XII. LABOR STUDIES/LABOR UNIONS,
COLLECTIVE BARGAINING, DISPUTE
RESOLUTION, AND LABOR AND
EMPLOYMENT LAW REFEREED PAPERS

Short Circuiting Circuit City?
Judicial Enforcement of Mandatory
Employment Arbitration

Michael H. Leroy and Peter Feuille
University of Illinois at Urbana-Champaign

Abstract

     Our research provides an empirical assessment of judicial enforcement of mandatory employment arbitration agreements. Results from a sample of 264 districts and 96 appellate decisions for the period 1954 through September 1, 2002, show that enforcement of arbitration agreements was lowest before the Supreme Court's Gilmer decision. After Gilmer, district court enforcement of arbitration agreements substantially increased but appellate court enforcement substantially declined. Following the Court's Circuit City decision, district court enforcement of arbitration agreements remained unchanged, while the appellate court enforcement rate substantially increased. In addition, although enforcement of arbitration agreements varied widely by circuits, there was no discernible geographic pattern. In sum, federal courts enforce a majority of the contested arbitration agreements, but they are willing to deny enforcement to arbitration agreements they believe are unfair to employees.

     The displacement of discrimination lawsuits by mandatory arbitration is arguably the most significant and controversial employment law development since the early 1990s. Critics charge that employers control too much of this private dispute-resolution system (Moohr 1999; Sternlight 1996). In contrast, supporters argue that the court system does not serve employees well. Plaintiff lawyers take only 5 percent of employment discrimination complaints they receive (Howard 1995). Among accepted complaints, only about 3 percent go to trial and result in a verdict (Litras 2000). Even then, federal appeals courts reverse 44 percent of appealed cases won by employees (Bravin 2001).

     For decades, the U.S. Supreme Court has endorsed the use of arbitration to resolve workplace disputes. Initially, it supported the voluntary form of arbitration adopted by unions and employers in their collective bargaining agreements--"labor arbitration." More recently, it approved mandatory arbitration systems that employers imposed on nonunion employees--"employment arbitration"--primarily to avoid litigation over workplace disputes. Largely as a result of these Supreme Court decisions, employment arbitration has expanded substantially and now covers many millions of employees. As a result, it is important to understand how much deference the courts demonstrate to these employer-promulgated dispute-resolution mechanisms. Accordingly, in this study we measure the extent to which federal courts enforced mandatory employment arbitration agreements during the 1954-2002 period.

Legal Overview of Mandatory Employment Arbitration

     Congress enacted the Federal Arbitration Act ("FAA") in 1925. This occurred at a time when businesses increasingly used arbitration as a method to resolve commercial contract disputes but were frustrated by judges who denied enforcement to arbitrator rulings. The FAA directed courts to enforce awards, with few exceptions. The federal courts expanded the use of the FAA in the 1940s and 1950s to enforce labor arbitration awards. The Supreme Court ended this role for the statute in 1960 by shifting the legal basis for enforcement of labor arbitration awards to the National Labor Relations Act. In the famous Steelworkers Trilogy decisions that year, the Court resoundingly endorsed the labor arbitration process.

     The FAA reemerged as an employment law in 1991. Gilmer, a fired securities broker, sued his employer for age discrimination. His former employer countered that Gilmer had signed a contract to arbitrate any dispute with them and asked the court to order that arbitration be used to resolve this age discrimination claim. In Gilmer v. Johnson/Interstate Lane Corp. (1991), the Supreme Court enforced this arbitration agreement. As a result, Gilmer's age discrimination dispute was arbitrated rather than litigated.

Gilmer had great impact. In subsequent years, many employers adopted arbitration policies for their nonunion employees, to the point where the number of nonunion employees covered by these employment arbitration agreements may soon be approaching the number of employees covered by labor arbitration procedures in union contracts; however, two major uncertainties clouded Gilmer. The arbitration agreement signed by Gilmer was not drafted with employment discrimination claims in mind. It was intended to resolve commission and client disputes. Thus, it could be characterized as a business contract and therefore not apply to arbitration agreements for employment claims. How far would courts extend Gilmer to ordinary employment disputes?

     Second, Gilmer polarized federal judges. Some judges followed the Gilmer precedent because they believed arbitration improved employee access to an adjudicatory process. These judges extended Gilmer to require the arbitration of discrimination claims arising under Title VII of the Civil Rights Act, the Americans with Disabilities Act (ADA), and other federal statutes. Other judges disagreed. Ruling for employees who petitioned to proceed with discrimination lawsuits and avoid arbitrations, they found that mandatory arbitration agreements were not enforceable contracts.

