The concept of “union security” or “maintenance of membership” — more commonly called the “union shop” — means that being a union member “in good standing” becomes a condition for continued employment. If you cease to be a member of the union, the company is required to fire you. In the postwar era this is usually implemented by the company simply deducting union dues from one’s paycheck.

“Union security” had long been an idea advocated by “business union” leaders as a way of maintaining their cherished dues income despite the ups and downs of member enthusiasm. Top CIO leaders were certainly as fond of this concept as were their brethren in the AFL — but in the early days the companies were, in most cases,[1] simply not willing to agree to this demand in the late ‘30s.

Though the typical union contract nowadays contains some sort of union shop provision, union membership was voluntary under almost all CIO contracts prior to 1942. The dues “check off” was virtually unknown in the late ‘30s and dues were collected on the shop floor by shop stewards and committeemen.

Critics of the earlier industrial union organizing efforts of the Industrial Workers of the World have long pointed to the fact that membership in an IWW branch would often drop dramatically after a period of intense struggle had passed and union membership was reduced to a more committed hardcore.

Yet, the early CIO was subject to the same dynamic. Workers had poured into the CIO industrial unions during the sitdown strike wave of 1937 but many dropped out when the recession of the late ‘30s made it difficult for unions to make gains against the employers. For example, 8,027 workers had joined the UAW local at the Fisher Body plant in Lansing by late 1937, after the sitdown strikes and the first GM contract. But a year later only 1,078 were still paying UAW dues. The entire membership of the CIO had dropped to a mere 1.35 million just prior to Roosevelt’s military buildup on the eve of the U.S. entry into World War II. “To remain solvent,” writes Nelson Lichtenstein, “SWOC had to resort to monthly dues picket lines that cajoled or threatened delinquent workers to pay up before they could enter the mill. Often steelworkers balked, and absenteeism ran as high as 25% on the days dues pickets patrolled mill gates.”[2]

These fluctuations in worker membership are probably inevitable in any mass workers movement that is autonomous, genuinely controlled by the workers themselves, since their participation would reflect their perceptions of what is needed and winnable at the moment.

American leftists typically assume that the “union shop” is definitely a valuable asset to workers. It could be argued that this position implicitly accepts the bureaucratic outlook. For, it assumes that the union has a value independent of its usefulness to the workers themselves and that it should be maintained no matter whether the workforce is sufficiently motivated on its own to keep it going. Moreover, the favorable view of the union shop ignores the role that it played in converting the unions into organs of employer discipline over the workers (see below).

The “open shop” situation of the CIO unions in the late ‘30s meant that local union officials and activists were in the position of having to justify support for the union every day if they wanted to maintain rank-and-file support and dues income. Thus, “grievance battles were the order of the day,” writes Lichtenstein[3], “and local officers went about their jobs in an aggressive and energetic manner. Although all SWOC contracts formally prohibited strikes for the duration of the contract, a form of guerrilla warfare nevertheless continued in the mills.”

So long as the union’s continued existence depended upon voluntary rank-and-file support, the local union organization was under pressure to continually mobilize to get results. Grievances were pursued whether or not they were clearly justified by language in the contract, and stewards or local officers supported slowdowns or short wildcat strikes if they thought they might work.

Even when they didn’t approve of wildcat strikes or other direct action, local union officials were reluctant to condone company repression of such actions. The most active participants were almost always key union supporters in the plants. If they simply abandoned them to the company, the local officials were afraid this would discredit the union in the eyes of the workers.

Once the “union shop” had been achieved, however, the local union organization would no longer be under such immediate pressure to mobilize a constant struggle with the employers in response to worker grievances and concerns.

But how did the “union shop” become so widespread in unionized industries? As the U.S. drifted towards war in 1941, “war hysteria” gripped the Roosevelt administration in Washington, and heavy pressures were put on the CIO leaders to prevent disruption of war production. On the other hand, workers and CIO union activists viewed the war buildup as an opportunity to press for more concessions in shopfloor struggles, and the huge strike wave of 1941 thus threatened to slowdown FDR’s head-long rush towards war.

This conflict came to a head with the Roosevelt administration’s use of military power to crush a strike, a walkout by 4,000 workers in June of 1941 at the North American Aviation plant in Inglewood, California. This was the largest strike in California since the maritime general strike of 1934. A period of shopfloor organizing and worker/employer polarization was brought to a head when the night shift suddenly walked off the job. Even the Communist-leaning officers of the local were caught by surprise. The workers did not resort to a sitdown strike because such actions had just been declared illegal by the Supreme Court; instead, they organized a mass picket line that surrounded the plant.

