Everyone agrees that labor law is broken. Under the auspices of the National Labor Relations Act (NLRA) — which was passed in 1935 at the height of the New Deal and laid the foundation for our current regime of collective bargaining — union membership rates have declined to existentially low levels. Though the weaknesses in labor law have been glaringly apparent for some time, and intermittent attempts have been made to reform it, discussion about labor law reform is now reaching a critical mass.

Labor law reform has been central to the campaign promises of both Bernie Sanders and Elizabeth Warren. There is much in common between the Sanders and Warren plans, though the level of detail in the Warren plan burnishes her reputation as a technocrat. Liberal think tanks have jumped on board. Left-leaning publications have also directed their attention to labor law reform.

What unites most of these proposals is the idea of strengthening labor rights. I wrote an essay recently in Catalyst arguing that this approach is wrong. The labor movement should be wary of labor rights and instead seek to expand labor freedoms.

A right is some legally enforceable claim, backed through the coercive machinery of the state (fines, injunctions, imprisonment, etc.), that one legal subject has against another because of some interference caused or threatened by that other. A freedom, in contrast, is the absence of a legally enforceable duty to refrain from some action.

A “right to strike,” for example, means that workers are protected from any interference an employer might take against an employee for engaging in a strike. During a strike, hiring permanent replacement workers counts as the most obvious form of interference, and indeed such replacements have had a devastating impact on the effectiveness of strikes. A fully recognized right to strike would prohibit the hiring of permanent replacements and legally compel employers to discharge their replacements when striking workers decide to call off the strike and return to work.

All well and good, except that this rights approach overlooks the most important reason employers get away with hiring permanent replacements: labor law effectively bans mass picketing, the picketing of large numbers of workers near the struck business. Before mass picketing was banned, it was the most potent weapon in labor’s arsenal in the 1940s, and its repeated use established an “unofficial norm” against hiring permanent replacements, a norm that lasted until employers started defying it in the 1980s.

Elimination of the ban on mass picketing would give workers a labor freedom rather than a labor right. With the labor freedom, it is workers themselves, through mass picketing, who enforce their strike power; with the labor right, it is the state, through the ban on permanent replacements, that does the enforcement.

One might ask, “What’s the difference, if workers win the strike in the end?” Part of the answer comes from asking yourself, “Which of the two will build stronger and longer-term working-class solidarity?”

The other part of the answer is that in numerous other cases, the effect of labor rights has been far more insidious. Labor rights, unfortunately, have been frequently used by judges, politicians, and bureaucrats as reasons for prohibiting or eliminating protection for strikes and other forms of collective activity. One example of this is the NLRA’s ban on organization and recognition picketing. Labor law prohibits any picketing (or even threats of picketing) “where an object thereof is forcing or requiring an employer to recognize or bargain with a labor organization as the representative of his employees.”

This provision exists not because of some cynical, ideologically motivated, anti-union impulse. Rather, it exists because the NLRA already provides workers with a “fair” and “neutral” administrative method for choosing a bargaining representative and establishing a bargaining relationship: the National Labor Relations Board’s election procedure. In practice, however, these provisions virtually compel workers to make use of the board’s election procedure, which is characterized by legal-bureaucratic delay and employer intimidation.

It should be pointed out that this prohibition was added by the Taft-Hartley Act, which was called a “Slave Labor Bill” by unions when passed in 1947. However, even before Taft-Hartley’s passage, the National Labor Relations Board (NLRB) through its case decisions had already withdrawn protection for organization and recognition picketing for precisely the reason identified: because it was inconsistent with the board’s preferred “neutral,” bureaucratic procedure for settling issues of representation. The ban was therefore inspired not by the rabid right wing of the postwar Republican Party, but the progressive-liberal vision of the likes of Louis Brandeis and others, for whom labor justice resided not with workers and unions themselves but with experts ensconced in the administrative state.

