Janus and the Future of Unions
Barbara Nachtrieb Armstrong Professor of Law, University of California at Berkeley School of Law
For nearly 100 years, union relationships with employers and with the workers whom the union represents have operated on the model of electoral democracy. A union, chosen by a majority in an election, represents all workers in the unit, just as a legislative or executive official represents everyone. But unlike a political leader, a union owes a duty of fair representation to every employee in the unit and cannot act arbitrarily or discriminatorily in deciding whose interests to prioritize. And in enforcing a contract, the union must treat all workers—union members and nonmembers alike—adequately and without discrimination. Democracy is foundational to everything unions do, from the way they govern their internal affairs to their efforts on behalf of workers to create workplace democracy to their role in civil society. Their responsibility to respect the interests and rights of minorities is what makes unions different from political leaders and what has made the contemporary fight over how unions fund their work so significant.
Like other large groups, union employees face a collective action problem; indeed, theirs is the paradigmatic collective action problem—the one used by economists to explain the theory of collective action. It is in the interest of every member of the large group to engage in collective action to improve wages and working conditions. It is equally in the interest of every member of the group to let others incur the costs of engaging in the collective action. But if every person acts rationally as an individual in free-riding on the efforts of other, all will be worse off because no one will get the benefit of collective action. Union security provisions were the ingenious contractual solution to this collective action problem: requiring everyone to support the collective representative prevents the individually rational decision to free-ride, thus promoting the economically optimal collective action.
Courts long ago prevented unions from solving their collective action problem by requiring workers to actually join the union. Rather, the most that unions could do is to require represented workers to share in the cost through payment of what was known as an “agency” or “fair share” fee. In Janus v. American Federation of State, County and Municipal Employees Council 31, the Court declared unconstitutional fair-share fee provisions in the labor laws and union contracts of twenty-two states, the District of Columbia, and Puerto Rico. The case is the latest, though probably not the last, in a series of cases in which the Roberts Court, always split on ideological 5-4 lines and always with the majority opinion written by Justice Alito, continued its attack on public employee unions. Finding the contractual solution to the collective action problem to violate the First Amendment, the Supreme Court continued its use of the First Amendment to invalidate well-established regulations of economic and social legislation.
Part I of this Article describes the legal background to Janus. Part II explores the reasoning of the majority and its possible implications for labor law. Part III explores the conceptualizations of complicity and compulsion embraced or at least hinted at in Janus and their troubling implications for an array of regulations. Part IV describes the efforts of anti-union litigation and lobbying to extend their win to other areas of labor law. Part V describes legislative efforts to provide for union security after Janus.
I. Union Security Before Janus
Republican Bruce Rauner was elected governor of Illinois on a pledge to destroy public sector unions. Rather than negotiate with the state’s public employee unions to address the state’s budget issues, or work with the Illinois legislature to repeal laws he considered an obstacle to that goal, Governor Rauner filed a federal suit to get fair share fees declared unconstitutional. That says quite a bit about Janus—it was from the start an effort to use the federal courts in a political fight that Governor Rauner felt he could not win on his own.
The state laws and contracts that Governor Rauner asked the federal courts to invalidate were settled principles of federal, state, and local labor law. For well over a hundred years, labor unions have sought contract terms that require all employees represented by the union to either join the union or pay their fair share of the costs the union incurs in negotiating and enforcing a labor contract. Administering a fair personnel process is expensive. In a unionized workplace, employees have some say in the process, unlike nonunion employees. But if workers are partly responsible for HR, the union must raise the money to support it. Unions are, in this respect, just like governments. A city or state requires every resident to pay taxes to support schools, parks, police, firefighters, and prisons. Some people deeply oppose policing and prisons, or don’t use schools or parks, but they pay taxes to support them. These are what economists call common goods, and an elementary principle of economics is that no economically rational person will voluntarily choose to pay for them so long as others pay to support them.
Before 1947, federal law allowed unions and employers to require employees to pay dues or fees to a union, as well as contractual terms requiring employees to be union members at the time of hire, not merely to pay fees to it. These were known as closed shops. An organization was founded by business groups to combat these closed shops, arguing that they violated employees’ “right to work.” In 1947, the right to work group scored a big legislative win, as Congress amended § 8(a)(3) of the National Labor Relations Act (NLRA) to prohibit these “closed shop” agreements, instead allowing employers and unions to agree only that employees must become union members within thirty days of hire (a “union shop” agreement). In addition, Congress added a new § 14(b) to the statute, which allowed states to choose whether to allow contracts with fair share fees, though it’s ultimately up to each union and agency to decide whether to require them.
When the Supreme Court in the 1960s expanded individual freedoms of speech and association through new interpretations of the First Amendment, the Court further restricted the ability of unions and employers to require employees to join unions. It did so both as a matter of interpretation of the NLRA and under the First Amendment.
As to the NLRA, it adopted a broad reading of what was prohibited by § 8(a)(3)’s and § 14(b)’s restrictions on compulsory union “membership.” A strictly literal reading of these two sections would allow employers and unions to negotiate contracts (under § 8(a)(3)), or states to forbid such contracts (under § 14(b)), that require a worker actually to become a member of the union, but would not prohibit anything else. But in a pair of cases from 1963, the Supreme Court held that the sections’ definition of “membership” is broader. In NLRB v. General Motors, the union had proposed a contract provision that did not require workers to join the union but only to pay a fee for the union’s representation services (an “agency fee”). The employer insisted that the only form of union security device permissible under the NLRA was one requiring workers actually to join the union; it was all or nothing. The Court rejected the employer’s argument and held that the Taft-Hartley Act changed the “meaning of ‘membership’ for the purposes of union security contracts” so that § 8(a)(3) “‘membership’ as a condition of employment is whittled down to its financial core.” That is, workers could be required to pay full dues, but they could not be required to join. In the same year, in Retail Clerks v. Schermerhorn, the Court extended the General Motors narrowing of “membership” to the preemptive scope of § 14(b).
But Schermerhorn did not hold that states may forbid contract provisions that require less than full membership. In fact, it suggested that states could not forbid such provisions. The union in Schermerhorn had argued that its agreement was distinguishable from the agreement in General Motors because it confined the use of nonmember dues to collective bargaining, rather than other union institutional goals (such as political activity). But the Court explained that the union’s contract did not limit how nonmember payments could be used. And, the Court pointed out, the union charged nonmembers the same as members, which suggested that the union would use nonmember fees to fund union activities other than contract negotiation and enforcement. This matters, because if states could ban any compulsory nonmember fees, it would have been unnecessary for the Court to emphasize that the contract at issue in the case didn’t restrict the use of nonmember fees and charged nonmembers the same as members.
