Bargaining Rights Gone Wrong: How State Courts Invented a Constitutional Duty to Bargain and How It Harms Individual Workers

Constitutions often give you the right to do things. They give you the right to speak your mind, to petition the government, to “bear arms”—the list goes on. But do they also require other people to help you do those things? Do I have to help you exercise your constitutional rights?

Traditionally, the answer has been no. Constitutions usually bind only state actors,[1] and the state generally has no duty to help you exercise your rights.[2] It doesn’t have to give you a platform for your speech.[3] It doesn’t have to give you a gun to bear.[4] It doesn’t even have to listen to your petitions.[5] It only has to refrain from interfering with your right to do the constitutionally protected thing.[6]

But that principle hasn’t always held true for collective bargaining. In a few states, collective bargaining has been enshrined as a constitutional right.[7] Some courts have applied that right in the traditional way. They have protected it from interference, but not required others to facilitate it. They have safeguarded employees’ right to form unions, join them, and demand recognition. But they have not forced employers—public or private—to bargain in return. For these courts, bargaining remains a matter of consent.[8]

Other courts, however, have gone further. Not only have they protected collective-bargaining rights from interference, but they have also ordered employers to bargain in return. That is, they have required employers to participate in and facilitate the exercise of their employees’ rights.[9]

From a constitutional perspective, this was an unusual step.[10] Constitutions in the United States usually create negative rights, not positive ones.[11] But these courts justified it as a necessity. They reasoned that if they didn’t require employers to bargain, bargaining rights would be meaningless. After all, bargaining takes two parties. If an employer has no enforceable duty, it can frustrate bargaining with “surface” negotiating tactics. Or worse, it can refuse to bargain in the first place. Employees are left with only the right to demand bargaining, which isn’t much of a right at all. So, these courts reasoned, the right to bargain must include an implicit duty to bargain as well.[12]

That rationale makes some intuitive sense. When a constitution guarantees the right to bargain, we might reasonably assume the right involves at least two parties. But the more we think about it, the more we can see the holes in that assumption. For one, constitutional bargaining rights can do a lot more than force one party to the table. They can protect people who engage in traditional labor activities, such as forming and joining unions.[13] They can prevent states from outlawing bargaining altogether, as some other states have done.[14] They can even guard against more subtle anti-labor tactics, such as “yellow dog” contracts.[15] What’s more, they can do all those things without imposing any affirmative bargaining duty. So no, imposing that duty isn’t necessary to give bargaining rights meaning. Bargaining rights have meaning all on their own. If courts want to impose a duty, they have to justify their decision on some other ground.

That point becomes especially clear when we look at the duty’s collateral effects. Two of those effects deserve mention here. First, there is the problem of judicial administration. When courts announce a duty to bargain, they also have to explain how bargaining will work. And that explanation requires them to resolve dozens of mundane administrative issues. For example, what are the subjects of bargaining? Whom does bargaining cover? And what happens if the parties can’t agree? Courts are ill-suited to decide these questions de novo. Without guidance, they have to make up the rules as they go along. And as they make up the rules, they slowly creep outside their comfort zones into areas where, from the perspective of institutional competence, we’d probably prefer they didn’t go.[16]

That leads us to the second collateral effect. Among the problems courts have to solve is the risk of multiple unions. If the employer has to bargain with every union chosen by an employee, the employer might easily find itself bargaining with two, three, or a dozen unions in the same workplace. That kind of bargaining isn’t industrial democracy; it’s industrial chaos. The natural solution is to require an employer to recognize only one union. Once a union gets support from a majority of employees, it represents all the employees, even the ones who don’t want its services. It’s a neat solution—one well established in federal law. The only problem is that it has no basis in state constitutional text or history. So to adopt it, courts had to effectively rewrite their constitutions.[17]

This not only oversteps the judicial role; it also affects individual rights. When courts impose exclusive representation, they necessarily deny some employees their choice. Some employees get the representative they want, but others have one foisted upon them. And that means some employees lose the only right guaranteed by the text—the right to bargain collectively. Ironically, by adopting an unwritten employer duty to bargain, courts subverted the written right guaranteed to workers.

None of this had to happen. Some courts avoided the conflict by interpreting their constitutions in the traditional way: They protected bargaining rights from interference, but imposed no new duties. They stayed mostly on the sidelines, leaving questions about bargaining duties to legislatures. And by proceeding in that way, they neither trampled on individual rights nor stretched themselves beyond their sphere of competence.[18]

This article argues that the traditional approach is the right one. It supports that position by tracking developments in three states: New York, New Jersey, and Missouri. New York courts followed the traditional approach, protecting employees from interference while leaving bargaining duties to the legislature. New Jersey and Missouri courts, by contrast, sidelined legislatures and read bargaining duties into their constitutions. And as a result, these courts ran straight into the practical and theoretical problems inherent in court-ordered bargaining.

Those problems deserve our attention. State labor law is a much-neglected practice area—practically a doctrinal backwater. It gets little attention from the academy or press.[19] It almost always takes a back seat to its federal counterpart, which admittedly has a wider reach. But for millions of workers, state law is the only source of bargaining rights. So when state courts get the law wrong, their errors affect real people. We should study their decisions just as closely—and call out their errors just as vigorously—as we do with federal courts.

I. States’ Role in Regulating Collective Bargaining

When modern lawyers think of collective bargaining, they usually think of federal law. For decades, bargaining in the United States has been governed by the National Labor Relations Act (NLRA).[20] The NLRA regulates labor relations from beginning to end: from elections to decertification, and everything in between.[21] Because the NLRA is so comprehensive, it casts a wide shadow over state law.[22] Any state law regulating the same subjects as the NLRA—or even subjects Congress meant to leave unregulated—is preempted.[23]

Given this state of affairs, one could reasonably wonder what role is left for states. If the NLRA already regulates labor relations from front to back, what can states add? How much does state labor law matter?

