case that a civil proceeding deals with conduct that could lead to criminal offences, such as fraud, the standard remains in the balance of probabilities, as now has been clearly defined by the House of Lords in Re B (Children) (Care Proceedings: Standard of Proof) ⁶ and confirmed by the Supreme Court in Re S-B (Children) (Care Proceedings: Standard of Proof). ⁷ This approach demonstrates that one of the most important issues that is taken into account in the elaboration of a standard of proof is the seriousness of the consequences of the proceedings and not the seriousness of the allegation.

Arbitration and evidence

Arbitration in England and Wales has mostly been considered as a private matter (Brunet, 1992). By entering into an arbitration agreement, the parties involved submit the dispute to an impartial third person, which is entitled to solve it using broad faculties to conduct the proceedings (Brunet, 1992). As a general rule, evidentiary law is not vastly considered in this regard, by either the legislation or the parties, given the fact that one of the main features of arbitration relies on its flexibility and the possibility to avoid over-strict rules of proceedings. Thus, arbitrators practising their functions usually decide on the main evidence and proof matters, such as exclusion and admissibility, without specific or clear rules. In addition, they must establish a procedure for the gathering of evidence, which might or might not be similar to a disclosure process in formal litigation. Finally, they must weigh up and assess all the evidence at their disposal to make the final decision.

As part of the aforementioned process, the arbitrator must deal with more abstract rules of evidence. In doing so, he/she will have to apply rules for bearing the burden of proof, and, as a consequence, the rules of the standard, to determine whether the burden has been discharged. It is not our purpose to argue that the arbitrator should apply the rules of burden and standard contained in the law, or to suggest that he/she should use, necessarily, rules equivalent to those that the civil court would use when hearing the case. Nonetheless, inside the adjudicative process, it seems to be unavoidable for epistemological reasons—as part of the decision-making process—that the adjudicator will assign a burden to one of the parties and apply a given standard, although he/she may not explicitly mention it (Scheinman, 1977). Otherwise, the arbitrator could not arrive at a rational conclusion regarding the veracity of the facts proposed by the parties, to subsequently apply the proper law to those facts.

Nevertheless, we believe that the application of the standard of proof is independent of the possibility of applying the rules of burden of proof in the event that we are not before a plaintiff and defendant. In this type of case, the judge will also have to make a decision regarding a certain narrative, and therefore, background information will be introduced into the process in order to prove it. Taruffo (2008) has pointed out that there is no real subjective burden of proof. Hence, a decision must be made and an evidentiary standard applied.

The domestic arbitration process, governed by AA1996, entrusts the arbitrator with significant powers to deal with the proceeding and evidence. Some scholars define the arbitrator as the ‘master of his own procedure’ (St John Sutton, 2015: 242). Yet the arbitrator’s powers can always be restricted and the rules shifted by agreements between the parties involved, as stated both in s. 34 of AA1996, which declares that: ‘It shall be for the tribunal to decide all procedural and evidential matters, subject to the right of the parties to agree any matter’; and letter f) of the rule, which adds: ‘whether to apply strict rules of evidence as to the admissibility, relevance or weight of any material…’ ⁸ . This approach, as Russell notes, was taken while the Act was being passed and it was decided to exclude formal and sophisticated doctrines of evidence law in order to facilitate arbitration (St John Sutton, 2015), unless they are expressly introduced by the parties. Otherwise, they fall into the scope of the arbitrator’s powers, as established by s. 34(2)(f). The courts have adopted a rather broad approach to evidence in arbitration, limiting their intervention to the application of s. 68 of the Act, and thus confirming the approach outlined by the statute, which entrusts arbitrators with the duty of control of evidence and proceedings. ⁹ As a general rule, with the exception of a major breach of the duties of the arbitrator, which would constitute a serious irregularity, the court has declared consistently that arbitrators should ‘rule their own evidence’. ¹⁰

