In 1936 there were forty-eight sitdown strikes. In 1937 there were 477: electrical workers in St. Louis; shirt workers in Pulaski, Tennessee; broom workers in Pueblo, Colorado; trash collectors in Bridgeport, Connecticut; gravediggers in New Jersey; seventeen blind workers at the New York Guild for the Jewish Blind; prisoners in an Illinois penitentiary; and even thirty members of a National Guard Company who had served in the Fisher Body sit-down, and now sat down themselves because they had not been paid.

The sit-downs were especially dangerous to the system because they were not controlled by the regular union leadership. An AFL business agent for the Hotel and Restaurant Employees said:

You'd be sitting in the office any March day of 1937, and the phone would ring and the voice at the other end would say: "My name is Mary Jones; I'm a soda clerk at Liggett's; we've thrown the manager out and we've got the keys. What do we do now?" And you'd hurry over to the company to negotiate and over there they'd say, "I think it's the height of irresponsibility to call a strike before you've ever asked for a contract" and all you could answer was, "You're so right."

It was to stabilize the system in the face of labor unrest that the Wagner Act of 1935, setting up a National Labor Relations Board, had been passed. The wave of strikes in 1936, 1937, 1938, made the need even more pressing. In Chicago, on Memorial Day, 1937, a strike at Republic Steel brought the police out, firing at a mass picket line of strikers, killing ten of them. Autopsies showed the bullets had hit the workers in the back as they were running away: this was the Memorial Day Massacre. But Republic Steel was organized, and so was Ford Motor Company, and the other huge plants in steel, auto, rubber, meatpacking, the electrical industry.

The Wagner Act was challenged by a steel corporation in the courts, but the Supreme Court found it constitutional-that the government could regulate interstate commerce, and that strikes hurt interstate commerce. From the trade unions' point of view, the new law was an aid to union organizing. From the government's point of view it was an aid to the stability of commerce.

Unions were not wanted by employers, but they were more controllable-more stabilizing for the system than the wildcat strikes, the factory occupations of the rank and file. In the spring of 1937, a New York Times article carried the headline "Unauthorized Sit-Downs Fought by CIO Unions." The story read: "Strict orders have been issued to all organizers and representatives that they will be dismissed if they authorize any stoppages of work without the consent of the international officers…" The Times quoted John L. Lewis, dynamic leader of the CIO: "A CIO contract is adequate protection against sit-downs, lie-downs, or any other kind of strike."

The Communist party, some of whose members played critical roles in organizing CIO unions, seemed to take the same position. One Communist leader in Akron was reported to have said at a party strategy meeting after the sit-downs: "Now we must work for regular relations between the union and the employers-and strict observance of union procedure on the part of the workers."

Thus, two sophisticated ways of controlling direct labor action developed in the mid-thirties. First, the National Labor Relations Board would give unions legal status, listen to them, settling certain of their grievances. Thus it could moderate labor rebellion by channeling energy into elections-just as the constitutional system channeled possibly troublesome energy into voting. The NLRB would set limits in economic conflict as voting did in political conflict. And second, the workers' organization itself, the union, even a militant and aggressive union like the CIO, would channel the workers' insurrectionary energy into contracts, negotiations, union meetings, and try to minimize strikes, in order to build large, influential, even respectable organizations.

The history of those years seems to support the argument of Richard Cloward and Frances Piven, in their book Poor People's Movements, that labor won most during its spontaneous uprisings, before the unions were recognized or well organized: "Factory workers had their greatest influence, and were able to exact their most substantial concessions from government, during the Great Depression, in the years before they were organized into unions. Their power during the Depression was not rooted in organization, but in disruption."

Piven and Cloward point out that union membership rose enormously in the forties, during the Second World War (the CIO and AFL had over 6 million members each by 1945), but its power was less than before-its gains from the use of strikes kept getting whittled down. The members appointed to the NLRB were less sympathetic to labor, the Supreme Court declared sit-downs to be illegal, and state governments were passing laws to hamper strikes, picketing, boycotts.

The coming of World War II weakened the old labor militancy of the thirties because the war economy created millions of new jobs at higher wages. The New Deal had succeeded only in reducing unemployment from 13 million to 9 million. It was the war that put almost everyone to work, and the war did something else: patriotism, the push for unity of all classes against enemies overseas, made it harder to mobilize anger against the corporations. During the war, the CIO and AFL pledged to call no strikes.

Still, the grievances of workers were such-wartime «controls» meant their wages were being controlled better than prices-that they felt impelled to engage in many wildcat strikes: there were more strikes in 1944 than in any previous year in American history, says Jeremy Brecher.

The thirties and forties showed more clearly than before the dilemma of working people in the United States. The system responded to workers' rebellions by finding new forms of control-internal control by their own organizations as well as outside control by law and force. But along with the new controls came new concessions. These concessions didn't solve basic problems; for many people they solved nothing. But they helped enough people to create an atmosphere of progress and improvement, to restore some faith in the system.

The minimum wage of 1938, which established the forty-hour week and outlawed child labor, left many people out of its provisions and set very low minimum wages (twenty-five cents an hour the first year). But it was enough to dull the edge of resentment. Housing was built for only a small percentage of the people who needed it. "A modest, even parsimonious, beginning," Paul Conkin says (F.D.R. and the Origins of the Welfare State), but the sight of federally subsidized housing projects, playgrounds, vermin-free apartments, replacing dilapidated tenements, was refreshing. The TVA suggested exciting possibilities for regional planning to give jobs, improve areas, and provide cheap power, with local instead of national control. The Social Security Act gave retirement benefits and unemployment insurance, and matched state funds for mothers and dependent children-but it excluded farmers, domestic workers, and old people, and offered no health insurance. As Conkin says: "The meager benefits of Social Security were insignificant in comparison to the building of security for large, established businesses."

The New Deal gave federal money to put thousands of writers, artists, actors, and musicians to work-in a Federal Theatre Project, a Federal Writers Project, a Federal Art Project: murals were painted on public buildings; plays were put on for working-class audiences who had never seen a play; hundreds of books and pamphlets were written and published. People heard a symphony for the first time. It was an exciting flowering of arts for the people, such as had never happened before in American history, and which has not been duplicated since. But in 1939, with the country more stable and the New Deal reform impulse weakened, programs to subsidize the arts were eliminated.