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

Private Justice and Public Policy:
Whose Voice Prevails in Arbitration?

MICHAEL H. LEROY AND PETER FEUILLE
University of Illinois at Urbana-Champaign

Abstract

      Data and findings on federal court review of arbitration awards from 1990–2000 are presented. In the arena of voluntary labor arbitration awards, the results replicate findings of earlier research for the period 1960–1990. District court enforcement is approximately 70 percent, and only slightly lower for appellate courts (66 percent). Results for individual employment awards show a higher rate of enforcement, about 84 percent for district and circuit courts. This discrepancy may be explained by the fact that the grounds for reviewing employment awards are usually governed by the Federal Arbitration Act and not the Steelworkers Trilogy. Although a higher percentage of these awards are enforced, there is evidence that courts are moving in the direction of the labor arbitration model by exercising closer review of arbitrator rulings.

      Grievance procedures long have been considered “remedial voice” mechanisms for employees. Grievance arbitration has been viewed as a particularly employee-favorable component of these remedial voice procedures, for arbitration enables an aggrieved employee to seek redress from a decision maker independent of management.

      During the 1990s the arbitration of employee grievances expanded in American workplaces. This expansion did not occur in the unionized sector, for grievance arbitration provisions have been an almost universal feature in union contracts for decades (Eaton and Keefe 1999). Instead, this expansion occurred as a result of thousands of nonunion employers adopting arbitration as a condition of employment (U.S. GAO 1995). The result is that we now have systems of “labor arbitration” (grievance arbitration in the union sector) and “employment arbitration” (grievance arbitration in the nonunion sector) covering many millions of employees.

      Both types of arbitration have generated controversy. Labor arbitration is a very stable feature of unionized employment relationships, but the judicial review of appealed labor arbitration awards continues to generate sparks among union and employer advocates and arbitrators. Employment arbitration has generated a great deal of controversy, much of it via lawsuits over the legal status of this type of arbitration process and the awards it produces.

      Because these two types of arbitration have developed with little legislative guidance, the obligation to resolve these disputes has fallen upon the courts. In turn, an informed understanding of the public policy status of these two systems of private justice must involve a comprehensive analysis of how the courts have reacted to challenged arbitrator decisions. Accordingly, below we present the results of our analysis of hundreds of decisions issued by the federal courts involving appeals of labor and employment arbitration awards. Our results should yield a more informed portrait of the evolving public policy toward these two private justice systems.

Arbitral Finality and Judicial Review: An Oxymoron?

      Both labor arbitration and employment arbitration procedures state that the arbitrator’s ruling (award, decision, etc.) is final and binding. This finality is arguably the key attraction of the arbitration process. Arbitration allows the parties to privately devise and operate an adjudication procedure in which the disputes referred to arbitrators will be resolved once and for all--at least in theory. In this manner, labor arbitration provides unions and employers a substitute for the use of work stoppages to resolve grievances, and a dispute decision system that is faster and cheaper than litigation. Employment arbitration is not adopted as a substitute for work stoppages but as an alternative to litigation. It also offers a faster and less expensive way to resolve disputes, and it arguably offers employees more access to arbitrators than they would have to the courts. However, these efficiency advantages are premised upon arbitrator decisions being final and binding. If arbitrator rulings are not final, the process would become merely an intermediate stop on the way to a final, and much more costly, decision in some other forum, most likely the courts.

      At the same time, these private judging systems cannot stand above our societal system of public law. For instance, it is difficult to persuasively argue that an award that directed one or both of the disputing parties to do something that violated a clearly articulated public policy should be unreviewable and allowed to stand. At a more prosaic level, there needs to be a method to seek redress from an arbitration award that is the result of the arbitrator’s collusion with one party at the expense of the other, or from an award in which the arbitrator clearly exceeded the decision authority provided to her by the procedure. As a result, there must be a balance between the parties’ need and desire for finality in the private judging system they have adopted, and the need for society to provide a safety valve through which arbitration users can seek redress if they are convinced the arbitrator’s decision is intolerable.