In comparative jurisdictions, a more court-like approach has been taken. The American Arbitration Association has introduced a very similar rule in its rules for Commercial Arbitration, which states that ‘The arbitrator shall determine the admissibility, relevance, and materiality of the evidence offered and may exclude evidence deemed by the arbitrator to be cumulative or irrelevant’. ¹¹ As an illustrative comparison, within international arbitration, there are no such clear rules for the gathering and assessment of evidence, mainly due to the influence of the different legal traditions taken into account when drafting the international instruments of arbitration tribunals (Pietrowski, 2006). Nonetheless, some main principles have been established to provide some basic ruling in the arbitration process, regardless of the jurisdiction where the arbitration has its seat. These principles tend to permit, as a general rule, a relative freedom to the parties to submit any kind of evidence they consider necessary to support their case, and a wide discretionary power to the arbitrator to assess that evidence. Furthermore, arbitrators do not face strict rules of admissibility and the allocation of the burden of proof necessary for the verification of their respective cases (Pietrowski, 2006).

Notwithstanding the above, some scholars have argued that the differences between common law and civil law traditions engender some material disparities in arbitral proceedings. Indeed, although the rules might be the same, the legal background of the arbitrator, the counsels and the seat of the arbitration will undoubtedly affect the application of evidence rules. Furthermore, the inquisitorial and/or adversarial-like approach of the arbitrator can have an impact on proceedings (Blackaby, 2015).

Considering this, some international efforts have been made to provide a more unified legal framework for arbitration. Thus, the UNCITRAL model regulation has established the possibility to use strict rules of evidence, but always enables the arbitrator to apply them in his/her own way, as stated in Article 27, 4. ¹² A similar rule can be found in the IBA Rules for Evidence in International Arbitration. ¹³

Arbitration and standard of proof

Having provided a brief introduction to the issues arising from evidence in arbitration, this essay will now analyse the role that the standard of proof plays. As mentioned above, in general, arbitrators do not specify the evidence rules they consider in their adjudicative processes. Moreover, although the rules establish that it is up to the parties involved to propose the principles or rules that are going to govern the arbitration proceedings, this agreement can be very hard to execute. This is particularly relevant in complex cases with parties that present an extremely adversarial approach to arbitration, unless some rules related to the proceedings and the evidence are incorporated into the arbitration agreement (St John Sutton, 2015). Yet that level of specification is hardly seen in the agreements and it is rather common to see parties coming to hearings without any clear idea of the proceeding they may like to be implemented (Goske, 1959). The above, in accordance with the minimum guarantees of a trial, is decisive, considering that the evidentiary process and, especially, the rules regarding the final resolution, allow the parties to control the judge’s decision.

AA1996 does not mention any specific standard rule under s. 34(2) of the Act. The power to decide rests in the hands of the arbitrator. In this context, it is well known that the arbitrator will be influenced by his/her legal background and training, being tempted to apply a similar rule in court (Pietrowski, 2006). This is particularly the case if, as is common in England, the arbitrator is a retired judge or an experienced barrister (Atlas et al., 2003; Sanders, 1999). In accordance, an English arbitrator will tend to apply roughly analogous rules to the common standard for civil and commercial cases within the English Civil Justice system (Beck, 1949), which, as noted, consists of a balance of probabilities (Pietrowski, 2006), even though he/she might have been appointed for the resolution of a dispute that does not involve English parties, who are not so familiar with that standard. Thus, in international arbitration, the tendency to apply the standard of civil evidence, despite the matter involved, leaves open the question of whether, within an arbitration procedure, the type of conflict submitted should be distinguished independently of whether or not it is an arbitration procedure.

It is commonly argued that one of the most significant features of arbitration is the flexibility that it provides, compared with formal civil litigation. It could be said that arbitrators have an opportunity to extend that flexibility in connection with the standard of proof. Hence, the arbitrator—who is not compelled by the court’s precedent in this regard (Roberts and Palmers, 2005)—could move the standard of proof away from the rigidness of Civil Justice adjudication.