     This conflict boiled over in the Ninth Circuit Court of Appeals, which openly rebelled against Gilmer and put forth an independent interpretation of the FAA. In Duffield v. Robertson Stephens & Co. (1998), the Ninth Circuit ruled that Congress did not intend in the 1991 Civil Rights Act to allow arbitration to preclude an employment discrimination lawsuit. This created a significant exception to Gilmer's general rule of enforcement of arbitration agreements, which most other federal circuit courts had adopted.

     The court took a bolder step a year later. In Craft v. Campbell Soup Co. (1999), the Ninth Circuit struck at the heart of Gilmer. The FAA specifically excludes from coverage the employment contracts of "seamen, railroad employees, or any other class of workers engaged in foreign or interstate commerce." Craft examined congressional regulation of the employment relationship under its commerce powers as of 1925, when the FAA was enacted. The Craft court concluded this FAA exclusion should be interpreted broadly, and thus it ruled that most employment arbitration agreements were not enforceable under the FAA. This ruling had the effect of nullifying Gilmer in most workplaces within the Ninth Circuit's geographic jurisdiction.

     The Supreme Court answered this provocation in Circuit City, Inc. v. Adams (2001). By a 5-4 vote, it held that the FAA's exclusion should be interpreted narrowly, and thus it applies only to transportation workers in interstate commerce. The majority reasoned that, if this exclusion was so broad as to cover all employment contracts, there would be no point in its specific reference to maritime and rail workers. This ruling sent a strong signal to the lower federal courts that the Supreme Court continued to support mandatory arbitration agreements.

Research Questions and Methods

     To date researchers have not systematically examined the track record of federal court enforcement of mandatory employment arbitration agreements. In this study, we attempt to partly fill this gap by providing a longitudinal portrait of federal district and circuit court rulings during 1954-2002. The first part of this time frame covers the pre-Gilmer period, ending with the Court's May 13, 1991, decision. We answer two questions in this part of our analysis. When did federal courts begin to rule on employee challenges to individual arbitration agreements that arise under the FAA? Second, though the Supreme Court provided much guidance about judicial review of labor arbitration since the late 1950s, it said very little about individual employment arbitration until Gilmer. Without this guidance, how did federal courts treat these individual arbitration contracts? Specifically, how often did they enforce them?

     The second period is bounded by the Gilmer and Circuit City decisions (May 14, 1991-March 21, 2001). This marks the first period of explicit Supreme Court regulation of individual employment arbitration. Scholarly commentary, observing that Gilmer sent lower courts a strong signal approving the use of this form of arbitration, assumes that lower court behavior was affected by this ruling (Estreicher 1997). Is there empirical support for this view? How much did judicial behavior change compared to the baseline (pre-Gilmer) period?

     As with Gilmer, there is a widely held view that Circuit City strengthened the arbitration signal. Only a short time has passed since this decision, but the courts have decided enough cases to enable us to assess the initial impact of Circuit City. Accordingly, we use the post-Circuit City period to determine whether the Court's second major embrace of mandatory employment arbitration made a noticeable difference in judicial enforcement rates of arbitration agreements.

     We constructed our sample of court rulings as follows. We examined only reported federal court decisions. Our sample includes only decisions involving the enforcement of arbitration agreements with individual employees. Thus, we exclude discrimination claims asserted by union-represented workers in which the employer sued to compel arbitration of these claims under the collective bargaining agreement.

     Using Westlaw's online reporting service, we began by locating and analyzing employment arbitration cases cited in the Gilmer and/or Circuit City landmark precedents. This online program creates a Web link to every court case cited in the body of the Gilmer and Circuit City decisions. This enabled us to work back in time by reading each of these earlier cases and including in our sample those that met our criteria. Within each of these decisions, another set of Web-linked decisions also was reported. We included those meeting our criteria and continued this iterative process until we identified and included all the nonduplicated pre-Gilmer and pre-Circuit City cases identified by this process. Then, looking forward in time, we read all cases that cited Gilmer and/or Circuit City through a similar iterative process until all leads were exhausted.

     From each case we extracted the following information: year of decision; type of employment; employee characteristics; legal claim of party resisting arbitration (e.g., Title VII, ADA, ADEA, etc.); legal argument to resist arbitration (contract was adhesive, waiver was inadequate, cost of arbitration was prohibitive, etc.); party who prevailed at district and circuit court; district and circuit court ruling; and the length of time to litigate the arbitrability issue.

