I. The Special Nature of CBAs
As the Court recognized in Warrior & Gulf, “the special nature of CBAs derives from their function in the system of labor relations envisioned in the National Labor Relations Act (NLRA). Under the NLRA, when an employer recognizes a union, either voluntarily or pursuant to NLRB order or certification, the employer’s ability to act unilaterally is at an end. It now has a duty to negotiate in good faith with the union before taking action with respect to wages, hours, and terms and conditions of employment. In theory, such negotiations could occur on an ad hoc basis prior to every decision concerning any of these subjects, with the parties’ bargaining power derived from the strength of the economic weapons they have and may threaten to deploy if agreement is not reached.
Such an ad hoc system is counterproductive for both parties. Unions lack the resources to negotiate every time an employer approves or denies a vacation request or personal day or assigns overtime, and threatening to strike over many of those issues amounts to using an elephant gun to swat a fly. Employers cannot maintain continuous and efficient production if they have to negotiate and run the risk of a strike each time they make such decisions. Consequently, as the Court recognized, parties enter into CBAs as “an agreed upon rule of law,” instead of “leaving each and every matter subject to a temporary resolution dependent solely upon the[ir] relative strength, at any given moment.” The CBA is thus “an effort to erect a system of industrial self-government.” “It is more than a contract; it is a generalized code to govern a myriad of cases which the draftsmen cannot wholly anticipate.”
Even so, it was never in dispute that a CBA is a contract. But as Archibald Cox pointed out more than sixty years ago, the world of contract is populated by a variety of contracts, of special contracts—leases, bills of sale, long-term requirements contracts, franchise contracts, and more—each requiring its own manner of interpretation. The challenge, he said, is to synthesize our notions concerning the interpretation of collective agreements with those and only those contract principles that assist in the effort.
That sentiment was echoed and expanded on by Clyde Summers. But Summers struck a cautionary note.
[T]he legal rules and principles developed to govern contractual relations generally should be useful in defining the rights and duties created by collective agreements. . . . [T]he stubborn fact [remains] that these rules and principles will often make mischief if imported into the collective bargaining context. There is a core of truth in the assertion that collective agreements are not “ordinary contracts,” and therefore that “ordinary contract law” is not applicable to them.
As Congress recognized in enacting the Labor Management Relations Act of 1947,
[t]he execution of an agreement does not by itself promote industrial peace. The chief advantage which an employer can reasonably expect from a collective labor agreement is assurance of uninterrupted operation during the term of the agreement. Without some effective method of assuring freedom from economic warfare for the term of the agreement, there is little reason why an employer would desire to sign such a contract.
A typical issue that can arise during the term of a CBA illustrates the point. Assume that the CBA designates specific holidays during which employees will not be required to work but will receive eight hours’ straight-time pay for the day. The CBA further provides that to receive holiday pay, employees must work their last regularly scheduled day before the holiday and their first regularly scheduled day after the holiday. An employee is injured on the job on the Wednesday of the week leading up to a holiday weekend and is taken to an urgent care center with whom the employer has a contract. The doctor instructs the employee not to work the next two days and releases the employee to return to work the following Tuesday, i.e. the day after the holiday. The employer denies holiday pay because the employee did not work Friday, the last regularly scheduled day preceding the holiday. The union demands that the employer pay the holiday pay.
At issue is eight hours of straight-time pay, an amount over which no reasonable party would file a lawsuit. But the issue is one in which the entire workforce is interested, as anyone could have received and in the future could be subject to the same treatment. It is easy to imagine the union mobilizing a job action to compel the employer to pay the holiday pay.
In the words of Warrior & Gulf, the requirement to work the last regularly scheduled work day before the holiday is a “more or less specific standard[] which require[s] reason and judgment in [its] application.” In other words, resolving this dispute requires further collective bargaining. The vehicle for the continued negotiations is the CBA’s grievance procedure. In the grievance procedure, the union’s claim is negotiated at successively higher levels within the union and the employer. If agreement is not reached, the union may submit the claim for resolution by an arbitrator mutually selected by the parties. The Court described the process in Warrior & Gulf:
[T]he grievance machinery under a collective bargaining agreement is at the very heart of the system of industrial self-government. Arbitration is the means of solving the unforeseeable by molding a system of private law for all the problems which may arise and to provide for their solution in a way which will generally accord with the variant needs and desires of the parties. The processing of disputes through the grievance machinery is actually a vehicle by which meaning and content are given to the collective bargaining agreement. . . . The grievance procedure is, in other words, a part of the continuous collective bargaining process. It, rather than a strike, is the terminal point of a disagreement.
In exchange for agreeing to abide by the result of the grievance and arbitration process, the employer secures a commitment from the union that it will not strike during the term of the CBA. Indeed, the Supreme Court has recognized that, in the typical CBA, the arbitration procedure is the quid pro quo for the union’s agreement not to strike during the contract’s term. David Feller, whose briefs as a lawyer successfully arguing the Trilogy cases were relied upon by the Court and who later became a renowned arbitrator and labor law professor, has theorized that the true essence of a CBA consists of the grievance-arbitration procedure and the no strike clause.
Arbitration enables the parties to avoid workplace strife in a second way. The availability of a grievance-arbitration procedure enables the parties to reach agreement on the terms of the CBA even though they do not agree precisely on what those terms mean. For example, a common CBA provision requires the employer to post vacancies and sets forth a period of time during which employees covered by the CBA may bid to fill the vacancies. The provision further states that where the qualifications of two competing bidders are relatively equal, the job shall go to the more senior of the two. The negotiations leading to such a provision probably began with the union demanding that vacancies be filled strictly by seniority and the employer demanding that they be filled based on qualifications as judged by management. The compromise probably did not resolve the parties’ differences over the interaction between qualifications and seniority. Each party realized that its meaning of “relatively equal” differed greatly from the other’s. However, the availability of the grievance-arbitration procedure allowed the parties to agree on the language, knowing that they disagreed on its meaning, and postpone refinement of the meaning to case-by-case negotiation through the grievance procedure with the understanding that if, in any given case, they were unable to agree, the language would mean in that case whatever the arbitrator would say it meant.
Arbitration is thus an integral part of the parties’ process of workplace self-governance and, therefore, is due considerable deference from courts. In Enterprise Wheel, the third of the cases in the Steelworkers Trilogy, the Court opined:
When an arbitrator is commissioned to interpret and apply the collective bargaining agreement, he is to bring his informed judgment to bear in order to reach a fair solution of a problem. . . . Nevertheless, an arbitrator is confined to interpretation and application of the collective bargaining agreement; he does not sit to dispense his own brand of industrial justice. He may of course look for guidance from many sources, yet his award is legitimate only so long as it draws its essence from the collective bargaining agreement. When the arbitrator’s words manifest an infidelity to this obligation, courts have no choice but to refuse enforcement of the award.
More than a quarter century later, in United Paperworkers International Union v. Misco, Inc., the Court elaborated:
The arbitrator may not ignore the plain language of the contract; but the parties having authorized the arbitrator to give meaning to the language of the agreement, a court should not reject an award on the ground that the arbitrator misread the contract. . . . [I]t must be remembered that grievance and arbitration procedures are part and parcel of the ongoing process of collective bargaining. It is through these processes that the supplementary rules of the plant are established. As the Court has said, the arbitrator’s award settling a dispute with respect to the interpretation or application of a labor agreement must draw its essence from the contract and cannot simply reflect the arbitrator’s own notions of industrial justice. But as long as the arbitrator is even arguably construing or applying the contract and acting within the scope of his authority, that a court is convinced he committed serious error does not suffice to overturn his decision.
In Eastern Associated Coal Co. v. United Mine Workers, the Court reiterated, “As long as an arbitrator is even arguably construing or applying the contract and acting within the scope of his authority,” the award must be enforced. A year later, the Court held that “[i]mprovident, even silly, factfinding” by an arbitrator does not provide a basis for refusing to enforce the award.
What constrains the arbitrator is not a court or other external legal authority. It is the expectations of the parties. In his seminal article quoted extensively by the Court in the Trilogy, former Yale Law School Dean Harry Shulman observed:
A proper conception of the arbitrator’s function is basic. He is not a public tribunal imposed upon the parties by superior authority which the parties are obliged to accept. He has no general charter to administer justice for a community which transcends the parties. He is rather part of a system of self-government created by and confined to the parties. He serves their pleasure only, to administer the rule of law established by their collective agreement. They are entitled to demand that, at least on balance, his performance be satisfactory to them, and they can readily dispense with him if it is not.