UAW leader Dick Frankensteen had previously told the workers that they shouldn’t worry about strike authorization since “it’s just a scrap of paper anyway.” However, the unauthorized nature of the work stoppage, and the visible presence of local Communists, were seen by the Roosevelt administration as providing a convenient opportunity for a show of force to discourage strikes in war industries.

Under pressure from the Roosevelt Administration, Frankensteen tried to order the strikers back to work at a meeting outside the plant, but was shouted down. He then put the local union under the control of an appointed administrator. Though he knew low wages, not “Communist agitation,” was responsible for the strike, Frankensteen denounced local Communist activists in a national radio broadcast — a virtual invitation to government intervention.

With the approval of top CIO leaders, troops broke up the mass picketlines around the plant and imposed virtual martial law in the immediate area. In smashing the strike with troops, and threatening strikers with induction into the army, FDR sent a strong message to the CIO leaders that their organizations were at risk if they allowed disruption of his program for imperialist war. The upshot was soon apparent as the CIO leaders capitulated to FDR’s demands for a total “no strike” pledge when the U.S. finally entered the world war.

The “No-Strike” Pledge

The no-strike pledge created a crisis for the CIO union bureaucrats. Who would voluntarily belong to a union that was unwilling to fight? That couldn’t get any results? Before the war, grievances had been settled by direct action methods, as we’ve seen. But with a no-strike pledge, management had less incentive to make concessions on shopfloor issues. The no-strike pledge thus threatened to undermine worker support for the CIO and erode the bureaucrats’ dues base. The CIO leaders then appealed to the National War Labor Board (NWLB) to grant “union security” (belonging to the union as a condition of keeping one’s job) in exchange for the no-strike pledge.

The case of the United Electrical Workers local at Walker-Turner in New Jersey illustrates the problem. After Hitler’s invasion of Russia in June, 1941, the Communist Party suddenly became very strong partisans of the no-strike pledge. Saving the Soviet Union from Hitler’s advancing armies was more important, in their eyes, than fighting for the interests of American workers. Thus, the CP leadership of the Walker-Turner UE local had abandoned efforts to improve wages or fight the bosses on other issues. By late 1941 the local had lost 25% of its members and most of the remaining members were delinquent in their dues. The UE leaders appealed to the NWLB to come to their defense.

Why should the bosses’ government come to the aid of the bureaucrats in this situation? Here we need to look at the other side of the labor crisis during World War II: the erosion of managerial authority and labor discipline.

The War-Time Labor Crisis

The buildup of the armed forces and the massive mobilization of American industry for the war effort created labor scarcity, reducing unemployment to a mere 1.3% by 1943. This situation gave workers more leverage with employers, which was reflected in a variety of ways, such as more people quitting their jobs when dissatisfied. In 1939, 71% of the instances where workers were separated from their jobs were due to layoffs. But by 1943, 72% of job separations were due to voluntary resignations. Between 1941 and 1944 average wages in the U.S. doubled, despite war-time controls.

This situation of labor scarcity, greater availability of jobs, and massive job turnover, due to war mobilization, created a shopfloor environment in the U.S. where it was very difficult for management to maintain the authoritarian discipline it was used to, especially given the practice of worker struggle that had developed during the ‘30s. By the end of World War II auto executives claimed that worker productivity had fallen by as much as 39%.

Workers and local union activists were able to press for greater concessions and worker control at the shopfloor level. For example, at the Dodge Main plant in Detroit aggressive shop stewards turned the union’s formal right to observe time-study procedures into the power to jointly set new rates with management. And in the turbine department at GE’s Erie works, management virtually abandoned control of piecework standards to the workers.

The favorable employment situation, worker efforts to press for greater control, and management efforts to maintain discipline created a situation of intense shopfloor struggle, and the years 1943–45 experienced an increasing level of wildcat strikes, especially in the plants that had been centers of CIO militancy in the ‘30s.