But that is not how lower courts and litigants have read Schermerhorn. Rather, it has been read to allow states to prohibit any payments from nonmembers. Now, twenty-eight states prohibit unions and employers from agreeing to require employees to pay any fees or dues to the union. In those states, the union is obligated to provide contract negotiation and administration services to nonmembers for free. The union is required to represent all workers in the bargaining unit equally, and may not discriminate between those who become union members and those who do not. The duty extends not just to collective bargaining—where the union cannot bargain terms that favor members over nonmembers—but to disciplinary matters as well. The union must grieve and arbitrate on behalf of nonmembers just as zealously (and, more to the point, as expensively) as they do on behalf of members. In non-right-to-work states, federal law enables unions to require that nonmembers pay for the services they receive. In right-to-work states, on the other hand, the union still bears the same federal duty to represent nonmembers, but state law precludes a requirement that the nonmembers pay for that representation.
Having held in General Motors that private sector employees cannot be required to join a union, for twenty-five years the law was settled that private sector employees could not be required to join a union. But they could be required to pay full dues, even if the union spent some portion of the revenue from dues on political activity, unless the employee worked in a state that had enacted a so-called right-to-work law prohibiting any payment of fees. But in Communications Workers v. Beck, the Supreme Court held that § 8(a)(3) permits a collective bargaining agreement to require nonmembers to pay fees only to support the union’s contract negotiation and enforcement functions, and not to support the union’s political operations. Thus, the most that a union and employer can require of an objecting employee is to pay an “agency” or “fair share” fee representing the employee’s fair share of the union’s costs of services “germane” to the union’s role as bargaining agent. The employee cannot be required to subsidize “political” expenditures. This “fair share” fee arrangement had its origin in two strands of cases involving railway and then public sector unions.
The first strand began with Railway Department Employees v. Hanson. In Hanson, the National Right to Work Committee argued that the Railway Labor Act (RLA) was unconstitutional because it preempted a Nebraska right-to-work law and therefore compelled employees to support unions. The Supreme Court made two significant holdings. First, the Court agreed with the plaintiffs that the RLA’s preemption of state right-to-work laws was sufficient state action to subject the union security agreement between the private union and the private railroad to constitutional scrutiny. Second, the Court held that the compelled subsidization of the employee’s exclusive bargaining representative did not violate the employees’ First Amendment rights. But the Court reserved the lower-court ruling that the expenditure of those employees’ dues or fees over their objection on political candidates violated their First Amendment rights.
The Court reached the issue it reserved in Hanson in International Association of Machinists v. Street. To avoid the question whether compulsory dues violated the First Amendment, the Court interpreted the RLA not to require dissenting employees to provide financial support for union political speech. Noting the importance of allowing self-governance of work on the railroads, the Court deemed it important to avoid a situation in which nonunion members “share in the benefits derived from collective agreements negotiated by the railway labor unions but bear no share of the cost of obtaining such benefits.” Moreover, the Court recognized the right of the majority of workers and their union to engage in political activity, “without being silenced by the dissenters.” The Court thought a sensible compromise was to prohibit the expenditure of agency fees on political expression not germane to the union’s role as bargaining agent.
The second strand of cases began with Abood v. Detroit Board of Education, the Court’s first case about the constitutionality of public sector unions. The Court held that government employees have First Amendment rights to refuse to join the unions that represent them and to refuse to provide financial support to their unions’ political activities unrelated to the union’s duties in negotiating and enforcing the collective bargaining agreement.
The notion embraced in both Street and Abood that contractually required union membership involved compelled speech in violation of the First Amendment produced sharp dissent from some justices. As Justice Frankfurter (joined by Justice Harlan) stated in dissent in Street, “what is loosely called political activity of American trade unions . . . indissolubly relat[es] to the immediate economic and social concerns that are the raison d’etre of unions.” The dissent went on to note the many examples throughout the nineteenth and twentieth centuries through which labor unions achieved improved working conditions through “an extensive program of political demands calling for compulsory education, an eight-hour day, employer tort liability, and other social reforms.” Justices Frankfurter and Harlan insisted that the use of union dues to support political activity did not constitute compelled speech.
Plaintiffs here are in no way subjected to such suppression of their true beliefs or sponsorship of views they do not hold. Nor are they forced to join a sham organization which does not participate in collective bargaining functions, but only serves as a conduit of funds for ideological propaganda. No one’s desire or power to speak his mind is checked or curbed. The individual member may express his views in any public or private forum as freely as he could before the union collected his dues.
The dissent also pointed out that payment of taxes, like payment of union dues, supports political speech “to propagandize ideas which many taxpayers oppose. . . . It is a commonplace of all organizations that a minority of a legally recognized group may at times see an organization’s funds used for promotion of ideas opposed by the minority.”
But that point of view was a dissent. Since 2012, the conservative majority of the Supreme Court twice expanded the protections for union dissenters, culminating in 2018 with overruling Abood.
II. The Janus Decision
The Janus majority, in an opinion by Justice Alito (joined by Chief Justice Roberts and Justices Kennedy, Thomas, and Gorsuch), held that the payments for representation services are speech protected by the First Amendment because the union uses the money to engage in speech (negotiating and administering a contract), and payment cannot be compelled. The case produced a stinging dissent by Justice Kagan, joined by Justices Ginsburg, Breyer, and Sotomayor. Each step of the Court’s reasoning has significant implications not only for the future labor law, but for the First Amendment more generally.
A. When Are Fees Compelled Speech?
The first step in the Janus reasoning is to equate paying a fee to compelling a statement of belief. The majority began with the proposition that “compelling individuals to mouth support for views they find objectionable violates [the] cardinal constitutional command” that (quoting the compulsory flag salute case) “no official, high or petty, can prescribe what shall be orthodox in politics, nationalism, religion, or other matters of opinion or force citizens to confess by word or act their faith therein.” Equating payment of a fee to forced confession of a belief is a bold and controversial move. All of us are compelled by law to pay money to entities that use it to engage in speech activities: taxes, homeowners association dues, health insurance premiums, pension plan contributions, licensing fees, public school and university student fees. The majority in Janus did not say anything about these other fees.
