HARDLY ANYONE remembers the yellow dog contract. Before the New Deal, employers commonly required workers to sign documents warranting that they were not union men. The courts regarded yellow dog contracts as a lawful exchange of considerations: the employer proffered the job, the employee handed over his freedom of association. The surrendered worker’s right became, in effect, the employer’s property and—as the Supreme Court declared in 1917—it, therefore, entitled him to an injunction against any union trying to organize his workers. In the wake of this monstrous decision, wide swaths of American industry became off limits to union organizers.
The Court had overreached itself. In 1932, the Norris-LaGuardia Act declared the yellow dog contract unenforceable and promulgated a public policy that favored freedom of association for “the individual unorganized worker [who] is commonly helpless to exercise actual liberty of contract.” The key words—“actual liberty of contract”—reappeared in Senator Robert F. Wagner’s 1935 National Labor Relations Act and underwrote the New Deal’s embrace of collective bargaining.
So it is a towering irony that the labor law that arose from the ashes of the yellow dog contract recapitulates that time in American history when injunctions barred unions from much of American industry. Consider a recent Supreme Court decision that preempted a California statute intended to prohibit private contractors from using state funds to fight union organizing. The California statute, the Court declared, offended the protection the federal law affords for employer free speech and “uninhibited, robust, and wide-open debate.” This invocation of an employer’s speech right performs the same validating function in our failing labor law that liberty of contract once performed for the yellow dog contract.
Employers can speak, but they cannot coerce. That’s the reigning doctrine, defined in this way: threats are coercive, predictions are not. The employer cannot say that he will shut down if the union wins, but he can suggest that a union contract will drive him out of business. If threats are needed, he can do that too, because—except in the most egregious of cases—he’s subject to nothing more than a National Labor Relations Board order to desist. He can even fire the key rank-and-file leaders since the NLRB penalty is only to reinstate them and pay lost wages. In most instances, the employer settles, years after the union has gone away. He has got his money’s worth (and writes off the cost as a business expense).
As for “uninhibited, robust, and wide-open debate,” here’s what typically happens when a representation election commences. The employees are called into meetings. Attendance is required, no back-talk permitted. (Free speech is strictly the employer’s right.) Employees will also be interrogated individually. Up close, their supervisors will be on their case, because their own jobs depend on holding the employee’s loyalty. The election stretches out for eight weeks, plenty of time to put the screws on, and while the employer is holding his captive-audience meetings, the organizers are out on the street, barred from company property.
HOW DID we ever arrive at this juncture? Not because Senator Wagner, the prime mover, didn’t see it coming. He was FDR’s point man on the New Deal’s first non-statutory labor board, and the main problem he confronted was the company union, which employers hawked as an alternative to independent unions. From that battle, he drew two conclusions: employers had no place in the unionizing process; and their words, however honeyed (as they often were in the company-union context), coerced workers. But fearful of a hostile Supreme Court—one that was challenging the entire New Deal—Wagner contemplated and then decided against adding the word “influence” to the unfair-labor-practice provision that forbade employers “to interfere with, restrain, or coerce” employees. In the end, the Court backed down, but it was too late for the Wagner Act.
The early NLRB, understanding Wagner’s intention, barred employer speech during representation campaigns, but the Supreme Court overruled the Board in 1941. Without a legislative prohibition to push up against, employer free speech slipped into the case-law canon almost by default: the original 1941 decision actually held against the employer, and the case first invoking the First Amendment was not about employer speech. In that case, Justice Robert H. Jackson remarked that if intimidation accompanied speech, “the constitutional remedy would be to stop the evil, but permit the speech, if the two are separable, and only rarely and when they are inseparable to stop speech or publication.” Justice Jackson’s incomprehension is in a direct line with the current Supreme Court’s invocation of “non-coercive” employer speech and “uninhibited, robust, and wide-open debate”
The representation system the Court celebrates is, in any case, grinding to a halt. NLRB elections have fallen by half in the past decade. Four-fifths of all newly unionized workers now gain their bargaining rights not via NLRB certification but as a result of corporate campaigns in which unions pressure employers into forgoing secret-ballot elections. So-called card-check/neutrality agreements conform to the law—they still eventuate in a demonstration of majority support—but they revert to power struggles that the law was meant, more than anything else, to end.
So the country has arrived, slow-motion, at another Norris-LaGuardia moment, when the dissonances are so jarring, the consequences so unpalatable, that something must be done. That something is the Employee Free Choice Act. In a sense, the Employee Free Choice Act turns back the clock, revisiting choices Senator Wagner made in the face of a hostile Court and robustly enforcing workers’ rights and the employer’s duty to bargain in a different constitutional era.
