On June 13 the Supreme Court once again sided with a multibillion-dollar corporation over its workers. The case of Starbucks Corp. v. McKinney concerns seven employees, now known as the Memphis Seven, whom Starbucks fired in February 2022 as they tried to unionize their store in Tennessee. (Because federal law prohibits employers from retaliating against organizing, the company naturally claims they were let go for violating workplace policies.) The National Labor Relations Board (NLRB), the agency tasked with guaranteeing workers’ rights to join unions and negotiate contracts, was quick to intervene. Directed by the Biden-appointed General Counsel Jennifer Abruzzo, NLRB staff filed for a preliminary injunction to force Starbucks to reinstate the fired activists while the case was fully litigated.

Such requests are rare. The NLRB only makes them when companies glaringly violate labor law and the agency is confident that courts will decide the case in the workers’ favor. In August 2022 a lower-court judge agreed and granted the injunction.

That six-month wait for a modicum of justice was blazing speed by the standards of United States labor law. It was too fast for the Supreme Court, which, in an 8-1 decision, reversed the injunction. The justices ruled that when courts consider the NLRB’s injunction requests, rather than using a legal standard specific to labor disputes that gives the board relative deference, they must use a more restrictive standard known as “the traditional four-factor test,” as articulated in the 2008 case Winter v. Natural Resources Defense Council, Inc. Among the factors considered are the “balance of equities” (meaning that the ruling is fair to both sides) and that an injunction serves the “public interest.” Both factors are a matter of opinion. Put simply, the Court is placing its own views over the expertise of the NLRB’s professional staff, in the name of a “traditional” test that’s as old as a teenager.

The NLRB still needs to decide the case, which might take a year or more. A majority of its members—the Democrats—will likely agree that Starbucks violated the plain language of the law forbidding “discrimination in regard to hire or tenure of employment or any term or condition of employment to encourage or discourage membership in any labor organization.” When the agency returns to court for an enforcement order, the judges should rule in its favor once more, though Starbucks can appeal that decision as well.

All this delay favors employers. Union representation cases are usually won or lost in the years it takes to finish adjudicating a wrongful termination charge—the fired Starbucks workers included five of the six members of the organizing committee. Many cases are settled for cash payouts and a mutual agreement to call the matter a “resignation.” Sometimes unions trade settlements to get a company to withdraw its own charges or bargain in good faith. Starbucks is apparently negotiating under just such a brokered framework with the union, Workers United.

In one sense the consequences of Starbucks v. McKinney are relatively minor. Many Republican-leaning circuit courts were already using the four-factor test. In her partial dissent, Ketanji Brown Jackson agreed that it was the appropriate rubric but argued that it should be applied in a way that recognizes the NLRB’s authority, and that courts shouldn’t fully relitigate such cases: an “injunction request simply does not present the district court with an opportunity to wade into the midst of an ongoing labor dispute (over which it otherwise has no say) and offer its own take about how the merits should be decided.”

But in another, deeper sense, the Starbucks decision is a dispiriting sign that the courts will only allow labor rights to be revised downwards. The National Labor Relations Act, which Congress passed in 1935 to protect workers’ right to form unions and bargain collectively, may no longer be adequate to that task in a court system that has been historically pro-corporate but is especially conservative today in the aftermath of the Trump administration (though even two of the liberals, Elena Kagan and Sonia Sotomayor, joined the Starbucks majority). In another blow to unions, on June 28 the Court, in a more nakedly partisan 6-2 vote, eviscerated the forty-year-old Chevron deference, rejecting the subject-matter expertise and statutory interpretation not just of the NLRB but of all federal regulatory agencies. How might union supporters hoping to curb inequality wrest control back from the legal system?

Judicial hostility to labor is hardly a new phenomenon. As Jackson noted in her partial dissent, “To put it bluntly, courts exercising their equitable discretion amidst labor disputes today do so against the backdrop of an ignominious history of abuse.” In his classic study, Law and the Shaping of the American Labor Movement, the historian William E. Forbath showed how, in the nineteenth century, “judge made law”—that is, legislating from the bench—forced unions to adopt an essentially conservative political strategy.

In an era-defining decision in the case Lochner v. New York (1905), the Court overturned a New York law limiting working hours in bakeries, arguing that the legislation interfered with the freedom to contract under the due process clause of the Fourteenth Amendment.