The implications of Janus will depend on how the Supreme Court and lower courts expand on the idea that compulsory payments are tantamount to compulsory professions of belief. Although Janus said nothing about other fees, in Harris v. Quinn, the same five-justice majority (who held unconstitutional fair share fees for unionized home health aides paid by Medicaid) said they thought that the government has a more compelling interest in requiring attorneys to pay state bar dues and public university students to pay activity fees. But it did not say why, which makes it hard to discern how the Court expects other fees to be analyzed. Harris said that bar dues “served the State’s interest in regulating the legal profession and improving the quality of legal services” and “[s]tates also have a strong interest in allocating to the members of the bar, rather than the general public, the expense of ensuring that attorneys adhere to ethical practices.” It did not say why the state’s interest in requiring lawyers to pay the cost of regulation is greater than its interest in requiring public employees to pay the cost regulation. As to university fees, Harris said that universities “have a compelling interest in promoting student expression in a manner that is viewpoint neutral.” That is not a principle that could explain other forms of compulsory fees, because there is no requirement that health insurers spend employee contributions in a viewpoint neutral manner. Harris also said that creating an opt-out regime for public-university student fees would create “administrative problems [that] would likely be insuperable.” It is unclear what principle is used to decide when administrative problems are “insuperable” or when that becomes the basis for upholding a compulsory fee system.
The failure to distinguish contrary authority is, of course, common in constitutional and common law adjudication. What’s troubling about this line of cases, though, is the fact that the only thing that distinguishes one case from another is whether the Court believes the government’s interest is compelling. If every regulation is subject to strict scrutiny, constitutional analysis in the end will turn on whether the Court’s majority thinks the government’s interest is compelling.
B. Misunderstanding the Collective Action and Free Rider Problem
The next step in the majority’s reasoning addressed the collective-action problem, which the majority incorrectly reduced simply to a free-rider problem involving fees for representational services. The collective-action problem is not simply whether an employee must pay for the cost of contract enforcement once a contract is negotiated. Rather, it is that without being able to make a credible commitment to stick with the group, a large group will be unable to engage in economically optimal collective action (such as achieving a union contract) in the first place.
The difference is simple to illustrate. When I first took a job at the University of California in 2008, I agreed to work in a non-union position for a salary. When the state’s budget crashed in 2009, the university unilaterally reduced my salary by ten percent, though none of my responsibilities changed. (To be clear, I’m not complaining about my salary: I was well paid and I loved my job.) It could do so because faculty had never joined together to negotiate collectively for enforceable contracts. Unionized employees, by contrast, had enforceable contracts, so the state could not unilaterally reduce their pay. The significance of unionization is not just that employees shouldn’t free ride on the union dues paid by their co-workers so that when bad things happen they have the union to handle their grievance. Rather, it’s that without unionizing, they never get the contractual protection in the first place. The Janus majority assumed that, even without union security, public employees would still have labor contracts and the only question is whether employees should have to pay to get union representation in enforcing the contract. It’s like health insurance: the majority assumed that the problem is whether an employee will get his medical bills paid by the insurance company after he gets sick. But without the union to bring everyone together, or without the ability to require employees to join a large group to form a risk pool, there would be no contract, no insurance policy, in the first place. So when the majority said that unions could solve the free rider problem by declining to represent nonpayers or by charging them for handling grievances, it misunderstood that, without the union’s ability to require everyone to support the union, nobody gets the contractual protections.
The benefits of collective action accrue to the employer, too. California teachers before 1976 did not have a majority representative system. Rather, teachers could choose the union they wanted, or no union at all, and school districts had limited bargaining obligations with each different union, depending on the size of the membership. School districts had multiple different contracts with different groups of teachers. And, not surprisingly, they hated it. The system was cumbersome, confusing, and expensive to administer. It generated enmity among teachers at the same school. The Court recognized all this in Abood in explaining why union security is constitutionally required:
designation of a single representative avoids the confusion that would result from attempting to enforce two or more agreements specifying different terms and conditions of employment. It prevents inter-union rivalries from creating dissension within the work force and eliminating the advantages to the employee of collectivization. It also frees the employer from the possibility of facing conflicting demands from different unions, and permits the employer and a single union to reach agreements and settlements that are not subject to attack from rival labor organizations.
Union critics seem to assume that the alternative to majority unions is either no union or a plethora of weaker unions and that will strengthen the ability of employers to set policy, to fire poor workers, to lower salaries and benefits costs, and to avoid periodic strikes. That is true only in times of labor shortage. As everyone witnessed in the spring of 2018, weak unions did not lead to better schools, better teachers, or better educational outcomes in West Virginia, Oklahoma, and the four other states that experienced massive strikes.
C. The First Amendment and Public Employee Speech
The third fundamental point in the Janus opinion concerns the speech rights of public employees. Under 50-year-old Supreme Court precedent, public employees have relatively few First Amendment rights. In particular, they have no First Amendment right to speak to their supervisor about their working conditions. In Garcetti v. Ceballos, the Court upheld the discipline of an assistant district attorney who raised concerns with his supervisor about false police testimony that could lead to a wrongful conviction. That speech, the Court held, is not protected by the First Amendment no matter how important the issue, because it was speech about the DA’s work. On matters outside of their job duties and in forums outside the workplace, the Court held in Pickering v. Board of Education that public employees have a right to speak out as citizens, but only on matters of public concern, and only if their speech does not disrupt the government’s interests as employer.
Janus dismisses this area of law. First, explained the majority, Pickering and Garcetti are a different strand of First Amendment doctrine, and “[w]e see no good reason, at this late date, to try to shoehorn Abood into the Pickering framework.” The notion seems to be that one strand of cases holding that employees have few speech rights need not be consistent with another. That is not convincing; there aren’t multiple First Amendments. All of these cases concern public employees’ free speech rights, and the Court needs to explain why employees can get fired for complaining about their work, but not for refusing to pay fair share fees.
The second reason for dismissing Pickering and Garcetti was that they concern the rights of individual employees, whereas fair share fees are “a blanket requirement,” and a “speech-restrictive law with widespread impact . . . gives rise to far more serious concerns than could any single supervisory decision.” That is not the law and never has been. Nor was it the facts of Pickering, Garcetti, or Janus. In Garcetti, for example, the Los Angeles County District Attorney justified the discipline of Richard Ceballos in terms of policy: his supervisors had decided not to dismiss the case that Garcetti thought should be dismissed, and his defiance of their judgment was the basis for discipline. Similarly, when the Court rejected a First Amendment challenge to the abortion counseling gag rule in Rust v. Sullivan, it upheld a blanket policy (prohibiting any recipient of Title X family planning funds from providing information about abortion).