Results

     We found a sample of 264 usable district and 96 appellate decisions for the period 1954 through September 1, 2002. In Tables 1 and 2, we present our summaries of court rulings that ordered or denied arbitration. Table 1 reports data for 264 district court decisions, and Table 2 covers 96 circuit court rulings.

     1. Enforcement of arbitration agreements was lowest before Gilmer. Before Gilmer, federal district and circuit courts enforced 51 percent and 60 percent of individual arbitration agreements, respectively. Often, the facts and issues in these cases differed from those after Gilmer. Many were commission or bonus disputes involving securities brokers. Also, a few cases involved a role reversal. Employers sought to escape their own arbitration agreements in favor of litigation when an employee quit to join a competitor. The employer, often a securities brokerage, sued to restore the former employment relationship, prevent direct competition, or order the broker not to take clients to a competing firm.

     2. After Gilmer, district court enforcement of arbitration agreements substantially increased, but appellate court enforcement substantially declined. Unlike pre-Gilmer cases, most of these post-Gilmer decisions involved discrimination claims. Although the observation period was much shorter (1991-2001), there were many more cases (171 compared to 39 district decisions, and 61 compared to 20 appellate decisions). Whereas the district court enforcement rate rose from 51 percent to 66 percent, the rate at which these courts allowed lawsuits to proceed declined only slightly (e.g., district court dismissals of arbitration fell from 31 percent to 28 percent). The gain in contract enforcement came from a sharp reduction in partial arbitration rulings. Before Gilmer 10 percent of district court decisions denied enforcement of an agreement to arbitrate a federal employment claim, but compelled arbitration of a companion state law claim. These mixed rulings, however, account for only 1 percent of the post-Gilmer sample.

     In contrast, circuit courts registered a significant decline in enforcing arbitration agreements. Employees were ordered to arbitrate their disputes in only 49 percent of these cases. This was an 11 percentage point drop compared to pre-Gilmer decisions. In spite of the strong proarbitration signal the Supreme Court sent to the lower courts in Gilmer, many appellate courts found ways to partly or completely reject half of the employer motions to enforce arbitration agreements in the 10 years following Gilmer.

     3. Following Circuit City, district court enforcement of arbitration agreements remained unchanged, whereas the appellate court enforcement rate substantially increased. The post-Circuit City sample is much smaller, of course, because of the very short measurement period (March 22, 2001-September 1, 2002). Still, this subsample contained 69 decisions, or nearly 20 percent of all cases reported for the entire post-Gilmer period. District court enforcement was essentially unchanged, at 67 percent, while the rate dropped slightly for court decisions that allowed employee complaints to proceed as lawsuits (see dismissal of arbitration, at 24 percent). Circuit courts registered a sharp increase in arbitration enforcement decisions, from 49 percent to 73 percent. It is important to keep in mind, however, that this finding is based on only 15 appellate decisions. Only three circuit court rulings (20 percent) completely denied arbitration.

     4. Enforcement of arbitration agreements varied widely by circuits, with no discernible geographic pattern. An analysis compared court rulings grouped by appellate circuits. We found a wide variation in arbitration enforcement rates (not reported in our tables).

     Looking first at district court decisions as arranged by their respective circuits, the arbitration enforcement rate ranged from 31 percent in the Tenth Circuit to 78 percent in the Eighth Circuit. Surprisingly, a low enforcement rate occurred among district courts in the Fourth Circuit (47 percent), a jurisdiction regarded as conservative and proemployer. In addition to the high district court enforcement rate in the Eighth Circuit, the next highest rate occurred in the Second Circuit (72 percent).

     Appellate trends were even harder to identify, because the sample of 96 circuit decisions was spread over nearly 50 years. Remarkably low arbitration enforcement rates occurred in the Tenth (25 percent), Ninth (39 percent), District of Columbia (40 percent), and First (50 percent) Circuits. Shortly after we ended our data collection, the Ninth Circuit, in E.E.O.C. v. Luce, Forward, Hamilton & Scripps (2002), reversed its course when it declared, "In Circuit City, the Supreme Court so directly undermined the reasoning behind Duffield, that we conclude it has lost its status as valid precedent" (p. 1002). In a rare moment of humility, this court added, "Since our Duffield decision in 1998, our Sister Circuits as well as the Supreme Courts of California and Nevada have unanimously repudiated its holding. Duffield, like Bikini Atoll, now sits ignominiously alone awaiting remediation" (id.). In short, although Duffield is no longer the law in the Ninth Circuit, its impact on our results cannot be overlooked. Nevertheless, this major shift implies that the Ninth Circuit's enforcement rate will likely increase in the future.