The representation election, however, offers no constitutional remedy. Once the Supreme Court upheld employer free speech, there was no turning back. Taft-Hartley added its own free-speech clause. Even if it was repealed, the underlying jurisprudence would remain. At best we’d see some slight shift in coercive/non-coercive distinctions (threat/prediction derives from Taft-Hartley). Besides, the ground rules for “wide-open” debate—captive-audience meetings, interrogations, organizers barred from company property—all once unfair labor practices, now with a Supreme Court imprimatur, make legislative fixes problematic.
So, on this issue, the Employee Free Choice Act follows the path of Norris-LaGuardia. It does not challenge employer free speech (any more than Norris-LaGuardia challenged liberty of contract). Instead, the signing of authorization cards—card check, so-called—becomes the default method. Employer free speech is untouched. It’s just deprived of the platform—the representation election—for its exercise.
NATURALLY EMPLOYERS are up in arms, not on their own behalf, but on behalf of their employees. The pitch, now heard incessantly, is that Big Labor wants to deprive workers of the “basic, democratic right” to the secret ballot. This is equivalent to the pitch once heard that enemies of the yellow dog contract wanted to deprive the worker of the liberty “of every man to work for whom he pleases and under conditions chosen by himself.” No fooling: employers really said that. So we need an equivalent answer, a crystallizing argument like Norris-LaGuardia’s declaration of actual liberty of contract.
The place to look is deep within the law. In the array of labor’s rights, the one that applies to the NLRB election is “representatives of their own choosing.” But “representatives of their own choosing” also outlaws company domination of a labor organization. It is not obvious why—since both derive from “representatives of their own choosing”—labor organizations should be subject to a company-domination test, but the process for choosing labor organizations is not. The explanation is wrapped in history. Employee choice, although always implicit in “representatives of their own choosing,” arrived unbidden. It was the byproduct of the company union battle in the run-up to the Wagner Act and its characteristics—majority rule, exclusive representation, duty to bargain—were defined by that battle. When the battle ended abruptly with the company unions being outlawed, the majority-rule system remained in place but with no requirement that elections be used.
In 1939, however, the NLRB ruled that employers could call for an election when faced rival unions, and then, almost immediately, even absent union rivalry. It was labor’s doing—the AFL-CIO war had broken out—but the election became mandatory at the employers’ behest. Their mutterings about free-speech rights erupted into a clamor. They had been driven out of the unionizing space by the defeat of company unionism. The representation election, with speech rights as their lever, gave them a way back in. How they exploited that opportunity is another story, but the end result is as I’ve described it, and for which we now have language: the NLRB representation system is company-dominated.
This is the time to restore the symmetry lost in history. The law prohibits company domination of a labor organization. It should not countenance company domination of the representation process. That’s the argument for the Employee Free Choice Act. Not perhaps as potent as Norris-LaGuardia, but good enough. Employers say “secret ballot.” Labor should say “company-dominated election,” every time. “Company-dominated” may not have the zing of “yellow dog.” But it’s good enough.
The labor side has chosen to evade the secret-ballot issue. All that the Employee Free Choice Act says on behalf of card check is that it would make for a more “efficient” certification system. In public, unions talk mainly about the benefits of collective bargaining, about saving the middle class. That’s true, but in these days of economic meltdown, it’s not enough.
And now, as Congress takes up the Employee Free Choice Act, we can see why. The Senate is the battleground, and in the face of tremendous industry pressure, Democrats are wavering. On top of that, Senator Arlen Specter, the bill’s lone Republican co-sponsor in the last Congress—his would have been the sixtieth vote for cloture—caved. His excuse? While acknowledging employer coercion, he couldn’t accept “the elimination of the secret ballot, which is the cornerstone of how contests are decided in a democratic society.” This is the price of labor’s intellectual failure. The answer to Specter—that he’s defending company-dominated elections—might have kept him honest. In any case, the secret-ballot dodge must be faced because, beyond the cover it offers to fence-sitters, it in fact works all too well with those who don’t know how the law operates–which is to say, most people, including economists and journalists.
Specter’s defection inaugurated a new phase in the battle. The question is no longer whether the Employee Free Choice Act has the votes—it doesn’t—but whether the votes could be collected by compromise. Specter himself got the ball rolling by appending to his floor statement a wide-ranging list of proposed revisions. And here, amid the give-and-take on Capitol Hill, the concept of company domination has a second use: it’s a standard for testing what the labor side can settle for. In an ideal world, of course, we would aim for restoring the law’s original intent and bar employer speech as inherently coercive. Second best is card check, which leaves employer speech unimpaired but—imperfectly, to be sure—insulates workers from employer coercion and re-opens the unionizing process. We should not let card check slip away. This just might be a second New Deal moment, when the unachievable suddenly comes within reach. If next year’s mid-terms bring a second Democratic wave, we will have a once-in-a-generation chance to rid the country of an employer-dominated labor law. That’s worth waiting for.
David Brody is a professor emeritus of history at the University of California, Davis.This is an updated version of Brody’s April 1, 2009 article.