Moreover, finding greater First Amendment protection for speech compelled by a blanket policy than for the speech of an individual isn’t even the law that the Court itself applied in a First Amendment case handed down the day before Janus. In National Institute of Family and Life Advocates v. Becerra, the same conservative five justices held that the First Amendment invalidated California laws requiring women’s health clinics to give accurate information. This involved a blanket policy (a California state law) regulating the speech of all health providers, not individual restrictions. Similarly, in Masterpiece Cakeshop, Ltd. v. Colorado Civil Rights Commission, a bakery argued that a state law prohibiting discrimination in places of public accommodation violated its alleged free-speech right to refuse to serve gay couples. The Court did not decide the First Amendment issue, sending the case back to the lower court to reconsider, but the Court never suggested that the First Amendment wouldn’t apply because the law affects all businesses rather than just Masterpiece. First Amendment rights don’t get stronger when it is one person rather than many whose speech is restricted.
The Janus majority also distinguished Pickering and Garcetti on the ground that they were cases in which employees were prevented from speaking, not compelled to speak. Outside a situation involving speech as part of an employee’s official duties, the majority said, “it is not easy to imagine a situation in which a public employer has a legitimate need to demand that its employees recite words with which they disagree.” Leaving aside the fact that nobody was forced to “recite words,” the First Amendment generally has not distinguished between compulsion and restriction of speech. Take the Court’s original compelled speech case, West Virginia Board of Education v. Barnette, which struck down discipline of students with a religious objection to reciting the Pledge of Allegiance: would it matter if a student were disciplined for refusing to recite the Pledge, or for reciting it but changing the words to avoid offending the student’s beliefs? Of course not. Moreover, the Court has upheld laws requiring speakers to “recite words.” As Justice Breyer pointed out in his dissent in NIFLA, in Planned Parenthood v. Casey the Court upheld a law requiring medical personnel providing abortions to inform patients about the nature of the procedure, the health risks of abortion and childbirth, and the “probable gestational age of the unborn child,” and to inform the patients of printed materials “describing the fetus, medical assistance for childbirth, potential child support, and the agencies that would provide adoption services (or other alternatives to abortion).”
This aspect of the Court’s reasoning in NIFLA raises a point the Janus Court never considered. It has never before held that there is a constitutional right to refuse to subsidize speech where the underlying speech is not protected by the First Amendment. The only basis on which Janus could resist paying fees to subsidize collective bargaining would be if the collective bargaining is speech that is protected by the constitution. One doubts that the Court would hold that a group of employees who demanded collective bargaining over terms of employment would be protected by the First Amendment. But why not? If subsidies for collective bargaining are protected, why on earth not the bargaining itself? Many states do have a right to bargain collectively in their state constitutions.
Perhaps the argument would be that compelled subsidies for collective bargaining are unconstitutional even though collective bargaining is not constitutionally protected speech, because public sector bargaining is bad. The attack on union security, both in Janus and in political critiques of unions and collective bargaining, is fueled, in part, by the sense that unions in both the public and private sectors are political organizations that should not be able to intervene in the policy-making process through collective bargaining. The critique is that union security gives unions too much power over the workers they represent and too much power over employers. Much of the Janus critique of union political activity focuses on the notion that unions do not reflect the interests of the employees they represent. Union supporters believe union security arrangements are democratically devised solutions to collective-action problems facing democratic organizations. Union critics see union security provisions, including fair-share fees, as perpetuating the power of more-or-less corrupt and oligarchic special interest groups that use money coerced from dissenting employees and taxpayers to pursue a political agenda closely tied to the values of the Democratic party and the unions’ leadership. The legitimacy of union security thus turns, in part, on whether laws and union rules effectively promote democratic self-governance within labor unions.
Justice Alito’s third way of reconciling the Court’s newly-invented First Amendment right with precedent was that some forms of political-patronage hiring are unconstitutional and, therefore, union fees should be unconstitutional too. Apart from the false syllogism, the reasoning falls apart because the Court has upheld many laws prohibiting political activity of government employees. It upheld the federal Hatch Act (which prohibits certain partisan political activity on federal government employees’ free time). It upheld state laws that prohibit patronage appointments in low-level jobs. The only anti-patronage law the Court has declared unconstitutional is a restriction on awarding policymaking jobs based on political affiliation. (There might be many reasons why Democrats wish Jeff Sessions were not the Attorney General, but one cannot argue that his early political support for Donald Trump makes it unconstitutional for Trump to choose him.)
III. Complicity, Compulsion, and Democracy
Although, as will be explained below, the impact of Janus is enormous for governments and their employees, those outside the labor field may think it will have little impact on other areas of free-speech law. But there are disturbing aspects to the Court’s reasoning in Janus and in its other compelled speech case, handed down the day before, NIFLA. Those may portend much more radical changes in how the Court considers the constitutionality of the what Justice Kagan in her Janus dissent: “Speech is everywhere – a part of every human activity (employment, health care, securities trading, you name it).”
First, the Court appears to have launched itself on its own form of viewpoint discrimination in how it handles speech restrictions. The day before the Court handed down its decision in Janus, the same five conservative justices decided that California lacks a compelling interest in requiring so-called crisis pregnancy centers to inform pregnant women that low-cost abortions are one available alternative. Yet, the Court did hold several years ago that states do have a compelling interest in requiring health care providers to inform patients seeking abortions about certain facts about fetuses and about certain debatable propositions about abortion. As Justice Breyer remarked in dissent, the Court essentially believes that the government has a compelling interest in forcing health care providers to try to talk women out of having abortions but no compelling interest in forcing them to identify abortion as one way of dealing with an unwanted pregnancy.
It upheld compelled speech warnings attempting to dissuade women from having abortions (including compelling health care providers to tell patients information that science considers misleading or wrong). But it struck down notices alerting women to the availability of abortions in NIFLA. If a legislature were to adopt the rule the Court has created, under the Court’s own precedents condemning viewpoint discrimination in law, the rule would be unconstitutional. Moreover, the Court has compelled speech in the labor area too. In Hudson, the Court invented a constitutional rule compelling unions to notify workers they represent of their right to free-ride on the fees paid by their co-workers. It’s worse than ironic that the Court engages in viewpoint discrimination while invalidating laws on the grounds of viewpoint discrimination.
The tension between Janus and NIFLA is troubling. Compelling a private organization to give a message to its members that is antithetical to the organization’s own deeply held values is precisely what Justice Thomas found objectionable in NIFLA. Yet, Janus dramatically expands public sector unions’ obligation to undermine their own mission by doing exactly that. If it is unconstitutional to require a pregnancy service provider “to inform women how they can obtain state-subsidized abortions—at the same time petitioners try to dissuade women from choosing that option” it should be equally unconstitutional to require unions to notify their members of their right to quit the union “at the same time [unions] try to dissuade [workers] from choosing that option.”
