SPRING 2026
ISSUE 06

₁ For Part 1, see Long-Haul, no. 2 (Spring 2025), available in print and online.

₂ Bruce Lee, “The GM–Toyota Team: Worker Harmony Makes Nummi Work,” New York Times, December 25, 1988.

₃ Lee, “The GM–Toyota Team.”

₄ Paul S. Adler, The “Learning Bureaucracy”: New United Motor Manufacturing, Inc. (Center for Effective Organizations, 1992), 11.

₅ Mike Parker and Jane Slaughter were both longtime members of International Socialists (IS), an unorthodox Trotskyist organization with roots in the Bay Area. Like many members of IS, they “industrialized” in the 1970s, getting jobs in heavy industry – although not at GM-Fremont, where some IS members became active – in hopes of organizing rank-and-file workers at the point of production. Parker and Slaughter went on to publish the journal Labor Notes.

₆ Mike Parker and Jane Slaughter, Choosing Sides: Unions and the Team Concept (South End Press, 1988), 16.

₇ Lee, “The GM–Toyota Team.”

₈ Lee, “The GM–Toyota Team.”

₉ Henry Weinstein, “Fremont UAW Members Buffeted by Forces They Can’t Control,” LA Times, September 5, 1983.

₁₀ Carol Brydolf, “Ex-GM Workers’ Bad Times Linger,” Oakland Tribune, February 2, 1983.

₁₁ “GM, Toyota Deny Deal Struck,” The Pantagraph, June 13, 1982.

₁₂ In the United States, the three large auto manufacturers are General Motors, Ford, and Stellantis (formerly Chrysler).

₁₃ Toyota labor relations specialist Mike Furuhashi, quoted in Adler, The “Learning Bureaucracy,” 11.

₁₄ Fred Neugent, “Life Inside New GM–Toyota Plant in California,” The Militant, February 15, 1985.

₁₅ Parker and Slaughter, Choosing Sides, 16. One aspect of TPS was notably absent in the American context: job security. See Beverly J. Silver, Forces of Labor: Workers’ Movements and Globalization Since 1870 (Cambridge University Press, 2003), 67.

₁₆ Parker and Slaughter, Choosing Sides, 17.

₁₇ Parker and Slaughter, Choosing Sides, 24.

₁₈ Parker and Slaughter, Choosing Sides, 116.

₁₉ Harry Braverman, Labor and Monopoly Capital: The Degradation of Work in the Twentieth Century (1974; repr., Monthly Review Press, 1998), 81.

₂₀ Braverman, Labor and Monopoly Capital, 96.

₂₁ Karl Marx, Capital: A Critique of Political Economy, Volume 1 (Penguin, 1976), 482.

₂₂ Neugent, “Life Inside New GM–Toyota Plant in California.”

₂₃ Takemae Eiji, Inside GHQ: The Allied Occupation of Japan and its Legacy (Continuum, 2002); Robert Brenner, The Economics of Global Turbulence: The Advanced Capitalist Economies from Long Boom to Long Downturn, 1945–2005 (Verso, 2006); Susan Cockerill and Colin Sparks, “Japan in Crisis,” International Socialism 2, no. 72 (1996); Rafael Aguayo, Dr. Deming: The American Who Taught the Japanese About Quality (Carol, 1990). Japanese trade unions survived the Red Purge as highly acquiescent “partners” with management. Toyota cars are still produced by the nearly 800,000 member Confederation of Japan Auto Workers. Japan’s labor force is one of the least strike-prone of any OECD country.

₂₄ Satoshi Kamata, Japan in the Passing Lane: An Insider’s Account of Life in a Japanese Auto Factory (Counterpoint, 1983), 42.

₂₅ Kamata, Japan in the Passing Lane, 54.

₂₆ James P. Womack, Daniel T. Jones, and Daniel Roos, The Machine that Changed the World (Rawson Associates, 1990), 83.

₂₇ Adler, The “Learning Bureaucracy,” 4, 50.

₂₈ John Holusha, “For Good or Bad, Autoworkers Get a Taste of the Future,” Dayton Daily News, February 5, 1989.

₂₉ Holusha, “For Good or Bad.”

₃₀ “Union Begins New Life Based on Cooperation, Not Confrontation,” Dayton Daily News, March 1, 1987.