Union/Employee Voice and the Finality of Labor Arbitration Awards

      The arbitration of disputes over the interpretation of collective bargaining agreements became widely established in unionized American workplaces during the 1940s and 1950s (Nolan and Abrams 1983). In 1960 the U.S. Supreme Court strongly endorsed the use of the labor arbitration process in a series of three decisions issued on the same day known as the Steelworkers Trilogy. For our purposes, the most pertinent of the Court’s rulings emerged in the third of these decisions, Steelworkers v. Enterprise Wheel and Car Corp. (363 U.S. 593, 1960). In Enterprise, the U.S. Supreme Court issued strongly worded instructions to the lower courts directing them to refrain from reviewing the merits of arbitrators’ rulings.

      At the same time, however, the Enterprise Court said that arbitral authority is not absolute and that lower courts may vacate appealed awards when arbitrators’ awards do not draw their “essence” from the collective bargaining agreement. In other words, the Court said that courts reviewing appealed labor arbitration awards generally should follow a “hands-off” policy in deference to the arbitration decision process chosen by the parties but that there are some lines that arbitrators cannot cross. The result is that the lower courts asked to review arbitration awards need to engage in a “hands-off/hands-on” balancing act that is easier to describe than to follow (Sharpe 2000).

      The available evidence indicates that the vast majority of labor arbitration awards are indeed final. Researchers have estimated that less than 1 percent of the labor arbitration awards issued in the private sector are appealed to the courts (Estreicher and Harper 1993; Feuille and LeRoy 1990). This high rate of award compliance seemingly indicates that arbitral finality is widespread and hence any accompanying controversy should be minimal. However, judicial review of labor arbitration awards continues to be a vexing problem, as seen in two continuing developments.

      First, three times in the past 18 years the Supreme Court has believed it necessary to issue decisions reaffirming the “hands-off” portion of the balancing act lower courts should use when reviewing appealed labor arbitration awards. Specifically, in W.R. Grace and Co. v. Rubber Workers Local 759 (461 U.S. 757, 1983), Paperworkers International Union v. Misco (484 U.S. 29, 1987), and Eastern Associated Coal Corp. v. United Mine Workers (531 U.S. 57, 2000), the Court tackled the problem of how the courts should respond to appeals in which the disgruntled party sought to nullify awards on grounds that the offending awards violated “public policy.” In each of these rulings the Court said that (1) federal courts could vacate awards that violate public policy but (2) only in very limited circumstances.

      Second, ever since the issuance of the Trilogy decisions in 1960, and continuing to the present, commentators continue to criticize the courts for being too willing to second-guess arbitrators and undo their decisions (Feller 1993; Gottesman 1989; Sharpe 2000). Much of this criticism understandably comes from labor arbitrators, who don’t like to see their rulings reversed.

      These developments indicate a need for a careful analysis of the judicial review phenomenon. In contrast to most of the other commentary on this subject, which usually consists of a close textual analysis of a handful of cases, we use a previously developed methodology for locating and analyzing reported federal court decisions involving appealed labor arbitration awards (LeRoy and Feuille 1991). Specifically, we used the online version of Westlaw to locate the reported federal court decisions issued during the 1991–2000 period that involved a challenge to a labor arbitration award. After locating each court decision, we used a survey form to extract standardized information about the appealed award and about the court’s ruling. We present these results in Table 1.

      The Table 1 results confirm the findings offered by LeRoy and Feuille a decade ago. First and foremost, challenges to labor arbitration awards are overwhelmingly driven by employer dissatisfaction with arbitrator decisions favoring unions/employees. Eighty-five percent of the appealed awards in this sample period favored the union. This is a continuation of the phenomenon noted in their analysis of challenges to awards during the 1960–1990 period, when 80 percent of the appealed awards favored the union (Feuille and LeRoy 1990). Second, the district courts rejected 70 percent of these challenges during the past decade, a result that is very similar to the 71.8 percent rate at which the district courts rejected challenges to awards during the 1960–1990 period (LeRoy and Feuille 1991). Third, however, the federal courts show substantial regional variation in their willingness to overturn awards. Leaving aside the judicial circuits in which fewer than 10 decisions are reported, the district courts in the federal Second Circuit upheld awards 86 percent of the time, while the district courts in the Fourth Circuit upheld awards only 45 percent of the time. In other words, Table 1 confirms the earlier finding that challenges to labor arbitration awards have significantly greater chances of succeeding depending upon the region of the country where the challenge is lodged.