The answer is more than you might expect. Though the NLRA is broad, it’s also full of holes. Probably the biggest hole includes government workers.[24] States and their political subdivisions, such as towns and counties, are exempted from the NLRA’s coverage.[25] As of 2019, more than 5 million people worked for state governments, plus another 14 million for local ones.[26] So all told, the NLRA’s government exemption carves out nearly 20 million workers.[27]

The NLRA also exempts some private-sector workers. For example, the statute excludes agricultural workers,[28] domestic workers, supervisors, and independent contractors.[29] Courts have likewise carved out managers and “confidential personnel” workers.[30] Similarly, some workers are left out because the National Labor Relations Board has declined to regulate them. While this category can shift depending on the Board’s views, it has at times included student athletes, teaching assistants, and racetrack employees.[31] And still other workers have been excluded on constitutional grounds. In one notable example, the Supreme Court carved out a First Amendment exception for religious schools.[32]

In each of these cases, the carved-out workers fall into a regulatory gap. Federal law does not apply to them, nor does the NLRA’s broad preemptive effect.[33] And with no preemption, states are free to apply their own law.[34] For these workers, then, state law matters a great deal.

So it is meaningful when a state enshrines collective-bargaining rights in its constitution. For a large group of the state’s workers, the state constitution serves as their primary, and maybe even only, source of bargaining rights.[35] Workers will look to the constitution to understand their rights and responsibilities.[36] And that means they will often look to courts.[37]

Courts, however, haven’t always treated bargaining rights uniformly. Some courts have interpreted them modestly, leaving room for legislatures to design bargaining systems within constitutional boundaries. But others have taken it upon themselves to write the rules of bargaining, largely sidelining legislatures. We can see this divide play out in three states: New York, New Jersey, and Missouri. The former took the modest, traditional approach, while the latter two staked out more aggressive positions. The primary difference between them was how they treated a single issue: an employer’s duty to bargain in good faith.

II. From Employee Rights to Employer Duties: The Judicial Creation of Mandatory Bargaining

A. New York

In 1938, New York’s voters approved a new constitution.[38] The new constitution included at least one provision that was the first of its kind: article I, section 17. That section guaranteed, as a matter of constitutional law, the right to bargain collectively:

Labor of human beings is not a commodity nor an article of commerce and shall never be so considered or construed. . . . Employees shall have the right to organize and to bargain collectively through representatives of their own choosing.[39]

Though new as a matter of constitutional law, these two sentences would have sounded familiar to many voters. The first echoed the Norris–LaGuardia Act.[40] Passed by Congress only six years earlier, that act mostly barred federal courts from issuing injunctions in local labor disputes.[41] The second sentence, meanwhile, echoed the NLRA’s section 7.[42] Section 7 had been adopted even more recently—only three years earlier. And it had used mostly the same words to guarantee bargaining rights. [43] But it did not contain an explicit duty to bargain.[44] Mirroring section 7, New York’s new constitution was likewise silent on bargaining duties.[45]

That silence would prove meaningful. In 1945, in Trustees of Columbia University v. Herzog,[46] the New York Appellate Division held that section 17 created no affirmative bargaining duty. The case involved a dispute between the New York Labor Board and Columbia University. The Board wanted to force Columbia to bargain with a union.[47] At the time, a state statute required most employers to bargain.[48] But it exempted some employers, including educational institutions.[49] Columbia argued that, as an educational institution, it was exempt under the statute and so had no bargaining duty. The Board disagreed. It claimed that even if Columbia fell within the statutory exemption, the university was still covered by section 17. And section 17 imposed a duty on all employers, whether or not covered by the statute.[50]

The Appellate Division rejected that argument.[51] It pointed out that, if the board were correct, section 17 would effectively nullify all statutory exemptions.[52] Every employer, regardless of the exemptions, would have a duty to bargain.[53] And there was no evidence that section 17’s drafters meant to override existing law. To the contrary, it appeared that they had wanted only to affirm “the right of labor to organize and bargain collectively which, in 1935, had found expression in the [statute].”[54]

The New York Court of Appeals affirmed without an opinion.[55] As a result, the Appellate Division’s opinion has been the definitive word on the right to bargain in New York. It has been cited to show that while section 17 protects the right to organize and select a representative, it imposes no reciprocal duty on employers.[56] In other words, section 17 is a shield against interference, not a sword to enforce bargaining.[57]

B. New Jersey

Just to the south, courts followed a different path. New Jersey rewrote its constitution in 1947, about a decade after New York.[58] Like New York, it included a new provision, section 19, guaranteeing the right to bargain collectively:

Persons in private employment shall have the right to organize and bargain collectively. Persons in public employment shall have the right to organize, present and make known to the State, or any of its political subdivisions or agencies, their grievances and proposals through representatives of their own choosing.[59]

Section 19 shared much with its northern predecessor. Both sections appeared in their respective bills of rights. They each gave private employees the right to “bargain collectively.” And they each closely followed NLRA section 7, which framed bargaining rights in similar terms. [60] Given those similarities, one might have expected New Jersey’s courts to read section 19 with the way New York courts read their section 17.[61] They might naturally have looked for guidance in New York’s caselaw, including Herzog.[62] This would have led them to conclude that section 19 protected bargaining rights without also foisting new obligations on employers.[63]

But instead, New Jersey courts diverged. The split emerged in Johnson v. Christ Hospital. There, a union sought to represent a nonprofit hospital’s employees. At the time, no statute required the hospital to bargain.[64] The only possible source of a bargaining duty was the new constitution.[65] So when the hospital refused to recognize the union, the union sued under section 19.[66]

The Chancery Division[67] accepted the union’s claim. It reasoned that although section 19 never mentioned a duty to bargain, it still implied one.[68] The constitution gave employees the right to bargain.[69] And for that right to mean anything, employers had to have a duty to bargain in return.[70] Any other interpretation would make the right to bargain “impotent.”[71] So the court ordered the parties to hold an election.[72] If a majority of employees voted for the union, the union would become the employees’ exclusive bargaining agent.[73] And the hospital would have a duty to bargain.[74]

The hospital appealed. Citing Herzog, it argued that section 19 protected the right to organize and bargain through a chosen representative.[75] But the constitution imposed no duty on an employer to recognize that representative, much less to sit down and bargain.[76]

