Professor Peter Rutledge

December 12, 2007

Testimony of Professor Peter B. Rutledge
Associate Professor of Law
Columbus School of Law, Catholic University of America

December 12, 2007

Hearing: "S. 1782: the Arbitration Fairness Act of 2007"

Senate Judiciary Committee
Subcommittee on the Constitution

Chairman Feingold, Ranking Member Brownback and Members of the Subcommittee. Thank you for the invitation to testify today. My name is Peter B.Rutledge, and I am an Associate Professor of Law at the Columbus School of Law at the Catholic University of America here in Washington, D.C. I am co-author of the book International Civil Litigation in the United States and author of several articles in the field of arbitration. I am pleased to offer my thoughts on S. 1782, the Arbitration Fairness Act.

In October of this year, I had the privilege of testifying before a subcommittee of the House Judiciary Committee on H.R. 3010, the House companion bill to S. 1782. I trust that your staffs have reviewed that testimony, so I do not intend simply to rehash its contents here. Instead, I hope to highlight some of the most critical points that I sought to make in the House, to address some of the issues that arose during that hearing, and to discuss some intervening developments since that hearing.

SUMMARY

At bottom, I wish to convey four critical points to the subcommittee today:

? First, a thorough understanding of the available data and gaps in the data should drive the policy debate over the future of arbitration. Otherwise, there is a risk that the policy debate will be driven by a mixture of unrepresentative cases and untested hypotheses. The available empirical data on arbitration is growing. In important respects, that data demonstrate that individuals, in the aggregate, often are better off in a
world with enforceable predispute arbitration agreements than in a world without one. In this regard, the data either are inconsistent with or flatly contradict some of the premises that appear to underpin efforts to abolish predispute arbitration.

? Second, to the extent there are particular instances of problematic arbitrations, the solution should not be to jettison the system altogether. In several respects, existing mechanisms serve to filter out truly unfair arbitrations. To the extent those mechanisms do not suffice, Congress would be better off analyzing structural deficiencies in the civil justice system than in abolishing predispute arbitration agreements for vast segments of the population.

? Third, if Congress were to eliminate predispute arbitration agreements, it ironically may make worse off the very parties whom it would be trying to protect. Individuals would find it more difficult to obtain a lawyer, would realize worse outcomes, and would receive justice at a far slower rate. For society as a whole, the costs of resolving these disputes without arbitration likely would rise, and individuals ultimately would bear those higher costs - whether in the form of higher prices, lower wages or lower share prices. The only people who, with certainty, benefit from this bill are the lawyers.

? Fourth, postdispute arbitration does not present a viable alternative to a system of enforceable predispute agreements.

With that summary, I will now elaborate on each of these points.

I. The State of the Empirical Research

As I explained to the House back in October, empirical research on arbitration has advanced greatly in the last fifteen years.1 At the most general level, we have solid studies on questions such as whether arbitration benefits the repeat players2 or whether arbitration leaves individuals better off or worse off than litigation.3 More recently, Chris Drahozal at the University of Kansas recently edited an excellent volume synthesizing the available empirical research in the field of international arbitration.4 Researchers at Cornell University and New York University, among others, are undertaking pathbreaking research in the area of employment arbitration.5 Next month, participants at the annual meeting of the Association of American Law Schools will convene a panel addressing the state of the empirical literature on arbitration.

At the same time, important gaps remain in the empirical record. At the industry level, we probably have the best data about employment arbitration; the empirical record on consumer arbitration and, especially, franchise arbitration is far less developed.6 Moreover, the amount of available data varies with the arbitration provider. Based on my review of the literature and my own data-gathering efforts, the American Arbitration Association and the organization JAMS have been quite willing to provide access to data about their caseloads; the amount of publicly available data for other organizations is more limited. Lastly, and perhaps most importantly, we are only beginning to get a good handle on the economics of arbitration and, particularly, the economic impact of a prohibition against predispute arbitration.7 (I return to this last point later in my
testimony.)

With the empirical record in this state, I would urge Congress to proceed cautiously. The risk is that the untested hypotheses or unrepresentative horror stories about arbitration will drive the discussion at the expense of the empirical reality. In my view, some of the "Findings" contained in Section 2 of S. 1782 reflect some of these biases about arbitration. Permit me to address several of them briefly:

Finding: "A large and rapidly growing number of corporations are requiring millions of consumers and employees to give up their right to have their dispute resolved by a judge or jury, and instead submit their claims to binding arbitration."8

The Empirical Record:

As to the frequency of arbitration, the record is far
more mixed than this finding suggests. One recent study of consumer arbitration in over thirty industries found that the frequency of clauses varied greatly across industries.
Approximately 33% of surveyed companies employed arbitration clauses; nowhere were they universally used; industries such as the financial services sector used them 69.2% of the time while other industries such as food and entertainment never used them.9 Another very recent study of corporate 8-K filings found that companies used arbitration clauses only about 11% of the time; the precise data again varied with the type of contract
(with clauses more frequent in licensing and employment contracts), but with respect to no category of contracts did companies use arbitration clause a majority of the time.10

As to the waiver of the right to a jury trial, it is certainly true that arbitration does not involve a jury. But eliminating arbitration would not suddenly cause all of those disputes to be decided by a jury. Numerous studies have documented how most civil litigation is resolved far before a case ever reaches a jury - whether through voluntary dismissal, settlement or dispositive rulings by the judge.11

Finding: "Most consumers and employees have little or no meaningful option whether to submit their claims to arbitration."12

The Empirical Record: Again, as I've noted above, the record on the frequency of arbitration clauses is more mixed than this finding suggests. Even where arbitration clauses are used, it is important to emphasize that only a fraction of the disputes subject to arbitration clauses actually reach arbitration. At least in the employment context, we know that the arbitration clause is part of a broader "multi-stage" dispute resolution system whereunder many of the disputes resolve at an earlier stage.13 This does not mean, of course, that arbitration is irrelevant. Rather, it serves as an essential piece of a broader tapestry of alternative dispute resolution processes.

Finally, as to the notion that individuals have "no meaningful option," it is important to place that comment into context. Individuals are presented with a variety of terms on a take it or leave it basis.14 For example, my employer presents me with only a single health insurer and a single 401(k) plan.