Faced with this crisis, some business and government leaders began to see the potential of the union bureaucracy as a means of enforcing industrial discipline. San Francisco shipping baron Roger Lapham — speaking for the “liberal” wing of the business class — argued:

“Can union leaders be held accountable for labor troubles if because of a falling off in their membership, they find they control a minority rather than a majority in the plants where they are the bargaining agents? If one is realistic, it is hard to reconcile the views of those who wish to hold union leaders responsible for more stable labor relations and yet will not help them in some practical way to attain responsibility.[4]”

“Responsibility” is here a euphemism for “does not actively fight management.” Frank Graham, another FDR appointee on the NWLB, argued, “Too often members of unions do not maintain their membership because they resent the discipline of a responsible leadership. A rival but less responsible leadership feels the pull of temptation to obtain and maintain leadership by relaxing discipline, by refusing to cooperate with the company, and sometimes by unfair and demagogic agitation.”[5]

With Graham taking the lead, the NWLB reached the conclusion that “only a general union security clause in every contract would give labor officials the ‘self-confidence’ and ‘firmness’ to deal with their members and enforce their contracts,” writes Lichtenstein.[6]

In other words, once a “union shop” provision is granted, if the organization refuses to fight and acts to squelch rank-and-file actions, the resulting disaffection of the ranks will not lead to the erosion of membership and decline of the organization because the workers will be forced to still belong to it. Thus, officials will have the “firmness” needed to act against the wishes of the rank and file. With this as their motive, the NWLB issued an edict in June of 1942 granting the union shop to every union that accepted the no-strike pledge.

The NWLB explicitly recognized that the unions could be used to discipline the workforce. When workers at a steel fabrication plant in New England conducted wildcat strikes in 1942, against the advice of the local United Steel Workers leadership, the NWLB held this was just a “spontaneous and unplanned demonstration” and that the union shop would help this union to become “responsible”:

“If this union is to become a responsible organization acting through its leaders, it is necessary that it have some power over its members....The inclusion of a maintenance of membership provision [i.e. union shop] in the contract...will have that result because an employee who elects to be bound by the maintenance clause must remain a member in good standing to keep his job.”

This position was demonstrated on a much larger scale in dealing with wildcat strikes in the Akron, Ohio, tire industry. The head of the United Rubber Workers was one Sherman Dalrymple, supported by an alliance of conservatives and Communists who backed the no-strike pledge and labor/management cooperation during the war. In January, 1944, when a strike of tire-builders at General Tire protested reductions in piece rates, Dalrymple expelled 70 strikers and invoked the union shop clause to have them fired by the company. He also imposed an appointed trustee to run the local union. When two ex-presidents of the local championed the cause of those who had been fired, Dalrymple had them fired, too.

In September of 1944 Dalrymple moved against a strike of workers at US Rubber in Detroit by fining 1,000 members $12.50 each. Most refused to pay and Dalrymple then expelled 572 of them and — again using the union shop clause — demanded that the company fire them. When the company refused to fire these workers, saying this would be too disruptive to production, Dalrymple then appealed to the NWLB to enforce his authority, and the NWLB ordered the fines taken out of the workers’ paychecks. Shortly thereafter 2,000 members of the local attempted to join MESA and conducted a 17-day strike, which was only ended when the army seized the plant.

Right-wing opponents of the union shop, such as the National Right to Work Committee, argue against the “union shop” on the grounds that it is “coercion” of the individual (and we know how the employers have such a tender regard for the rights of the individual!). But the problem with the union shop is not just “coercion” — the heart of the problem is who controls that coercion. Even in an “open shop” situation, workers may use various sorts of peer pressure to discourage actions of individuals that show a lack of solidarity or undermine conditions. (And rightly so — the individual does not have the right to undermine collective conditions by collaboration with our exploiters.) But in that situation the “coercion” is directly controlled by the workers themselves, it is not imposed by an institution outside the direct control of the workers on the shopfloor.

The union shop, on the other hand, is one element in a set of institutional controls on worker’s freedom of association, which hinder the development of association more directly controlled by the rank and file in the shop. If workers could not only drop out of a bureaucratized union, but easily form new workplace-based associations to carry on struggles more directly under their own control, this would undermine the control of officials. Unpopular actions by officials could easily lead to an exodus of workers from the old union to a new association. Apart from the “union shop,” two other restrictions on worker freedom of association are:

the Taft-Hartley Act ban on “jurisdictional strikes” (i.e., using direct action to force an employer to recognize workers’ choice of a new organization), and the institutionalization of “exclusive bargaining rights.”

