Anxiety that automation is coming for workers’ jobs has reached a fever pitch. But talk of robots replacing humans often conceals a less complicated reality: management uses technology to undemocratically reorganize and intensify labor.
Machining of aluminum pistons at the Ford Motor Company’s Cleveland engine plant through automated equipment, 1955. (Bettmann / Getty Images)
Just last month, IBM’s chief executive warned that AI will automate white collar work, despite the fact that the Bureau of Labor Statistics reports that there is little evidence to support the claim that technology is accelerating job loss. Labor productivity, which should be skyrocketing if machines really were rapidly replacing workers, has been exemplary for below-average growth. This mismatch between automation’s hype and its reality has a history.
In reality, automation has always been more ideology than technology. Coined in the 1940s, the idea of automation was first articulated by managers. Its purpose was to provide rhetorical cover for union busting by depicting human labor as technologically obsolete, even as workers continued to toil under degraded conditions. In the not-too-distant past, some on the Left have called for an automated and workless society, while others have warned of impending doom for working people at risk of replacement by machines.
What unites techno-optimists and dystopians about automation is that both look at the workplace from the perspective of management. From the corner office they equate increases to productivity — made possible by technologies that speed up and intensify the pace and quality of work — with laborsaving technology. But if the history of automation discourse is any indication, the Left has not benefitted from uncritically embracing the notion of automation to understand the investment and managerial decisions of capitalists. A clear-eyed look at the history of the term reveals that behind its rhetorical justifications, automation is simply another way of explaining away the control that capitalists have over the conditions and compensation of workers.
Automation: An Anti-Union Philosophy
In 1950, over a thousand workers struck the Ford Motor Company’s Buffalo stamping plant to protest a factory-wide speedup that forced workers to complete more tasks per hour than in any other factory in the Ford empire. In many ways, the strike was unremarkable; speedup was the rule in the postwar automobile industry, and workers everywhere hated it. What set the action apart was the fact that, in theory, no one was supposed to be working in the Buffalo plant at all. According to Ford, the factory possessed a revolutionary new technology: automation.
The word automation stood in for a story: that all technological development inevitably bent toward the elimination of human labor.
In 1950, the word automation was brand new, and like the Buffalo plant, it was a Ford product. The word’s first public use dated from 1947, when Ford opened its Automation Department. There, managers used automation to describe the introduction of the transfer machine, a mechanism that bored holes in engine blocks.
However, the transfer machine was not a new technology; it dated from the 1880s. What was actually new at Ford in 1947 was the industrial labor movement, in the form of the United Auto Workers (UAW). Famously hostile to organized labor, Ford had only recognized the union in 1941, and was the last major American automobile manufacturer to do so. Forced to bargain with the UAW, Ford executives sought to limit the union’s power, and they used the widespread technological enthusiasm following World War II to do so.
After the war, American industry appeared not only to have overcome the economic disaster of the Great Depression (“The capitalist dog has returned to his old vomit,” the historian Lewis Mumford rued in 1944), but also to have issued victory ready-made from its factories and laboratories. Breakthroughs like the development of the electronic digital computer and the splitting of the atom seemed to promise a new world where technology could solve even the most difficult problems of human existence. Ford vice president of production D. S. Harder called automation “a new concept — a new philosophy — of manufacturing.” According to management, technological progress itself, and not the profit motive, required that shop-floor power shift away from the worker (along with the unions employers were now legally compelled to recognize) and into the hands of the engineer and the manager.
The word automation stood in for a story: that all technological development inevitably bent toward the elimination of human labor, and that production would demand more technical skill than physical effort from the workforce of the future.
Automation and the Degradation of Work
So compelling was this new narrative of automation that even union officials took it for granted that the organization of the workforce had become as much a technical, as a political, issue. Deferring to the discourse in 1955, Walter Reuther, president of the UAW, remarked that “we stand on the threshold of a development of science and technology that will enable us to solve the problems that have plagued the human family for these many thousands of years. . . . We welcome any development that will lighten the burden of human labor.”
The rest of the automobile industry would follow Ford, using the pretext of automation not to abolish human labor, but to degrade and cheapen it.