More generally, as many have observed, the Court has suddenly cast into constitutional doubt an enormous array of regulations of lawyers and a host of other service providers. Laws require sellers of goods and services routinely to warn prospective consumers about their goods and services because the government, sometimes controversially, deems certain uses of a product or service to be objectionable. Purveyors of alcoholic beverages are required to warn pregnant women of the dangers of drinking because the government condemns drinking while pregnant, even though some research suggests that modest consumption is not hazardous. Most of the law regulating lawyers operates as restrictions and compulsions on speech. When the majority in NIFLA explained that most professional speech is protected by the First Amendment, and the Court’s decision upholding certain restrictions on lawyer solicitation and advertising in Zauderer does not apply because “[t]he notice in no way relates to the services that licensed clinics provide,” it was ignoring most of the law of professional responsibility.
The tension between how Janus treats unions and how state rules of professional conduct treat lawyers extends beyond the bar-dues issue noted above. Lawyers are required to counsel clients about the wisdom of hiring another lawyer to give a second opinion on a business transaction between lawyer and client and on a retainer agreement that limits malpractice liability. Lawyers must make elaborate disclosures about contingency fees. Organizations receiving funding from the Legal Services Corporation must warn prospective clients that the lawyers will have to terminate the representation if the client loses her lawful immigration status or starts to reside with someone who has ever been convicted of certain crimes, even if the representation is not funded by the LSC. While the Court in NIFLA suggested that compulsory disclosure is acceptable when it takes the form of an informed consent law, these are not informed consent laws, they are disclosure laws. Some lawyers feel these warnings are irrelevant or inimical to their effective representation, just as NIFLA objects to the disclosure laws, but that has never made them unconstitutional. Justice Thomas’s notion that abortion is different because it is controversial (unlike, presumably, restrictions on lawyer speech) ignores the controversy surrounding many rules prohibiting or compelling lawyer speech. In some states, lawyers cannot report a client’s likely harmful conduct toward a third party; in some states lawyers may report; but in most states psychotherapists must report. These are highly controversial rules and they do not fit within the categories of informed consent and professional conduct that the NIFLA majority suggests are the only exceptions to First Amendment protection for professional speech.
The justices have snuck two especially baffling statements about general free speech rules into its compelled speech cases from June 2018. First, in Janus, the Court created confusion about the degree of constitutional scrutiny for compelled speech in the commercial area. The Janus Court said that Knox, a prior fair-share fee case, held that “exacting scrutiny” should be applied to such restrictions, even if they are commercial speech and not subject to strict scrutiny. The Janus majority continued:
Even though commercial speech has been thought to enjoy a lesser degree of protection, see, e.g., Central Hudson Gas & Elec. Corp. v. Public Serv. Comm’n of N.Y., 447 U.S. 557, 562-563 (1980), prior precedent in that area . . . had applied what we characterized as “exacting” scrutiny, Knox, 567 U.S. at 310, a less demanding test than the “strict” scrutiny that might be thought to apply outside the commercial sphere. Under “exacting” scrutiny, we noted, a compelled subsidy must “serve a compelling state interest that cannot be achieved through means significantly less restrictive of associational freedoms.” Ibid. (internal quotation marks and alterations omitted).
That passage makes no sense. Central Hudson applies intermediate scrutiny to commercial speech. But the “exacting scrutiny” standard that the Court articulates here is identical to strict scrutiny. The majority in Janus then continued:
[P]etitioner in the present case contends that the Illinois law at issue should be subjected to “strict scrutiny.” The dissent, on the other hand, proposes that we apply what amounts to rational-basis review, that is, that we ask only whether a government employer could reasonably believe that the exaction of agency fees serves its interests. This form of minimal scrutiny is foreign to our free-speech jurisprudence, and we reject it here. At the same time, we again find it unnecessary to decide the issue of strict scrutiny because the Illinois scheme cannot survive under even the more permissive standard applied in Knox and Harris.
The “more permissive standard” appears to be the same as strict scrutiny, and it is indeed fatal in the union dues cases. Is that something that the Court will pick up on next year or thereafter to say that it has already decided that regulation of commercial speech is no longer subject to intermediate scrutiny?
The second disturbing aspect of the holdings in Janus and NIFLA is the statement in NIFLA that the disclosure requirements are content regulation subject to strict scrutiny and invalid if they are under- or over-inclusive or if the interests they serve are not compelling or not real or if compliance with the disclosure rule is “unduly burdensome” or if the government could convey the required information itself. All disclosure rules are, by definition, content regulation. Most disclosure rules are under- or over-inclusive because some people will not need or want the disclosure, and some will require more disclosure. Compelled disclosure in English will not help those who do not understand English. And the government can always convey the message itself rather than require the private entity to convey it.
Moreover, compelled subsidies of others’ speech are ubiquitous. Drivers and employees are compelled to purchase liability and health insurance, and insurers use some part of that money for speech. Employees covered by pension plans subsidize speech of the financial-services companies that collect pension contributions. The Janus majority suggested that union fair-share fees become unconstitutional because, in the aggregate, the effects of public-sector bargaining are substantial for public policy. But of course the effect of insurance and financial-services industry lobbying (using the insureds’ compulsory payments) is equally substantial and equally controversial.
The conservative majority’s answer to all of these examples is that the government has a compelling or substantial interest in some restrictions but not in the ones it struck down. Justice Alito said exactly this in Harris v. Quinn in explaining why requiring lawyers to pay bar dues and public university students to pay activity fees serves an overriding government interest, while fair-share fees do not. But it is not enough simply to posit that states need the bar to charge dues to fund bar admission and discipline more than they need to maintain public-employee personnel systems. That is like saying chocolate ice cream is better than vanilla. It’s self-evident only to those who share the justices’ values. And even if it were true, can it really be said that the Federal Aviation Administration has a substantial interest in requiring airlines to tell passengers how to fasten a seatbelt? Is there anyone on a plane who has never seen or used one? And how many people who can afford to buy a plane ticket are unaware of questions about alcohol consumption during pregnancy?