Similarly, as a consumer, I may be presented with a variety of "take it or leave it" terms ranging from the interest rate at my bank to the price of the car that I rented last month. Yet no one would deny there are valid economic reasons, some of which directly benefit me as an employee or a consumer, why my counterparty does not dicker over those terms. The same economic rationale that justifies these sorts of "take it or leave it" policies applies to arbitration.

Finding: "Private arbitration companies are sometimes under great pressure to devise systems that favor the corporate repeat players who decide whether these companies will receive their lucrative businesses."15

The Empirical Record: Some studies have found evidence of a repeat player phenomenon while others have found no demonstrable effect.16 Furthermore, even where the repeat player effect exists, the cause is not clear. Most research suggests that the repeat player effect - if it exists - is not due to the arbitrator's financial incentives but, instead, to the "learning effects" from the repeat player's experiences.17 That is, the
repeat player learns what sorts of cases can be won and, therefore, is more likely to settle those, leaving for arbitration those where the repeat player is relatively confident it can win outright (or at least where the costs of taking the case through arbitration are lower than the minimum amount that the claimant is prepared to accept in settlement).

Finding: "Mandatory arbitration undermines the development of public law for civil rights and consumer rights, because there is no meaningful judicial review of arbitrators' decisions."18

The Empirical Record: Public law can still develop in arbitration whether through publication of the awards or judicial decisions in actions to confirm the awards.
Moreover, a variety of other mechanisms have a far greater impact on the development of public law. The most obvious one is settlement, which I would safely suspect occurs far more frequently than arbitration. Settled cases generally do not result in the creation of binding precedent.

Finding: "Mandatory arbitration is a poor system for protecting civil rights and consumer rights because it is not transparent."19

The Empirical Record: This criticism is mistaken for three reasons. First, it misapprehends arbitration: there are at least two junctures where the merits of arbitration can be publicly aired: the enforcement of the agreement and the enforcement of the award. Second, like several of the other criticisms noted here, it unfairly singles out arbitration: a variety of other mechanisms, judicial or otherwise, are not transparent. Settlement again is the most obvious - a claim of threatened litigation may settle with even less public disclosure than arbitration. Even when claims are litigated, the opportunities for transparency are limited. The judge may enter an order on the record without elaboration, or an appellate court may summarily affirm a lower court judgment on some issue without elaborating on its reasoning. Third and finally, the criticism over transparency has a flipside - namely confidentiality. Parties may well prefer arbitration precisely because, relative to civil litigation, the proceedings take place in a less public
setting and, thereby, avoid the more open hostility that can be engendered when the parties stake out their position in public. Indeed, one of the great benefits of arbitration is a psychological one - it enables parties to sort out their differences before their dispute spills out into the court of public opinion and causes parties to dig into their positions.

Finding: "Many corporations add to their arbitration clauses unfair provisions that deliberately tilt the systems against individuals, including provisions that strip individuals of substantive statutory rights, ban class actions, and force people to arbitrate their claims hundreds of miles from their homes."20

The Empirical Record: Here, it is important to unpack several propositions. As to the claim that arbitration clauses "deliberately tilt the systems against individuals," nearly all of the available academic studies, most of which concern employment arbitration, demonstrate precisely the opposite outcome.21 That is, by most measures, the party with the inferior bargaining position achieves better, or at least comparable, outcomes in arbitration compared to litigation.22 Eliminating predispute arbitration might well make these individuals worse off - I return to this topic later in my testimony.

As to the claim that arbitration bans class actions and forces people to arbitrate far from their homes, the empirical evidence again suggests that these practices are not as widespread as the finding suggests. For example, a 2004 study of consumer arbitration clauses by Demaine and Hensler found that only 30.8% prohibited class actions.23 As to the situs, 50% of the clauses they surveyed specified the situs of the arbitration, and in all
but three cases was the specified situs near the individual's residence or place of service.24 This led the authors to conclude that "few of the 52 clauses reflect the type of egregious self-dealing that has been identified in publicized cases. Most of the clauses appear in many respects to put consumers on equal terms with the businesses that drafted
them ... ."25

The September 2007 Public Citizen Report:

The House hearing in October highlighted a recent report by Public Citizen on arbitrations in the credit card industry.26 I commend Public Citizen for its attempt to contribute to a more systematic, rather than anecdotal, assessment of arbitration. But even here I would urge Congress to evaluate the findings with care.

The report contains sensational allegations about how a small cadre of arbitrators overwhelmingly favor the credit card industry over their customers and churn out a high
number of awards, seemingly with little consideration or oversight. When I initially skimmed the report, I confess that some of these allegations gave me pause. But when I
dug into the guts of the report, two things struck me.

First, while the allegations at times are quite general, the underlying data focus largely on the work of a single arbitration association (the National Arbitration Forum) for a single company (MBNA) in a single industry (credit card collection actions). As you know, arbitration involves a far greater number of associations, companies and industries than the ones addressed in the Public Citizen report. Thus, the report does not provide a particularly helpful set of data upon which to base a decision about the future of arbitration.

The second striking feature of the Public Citizen report is that most of the arbitrations addressed in it appear to be default collection actions - that is, relatively straightforward arbitrations commenced by a bank when someone does not pay their credit card bill. Seen in this light, it is perhaps unsurprising that the company's win rates are so high (apart from cases of outright fraud or identity theft, the matters presumably are straightforward). Thus, the lopsided win-rates described in the Public Citizen report also are a poor metric upon which to make any decision about whether arbitration generally favors corporate interests.27 Moreover, due to this focus on default collection actions, many of the more generalized complaints about arbitration are misplaced. For example, the report complains about how arbitration may deprive consumers of the opportunity to maintain a class action or to seek punitive damages. Yet procedural rights and remedies of this sort would be unavailable to an individual debtor, regardless of whether the case is heard in arbitration or in a court. And to the extent the judicial proceeding may be slower, more cumbersome and more expensive, it may well leave the individual debtor worse off.

II. Existing Mechanisms To Address Problems

My testimony should not be understood as an uncritical acceptance of the status quo. Surely there are instances of indefensible arbitration agreements.28 But the question is not whether arbitration is perfect; like any system, it is not. Rather, the question is whether Congress should jettison the entire enterprise of predispute arbitration agreements in order to combat these difficulties. I would submit that it should not do so, and part of the reason is my trust in the existing mechanisms that have evolved to address this problem.