But what management and union leaders called automation, workers described as a speedup.
When Soviet engineers visited Ford’s “automated” engine plant in Cleveland (which employed seven thousand people), they reported that the machinery was standard, but that the speed with which workers labored was unparalleled. Watching workers, they said, was “like looking at a speeding motion picture.” The work was degraded in other ways, too. The Cleveland engine plant paid workers, on average, eleven cents less an hour than their coworkers at the famous River Rouge facility in Dearborn, Michigan. “[A]utomation and down-grading as a result of job-dilution had gone hand-in-hand at the Cleveland plant,” the UAW found. According to labor organizer Nat Ganley, despite the fanfare around “the most modern push button machines, the most wonderful technical improvements,” workers were “compelled to turn out more production on the same machines, and the same job, than they were a dozen years ago.”
The rest of the automobile industry would follow Ford, using the pretext of automation not to abolish human labor, but to degrade and cheapen it. In the words of Simon Owens, a longtime worker at Chrysler, “automation has not reduced the drudgery of labor. Whatever Automation means to management, labor bureaucrat, or engineer, to the production worker it means a return to sweatshop conditions, increased speedup and gearing the man to the machine.”
Managers sped up and degraded black workers with particular intensity. Leaders of the League of Revolutionary Black Workers found that for “the black worker the pressure of production never ceases . . . . [they] are now producing at least twice as much as auto workers twenty years ago.” In particular, because of the “super exploitation of black labor, profits in auto have soared. . . . Often new black workers are forced to do the work of two white men.” (The postwar automation discourse in fact universalized an older, racist argument, applying it to the entire industrial working class: white supremacists a century earlier had depicted black labor as necessarily menial, and therefore prone to mechanization. Slavery and steam, said a former US Commissioner of Patents in 1859, were practical equivalents.)
None of what was going on at Ford was unprecedented in the history of industrial capitalism. When contemporary economists write about automation, it is usually as shorthand for labor productivity, or the amount produced per hour worked. The more productive labor, the argument goes, the more work vested capital — technology — must be contributing to production. There were certainly increases in productivity during the postwar period. Machines can increase productivity by speeding up labor, intensifying it, and even, at times, yes, replacing it. For that matter, managers compelling workers through nonmechanical means — like coercion, fear of unemployment, or wage theft — also wrest more “productivity” out of their employees. The postwar automation discourse, however, was distinct from the specific dynamics of capitalist competition that drive the relentless transformation of productive forces.
Since the dawn of industrialization, technological changes have, paradoxically, both increased productivity and generated work under newly degraded conditions. Increases in productivity have not resulted in less human effort, because the “work” of capitalism is limitless, and no breakthrough in productivity will persuade capitalists to let a resource to sit idle. It will both overuse human labor in one place and let it go to seed somewhere else; it will countenance huge reserves of untapped human potential for a time while working it to the bone moments later.
Rather than a general decline in human labor and the rise of a leisure society — a through line to some endgame where human labor will be pointless — we see a circular trend, where capital rushes from place to place, exploiting labor for a time, using it up, and then moving on to use it somewhere else. (“The bones of the cotton-weavers are bleaching the plains of India,” the British statesman William Bentinck averred in the 1830s, as English factory owners mechanized textile production and accelerated the expansion of English wage laborers, that proletariat of proletariats.) This fact is perhaps the most important difference between the enlargement of humanity’s powers to manipulate the natural world, and the realities of working under conditions of capitalism: no amount of power is ever enough.
By the 1950s, employers across the United States deployed the automation discourse to degrade working conditions. At the same time, union leaders, unable or unwilling to resist claims that human labor was becoming obsolete, attempted to bargain under the terms of automation, often with poor results. When the International Longshore and Warehouse Union (ILWU) embraced the containerization of shipping in the infamous 1960 Mechanization and Modernization Agreement, it signed off on the creation of a bifurcated workforce of A-men who enjoyed full union benefits and B-men who did not, foreshadowing what by the turn of the twenty-first century would become typical in unionized industrial workplaces.
Rather than a decline in labor and the rise of a leisure society, we see a circular trend, where capital rushes from place to place, exploiting labor for a time, using it up, and then moving on to use it somewhere else.