The New Jersey Supreme Court rejected that argument. In a summary opinion, it affirmed the Chancery Division’s order. It recognized that imposing a bargaining duty was not a matter of simply applying the text.[77] Section 19 was not a comprehensive labor-relations statute; it cast bargaining rights only in “general terms.”[78] The text offered no details on implementation and no mechanism for enforcement.[79] No cause of action appeared, nor did any remedies. Nowhere could the court find guidance on how to put meat on the new right’s bones.[80] But even so, the court had a duty to enforce the constitution, whatever its ambiguities.[81] It would be “derelict” in that duty if it left the union without a remedy.[82] Section 19 left gaps, but someone had to fill them.[83] The Chancery Division had therefore been right to order an election and, if the union won, to order bargaining.[84]

In the same breath, however, the court recognized that judge-fashioned solutions were less than ideal in this space. Courts were not well suited to write day-to-day bargaining rules. The proper body to write those rules was the legislature. So the court called on the legislature to fill out section 19’s skeletal structure:

In the present state of the law the courts have the general power and the duty to determine justiciable labor disputes between nonprofit hospitals and their employees. At the same time we recognize that it is more expedient to have the day-to-day problems arising out of disputes concerning wages, hours and conditions of employment regulated by over-all legislation, than for the courts to set about the establishment of procedural and substantive precedents on a case-to-case basis.[85]

But the court’s call went unheeded. In the years following Johnson, no general labor law was forthcoming. Nor did any other guidance emanate from the legislature. Section 19 remained elliptical, and key questions remained unanswered. Most important, with whom did the employer have a duty to bargain? The Chancery Division had embraced majority rule, with majority status determined by an election.[86] And by affirming, the supreme court seemed to endorse that approach.[87] But the question remained without a definitive answer. By getting support from a majority, did the union win the right to represent everyone? Or did dissenters keep some bargaining rights for themselves?

The New Jersey Supreme Court eventually answered that question, but again, only implicitly. In Comite Organizador v. Molinelli, the court considered a suit by a group of farmworkers and their union. A majority of the workers signed cards designating the union as their representative.[88] When the union presented the cards to the farm, the farm refused to bargain. Instead, it fired the whole group of workers.[89] The union sued, and the Chancery Division ordered an election.[90] The union won the election and again demanded bargaining. But the farm refused a second time, and the union sued again.[91] The Chancery Division again found that the farm had unlawfully failed to bargain and ordered it to recognize the union.[92] But again, rather than sit down with the union, the farm appealed. It argued that despite the outcome in Johnson, section 19 imposed no duty to bargain on employers.[93]

The supreme court disagreed. Section 19, it explained, gave all employees the right to bargain. That right would be “emasculated” if the employer had no duty to bargain in return.[94] A majority of the workers had twice chosen a representative, and the farm had twice refused to respect their choice.[95] So again, echoing Johnson, the court held that the Chancery Division had been right to order an election, and following the election, to order the farm to bargain.[96]

Some of this reasoning may have been driven by bad facts. The farm explained its decision to fire the workers by saying that it had closed its operations.[97] But parts of those operations remained open, and the farm continued to employ other workers.[98] So its explanation carried more than a waft of mendacity. The farm had also failed to pay Social Security taxes on the workers’ wages—a fact irrelevant to the issue at hand, but one that still made its way into the court’s opinion.[99] And maybe worst of all, at least in the court’s eyes, the farm had never asked to set aside the Chancery Division’s original order.[100] It had simply ignored the order and continued to reject the union.[101] The supreme court could hardly stomach such brazen disregard for judicial authority.

Whatever role these facts played, the result was clear. Section 19 not only protected employees from interference, but also imposed an affirmative bargaining duty on employers.[102] And that duty attached to a single union, chosen by a majority vote.[103] The duty to bargain and exclusive representation had been lashed together.[104] As a unit, they had been firmly embedded in the state’s constitutional law.

C. Missouri

Missouri would reach the same conclusion, if only belatedly. The state rewrote its constitution in 1945.[105] Among the new constitution’s features was a right to bargain collectively.[106] The right appeared in article I, section 29, which read: “[E]mployees shall have the right to bargain collectively through representatives of their own choosing.”[107]

Like New York’s and New Jersey’s constitutions, Missouri’s new constitution mirrored section 7 of the NLRA. Like section 7, it guaranteed a right to bargain collectively through a chosen representative. Also like section 7, it said nothing about an employer’s duty to bargain.

That omission soon became important. In Quinn v. Buchannan, the Missouri Supreme Court held that section 29 created no new bargaining duties.[108] The case concerned a group of truckers, who had designated a local chapter of the Teamsters as their representative.[109] The Teamsters collected authorization cards from the truckers and presented the cards to the employer.[110] The employer responded by insisting that he would never have a union in his company.[111] He refused to recognize the Teamsters and, instead, fired the truckers.[112]

The truckers sued.[113] They argued that by firing them, the employer had violated their rights under section 29.[114] They sought damages, reinstatement, and an order forcing the employer to bargain.[115]

The Missouri Supreme Court agreed with them in part. It reasoned that section 29 gave the truckers a right to choose a bargaining representative.[116] That right implied the right to make an uncoerced choice.[117] And coercion was exactly what they had experienced: they had been fired for joining a union and demanding bargaining.[118] So section 29 entitled them to relief, including damages and reinstatement.[119] It did not, however, entitle them to a bargaining partner.[120] It said nothing about an employer’s duty, and nothing in the records of the constitutional debates suggested that the drafters meant to create one.[121] The drafters had not written section 29 as a “labor relations act, specifying rights, duties, practices, and obligations.”[122] If those obligations and duties were to exist, they would have to come from legislation.[123]

In this way, Quinn followed the Herzog model.[124] It found gaps in the constitutional language, but refused to fill them with judge-made solutions.[125] It instead applied the words as written and left legislators to fill in the rest.[126] It recognized that legislators, not courts, should decide whether and when to require bargaining.[127]

But unlike Herzog, Quinn would not stand the test of time. In 2012, some sixty years after Quinn, the Missouri Supreme Court reversed course. In a pair of decisions, it repudiated Quinn and held that section 29 did, in fact, create a duty to bargain.

