This Article theorizes that college football players participate in an invisible labor market, where the NCAA is a monopsony purchaser of their labor and strictly allocates these market inputs to regulate competitive balance between schools. This labor market theory is supported by analyzing the U.S. Tax Code, Fair Labor Standards Act, and worker’s compensation laws. These laws suggest that NCAA football programs are employers because they control so much of a player’s activities, and derive a direct financial benefit from this labor.

The Article proposes a unique and limited form of collective bargaining for college football. This conception accounts for the fact that NCAA football differs from the NFL model because college players are full-time students who must adhere to a code of amateurism.

Thus, the proposed model of collective bargaining does not involve wage negotiations or strikes. This analysis shows, however, that players have interests apart from wages. These include: (a) scholarship shortfalls (the difference between their grant-in-aid and true cost for attending college), (b) extended or improved educational benefits, (c) complete medical and hospital insurance for football related injuries, (d) long term disability insurance for injuries such as brain trauma, (e) transfer and eligibility rights not inconsistent with NCAA rules, and (f) a grievance process to challenge abusive treatment by coaches and administrators. For negotiations between players and schools that deadlock on these subjects, the proposal draws from public sector labor laws that bar strikes but allow final offer interest arbitration.

This proposal also draws from industrial relations research on the “union substitution” effect, which shows that when employers face a credible threat of unionization they provide more voice and benefits to employees. This Article’s use of the union substitution theory is an acknowledgement that collective bargaining for college football is an improbable idea. However, if the concept is only proposed as state or federal legislation, this would be enough to stimulate a robust union substitution effect— that is, the NCAA would likely be induced to provide players a voice in their affairs, and be more responsive to their concerns. This view is supported with historical and current union-substitution examples.

The status quo is ripe for change. The NCAA is a monopoly that generates new revenue but reforms itself slowly. Without a credible threat of unionization, the NCAA has no reason to confront the fact that it is professionalizing college football. College football exploits an invisible labor market, but traditional collective bargaining is impractical for players. An invisible union, derived from the union substitution effect, is a plausible way to address their interests.

The dual engagement strategy has created jurisdictional and choice-of-law conflicts for federal judges. District courts often ruled that league restrictions on free agency are anti-competitive business practices. Seeing no conflict between antitrust and labor law, they issued injunctions or found antitrust violations.

But the data in this study show a different side to the jurisdiction and choice-of-law story. After passage of the Sherman Act, businesses used this law ironically to challenge unions. Congress concluded that federal judges subverted antitrust law by enjoining union activities in labor disputes; and consequently, enacted a “labor exemption” to protect workers from this potent economic regulation. But the Clayton Act stated the labor exemption vaguely, referring to labor organizations but not their activities. When courts continued to issue injunctions in labor disputes, Congress responded again by stripping their jurisdiction in the Norris-LaGuardia Act.

This history came to life in my sample. Some courts perceived an intolerable conflict between antitrust and labor law. Thus, they denied jurisdiction to player complaints, or ruled that league-imposed labor restraints are immune under the antitrust “labor exemption.” This background provides essential context for the main findings in my statistical analysis:

(1) When leagues pleaded the Norris-LaGuardia Act as a jurisdictional defense to an antitrust injunction, district and appellate courts differed significantly in their rulings. Trial judges rejected Norris-LaGuardia arguments in 5.0% of cases in the district court sample, and accepted it in 1.7%. But appellate courts reversed this pattern by applying the Norris-LaGuardia defense in 13.0% of their decisions—meaning that these judges ruled that lower courts lacked jurisdiction.

(2) Ruling on the antitrust labor exemption, district courts rejected this league defense in 15.0% of cases, and applied it in 8.3%. Appellate courts treated the issue differently, applying the exemption in 26.1% of their cases and rejecting it only once (4.3%). Thus, appellate courts were more likely to immunize the disputed restriction from antitrust enforcement. In other words, appellate courts chose to apply the NLRA over the Sherman Act more often than district courts.