₃₁ Michelle Levander, “Labor-Management Cooperation Sputtering at Calif. NUMMI Plant,” Detroit Free Press, June 28, 1991.

₃₂ Lund was a lifelong Trotskyist militant. She became involved with the Socialist Workers Party (SWP) while attending college in Minnesota in the late 1960s. She relocated to Columbia University, where she was deeply involved in campus anti-war activity, working closely with David Gilbert and others who would go on to form the Weather Underground. Over the course of the 1970s, Lund and her partner Barry Sheppard rose high in the ranks of the SWP. At the party’s urging, she industrialized in the early 1980s (the SWP’s industrial turn came much later than most Left groups of the era). Toward the end of the decade, she and Sheppard resigned from the party over deep strategic disagreements and relocated to the Bay Area, where Lund took a job at NUMMI. She remained an independent socialist agitator at the plant, publishing the Barking Dog and serving as an elected trustee of Local 2244, until her death in 2006. In 2014, Sheppard compiled Lund’s newsletters into a book, The Barking Dog: A Collection of Auto Plant Newsletters.

₃₃ “Oppression by Intimidation,” The Barking Dog, August 2, 1999; “Puts Heart into Work,” The Barking Dog, December 7, 2001; “Speed-Up and Cut Jobs,” The Barking Dog, March 10, 2003.

₃₄ Labor Video Project, “The State of the Union: UAW 2244 NUMMI Toyota–GM Workers Fed Up,” posted February 4, 2010, by laborvideo, YouTube, 11 min., https://www.youtube.com/watch?v=B3yxVuLi2Sk.

₃₅ Labor Video Project, “The State of the Union.”

₃₆ Labor Video Project, “The State of the Union.”

₃₇ Walter Isaacson, Elon Musk (Simon & Schuster, 2023), 114.

₃₈ Isaacson, Elon Musk, 220.

₃₉ Paolo Confino, “During ‘the Most Concentrated Pain’ in His Life, Elon Musk Saved Tesla and Created ‘the Algorithm’ that Would Become His Manufacturing and Management Philosophy,” Fortune, September 13, 2023.

₄₀ Confino, “During ‘the Most Concentrated Pain’ in His Life.”

₄₁ Isaacson, Elon Musk, 282.

₄₂ Isaacson, Elon Musk, 281.

₄₃ Black former employee Owen Diaz took the company to court over the treatment he endured at the Fremont Factory, settling for an undisclosed sum after a jury initially awarded him $137 million.

₄₄ “Fired By Elon Musk for Unionizing at Tesla,” More Perfect Union, June 4, 2021.

₄₅ John Shook, “Was NUMMI a Success?” Lean Enterprise Institute, September 14, 2009.

₄₆ Matt Ray and Matthew Wranovics, “The Worst Among the Bad: Auto Workers and Class Power in Fremont, California (Pt 1),” Long-Haul, no. 2 (Spring 2025).

On Christmas Day, 1988, readers of The New York Times opened their morning papers to a startling announcement. “The workers’ revolution has finally come to the shop floor. The people who work on the assembly line have taken charge,” a full-page op-ed blared. “The revolution isn’t happening everywhere – yet. But it is happening at NUMMI.”² Bruce Lee, the western regional director of the United Auto Workers (UAW), took to the pages of the Times to defend his union’s “historic” achievement in setting “productivity and quality levels that exceed anything in the American auto industry.”³

Four years after the transformation and reopening of General Motors’ Fremont, California factory as a GM–Toyota joint venture called New United Motor Manufacturing Inc. (NUMMI), a revolution had taken place. The plant was transformed from the “worst plant in the world” to the most efficient auto manufacturing facility in the United States.⁴ Industry executives celebrated NUMMI as the proof of concept for importing Japanese management techniques to the United States. The old hierarchical American system reduced workers to mere cogs, generating conflict between management and a miserable, unproductive workforce prone to destructive slowdowns, walkouts, and strikes. The new Japanese framework promoted equality, harmonious relations, and mutual respect, valuing workers’ intelligence and contributions. In return, satisfied Japanese laborers had rewarded companies like Toyota with world-leading quality and productivity. NUMMI demonstrated that US workers, if managed correctly, could match their Japanese counterparts.

