The problem with arbitration agreements.

Arbitration procedures today have become highly standardized. Institutions such as the International Chamber of Commerce (ICC), (1) the London Court of International Arbitration (LCIA), (2) and the American Arbitration Association Center for International Dispute Resolution (AAA) (3) each have detailed provisions for administering arbitration proceedings (often involving parties of different nationalities). (4) Parties entering into arbitration can expect to have limited discovery, a hearing, and the ability to bring attorneys to the proceedings. (5) While typically providing less process than formal court proceedings, (6) the standardized nature of arbitration can lead parties to view arbitration much like court proceedings--a fixed, predetermined process to settle disputes. Thomas Carbonneau's article, The Exercise of Contract Freedom in the Making of Arbitration Agreements, reminds us of the contractual roots of arbitration. While arbitration has become standardized, parties retain the ability to vary aspects of arbitration. Carbonneau points out various ways in which parties may modify arbitration terms and makes the normative claim that sophisticated parties should affirmatively consider doing so. (7)

In Carbonneau's idealized world of contracting, parties would take into account a variety of factors in tailoring their agreement to fit the specific needs of the parties. (8) Among other things, Carbonneau describes how parties may wish to customize the following: the governing law; (9) the venue for arbitration (and the corresponding rules of practice for foreign attorneys); (10) whether arbitration is to take place on an ad hoc basis or under the auspices of one of the institutional arbitration providers (including the ICC and the LCIA); (11) how the arbitrators are to be selected (and their compensation and qualifications); (12) what devices to employ to foster collegiality among multiple arbitrators and provide for their accountability; (13) who is to have ultimate authority in the arbitration proceedings (the arbitral institution, the arbitrators, or the parties to the dispute); (14) procedural aspects of the arbitration including discovery, evidentiary rules, and how witnesses are handled; (15) whether arbitrators must provide reasons with their decision; (16) and finally the standard of judicial review for arbitration awards. (17)

As a starting point to assess Carbonneau's article, assume that different contracting parties would in fact desire varying arbitration agreement terms. Some parties might prefer extensive discovery while others might not. Some contracting parties might wish to have attorneys present during an arbitration hearing while others might prefer to represent their own interests more informally. Given varying preferences for arbitration terms, what might cause parties not to negotiate for different arbitration agreement terms? One possibility is that parties may lack the sophistication to negotiate an agreement in their best individual interests. For consumers faced with mandatory arbitration, imposed through contracts of adhesion, such a concern is certainly material. However, Carbonneau has a different set of parties in mind: sophisticated business parties often contracting across international boundaries. (18) Carbonneau is correct in arguing that such parties--should they desire--possess the ability to negotiate for highly specific arbitration contracts.

A fundamental question therefore arises from Carbonneau's analysis: why do sophisticated parties not already internalize Carbonneau's advice and tailor arbitration agreements? Put another way, why do sophisticated parties (presumably represented by attorneys) need Carbonneau's article to remind them of the ways in which they can modify arbitration agreements? (19) Simply possessing the ability to negotiate complex arbitration agreements does not necessarily mean that parties will desire to do so. Even for sophisticated parties, contracting is an expensive proposition for several reasons. (20) First, parties must consider how others (including courts) may interpret the wording of newly negotiated contractual terms. Michael Klausner, among others, has written on the value of certainty in contract terms. (21) Parties attempting to craft variations to standardized terms run the risk of having a court (or other interpreting body) later misinterpret the terms. The more parties use a standardized term over time, the more certainty the term achieves--giving rise to a network externality effect. (22) Even substandard arbitration terms may then persist, to the extent parties fear the uncertainty from deviating from such terms.

Second, network externalities may also arise due to the interaction of a particular arbitration agreement term with the arbitration industry. (23) Many different parties play varying roles in an arbitration proceeding, including both the arbitrators themselves and attorneys. These different role players develop a set of skills specific to particular types of arbitration. Modifications to standardized arbitration agreements may generate higher costs for these related parties. Parties who employ a different set of discovery rules, for example, may force attorneys to expend time and effort in determining exactly how the new set of discovery rules are different from the standardized rules--leading to higher fees for the parties to the arbitration. (24)

Third, the very act of negotiating for a specific contract term may signal negative information to the other party. Where the norm is for most parties to accept the standardized arbitration terms, (25) other parties may view with suspicion a party who attempts to deviate from the standard. Because modification is expensive, parties who seek to negotiate such a change will do so only if the benefits exceed this expense, (26) such as in the case of parties who believe that arbitration is relatively more likely. Why bother negotiating over an arbitration term if a party believes that a dispute is unlikely? Other parties may therefore view an attempt to negotiate over an arbitration term as a signal that the party seeking to negotiate believes that arbitration is more likely (perhaps because the party attempting to negotiate is more contentious than the norm). Other parties may then decide not to contract at all with the negotiating party (or demand a price adjustment for the added risk involved).