Bargaining Monopolies

“Exclusive bargaining rights” means that a single organization is granted a monopoly on the right to negotiate with the employers on behalf of a particular group of workers, not defined in terms of who chooses to belong to the organization but in terms of where the plant is located or what type of job they have. Thus, if a workforce is divided into two organized groups that advocate a different approach in dealing with the employers, but one of them wins a majority in a government election, the other group is completely frozen out.

An example of a problem that can arise is the situation that has existed for a number of years among the classroom teachers here in San Francisco. Two groups exist that are aligned with the two separate teachers’ unions that exist at the national level — the American Federation of Teachers and the National Education Association. At one point the courts denied an AFT shop steward the right to pursue a grievance on behalf of a pro-AFT teacher because the NEA group had recently won a labor board election granting it “exclusive bargaining rights.” Without taking sides in this inter-union rivalry, we can see that a group of workers were being denied their right to freedom of association.

This institution of “exclusive bargaining rights” can at times make it more difficult for a militant minority in a particular workplace to have an influence on the course of events, given the monopoly of a more conservative union machine on the legal right to take concerted action and negotiate with employers.

The alternative to “exclusive bargaining rights” would be some sort of system of shared bargaining rights, such as typically exists in southern Europe. From a rank-and-file point of view, the advantage to such a system is that the freedom to leave one union and join another, or set up a completely new one, puts pressures on the unions to fight for workers’ concerns and puts a check on leaders. The existence of competing unions does not prevent worker unity. As shown in Spain and elsewhere when workers periodically unite assemblies to discuss contract struggles or conduct strikes, bringing together members of the competing unions as well as workers not belonging to any union. The point is that unity needs to be developed by the workers themselves; genuine unity is not created top-down, by legal, institutional means.

The U.S. almost did get a system of shared bargaining rights in the ‘30s. In a settlement of workers’ demands in the auto industry, worked out by the Roosevelt Administration in March, 1934, a system of “proportional representation” was provided for in order to give the various competing unions the right to participate in negotiations with the employers.

However, it seems that this idea was discredited among workers at the time because the system was designed so as to give representation to the “company unions” — fake unions that were run by managers or bankrolled by the company. However, the “company unions” would not get workers votes in elections of joint bargaining committees insofar as they became truly discredited among the workforce and insofar as workers had freedom to develop associations they really controlled. And, in any case, “company unions,” in this sense, were outlawed by the Wagner Act in 1935.

Nonetheless, the Wagner Act, in addition to banning “company unions,”[7] also got rid of the system of “proportional representation” in favor of the setup first worked out in a Roosevelt Administration settlement of a dressmakers’ strike in Reading, Pennsylvania, in July of 1933. That settlement — the so-called “Reading formula” — provided for secret ballot elections of “exclusive bargaining agents” to “represent” groups of workers called “bargaining units.”

Top Officials Consolidate Their Control

Championed by the National War Labor Board, the development of centralized grievance systems in the ‘40s, beginning with the UAW’s 1941 GM contract, and the spread of no-strike clauses during the war, contributed to the bureaucratization of the union movement. The use of wildcat strikes to circumvent the slow and bureaucratic grievance procedures was a commonplace in CIO workplaces of that era, however. The NWLB was aware of this rank-and-file resistance:

“The elimination of habits, nurtured in successful practice, from thousands of workers is no overnight task...”[8]

“The NWLB understood that innumerable grievances would arise in the day-to-day life of the workplace,” observes Lichtenstein, “but the board sought to build a system of shop governance that would settle these disputes and at the same time prevent them from either interfering with production or challenging wholesale the necessary authority of shop management. To this end, the board elaborated a system...that removed industrial disputes from the shop floor and then provided a set of formal, bureaucratic procedures to resolve them.” However, the crescendo of wildcat strikes that spread through industry in the last years of the war made the implementation of this goal no easy task. Between 1943 and V-J Day 12% of all American workers participated in wildcat strikes.

As Lichtenstein observes, the de facto practice in the United Auto Workers union, prior to the North American Aviation strike, was to grant sanction to virtually any strike that had a chance of success or the support of influential segments of the union. Under the pressures of war-time, and with the full support and encouragement of the National War Labor Board, the CIO leaders were increasingly willing to act directly against the rank and file in consolidating their own control.