Lee’s piece in the Times was a response to Mike Parker and Jane Slaughter’s recently published Choosing Sides: Unions and the Team Concept (1988). Parker and Slaughter were militant shop-floor intellectuals who had a history of bruising conflict with UAW officialdom.⁵ Published by the small left-wing cooperative South End Press, their book made unusually large waves for an anti-capitalist text written by avowed revolutionary socialists. Meticulously researched and sharply argued, Choosing Sides cut through an otherwise strictly laudatory national discussion about NUMMI. Where mainstream accounts of Toyota’s management revolution described it as humane and democratic, Parker and Slaughter characterized the Toyota Production System (TPS) as an experiment in “management-by-stress” (MBS).⁶ Far from signaling a new era of cooperation between labor and management, MBS was merely a novel form of labor discipline, a new advancement by capital in its age-old war on workers’ power, with NUMMI as its vanguard.

To labor bureaucrats like Lee, Choosing Sides was a declaration of war. Defending the plant’s achievements, he wrote that NUMMI’s “spectacular success . . . is due to a revolutionary team production system run by the workers themselves. The system was worked out jointly with the plant’s Japanese and American managers by the third party in the venture – the UAW. And it is the UAW’s rank-and-file workers who are making the team system work.”⁷ For Lee, the wave of 1970s plant closures that shuttered GM-Fremont had proven the futility of “‘traditional’ us-against-them factory practices. Arrogant managers and angry workers, constant battles over the details of every worker’s job assignment, a maze of restrictive job classifications barring one worker from doing another’s job, crippling absenteeism, firings, strikes, sick-outs, sabotage.”⁸ The very survival of American organized labor was contingent on a restructuring of the labor–management relationship. Unions had either to make company profitability their highest priority or die.

The techniques of labor discipline introduced by Toyota in Fremont proved highly successful. In the 19-year history of GM-Fremont, its workers struck at least nine times, five of which were “unauthorized” wildcats. In NUMMI’s 25 years, they struck only once – for ninety minutes. These techniques, pioneered by Japanese firms but Americanized in Fremont, are now standard across much of the US economy. The fingerprints of the TPS are visible everywhere from the fashion industry to the healthcare system to the ever-growing logistics sector. NUMMI’s legacy is a US working class with less free time, less power, less autonomy, and more work.

Ultimately, Lee’s vision for organized labor’s future would prove illusory. The UAW did not cruise to success by hitching a ride with Japanese capital. Instead, Toyota, having used the UAW and GM to establish itself in American manufacturing, would leave the union behind as it expanded across the country. To this day, only Fremont workers have burdened Toyota with the inefficiencies and headaches of a UAW contract. In 2009, GM – which by then accounted for only a fraction of the cars produced at NUMMI – pulled out of the joint venture amid the chaos of the Great Recession. Unsurprisingly, Toyota wasted no time shuttering its only union plant in the United States. NUMMI closed for good the following year before Toyota effectively gifted the facility to a fledgling tech company-cum-boutique electric auto manufacturer: Silicon Valley-based Tesla Motors.

Under Tesla, another shift in management style would take place at the former GM-Fremont. Most significantly, the plant would operate without union labor for the first time. A precisely managed veneer of labor harmony would be replaced by open, sneering violation of federal labor, civil rights, and environmental law. Toyota’s enlightened managerialism at the corporate level would give way to a highly personalized shop-floor despotism. While a great deal of the carefully crafted myths and industry hype surrounding one-time liberal darling Tesla has recently collapsed under the weight of “technoking” Elon Musk’s public forays into reactionary politics, the myth of NUMMI’s “workers’ revolution” remains unchallenged business school gospel.

*

When General Motors cut out of Fremont in April 1982, UAW Local 1364 stayed behind. The local bore little resemblance to the mess of dysfunctional, drunken entitlement depicted in popular accounts of GM-Fremont. Decades of shared struggle, internal debate, and political ferment had produced a durable institution its members boasted was “the most vibrant local [in] the US,” a “symbol of progressive, racially-integrated grassroots unionism.”⁹ Longtime auto workers, cast unceremoniously into the ranks of the unemployed, continued to haunt their union hall directly across the street from the idled factory to vent shared feelings of abandonment and await better news.