First, there has been a good deal of self-regulation in this area. In the securities industry, for example, the major arbitration services promulgate and revise their rules under the auspices of the Securities and Exchange Commission. On the commercial side, several of the major arbitration organizations have signed on to "Due Process Protocols".

For example, the employment protocol sets forth a variety of rights including:

- the employee's right to be represented by a person of her own choosing;

- the employer is encouraged to pay at least a share of the employee's fees;

- employees should have access to all information reasonably relevant to their claims;

- before selecting an arbitrator, parties should have sufficient information to contact
parties who previously have appeared before her;

- arbitrators should have sufficient skill and knowledge;

- arbitrators should be drawn form a diverse background;

- arbitrators should be free of any relationships that would create an actual or apparent conflict of interest;

- the employee's entitlement to the same array of remedies in arbitration as she would be entitled to in a judicial proceeding.

Subsequent protocols governing consumer disputes and health care disputes differ in some of the specifics but contain the same basic protections. Many of the major arbitration associations have committed to administering arbitrations in the consumer and employment areas only if the parties agreed to be bound by the protocols.29

To be clear, not all arbitral institutions have signed onto the protocols. But even where they do not bind the organizations, that does not mean they are wholly irrelevant.

As I have explained elsewhere, some courts, including several justices on the Supreme Court, have looked to the protocols as a benchmark by which to assess the procedural fairness of a particular arbitral scheme.30 In other words, while the protocols technically do not have the binding force of a legal rule, they nonetheless have exerted a persuasive influence on how some courts have interpreted existing doctrine governing the enforceability of arbitral agreements and awards.

Even where the protocols or the judicial reliance on them is inadequate, the FAA provides several mechanisms for regulating arbitration. Section 2 of the FAA, as interpreted by the Supreme Court, authorizes courts to deny enforcement of arbitration agreements when, for example, the agreement is deemed to be substantively or procedurally unconscionable. Several courts have relied on these doctrines to invalidate agreements that, for example, cede too many of the claimant's procedural rights or impose too heavy a financial burden on arbitration.31

Additionally, Section 10 of the FAA sets forth several grounds upon which courts can vacate awards, and the federal courts have articulated several other grounds, such as the manifest disregard of the law doctrine.

Finally, in certain contexts, administrative agencies perform an important role to check imperfections in the system. Agencies such as the Equal Employment Opportunity Commission have responsibility for the enforcement of federal laws such as the employment laws. Only recently, the Supreme Court made clear that these agencies retain the right to commence litigation against an alleged violator even where the claim is on behalf of individual or a group who, due to an arbitration clause, may be unable to pursue litigation themselves.32

III. The Effect of Eliminating Pre-Dispute Arbitration

What would happen if Congress prohibited predispute arbitration agreements? In my view, several things would occur:

? Many individuals will find it harder to obtain a lawyer willing to take their case;

? For those who actually find a lawyer willing to take their case, justice will not come quickly;

? When it does come, the outcome of litigation may be inferior to that in arbitration;

? The net social costs of resolving these disputes will rise, and these costs would be passed onto employees (in the form of lower wages),consumers (in the form of higher prices) and investors (in the form of lower share prices).

Ironically, then, banning arbitration agreements may end up hurting some of the very groups that Congress is trying to protect. The only individuals who benefit from the ban
are the lawyers, who reap higher fees engaging in more expense, more protracted litigation. Permit me to elaborate on each of these points.

A. Difficulty In Obtaining a Lawyer

Several scholars have documented how difficult it is for a plaintiff such as an employee to find an attorney willing to take her case in the civil justice system unless the
amount in controversy is sufficiently high and the merits sufficiently strong.33 According to one study, an employee needed to have a meritorious claim of at least $60,000 in order for an employment lawyer to be willing to litigate her case,34 and a founder of the National Employment Lawyers' Association testified a few years ago that employment attorneys turned away at least 95% of employees who sought representation.35 The upshot is that by eliminating predispute arbitration, Congress may well worsen an access to-justice problem for the average claimant.

These ideas arose in the House hearing back in October, and Chairwoman Sanchez picked up on their centrality. She made the appropriate point that part of the debate here stems from the difficulties that individuals have had obtaining access to legal services. I responded that I could not agree more - it is something that we regularly consider in legal education in the context of clinical and pro bono work. But the solution, I would submit, is to improve access to legal services, not to jettison a system of dispute resolution that, based on the available empirical evidence, seems to yield a net benefit to the individuals whom Congress is trying to protect.

B. The Speed of the Civil Justice System

For those who actually find a lawyer willing to take their case, justice will come far more slowly in a world without arbitration. The comparative speed of recovery with respect arbitration and litigation is one area where we have especially good data and where the import of the data is clear. Virtually every study considering the issue has concluded that results in arbitration are far swifter than those in litigation. In fact, of all the empirical premises about arbitration that has been subject to empirical study, this is the one where the data results are most consistent.36 For example, Delikat and Kleiner concluded in their study of securities arbitrations that mean and median times between filing and judgment were approximately 50% longer in litigation compared to arbitration.37 Thus, for those claimants who desire speedy resolution of their claims(whether for financial reasons, psychological ones or others), arbitration is far superior.

C. Comparative Outcomes

As I've briefly alluded to above, most of the available empirical evidence suggests that arbitration actually leaves individuals better off than in litigation. If that is true, then conversely eliminating predispute arbitration may well leave those individuals worse off in terms of their recoveries. Allow me to elaborate on that claim here.

Scholars studying this issue ("Are individuals better off in arbitration?") employ a variety of methodologies for evaluating that question. Some look at raw win rates - that is, how often does the individual prevail in arbitration compared to the company? Others look at comparative win rates - how often does the individual recover in arbitration compared to litigation? Yet others look at comparative recovery rates - asking how much does the individual recover in arbitration compared to litigated?