Similarly, John L. Lewis of the United Mine Workers (UMW) supported the introduction of the continuous miner in the 1950s under the assumption that technological change would lead to better-paid, secure jobs. It did not. And the United Packinghouse Workers of America (UPWA) understood the introduction of power tools to the disassembly line as “automation” (and therefore inevitable), only to discover in the early 1960s that the Automation Committee set up between the Amalgamated Meat Cutters (AMCBW), the UPWA, and Armour and Company had been, in the words of the UPWA’s president, a “façade of humaneness and decency that would conceal a ruthless program of mass termination of employees of long service and cynical manipulation of the natural fears of its employees to accomplish drastic cuts in wages and working conditions.”
Today, far from laborless, the meatpacking industry depends on an army of workers, whose jobs are in fact among some of the most fast-paced and dangerous in the United States.
The Left Adopts the Automation Discourse
While it may have begun as a story told by management, the automation discourse appealed to many on the American left. While classic liberal political thought had held that a free person might work, so long as they worked their own property, the mass dispossession of ordinary people through industrial capitalism now meant that the vast majority worked for wages on property they did not, and would never, own. The automation discourse solved this problem by erasing half of the equation: workers did not need to overcome the alienation caused by industrial modernity by finding freedom in their work, because control over the organization and distribution of labor was decided by the constant progress of science and technology.
But when left tendencies folded the idea of “automation” into their politics, the result was often that the labor process itself, as well as the workplace, lost a great deal of political importance in their analyses and activism. Students for a Democratic Society (SDS) set up cells in northern cities to organize what they believed would be masses of white youth unemployed by automation. When they failed to appear, SDS adopted the “New Working Class” theory, calling for its members on college campuses to organize what it believed would be the workers of the future — skilled white collar professionals who would service the machines that replaced industrial workers. (The idea has had staying power as a liberal nostrum, with centrists calling for job-training and “learn-to-code” panaceas under the mistaken assumption that technological change inevitably creates demand for skilled labor.)
James and Grace Lee Boggs, closer to the labor movement than SDS, argued that “automation” would put the lever of revolution into the hands of black Americans by upskilling the white working class and marginalizing black workers to low-paying “scavenger” jobs. People who sought to bring about the revolution by organizing the workplace, the Boggses argued, had misplaced their energies. “These militants who are so advanced,” James Boggs wrote, “are really behind the average worker who has reconciled himself to eventual oblivion.”
Among some radical feminists, the allure of “cybernation” (a late-1960s synonym for automation) premised women’s liberation on an escape from biological reproduction. According to Shulamith Firestone, biology was the cause of sexism, and to liberate women from oppression, society needed first to “free humanity from the tyranny of its biology.” The demands of the body itself, rather than social relations, needed to be mechanized away. A political revolution, according to this view, required a technological one, shifting the site of struggle from the present to the future.
When left tendencies folded the idea of ‘automation’ into their politics, the result was often that the labor process itself, as well as the workplace, lost a great deal of political importance in their analyses and activism.
While Firestone called for escaping reproduction, the National Welfare Rights Organization (NWRO) took a different tack. It called on the federal government to compensate mothers for their labor raising the next generation of Americans. Rather than dismiss reproductive labor as inherently degraded, activist Johnnie Tillmon called on President Richard Nixon to issue “a proclamation that women’s work is real work.” If motherhood was properly acknowledged and remunerated, it could be made coherent with a liberated subject. The NWRO, at least, could imagine a social revolution that did not require a technological breakthrough ushered into existence by capital.
A misplaced faith in technological process left both unions and revolutionaries intellectually disarmed. It absolved them of the admittedly difficult task of imagining how to reconcile work and freedom. This remains the task of the Left today, which like the rest of the culture has (understandably) become increasingly dystopian. Technological dystopianism, like technological utopianism, is a kind of fetish and a distraction. It treats capital like a capitalist would — as the decisive force in history.
We do not require technological wonders and an escape from labor to make our world more just. A left politics of work must call not for new powers, but for a more equitable sharing of those powers it currently possesses. That isn’t a question of what the machines will do tomorrow, but what people should do today.