(3) District courts usually did not rule on a motion for an injunction (73.3% of the cases). But when they ruled on this motion, they were more likely to grant (20.0%) or affirm (1.7%) an injunction than deny this order (5.0%). Appellate courts behaved differently. More than half their rulings did not involve an injunction (60.9%). However, when they ruled on an injunction, they stayed or vacated 8 injunctions (34.8% of their cases), and affirmed only 1 order (4.3%).

In sum, district and appellate judges behaved differently in these hybrid labor law-antitrust cases. Some cases involved the highly technical antitrust labor exemption. In those cases, district judges committed “technical fouls” by failing to apply this exemption as required by federal labor law. In injunction cases, where district courts often granted the motion but appellate courts routinely overruled them, the former were guilty of “false starts.” On the one hand, the data suggest that appellate courts have cleaned up these district court “fouls,” and therefore, no legislative remedy is indicated. But the cases also imply that by the time appellate courts corrected lower courts, a judge’s “false start” or “technical foul” altered the balance of bargaining power, enabling players to gain an advantage that they could not achieve through collective bargaining. Regardless of whether federal judicial power should be legislatively curbed in sports labor disputes, or left to appellate correction, my study shows that district courts often misuse their power in these antitrust cases.

My study draws from legal and industrial relations theories to explain how labor agreements in pro sports are settled by collective bargaining or antitrust litigation. First, when courts do not define the antitrust-labor law boundary so that labor disputes are exempt from their jurisdiction, they open an alternative path to bargaining these agreements. Second, when courts entertain antitrust lawsuits, they raise the odds that economic weapons under the NLRA will not be used because of judicial inclination to protect players from irreparable harm and injury resulting from league-imposed labor market restrictions. Third, as this behavior becomes a pattern, collective bargaining is disrupted by faulty information as players, unions, and leagues guesstimate the odds that their differences will be settled at a collective bargaining table or in a court supervised negotiation. Fourth, as players negotiate better agreements in court compared to the bargaining table, they become addicted to this settlement process.

To apply these theories, I use data from 82 federal antitrust court opinions from 1965-2011. Individual players are the most common antitrust plaintiff (65.5%), compared to player unions (8.6%). This means the dispute resolution processes of collective bargaining are supplanted by litigation in federal courts. And except for baseball players, pro athletes often lose labor disputes when economic weapons are used. Their dismal bargaining experience substantially improves, however, by suing under the Sherman Act. In court, players win 43.9% of the rulings, compared to 46.3% for the leagues. These rulings—for example, an injunction that ends a league’s restrictions on free agency— can have dramatic consequences for antitrust settlements that are later codified in a collective bargaining agreement. Textual analysis of cases supports this conclusion.

Applying the “narcotic effect” theory from industrial relations, I conclude that antitrust litigation addicts players in football and basketball to the adjudicatory procedures of the Sherman Act— thereby replacing collective bargaining. This is undesirable because Congress intended, under the NLRA, to leave labor and management free from government interference as they adjust their differences. In contrast, baseball’s total exemption from antitrust law, combined with its high frequency of work stoppages, shows what happens when the opiate of antitrust litigation is not available to players: In time, labor and management establish an informed bargaining protocol, and work through their issues by making difficult concessions on their own. As long as courts entertain these sports lawsuits under the Sherman Act, collective bargaining will be subverted.

Hall Street Associates v. Mattel, Inc., 552 U.S. 576 (2008), failed to answer this question. The parties asked a court to review their award for errors of law. This standard is not in the Federal Arbitration Act (FAA). Hall Street ruled that courts cannot review awards beyond the FAA’s express terms. The parties’ standard prompted Hall Street to ask whether courts may apply “manifest disregard of the law,” even though it is not in the FAA. Inscrutably, Hall Street answered: “Maybe the term ‘manifest disregard’ was meant to name a new ground for review, but maybe it merely referred to the [FAA’s] § 10 grounds collectively, rather than adding to them.”