Fourth, even sophisticated parties may suffer from a variety of behavioral biases which may hinder their ability to negotiate for contract terms. People, for example, suffer from the well-documented overoptimism bias. (27) Entering into a contract, sophisticated parties may over-optimistically believe in their ability to enter into value-increasing business deals and that conflict will not arise in the future. With such a belief, parties may decide that the benefit of specifically tailoring an arbitration clause is simply not worth the cost.

Despite the lack of incentive of individual sophisticated parties to negotiate over new contract terms, the market may devise ways to generate real choice in arbitration agreements. Institutions such as the ICC (28) and LCIA, (29) for example, already design their own standard set of arbitration terms for use within their own sponsored arbitrations. (30) To the extent variations exist between ICC and LCIA terms, parties have some choice in selecting between the two. (31) However, market forces work to keep arbitration choices limited. Institutions that take it upon themselves to set up arbitration proceedings and the underlying rules governing the arbitrations must convince parties that the organizations (and the arbitrators associated therewith) will not take advantage of the parties. Hew can one party, for example, be assured that an arbitral institution will not favor the other side unduly (perhaps in the selection of the specific arbitrators for a proceedings)? (32) Similarly, what assurances do parties have that an arbitral institution will faithfully adhere to its rules of arbitration in administering a proceeding or apply agreed upon substantive law? (33)

Reputation provides one answer. Arbitral institutions (and arbitrators) may develop a reputation for unbiased, competent, and efficient decision making over time. This reputation allows the arbitral institution to charge a higher fee for its services. The prospect of receiving a flow of elevated fees then works to deter the arbitral institution from sacrificing its reputation in any specific arbitration proceeding. (34) Reputation, however, is not cost free. Because reputational bonding relies on high future profits, arbitral institutions with a large market share (and correspondingly greater profits) will provide the greatest assurance for parties. (35) Greater market share and the need to sustain high profits to provide a reputational bond necessarily limit competition among arbitral institutions. (36) Upstart providers of arbitration services will have a difficult time convincing consumers of their expertise and credibility. Consumers of arbitration services therefore may end up with both high fees and fewer choices among arbitration bodies.

So how can we generate greater choice in arbitration proceedings while at the same time keeping fees down and preserving the credibility of arbitral institutions? One possibility is that lawmakers may seek to reduce the bonding costs for potential providers of arbitral services. Imposing liability on arbitrators and arbitral institutions--something typically not allowed (outside of narrow exceptions) under current law in the United States and other states (37)--provides one possible mechanism of ensuring arbitrator fidelity without resorting to high-cost reputational bonding. Indeed, for arbitral institutions and arbitrators otherwise unable to convince contracting parties of their credibility, government imposed liability may provide a low-cost alternative to reputational bonding. Significantly, increased liability does not have to be mandatory. One could imagine a system in which lawmakers allowed arbitral institutions and arbitrators to "self-tailor" the liability they would face if they took unfair advantage of one of the arbitrating parties, allowing institutions and arbitrators to increase liability above some minimum floor. (38) Alternatively, lawmakers may allow arbitral institutions and arbitrators to opt into a higher level of judicial review of arbitration decisions (for a fee to cover the cost of the increased judicial review) (39) or a particular level of enforcement of arbitration awards. Upstart arbitral institutions in particular may benefit the most from such a self-tailored liability regime. By opting for relatively high levels of liability to signal their credibility to contracting parties, upstarts may enhance their ability to enter the arbitration market and compete with more established arbitrators and institutions. (40)

Government assisted bonding for arbitrators extends to the international arena. No single government has the ability to monitor and enforce liability awards against arbitrators and institutions located in disparate locations worldwide. Nonetheless, to the extent at least some arbitrators and arbitral institutions value the low-cost bonding possible through government imposed liability, these private actors will voluntarily adjust their behavior to maximize the impact of such liability. Arbitrators and arbitral institutions that choose to face a higher degree of liability imposed by state X, for example, may shift greater amounts of assets into state X to increase the enforceability of a potential liability award (and thereby the credibility signal sent by placing themselves under state X's liability regime).