It might be that the Court will not follow the broadest implications of its compelled-speech reasoning to all forms of compulsory disclosure or compulsory subsidies, and will instead limit these cases to speech or subsidies that it thinks requires speakers to be complicit in speech they abhor. What underlies the Court’s hostility to speech in NIFLA and Janus is the notion that the anti-abortion “crisis pregnancy centers” and Mark Janus were philosophically opposed to giving the notice or paying fees and that the speech made them complicit in conduct they abhor. In other words, Janus and NIFLA are not necessarily broad pronouncements on compelled speech but, rather, are narrower condemnations of speech or subsidies that require complicity. Even if the opinions are thus narrowed, one may be troubled by the Court’s approach to complicity.
Justice Kennedy said in his concurring opinion in NIFLA: “Governments must not be allowed to force persons to express a message contrary to their deepest convictions.” He found the claim of complicity so weighty in Masterpiece Cakeshop that a baker who refused to bake a cake for the wedding reception of a gay couple was entitled to have his claim judged by a state agency that hadn’t had a member say that atrocities have been committed in the name of religion. The NIFLA and Janus majority’s approach to the complicity theory of compelled speech, like the Masterpiece approach to free speech and free exercise claims, pick up the idea that the Court rejected in Employment Division v. Smith but revived in Hobby Lobby, when it found a statutory right to freedom of religion violated by requiring employers to provide contraceptive coverage in their employee benefit plans. These are extravagantly broad claims about what makes one complicit in the actions of another. This isn’t requiring people to have an abortion or use contraception or marry someone of the same sex, and it’s not even asking someone to officiate at a marriage or attend or help someone get an abortion or contraception. Rather, it is simply having to make a statement of fact or decorate a cake or include coverage in a benefits plan. To turn disclosure, nondiscrimination laws, and employee benefits into complicity is to render religion or conscience a basis for opting out of life in a diverse community.
IV. Implications of Janus on the Ground and in the Courts
Janus poses four very significant problems for labor. First, and most obvious, is the question whether it will lead to a dramatic drop in membership. Anti-union groups have already hired canvassers to go door to door in blue states encouraging employees to quit their union and quit paying dues by convincing them that they can get the benefits of the union contract without paying for it. Their goal is to get so many teachers, home health aides, and others to leave the unions that the unions lack the money to provide services to all the workers they represent. As the quality of services falls, more will leave the union, until ultimately the union will wither away. Unions have anticipated this and have been vigorously signing up fee-payers as full members.
Second, dozens of suits have been filed seeking repayment of fees paid going back for as many years as the statute of limitations will allow. The potential liabilities are staggering if courts conclude that Janus applies retroactively and requires refund of fees paid for many years past.
Third, litigation has been filed seeking to build on the Janus ruling to extend the prohibition on agency fees to the private sector. Some cases focus on the Railway Labor Act, and some on the National Labor Relations Act.
Fourth, litigation has also been filed to build on Janus to extend the prohibition on agency fees to a prohibition on union representation based on majority rule. The plaintiffs in these cases argue that, if paying fees to support bargaining is unconstitutional, it should be unconstitutional for a union to negotiate a contract on behalf of those who do not want union representation at all. Anti-union advocates have lost those cases before. The majority stated that it was not calling majority-rule representation into doubt.
Whether Janus will change that ultimately depends on how the Supreme Court chooses to build on a few passages in Janus. Justice Alito said Mark Janus was not free-riding “on a bus headed for a destination that he wishes to reach but is more like a person shanghaied for an unwanted voyage.” Elsewhere, the majority said that majority rule representation is “a significant impingement on associational freedoms that would not be tolerated in other contexts” and “[d]esignating a union as the employees’ exclusive representative substantially restricts the rights of individual employees.” In a footnote, the majority noted that “under common law, collective bargaining was unlawful, and into the twentieth century, every individual employee had the liberty of contract to sell his labor upon such terms as he deemed proper,” but then said “[w]e note this only to show the problems inherent in the Union respondent’s argument; we are not in any way questioning the foundations of modern labor law.”
V. Unions Security for the Twenty-First Century
The outcome in Janus was anticipated from November 2016, when it became apparent that Donald Trump would fill the Scalia seat rather than Hillary Clinton. Unions and worker advocates have had time to think about union security in a world in which all forms of union security as it has been known for a century have been declared unconstitutional (in the public sector) and are in the sights of the five conservative justices (in the private sector).
The most important and most promising set of ideas are going by the name “Together We Rise.” They focus on legislation and executive action to facilitate communications among workers about unions, to ease membership sign-up, to promote stability of bargaining units, and to create alternative dues-payment systems. Some of these provisions have been enacted into law in California, New York, and New Jersey. As for membership, unions insist that their best protection is to enable them to recruit and retain members. Unions seek, and in California and other states have obtained, mandatory access to new employee orientations. Under these rules, unions have a right to receive notice of when all public-sector employers conduct all new employee orientations and have a right to meet with new employees at the orientations without the presence of supervisors or other entities (such as anti-union organizations). Unions also seek legislation and contract terms requiring delivery of new hire and bargaining unit lists to the exclusive representative and also protections against disclosure of such lists to other organizations. To enable unions to keep their members engaged and to ensure effective communications, unions seek opportunities to communicate with workers at employer trainings or other in-service meetings and to be allowed to use the email system or website of the employer to communicate with workers. Finally, unions seek to ensure the employer remains neutral about unionization by restricting the ability, especially the differential ability, of employers to use supervisors to dissuade employees from joining the union.
Even with rights to communicate with the employees, still unions must have the ability to collect member dues in order to have a stable source of funding to support the union’s role as representative. To that end, unions seek to ensure, through statute or contract, that the employer remains required to deduct dues from the employees’ paycheck if the employee allows it and to allow employees to authorize payroll deduction through electronic signatures. To some extent, the loss of agency fees could be partially offset by allowing union representatives to adjust grievances or engage in other contract enforcement tasks on paid time. (This is the system that is used in the federal service.) But many unions are attentive to the need to ensure that this “official time” does not become the union’s only source of support, because it may have the unintended consequence of making union representatives less responsive to their membership and more responsive to the employer to continue the union representative in his or her job.
These proposals also insist that unions should remain democratically chosen organizations that are member-driven and member-governed. They reject ideas that unions should disclaim their obligation to represent all workers in a bargaining unit; in other words, they reject members-only bargaining.
The most controversial set of proposals to protect union finances without agency fees are proposals to require government employers to pay money directly to the union for representational services. If governments fund unions directly, the money never comes out of the employees’ paycheck, and this would avoid the First Amendment problem the Court perceived in Janus. But this kind of system has serious problems.