My research analyzes “manifest disregard” by using historical and empirical methods. Common law courts vacated awards for “fraud,” “corruption,” “partiality,” or if arbitrators “exceeded powers.” The FAA enumerates these as grounds to vacate awards. In the same sequence with these terms, nineteenth century courts vacated awards for “manifest mistake” or “palpable mistake” of the law. I contend that Congress inadvertently omitted “manifest disregard” from the FAA. To answer Hall Street’s question: “Manifest disregard of the law” is part of the FAA.

This sets the stage for my empirical question: Has Hall Street led courts to confirm more arbitrator rulings, thus promoting finality? The answer is yes. In 46.4% of federal cases and 21.8% of state cases, parties argued that an award manifestly disregarded the law. Still, state appellate courts confirmed more employment awards after Hall Street was decided on March 3, 2008— 88.9% (16/18), compared to 70.9% (73/103 awards) from 1975 until Hall Street. Federal district courts confirmed 93.7% of awards (164/175) before Hall Street, and 90.9% (30/33) after. Federal appeals courts confirmed awards at a high rate before and after Hall Street (87.8% and 85.7%).

Unfortunately, Hall Street’s muddled analysis has split federal circuits. The Fifth and Eleventh Circuits ruled that Hall Street ended “manifest disregard,” but the Second, Sixth, and Ninth Circuits still treat it as part of the FAA. The First, Third, Fourth, and Tenth Circuits have not ruled on the standard. In addition, state courts have differed in their reactions to Hall Street.

This fractured approach implies that the Supreme Court may reconsider its vague treatment of “manifest disregard.” The Court should affirm this standard. My findings show that review for “manifest disregard” does not erode finality. It translates to nanoscale limits on judicial review of awards. As one court put it: “There is . . . a way to understand ‘manifest disregard of the law’ that preserves the established relation between court and arbitrator. . . . It is this: an arbitrator may not direct the parties to violate the law.” Judges have applied this concept for two centuries to ensure that private tribunals conform to the laws. This rationale is particularly relevant because so much arbitration has changed from a voluntary to mandatory process. Judicial review must be allowed to correct an arbitrator’s intentional flouting of the law. If “manifest disregard” is eliminated, arbitral finality will rise above the crowning principle of the American constitutional system: “No man in this country is so high that he is above the law.” (U.S. v. Lee, 106 U.S. 196, 220 (1882)).

Why are the rulings of Article III judges scrutinized more than the awards of citizen-arbitrators? What does this mean when companies can avoid Article III court rulings by requiring employees to arbitrate their claims? Judicial review of awards based on statutory claims is inadequate, and undermines the constitutional role of federal courts.

I explore these empirical findings from a historical perspective. English kings and merchants helped to fashion modern arbitration. Nearly 700 years ago, small merchants traded goods at fairs that operated under a royal franchise. Arbitrators improved the efficiency of these markets by adjudicating transactional disputes. This role was codified by the Statute of the Staple of 1353, where the king delegated his sovereign power to ensure the success of the fair.

I point to two prominent junctures— in 1698, and again in 1925— when lawmakers in England and the U.S. believed that court litigation hampered commerce. They enacted similar statutes to authorize courts to confirm disputed awards, unless these private rulings resulted from corruption or misconduct. This deference grew out of practical considerations. The parties had chosen the arbitrator, agreed to the private process, and bound themselves to an industry norm.

Courts deferred so heavily to awards because William III wanted these merchant tribunals to be autonomous. His law, the 1698 Arbitration Act, did not allow courts to vacate awards for fact finding or legal errors. Great deference in its reviewing standards reflected the king’s infallibility.

My textual research shows that the FAA’s reviewing standards descended from William III. I suggest that our law crowns today’s statutory arbitrator with the king’s mantle of infallibility. But this deference is too extreme for awards that rule on statutory claims. In Gilmer v. Johnson/Interstate Lane Corp., the Supreme Court ignored the commercial history of arbitration when it broadly approved a theory of forum substitution. Gilmer said that arbitrators may decide statutory claims, even if one disputant objects to the forum and wishes, instead, to be heard by a court. The result is that the ruling of the arbitrator is subject to a narrower standard for review than an Article III judge’s order. Epitomizing this regal deference, a contemporary court said: “The arbiter was chosen to be the Judge. That Judge has spoken. There it ends.” In textual and empirical analysis, I show that statutory arbitrations enjoy a presumption of royal infallibility. I conclude with two solutions for aligning the review of rulings by statutory arbitrators and Article III judges.