The first decision came in Eastern Missouri Coalition of Police v. City of Chesterfield.[128] Chesterfield involved a drive to unionize police officers.[129] At the time, a Missouri statute gave bargaining rights to some state and local employees, but exempted police.[130] On the strength of that exemption, the city refused to recognize the officers’ union.[131] The union responded by suing.[132] A trial court held that despite the exemption, the officers had a right to bargain under section 29—and that the city had a duty to bargain in return.[133]

On appeal, the supreme court affirmed. It announced that the city had a duty to “meet and confer” with the union—despite the contrary conclusion in Quinn.[134] Quinn, it said, rested on an unstated assumption: that constitutional rights were inherently negative.[135] But that assumption was not invariably true. Nothing in doctrine or structure prevented constitutions from creating positive rights.[136] In fact, many other states had found positive rights in their constitutions. For example, Connecticut courts had found a positive constitutional right to an equal educational opportunity.[137] Montana courts had found a positive right to observe government meetings and inspect certain public documents.[138] Those decisions showed that constitutions could, in some cases, create positive rights. Nothing prevented section 29 from doing the same.[139] Quinn’s contrary assumption was flawed, and that flaw undermined the rest of its rationale.[140]

The court continued in this vein in a second case, American Federation of Teachers v. Ledbetter. Ledbetter involved negotiations between a teachers’ union and a board of education.[141] Like police officers, teachers were exempted from Missouri’s public-sector labor law.[142] Even so, the board voluntarily recognized the union and bargained with it for about a year.[143] The negotiators met nearly twenty times and came to a tentative agreement.[144] But when the deal was presented to the board, the board rejected it.[145] The board was particularly unhappy about the tentative deals on tenure and pay.[146] Because negotiations had carried on so long, there was little time to go back to the table; the next school year was fast approaching.[147] So rather than make a counterproposal, the board announced new salaries and sent individual contracts to the teachers.[148]

The union sued.[149] It argued that by making individual offers, the board had failed to bargain in good faith.[150] It conceded that the teachers were excluded by the statute. Even so, it claimed that they had a right to bargain under section 29, and that the school board had a reciprocal duty to bargain.[151]

Relying on Quinn, a trial court rejected that argument.[152] But the supreme court reversed.[153] Building on Chesterfield, the court reasoned that section 29 gave every employee, public or private, the right to bargain collectively.[154] That right would be empty if the employer had no duty to bargain in good faith.[155] Were it otherwise, the employer could stymie bargaining by using “surface” negotiation tactics.[156] Or worse, it could refuse to bargain at all.[157] Employees would have no way to force the issue; their bargaining rights would mean little more than the right to present grievances.[158] And for public employees, that would make their bargaining rights no different from the right to petition the government—i.e., their employer—making their rights would be effectively redundant.[159] The drafters surely hadn’t meant to duplicate existing rights.[160] They must have meant to give employees something meaningful. So the right to bargain must imply a reciprocal duty.[161]

To bolster that conclusion, the court looked to federal law.[162] It argued that while nothing in the constitutional debate records showed that the drafters meant to create an affirmative duty, contemporary federal authorities suggested that they may have assumed they were creating one anyway.[163] In the early 20th century, the War Labor Board,[164] the Railway Labor Board,[165] and the National Labor Board[166] had each concluded that the right to bargain collectively implied a reciprocal duty.[167] Section 29’s drafters must have been aware of those decisions and would have known that “collective bargaining” had become a term of art.[168] So they likely knew they were creating a duty to bargain, even if they never acknowledged it.[169]

The court’s decision was not unanimous. It drew a spirited dissent from Judge Zel M. Fischer, who accused the majority of reading new words into the constitution.[170] He pointed out that section 29 never mentioned a duty to bargain.[171] It used no words like duty, recognition, or good faith.[172] Instead, it simply gave employees a right to choose their bargaining representatives.[173] By expanding that right to include a reciprocal duty, he said, the majority was glossing the text with its own policy judgments.[174]

Judge Fischer also found it odd that the majority put so much emphasis on federal law while ignoring what section 29’s drafters actually said.[175] Looking to the convention debates, he found no evidence that the drafters meant to create a new bargaining duty.[176] Instead, by their own words, they were trying only to protect bargaining rights against legislative interference.[177] They said that section 29 would “preclude the possibility and the probability . . . [of] many bills being introduced seeking to destroy collective bargaining.”[178] But they never said they wanted to expand bargaining beyond its current status under state law, much less adopt federal law in its entirety.[179]

He likewise disagreed that a duty was necessary to make bargaining rights meaningful. Even without a duty, section 29 would protect employees from interference.[180] For example, in Quinn itself, the court granted relief to employees fired for joining a union.[181] That kind of retaliation was illegal under section 29 and well within the court’s power to remedy.[182] But it was not within the court’s power to convert section 29 into a full-blown labor-relations act.[183] The court had no authority to impose a new duty, much less to create the rules that went along with it.[184] That kind of detailed rulemaking could be done only by the legislature.[185] By taking that task upon itself, the majority had overstepped its proper judicial role.[186]

III. The Textual, Historical, and Structural Flaws of Judicially Imposed Bargaining

These decisions show how rights can develop in unexpected directions. In New York, courts read bargaining rights modestly and fit them into existing law.[187] Missouri courts did the same for nearly sixty years.[188] But later, Missouri courts abandoned that approach and imposed an unwritten bargaining duty. [189] And in New Jersey, courts took the duty-to-bargain approach from the beginning. [190]

What emerges is a lesson in unintended consequences. When planted in a governing document, even the barest of texts can sprout a tangle of judicial glosses.[191] And that kind of tangle can be hard to pull back from. Having waded into the thicket, courts struggle to extract themselves. Witness how New Jersey’s and Missouri’s courts took one step beyond the text—inferring an unwritten duty to bargain—and quickly found themselves straying even further. Lacking any statute or administrative system to fall back on, they reached for federal concepts—good faith, majority elections, and exclusive representation—rather than simply hewing to the plain text and the original meaning of their constitutions.[192] 

A. Text

Nothing in the text of either Missouri’s or New Jersey’s constitution supported a duty to bargain. The relevant texts said that employees had the right to bargain collectively through representatives of their choosing.[193] But they said nothing about an employer’s duty. They used no words like “duty to bargain,” “meet and confer,” or “good faith.” They offered no textual hook for an affirmative bargaining duty.[194]