In the mind of a judge or bureaucrat, there is no harm in this substitution of legal procedure for collective action. If workers designate a union to represent them, the legal process triggers a duty to bargain on the part of the employer (as well as the union), a duty that the board can enforce. From the state’s point of view, allowing workers to enforce that duty through concerted activity is legally redundant. From a class perspective, however, legal enforcement and worker self-help are not sociologically redundant.

Strikes and other concerted activities are not merely means by which workers achieve gains in the workplace. Rather, they are moments in the process by which workers constitute themselves as a class — building solidarity, raising class consciousness, creating their own norms and institutions, and discovering their own forms of class power.

Not only is the legal point of view blind to these processes, they are viewed as subversive to the law. As one Supreme Court opinion put it, “To justify such conduct because of the existence . . . of an unfair labor practice would be to put a premium on resort to force instead of legal remedies and to subvert the principles of law and order which lie at the foundations.”

It is easy to find additional examples of the way rights are used to curtail workers’ concerted activity. In NLRB v. Fansteel Metallurgical Corp., the Supreme Court denied the board’s order to reinstate fired sit-down strikers, who struck in response to the employer’s violation of the NLRA, because the act already “provided a remedy.” Thus, the employer’s interference with the “right of self-organization could at once have been the subject of complaint to the Board.”

In Boys Markets, Inc. v. Retail Clerks Union, Local 770, the Supreme Court upheld an injunction against workers striking to enforce their legitimate claims under a collective agreement because federal law already recognizes collective bargaining agreements as legally enforceable instruments. The provision for legal enforcement would be “obviously largely undercut” if there was no way to stop the collective action that legal enforcement was “designed to obviate.”

In American Communications Assn. v. Douds, the Supreme Court upheld the constitutionality of the Taft-Hartley’s anti-communist oath for union leaders because “when Congress clothes the bargaining representative ‘with powers comparable to those possessed by a legislative body both to create and restrict the rights of those whom it represents,’ . . . the public interest in the good faith exercise of that power is very great.” In each case, the statutory provision of rights (or powers) justifies limits on labor’s own forms of collective autonomy.

The conclusion is: rights limit strikes. In this view, collective action is unnecessary and disruptive in light of existing statutory rights and procedures. But this outcome hampers the formation of working-class solidarity and makes workers and unions subordinate to and dependent on the state. This is not the path to building working-class power or socialism.

What sort of labor law should socialists and the rank-and-file labor movement be fighting for? If labor rights present potential pitfalls, the answer lies in seeking to expand labor freedoms instead.

The affirmative bans on worker collective activity are abundant in the NLRA. Removing these restrictions would alone do much to free up working-class power.

At the top of the list should be the elimination of the bans on (1) mass picketing in Section 8(b)(1)(A) of the NLRA, (2) organization and recognition picketing in Section 8(b)(4), and (3) secondary boycotts and picketing in Section 8(b)(7). We should also challenge the numerous bans on mass picketing that exist at the state level — a curious exception to the otherwise extremely broad principle that federal labor law preempts local and state labor laws.

The ban on mass picketing has done the most to destroy the power of the simple, direct strike and open the door to the employer’s use of permanent replacement workers. The ban on organization and recognition picketing has done the most to frustrate work self-organization, organizing that is independent of the established procedures of the NRLA. And the ban on secondary activity has done the most to inhibit and undermine working-class solidarity and coordination across workplaces and industries.

There are several ways to eliminate these restrictions. The US Congress could amend the National Labor Relations Act and remove the offending provisions. Some labor law scholars have argued that prohibitions on picketing violate the First Amendment’s protection of free speech, and therefore constitutional litigation could challenge them in the federal courts. However, I am not optimistic about either of these strategies alone, without workers themselves confronting these bans in acts of civil disobedience.

In the end, the judgment of a good labor law should not be whether it serves the immediate interests of particular workers or even groups of workers. The question, instead, should be what kinds of labor law will permit the working class to forge itself into the most cohesive, inclusive, and self-conscious class for itself, capable of independent and autonomous action. Only a fight for labor freedoms can obtain that goal.