While none of these methodologies is flawless, each undercuts the idea that arbitration is somehow stacked against the individual. For example, employing a raw win-rate methodology, a review of consumer arbitrations in California by the California
Dispute Resolution Institute found that the consumer prevailed 71.2% of the time.38

Under the comparative win-rate methodology, most studies find no significant difference between arbitration and litigation in terms of the frequency with which the individual prevails.39 As to the comparative recovery methodology, the available studies reach conflicting conclusions,40 but the best research appears to conclude that higher-compensated employees recover at least as much in arbitration as litigation whereas lower-compensated employees (those with a gross income of less than $60,000) might not.41

This last bit of data is an important. It might well suggest that arbitration leaves employees in a certain income category worse off. But the consensus on this data suggests a much more complex picture. Both the authors themselves and subsequent
commentators noted that the sample size was small and the standard deviations significant. Moreover, the authors noted that the recovery differential might not be due to some flaw in the system of arbitration. Rather, the more likely cause might be the difficulty that lower-compensated employees encounter in obtaining legal counsel to litigate their case. Because their claims are likely to be lower on average, only employees with high-value, high-merit claims are going to be able to find attorneys willing to try their case, thus skewing the average recoveries from litigation. By contrast, the lower cost of arbitration may make it easier for them to resolve their claims without an attorney or to afford an attorney, thus lowering the average recoveries from arbitration in part due to the fact that more employees in this income echelon are able to have their
case heard.

The bottom line is that eliminating arbitration will not meaningfully enhance the outcomes for individuals and may well produce inferior ones. To the extent there is a problem, the data suggest that the problem is not a systemic one with arbitration but rather stems from difficulties that an individual encounters obtaining access to counsel in
our civil justice system, as suggested during the dialogue with Chairwoman Sanchez back in October.

D. Increase Costs of Resolving Disputes

As I noted earlier, I believe that eliminating predispute arbitration agreements would increase the costs of resolving disputes. Here it is important to recall precisely why arbitration grew in popularity. A 1997 report by GAO evaluated the experiences of five employers with their alternative dispute resolution programs and helped to document how their interest with arbitration came about.42 All five companies had adopted ADR programs, which included an arbitration component, after spending exorbitant fees defending against employment lawsuits. Brown & Root adopted its ADR program after spending $400,000 in legal fees to defend against an employment discrimination suit
which it won.43 Similarly, Rockwell adopted an ADR program after it spent over $1 million in legal fees defending against a wrongful discharge/disability discrimination suit which it too won.44 The GAO went on to report that, after implementing the ADR programs, the companies' legal costs dropped sharply. Brown & Root, for example, reported a 90% reduction in its legal fees during the first three years of its ADR program.45 Even factoring in the additional costs of the ADR system, Brown & Root's overall costs of dealing with employment conflicts, including ADR costs, were now less than half of what the company used to spend on legal fees for employment-related lawsuits.46

If arbitration, as part of a larger fabric of ADR programs, can reduce corporate legal costs, do those savings actually benefit the individual? By the early 1990's, some government studies suggested that the answer was "yes." One early indication of the relationship between dispute resolution and individual wealth came in report of the Dunlop Commission, created by President Clinton.47 As part of its work, the Commission considered the impact of employment litigation and dispute resolution.

It concluded:

For every dollar paid to employees through litigation, at least another dollar is paid to attorneys involved in handling both meritorious and non-meritorious claims. Moreover, aside from the direct costs of litigation, employers often dedicate significant sums to designing defensive personnel practices (with the help of lawyers) to minimize their litigation exposure. These costs tend to affect compensation. As the firm's employment law expenses grow, less resources are available to provide wage [sic] and benefits to workers.48

This "dollar for dollar" statistic derives from a report of factual findings issued by the Secretaries of Labor and Commerce.49 Those findings in turn trace to a 1988 study of
wrongful termination litigation in California conducted by the Rand Corporation's Institute for Civil Justice.50 In that study, researchers reviewed a sample of jury trials over an eight-year period in California. The authors surveyed counsel in each case to gather information about litigation costs. Based on their analysis of counsel's answers and the final recovery by prevailing claimants, they determined that a claimant's legal fees were more than one-third of her final payment and that the sum of the claimants' legal fees and the defendant's legal fees represented over 75% of the final payment received by the claimant.51

More recent research confirms that the cost savings generated through arbitration result in benefits passed on to employees. One survey of thirty-six employers who had alternative dispute resolution programs found that several employers provided certain benefits such as the right to participate in a corporate profit sharing plan in return for the employees' willingness to participate in an ADR program that included arbitration.52

Finally, one case suggests that the distributive benefits of cost savings might extend to the credit industry as well.53 In one case, a finance company varied the interest rate on its credit facility with a consumer's willingness to agree to arbitration.54 If the borrower did not agree to arbitration, the APR was 18.96%; if the borrower agreed to arbitration, the interest rate dropped to 16.96%. In other words, arbitration generated some unspecified quantity of cost savings for the lender, a portion of which was passed on to the customer in the form of a 2-point drop in the interest rate.

All of these anecdotes provided some indications that litigation was not only expensive for American companies but had identifiable negative wealth effects for their employees and price effects for their consumers. Recognizing that such anecdotes only have so much explanatory value, I have endeavored to take the empirical record one step further. Employing a comparative cost recovery framework, I analyzed the data on arbitration caseloads, the cost of resolving those disputes in arbitration, the costs of various forms of dispute resolution outside arbitration and the frequency with which alternatives to arbitration are used. Here, I wish to be very cautious because the data sets are incomplete, the analysis rests on several assumptions and the figures require further testing. But based on the data that I have been able to generate, it is my present belief that eliminating the employment arbitration docket of just one of the nation's leading arbitration associations

- the American Arbitration Association

- would increase aggregate dispute resolution costs approximately fourfold or approximately $88 million.

Let me be clear, this figure does not reflect any changes in the amount actually recovered by the claimant. Rather, it reflects simply an estimate of how much more it will cost society to resolve these same disputes that, under current law, are arbitrable.

This is why I say that the only people who come out ahead from the abolition of arbitration are the lawyers. Companies will have higher litigation costs, which they must pass on to individuals in the form of lower wages, higher prices or reduced share value. If $88 million is the net increased cost from eliminating the employment docket of a single arbitration institution, then imagine the cost of eliminating predispute arbitration in all consumer, employment and franchise contracts altogether.

I said at the October House hearing, and I reiterate here: this is a tentative conclusion. The important topic that it addresses - the economic impact of arbitration - is largely unexplored terrain, and one that would benefit from serious, open debate among academics, policymakers and interested parties. That debate is now beginning to occur. I would urge Congress to let that debate run its course so that it has a more
complete and accurate picture of the economic impact of this proposal.