Without such a hook, courts had to infer a duty. And to justify that inference, they had to borrow from federal law.[195] They pointed out that federal law required employers to bargain in good faith—i.e., with a genuine desire to make an agreement.[196] The state constitutional drafters were writing against the backdrop of federal law and surely would have been aware of it. So, they reasoned, the drafters must have assumed a similar requirement would apply under the state law provisions they were drafting.[197]

But that analysis glides over differences in the relevant texts. Both constitutions mirrored section 7 of the NLRA, which gave employees a right to bargain collectively.[198] Section 7 predated the two constitutions, so it was reasonable to assume that the drafters thought of it as a rough model. But that assumption doesn’t lead us to a bargaining duty, because there is no bargaining duty in section 7. The federal bargaining duty comes from a different part of the NLRA—section 8(a)(5).[199] That section makes refusing to bargain an unfair labor practice.[200] Likewise, section 8(d) (added later) explains that the duty to bargain includes the duty to “meet at reasonable times and confer in good faith.”[201] Both 8(a)(5) and 8(d) would be redundant if section 7 imposed a duty to bargain on its own.[202] They are necessary only because section 7’s language doesn’t do the job.

At least, that’s how the NLRA’s drafters saw it. When the NLRA was first proposed, some in Congress wanted to let section 7 stand alone.[203] Chief among them was the statute’s sponsor, Senator Robert Wagner.[204] Wagner told various Senate committees that section 7 implicitly required employers to bargain in good faith.[205] But the broader Senate rejected that interpretation.[206] Several senators observed that, without a clear textual hook in section 7, employers might have no enforceable duties.[207] So they added section 8(5), and later section 8(d), to spell the duty out explicitly.[208]

No similar additions were made in Missouri or New Jersey. The drafters gave employees the right to bargain collectively, but they said nothing about an employer’s duty. Had they wanted to create such a duty, they could easily have done so. They only had to look at what the U.S. Senate had done with section 8(5). But they chose not to do that. Instead, they adopted language mirroring section 7—and only section 7.

Sound constitutional interpretation should treat that choice as meaningful.[209] As Judge Fisher observed in his Chesterfield dissent, courts have no power to amend constitutional text.[210] They cannot add words simply because they think the text would work better with a little embellishment. They must apply the text as written. And as written, sections 19 and 29 impose no duty on employers.[211]

B. History

If the mandatory-bargaining approach makes little sense as a matter of text, it makes even less sense as a matter of history. To understand why, we have to start with the common-law baseline. At common law, a party had no duty to bargain.[212] She was free to bargain or not bargain with whomever she chose.[213] The concept of “good faith” came into play in only limited circumstances. A party could not entice another person into a contract through fraud, nor could she deny a contract when she had induced the other party to rely on her representations. But outside those situations, no “good faith” obligation attached. No one had to negotiate with a genuine desire to make an agreement. If a party didn’t want an agreement, she could walk away—or never bargain to begin with.[214]

These same principles applied to bargaining between employers and unions.[215] The common law recognized no duty for either party to bargain, much less bargain in good faith.[216] Instead, the law respected an employee’s right to choose a bargaining representative, typically a union.[217] It then left the union with the normal tools for extracting an agreement. The union could approach the employer through persuasion, protests, or displays of economic strength.[218] It could use skill or tact or weight of numbers.[219] But it could not sue to force the employer to the table.[220] Likewise, the employer could respond with its own economic weapons, such as hiring replacements. Or it could take a less drastic approach and bargain voluntarily.[221] But it had no legal right to dictate the union’s bargaining behavior.

The NLRA, of course, changed all that.[222] It deliberately departed from the common law and imposed a duty to bargain on both parties.[223] That mandate marked a sea-change in the law of labor relations, and it still colors our views of bargaining today.[224] But in the 1940s, when their constitutions were drafted and passed, neither Missouri nor New Jersey had anything like the NLRA. They had no general statute imposing a duty to bargain.[225] So for them, the relevant baseline was still the common law.[226]

The constitutional drafters had no intent to upset that baseline. In the contemporary debates, supporters of both provisions assured their convention colleagues that they were creating no new rights.[227] They did not mean to give any special privileges to labor.[228] Instead, they were trying to do only two things: (1) recognize employees’ preexisting right to bargain collectively through their chosen representatives; and (2) protect that right from legislative and judicial erosion.[229]

In New Jersey, this position was articulated by multiple delegates. Frank Eggers, the mayor of Jersey City, insisted that section 19 gave nothing “special” to unions.[230] It was “merely declarative” of employees’ “inherent right” to bargain collectively.[231] Likewise, Spencer Miller, an NYU professor of industrial relations, said that section 19 would protect employees’ rights to “unite” and “exert influence” on their employers.[232] It would allow them to “withhold[] their labor of economic value” and force the employers to “pay them what they thought it was worth.”[233] Such a right would not be new—it had already existed for “nearly a hundred years.”[234]

Reacting to these arguments, some delegates questioned section 19’s utility. If it would create no new rights, why put it in the constitution?[235] In response, supporters pointed to hostile legislators.[236] True, the supporters said, the right to bargain already existed; it was inherent and longstanding.[237] But that hadn’t stopped it from coming under constant attack.[238] Lawmakers—federal and state—had shown they could not be trusted to leave bargaining rights untouched.[239] Section 19 would forever put bargaining rights beyond their reach. It was intended, in short, to prevent legislative backsliding.[240]

The same position was taken by supporters in Missouri. Section 29’s chief advocate, R.T. Wood, explained that by enshrining bargaining rights in the constitution, labor was not gaining anything new.[241] It already had the right to bargain collectively. Section 29 would merely recognize labor’s rights and elevate them to constitutional status.[242]

As in New Jersey, skeptics pounced on this line of reasoning. If employees already had the right to bargain collectively, the skeptics asked, what would section 29 do?[243] Why put it in the constitution at all? In response, Wood gave a now familiar answer:

Mr. Darmon: Mr. Wood, does the labor [sic] now have the right to organize and bargain collectively in this state?

Mr. Wood (of Greene): Oh yes. Everyone knows that. That’s true.

Mr. Darmon: What rights or powers if any, do you seek to obtain by this proposal that you do not now have?