The lopsided results suggest a moral hazard. When courts vacate many awards that rule for employees, the individual must either return to a lengthy and costly “do over” arbitration— or worse, be stuck with a useless award, and no other recourse because Gilmer v. Interstate/Johnson Lane Corp. bars employees from suing.

Throughout this Article, the reader will sense snowballing futility for employees. The problem is that the number of award reviewing standards is growing, due to new state laws and creeping expansion of common law standards. This causes judges to deviate from the FAA’s extremely deferential principles. As a result, court review is becoming an insurance program that protects employers from costly awards. This poses a moral hazard, as judges reduce incentives for employers to be responsible for their actions. Because Gilmer permanently bars employees from suing, employers are double-insulated from being held to account for their unlawful conduct.

This trend undermines the purpose of the Federal Arbitration Act (FAA). The law meant to end court interference with awards. Gilmer reinforced this view by stating a theory of forum substitution— the idea that by “agreeing to arbitrate a statutory claim, a party . . . submits to their resolution in an arbitral, rather than a judicial, forum.” After Gilmer, employers adopted mandatory arbitration to control litigation costs and liability. But the data raise new doubts about forum substitution. Courts interfere with arbitration more than the FAA or Gilmer envisioned.

I propose two solutions. Judicial review of awards in all courts should be limited to the four explicit FAA standards. This would free arbitration from court interference, just as Congress intended. Indeed, the gist of my legislative proposal was supported in the current term of the Supreme Court, when Preston v. Ferrar said: “The Act (FAA), which rests on Congress’ authority under the Commerce Clause, supplies not simply a procedural framework applicable in federal courts; it also calls for the application, in state as well as federal courts, of federal substantive law regarding arbitration.” Second, arbitration losers who incur liability should be required to post judgment bonds if they challenge an award. This idea borrows from civil procedure codes, and is therefore consistent with the goal of forum substitution. The current practice erodes award finality by allowing employers to make cost free appeals. Arbitration offers attractive benefits: reduced cost, simplicity, and easy accessibility to disputants. But these benefits will not be achieved until the growing vacatur problem— and its attendant quality of relieving employers of liability for unlawful behavior— is addressed.

This change is affecting how court reviews arbitrator awards. I collected data in 426 federal and state court rulings in employment disputes from June 1975 through April 2007. Federal district judges confirmed 92.7% of arbitrator awards, compared to 78.8% for state courts. This statistically significant difference was also observed for appellate courts, where the confirmation rate in federal courts was 87.7%, contrasted to 71.4% for state courts.

The anomalous structure of the FAA explains why federal courts are more deferential than state courts. The statute allows litigants to choose federal or state court to challenge an award. In federal court, the standard of review is among the narrowest in the law. Oddly, however, the FAA also allows states to define their own standards. Fewer states now replicate the FAA, as a growing number adopt more intrusive standards.

The Revised Uniform Arbitration Act is exacerbating this situation. Passed in 2000, RUAA has been enacted by 12 states— and the number is likely to grow. RUAA aims to curb recent abuses in arbitrations by legislating fairness into arbitration. However, by strictly regulating arbitrator disclosures, and limiting arbitrator powers to order attorney’s fees and other relief, RUAA creates new and more conditions that erode the finality of arbitration decisions.

Fairness principles in RUAA are laudable, but RUAA drafters ignored an ancient maxim of fairness— the Magna Carta’s injunction that justice delayed is justice denied. They never imagined that so many arbitrations at the state level would leave the underlying arbitration agreements, which secure a promise for a final and binding award, in tatters. The apparent winner in arbitration is therefore deprived of justice.