IV. Postdispute Arbitration Is Not a Viable Alternative.

Let me close by addressing the idea that postdispute arbitration can simply reap all the benefits of predispute arbitration while preserving some greater modicum of "choice" for the individual. Opponents of predispute arbitration often argue that they do not reject arbitration, only agreements that bind a party to arbitration before a dispute has arisen; parties remain free to agree voluntarily to arbitrate after the dispute has arisen. The explanation for this proposal is deceptively simple: if defenders of arbitration are correct that arbitration offers so many advantages, then those advantages must also apply after a dispute has arisen; consequently, eliminating predispute arbitration agreements should not have much impact.

Postdispute arbitration has several problems, but let me focus on the central one: the parties' incentives in the postdispute context fundamentally differ from their incentives in the predispute context.55 Specifically, parties have more information in the postdispute context about the likely contours of the dispute. This superior information enables them to make more strategic calculations about which form of dispute resolution better advances their interests (or more effectively hinders the individual's interests). If a company knows that an individual's claim is below a certain amount, it may calculate that the individual could have difficulty obtaining a counsel willing to represent her. In those cases, a company may be less likely to agree to arbitration precisely because it knows that, effectively, its holdout will prevent the individual from pursuing her claim.56

Now contrast this state of affairs with those in the predispute context. In this setting, neither the company nor the individual knows in advance the terms or nature of a dispute.57 Yet each has an incentive to enter into arbitration - from the individual's perspective, arbitration provides an affordable forum with superior chances for obtaining a favorable result; from the company's perspective, arbitration can lower the company's litigation costs.

Experience under the recently enacted ban on predispute arbitration clauses in automobile dealer agreements lends some support to this hypothesis. As you know, in 2002, Congress amended the FAA and, for the first time since the FAA's enactment, explicitly banned predispute arbitration in a category of cases. The stated purpose of the law was to level the playing field between automobile manufacturers and their dealers(while leaving open the possibility of postdispute arbitration).58

Yet, in a recent case from the Seventh Circuit, an automobile dealer actually sought to compel arbitration, and the manufacturer successfully resisted it with respect to part of their dispute - effectively forcing the dealer to resolve the dispute in multiple forums.59 Had the parties been able simply to enter into a predispute arbitration agreement, such strategic behavior likely never would have arisen.

To be sure, parties to predispute arbitration agreements are engaging in some tradeoffs - the individual may be trading greater forum accessibility off against higher recoveries in litigation(assuming, of course, she can find a lawyer willing to take her case); the company is trading lower litigation costs off against a reduced likelihood of prevailing in the dispute. But that is the nature of any contractual bargain. The comparative advantage of arbitration is that it enables both parties to enter into an arrangement to manage some of the ex ante uncertainties about disputes before they arise, a possibility that is lost once the dispute arises and its terms are better known.60 Samuel Estreicher has used a very memorable metaphor to describe this essential bargain in predispute arbitration. According to Estreicher, "in a world without employment arbitration as an available option, we would essentially have a Cadillac system for the few and a rickshaw system for the many."61

Cadillacs represent the high-level recoveries for those few individuals with high-value, meritorious claims who find representation; the rickshaws represent the majority of individuals who struggle to find counsel willing to take their lower-stakes or more questionable claim. In a world with predispute arbitration, people substitute their Cadillacs and rickshaws for Saturns. In other words, individuals as a whole achieve the greater access to justice afforded by arbitration, even if a few individuals with high-stakes claims experience a marginal reduction in recoveries.62

CONCLUSION

In sum, Mr. Chairman, thank you for the opportunity to offer these views on S.1782. At bottom, it is my view that Congress should not prohibit predispute arbitration agreements in employment, consumer and franchise contracts. Rather, it should both encourage and await additional empirical research. That research may well show that minor additions to the existing regulatory repertoire are necessary. But eliminating predispute arbitration agreements would make worse off the very people whom Congress, through this legislation, is seeking to protect.

1 In addition, a number of governmental studies have looked at various aspects of arbitration. See GAO, Alternative Dispute Resolution: Employers' Experiences With ADR in the Workplace (1997); GAO, Employment Discrimination: Most Private Sector Employers Use ADR, 7 (1995); U.S. Commission on the Future of Worker-Management Relations, Final Report (1994); GAO, Securities Arbitration: How Investors Fare, 7-8 (May 1992). One government commissioned study did provide some valuable
empirical evidence in the field of securities arbitration. See Michael Perino, Report to the Securities and Exchange Commission Regarding Arbitrator Conflict Disclosure Requirements in NASD and NYSE Securities Arbitrations 32 (November 4, 2002);

2 Lisa Bingham, Employment Arbitration: The Repeat Player Effect, 1 EMP. RTS. & EMP. POL'Y J. 189, 208-09 (1997); Lisa Bingham & Shimon Sarraf, Employment Arbitration Before and After the Due Process Protocol for Mediation and Arbitration of Statutory Disputes Arising Out of Employment: Preliminary Evidence that Self-Regulation Makes a Difference in ALTERNATIVE DISPUTE RESOLUTION IN THE EMPLOYMENT ARENA: PROCEEDINGS OF THE NYU 53RD ANNUAL CONFERENCE ON LABOR 303 (Estreicher & Sherwyn eds. 2004); Lisa Bingham, On Repeat Players, Adhesive Contracts and the Use of Statistics in Judicial Review of Employment Arbitration Awards, 29 MCGEORGE L. REV. 223, 234 (1998).

3 Elizabeth Hill, Due Process At Low Cost: An Empirical Study of Employment Arbitration Under the Auspices of the American Arbitration Association, 18 OHIO STATE J. ON DISP. RES. 777, 814-16 (2003); Theodore Eisenberg & Elizabeth Hill, Arbitration and Litigation of Employment Claims: An Empirical Comparison, 58 DISP. RES. J. 44, 44 (Nov. 2003-Jan. 2004); Lewis Maltby, Employment Arbitration and Workplace Justice, 38 U.S.F. L. REV. 105, 107 (2003); Lewis Maltby, Private Justice: Employment
Arbitration and Civil Rights, 30 COLUM. HUM. RTS. L. REV. 29, 56-58 (1998); Lewis Maltby, Out of the Frying Pan, Into the Fire: The Feasibility of Post-Dispute Employment Arbitration Agreements, 30 William Mitchell L. Rev. 313 (2003).