In sum, Carbonneau deserves praise for providing a roadmap for parties interested in modifying their arbitration agreements. Underlying Carbonneau's seemingly intuitive points about the contractual nature of arbitration, however, is the puzzle of why sophisticated parties typically choose not to tailor their agreements, instead opting into standardized arbitration procedures. Several obstacles stand in the way of even sophisticated parties seeking to tailor their arbitration agreements. Arbitral institutions provide one solution to the high cost of contracting through the provision of off-the-rack standardized arbitration proceedings. However, the need to maintain a high reputation necessarily results in reduced competition among arbitral institutions to provide new contract choices for parties. To generate greater choice, governments may therefore have a role in assisting competition in the market for arbitration services, through, for example, a regime of self-tailored liability for arbitrators and arbitral institutions.

(1.) See http://www.iccwbo.org/index_court.asp (last visited July 20, 2003) [hereinafter ICC].

(2.) See http://www.lcia-arbitration.com (last visited July 20, 2003) [hereinafter LCIA].

(3.) See http://www.adr.org (last visited July 20, 2003).

(4.) For a description of the various international arbitral institutions, see Thomas E. Carbonneau, The Ballad of Transborder Arbitration, 56 U. MIAMI L. REV. 773, 796-801 (2002).

(5.) See generally ICC Rules of Arbitration, available at http://www.iccwbo.org/english/arbitration/pdf_documents/rules.

(6.) Thomas E. Carbonneau, The Exercise of Contract Freedom in the Making of Arbitration Agreements, 36 VAND. J. TRANSNAT'L L. 1189.

(7.) Carbonneau writes: "Custom-fitting arbitration to the parties and the transaction, while maintaining the functionality of the process, is the cardinal objective." Id. at 1203.

(8.) Id. at 1204.

(9.) Id. at 1218-20.

(10.) Id. at 1225-30.

(11.) Id. at 1206-09.

(12.) Carbonneau provides an example of a possible qualification stipulation: "A Muslim party, for example, might insist that all arbitrators be male because an arbitral award rendered by a tribunal which includes a female arbitrator is generally not enforceable in a Muslim country." Id. at 1210.

(13.) Carbonneau writes: "Whatever the parties' other aims, appointing a group of decision-makers who can work well together may be the most critical objective of all." Id. at 1215.

(14.) Id. at 1217-19.

(15.) Id. at 1221.

(16.) Id. at 1221-22.

(17.) Id. at 1222-24.

(18.) Id. at 1191.

(19.) As Carbonneau puts it: "Because the standard clause is simple, straightforward, and well-known, there is an understandable reluctance among legal practitioners and business parties to deviate from its established pattern." See id. at 1202-03.

(20.) Carbonneau himself recognizes this expense writing: "Customization requires negotiations and greater cost at the outset of the transaction. The additional expense and frustration could undermine the deal or compromise its implementation." Id. at 1231.

(21.) See Michael Klausner, Corporations, Corporate Law, and Network of Contracts, 81 VA. L. REV. 757, 774-79 (1995).

(22.) Network externalities in contracts involve spillover effects from the use of a particular contract term by other contracting parties. Network externalities in contracts may cause parties to choose even inefficient terms (due to their certainty value from widespread use). See id. at 780-82.

(23.) Klausner terms this a "common practice" network externality:

   The prospect of common practices that develop in the future to
   implement a contract term is thus a network benefit. Such a benefit
   can be expected to arise only if a term is commonly used. The more
   commonly firms use a particular term, the more reliable, and hence
   more valuable, common practices implementing the term can be
   expected to be.

Id. at 781.

(24.) See id. at 782.

(25.) A norm of simply accepting standardized terms may arise, for example, due to the generally high cost of negotiating for new terms (and the uncertainty of such terms) as discussed earlier in this comment. See id. at 785.

(26.) Id.

(27.) Over-optimism may affect experts (or those who perceive themselves as experts) in particular. See generally John Jacob et al., Expertise in Forecasting Performance of Security Analysts, 28 J. ACCT. & ECON. 51 (1999); Michael Mikhail et al., The Development of Expertise: Do Security Analysts Improve their Performance with Experience?, 35 J. ACCT. RES. 131 (Supp. 1997); Dale Griffin & Amos Tversky, The Weighing of Evidence and the Determinants of Confidence, 24 COGNITIVE PSYCHOL. 411 (1992).

(28.) ICC, supra note 1.

(29.) LCIA, supra note 2.

(30.) ICC Rules of Arbitration, supra note 5.

(31.) See, e.g., Carbonneau, supra note 4, at 797:

   The LCIA Rules of Arbitration are more accessible than their ICC
   counterpart, which seems to be mired in the franglais foibles of
   French-to-English translation and composition. Historically, among
   international lawyers, LCIA arbitration has been perceived as
   English arbitration and linked to English court supervision of the
   merits of arbitral awards.