The more spontaneous wildcat strikes in the earlier years of the war tended to be small actions limited to a particular department. Towards the end of the war, however, a number of local CIO unions began to “bunch” grievances and actively coordinate wildcat actions with the result that these tended to become a more organized affair, with mass picket lines and formally articulated demands. Often these actions were directed against the International union as much as against the employer. This reflected the increasing conflict between the movement on the shopfloor and the CIO leaders.

Not long after America’s entry into the world war, the UAW International Executive Board made its first decisive moves to quell the waves of shopfloor wildcat strikes that had been common in the auto industry since the sitdowns of the ‘30s. In February of 1943 the UAW tops enacted a decision to “withhold all services to members” guilty of causing unsanctioned strikes and to prohibit “all intervention on their behalf, in the event of disciplinary action against them by management.”

This decision was first put into effect against workers at Ford in March of 1944. When two veterans were fired for smoking, several hundred aircraft workers surged into the personnel office, roughed up managers and destroyed work records. A further effort to get fired workers reinstated included the use of auto barricades. Though local leaders characterized this as “a spontaneous reaction of the rank and file against inhuman and dictatorial treatment,” the UAW tops successfully ordered those fired in this dispute deprived of any union support.

The consolidation of the bureaucracy, in the UAW and other CIO unions, was bolstered by the union’s increasing resort to top-down seizure of control of local unions to discourage unsanctioned struggles initiated by the ranks. The trusteeship imposed to break the North American Aviation strike was a first taste of things to come. When Chrysler Local 490 organized a week-long strike in defiance of the no-strike pledge in May of 1944, the UAW International suspended all 14 members of the local executive board. A similar dictatorship was imposed that year on a Chevrolet local that struck for 11 days in defense of two members fired by GM.

Bureaucratic domination of the CIO Internationals had been implicit from the very beginning in the hierarchical union constitutions and paid officers which John L. Lewis and the other professional union leaders instituted when they founded the CIO.

But in the early CIO, bureaucratic domination was not yet solidly entrenched. As I said in Part 2, the CIO of the ‘30s was contradictory in that it contained both revolutionary and bureaucratic aspects. The hierarchical constitutions and paid officers at the top were one factor. But, on the other hand, there was a mass movement of workers that showed widespread disrespect for corporate “property rights” and “managerial prerogatives” — through sitdown strikes, mass participation of rank-and-file workers, the daily practice of on-the-job direct action, links between employed and unemployed and other forms of expanded solidarity. In the mass events of daily union practice in the ‘30s, the union hierarchy was not yet as dominant a factor as it would later become.

The institution of the union shop, the “no-strike” pledge, and the development of centralized grievance systems, were major steps on the path towards consolidation of the bureaucratic, top-down control that had been implicit in the hierarchical structures of the CIO.

The mass participation in war-time sitdown strikes eventually merged into the massive postwar strike wave that saw 1946 surpass every previous record of strike activity since 1919. Nonetheless, the postwar strike actions in the various industrial sectors were fought out largely in isolation. A cross-industry alliance was not built although there were community-wide general strikes in several smaller cities. The CIO’s industrial unions failed to overcome the sectoralism that has always been the major weakness of the American labor movement. Ultimately, the CIO merely expanded the definition of a sector from an individual craft to an industry or company.

[1] Ford was an exception. Despite Ford’s long hostility to unionism, they apparently saw the corrupting influence inherent in bureaucratic privilege. The first Ford UAW contract in 1941 granted the union shop and special prerogatives (offices, right to leave the plant, etc.) for union committeemen.

[2] Labor’s War At Home: The CIO in World War II, p. 14

[3] Ibid, p. 22.

[4] Quoted in Labor’s War At Home, p. 75.

[5] Ibid, p. 78.

[6] Ibid, p. 79.

[7] The ban on “company unions” was not always enforced, however. In the ‘60s I worked as a gas station attendant for Standard Stations, Inc. — the large chain then operated by Standard Oil of California (now called Chevron). A naive 18-year-old new-hire in 1964, I signed the form approving dues checkoff for the Western States Service Station Employes Union. (Membership in the union was voluntary.) I thought a union sounded like a fine idea. An unsuspecting rank-and-file worker who went to union meetings would get a rude awakening, however. WSSSEU was run by station managers and assistant managers — a clear violation of the Wagner Act.

[8] Quoted in Lichtenstein, p. 181.