When the union receives money directly from the government, it is at risk of becoming less responsive and accountable to its members and more responsive to its funder. Moreover, turning the union into a line item in the government’s budget risks catastrophic cuts to the union’s budget whenever a new government is elected or whenever the public-sector agency budget is cut, and these are the times when government employees may most need the union to advocate for the interests of government workers.
The four major public-sector unions (not including those representing public safety workers)—American Federation of State, County and Municipal Employees (AFSCME), American Federation of Teachers (AFT), National Education Association (NEA), and Service Employees International Union (SEIU)—have all endorsed policies that encourage and enable unions to be responsive to and to represent all workers in the bargaining unit. They unite in opposing members-only bargaining or any other incursion on the principles of majority-rule representation in both contract negotiation and contract enforcement. They unite in opposing authorization of representation of bargaining unit employees by attorneys or other representatives not appointed by the union. They also unanimously oppose creating fee-for-service arrangements for non-members, including the system of per-capita payment by the government to the union.
Janus was one of fourteen 5-4 decisions this Term in which Justice Anthony Kennedy joined with Justices Roberts, Alito, Thomas and Gorsuch in reaching a conservative result. In no case did Justice Kennedy join with the liberals against the conservatives. Therefore, one may say that October Term 2017 marks the beginning of what will surely be a more conservative Supreme Court era. If Brett Kavanaugh is confirmed, he is unlikely to adopt a more tolerant or egalitarian interpretation of the Constitution. Those concerned about the growing divide between rich and poor and who support workplace democracy on the principle of majority rule are, to paraphrase Janus, on an unwanted voyage to a destination they would rather not reach. And the Republican-appointed justices are the captains of the ship, navigating with only their own personal policy preferences as a guide.
* Catherine L. Fisk is the Barbara Nachtrieb Armstrong Professor of Law at the University of California, Berkeley School of Law.
 Mancur Olson, The Logic of Collective Action: Public Goods and the Theory of Groups (1965).
 Janus v. Am. Fed’n of State, Cnty., & Mun. Emps., Council 31, 138 S. Ct. 2448 (2018).
 See Monica Davey and Mitch Smith, Illinois Governor Acts to Curb Power of Public Sector Unions, N.Y. Times (Feb. 9, 2015), https://www.nytimes.com/2015/02/10/us/illinois-governor-bruce-rauner-acts-to-curb-power-of-public-sector-unions.html; Lydia DePillis, Why Public-Sector Unions Lost Big in Illinois, Wash. Post (Nov. 14, 2014), https://www.washingtonpost.com/news/storyline/wp/2014/11/14/why-public-sector-unions-lost-big-in-illinois/?utm_term=.db4520a9b218; Steven Greenhouse, Illinois Governor Bruce Rauner: Organized Labor’s Public Enemy No. 1?, Guardian (Feb. 17, 2015, 7:00 AM), https://www.theguardian.com/us-news/2015/feb/17/illinois-governor-bruce-rauner-unions-labor.
 Rauner v. Am. Fed’n of State, Cnty. & Mun. Emps., Council 31, AFL-CIO, 2015 WL 2385698 (N.D. Ill. May 19, 2015).
 See Olson, supra note 1.
 Section 8(3) of the National Labor Relations Act, as originally written, provided “[t]hat nothing in this Act . . . shall preclude an employer from making an agreement with a labor organization . . . to require as a condition of employment membership therein.” 49 Stat. 499 (1935), as codified at 29 U.S.C. § 158(3).
 Sophia Lee, The Workplace Constitution from the New Deal to the New Right 59 (2014) (“deep-pocketed executives like the du Ponts, anti-New Deal activist groups like [the National Association of Manufacturers], and populist mobilizers ... formed a loose and hazily defined movement in the early 1940s. ‘Right to work’ was its emerging slogan.”).
 As amended in 1947, § 14(b) of the NLRA provided: “Nothing in this Act shall be construed as authorizing the execution or application of agreements requiring membership in a labor organization as a condition of employment in any State or Territory in which such execution or application is prohibited by State or Territorial law.” 29 U.S.C. § 164(b) (2018). Section 8(a)(3) as amended in 1947 prohibited agreements requiring membership at the time of hire, but authorized agreements requiring employees to become members within 30 days after hire. However, it also stated that employees cannot be disciplined if they did not become members if denial of membership was for reasons other than failure to pay dues. 29 U.S.C. § 158(a)(3) (2018).
 NLRB v. Gen. Motors Co., 373 U.S. 734 (1963).
 Id. at 742.
 Retail Clerks Int’l Ass’n, Local 1625, AFL-CIO v. Schermerhorn, 373 U.S. 746 (1963).
 Id. at 752.
 Id. at 753-54.
 This analysis is developed at further length in Catherine L. Fisk & Benjamin I. Sachs, Restoring Equity in Right-to-Work Law, 4 U.C. Irvine L. Rev. 857 (2014).
 Furniture Workers Div., 291 N.L.R.B. 182, 183 (1988); Columbus Area Local, Am. Postal Workers Union, 277 N.L.R.B. 541, 543 (1985); Int’l Ass’n of Machinists & Aerospace Workers, Local Union No. 697, 223 N.L.R.B. 832, 835 (1976).
 Vaca v. Sipes, 386 U.S. 171 (1967).
 Commc’n Workers of Am. v. Beck, 487 U.S. 735 (1988).
 Ry. Emps. Dep’t v. Hanson, 351 U.S. 225 (1956).
 Id. at 238.
 Int’l Ass’n of Machinists v. S.B. Street, 367 U.S. 740 (1961).
 Id. at 762.
 Id. at 773.
 Id. at 768.
 Abood v. Detroit Bd. Of Educ., 431 U.S. 209 (1977).
 S.B. Street, 367 U.S. at 780 (Frankfurter, J., dissenting).
 Id. at 806.
 Id. at 808.
 Knox v. Serv. Emps. Int’l Union, Local 1000, 567 U.S. 298 (2012); Harris v. Quinn, 134 S. Ct. 2618 (2014).
 Janus v. Am. Fed’n of State, Cnty., & Mun. Emps., Council 31, 138 S. Ct. 2448, 2486 (2018).
 Id. at 2463 (quoting W. Va. Bd. of Ed. v. Barnette, 319 U.S. 624, 642 (1943) (emphasis in Janus)).
 Harris, 134 S. Ct. at 2644.
 Actually, it did. But at least it was compelled to give them a corresponding ten percent reduction in work hours.
 Abood v. Detroit Bd. of Educ., 431 U.S. 209, 220-21 (1977).
 Garcetti v. Ceballos, 547 U.S. 410 (2006).