Mr. Wood (of Greene): Well, Mr. Darmon, if it is in our state constitution we will preclude the possibility as has happened in the past, in future sessions of the Legislature, many bills being introduced seeking to destroy collective bargaining.[244]

This exchange shows that section 29 was supposed to protect existing bargaining rights from legislative sabotage.[245] It created no new rights, but rather “recognized” that “the members of organized labor [had] the same right to organize and bargain collectively in [their] own interest as every other organization and every other group.”[246]

On the flip side, neither Wood nor any other supporter said that section 29 would force employers to bargain.[247] And if they had meant to force employers to bargain—a dramatic departure from the common-law baseline—one would have expected them at least mention it.[248] That no one did is a strong indication that we should take them at their word: they really did mean to preserve the status quo.[249]

It may come as a surprise that the drafters saw their handiwork in such modest terms. After all, they were inscribing rights into their fundamental organizing documents. Surely, they meant those rights to be meaningful. But their accomplishments seem modest only if we view them through a modern lens. We now have nearly a century’s worth of experience with the NLRA; given that experience, we take bargaining rights for granted. The duty to bargain is baked into modern labor law. It is as natural to us as the common-law standard was to our predecessors. But to convention delegates in the mid-1940s, bargaining rights seemed far less secure. The delegates had only a few years’ experience with the NLRA, and they could still remember a time when bargaining rights seemed quite precarious. Indeed, it was not so long before that labor unions faced threats to their very existence.

The most acute threat had come from the aggressive use of labor injunctions.[250] The injunction was an ancient remedy with deep roots in the common law.[251] Courts originally used it to prevent irreparable injuries to land.[252] But they gradually expanded it over time to cover other kinds of property, including business interests.[253] In the late 1800s, some courts started using it to tamp down labor disputes.[254] They reasoned that labor disputes posed serious threats to business interests, and so could be enjoined when paired with an “unlawful” motive—e.g., the intent to harm another person’s property.[255] With an injunction came potentially heavy penalties, including criminal contempt.[256] If labor leaders violated one, they could go to jail.[257]

Though rare at first, labor injunctions expanded over the century.[258] And more and more, they were used to shut down union activities.[259] They became particularly effective when the U.S. Supreme Court coupled them with antitrust law. In a series of decisions, the Court permitted legal attacks against unions under section 1 of the Sherman Act.[260] Under that act,[261] any group formed to restrain trade was an illegal combination.[262] Unions, of course, existed to control a particular kind of trade: the sale of labor.[263] So like other combinations in restraint of trade, they faced potential dissolution.[264] Antitrust law had thus become a dagger aimed squarely at their collective heart.[265]

Nor were labor injunctions the only threat facing unions. Another prime example was the “yellow dog” contract.[266] Yellow-dog contracts were common early in the 20th century. They required workers to agree, as a condition of employment, not to join a union.[267] They presented obvious obstacles to unionization, and they were a key union-avoidance tool for much of the early labor movement.[268] The constitutional drafters knew them well and cited them as a threat to bargaining rights.[269] They clearly expected sections 19 and 29 to abolish them.[270]

Unions eventually beat back these threats with the help of Congress. In 1914, Congress passed the Clayton Act,[271] which declared that labor unions were not illegal combinations in restraint of trade.[272] In 1932, Congress passed the Norris–LaGuardia Act,[273] which both stripped federal courts of much of their power to enjoin labor disputes and outlawed yellow-dog contracts.[274] These laws effectively neutralized the threat of labor injunctions in federal court.[275] They also spurred states to pass copycat laws, which put state courts on the sidelines as well.[276]

The Norris-LaGuardia Act took effect about fifteen years before Missouri and New Jersey convened their constitutional conventions. But in both states, the drafters could still remember the bad old days. They cited this history of judicial hostility as a reason for elevating bargaining rights to constitutional status.[277] They knew what it was like to feel the full force of judicial pressure and to wonder whether unions were lawful at all.[278] So to them, protecting the right to form unions and demand recognition would not have seemed so modest an accomplishment. It would have seemed instead like an epochal moment in the history of labor.[279]

C. Structure

The duty-to-bargain approach necessarily imbeds judges in overseeing day-to-day bargaining, in part because the relevant texts are so sparse. Neither section 19 nor section 29 mentions a duty to bargain, much less says how to implement one. So when courts declare that a duty exists, they have no choice but to define it as well.[280] And that kind of definitional work pulls them far outside their comfort zones.[281]

It’s one thing for courts to protect the right to engage in specified conduct. They have a lot of experience doing that. Every year, they handle thousands of claims under federal and state antidiscrimination laws.[282] Among other things, those laws protect employees who oppose discrimination.[283] Similarly, courts adjudicate claims under laws protecting free speech in the workplace and whistleblowing activity.[284] Courts have well-worn tools for adjudicating these claims and safeguarding protected conduct. They can muster a range of remedies, including reinstatement, backpay, and attorneys’ fees.[285] They are comfortable with these remedies and more than competent to deploy them.[286]

But courts have no similar toolkit to enforce an affirmative bargaining duty.[287] To start, they can’t award any remedy until they figure out what the duty entails. What subjects does it cover?[288] Where and when does bargaining take place?[289] With whom does the employer bargain?[290] Can the employer declare an impasse at some point, or does it have to bargain in perpetuity?[291] And if the parties reach an agreement, does the agreement satisfy the duty? Or do the parties have to keep bargaining even after they sign a contract?[292]

The problems don’t stop there. Once courts sketch out the duty’s contours, they still have to police it.[293] And that requires them to resolve yet another set of issues. For example, can an employer demand proof that the union represents the employees? If so, what kind of proof is enough?[294] Can the union use signed authorization cards? Or does it have to ask for an election?[295] If it needs an election, who supervises the voting? Who is eligible to vote? Who counts the ballots?[296] What should the court do about misconduct during a campaign? What happens if someone intimidates voters? Is the election voidable, or is it void? And more fundamentally, where is the line between intimidation and good-old-fashioned hard campaigning?[297]

Even if a court can work all these issues out, it still has to circle back to the remedy question. And there, it will find no clear answers. How does one remedy a failure to bargain? Yes, the court could order the employer to back to the table. But that’s no better than ordering the employer to follow the law. It creates no real penalties for noncompliance, and so leaves the bargaining duty with no teeth.[298] So the court could instead award damages. But how does it measure damages in this context? What loss does the union suffer from a refusal? The union possibly suffers a delay in getting an agreement; but the costs of delay are hard to measure. The court has to reconstruct a counterfactual scenario with multiple variables; its conclusion is almost inherently speculative.[299] And in any event, no one is entitled to an agreement. The duty to bargain in good faith does not include the duty to accept any particular proposal, or to reach an agreement at all.[300] So there’s no basis for awarding damages in the expectation that the union would eventually get an agreement. Taken together, these problems make the remedy hard to pin down. At the federal level, the NLRB has struggled with remedies for decades.[301] State courts are unlikely to find them any easier.