These recent developments trace to 1925, when Congress transferred oversight of arbitration from courts to legislatures. Today, this means that arbitration is more exposed to lobbying efforts by special interests who want to tailor ADR processes— under the aegis of the FAA— to suit their preferences. My data show that the U.S. has two different justice systems for enforcing arbitration awards. The federal domain is minimalist, rooted in centuries of common law experience, and also devoid of political maneuvering. Many states, however, naively promise more fairness in arbitration while subjecting this private process to special deals that are worked out in obscure legislative committees.

State expansion of reviewing standards poses three significant problems: (1) Foremost, this trend is fragmenting the national policy that favors arbitration by allowing a plurality of judicial review standards.

When Congress and the Supreme Court beat the drum to proclaim that the FAA provides a national arbitration policy, they are wrong. (2) The results imply that RUAA will stimulate forum shopping. Award winners should run to federal court to confirm their awards, while challengers should race to state court to improve their odds of vacatur. In time, the judicial inconsistency that I document will undermine cost, efficiency, and time saving advantages of arbitration over court adjudications. (3) A moral hazard is created when losers at arbitration are tempted to renege on their contractual promise to submit to binding arbitration in order to pursue “do-over” adjudication. As more courts vacate awards, more disputes remain unresolved; and as the vacatur rate increases in state courts, this will lead to more award challenges. Congested court dockets will become a bigger problem if parties cannot rely on contractual promises for final and binding arbitration. Losing confidence in award finality, more parties will bypass arbitration and resort to filing more lawsuits. The Congress that passed the FAA sought a quick, efficient, and low-cost alternative to court adjudication. These lawmakers would be alarmed by empirical trends that depict the unbinding of arbitration.

Until now, researchers have examined the fairness and legality of Gilmer agreements and other aspects of employment arbitration. Our timing is significant because employment arbitration has matured beyond the initial phase of pre-arbitration challenges to this forum. By now, a critical mass of individuals and their employers have been to arbitrations and appealed arbitrator rulings to courts.

But little is known about why parties appeal adverse arbitration rulings. There is no empirical information about the legal arguments that parties use to challenge awards. No data exist to explain how federal and state courts rule on these challenges.

We extracted data from 212 federal and 124 state court decisions from 1975-2006. These 336 cases show that the grounds for challenging employment awards are far more numerous than those for labor awards that are reviewed under the Supreme Court’s Trilogy standards. Too many courts and legislatures have presented arbitration losers with an abundance of choices to sue on the award. These grounds include federal bases such as the four statutory grounds under the Federal Arbitration Act (FAA), common law standards, all the Trilogy arguments that are used to challenge labor arbitration awards; and numerous state statutes and common law grounds.

Federal district courts confirmed awards in 93.4% of the cases, compared to 88.0% in appellate cases. State courts were less deferential, confirming 86.4% of awards in initial reviews and 74.1% at the appellate level. While state courts confirmed a high percentage of awards, these rates were quite variable. Texas courts confirmed 100% of awards, followed by Louisiana, Florida, and Arkansas. Two Midwestern states, Michigan and Ohio, had much lower confirmation rates (respectively, 58.3% and 64.2%).

We also observed a recent spurt of award-review cases— exemplified by the finding that 62.5% of our federal district courts decisions occurred since 2000. This means that courts will likely face a growing docket of post-arbitration appeals as parties seek to re-litigate the claims that were privately adjudicated. This upsurge also suggests that employment arbitration serves too many masters— an uncoordinated array of legislatures and courts that regulate this process— despite the fact that the FAA appears to legislate uniform standards.

Comparing these findings with our companion studies on labor arbitration, we believe that unions and employers limit their award appeals to maintain a stable relationship. But fewer parties in employment arbitration seem committed to the norm of arbitral finality. We see more cases of a winner-take-all mentality in post-award appeals, suggesting that disputants are destined to be “happily never after” their arbitrations.