Conclusions

     Recent research provides a context for interpreting these results. Malin (2001) and Green (2000) observe that Gilmer created numerous "fallout" issues that are now occupying many lower courts, such as repeat-player bias, discovery, filing deadlines, remedies, and cost allocation. They believe that Gilmer approved a wide-ranging dispute-resolution system without defining due process safeguards. Silverstein (2001) concludes that Gilmer displaced statutory employment regulation with contract law. She shows that employment and commercial relationships are not the same, because workers suffer power and information imbalances compared with their employers. The post-Gilmer results for district and appellate courts, and post-Circuit City results for district courts, are consistent with Malin's and Green's fallout theories. Numerous decisions in this sample attempt to re-create the procedural and substantive protections in arbitration that apply in employment litigation.

     These results paint a mixed picture of the future of mandatory employment arbitration. The good news for arbitration proponents is that the federal courts enforce a majority of the contested arbitration agreements. At the same time, this research indicates that the courts are willing to deny enforcement of arbitration agreements they believe are unfair to employees. This is not surprising, in light of the fact that Gilmer and Circuit City were long on judicial signals but short on specific guidance. The longitudinal findings in this study suggest that judicial enforcement of employment arbitration is not nearly as certain as the leading precedents imply. As a result, the Supreme Court's proarbitration signals have been short-circuited in many lower courts. Our findings also suggest that many years will pass before judicial regulation of employment arbitration achieves consistency across the country. An alternative to this protracted morass of conflicting court rulings is for Congress to pass rules for mandatory employment arbitration. The legislative history of such attempts in the post-Gilmer period, however, indicates that statutory regulation of employment arbitration is highly unlikely. As a result, the evolution of this controversial dispute-resolution process promulgated to avoid litigation will unfold--ironically--through many hundreds of lawsuits.

References

Bravin, Jess. 2001. "U.S. Courts Are Tough on Job-Bias Suits." The Wall Street Journal, (July 16), p. A2.

Circuit City Stores, Inc. v. Adams. 2001. 121 S.Ct. 1302.

Craft v. Campbell Soup Co., Inc. 1999. 177 F.3d 1083 (9th Cir.).

Duffield v. Robertson Stephens & Co. 1998. 144 F.3d 1182 (9th Cir.).

E.E.O.C. v. Luce, Forward, Hamilton & Scripps. 2002. 303 F.3d 994, 1002 (9th Cir.).

Estreicher, Samuel. 1997. "Predispute Agreements to Arbitrate Statutory Employment Claims." New York University Law Review, Vol. 72 (December), pp. 1344-75.

Gilmer v. Johnson/Interstate Lane Corp. 1991. 500 U.S. 20.

Green, Michael Z. 2000. "Debunking the Myth of Employer Advantage from Using Mandatory Arbitration for Discrimination Claims." Rutgers Law Journal, Vol. 31 (winter), pp. 399-471.

Howard, William M. 1995. "Arbitrating Claims of Employment Discrimination." Dispute Resolution Journal, Vol. 50 (October), pp. 40-49.

Litras, Marika. 2000. "Bureau of Justice Statistics Report on Civil Rights, Complaints Filed in U.S. District Court." Bureau of National Affairs, Daily Labor Report, No. 14 (January 20), p. E-10.

Malin, Martin H. 2001. "Privatizing Justice But By How Much? Questions Gilmer Did Not Answer." Ohio State Journal on Dispute Resolution, Vol. 16, pp. 589-631.

Moohr, Geraldine Szott. 1999. "Arbitration and the Goals of Employment Discrimination Law." Washington and Lee Law Review, Vol. 56 (spring), pp. 395-460.

Silverstein, Eileen. 2001. "From Statute to Contract: the Law of the Employment Relationship Reconsidered." Hofstra Labor and Employment Law Journal, Vol. 18 (spring), pp. 479-5'.

Sternlight, Jean R. 1996. "Panacea or Corporate Tool?: Debunking the Supreme Court's Preference for Binding Arbitration." Washington University Law Quarterly, Vol. 74 (fall), pp. 637-71.