Consequently, according to the Court in Warrior & Gulf, an arbitrator is not as confined as a judge when interpreting a CBA:
The labor arbitrator’s source of law is not confined to the express provisions of the contract, as the industrial common law—the practices of the industry and the shop—is equally a part of the collective bargaining agreement although not expressed in it. The labor arbitrator is usually chosen because of the parties’ confidence in his knowledge of the common law of the shop and their trust in his personal judgment to bring to bear considerations which are not expressed in the contract as criteria for judgment. The parties expect that his judgment of a particular grievance will reflect not only what the contract says but, insofar as the collective bargaining agreement permits, such factors as the effect upon productivity of a particular result, its consequence to the morale of the shop, his judgment whether tensions will be heightened or diminished. For the parties’ objective in using the arbitration process is primarily to further their common goal of uninterrupted production under the agreement, to make the agreement serve their specialized needs. The ablest judge cannot be expected to bring the same experience and competence to bear upon the determination of a grievance, because he cannot be similarly informed.
Thus, since 1957, the Supreme Court consistently regarded CBAs as significantly different from ordinary commercial contracts and arbitration under CBAs as significantly different from ordinary litigation: commercial contracts commonly reflect terms geared to discrete transactions and in which arbitration substitutes for a judge in applying normal principles of contract law; collective agreements are instruments of industrial self-government, a process of which arbitration is an integral part as the arbitrator substitutes not for a court but for a strike and in which the arbitrator may be guided by sources extrinsic to the contract. Led by Justice Thomas, the view began eroding on April Fools Day in 2009. The next part chronicles this erosion.
II. Justice Thomas’s Mischief
14 Penn Plaza LLC v. Pyett holds that an employee may be compelled to grieve and arbitrate an Age Discrimination in Employment Act (ADEA) claim where the CBA clearly and unmistakably waives the employee’s right to sue. Although Justice Thomas quoted the statement in Lincoln Mills that grievance arbitration is the quid pro quo for the no-strike clause, he ignored the more than six decades of Court jurisprudence that Lincoln Mills ushered in. He failed to give even lip service to the grievance arbitration process as a continuation of the parties’ collective bargaining process and as a vehicle for workplace self-governance. Because of that, he refused to come to grips with the stark fact that what was involved was not the situation of a single, non-unionized employee governed by an employer’s contract of adhesion requiring her to arbitrate her civil rights. He declined to face the fact that the commitment to arbitrate civil rights claims was mapped on to a contractual grievance arbitration procedure that allocated exclusive power to the union to decide what claims to bring to arbitration; as the National Academy of Arbitrators argued as amicus curiae, the union’s adherence to its duty of fair representation in declining to pursue a grievance over a civil right, which is fitting when the arbitration process is viewed as an extension of collective bargaining, is not a fitting substitute for a neutral judgment on the merits of the civil rights issue.
Instead, Justice Thomas regarded grievance arbitration, like commercial arbitration, as a substitute for litigation, divorced from the union’s role in a procedure that substitutes for strikes and other job actions. He wrote, “Parties generally favor arbitration precisely because of the economics of dispute resolution. As in any contractual negotiation, a union may agree to the inclusion of an arbitration provision in a collective-bargaining agreement in return for other concessions from the employer . . . .” Instead of viewing the arbitration provision as a concession given by the employer for the no-strike clause, Justice Thomas viewed it as a concession given by the union in exchange for unspecified concessions from the employer. He expressly declared that the arbitration provision of a CBA is no different from the arbitration provision of any other contract.
Justice Thomas also authored the opinion in M & G Polymers USA, LLC v. Tackett, a suit by retirees to enforce their claimed right to lifetime medical benefits guaranteed by a CBA. The collective agreement’s medical benefits provision made no mention of duration; but the larger CBA did, and it had expired. Arbitration was not involved. The district court dismissed the plaintiffs’ complaint for inter alia failure to state a claim, but the Sixth Circuit reversed. It relied on circuit precedent setting out factors to be considered in addition to the “explicit language” of the contract, notably the context of collectively negotiated medical benefits, which, the court said, are typically understood to be a form of delayed compensation for past services and which are unlikely to be left subject to the vagaries of future negotiation. On remand, following a bench trial, the district court entered judgment for the plaintiffs, and the Sixth Circuit affirmed.
The Supreme Court vacated the decision, rejecting the precedent on which the Sixth Circuit relied. Justice Thomas, writing for four of his brethren, held that the approach the Sixth Circuit took “violates ordinary contract principles by placing a thumb on the scale in favor of vested retiree benefits in all collective bargaining agreements. That rule has no basis in ordinary principles of contract law.” He then set out six grounds in support of the Court’s conclusion.
First is the governing precept that CBAs are to be interpreted “according to ordinary principles of contract law” unless they are inconsistent with federal labor policy, citing Lincoln Mills. Justice Thomas acknowledged that these “ordinary principles” allow extrinsic evidence to be looked to, if a contract term is ambiguous, citing the fourth edition of Williston on Contracts. Nor did the Court disagree that the benefits provision was ambiguous: it was silent regarding duration, something more than the text had to be looked to. But, that would only allow for resort to extrinsic evidence regarding the particular contract, not resort to a general presumption about collective bargaining.
Second, the context of “labor negotiations” was “too speculative and too far removed from the context of any particular collective agreement.” There was no basis for the inference of intent to have the benefits continue by virtue of their having been collectively bargained for, no warrant to believe it “unlikely that [retiree] benefits . . . would be left to the contingencies of future negotiations.” That unlikelihood was belied by the very fact that the parties had bargained about retiree benefits; in fact, they had “voluntarily” made benefits a subject of “mandatory bargaining.”
Third, equally unwarranted was the characterization of these benefits as deferred compensation. Under the Employee Retirement Income Security Act (ERISA), pension benefits (which vest as a matter of law) are understood to be a form of deferred compensation; but welfare benefits (which do not vest as a matter of law), including medical benefits, are not.
Fourth, the Circuit Court’s refusal to apply the collective agreement’s duration clause to the provision for retiree benefits conflicted with “the principle of contract law that the collective agreement is presumed to encompass the whole agreement of the parties,” again, citing Williston—that is, the “entire contract” rule.
Fifth, the Sixth Circuit failed to apply or even to consider the “traditional principle that courts should not construe ambiguous writing to create lifetime promises,” this time citing Corbin on Contracts. The Court found the Sixth Circuit inconsistent, presuming that retiree welfare benefits in CBAs were vested but opining that in nonunion settings the vesting of such benefits is not lightly to be inferred. The Sixth Circuit’s favoring of vesting only underscored the deviation from “ordinary principles of contract law.”
Finally, the Sixth Circuit failed to take account of the “traditional principle” announced in Litton Financial Printing that, ordinarily, the obligations of a CBA terminate with the agreement’s termination. This was not to deny that some provisions could survive termination; but it disallowed that an intention for a benefit to vest for life could be inferred from silence.
On closer examination, Justice Thomas’s reasoning becomes inconsistent with or contradictory to the established body of law, which dissonance passes unnoticed by the other members of the Court. This becomes clear once one works through each of the six points of Justice Thomas’s opinion. First, at the very outset, the Court stood extant law on its head. The Lincoln Mills Court held that section 301 conferred jurisdiction on federal courts to construe CBAs, but section 301 provided no guidance in terms of what substantive law governed the contract. The Lincoln Mills Court answered that question categorically:
We conclude that the substantive law to apply in suits under § 301(a) is federal law, which the courts must fashion from the policy of our national labor laws. The Labor Management Relations Act expressly furnishes some substantive law. It points out what the parties may or may not do in certain situations. Other problems will lie in the penumbra of express statutory mandates. Some will lack express statutory sanction but will be solved by looking at the policy of the legislation and fashioning a remedy that will effectuate that policy. The range of judicial inventiveness will be determined by the nature of the problem. . . . But state law, if compatible with the purpose of § 301, may be resorted to in order to find the rule that will best effectuate the federal policy.
Nothing is said here about “ordinary contract principles” being directive save only insofar as federal policy is not offended; in fact, “ordinary” or “traditional” contract law makes no appearance in Lincoln Mills. Rather, it is the other way around: federal policy directs the judiciary to manufacture—even to invent—a law of CBAs out of labor policy in which the courts are free to range wide for suitable guiding sources including contract law.