Consequently, arbitrators could use a higher or lower standard than the court, depending on the complexity and sensitivity of the allegations and the seriousness of the allegations against the grievant (Scheinman, 1977), e.g. fraud and criminal conduct, which have been debated in the court, and from which conclusions have been already examined early on this article. Therefore, in some cases, e.g. where the burden of proof is not clearly assigned, its discharge might be met with a lower quantum of proof, or, on the contrary, the threshold could be risen.

Relevance of the standard applied

As stated previously, the standard for civil cases in court is the balance of probabilities. The balance of probabilities entails a rule of an equalitarian distribution of errors between the parties, taking into account that the cost of those errors is distributed equally between them (Porat and Stein, 2001). Thus, the rule that establishes a standard of proof is not neutral (Porat and Stein, 2001), from a political or a moral perspective, being, by consequence, an eminently normative decision (Tuzet, 2019).

The standard of proof, ergo, fulfils a function in the equalitarian regulation of the adjudication process and, in doing so, carries the due process principle through the equality of arms between the parties, providing equal grounds for both of them when facing civil litigation. A standard that normatively distributes the probabilities on a 50% basis for each party promotes the same prerogatives for the claimant and the defendant, allowing the process to be conducted, pursuant to procedural fairness standards.

For its part, arbitration has been described as a mechanism that concerns an adjudicative decision-making process, which is final and binding, in which, when taking that decision, the adjudicator must, necessarily, use a standard of proof. Arbitration is a more flexible form of litigation than formal justice, which means that this mechanism permits the adaptability of the standard rules, either increasing or decreasing the threshold. With this modification, the formal equality that a balance of probabilities standard provides to the parties could be altered in arbitration, inclining the balance towards one of them, and away from the symmetry of its civil justice counterpart.

Another relevant issue that might arise related to the default function of the balance of probabilities is challenging the award. The problem relies on whether the fixation of a specific standard could tend to increase the arbitral awards to be challenged under s. 68 and s. 69. The response might seem unclear, yet our position is that every decision-making process involves the use of a standard of proof. Accordingly, arbitrators use a standard when they deliver their decision. Thus, the standard is actually being used, yet it is not clearly known which standard. That is to say, epistemically the decision needs a quantum of proof, which often remains uncertain. Should the standard of proof be clarified, it would not necessarily lead to more challenges than currently. It would, however, lead to more controllable challenges, founded on a specific normative threshold whereby the Court could review it more thoroughly.

That being said, an important question still needs to be asked: according to the contractual and private nature of arbitration in England, should the alteration of the equality of arms to which the parties are entitled, with regard to the standard, be allowed or not, from the perspective of the minimum principles of due process which, as explained, participate in arbitration?

Due process principles, as Flick (1984) describes, apply to all public decision-making processes, either judicial or administrative. In addition, in England, the court has stated that natural justice is applicable to any subject, public or private, acting in a quasi-judicial role, ¹⁴ which includes private adjudication. As Hörnle (2009: 98) points out: ‘rules of natural justice not only apply to public bodies or bodies exercising a public function but also to arbitration’. Furthermore, in s. 33 ¹⁵ the AA1996 acknowledges the necessity and applicability of procedural fairness to an extent that is not very common in arbitration enactments, a feature that Carbonneau refers to as a ‘concise and compelling description of the purpose and rationale of arbitration’ (Carbonneau, 1998: 146). The act enshrines what it denominates a ‘general duty’ to the arbitrator, a concept that includes an order to ‘act fairly and impartially’ and give each party ‘a reasonable opportunity’ to present its case and to respond to the opponent ¹⁶ . Moreover, the aforementioned s. 33 must be observed in accordance and correlation with s. 1 of the Act and, particularly, with s. 68, that allows the challenge of the award on the basis of a ‘serious irregularity’, which explicitly incorporates ‘(a)failure by the tribunal to comply with s. 33 (general duty of tribunal)’.