      Most important, the data in Table 1 show that criticisms of the federal judiciary as being increasingly more willing over time to overturn labor arbitration awards are without merit. There are always selected court decisions that arbitration proponents can point to as justification for claims that this or TABLE 1 that federal court is impermissibly intruding upon the labor arbitrator’s authority (Sharpe 2000). However, the Table 1 results, when combined with the results presented by LeRoy and Feuille (1991) a decade ago, show that during the past four decades the federal courts have been quite stable in their responses to requests to overturn labor arbitration awards. Specifically, they have approved less than one-third of these requests to vacate the appealed awards.

Employer Voice and the Finality of Employment Arbitration Awards

      Observers agree that union-negotiated/supported labor arbitration provides employees with an institutionalized and stronger form of remedial voice to challenge adverse managerial decisions than is customarily available to their peers in nonunion workplaces. As this description implies, unions typically propose to employers that arbitration be adopted (in return for the unions’ agreement to a no-strike clause), and unions are the moving party in seeking arbitral determinations of disputed grievances.

      During the past 10 or so years, employment arbitration has become much more widespread in nonunion workplaces (Bickner, Ver Ploeg, and Feigenbaum 1997; Howard 1995; McDermott 1995; U.S. GAO 1995). In contrast to the union sector, in nonunion establishments arbitration is not adopted in response to employee preferences. Instead, it is unilaterally imposed by the employer, often as a mandatory condition of employment. (Indeed, its critics often call it “mandatory arbitration”.) These employers have adopted arbitration primarily as a litigation avoidance mechanism. This behavior is a rational response to an increased employee willingness to sue their employers, as seen in the fact that the number of employment discrimination lawsuits filed in federal district courts more than tripled between 1990 and 1998 (BNA 2000).

      The employer incentive to adopt arbitration received a big boost in May 1991 when the U.S. Supreme Court said in Gilmer v. Interstate/Johnson Lane Corp. (500 U.S. 20, 1991) that a mandatory arbitration provision could be used to resolve an age discrimination complaint arising under the Age Discrimination in Employment Act. Since then, the federal courts have extended Gilmer to other kinds of discrimination complaints (race, sex, disability, etc.), and they usually have compelled plaintiff employees to pursue their discrimination claims via arbitration if the employer demonstrates the presence of a valid arbitration agreement between the parties. The court issued another pro-arbitration statement when it ruled in Circuit City Stores, Inc. v. Adams (121 S.Ct. 1302, 2001) that the Federal Arbitration Act covers most employment contracts. This ruling has the effect of affirming Gilmer by facilitating the employer-sought enforcement of mandatory arbitration provisions over employee objections. A subsequent court ruling in E.E.O.C. v. Waffle House (121 S.Ct. 754, 2002) limited the reach of Gilmer and Circuit City, but only modestly, by holding that an agency charged with enforcing discrimination laws (the EEOC) is not a party to a mandatory agreement and therefore is not barred from seeking relief for employees who are parties to such agreements.