Nor is Williston of help. The treatise does say that a CBA “is to be treated like any other,” that it is “to be interpreted in the same manner as any other contracts, and traditional rules of contract law apply to their interpretation.” But then it provides:
There are situations where the courts, including the Supreme Court of the United States, have not hesitated to go beyond the bounds of normal interpretation to give collective bargaining agreements added force and effect, seemingly without regard to the intention of the parties. The pronouncements of the Supreme Court support the view that collective bargaining agreements are contracts of a rather special type.
Second, on the Sixth Circuit’s impermissible reliance on the context of collective bargaining, the Court itself, in American Manufacturing, rejected a principle of contract law that maintained that, if a disputed contractual provision was beyond dispute, there would be nothing to arbitrate. The Court dismissed this as a “preoccupation with ordinary contract law”: “In our role of developing a meaningful body of law to govern the interpretation and enforcement of collective bargaining agreements, we think special heed should be given to the context in which collective bargaining agreements are negotiated and the purpose which they are intended to serve.” That is just what the Sixth Circuit did. It saw reason in the collective bargaining context to be solicitous of the rights of retirees. The parties did bargain about retiree benefits, but their doing so did not make these benefits a mandatory bargaining subject. By the Court’s own hand, retirees are not statutory employees; their benefits are not statutorily mandatory bargaining subjects—whence the Court’s artful interjection of the qualifier “voluntarily,” which, simply put, only means that the parties bargained about them. But the “bargaining context” has deeper meaning upon which the Sixth Circuit drew. When retiree benefits are contracted-for collectively, were they not to vest, were they to expire with the collective agreement’s termination, the employer may refuse to discuss them further with the Union and proceed unilaterally to reduce or abrogate them, those so adversely affected having no say in the matter. Nor, inasmuch as the subject is statutorily non-mandatory, could the union strike to restore them. The Sixth Circuit thought this, the “collective bargaining context,” should weigh in the decision, just as American Manufacturing said.
Third, Justice Thomas was quite right: ERISA welfare benefits, including medical benefits, do not vest under that statute. What remains to be seen is the relevance of that fact. It might be relevant if the collective agreement incorporated ERISA, in express terms, for example. But it did not. As the contract did not direct the contract reader to that statute, the contract reader was free to find that the contract treated these benefits as a form of deferred compensation. Decades ago, long before ERISA, the NLRA was held to treat “a retirement and pension plan” as a wage, i.e., as deferred compensation. A collective agreement could treat medical benefits as such as well; at least, the reader who speaks the contract could say it did.
In other words, it would seem open to a court, exercising the scope accorded it by Lincoln Mills, to fashion a law of collective agreements and, taking the foregoing contextualizing factors into account, to put its “thumb on the scale” in reading CBAs in favor of retirees as a matter of national labor policy. Note that, in practical terms, the Sixth Circuit’s quasi-presumption only placed the burden of securing contractual clarity on the employer. This would seem to be well within the judicial function. For example, a union can eschew arbitration and reserve the right to strike over a grievance, but the Supreme Court held that the burden rests on the union to have the collective agreement say so in unmistakable terms. The Court shifted the burden onto the union drawing on its disfavoring industrial strife, i.e., on the Court’s understanding of national labor policy.
Fourth is the Court’s reliance on that general principle of contract law that contracts are presumed to encompass the entire agreement of the parties—the “entire contract” rule—which principle the Court automatically extends to CBAs. But this general principle of contract law cannot be reconciled with the Court’s own conception of CBAs set out in Warrior & Gulf. As discussed previously, the Court, citing Harry Shulman’s famous Holmes’s Lecture, observed that a CBA “is more than a contract; it is a generalized code to govern a myriad of cases which the draftsmen cannot wholly anticipate.” The Court then expanded on that by reference to another classic piece, this time by Archibald Cox:
There are too many people, too many problems, too many unforeseeable contingencies to make the words of the contract the exclusive source of rights and duties. One cannot reduce all the rules governing a community like an industrial plant to fifteen or even fifty pages. Within the sphere of collective bargaining, the institutional characteristics and the governmental nature of the collective-bargaining process demand a common law of the shop which implements and furnishes the context of the agreement.
Even an “entire contract” provision the parties have agreed to put in a CBA is commonly discounted by arbitrators when facing contractual gaps, conflicts, and even established past practices. In other words, by the Court’s antecedent account, the “entire contract” rule, an archetypical “ordinary principle of contract law,” is fundamentally discordant with CBAs.
Fifth is the Court’s reliance on “the traditional principle that courts should not construe ambiguous writings to create lifetime employment contracts.” This time Justice Thomas cited Arthur Corbin’s treatise “explaining” that contracts silent as to duration are operative only for “a reasonable time.” However, Corbin laid out no such “principle.” The language that Justice Thomas paraphrased—as a “traditional principle”—was simply an account of the holding in a single case about a sewage system. Corbin did address cases involving a promise “to pay a specified pension, or retirement allowance, or other compensation ‘for life.’” Of this, he observed that “the problem in such cases is usually one of sufficiency and consideration,” which is inapplicable to CBAs. In the event, the most serious disconnect is the fact that Corbin thought that that none of this applied to CBAs anyway: “ This treatise does not attempt the analysis and discussion of collective bargains. They cannot be treated with advantage separately from the general subject of Labor Relations and Labor Legislation.” The collective agreement is a contract, he opined, but, echoing Williston, “a contract of a very special kind.”
This leaves the Sixth Circuit’s failure to have considered the “traditional principle” stated in Litton Financial. How traditional that is requires a quick look at two cases preceding Litton Financial.
First in line is John Wiley & Sons v. Livingston. Wiley, a large, non-unionized company acquired Interscience, a smaller, unionized company. The collective bargaining agreement at Interscience, which Wiley did not assume, provided inter alia for severance pay. The union sought to secure severance pay and other benefits provided for in its CBA with Interscience from Wiley via the collective agreement’s arbitration clause. Thus two issues were presented: whether a company that was not a party to a collective agreement might be bound to arbitrate under it; and, if so, whether it was bound to arbitrate actions taken after the collective agreement had expired. The Court answered in the affirmative to both.
Regarding the obligation to arbitrate, the Court looked not to “ordinary” or “traditional” principles of contract law, which were, obviously, to the contrary, but to the “policy of a national labor law,” citing Lincoln Mills. While “the principles of law governing ordinary contracts would not bind to a contract an unconsenting” contracting party, the Court said, these do not control when collectively bargained terms are at stake. Regarding the determination of which if any of the contract’s terms survived termination, the Court saw “no reason why parties could not if they so chose agree to the accrual of rights during the term of an agreement and their realization after the agreement had expired.” Whether that was so and of which provisions was for the arbitrator to decide.
Next in line is Nolde Bros., Inc. v. Bakery Workers. Here, a union claimed severance pay for plant closing that occurred after the termination of the collective agreement. There was no dispute that, were the event to have arisen during the term of the agreement, the agreement’s later termination would not affect the viability of the claim. Further, there was no disagreement that the expiration of the contract had no effect on the obligation to arbitrate post-termination claims to rights that vested. Accordingly, consistent with Wiley, whether severance pay had vested was for the arbitrator to decide.
A decade later, came Litton Financial. It posed the question of the duty to arbitrate layoff decisions alleged to violate a collective agreement’s seniority provision, which lay-off decisions were made after the collective agreement’s termination. The Court turned to Nolde Bros.:
A postexpiration grievance can be said to arise under the contract only where it involves facts and occurrences that arose before expiration, where an action taken after expiration infringes a right that accrued or vested under the agreement, or where, under normal principles of contract interpretation, the disputed contractual right survives expiration of the remainder of the agreement.
Whence “normal principles of contract interpretation” enters the stage, for the first time. The Court decided that the seniority clause, requiring comparison of worker aptitude and ability, did not survive the collective agreement’s termination. Unlike the severance pay “issue in Nolde Bros., which was viewed as a form of deferred compensation, the layoff provision here cannot be so construed and cannot be said to create a right that vested or accrued during the term of the agreement or a contractual obligation that carries over after expiration.”