On the bright side, the exact answer is somewhat irrelevant. As with all these questions, the answers are less about abstract principles than they are about arbitrary line-drawing. It’s more important that the questions are settled than that they’re settled right.[302] But that doesn’t mean courts can just make the answers up. Courts aren’t supposed to be arbitrary line drawers.[303] They’re supposed to apply abstract legal principles in concrete cases.[304] And while they of course develop some rules through their decisions, those rules grow over time, case by case, through the slow accretion of precedent.[305]

That’s a hardly the best way to create a bargaining system. In an ideal world, the system’s rules would be announced in advance.[306] The parties would review the rules and conform their behavior accordingly.[307] But court decisions don’t work like that. They announce rules only after the parties have fallen into dispute. And disputes by definition involve contestable questions. So parties often won’t know they’ve violated a rule until a court tells them they have.[308]

There is, of course, a way to lay down rules ahead of time: legislation. Unlike courts, legislators can address issues comprehensively and in advance. Over a period of months or years, they can study problems, hold hearings, solicit public input, and craft solutions. They can carve out exceptions and assign administrators to oversee new processes. They can allocate resources to make sure the job gets done. And even better, when their solutions don’t work, they can try different ones. Unlike courts, they owe no deference to doctrinal consistency.[309]

But when courts constitutionalize an issue, they take it out of legislators’ hands.[310] They prevent the kind of advance line-drawing you need to make a bargaining system work.[311] They replace political compromise with judicial artifice—an artifice constructed on the fly, with no foundation in the democratic process.[312]

Courts in New Jersey and Missouri, lacking any legislative guidance, looked to federal law.[313] Federal law was the obvious gap-filler; over the decades, the NLRA had collected an immense body of caselaw.[314] Federal courts and the NLRB had decided most, if not all, of the issues state courts were likely to face.[315] So federal law offered state courts a ready-made repository of solutions.[316] State courts could use federal law to reduce their decisional loads and, at the same time, keep their caselaw consistent with national norms.[317]

But those benefits came with two major drawbacks. First, federal law could not answer the threshold question: whether employers had a duty to bargain. In fact, to the extent federal law even suggested an answer, it seemed to be no.[318]

The second downside was a loss of state autonomy. By importing federal precedent, state courts sacrificed the independence of state law.[319] The genius of the American system is that we have fifty-one different sovereigns.[320] Each sovereign is free to experiment in its own sphere and develop its own solutions.[321] That flexibility spurs competition and innovation. When states try different things, they sometimes land on good policies. Those policies then get picked up and spread through the marketplace of ideas.[322] But states can’t generate new ideas if they interpret their laws in lockstep with federal law.[323] If they merely parrot federal principles, they make themselves junior partners in what is supposed to be a system of co-sovereigns.[324] They abdicate their duty to develop state law as an independent source of rights and protections.[325]

In fact, federal law itself recognizes the value of state independence in this field. When Congress wrote the NLRA, it carved out large swaths of the American workforce.[326] It excluded agricultural and domestic workers because it thought federal bargaining standards were too onerous for their employers, who tended to be small enterprises or even individual people.[327] Likewise, it excluded public employees because it didn’t want to interfere with state law, which usually denied employees the right to strike.[328] Those carveouts would have been meaningless if Congress had wanted states to simply copy the federal framework.[329] Yet by importing federal standards, Missouri’s and New Jersey’s courts did just that. They effectively erased the NLRA’s carveouts and nullified Congress’s judgment.[330]

Admittedly, none of these downsides would matter if the Missouri and New Jersey constitutions explicitly created a duty to bargain. Had the drafters written a duty into sections 19 or 29, courts would have had no choice but to enforce that language.[331] But the drafters didn’t do that. Instead, they left bargaining rights open-ended and undefined. Courts could have read the language modestly and left space for legislative solutions, as New York’s courts did.[332] Or they could have read the text broadly, expanding bargaining rights to encompass unwritten duties as well.[333]

Missouri’s and New Jersey’s courts chose the second path. They thought it was the only way to protect bargaining rights.[334] But as we’ve already seen, that line of reasoning was too facile. Even if bargaining rights meant only the right to demand bargaining, they would still be meaningful. They would still embody a “no interference” principle, which would include protection from retaliation.[335] We can see such a principle at work in Herzog and Quinn, where courts respected employees’ right to bargain collectively without also imposing a duty to bargain.[336]

But rather than adopt a no-interference principle, New Jersey and Missouri instead chose a “duty to bargain” principle.[337] Courts in those states chose the latter principle, they said, because it was the only way to protect employees’ rights.[338] But ironically, that approach led them to weaken employees’ rights. It made them overvalue group rights and undervalue individual ones. And it left individual employees with fewer rights than when they started. We now turn to that consequence.

IV. The Consequences of Court-Mandated Bargaining

Labor disputes often devolve into fights between unions and management. In the popular mind, these two sides are the yin and yang of labor relations. The history of the labor movement can be told as a Manichean tug of war between them.[339] But what gets lost in the telling are the very people whose rights are at stake: individual employees. After all, labor law doesn’t exist to protect unions or management; it exists to protect workers.[340] It is the choices of workers, not management or unions, that labor law is supposed to respect.[341]

When imposing a constitutional duty to bargain, courts paid lip service to that principle. They said that bargaining rights belonged to employees; and for those rights to mean anything, the employer must have a duty to bargain in return.[342] To them, a bargaining duty was the only way to respect employees’ rights.[343]

But that conclusion led them to another one—one far less solicitous of individual employees. In both Missouri and New Jersey, courts concluded that for a bargaining duty to work, the employees must have only one representative.[344] In other words, courts adopted the principle of exclusive representation. They reasoned that without exclusivity, the employer might have to bargain with multiple representatives in the same workplace.[345] And with multiple representatives, mandatory bargaining would be little better than industrial anarchy.