A cloud hung over the laid-off auto workers of Fremont as it did over other communities suffering plant closures in the Midwest and elsewhere. Many sank into severe depressions and alcoholism. In the months following GM’s departure, eight killed themselves. “My doctor says I should see a psychiatrist,” one union member explained to a journalist while sipping a rum and coke, “but what I really need is a job.”¹⁰

Other than the continued existence of their local, one thing buoyed Fremont auto workers’ hopes. From almost the moment GM-Fremont closed, rumors swirled about a potential reopening of the plant under the auspices of an unprecedented collaboration between GM and its fiercest international rival, Toyota. Talks between the two companies to “discuss the possibility of cooperation in production of small cars in the United States” had been made public in March, the month before Fremont Assembly went dark.¹¹

Over the previous decades, American automakers had increasingly lost their grip on the domestic market to foreign, and especially Japanese, competitors. Japanese cars were cheaper, more efficient, and of a consistently higher quality. Outcompeted, American automotive capital – traditionally committed advocates of free trade – cried for government protection. After Japanese auto companies rebuffed a formal request by the UAW that they open plants in the United States, the union joined the employers’ campaign for import restrictions on Japanese vehicles. In 1981, newly elected president Ronald Reagan took action, reaching a “voluntary export restriction” agreement with the Japanese government that protected auto workers by safeguarding their bosses’ profits.

Facing these import restrictions – and a threat of worse to come – Toyota recognized the need to manufacture cars on American soil and began searching for an established American company to help give them a foothold in US manufacturing. In exchange, they would promise to teach their hosts the vaunted “Toyota Way,” enabling them to efficiently produce the high-quality small cars that had allowed Japanese automakers to so rapidly outpace the Big Three.¹² After preliminary negotiations with Ford broke down, Toyota turned to GM. Fremont, the largest American plant on the Pacific Rim, quickly became the focus of negotiations. But there was one major sticking point: the union.

Toyota was hesitant to work with any American labor union, but GM, knowing it could not afford to reopen Fremont with a non-union labor force, had been careful to include UAW international leadership as a third party in the talks. Toyota was eager for a deal but balked at the idea of employing members of “the UAW Local 1364 . . . the terrible, the worst, the most militant UAW in the United States.”¹³ The union was determined not to allow the workers that had produced the Black Panther Caucus and the 1970 riot to become a roadblock to any pact with Toyota. Under Bruce Lee’s leadership, the UAW was careful to negotiate without the interference or involvement of the Fremont local. With a deal to reopen the plant imminent, the UAW stripped Local 1364 of its charter. The local sued the UAW, challenging the removal of their charter, the seizure of their assets, and the international’s failure to consult Fremont workers on the terms of the GM-Toyota-UAW deal. A federal judge dismissed the suit.

In exchange for recognition from Toyota, the UAW agreed to relinquish a host of rights. Former GM-Fremont employees would be allowed to reapply for their jobs sans seniority status. If hired, they would become members of newly chartered Local 2244. Many workers found this reapplication process humiliating.

To be hired as a production worker, you must do the following: First, mail an application to NUMMI which requires a detailed report on how many days you missed at work for the last 10 years and why. Then, there is the first of two physicals. There are also a couple of hours of assembly tests. There is a nine-hour workshop and evaluation, which includes a series of slideshows and lectures on how NUMMI will be operated. It includes a lot of promises that NUMMI will be based on “mutual trust” – not confrontation – between labor and management . . . Crucial to getting the job is using the vocabulary and approach taught in the workshop.¹⁴

NUMMI’s new vocabulary did more than communicate new concepts. It demanded a fundamental change in workers’ self-conception. Any language that suggested workers were a group with interests separate from management was to be eliminated. The new contract removed references to “grievances,” instead substituting the notion of “problem solving.” A worker was no longer a worker but, in a now-familiar turn of phrase, a “team member.”

In the months leading up to the plant’s reopening, GM selected 250 formerly laid-off union members to serve as “team leaders.” These workers were flown in groups of 30 to Toyota City, the corporation’s sprawling company town in Japan’s Aichi Prefecture, to receive intensive one-on-one coaching in the Japanese art of auto manufacturing.