(32.) See William M. Landes & Richard A. Posner, Adjudication as a Private Good, 8 J. LEGAL STUD. 235, 248 (1979). William Landes and Richard Posner note that parties entering into an arbitration face a dilemma in the selection of arbitrators. Id. A party who knows she is in the wrong will delay as much as possible in selecting an arbitrator in order to avoid the proceedings, undermining the value of arbitration in the first place. Id. Arbitral institutions such as AAA avoid such a problem by making the final arbitrator selection itself (after receiving some input from each party). Id.

(33.) See Stephen J. Ware, Default Rules From Mandatory Rules: Privatizing Law Through Arbitration, 83 MINN. L. REV. 703, 721-27 (1999) (discussing the privatization of substantive law through arbitration and noting that "[i]n most cases in which an arbitrator does not apply the law, it will be virtually impossible for a court to discover that the arbitrator did not apply the law"); see also Cameron L. Sabine, Adjudicatory Boat Without a Keel: Private Arbitration and the Need for Public Oversight of Arbitrators, 87 IOWA L. REV. 1337, 1343-44 (2002):

   [A]rbitrators often disregard arbitration agreements or contract
   terms. Most problematic, however, is the charge that arbitrators
   are biased, partial, or influenced by conflicts of interest, or that
   arbitration awards are procured through fraud or corruption.
   Connected to these allegations are claims that arbitrators pander to
   repeat players, an allegation supported by recent studies.

(34.) But see Sabine, supra note 33, at 1368 (arguing that status quo exemplifies that "regulation by the market does not provide an adequate check on arbitrator behavior. Nor can a solution depend on private arbitration associations policing arbitrators because enforcement is discretionary, inconsistent, and arguably fictional").

(35.) Id. at 1368-69.

(36.) See Carbonneau, supra note 4, at 796 (noting that "[a] number of traditional service-providers have a virtual lock upon the transborder arbitration service industry ... [making it] difficult for newcomers to establish a foothold in the area").

(37.) See Sabine, supra note 33, at 1350 (discussing how courts in the U.S. have given arbitrators the equivalent of common law judicial immunity for judges); David I. Bristow, Q.C. & Jesmond Parke, The Gathering Storm of Mediator & Arbitrator Liability, 55 DISP. RESOL. J. 14, 16 (2000) ("In the United States, the view of arbitrators as possessing a level of arbitral immunity is so well established in tort law that it is rarely challenged"). But see Susan D. Franck, The Liability of International Arbitrators: A Comparative Analysis and Proposal for Qualified Immunity, 20 N.Y.L. SCH. J. INT'L & COMP. L. 1, 4 (2000) (noting that "[t]raditionally, civil-law and several Arab countries emphasize the contractual nature of the arbitrator's receptum arbitri and use this as a baseline for establishing potential liability").

(38.) Policy reasons nonetheless exist for maintaining the present level of arbitrator immunity. See, e.g., Bristow & Parke, supra note 37, at 16 (noting two arguments in support of arbitrator immunity: "1. adjudicators perform an important societal function which deserves all possible encouragement and support; and 2. adjudicators need immunity to protect them from reprisals that could otherwise undermine the integrity of the decision making process"). Making increases in liability voluntary on the part of arbitrators (and arbitral institutions) allows the arbitrators to choose the level of liability that best balances the need to bond their credibility while also protecting the independence of their decision making.

(39.) More judicial review, on the other hand, may undermine the finality of arbitration awards and raise the expected costs to parties faced with arbitration. See, e.g., Ware, supra note 33, at 722-23. If given a choice, nonetheless, institutions will weigh the downside of greater judicial review against the upside of the increased credibility lent by it. Where institutions take into account their own direct interests as well as the interests of disputing parties (to the extent the fees the parties are willing to pay turn on the arbitral institution's decision with respect to judicial review), such a choice maximizes the joint welfare of arbitral institutions and the disputing parties.

(40.) Other problems, of course, exist with arbitration contracts apart from providing private parties greater choice (and lower costs). Parties choosing to arbitrate ignore the loss to the public from not having a case litigated in court where it may generate precedential value for other cases. See Landes & Posner, supra note 32, at 238-40 (noting the advantage of state provided adjudication in avoiding the underproduction of precedents in private adjudication). On the other hand, parties who settle also impose the same loss of precedence on other parties. As well, parties who choose to continue through the courts (at least in the United States) are subsidized to the extent they do not bear the full cost of operating the judicial system (including the salary of judges, among others).

Stephen J. Choi, Professor of Law, University of California, Berkeley (Boalt Hall). Special thanks to Un Kyung Park.