      The heated controversy over employment arbitration centers around two key dimensions. The first is the mandatory, take-it-or-leave-it nature of the arbitration agreement signed by the employee. Many employers present arbitration to prospective and current employees as a mandatory condition of employment, and employee advocates believe this is an inherently coercive and hence unacceptable method for adopting what should be a voluntary mechanism for resolving disputes. Second, these arbitration provisions are broadly written to cover “any dispute” arising from a person’s employment or termination of employment, and the breadth of the typical arbitration clause encompasses claims of employment discrimination. Critics argue that the public interest in the elimination of employment discrimination can be ensured only if the nation’s antidiscrimination statutes are enforced via public enforcement mechanisms. Expressed another way, this criticism of employment arbitration is based on the belief that public justice should not be privatized. The critics are not persuaded by the pro-arbitration arguments that arbitration provides greater access to a third-party adjudicator than does litigation, is faster and less expensive, and can offer the prevailing employee the same remedies available in court. These arguments have been flourishing ever since the Supreme Court’s 1991 ruling in Gilmer, and this controversy shows no signs of diminishing (Bales 1997; Stone 1999; Zack 1999).

      But what is happening closer to the workplace with employment arbitration? The available evidence is sketchy, but it suggests that (1) employment arbitration provisions now cover several million nonunion employees; (2) where valid arbitration agreements exist, the lower federal courts have generally followed Gilmer and compelled the use of arbitration over employee objections when employees have filed discrimination lawsuits (Cole 2000); and (3) the number of employment arbitration awards issued each year is increasing.

      There has been only a modest amount of research into how employment arbitration has actually worked in practice. Looking at outcomes, one study of awards issued in 510 cases processed by the American Arbitration Association involving a claim of employment discrimination during the 1992– 1994 period found that (1) employees won something in 68 percent of these awards, and (2) the median monetary award to prevailing employees was $32,950 (Howard 1995). For comparison purposes, in 21,518 employment discrimination lawsuits litigated in federal district court during that same period, the author found that employees recovered something 71 percent of the time if they settled, but only 28 percent of the time if they received a trial verdict (38 percent of the time if they received a jury verdict and only 19 percent of the time in bench trial verdicts; Howard 1995).

      Most of the research on the legality of employment arbitration has focused on pre-arbitration disputes of the type addressed by the Supreme Court in Gilmer and Circuit City: Should an employee with a discrimination complaint be compelled to use the employer-imposed mandatory arbitration procedure when the employee prefers to pursue her claim in court? This research shows that most courts compel arbitration (Cole 2000), but it offers no insight into what happens after the arbitration process has been completed. In particular, what happens when one party in an employment arbitration is sufficiently disgruntled with the award that they ask a court to vacate it? Our research has not uncovered any prior studies of how the courts have reacted to appeals of employment arbitration awards. Accordingly, we used the Westlaw electronic search capability described above to locate reported federal court decisions involving an appealed employment arbitration award (awards issued in nonunion workplaces). After locating these cases, we extracted standardized data from each judicial opinion about the nature of the dispute that had been arbitrated and the nature of the legal appeal. We report our results in Table 2.

      Because employment arbitration is a more recent and less widespread phenomenon than labor arbitration, it is not surprising that our search uncovered only a small number of court rulings involving appealed employment arbitration awards. With the “small sample” caveat in mind, the Table 2 data show that most of these employment arbitrations were triggered by a termination, that employers prevailed in these awards more often than employees, that the disgruntled party challenging the award is usually the employee, and that the federal district and circuit courts deny most of these appeals and uphold the award.

      How should we interpret the arbitration outcomes in these thirty-three cases? In particular, the fact that employees win something only 39 percent of the time in these cases [thirteen of thirty-three awards in which employees prevail (7) or receive a “split” decision (6)], compared with the fact that unions/ employees prevail fully or partly about 50 percent of the time in labor arbitration (American Arbitration Association 1993), may suggest that employment arbitration is a process biased in favor of employers. Employment arbitration indeed may be a process that favors employers more than employees, but there are at least two persuasive reasons why the arbitration outcome data presented here should not be used to support that assessment. First, the sample of arbitration decisions portrayed in Table 2 is highly selective and consists only of those taken to court. We have no data reporting the rate at which employees prevailed in the thousands of employment arbitration rulings issued during the past decade.