Four Justices dissented on the ground that, under Nolde Bros., the question of whether the seniority provision survived the contract’s termination was for an arbitrator, not the court. Justice Kennedy replied by reminding his brethren that, under Warrior & Gulf, arbitrability was for a court to decide, and, in doing so, he expressed a prudential concern that absent judicial supervision up-front, so to speak, post-termination arbitration claims would be otherwise “limitless”; this, without a nod to American Manufacturing’s acknowledging of no limit to the bringing of even frivolous claims into arbitration.
Neither of the dissenting opinions in Litton Financial addressed the majority’s novel use of “normal principles of contract interpretation” to qualify how a CBA is to be read. Even so, Litton is easily distinguished from Tackett, for, unlike Litton, the Court concluded that the collective agreement in Tackett could be read to create a vested right, whence a remand to the Sixth Circuit to apply ordinary principles of contract law. Nevertheless, Litton Financial is important here for its interjection of “normal principles of contract interpretation,” which provided the jurisprudential axis around which Tackett’s later reasoning turns.
To sum up, the Sixth Circuit read national labor policy governing CBAs that deal with retiree medical benefits to favor retirees. The court put the onus—its “thumb”—on employers who wish to have these contractual obligations terminate to bargain for express provisions having that effect. The Court’s doing so would seem to be all of a piece with the teaching of Lincoln Mills. The Supreme Court was free to disagree with the Sixth Circuit’s decision as a matter of labor policy. It could have held that the general context of collective bargaining had to accommodate a national labor policy expressed in ERISA that declined to vest medical benefits. Were it to have done so, the law on the vesting of collectively bargained terms would continue to be a matter of substantive law, just as Harold Katz said in his survey of the subject more than a half century ago. But the Court did no such thing. It set aside the Sixth Circuit’s precedent not as a matter of labor policy, but on novel methodological grounds, because the lower court had ignored “ordinary principles” of contract law that are contrary to and inconsistent with the Court’s long-standing understanding of the CBA.
The Court doubled down on its view equating interpretation of a CBA to interpretation of a commercial contract in CNH Industrial, N.V. v. Reese. Like Tackett, the case concerned whether retiree medical benefits provided for in a CBA vested for life. The CBA had a general duration clause saying that it would expire in May 2004. It also provided for medical benefits to employees who retired under the pension plan and that all other benefits, such as life insurance, would cease upon retirement. The Sixth Circuit found the agreement ambiguous because it was silent as to when medical benefits would cease, had other benefits such as life insurance expiring at times other than the expiration of the CBA, and tied medical benefits to pension eligibility. The ambiguity, the Sixth Circuit reasoned, enabled it to consider evidence extrinsic to the written agreement, and, based on that consideration, it held the medical benefits vested for life.
The Supreme Court would have none of that. It chided the Sixth Circuit for failing to follow ordinary principles of contract interpretation, the same principles that the Court relied on in Tackett. There was nothing express in the CBA stating that retiree medical benefits were not subject to the general duration clause, and so, under the unambiguous contract language, they expired with the CBA in May 2004.
III. Tackett-Inspired Mischief
Tackett’s view that CBAs are to be interpreted in accordance with ordinary principles of contract law is not surprising in light of Pyett’s treatment of the arbitration provision of a CBA as no different from arbitration provisions in commercial contracts. Although in both cases, the Court ignored more than a half century of its jurisprudence on CBAs, it also did not repudiate its vast case law back to Lincoln Mills and the Steelworkers Trilogy. It is tempting to dismiss Pyett and Tackett as anomalies destined to be limited to their peculiar circumstances. Pyett may be dismissed as part of the continuing evolution of the Court’s jurisprudence over the arbitration of statutory claims which now emphasizes the enforcement of arbitration agreements in accordance with their terms. Tackett and CNH were cases where the court, rather than an arbitrator, had to interpret the CBA because the plaintiffs were retirees who were no longer bound by the CBA’s grievance and arbitration procedures. Arguably, an arbitrator facing the same issue may not be bound by the Court’s approach.
Even so, the view that CBAs are no different from ordinary commercial contracts is gaining traction. We focus below on one area where Tackett has already had a direct effect—the duty under the NLRA to bargain in good faith. We then turn to interpretation of CBAs in arbitration and judicial review of labor arbitration awards where the potential for Tackett-inspired mischief is even greater.
A. The Duty to Bargain
The NLRA imposes on unions and employers a duty to bargain in good faith. A party, however, may waive its right to bargain, and such waiver will provide a defense to a charge that the other party failed to bargain in good faith. An issue of waiver often arises when an employer unilaterally changes a term or condition of employment and defends against the union’s charge of failure to bargain by claiming that the union waived the right to bargain in the CBA. The argument often centers on the CBA’s management rights clause.
The Warrior & Gulf Court noted that collective agreements are commonly a compilation of diverse provisions, some providing specific objective criteria, others providing less specific standards that require judgment, some little more than to leave problems for future consideration—that is, for further bargaining should the occasion arise. The latter was often seen to be in conflict with management rights clauses that, in equally general terms, reserved matters not contained in the collective agreement for unilateral management action. The view of the NLRB and of a majority of the courts was that, to be effective, a waiver of the right to require an employer to bargain had to be “clear and unmistakable.” In 2007, the NLRB reaffirmed this doctrine, tracing it back to 1949. In MV Transportation, the Board majority joined a minority of U.S. Courts of Appeals abandoning the clear and unmistakable waiver standard in favor of a “contract coverage” test. “Under contract coverage, the NLRB will examine the plain language of the collective-bargaining agreement to determine whether action taken by an employer was within the compass or scope of contractual language granting the employer the right to act unilaterally.” Heretofore, a reservation of managerial rights had to acknowledge that the specific subject as having been resolved. As Member McFerran opined in dissent, under the majority’s decision a general reservation of management rights, if sufficiently broad, would truncate the duty to bargain as to any more specific term and condition of employment, this under the “plain meaning” rule of ordinary contract construction embraced by the Tackett Court.
Of interest here is not the wisdom of the decision, but rather its treatment of the law of collective agreements, which the majority addressed in two different parts of the opinion. The first is found in the majority’s section devoted to criticism of the clear and unmistakable test. Citing Lincoln Mills and the Trilogy as the law governing CBAs, the majority argued that the clear and unmistakable standard put the Board crosswise with the courts because arbitrators, not the Board, are the primary source of authority on how CBAs are to be read. As the courts (actually a minority of them) did not defer to the NLRB’s requiring clarity and unmistakability, adoption of the contract coverage test would resolve the conflict by aligning the Board with the courts. As arbitrators are less likely than the NLRB to require express reference to a matter claimed to have been waived, the Board claimed, citing arbitral awards favorable to the claim, continued insistence on the clear and unmistakable test only encourages unions to resort to the Board over use of the grievance-arbitration process contrary to the national labor policy strongly favoring arbitration. How favored played a later role in the decision.
The second part in which the NLRB addressed the collective agreement is found in the section defending the contract coverage test. This time, the majority made no reference to Lincoln Mills or the Trilogy. Instead it quoted Tackett, according to which holding the majority grounded the test as being driven by “ordinary principles of contract law.”
Under MV Transportation, Tackett now applies to the NLRB when it, as contract-reader (i.e., interpreter of the CBA) decides a question of waiver. But the Board deployed Tackett to get contractual waiver disputes before arbitrators. In fact, the General Counsel issued guidance that these section 8(a)(5) cases are to be deferred to arbitration so long as the change is material, substantial, and significant without application of the contract coverage test at this stage. On its face, MV Transportation says nothing about how arbitrators must read collective agreements. To that extent, so far the arbitral role has not changed.
Whether it has changed can be tested by a real (but anonymized) case involving employees at a specific company. Before the affected employees were unionized, they recorded their time by a paper record. Error in timekeeping, from neglect, mistake, and misuse, was rampant. Once unionized, one of management’s demands was for the installation of an electronic timecard system, to which the union agreed. In addition to a general reservation of management rights, the collective agreement read as follows:
4.11 TIME AND ATTENDANCE MONITORING
(a) Time-keeping
All employees shall be required to verify and submit their attendance and hours of work during each pay period using time-keeping methods such as electronic timecards, certificates of attendance, or time clocks. Employees’ regular paychecks, including wages and salary for time worked, holiday pay and other paid leaves, shall be generated based upon time submitted.