4 Toward a Science of International Arbitration: Collected Empirical Research, Chris Drahozal, ed. (2005).

5 Sherwyn et al., Assessing the Case for Employment Arbitration: A New Path for Empirical Research, 57 STAN. L. REV. 1557 (2005); Samuel Estreicher, Saturns for Rickshaws: The Stakes in the
Debate Over Predispute Employment Arbitration Agreements, 16 OHIO ST. J. ON DISP. RES. 559, 567-68 (2001).

6 For some of the available research on franchise and consumer arbitration, see Keith Hylton & Chris Drahozal, The Economics of Litigation and Arbitration: An Application to Franchise Contracts, 32 J Legal Stud 549 (2003); Linda Demaine & Deborah Hensler, Volunteering to Arbitrate Through Predispute Arbitration Clauses: The Average Consumer's Experience, 67 LAW & CONTEMP. PROBS. 55 (Winter/Spring 2004).

7 Peter B. Rutledge, Whither Arbitration?, 6 Geo. J. Law & Pub. Pol'y __ (2008). In the interest of full disclosure, I should note that the Institute for Legal Reform provided funding for this study.

8 S. 1782 §2(2).

9 Demaine & Hensler, Volunteering to Arbitrate Through Predispute Arbitration Clauses: The Average Consumer's Experience, 67 L & Contemp Probs 55 (Winter/Spring 2004).

10 In a similar vein, a somewhat dated study of arbitration clauses in the securities industry found that clauses were used more frequently for higher-risk accounts. GAO: How Investors Fare at 28 (May 1992). Likewise, a survey of the telecommunications industry published in 2001 found that 16% of firms surveyed used external arbitration procedures in their employment disputes. See Colvin, Relationship Between Employment Arbitration and Workplace Dispute Resolution Procedures, 16 Ohio St J Disp Res 643 (2001).

11 Sherwyn et al., Assessing the Case for Employment Arbitration: A New Path for Empirical Research, 57 STAN. L. REV. 1557, 1566 (2005) (Vast majority of cases dismissed or resolved without court

12 S. 1782 §2(3). action undermines claim that arbitration will stagnate development of the law); Lewis Maltby, Private
Justice: Employment Arbitration and Civil Rights, 30 COLUM. HUM. RTS. L. REV. 29, 47(1998) (Cases never reach jury - of 3419 employment discrimination cases filed in 1994 that led to definitive judgment, 60% were disposed of by pretrial motion, with employers prevailing in 98% of those); Michael Delikat &
Morris Kleiner, Comparing Litigation And Arbitration Of Employment Disputes: Do Claimants Better Vindicate Their Rights In Litigation?, American Bar Association Litigation Section Conflict Management, Vol. 6, Issue 3, 8 (2003).

13 See Lewin, Dispute Resolution in NonUnion organizations, in Alternative Dispute Resolution in the Employment Arena (53rd Annual conference on Labor at NYU) at 379 (discussing case study at Northrop-Grumman); Lipsky, An Uncertrain Destination, in U.S. Corporations and ADR in Employment (Estericher & Sherwyn, eds.)(2004); GAO, Alternative Dispute Resolution:- Employers' Experiences (1997).

14 Sherwyn et al., Assessing the Case for Employment Arbitration: A New Path for Empirical Research, 57 STAN. L. REV. 1557, 1563-64 (2005).

15 S. 1782 §2(4).

16 Compare Lisa Bingham, Is there a Bias in Arbitration of Nonunion Employment Disputes? An Analysis of Actual Cases and Outcomes, 6 INT'L J. ON CONFLICT MGMT. 369, 380 (1995) and Lisa Bingham & Shimon Sarraf, Employment Arbitration Before and After the Due Process Protocol for Mediation and Arbitration of Statutory Disputes Arising Out of Employment: Preliminary Evidence that Self-Regulation Makes a Difference in ALTERNATIVE DISPUTE RESOLUTION IN THE EMPLOYMENT ARENA: PROCEEDINGS OF THE NYU 53RD ANNUAL CONFERENCE ON LABOR 303, 323 & Table 2(Estreicher & Sherwyn eds. 2004), with Elizabeth Hill, Due Process At Low Cost: An Empirical Study of Employment Arbitration Under the Auspices of the American Arbitration Association, 18 OHIO STATE J. ON DISP. RES. 777, 816 (2003).

17 Lisa Bingham & Shimon Sarraf, Employment Arbitration Before and After the Due Process Protocol for Mediation and Arbitration of Statutory Disputes Arising Out of Employment: Preliminary Evidence that Self-Regulation Makes a Difference in ALTERNATIVE DISPUTE RESOLUTION IN THE EMPLOYMENT ARENA: PROCEEDINGS OF THE NYU 53RD ANNUAL CONFERENCE ON LABOR 303, 323 & Table 2 (Estreicher & Sherwyn eds. 2004). See also Sherwyn et al., Assessing the Case for Employment Arbitration: A New Path for Empirical Research, 57 STAN. L. REV. 1557, 1570-71 (2005); Elizabeth Hill, Due Process At Low Cost: An Empirical Study of Employment Arbitration Under the Auspices of the American Arbitration Association, 18 OHIO STATE J. ON DISP. RES. 777, 816(2003).