How far can the U.S. go in forcing reluctant civilians to perform essential jobs during a national emergency? I explore solutions to this question by hypothesizing a large release of radiation— whether by terror attack, or catastrophic accident, or major earthquake— in a vital Pacific port. These ports have a history of work stoppages that disrupt the nation’s economy. I examine federal government responses if dock workers refused assignments until conditions were safe: (1) The President could declare a national emergency labor dispute under the Taft-Hartley Act, and seek an 80-day back-to-work injunction. (2) Congress could re-enact Section 8 of the War Labor Disputes Act, making it unlawful for dock workers to discontinue production for 30 days and subjecting violators to coercive damages. (3) The president could issue strong executive orders, backed by imprisonment, that regulate employment in ports.

At the heart of my analysis, I ask: Would any of these responses violate the Thirteenth Amendment ban on involuntary servitude? Congress and the judiciary have broadened this law, and its enforcement counterpart in 18 U.S.C. § 1584, beyond the abolition of African slave-holding. The Supreme Court in Kozminski defined involuntary servitude as forcing a person to work by physical or legal coercion.

But the Supreme Court created 13th Amendment exceptions for transportation work. Robertson upholds a law that bars merchant seamen from quitting work, and imprisons deserters. Butler permits states to conscript citizens to work on highways, on pain of imprisonment. Dock work is similar because ports integrate ships and trucks in a transportation hub. Courts now apply these precedents to new compulsory activities, such as mandatory public service for graduation. Moreover, Kozminski reaffirmed Robertson and Butler as precedents.

Thus, the Constitution would be unlikely to shield dock workers from involuntary labor. This has troubling implications for employees who have recently worked in national emergencies, and may do so again. Employees who work to alleviate avian flu or other catastrophic health threats are also at risk for compulsory labor that exposes them to extraordinary hazards.

I conclude with a legislative proposal to strengthen individual rights. As my research shows, courts that are presented with national emergency disputes rarely side with the individual who stands in the way of the public’s welfare. Without a more balanced labor policy to address emerging crises, the nation may realize belatedly “that when we allow fundamental freedoms to be sacrificed in the name of real or perceived emergency, we invariably come to regret it.”

These results clearly suggest that the Supreme Court’s rebuke of lower courts in Eastern Associated Coal Corp. (2000) and Garvey (2001) have changed how judges review arbitrator rulings. But this uniformly positive development masks observable differences in the federal circuits where these cases were decided.

• The Sixth Circuit routinely applies a four part essence test. This standard seems more intrusive than the simple essence test in Enterprise Wheel. The four part essence test has become so controversial among these judges that they decided en banc in May 2006 to reconsider its use. Our data shed important new light on this development: The four part essence test yields outcomes that are similar to other Enterprise Wheel tests, but it stimulates an excessive amount of federal lawsuits.

• On the opposite end of the award deference spectrum, courts in the Second Circuit enforce more than 90% of challenged awards. This trend is so clear that virtually no one there appeals a district court’s enforcement ruling. Courts in the Seventh Circuit have a similar confirmation rate, arrived at by applying Rule 11 sanctions to award challenges that are judged to be meritless.

• Data show that judges in the Fifth Circuit are in an isolationist camp. At the district and appellate levels in this study, these courts confirmed awards in only 44.4% of the cases. This is similar to findings in our earlier studies. A correction is needed for the Fifth Circuit’s long deviation from the deferential posture that the Supreme Court has commanded. These data might enable Fifth Circuit judges to see their problem and correct it by exercising more restraint.

• We found that 16 appellate decisions confirmed awards that were vacated by district courts, while 5 awards were vacated by appellate courts after these arbitrator rulings were confirmed by lower courts. This 3:1 ratio favoring award confirmation means that new precedents are reinforcing Enterprise Wheel messages of deference to district judges.

Overall, the empirical results are healthy indicators for the national policy that favors arbitration. The Supreme Court’s on-going investment in promoting judicial deference to awards is paying dividends not only for the institution of labor arbitration, but by implication, for newer ADR applications in individual employment, commercial transactions, environmental disputes and others.