As a result, in its early decades the American Federation of Labor (AFL)—the country’s largest labor organization—focused narrowly on contractual negotiations and largely gave up on state intervention as a solution to working people’s problems, except to try to prevent courts from intervening in labor disputes. When Roosevelt’s 1933 National Industrial Recovery Act tentatively endorsed collective bargaining and the right to organize, the Supreme Court overturned it using the Lochner doctrine. New York senator Robert Wagner then responded with the 1935 NLRA. That law stripped the courts of nearly all jurisdiction on cases pertaining to labor relations—essentially telling judges to keep their opinions on the subject to themselves—and created the NLRB, a parallel system of civil tribunals, which hears cases of alleged violations of workers’ rights to organize unions and go on strike. If the NLRB finds that an employer broke the law and cannot get them to comply or settle, it then and only then turns to the courts for enforcement (including injunctions). The NLRA was intended to supersede common law.

Set beside the original mandate of the NLRB, the Court’s decision in Starbucks v. McKinney couldn’t be more wrong. But the federal agency’s authority has eroded over time. In Values and Assumptions in American Labor Law, the legal scholar James B. Atleson narrated how, after the passage of the NLRA, the class biases of patrician judges led them to push back against pro-worker laws. “The belief in the inherent rights of property and the need for capital mobility, for instance, underlie certain rules,” Atleson writes, “and some decisions turn on the received superior need for continued production or the fear of employee irresponsibility.”

More recently, in his book The Supreme Court on Unions (2016), Julius B. Getman described how, if anything, the highest court’s historical antagonism to labor has only gotten worse in the last four decades. “What has remained constant over the years has been judicial arrogance,” he writes: “the willingness of the Court to establish factual premises for its decisions with little basis in reality.”

Take the right to strike, which is both specifically enumerated under the NLRA and clearly identified as one of its policy goals: “Nothing in this Act shall be construed so as either to interfere with or impede or diminish in any way the right to strike.” But this was undone just three years later in NLRB v. Mackay Radio & Telegraph Co., a 1938 case concerning the firing of striking workers at a regional telecom firm. Justice Owen Roberts upheld strikes as protected activity, but he also made a careless aside in his decision, musing that an employer would be allowed to permanently replace his striking workers if it was necessary “to protect and continue his business.”

In the following decades, employers pressed the outer edges of Mackay: Can we offer replacement workers super-seniority protection from layoffs while strikers are not similarly protected? (No.) How about give them vacation pay that strike participants are denied? (Also no.) By the 1980s, however, their efforts turned Justice Roberts’s offhand comment into stare decisis, a disastrous precedent that held that replacing strikers and only offering them their jobs back when scabs retire is “proper under Mackay.”

Judges also watered down the “balance of equities” protections, which were once robust enough to protect even the worst employees from retaliatory firings for union activism. Consider Walter Weigand, the subject of the landmark 1943 case Edward G. Budd Mfg. Co. v. NLRB. “If ever a workman deserved summary discharge it was he,” concluded a bemused circuit court judge. “He was under the influence of liquor while on duty. He came to work when he chose and he left the plant and his shift as he pleased.” If I’m not misinterpreting the court’s demure transcript, Weigand also ran a prostitution operation in the employer’s back alley (on company time) and testified that he had no idea what his actual job entailed. But he was only fired after switching from the boss’ company union to the more militant CIO, and for that reason the NLRB and the courts ordered him back on the job as the law intended.

But since 1980, the NLRB has applied what’s called the “Wright Line” standard for judging “mixed motive” terminations of union activists. It requires that a union prove that a worker was fired while engaged in protected union activity, that the employer knew the worker was a union activist, and that the employer held animus against the union. These demands in turn require more time-consuming litigation, even when an employer’s actions—like Starbucks firing people for talking to reporters about their union activism—wouldn’t pass a common-sense smell test. It’s not surprising, then, that the Wright Line standard has become a go-to union-busting weapon for employers. These days union supporters are fired in the course of nearly one in three certification votes brought before the NLRB.

An implicit assumption of the 1935 NLRA was that employers would remain neutral about organizing campaigns, and recognize a union if a clear majority of workers wanted one. But soon the Supreme Court had granted employers a First Amendment right to campaign against unionization and force employees to attend mandatory captive audience presentations. In the 1950s the NLRB fought for unions’ right to respond to these presentations, or at least mandate some access to the workplace, but since 1956 the Supreme Court has consistently supported employer’s property interests over unions’ access and speech.

What do employers say in captive audience meetings? In 1969 Chief Justice Earl Warren allowed them to threaten that a successful union drive would lead to workplace closures or other negative effects, as long as these threats were presented as predictions “carefully phrased on the basis of objective fact to convey an employer’s belief.” In a 2009 report, the labor scholar Kate Bronfenbrenner found that between 1999 and 2003, employers threatened plant closure in 57 percent of NLRB elections, and in 15 percent of the cases they actually followed through.