18 S. 1782 §2(5).

19 S. 1782 §2(6).

20 S. 1782 §2(7).

21 Researchers use various methodologies to determine whether arbitration leaves the party with the inferior bargaining position "better off." For studies looking at raw win rates - that is, comparing how often each side wins, see Lewis Maltby, Employment Arbitration and Workplace Justice, 38 U.S.F. L. REV. 105, 108-11 (2003); Ernst & Young, Outcomes of Arbitration: An Empirical Study of Consumer Lending Cases 15 (2004), http://www.arbforum.com/rcontrol/documents/ResearchStudiesAndStatistics/2005ErnstAndYoung.pdf; California Dispute Resolution Institute, Consumer and Employment Arbitration in California: A Review of Website Data Posted Pursuant to Section 1281.96 of the Code of Civil Procedure, Figure 4 (2004); William Howard, Arbitrating Claims of Employment Discrimination, 50 Disp. Resol. J. 40, 44 (Oct.-Dec. 1995); William M. Howard, Mandatory Arbitration of Employment Discrimination Disputes: Can Justice be Served, 130-31 (May 1995) (unpublished Ph.D. dissertation, Arizona State University) (on file with author); GAO, Securities Arbitration: How Investors Fare, 7-8 (May 1992). For studies using a comparative win rate methodology (that is, comparing how often a party recovers in arbitration as opposed to litigation), see Samuel Estreicher, Saturns for Rickshaws: The Stakes in the Debate Over Predispute Employment Arbitration Agreements, 16 OHIO ST. J. DISP. RES. 559, 564-65 (2001); Lewis Maltby, Employment Arbitration and Workplace Justice, 38 U.S.F. L. REV. 105, 111-12 (2003); Sherwyn et al., Assessing the Case for Employment Arbitration: A New Path for Empirical Research, 57 STAN. L. REV. 1557, 1569 (2005); William M. Howard, Mandatory Arbitration of Employment Discrimination Disputes: Can Justice be Served, p. 130-31 (May 1995)(unpublished Ph.D. dissertation, Arizona State University) (on file with author). For studies using a comparative recovery methodology (that is, comparing the amount of recovery in arbitration as opposed to litigation), see Michael Delikat & Morris Kleiner, Comparing Litigation And Arbitration Of Employment Disputes: Do Claimants Better Vindicate Their Rights In Litigation?, American Bar Association Litigation Section Conflict Management, Vol. 6, Issue 3, 10 & Table 3 (2003); William M. Howard, Mandatory Arbitration of Employment Discrimination Disputes: Can Justice be Served, p. 132 & Table 12 (May 1995) (unpublished Ph.D. dissertation, Arizona State University) (on file with author); William Howard, Arbitrating Claims of Employment Discrimination, 50 Disp. Resol. J. 40, 45 (Oct.-Dec. 1995); Elizabeth Hill, Due Process At Low Cost: An Empirical Study of Employment Arbitration Under the Auspices of the American Arbitration Association, 18 OHIO STATE J. ON DISP. RES. 777, 791-92 (2003); Elizabeth Hill, Due Process At Low Cost: An Empirical Study of Employment Arbitration Under the Auspices of the American Arbitration Association, 18 OHIO STATE J. ON DISP. RES. 777, 788-89 (2003). For a good synthesis of the empirical record and a blueprint for future empirical research, see Sherwyn et al., Assessing the Case for Employment Arbitration: A New Path for Empirical Research, 57 STAN. L. REV. 1557, 1567-68 (2005).

22 There are two main exceptions to the dominant trend in the literature. First, the 1995 study by William Howard suggested that outcomes in arbitration were inferior to those in litigation, but subsequent scholarship has criticized the methodology that Howard employed. Second, more recent research by Hill and Eisenberg, cited above, suggested that arbitration may result in lower recoveries for employees earning less than $60,000. Yet as the authors themselves recognize, this study did not necessarily demonstrate that arbitration caused this outcome. Rather, given the well documented difficulties that this class of plaintiffs encounters in obtaining trial counsel, only very large meritorious suits ever actually reach court; by contrast, because arbitration is more cost-effective (or parties may elect to proceed pro se), a greater array of cases - both meritorious and non-meritorious - reach arbitration, creating the misimpression that arbitration is somehow responsible for these outcomes.

23 Demaine & Hensler, Volunteering to Arbitrate Through Predispute Arbitration Clauses: The Average Consumer's Experience, 67 L & Contemp Probs 55 (Winter/Spring 2004).

24 Id.

25 Id. at 72.

26 Public Citizen, The Arbitration Trap: How Credit Card Companies Ensnare Consumers (September 2007).

27 As I explain above, by various measures, other data on more typical disputes, such as employment arbitrations, shows that the system produces favorable outcomes for individuals.

28 See, e.g., Hooters of America, Inc. v. Phillips, 173 F.3d 933 (4th Cir. 1999) (refusing to enforce agreement with one-sided procedural rules).

29 Am. Arbitration Ass'n, Consumer Due Process Protocol (1998), available at http://www.adr.org/sp.asp? id=22019; Am. Arbitration Ass'n, Employment Due Process Protocol (1995), available at http://www.adr.org/sp.asp? id=28535; JAMS, JAMS Policy on Consumer Arbitrations Pursuant to Pre-Dispute Clauses: Minimum Standards of Procedural Fairness (2007), available at http://www.jamsadr.com/rules/consumer_min_ std.asp; JAMS, JAMS Policy on Employment Arbitration: Minimum Standards of Procedural Fairness (2005), available at http://www.jamsadr.com/rules/employment_Arbitration_min_stds.asp

30 See Peter B. Rutledge, Is Arbitration State Action? Does It Matter? (unpublished manuscript on file with author).

31 See, e.g., Walker v. Ryan's Family Steak Houses, Inc., 400 F.3d 370 (6th Cir. 2005) (refusing to enforce arbitration agreement where arbitral forum was nonneutral); McMullen v. Meijer, Inc., 355 F.3d 485 (6th Cir. 2004) (refusing to enforce arbitration agreement which granted exclusive control over arbitrator selection to employer); Murray v. United Food and Commercial Workers Intern. Union, 289 F.3d 297 (4th Cir. 2002)(refusing to enforce arbitration agreement after finding agreement unconscionable). Hooters of America, Inc. v. Phillips, 173 F.3d 933 (4th Cir. 1999) (refusing to enforce agreement with onesided procedural rules). See generally Born & Rutledge, International Civil Litigation in the United States 1106-08 (4th ed. 2006).

32 See EEOC v. Waffle House, Inc., 534 U.S. 279 (2002).

33 See, e.g., David Sherwyn et al., Assessing the Case for Employment Arbitration: A New Path for Empirical Research, 57 STAN. L. REV. 1557, 1574 & n.88 (2005); Samuel Estreicher, Saturns for Rickshaws: The Stakes in the Debate Over Predispute Employment Arbitration Agreements, 16 OHIO ST. J. DISP. RES. 59, 563 (2001); William M. Howard, Arbitrating Claims of Employment Discrimination, Dis. Resol. J., Oct.-Dec. 1995, at 40; William M. Howard, Mandatory Arbitration of Employment Discrimination Disputes (1995) (unpublished dissertation on file with author).

34 See William M. Howard, Arbitrating Claims of Employment Discrimination, DIS. RESOL. J., Oct.- Dec. 1995, at 40; William M. Howard, Mandatory Arbitration of Employment Discrimination Disputes (1995) (unpublished dissertation on file with author).