After ratification of the agreement, the employer installed an electronic card-swipe timekeeping system. Even then, however, some employees were swiping in cards for coworkers. Without notice to or bargaining with the union, the employer contracted with a biometric technology company to add a fingerprint identification appliance to the card-swipe system for identification verification. The union objected and demanded to bargain about it. The company refused, pointing not only to the general reservation of managerial power, but, in particular, to Article 4.11’s allowance of the power to deploy “electronic timecards.” This, it argued, necessarily included the electronic system’s biometric identification as a matter of the contract’s plain language.
The union filed a charge of violation of section 8(a)(5) with the NLRB and also filed a grievance under its contract, which grievance the company denied. In the event, under the General Counsel’s policy implementing MV Transportation, the charge would have to be referred to arbitration.
Now assume that the arbitrator sustains the grievance, as the arbitrator did in the actual case. The opinion reasons that the use of biometric identification, never contemplated by the parties in bargaining over Article 4.11, is qualitatively different from an electronic time-keeping device which the parties’ contract allowed: the latter merely records time; the former deploys a biometric to identify who is actually recording the time. That technology, akin to a drug or a genomic screen, identifies something someone is, a part of one’s biological make-up. Its use involves privacy interests far deeper and more substantial than a material possession, a swipe card. Indeed, the arbitrator notes how the law globally and in the United States evidences the depth of social concern in the matter, citing inter alia Illinois’s biometric identification law. The arbitrator reasons that, as the purpose of collective bargaining is to give employees a meaningful say in those critical elements of their working conditions that materially and substantially affect them, it should take more than a reservation of management’s right to install an electronic time-keeping system to waive their critical right to be heard on so significant a matter, such being the policy of the NLRA. Accordingly, the arbitrator orders the company to bargain. In this scenario the case would return to the NLRB.
The Board would now have to decide whether the award is “clearly repugnant” to the Act. And this poses a question of no small interest. In Babcock & Wilcox Construction Co., the NLRB held that the arbitrator’s award must evidence a reasonable appreciation of the statutory principles that would govern the Board’s decision were it to have decided the case de novo. It is obvious that were the Board to have decided the case de novo, the contract coverage test would compel it to dismiss the complaint. Consequently, because the statutory principle the NLRB would apply is now one of contract coverage, it would appear to be the case that under Babcock & Wilcox this award would be due no deference; the complaint should be dismissed. The rub is that nine months before MV Transportation, the Board overruled Babcock & Wilcox in United Parcel Service, to return the state of the law prior to Babcock & Wilcox. Under that approach, the Board would defer to the award unless it was “not susceptible of an interpretation consistent with the Act,” or was “palpably wrong.” That decision dwelt extensively on the special role of arbitration under the Trilogy even to the point of discounting “the need for uniformity of result or the correct resolution of the dispute in every case.” It is “only ‘when the arbitration award cannot be arguably reconciled with the policies of the Act’” that deferral will be declined.
Consequently, the case tests the impact of MV Transportation. On the one hand, the award is in keeping with the language of the Trilogy that the Board relied on in MV Transportation and on the emphasis the Board placed on the Trilogy in United Parcel, that the strong preference for arbitration outweighs the need for uniformity of result. On the other hand, if the contract coverage test were compelled by the policy of the Act, which is what MV Transportation says, then the NLRB could not defer to the award, it could not be “reconciled with the policies of the Act,” giving effect to the application of that most ordinary principle of ordinary contract law, the “plain meaning rule.” Lest there be any doubt in the matter, the arbitrator’s reference to drug testing, as implanting so significant a privacy interest that absent reference to it the bargaining history it could not be considered waived, was expressly rejected by the Board in MV Transportation.
It would seem reasonable to conclude that the NLRB, were it to follow its analysis in MV Transportation, would not defer to the award and that the union could not secure a bargaining order from the Board. Nevertheless, the union could seek to enforce the award in court. Should the award be enforced judicially?
According to the teachings of the Supreme Court prior to Pyett and Tackett, it should. Under Enterprise Wheel and even more emphatically under recent authority—Misco, Eastern Associated Coal, Garvey—mistakes of fact, of law, or of contract reading are not grounds to set an award aside. So long as the arbitrator read the contract and applied those considerations of industrial self-government applicable to how it is to be read, as so richly explained in Warrior & Gulf, the rightness or wrongness of the arbitrator’s decision is not to be second-guessed by the courts.
If the preceding analysis is correct, the result could be the emergence of two inconsistent bodies of law on the waiver of bargaining during the term of a CBA: one, based on the conception of the CBA as performing the special function constitutive an industrial government to which the labor arbitrator gives meaning and in which an arbitrator’s judgment that a collective agreement requires a matter to be bargained about by the parties is not to be second-guessed; the other, based on a conception of the collective agreement as subject to the ordinary principles of contract law no different from any other ordinary contract (whatever that may be), as a result of which unions are to be precluded from being heard on matters not known or even conceived of at the time a management rights provision was agreed to so long as the subject is within the “plain meaning” of the contract’s terms.
B. Tackett and Labor Arbitration
A core foundation of the United States’ system of labor relations is that labor arbitration, Pyett notwithstanding, is not a substitute for litigation but is a substitute for workplace strife in a system of continuous collective bargaining. Consequently, courts asked to review grievance arbitration awards subject them to one of the narrowest standards of review known to the law. As the Supreme Court has said, “As long as an arbitrator is even arguably construing or applying the contract and acting within the scope of his authority,” the award must be enforced. But the view expressed in Tackett that CBAs are to be interpreted in accordance with ordinary rules of contract interpretation threatens to upend that deference.
There has always been some tension between the Court’s admonition in Enterprise Wheel that courts not enforce the awards of arbitrators who do not confine themselves to interpreting and applying the CBA but instead dispense their own brand of workplace justice and the Court’s directive that awards be enforced as long as they draw their essence from the CBA. In Misco, the Court reinforced the broad deference courts are to pay to labor arbitration awards, declaring that, as long as the arbitrator is even arguably interpreting or applying the CBA, the award must be enforced but also admonishing that arbitrators may not ignore the plain meaning of CBA language. As Justice Stevens subsequently bemoaned, “Our cases . . . do not provide significant guidance as to what standards a federal court should use in assessing whether an arbitrator’s behavior is so untethered to . . . the agreement of the parties . . . so as to constitute an attempt to ‘dispense his own brand of industrial justice.’”
Misco’s gloss on Enterprise Wheel had a particular effect in the Sixth Circuit. In Cement Divisions, National Gypsum Co. v. United Steelworkers of America Local 135, the court held that an arbitration award fails to draw its essence from the CBA and will be vacated if it conflicts with express terms of the CBA, imposes requirements not expressly provided for in the CBA, cannot rationally be derived from the terms of the CBA, or “is based on general considerations of equity instead of the precise terms of the agreement.” In Michigan Family Resources, Inc. v. Service Employees International Union Local 517M, the court overruled National Gypsum, concluding that it conflicted with Misco’s directive that as long as the arbitrator is arguably interpreting or applying the CBA, the award must be enforced. The court noted that under Misco, “an arbitration decision could be so ‘ignor[ant]’ of the contract’s ‘plain language’ as to make implausible any contention that the arbitrator was construing the contract.” Such, said the court, would be “the rare case,” and “only the most egregious awards [will] be vacated.”
Courts have vacated arbitration awards for ignoring the plain language of the CBA but have largely done so only in egregious cases. Typically, the cases involved arbitrators whose decisions contradicted their express findings. For example, in Excel Corp. v. United Food & Commercial Workers Local 431, the employer terminated an employee after the employee had been on medical leave for an on-the-job injury. The CBA provided that employees would lose their seniority if, among other things, they were absent from work for twelve months. The arbitrator acknowledged that this contract language supported the employer’s position that it could terminate employees whose medical leaves extended beyond a year. The arbitrator also found that the record did not support a conclusion that the employer had discriminated against “handicapped” individuals in violation of the CBA’s non-discrimination provision. Despite these findings, the arbitrator said that the loss of seniority provision conflicted with the non-discrimination provision and turned to evidence of bargaining history to resolve the conflict in favor of an exception for employees on medical leave due to on-the-job injuries. The incongruity between the arbitrator’s findings and conclusion led the court to vacate the award. The court wrote:
The arbitrator expressly states that “it is impossible to make a definitive finding as to whether the Company has discriminated against employees in contravention of Article III’s handicape [sic] ban.” The arbitrator nevertheless concludes that “the Company has violated Article III and Article XIII, Section 7E, of the contract by terminating such employees after their twelve-month medical leaves of absence expired.” Therefore, the arbitrator’s own factual findings not only directly contradict with the arbitrator’s ultimate ruling, but also support our conclusion that the seniority provision has not been shown to be discriminatory against handicapped individuals as applied in this case.