*

Although it is often reduced to the “quality concept” or the “team concept,” the TPS cannot be boiled down to a single idea or practice. In Choosing Sides, Parker and Slaughter delineate seven closely related key practices that together define the system:

1. Speed-up – ways for workers to do more in less time.
2. “Just-in-time” (JIT) organization of inventory and production.
3. Extensive use of outside contracting.
4. Technology designed to minimize indirect labor.
5. Design-for-manufacture – products specifically designed to reduce labor costs.
6. Methods to reduce scrap and rework.
7. Tighter management control.¹⁵

Of these concepts, perhaps the most central is JIT. JIT production (sometimes called “lean production”), is a “demand-pull” system, meaning that no operation is performed until a step further along in the production process demands it. In this way, no excess material is produced nor unnecessary operations performed, nor is any extra inventory held for emergencies. If one operation in the process fails, none of the subsequent operations can be performed, bringing production to a halt. While it may seem counterintuitive, this fragility is the point: “When a single point experiences trouble of any kind – whether difficulty meeting production or quality – there is no hiding. It becomes instantly apparent to all and is likely to affect operations far beyond the immediate trouble spot. Management at all levels will focus attention on the weak spot.”¹⁶ To facilitate the immediate isolation of problems, Toyota equips its assembly lines with the andon cord. When pulled, the andon cord brings the entire line to a grinding halt – an unthinkable sin in classical Detroit-style auto plants, where the constant forward motion of the line trumps all other concerns, including injury and death. By eliminating slack, keeping the system as taut as possible, JIT allows manufacturers to quickly identify and correct errors or defects before they can become chronic. This enables the continuous improvement (kaizen) of the production process. The ultimate goal of TPS is, simply, to reduce variation in output by eliminating waste (muda) and standardizing production processes down to the smallest detail.

TPS intentionally exerts maximum stress on a given production process. But as well as wasted material, muda also refers to wasteful activity on the part of the worker. Any movement of the human body not strictly necessary for production ought to be eliminated. Through careful observation and statistical analysis, each task is rigidly “charted.” There exists a scientifically determined set of physical steps to perform an operation as efficiently as possible, which workers are expected to memorize. All workers on a team are required to periodically rotate jobs, which means they must be able to easily and efficiently perform each other’s tasks. “Management calls it ‘multiskilling,’” Parker and Slaughter write,

but this is a misleading term. The abilities required in performing several related jobs of very short duration which have been carefully broken down by the charting process are manual dexterity, physical stamina, and the ability to follow instructions precisely . . . These are not ‘skills’ in the usual sense . . . The essence of ‘multiskilling’ is actually the lack of resistance, on the part of the union or the individual worker.¹⁷

As this system requires maximally interchangeable workers, hiring in Toyotist firms is relatively indifferent to an applicant’s past work experience or skills, instead emphasizing the candidate’s attendance and disciplinary histories.

The element of TPS that received the most glowing press in the United States was the team concept. In the Toyota Way, workers are divided into small “teams” with an appointed worker-leader. The team concept was codified in the NUMMI contract: “Employees will be organized on teams of 5–10 members. All members of a team share responsibility for the work performed by the team, and for participation in quality/productivity improvement programs such as QC [Quality Control] Circles and kaizen.”¹⁸ Poor performance by any single team member becomes the responsibility of the entire team. Coworkers exert pressure on each other not to fall behind, and workers exert pressure on themselves to avoid letting their coworkers down. This highly stressed system leads workers to self-discipline.

In QC Circles, team members are encouraged to share their observations and criticisms of work processes and suggestions for how tasks can be improved with managers, who may choose to implement employee suggestions. The solicitation of worker input was a novel development in industrial relations. The “fundamental” idea that “improvement of work methods by workers brings few benefits to management” was at the heart of Frederick Taylor’s scientific management revolution, which was premised on a total division between mental and manual labor.¹⁹ TPS breaks with Taylorism on this question, due less to an inherently Japanese belief in the dignity of individual workers and more to a recognition by capital that, as Harry Braverman writes, because “workers are not destroyed as human beings . . . their critical, intelligent, conceptual faculties, no matter how deadened or diminished, always remain in some degree a threat to capital.”²⁰

Toyota wagered that it would be better to try and appropriate those faculties. Workers’ intimate, highly particular understanding of their own work has always been a source of profound anxiety for bosses, who are well aware of the potential leverage that superior knowledge provides labor. In TPS, workers are incentivized to transfer that knowledge to management as soon as it develops, mitigating an unavoidable hazard of the capitalist mode of production. QC Circles allow workers’ knowledge to filter upwards without decision-making power filtering downwards. Under no circumstances are workers permitted to implement their own ideas; they can only do so when given the order by management. The division of manual and mental labor, in which “the worker is brought face to face with the intellectual potentialities of the material process of production as the property of another and as a power which rules over him,” remains fully intact.²¹