35 Lewis Maltby, Employment Arbitration and Workplace Justice,38 U.S.F. L. REV. 105, 107 (2003).

36 California Dispute Resolution Institute, Consumer and Employment Arbitration in California: A Review of Website Data Posted Pursuant to Section 1281.96 of the CCP (August 2004); David Sherwyn et al., In Defense of Mandatory Arbitration of Employment Disputes: Saving the Baby, Tossing out the Bath Water, and Constructing a New Sink in the Process, 2 U. PA. J. LAB. & EMP. L. 73, 99 (1999); GAO, Alternate Dispute Resolution: Employers' Experiences With ADR in the Workplace 19 (1997); Garry Mathiason & Pavneet Singh Uppal, Evaluating and Using Employer Initiated Arbitration Policies and Agreements: Preparing the Workplace for the Twenty-First Century, C902 A.L.I.-A.B.A. 875, 894 (1994)citing Rand Corporation's Institute for Civil Justice study indicating that average processing time from complaint to decision in arbitration = 8.6 months plus 20% cost savings to parties); GAO: How Investors Fare (May 1992).

37 Michael Delikat & Morris Kleiner, Comparing Litigation And Arbitration Of Employment Disputes: Do Claimants Better Vindicate Their Rights In Litigation?, American Bar Association Litigation Section Conflict Management, Vol. 6, Issue 3, 9(2003).

38 See California Dispute Resolution Institute, Consumer and Employment Arbitration in California: A Review of Website Data Posted Pursuant to Section 1281.96 of the Code of Civil Procedure (August 2004). See also Mark Fellows, The Same Result As In Court, More Efficiently: Comparing Arbitration and Court Litigation Outcomes, METRO. CORPORATE COUNSEL 32 (July 6, 2006) (discussing results of California report in light of other studies of win rates in arbitration).

39 Sherwyn et al., Assessing the Case for Employment Arbitration: A New Path for Empirical Research, 57 STAN. L. REV. 1557, 1569 (2005).

40 Compare Michael Delikat & Morris Kleiner, Comparing Litigation And Arbitration Of Employment Disputes: Do Claimants Better Vindicate Their Rights In Litigation?, American Bar Association Litigation Section Conflict Management, Vol. 6, Issue 3, 10 & Table 3 (2003) (arbitration yields higher recoveries), with William M. Howard, Mandatory Arbitration of Employment Discrimination Disputes: Can Justice be Served, p. 132 & Table 12 (May 1995) (unpublished Ph.D. dissertation, Arizona State University) (on file with author); William Howard, Arbitrating Claims of Employment Discrimination, 50 DISP. RESOL. J. 40, 45 (Oct.-Dec. 1995) (reaching opposite conclusion).

41 Theodore Eisenberg & Elizabeth Hill, Arbitration and Litigation of Employment Claims: An Empirical Comparison, 58 DISP. RES. J. 44, 44 (Nov 2003-Jan 2004).

42 GAO, Alternate Dispute Resolution: Employers' Experiences With ADR in the Workplace (1997).

43 Id. at 39.

44 Id. at 50.

45 Id. at 40.

46 Id. at 4, 19, 40.

47 U.S. Commission on the Future of Worker-Management Relations, Final Report (1994) (hereinafter "Dunlop Commission Report").

48 Id. at 50.

49 Factual Findings at 109-110 ("A conservative estimate is that for every dollar transferred in litigation to a deserving claimant, another dollar must be expended on attorney fees and other costs of handling both meritorious and non-meritorious claims under the legal program.") (footnote omitted).

50 Dertouzos et al., The Legal Consequences of Wrongful Termination (Rand Institute for Civil Justice 1988).

51 Id. at 38. To clarify the terminology, the final payment is the amount actually received by the claimant (which may be lower than the verdict due to post-verdict negotiations between the parties). The net payment represents the difference between the final payment and the claimant's legal fees.

52 Bickner et al., Developments in Employment Arbitration, 52 DISPUTE RES. J. 10, 78 (1997).

53 Christopher Drahozal, Unfair Arbitration Clauses, 2001 U. Ill. L. Rev. 695; Christopher Drahozal, Privatizing Civil Justice: Commercial Arbitration and the Civil Justice System, 9 KAN. J. L & PUB. POL'Y 578, 584-85 (2000). See also Stephen Ware, Case for Enforcing Adhesive Arbitration Agreements, 5 J. AM. ARB. 251, 256 n. 8 (2006).

54 Stiles v. Home Cable Concepts, Inc., 994 F. Supp. 1410 (M.D. Ala. 1998).

55 Lewis Maltby, Out of the Frying Pan, Into the Fire: The Feasibility of Post-Dispute Employment Arbitration Agreements, 30 WM. MITCHELL L. REV. 313, 320 (2003); Samuel Estreicher, Saturns for Rickshaws: The Stakes in the Debate Over Predispute Employment Arbitration Agreements, 16 OHIO ST. J. ON DISP. RES. 559, 567-68 (2001);

56 Lewis Maltby, Private Justice: Employment Arbitration and Civil Rights, 30 COLUM. HUM. RTS. L. REV. 29, 58 (1998).

57 They may be able to predict a likely dispute to a degree. They could base these predictions on their past experiences and the nature of the relationship between the parties.

58 15 U.S.C. §1226(a)(2).

59 Volkswagen Of America, Inc. v. Sud's Of Peoria, Inc., 474 F.3d 966 (7th Cir. 2007).

60 See Lewis Maltby, Out of the Frying Pan, Into the Fire: The Feasibility of Post-Dispute Employment Arbitration Agreements, 30 WM. MITCHELL L. REV. 313, 317 (2003).

61 Samuel Estreicher, Saturns for Rickshaws: The Stakes in the Debate Over Predispute Employment Arbitration Agreements, 16 OHIO ST. J. ON DISP. RES. 559, 563 (2001) (internal quotations omitted) (noting that employers are willing to agree to predispute arbitration because they "are willing to create a risk of liability in many cases they could have otherwise ignored in order to decrease the risk of a ruinous punitive damages award.")

62 See also Richard A. Bales, Normative Consideration of Employment Arbitration at Gilmer's Quinceanera, 81 Tulane L. Rev. 331, 357-58 (2006).