Despite these extreme facts, Circuit Judge McMillan dissented, urging that the award still drew its essence from the CBA.
A similarly divided Third Circuit, over the dissent of then-Circuit Judge (now Supreme Court Justice) Alito, vacated an arbitrator’s award in Pennsylvania Power Co. v. Local 27, International Brotherhood of Electrical Workers. The agreement at issue provided for early retirement benefits if the employees cooperated with management efforts to improve efficiency. The arbitrator found that the employees failed to cooperate but awarded the early retirement benefits anyway because the employer had provided the benefits to its supervisors. The arbitrator relied on a different provision of the CBA, which stated that the employer and union would “not discriminate, coerce nor intimidate any employee because of membership or non-membership in the Union.” The court majority reasoned that the arbitrator impermissibly amended the CBA to require the employer to provide the same benefits to employees as it provided to its supervisors and branded the arbitrator’s reliance on the non-discrimination clause as “unreasonable and impermissible.” Dissenting Judge Alito agreed with the majority’s view of the non-discrimination clause but would have enforced the award because, regardless of its flaws, the arbitrator’s opinion was arguably construing the CBA.
Several decisions from the Seventh Circuit illustrate how confined the “rare” and “egregious” case of arbitral ignorance of a CBA’s plain meaning has been. In Anheuser-Busch, Inc. v. Teamsters Local 744, the CBA provided that drivers-salespeople receive different commission rates if they operated alone or had a helper. Earlier CBAs provided a single commission rate regardless of whether the driver had a helper but, in their 1990 contract, the parties moved to a two-tiered commission system. The system was retained in their 1994 and 1998 contracts. Throughout the period governed by the 1990 and 1994 contracts, the employer continued to pay the higher commission rate even when the driver had a helper. The employer continued the practice for the first two months under the 1998 contract but then announced that it would switch to the two-tiered commission rate and did switch going forward, and the union grieved. The CBA also provided that it “supersedes all prior agreements and practices not specifically preserved in the contract.”
The stipulated issues presented to the arbitrator were, “Did the company violate the labor agreement by changing its practice to conform to the contract provision relating to two-person commission rates?” and “Is the company violating the contract when complying with the written terms of the most recent labor agreement?” The arbitrator sustained the grievance, reasoning that the employer’s payment of the higher commission rate in the first two months of the 1998 contract continued a practice that existed throughout the two prior contracts, that deviation from that practice was unfair, and that the parties had impliedly modified the written 1998 contract.As Judge Rovner observed in her concurring opinion, “[T]he arbitrator framed the question before him: ‘I must decide whether a long-standing past practice supersedes the contract language,’” a question “answered unequivocally by Article 20, Section 20—all prior practices not expressly preserved by the contract are displaced, and the written terms of the contract control.”
Judge Easterbrook dissented. He would have enforced the award because it was arguably based on the arbitrator’s finding of an implied modification of the CBA after the contract took effect, even if in doing so the arbitrator “made a whopper of an error.”
That the Anheuser-Busch majority regarded the case as the rare and egregious exception to deference to the arbitrator is clear when reading it in light of other Seventh Circuit opinions. For example, in Northern Indiana Public Service Co. v. United Steelworkers of America, the CBA provided for payment of bonuses in accordance with a chart showing bonuses to be paid each year if the employer’s earnings per share and pre-tax operating income exceeded specified amounts. When the employer’s pre-tax operating income exceeded the highest amount listed on the chart, the employer paid the bonus corresponding to the highest amount. The union grieved, and the arbitrator interpolated to the higher earnings to come up with a higher bonus than the highest provided in the CBA and awarded that bonus. The court enforced the arbitrator’s award. It reasoned:
[A]lthough the arbitrator was not empowered under the CBA to add terms to the PRP, arbitrators are empowered to fill gaps left in contracts. . . . In its appellate brief, the employer has persuasively argued that the arbitrator’s interpretation may well have been unsound. This is not enough. All this just amounts to saying that the arbitrator may have been wrong, maybe even clearly wrong; it does not show that he was doing something other than interpreting the contract. So long as the award is based on the arbitrator’s interpretation—unsound though it may be—of the contract, it draws its essence from the contract.
In International Union of Operating Engineers Local 139 v. J.H. Findorff & Son, Inc., the Seventh Circuit, following the Supreme Court’s admonition to enforce awards as long as the arbitrator is even arguably interpreting the CBA, resoundingly rejected the notion that a court may vacate an arbitration award because the award conflicted with the plain meaning of CBA language. The district court had vacated an award as contrary to the plain meaning of the CBA. The Seventh Circuit reversed. It reiterated its prior holdings that “the question for decision by a federal court asked to set aside an arbitration award . . . is not whether the arbitrator or arbitrators erred in interpreting the contract; it is not whether they clearly erred in interpreting the contract; it is not whether they grossly erred in interpreting the contract; it is whether they interpreted the contract.” The court reasoned:
[T]he judge applied a “plain-meaning exception” to the normal rule that an arbitrator’s power to decide includes the power to err. Apart from what the Supreme Court has had to say about the propriety of such an exception is the fact that what may seem “plain” to a judge is not necessarily plain to persons with greater experience in the business that the agreement is designed to cover. Arbitrators, often chosen because of their expertise in the industry, may see nuances that escape generalist judges. Persons steeped in the specialized language of a trade, or the business norms against which the language was written, often eschew “plain meaning” in favor of context, while generalists use a more text-bound approach because that is easier and less error-prone for outsiders.
After Tackett, however, some courts have aggressively vacated awards they believed conflicted with the plain meaning of the CBA. A year after Tackett was decided, the Seventh Circuit seemingly broadened the range of awards it will vacate beyond the rare egregious case. At issue in U.S. Soccer Federation v. U.S. National Soccer Team Players Ass’n were sponsor uses of the likenesses of six or more team members in non-spot (i.e., non-video) creatives. The CBA required that for spot uses of likenesses of six or more players, the Federation had to provide a copy to the Association for its approval. For non-spot uses, the CBA provided that the Federation would request but not require the sponsor to make a contribution to the Player Pool. For twelve years, the Federation submitted non-spot matters to the Association for its approval, but, when the Association disapproved one, the Federation declared that it would no longer submit them. The Association grieved.
The arbitrator sustained the grievance, reasoning that the CBA was silent as to whether non-spot material had to be submitted to the Association for approval. Therefore, the arbitrator looked to the consistent past practice and held that the CBA required the Federation to submit non-spot material to the Association for approval. The arbitrator’s reasoning was not rare or egregious. It fell within the mainstream of arbitrator approaches to interpreting CBAs. Although the Seventh Circuit gave lip service to the Supreme Court’s and its own prior decisions’ admonitions that awards should be enforced as long as they draw their essence from the CBA and as long as the arbitrator is even arguably interpreting the contact, the court held that the arbitrator disregarded the plain language of the CBA, which provided for the Federation to ask the sponsor to make a voluntary contribution and disregarded the CBA’s integration clause, and therefore the award did not draw its essence from the CBA.
The Third Circuit has also expanded the range of awards that it will vacate beyond the rare and egregious. In Monogahela Valley Hospital, Inc. v. United Steelworkers, the CBA provided:
Vacation will, so far as possible, be granted at times most desired by employees; but the final right to allow vacation periods, and the right to change vacation periods[,] is exclusively reserved to the Hospital. Any changes in vacation schedules may be realized by mutual consent. In the event the Hospital unilaterally changes a schedule causing the employee to suffer financial loss, the Hospital agrees to reimburse the employee for provable loss.