In all, it took the courts roughly thirty years to take workers out of a legal environment where union organizing was a fully enforceable right—with meaningful job protections and enforcement against threats, reprisals and the refusal to negotiate—to one where certification elections are conducted under manifestly crooked rules.

While corporations aggressively worked the courts, the AFL-CIO and its think-tanks pressed for legislative reform whenever the Democrats briefly controlled the White House and both chambers of Congress. But the Democrats only pursued narrow, technical fixes: a push to ban permanent replacements under Clinton; simple proof of majority, or “card check,” certification under Obama. Even these were vigorously resisted by the business lobby and eventually killed by filibusters. Biden’s is the first Democratic administration to meaningfully use the rules in place, and the NLRB’s rule-making authority, to encourage the practice of collective bargaining.

“Of all of the members of Biden’s administration,” the Nation has gushed, it’s Abruzzo “who has brought about the most significant changes for American workers.” Under Abruzzo the NLRB has sped up the timeline for conducting union representation elections. It has also expanded the instances in which employers must accept card check certification, has tried to limit the ability of employers to conduct mandatory captive audience meetings, and stretched for “make whole” financial penalties against employers who violate their workers’ rights. Even its decision to request an injunction in the Starbucks case was part of Abruzzo’s systemwide attempt to modernize operations. (The NLRB has not yet, I’ll note in case Abruzzo is reading, restored the right to strike by revisiting the Mackay standard and forcing employers to prove that they would go out of business if not allowed to hire permanent replacements.)

And yet any of these actions can be overturned if the Supreme Court finds they violate a precedent or employers’ First or Fourteenth Amendment rights. All of them, ultimately, cry out for new lawmaking. Restoring the promise of US labor law requires amending the NLRA to override unfair court precedent and reassert the NLRB’s supremacy over judges on routine enforcement of labor violations.

It is axiomatic on the left that lawmakers do not pass such prolabor legislation until a strike wave forces them to. Yet the historical record suggests otherwise. The “right” to organize was shoehorned into the National Industrial Recovery Act—the objectively pro-business “First New Deal” legislation that the Supreme Court overturned—to gain the AFL’s support for a controversial bill. It wasn’t won through workplace action, and lacking enforcement powers, it wasn’t worth the paper it was written on. But by offering just enough to raise workers’ expectations—and more to dash their hopes—it inadvertently birthed the very militancy needed to enshrine more meaningful reform.

This argument about the effect of the NIRA is still controversial. The labor activist and scholar Eric Blanc recently cut through decades of mythmaking and crunched the numbers. He found a 129 percent increase in union membership and a 260 percent increase in workers going on strike in the months that followed the act’s enactment. In my forthcoming book, We Always Had a Union: New York’s Hotel Workers Unions, 1912-1953, I show that the National Recovery Administration’s wage and hours code, in addition to failing to compel union recognition and reinstate fired union activists, directly caused a citywide hotel strike that the agency was attempting to mediate in January 1934. Though the unions didn’t “win” that strike, shaken hotel bosses negotiated a neutrality agreement so that the industry wouldn’t be hit with another work stoppage during the 1939 World’s Fair. The agreement resulted in the New York Hotel Trades Council, the union that remains a powerhouse in the city and state.

One lesson from this history is that unions need to be quicker to take advantage of favorable organizing environments. Today there is both a noticeable uptick of worker-led organizing (as evidenced by independent union wins at Amazon and Trader Joe’s, and wildcat job actions during the pandemic) and a pro-worker NLRB. And yet most unions are building up their savings accounts instead of spending down on new campaigns. Unions have collectively reduced their staff and organizers by an estimated 19 percent between 2010 and 2020, which translates to 23,440 fewer people that could potentially be put in the field. With the benefit of new leadership, the Teamsters and UAW are notable exceptions, joining a small handful of organizing champions like SEIU and UNITE HERE in investing significant resources on campaigns aimed at growing the labor movement, at huge employers like Volkswagen, Marriott hotels, and FedEx.

The other lesson is that opportunities for legal reform come rarely, briefly, and usually by surprise. There is no shortage of reform proposals like the PRO Act and the Clean Slate program. But too little attention is paid to counteracting judicial bias. If the Roberts Court’s judicial power grab, already a constitutional crisis, is to be met with more political brinkmanship like court-packing and Congressional overrides of politically motivated decisions, then why not center that fight on issues where judges’ historic tendency to favor corporations over workers cast them as perfect villains? Union supporters should be demanding that judges stop imposing their values and assumptions on accredited bodies like the NLRB. As the Transport Workers Union president Michael J. Quill once said of the man who sentenced him to jail for the 1966 New York City transit strike, “The judge can drop dead in his black robes.”

[This article originally appeared at New York Review of Books.]