      Second, the union/employee win rate in labor arbitration is not an appropriate benchmark for assessing outcomes in the employment arbitration arena. The employment arbitration cases reported in Table 2 served as substitutes for court adjudication of these employee claims. Accordingly, an appropriate assessment yardstick would be the rate at which plaintiffs prevail in similar disputes that are litigated to a verdict in court. Although not all of these appealed awards involved a claim of employment discrimination, at least 20 of them did (based on the issues raised on appeal to the courts). Accordingly, we compare the rate that employees prevailed in this sample of awards with the rate that employee-plaintiffs prevail in employment discrimination lawsuits. In that vein, the employees in the 33 awards portrayed here prevailed in the arbitration process at a somewhat higher rate (39 percent) than did the employment discrimination plaintiffs in two studies: (1) plaintiffs who received a trial verdict in federal district court during 1992–1994 (28 percent) as reported above in Howard (1995), and (2) plaintiffs in another study of almost 8,500 employment discrimination federal court trial verdicts issued during 1990–1998 (30 percent; BNA 2000).

      The monetary amounts awarded to prevailing employees in the nine awards where this information was available similarly should be cautiously interpreted. Some of these cases involved disputes over executive compensation agreements that arbitrators determined had been breached by employers. As the wide range of amounts awarded to prevailing employees suggest, and as the huge difference between the median and mean amounts awarded to prevailing employees suggest, a few of these rulings resulted in very large amounts awarded to highly paid employees. Among other things, the dollar amounts reported in Table 2 indicate that the median amount awarded to prevailing employees is a much more useful assessment yardstick than the mean amount awarded.

Conclusions

      Labor arbitration, as the final step in the union-negotiated/supported grievance procedure, is a classic example of an employee remedial voice mechanism by which employees, with union assistance, can challenge what they believe are adverse decisions by their employer. The evidence presented here indicates that when challenges are filed against labor arbitration awards, these challenges usually result from employer dissatisfaction with the labor arbitrator’s ruling. Perhaps this can be viewed as evidence of an employer belief that the challenged arbitrator provided too much voice to the employee-grievant. This evidence also indicates that the courts uphold less than one third of these challenges, which means that most of the time the courts abide by the arbitrator’s interpretation of whose voice should prevail in the resolution of the grievance.

      Employment arbitration ostensibly plays a similar remedial voice role by allowing nonunion employees to challenge adverse decisions by their employer. However, the fact that employment arbitration procedures are unilaterally designed and implemented by employers as a method to avoid employee-initiated litigation, and the fact that most legal challenges to employment arbitration awards are filed by employees, indicates that the remedial voice opportunity being provided to employees is occurring through the employer’s preferred voice mechanism. In other words, the two sets of legal challenges to arbitration awards examined here indicate that employers and employees have divergent views of whose interests are favored in these two kinds of arbitration systems.

      At the same time, the evidence reviewed here indicates that the federal courts generally treat labor and employment arbitration awards for what they are designed to be: a private judge’s binding ruling in a workplace dispute that the disputing parties could not resolve by themselves. Our two samples of reported federal court decisions from the past decade show that the federal courts are unlikely to accommodate the wishes of the disgruntled parties who seek to escape adverse arbitral rulings issued in either type of arbitration. Whatever one’s views of the appropriateness of the arbitral processes or decisions being disputed in these lawsuits, the public policy portrayed in the federal court decisions reviewed here demonstrates considerable deference toward the private justice voice of the arbitrators asked to decide these workplace disputes.

References

American Arbitration Association. 1993. Study Time (a quarterly newsletter for labor arbitrators), no. 4.

Bales, Richard. 1997. Labor and Employment Law Compulsory Arbitration: The Grand Experiment In Employment. Ithaca, NY: Cornell University Press.

Bickner, Mei L, Christine Ver Ploeg, and Charles Feigenbaum. 1997. “Developments in Employment Arbitration.” Dispute Resolution Journal, Vol. 52, no. 1 (January), pp. 8– 15, 78–84.