The grievant was denied her vacation request for Christmas week because her supervisor had also requested that week and both could not be away at the same time. The arbitrator sustained the grievance, reasoning that the “exclusively reserved to the hospital” language could not be read to negate the “so far as possible” language. In the arbitrator’s view, the hospital’s interpretation would mean that management could always deny employee vacation requests. In the words of Misco, the arbitrator was more than arguably interpreting the contract. This was a garden-variety contract interpretation award where the arbitrator had to reconcile seemingly contradictory language. The Third Circuit, however, did not see it that way. The court declared, “We begin with the obvious: an arbitrator may not ignore the plain language of the contract.” In the court’s view, the plain meaning of final and exclusive was to give the hospital absolute authority to deny vacation requests. As for the arbitrator’s concern that such an interpretation would negate the contractual requirement that vacations be granted so far as possible, the court reasoned:
“So far as possible” cannot hold hostage what follows here. It is a subordinate phrase clearly qualified by the superseding “but,” and what follows grants the Hospital the “final,” “exclusive[ ]” and “unilateral[ ]” right to schedule vacations. “It is . . . well established that contract language must be read in context and that a subsequent specification impliedly limits the meaning of a preceding generalization.”
Although the Seventh and Third Circuits did not cite Tackett, their reasoning parallels that of Justice Thomas. Just as the Court in Tackett rejected the Sixth Circuit’s analysis as inconsistent with ordinary principles of contract law, the Seventh and Third Circuits appear to have relied on ordinary principles of contract law in rejecting the arbitrator’s interpretation of the CBA. The courts ignored the admonition that courts should limit their inquiries to whether the arbitrator was even arguably interpreting the contact and vacated the awards, concluding that the awards failed to draw their essence from the CBA.
The courts are substituting their views of what the plain meaning of contract language is for the arbitrator’s view. This is in keeping with ordinary principles of contract law. Under ordinary principles of contract law, the interpretation of contract language is a question of law for the court, subject to de novo review on appeal. Whether written contract language is ambiguous is similarly treated as a question of law. Thus, under ordinary principles of contract law, arbitral interpretation of CBA language is entitled to no deference by a reviewing court.
The lower courts have not been unanimous in their abandonment of Trilogy-inspired approaches and their embrace of Tackett-inspired review. For example, in Exide Technologies v. International Brotherhood of Electrical Workers Local 700, the arbitrator sustained a grievance challenging the employer’s change from having employees submit Family & Medical Leave Act leave requests to its human resources office to submission to an outside third-party administrator. The arbitrator sustained the grievance, narrowly interpreting the CBA’s management rights clause, which vested in the employer the right “to conduct its business in all particulars except as expressly modified by [the CBA] and any written supplements to the [CBA] . . . [and] all management functions including the full and exclusive control, direction and supervision of operations and the working forces . . . .” The court reasoned:
In his findings and conclusions, the arbitrator distinguished FMLA leave administration from other management functions on the basis that “[a]dministrative processing and evaluating employees’ requests for statutory leave under the FMLA is not core entrepreneurial activity as it does not change the scope, direction or nature of the enterprise.” And later in the opinion, the arbitrator stated: “As for Article II, Management Rights, it does not specifically reference or identify leave administration as reserved solely to management.” The arbitrator thus considered the CBA’s management-rights section. We have “no business” second-guessing his interpretation of it.
Within the Third Circuit, there appears to be disagreement. In Independent Laboratory Employees Union v. ExxonMobil Research & Engineering Co., the CBA “allow[ed] the Company to ‘let independent contracts’ so long as ‘during any period of time when an independent contractor is performing work customarily performed by employees and employees qualified to perform such work . . . are available . . . the Company may not because of lack of work demote or lay off any employee(s) qualified to perform the contracted work.’”
The union grieved when the employer, in 2015, after internal advertising failed to fill a vacancy caused by the retirement of a bargaining unit employee, contracted out the work performed by the position. The arbitrator interpreted the subcontracting language “during any period of time when an independent contractor is performing work of a type customarily performed by employees” to limit the employer to subcontracting for a specific period of time rather than indefinitely as was done in the case before her. She also relied on the CBA’s recognition clause in finding that the subcontracting was eroding the bargaining unit, thereby eroding the union’s recognition as exclusive representative; a 1977 statement from an employer vice president in the course of a grievance proceeding that in the case under discussion there was no intent and that there would never be any intent to erode the bargaining unit; and two prior arbitration awards that had upheld subcontractings finding that they were supported by operational need but in dicta suggested that subcontracting to replace bargaining unit employees would not be permissible.
The employer asked the court to vacate the arbitration award because it contradicted the plain language of the subcontracting provision of the CBA allowing it to subcontract but precluding it from laying off or demoting any bargaining unit employee. Writing the opinion of the court, Judge McKee considered the arbitrator’s interpretation that the subcontracting language coupled with the recognition clause precluded permanent subcontracting as plausible. Judge McKee also found the award supported by the two earlier arbitration awards and by the 1977 vice president statement, citing the latter as distinguishing the case from Monogahela Valley Hospital.
But the other two judges on the panel saw the case as much closer. In a concurring opinion, Judge Bibas disagreed that the arbitrator’s interpretation of the subcontracting language was plausible, opining, “The text lets Exxon hire contractors indefinitely, so long as it does not fire or demote qualified Union employees during that time.” He observed, “If this were a contract case, I would stop there and reverse the award. The specific contracting-out power in Article XVIII means what it says: Exxon may hire contractors.” But he regarded the recognition clause, the two prior arbitration awards, and the 1977 vice president statement as providing sufficient support for the award to justify judicial enforcement of it, while cautioning, “This case is at the very outer edge of our deference [to the arbitrator]. The arbitral award is contrary to the fairest reading of the text.” The third member of the panel, Judge Cowen, concurred in the judgment “for the reasons set forth in Judge Bibas’s concurring opinion.”
We are also seeing movement expressly relying on Tackett to supplant traditional analysis with respect to compelling arbitration under CBAs. At least one circuit court judge appears to regard Tackett as supplanting the Trilogy and its progeny. Society of Professional Engineering Employees in Aerospace v. Spirit Aerosystems, Inc. concerned the union’s lawsuit to compel arbitration of a class-action grievance filed by an individual employee. The majority compelled arbitration, applying the Trilogy approach that a grievance is presumed arbitrable unless it can be said with positive assurance that the parties did not intend it to be subject to the arbitration procedure. Judge O’Brien concurred, but he disclaimed reliance on the Trilogy’s presumption of arbitrability and instead cited what he concluded was the unambiguous language of the CBA and Tackett.
Because of Tackett, the Third Circuit has overruled its precedent on arbitrability. In Luden’s, Inc. v. Local 6, Bakery, Confectionary & Tobacco Workers International Union, a Third Circuit panel held that an implied-in-fact contact may arise following the expiration of a CBA to arbitrate grievances and a court may compel arbitration despite Litton. In Pittsburgh Mailers Union Local 22 v. PG Publishing Co., a different panel of the Third Circuit overruled Luden’s. The Pittsburgh Mailers panel observed that the CBA had a general duration/expiration clause and that the arbitration provision did not have its own duration/expiration provision. The panel read Tackett as requiring it to apply ordinary contract principals and that one such principal was that a general duration provision controls when an article does not have its own specific duration provision. The panel further reasoned that Luden’s holding that an implied-in-fact, post-expiration agreement to arbitrate grievances could arise was inconsistent with ordinary contract principles. The panel wrote, “The rule does not preclude the parties from coming to an agreement after the expiration of the CBA to arbitrate disputes in accord with contract law principles. However, such an agreement must be a complete one, encompassing all necessary provisions. It cannot be merely ‘implied.’” The panel acknowledged the extraordinary nature of what it was doing—overruling precedent of a prior panel without an en banc hearing by the court. It justified its action on the ground that “[o]ur review of Tackett and Reese convinces us that the holdings in those cases undermine our opinion in Luden’s.”
Courts like the Third and Seventh Circuits have usurped from arbitrators the authority to read CBAs, at least where their view of the plain meaning of contract language differs from the arbitrator. Although they have not cited Tackett, the decision gives them ammunition to claim that they are acting as the Supreme Court has charged them to do. The Third Circuit has relied on Tackett to overrule a prior panel’s decision on arbitrability without an en banc hearing. And one circuit judge in the Tenth Circuit has turned to Tackett rather than the Trilogy in deciding whether to compel arbitration. Tackett is a time bomb that may not but easily could explode the Trilogy.