 Pickering v. Bd. of Educ., 391 U.S. 563 (1968).
 Janus v. Am. Fed’n of State, Cnty., & Mun. Emps., Council 31, 138 S. Ct. 2448, 2472 (2018).
 Brief for the Petitioners, Garcetti, 547 U.S. at 410 (No. 04-473).
 Rust v. Sullivan, 500 U.S. 173 (1991).
 Nat’l Inst. of Family & Life Advocates v. Becerra (NIFLA), 138 S. Ct. 2361, 2378 (2018); see Cal. Health & Safety Code § 123472(a)(1) (requiring certain primary care clinics to post a notice stating: “California has public programs that provide immediate free or low-cost access to comprehensive family planning services (including all FDA-approved methods of contraception), prenatal care, and abortion for eligible women. To determine whether you qualify, contact the county social services office at [telephone number]”).
 Masterpiece Cakeshop, Ltd. v. Colo. Civil Rights Comm’n, 138 S. Ct. 1719 (2018).
 Janus v. Am. Fed’n of State, Cnty., & Mun. Emps., Council 31, 138 S. Ct. 2448, 2473 (2018).
 W. Va. Bd. of Ed. v. Barnette, 319 U.S. 624, 642 (1943).
 NIFLA, 138 S. Ct. at 2384 (Breyer, J., dissenting) (quoting Planned Parenthood v. Casey, 505 U.S. 833, 881 (1992) (joint opinion of O’Connor, Kennedy, and Souter, JJ.) (quoting 18 Pa. Cons. Stat. § 3205 (1990)).
 See, e.g., NY Const. art. I § 17 (“Employees shall have the right to organize and to bargain collectively through representatives of their own choosing.”).
 Janus, 138 S. Ct. at 2470.
 United Pub. Workers v. Mitchell, 330 U.S. 75 (1947); U.S. Civ. Serv. Comm’n v. Nat’l Ass’n of Letter Carriers, 413 U.S. 548 (1973).
 Elrod v. Burns, 427 U.S. 347 (1976).
 Rutan v. Republican Party, 497 U.S. 62 (1990).
 Janus, 138 S. Ct. at 2502 (Kagan, J., dissenting).
 Nat’l Inst. Of Family & Life Advocates v. Becerra (NIFLA), 138 S. Ct. 2361, 2379 (2018).
 Id. at 2384 (Breyer, J., dissenting).
 See id. at 2371.
 Id. at 2372.
 Janus v. Am. Fed’n of State, Cnty., & Mun. Emps., Council 31, 138 S. Ct. 2448, 2464-65 (2018).
 “For commercial speech to come within [the First Amendment], it at least must concern lawful activity and not be misleading. Next, we ask whether the asserted governmental interest is substantial. If both inquiries yield positive answers, we must determine whether the regulation directly advances the governmental interest asserted, and whether it is not more extensive than is necessary to serve that interest.” Cent. Hudson Gas & Elec. Corp. v. Pub. Serv. Comm’n of N.Y., 447 U.S. 557, 566 (1980).
 Id. at 2465 (citations omitted).
 NIFLA, 138 S. Ct. 2361, 2377 (2018).
 Harris v. Quinn, 134 S. Ct. 2618, 2643 (2014).
 NIFLA, 138 S. Ct. at 2379 (Kennedy, J., concurring).
 Masterpiece Cakeshop v. Col. Civil Rights Comm., 138 S. Ct. 1719 (2018).
 Emp’t Div., Dep’t of Human Res. v. Smith, 494 U.S. 872 (1990).
 Burwell v. Hobby Lobby Stores, Inc., 134 S. Ct. 2751 (2014).
 See, e.g., Dirk Vanderhart, Following Decision On Union, Canvassers Target Public Workers in Oregon, NW News Network (July 2, 2018), http://nwnewsnetwork.org/post/following-decision-union-canvassers-target-public-workers-oregon; Bloomberg, Group Funded by Conservative Billionaires Launches Anti-Union Campaign Following Supreme Court Ruling, L.A. Times (June 28, 2018, 12:10 PM), http://www.latimes.com/business/la-fi-freedom-foundation-20180628-story.html.
 See J.I. Case Co. v. NLRB, 321 U.S. 332 (1944); Order of R.R. Telegraphers v. Ry. Express Agency, 321 U.S. 342, 346 (1944) (reasoning that if individual agreements could supersede collective bargaining agreements, “statutes requiring collective bargaining would have little substance, for what was made collectively could be promptly unmade individually”); NLRB v. Crompton-Highland Mills, 337 U.S. 217 (1949) (holding that, when an employer excluded the majority elected representative by increasing wages outside of the negotiation, the side agreement constituted unfair labor practices).
 Janus v. Am. Fed’n of State, Cnty., & Mun. Emps., Council 31, 138 S. Ct. 2448, 2466 (2018).
 Id. at 2478.
 Id. at 2460.
 Id. at 2471 n.7.
 See Cal. Gov’t Code §§ 3555-3559 (effective Jun. 27, 2018) (requiring California public employers to provide unions mandatory access to new bargaining unit employees at orientation); 2018 Sess. Law News of N.Y. Ch. 51 (S. 7501) (requiring public sector employers to furnish labor organizations with employee contact information and permitting union representatives to meet with new public sector employees within thirty days); A. 3686, 218th Gen. Assem., Reg. Sess. (N.J. 2018) (allowing labor representatives to meet with members during work hours and requiring public employers to furnish labor organizations with contact information of covered employees).
 See id.
 See id.
 See, e.g., Assem. Bill 119, 2017-2018 Reg. Sess. (Cal. 2017).
 See 29 U.S.C. § 158(a)(3) (2018) (“It shall be an unfair labor practice for any employer to . . . encourage or discourage membership in any labor organization.”); 29 U.S.C. § 152(2) (2018) (“The term ‘employer’ includes any person acting as an agent of an employer, directly or indirectly . . .”); but see NLRB v. CNN America, Inc., 865 F.3d 740, 762 (D.C. Cir. 2017) (finding that although supervisors’ statements reflected an anti-union animus, the “employer’s hiring process must be either upcoming or ongoing” because if hiring were completed a “no-union statement could not generally suggest coercion.”).
 5 U.S.C. § 7131 (2018).
 See Daniel Hemel and David Louk, Is Abood Irrelevant?, 82 U. Chi. L. Rev. Dialogue 227, 229 (2015) (“If a public sector employer wants to make sure that a labor union is compensated for the cost of representing nonmembers, the employer could just as easily reimburse the union for those expenses directly.”).