Bureau of National Affairs. 2000. “Employment Bias Cases in Federal Court Almost Tripled in Previous Decade, DOJ Says.” Daily Labor Report, January 20, pp. A1–A2, E5–E16. Circuit City Stores, Inc. v. Adams, 121 S.Ct.1302 (2001).

Cole, Sarah Rudolph. 2000. “Managerial Litigants? The Overlooked Problem of Party Autonomy in Dispute Resolution.” Hastings Law Journal, vol. 51 (August), pp. 1199– 1263.

E.E.O.C. v. Waffle House, 121 S.Ct. 754 (2002).

Eastern Associated Coal Corp. v. United Mine Workers, 531 U.S. 57 (2000). Estreicher, Samuel, and Michael Harper. 1993. Cases and Materials on the Law Governing the Employment Relationship, 2d ed. St. Paul: West Group.

Eaton, Adrienne E., and Jeffery H. Keefe. 1999. “Introduction and Overview.” In Adrienne E. Eaton and Jeffery H. Keefe, eds., Employment Dispute Resolution and Worker Rights in the Changing Workplace. Champaign, IL: Industrial Relations Research Association, pp. 1–26.

Feller, David E. 1993. “End of the Trilogy: The Declining State of Labor Arbitration.” Arbitration Journal, Vol. 48, no. 3 (September), pp. 18–29.

Feuille, Peter, and Michael H. LeRoy. 1990. “Grievance Arbitration Appeals in the Federal Courts: Facts and Figures.” Arbitration Journal, Vol. 45, no. 1 (March), pp. 35–47.

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

Gottesman, Michael H. 1989. “Enforceability of Awards: A Union Perspective.” Proceedings of the Forty-First Annual Meeting of the National Academy of Arbitrators. Washington, DC: Bureau of National Affairs, pp. 88–96.

Howard, William F. 1995. “Arbitrating Claims of Employment Discrimination.” Dispute Resolution Journal, Vol. 50, no. 4 (October), pp. 40–50.

LeRoy, Michael H., and Peter Feuille. 1991. “The Steelworkers Trilogy and Grievance Arbitration Appeals: How the Federal Courts Respond.” Industrial Relations Law Journal, Vol. 13, no. 1, 78–120.

McDermott, E. Patrick. 1995. “Survey of 92 Key Companies Using ADR to Settle Employment Disputes.” Dispute Resolution Journal, Vol. 50, no. 1 (January), pp. 8– 13.

Nolan, Dennis R., and Roger I. Abrams. 1983. “American Labor Arbitration: The Maturing Years.” University of Florida Law Review, Vol. 35, no. 4 (Fall), pp. 557–632.

Sharpe, Calvin William. 2000. “Judicial Review of Labor Arbitration Awards: A View From the Bench.” Proceedings of the Fifty-Second Annual Meeting of the National Academy of Arbitrators. Washington, DC: Bureau of National Affairs, pp. 126–51.

Stone, Katherine V.W. 1999. “Employment Arbitration Under the Federal Arbitration Act.” In Adrienne E. Eaton and Jeffery H. Keefe, eds., Employment Dispute Resolution and Worker Rights in the Changing Workplace. Champaign, IL: Industrial Relations Research Association, pp. '–65.

United Paperworkers International Union v. Misco, 484 U.S. 29 (1987).

United Steelworkers of America v. Enterprise Wheel & Car Corp., 363 U.S. 593 (1960).

U.S. Government Accounting Office. 1995. Employment Discrimination: Most Private-Sector Employers Use Alternative Dispute Resolution. Report GAO/HEHS-95–150, July.

W.R. Grace & Co. v. Rubber Workers Local 759, 461 U.S. 757 (1983).

Zack, Arnold M. 1999. “Agreements to Arbitrate and the Waiver of Rights Under Employment Law.” In Adrienne E. Eaton and Jeffery H. Keefe, eds., Employment Dispute Resolution and Worker Rights in the Changing Workplace. Champaign, IL: Industrial Relations Research Association, pp. 67–94.