IV. Does It Matter?
If Tackett’s cue is pursued, two unsettled consequences emerge straightaway. First is to exacerbate a collective action problem. To the extent the courts take adherence to “ordinary principles of contract law” to direct the way that arbitrators are to read the terms of collective agreements, the courts could understand their role to police arbitral faithfulness to that dictate. This would encourage the courts to second-guess arbitral contract reading, to substitute their judgment for arbitral ones. The immediate cost would be the erosion of arbitral finality, as losing parties would understandably be encouraged to seek to set aside awards with which they disagree. It is not in the long-term interest of either management or unions to erode arbitral finality: the parties’ long term interests are served in having disputes over contractual readings resolved quickly, in having certainty in what management may or may not do for the agreement’s term and which, at its conclusion, they are free to agree to change. Even so, a losing party in any individual case may seek the short-term advantage of contesting the outcome even though such is not to the collective interest of both parties in the longer run. Whence the collective action problem. If Tackett invites greater scrutiny, it heightens the prospect of that individual resort the short term with the negative longer-term institutional consequences.
The second is more speculative and more serious. Warrior & Gulf distinguished the arbitral role in a labor dispute from the judicial role in a commercial dispute: the arbitrator may consider “such factors as the effect upon productivity of a particular result, its consequences to the morale of the shop, [and] his judgment of whether tensions will be heightened or diminished . . . to further [the parties’] goal of uninterrupted production . . . .” This is in keeping with the Court’s view of the arbitrator as an extension of the collective bargaining process. To bind a labor arbitrator to the ordinary principles of contract law is to imply that the CBA should be treated as an “ordinary” contract, as a commercial transaction not for the provision of goods but of labor.
The evil instinct in this metamorphosis can be illustrated by pursuing that analogy. A purchaser of goods whose seller fails to meet the contractual time limit for delivery may cancel the contract for non-performance. The fact that other sellers having no contractual privity among themselves have been equally derelict in shipment whilst their contracts were not terminated is irrelevant to the purchaser’s legal right to terminate. Such is the ordinary principle of contract law.
In labor arbitration, it is a bedrock principle of arbitral practice that like cases must be treated alike: it is a defense in a termination case that other employees had engaged in the very same misconduct which the employer knew and took no disciplinary action for. This rule, contra ordinary contract principles, is grounded in the consequences to employee morale, tensions, and productivity: that tolerating one person’s act while discharging another for the very same act is arbitrary and likely to play out badly on the shop floor. This is in keeping with the Trilogy’s teaching of the arbitral role. It is inconsistent with Tackett’s teaching on how collective agreements are to be read.
A case in which one of the authors (Malin) presided as arbitrator markedly illustrates the differences between viewing a CBA as an ordinary commercial contract with arbitration as a substitute for litigation and viewing the CBA as special with arbitration a continuation of the collective bargaining process. A new CBA between a school district and the union representing its teachers provided:
Retirees shall have two hospitalization options as follows:
Option 1: Those who choose to remain in the District’s self-funded insurance program shall pay a $300.00 single/$600.00 family deductible and the following monthly contributions.
2004 –2005
$ 50.00 for single coverage
$150.00 for dependent coverage
2005–2007
$100.00 for single coverage
$200.00 for dependent coverage
2007–2009
$125.00 for single coverage
$250.00 for dependent coverage
Under prior CBAs, retirees paid lower premiums and had no deductibles.
The evidence established that the parties never discussed the meaning of the word “retirees” during their negotiations. It further established that the union’s negotiators reasonably believed that the word referred only to teachers who retired after the effective date of the new CBA and had no reason to even suspect that the school district’s negotiators had a different understanding. Similarly, the school district’s negotiators reasonably believed that the word referred to all retired teachers, including those who retired under earlier CBAs, and had no reason to even suspect that the union’s negotiators had a different understanding. Finally, the evidence showed that, in the past when new CBAs raised retiree health insurance premium contributions, the higher rates were applied only to new retirees but when new CBAs made changes to the benefits provided, the changes were applied to existing retirees as well.
The arbitrator awarded that the premium contribution increases applied only to teachers who retired after the effective date of the new CBA but that deductibles applied to existing retirees as well. In so doing, the arbitrator recognized that he was reading the word “retirees” to mean one thing with respect to the first part of the sentence using the word (i.e., with respect to deductibles) and something different with respect to the second part of the same sentence (i.e., with respect to premium contributions).
The award would be difficult to square with ordinary principles of contract interpretation. Under such principles, the same word would be hard pressed to have different meanings in different parts of the sentence in which the word is used. But, when the CBA is regarded not as an ordinary contract but as a generalized code whose terms are subject to further refinement through the grievance procedure with the arbitrator serving as an instrument of the collective bargaining process, the award makes unassailable sense.
Were the CBA an ordinary contract with the arbitration a substitution for litigation under that contract and the arbitrator obligated to rule in accordance with ordinary principles of contract law, a possible (and likely) result would have harkened back to the 1864 decision in Raffles v. Wichelhaus, with the declaration that the parties had no contract at all. But unlike the ordinary contract situation, the union and employer did not have the option to simply go their separate ways. They remained legally obligated to deal with each other. Consequently, an award consistent with ordinary principles of contract law would have sent the parties back to negotiations even though they had already shown that their negotiations, through the grievance procedure, could not produce an agreed-upon resolution leading to the need to have their mutually selected arbitrator impose a resolution. For these parties, resolution in accordance with ordinary principles of contract law would have been untenable.
As the matter currently stands, Pyett regards CBAs as ordinary commercial contracts with arbitration as a substitute for litigation, whereas the Trilogy and its progeny regard CBAs as generalized codes subject to continuous negotiation with arbitration as a part of the continuing bargaining process. Tackett requires the courts to take one approach to contract reading, while the Trilogy authorizes arbitrators to take a different one. The two views of the CBA stand in a shaky relationship to one another. If Pyett and Tackett are read to eclipse the Trilogy, the result could upend a half century of law on which our system of unionized industrial relations has rested. The better approach, historically and conceptually, would be to confine Pyett and Tackett to their peculiar circumstances, leaving the Trilogy and its progeny to continue to govern CBAs generally.
Conclusion
For more than half a century, the Supreme Court has regarded CBAs as different from ordinary commercial contracts. Unlike ordinary contracts that are entered into without legal compulsion, unions and employers are legally required to deal with each other. CBAs result from a legally required process of collective bargaining that continues after the formal agreement is executed. Arbitration is an important part of this process of continuous collective bargaining with the arbitrator resolving the dispute when the parties’ negotiations have been unable to do so.
The Supreme Court’s recent decisions in Pyett and Tackett deviate from this long-accepted view of CBAs. Instead, they treat CBAs as no different from other contracts with arbitration serving as a substitute for litigation, rather than a part of the collective bargaining process and contract language to be interpreted in accordance with ordinary principles of contract law. These ill-conceived decisions threaten to upend the firmly established system of collective bargaining and labor relations. They should be confined to their peculiar circumstances and should not be allowed to infect the general approach to CBAs and their application and interpretation through grievance arbitration.
Such confinement should not be difficult. At issue in Pyett was the enforceability of the union’s agreement that employees would arbitrate rather than litigate their statutory claims. In other words, the CBA in Pyett, with respect to claims for violation of specified statutes, used the grievance and arbitration procedure as a substitute for litigation, rather than a continuation of the collective bargaining process. Pyett should have no application to claims arising under the CBA as opposed to claims arising under a statute. With respect to claims arising under the CBA, the grievance and arbitration process remains a substitute for job actions and other workplace strife. The Trilogy, and its progeny, should continue to apply.
Tackett concerns the standards to be applied when a judge, rather than an arbitrator, interprets a CBA in the first instance. The retirees in Tackett were no longer in the bargaining unit represented by the union and, therefore, no longer obligated to take their claims through the CBA’s grievance and arbitration procedure. Indeed, it is likely that they did not have access to the contractual procedure. But, as the Court recognized in Warrior & Gulf, a generalist judge lacks the specific qualifications and knowledge of a mutually selected arbitrator. Moreover, a judge is imposed on the parties, whereas an arbitrator is mutually selected by the parties. A judge is responsible to the legal system as a whole whereas an arbitrator is responsible to the parties. Mutual selection by the parties empowers the arbitrator to act as part of the parties’ continuous process of collective bargaining. Tackett controls a judge’s interpretation and application of a CBA in the first instance but should have no applicability to an arbitrator. Confining Pyett and Tackett to their peculiar circumstances precludes applying them to the more common cases of grievance arbitration. Doing so avoids the potentially calamitous undermining of bedrock principles of labor relations in the United States.