Some scholars have gone as far as to say that due process must be applied to arbitration to the same extent as in civil litigation, arguing that arbitration cannot avoid the applicability of Article 6 of the ECHR (Jaksic, 2007). However, other experts have stated that the applicability of due process is direct only in cases of compulsory arbitration, where the state, through statute, might be delegating a public function in private hands (Jaksic, 2007: 103). The extent of this application remains unclear, yet the more basic principles seem to be the starting point: a sort of ‘measured’ due process in arbitration (Brunet, 1992). In this regard, the minimum principles that should apply are those embracing what Andrews denominates to be a ‘fair process’ (Andrews, 2013: 687), including a ‘fair hearing’ which comprehends the right to equality of arms, the right for an equal exhibition of evidence and equal access to information. ¹⁷ Additionally, other basic principles that should be applicable are judicial independence and impartiality ¹⁸ .

This discussion becomes even more relevant when arbitration is mandatory. In these cases, parties are obligated to enter into arbitration, although they might not consider this to be the best option to resolve that particular dispute. Consequently, they will solve their differences through the utilisation of rules that are—or at least could be—unknown to them, and to which they remain oblivious. The risk arises especially in cases of unrepresented parties, where the arbitration is conducted without counsel advice, or in self-representation situations, where the procedural fairness guarantees rely almost exclusively upon the arbitrator’s criteria. The scenario becomes even more complex when there is a marked asymmetric relationship between the involved parties, e.g. labour law disputes. This lack of balance in the negotiation power can be reflected in knowledge of the rules to be accorded or proposed before the arbitrator, or worse, in the mere fact of forcing the execution of the arbitration agreement, taking advantage of the fear of the weak party towards civil litigation delays and expenses (Hill, 2003), thus affecting the equality of arms through a breach in equal access to information.

If the arbitrator does not establish an objective and fixed standard of proof in the evidence rules prior to the commencement of the proceedings—or to the gathering of the evidence stage at the most—the parties would have to face the uncertainty of the limit that their propositions of fact would have to overcome (Andrews, 2013). Moreover, this uncertainty might depend on, or increase due to, the depth of the regulation regarding disclosure the parties had or had not agreed to during the arbitration process. Thus, as regards the importance of disclosure to the proceeding, if the parties have drafted an agreement that includes at least a certain type of disclosure-like evidence stage, it would be more likely for them to prepare and foresee evidence-related issues, being aware of the standard of proof. Additionally, they would provide more proof, particularly documents, for the arbitrator to arrive to a factual decision that overcomes the threshold fixed by balance of probabilities. Conversely, when the parties have remained silent on the issue of disclosure, experience shows that the arbitration would try to avoid over-structured evidence proceedings, such as disclosure or discovery (Guys, 2018), and consequently increase the uncertainty due to the doubt the parties might engender, including the standard of proof.

Furthermore, a risk of critical imbalance could arise if the arbitrator additionally uses a variable standard, depending on the seriousness of the allegation. For example, in medical negligence claims, the arbitrator may understand that the proof of the negligent behaviour of the medic requires a standard of proof above the 0.5 mark, and move it very close to the criminal standard of beyond any reasonable doubt, forcing the claimant to exceed a very high barrier. Conversely, if the arbitrator decides that the standard of proof for the same facts should be fixed under 0.5, the claimant would solely need to be more convincing than the defendant to obtain a favourable award, leaving the latter in a worse position, as a consequence of diminishing the threshold.

The situation in this example could lead to even more complex consequences. As Grando (2009: 129) suggests, a lower standard of proof could lead to bias towards claimants, affecting the impartiality of the system. Claimants file their cases first and, very commonly, their evidence. Therefore, if the standard is set very low, the arbitrator could be easily convinced and, eventually, transfer the burden of proof to the defendant (Grando, 2009). The aforementioned would contribute to a major breach in the essential rights of the parties in arbitration, even if considered as a private matter, because one of the parties would be dealing with an unforeseeable lack of impartiality that could not have been waived. Finally, a very low standard could lead, as Porat and Stein (2001) correctly argue, to the risk of aleatory results, making arbitration more closely resemble gambling than civil litigation.