That conclusion made some intuitive sense. It’s easy to see how multiple representation would devolve into chaos. Imagine that two unions represent employees on the same assembly line. Let’s even say the employees work on the same crew. One union wants to eliminate overtime. Its members want to work less because they value leisure time or time with their families. But the other union wants to expand overtime. Its members put more value on premium rates and higher take-home pay. These competing demands put the employer in a bind. The demands are mutually exclusive: the employer cannot run the line with half a crew. So if it mandates overtime for some, it has to mandate overtime for all. It therefore has to make a choice. But if it chooses to accept one union’s demand, it still has to bargain with the other.[346] And how can it bargain in good faith with the second union when it has already committed itself to the first? Isn’t bargaining with the second union futile? The employer might discuss the second union’s demands, but it cannot accept them. So it has no real expectation of an agreement. And doesn’t that kind of pro forma bargaining violate the duty?[347]

Exclusive representation solves this problem.[348] It gives one union the right to represent employees in the bargaining unit.[349] The union only has to gain support from more than half of the employees.[350] From there, it represents all the employees, not just the ones who support it.[351] The employer must bargain with this union, but it has no duty to bargain with other representatives; in fact, it cannot do so.[352] Bargaining with other representatives—or the employees themselves—would violate the good-faith requirement.[353]

But therein lies a new problem. Sections 19 and 29 give each employee the right to bargain through a chosen representative.[354] Exclusive representation respects that right for some employees, but not for others.[355] If a bare majority of employees selects a union, that union bargains for everyone.[356] Any employee who wants a different representative—or no representative at all—is stuck. She has a representative foisted upon her. For her, the right to select a representative is less than meaningless; it is reversed. She cannot even bargain for herself. Her rights are subordinated to the group’s preference.[357]

She is also at much greater risk of coercion. Consider the process for gaining majority status. Neither Missouri nor New Jersey has a statute covering bargaining for private-sector employees. So there is no clear procedure for establishing majority status.[358] A union might gain that status by winning an election.[359] But it might also do it by collecting authorization cards.[360] Employees sign these cards in private, but not anonymously: the union knows who signs the cards because it collects them in person.[361] And these in-person interactions come with a lot of pressure and potential confusion.[362] When approached, an employee may feel social pressure to sign a card, even if she has no desire to join. She may be told that others have signed, and she will be the odd one out if she refuses. She may want to feel like a team player. Or she may sign only to get the organizer to leave her alone. Even worse, she may sign under the mistaken impression that eventually, she will get a chance to cast a secret ballot. But in fact, she may have no opportunity to vote, much less to revoke her card if she changes her mind.[363]

What’s more, she may be surprised when, months later, she finds the union suddenly ensconced as her representative. The union does not have to set a deadline for presenting cards to the employer.[364] It can collect cards over weeks, months, or even years. It can then ambush the employer with cards from 51% of the workforce. Its support may have risen and fallen over the signature-collecting drive. It may never have enjoyed support from more than half the workforce at any one time. But as long as it has those cards, it’s locked in.[365] And the 49% of employees who never signed cards? They never had their voices heard, much less their wishes respected.[366]

By inventing a bargaining duty, courts were forced to adopt exclusive representation. And by adopting exclusive representation, they trampled on individual rights. They took a right given to each employee and subordinated it to popular rule.[367] That rule may work for employees in the majority, who get to impose their preferences on their coworkers. And it may likewise work for employers, who don’t have to deal with a cacophony of divergent demands. But it does nothing for employees in the minority. Those employees not only lose their right to choose a representative, but also the right to bargain for themselves. Their constitutional rights have been reversed.

V. Conclusion: A Path Forward?

Interpreting broad constitutional language often requires a degree of judgment. Answers rarely come with mathematical precision.[368] But while the edges may blur, we can identify guideposts to help us reach better results. We can produce faithful interpretations if we focus on the things we know.[369]

And when it comes to bargaining rights, we do know three things. First, we know that courts have interpreted bargaining language two ways: the no-interference approach and the duty-to-bargain approach. Second, we know that the latter approach was based on a policy judgment about the best way to protect bargaining rights. And third, we know that this interpretation contradicted text, history, and constitutional structure.

What we don’t know is whether there’s any way to move from the second approach to the first. Having inserted themselves into everyday bargaining, can courts find their way out? Having read rights maximally, can they revert to a more traditional position?

The answer is probably yes, but it will be difficult. Rights are like entitlement programs: once extended, they are hard to roll back.[370] The obstacles to a rollback are likely easier to overcome in Missouri, where courts at least have a history of reading bargaining rights modestly. Their turn toward a maximal interpretation is relatively recent, and so they can frame their return to the traditional position as a reversion to historical norms. But in New Jersey, courts have taken the maximal approach since the very beginning, so they cannot revert to a previous position. Courts, of course, are creatures of precedent: they like nothing less than undisguised innovation. So any reversal of bargaining duties would likely come through the people themselves—either with a new constitution or a political sea change. Neither path seems likely, but neither do they seem impossible. For either, the first step will be to bring attention to the problem. We have to see the wrinkles in our doctrine before we can start ironing them out.

That has been the modest goal of this article. Its aim has been to shed light on a much-overlooked corner of the law—one that could use a bit of sunshine. While most lawyers think of labor law in strictly federal terms, for millions of workers, the only source of rights is state law. So when state courts get things wrong, their errors matter in the real world for real people. Real people can lose their rights, including their right to choose a bargaining representative. The effect is the same whether the loss stems from an error in federal court or one in state court. The loss matters just as much—for workers, for courts, and for the rational development of the law.

Note from the Editor: The Federalist Society takes no positions on particular legal and public policy matters. Any expressions of opinion are those of the author. We welcome responses to the views presented here. To join the debate, please email us at [email protected].