For the authors of this essay, if the standard of proof, as an important aspect of evidence matters, can be located among due process guarantees and due process applied to arbitration—although to a limited extent—then civil standard of proof rules should apply, in the absence of an agreement between the parties, as a non-mandatory guidance for the arbitrator, as part of a suppletory minimal due process framework.

Conclusions

Throughout this essay, we have revised the nature of the standard of proof, its function in the decision-making process, and its application to civil litigation. Additionally, we have explained the standard of proof in civil justice, and dealt with the application of more flexible standards. Subsequently, we have investigated the applicability of the standard of proof in arbitration, making the statement that it does apply, as it does in any decision-making process. We have described which standard could be applied by the arbitrator pursuant to the contractual nature of this dispute-solving method, and the relevance of that distinction. In addition, we have highlighted the role of the equality of arms and other principles of due process that, as explained, must be taken into account in arbitration, due to its adversarial, adjudicative and quasi-judicial scheme. Accordingly, the standard of proof used by arbitrators in England and Wales under AA1996 should be addressed as follows:

As the main rule of arbitration, the parties should have control over the proceedings and evidence rules. This remains as the general rule enshrined in AA1996 s. 34(2), as arbitration has a predominantly private nature, and this rule maintains the supremacy of the Party Autonomy Principle. The parties should be free to agree to any standard, including standards that are not expressly recognised in domestic law and that may be higher or lower than the balance of probabilities.

Being aware that the agreements mentioned above are scarcely used, the focus should be fixed, where the parties do not expound any preference related to proceeding rules. In this context, the arbitrator should use the authority delegated by AA1996, to establish the evidence rules. The arbitrator should propose and determine the standard to be used to consider the burden of proof discharged, before any proceedings have commenced, and should not change it throughout any stage of the process, ideally in absence of an agreement with the parties. The proposed standard should be the balance of probabilities, whereby the risk of errors and possibilities of success are allocated equally between the litigants. In accordance with that, we indicate that the default function of balance of probabilities should be understood as a consequence of the symmetrical position of the parties to the arbitration, and in doing so arbitration schemes, whether domestic or international, should transit to the enshrining of this rule. On a domestic basis, the rule could be delivered by courts when revising arbitration cases, as well as Arbitration Centres’ guidelines. Internationally, it will depend on the legal traditions and background of each jurisdiction regarding arbitration. Thus, it could be enacted, for those countries with a rigid legalistic tradition, or case law constructed or suggested in more common law-like legal backgrounds. Nevertheless, in an effort towards unification, the international conventions, as the UNCITRAL Model Law, could endorse its inclusion as a method of setting the ground for some of the most relevant evidence issues of arbitration.

Therefore, regarding English domestic arbitration, considering that many parties select England as an arbitration destination for its flexibility as enacted—and as a well-known incentive in order to facilitate arbitration—in AA1996 the solution provided by the authors relied on a twofold possibility. First, as aforementioned earlier, through a case law doctrine regarding the establishment of a default rule of standard of proof in arbitration. Notwithstanding the fact that this would be the more appropriate approach, the limited number of challenges to arbitral awards could lead to a slow improvement, especially considering the non-binding effect of courts’ decision to consequent arbitral decision. Secondly, and in a attempt to tackle the difficulties of a case law-based solution, an amendment to AA1996 could be introduced, in order to provide a supplementary framework for evidence issues such as standard of proof, yet always considered as a non-mandatory rule, subject to parties and arbitral modifications according to the general rules of the Act. The same result could be achieved by enacting a provision encouraging a more detailed agreement on evidence matters at the very first stage of arbitration. Thus, through a limited scope modification, the necessity of a default rule of standard of proof would be compatible and would not trump the original ethos of the Act, namely, to attract domestic and international arbitration work.

Moreover, we think that the decision about the standard of proof should be taken as part of a wider set of rules, including explicit mention of the principles of due process to be considered in the arbitration. Notwithstanding that the basic principles of due process should always be applicable to arbitration, the express mention of them would provide clarifying guidance for the relevant parties, which is particularly significant in cases that involve litigation between private entities.