The interlocking principles of the TPS are often grouped together under the broad heading of “Japanese management techniques.” In the 1980s, this branding gave the concepts an attractive sheen in capitalist circles while encouraging workers to redirect their understandable anger in racist and xenophobic directions. But despite its exotic veneer, “Japanese management” wasn’t exactly a new import. If anything, it was returning home. “While the terms are Japanese,” a Fremont worker associated with the Socialist Workers Party rightly pointed out, “the ideas originated in the 1950s from U.S. management consultants sent to Japan.”²²

The “quality concept” that would later become synonymous with Toyota was first systematically elaborated by Walter A. Shewhart, a statistician for Bell Laboratories, the research and development arm of AT&T. In his Economic Control of Quality of Manufactured Product (1931), Shewhart argued that variation in industrial output could be dramatically reduced by rigidly standardizing production and work processes. The real-life application of Shewhart’s theories, developed in the theoretical hothouse of Bell Labs and academic statistical journals, would require a thoroughgoing transformation of the relationship between labor and capital. The postwar United States, where organized labor still held considerable clout, was hardly the place for such an experiment. But one of Shewhart’s most committed disciples, fellow Bell Labs statistician W. Edwards Deming, would be granted a much more hospitable testing ground when, in 1950, he was appointed assistant to General Douglas A. MacArthur, the Supreme Commander of Allied Powers (SCAP) in Tokyo, and de facto dictator of postwar Japan.

In the early years of the American occupation, MacArthur had encouraged the legalization and growth of Japanese labor unions and left-wing parties as part of the democratization process. But the onset of the Cold War and the precipitous rise of the Japanese Communist Party in the late 1940s led to a sharp change in direction from SCAP. The “reverse course” included the dismantling of major strikes, the suppression of student protests, and the firing of tens of thousands of suspected communists from public and private employment. In the aftermath of this “Red Purge,” which was much more thorough than the McCarthyite witch hunts in America, the Japanese labor movement was little more than a shell of itself.

In this context, MacArthur invited Deming to help rebuild the country’s industry with cutting-edge American management science. In Japan, Deming found a national bourgeoisie far more eager to learn new tricks than its American counterpart, and a defeated proletariat without institutional means to resist. With full support from MacArthur, Japanese industry rapidly remade itself in Deming’s image, and the “quality concept” took center stage in the postwar economic miracle. To this day, the Japanese state awards outstanding businesses with the Deming Prize. It was in this new arrangement, and on ground cleared by tremendous inter-class violence, that Toyota developed its “humanistic” production system.²³

It is unlikely that workers from Fremont ever had the chance to speak directly to their Japanese counterparts without the mediation of corporate managers. If they had, they may have had different expectations about their futures under Toyota. Japanese writer Satoshi Kamata’s diaries of his time on a Toyota assembly line in the early 1970s – published in English as Japan in the Passing Lane: An Insider’s Account of Life in a Japanese Auto Factory (1983), shortly before NUMMI opened – offer an alternative perspective to that found in the company’s promotional materials. Far from the creative, autonomous, multiskilled work promised by the team system, Kamata describes an atmosphere of stultifying rigidity. “Here at Toyota,” he writes,

“‘to learn’ means merely to acquire the ability to keep up with the line, quickly repeating physical movements while thinking absolutely nothing . . . If we don’t make the precise motions we’re ‘taught,’ it’s absolutely impossible to do the required work in the required time. Under such a system, all our movements must become mechanical and habitual. Only if we stop thinking and unconsciously follow the system can we keep up.”²⁴

In a particularly evocative passage, Kamata explores the psychological dimension of life on a Toyota assembly line.

What do the other workers think of this monotonous work as they repeat it in cycles of a minute and twenty seconds? I asked Takeda what goes through his mind at work. He looked a bit surprised, but quickly replied, “Time, I guess.” Me too. All I think of is time. I think: I’ve worked one hour . . . now two hours . . . four more hours to go . . . three-and-a-half hours more . . . three more hours to go . . . Do I have to work overtime today? . . . Should I do my laundry after work? I can’t think of anything else. Sometimes I think of something totally illogical: landscapes with towns I once visited suddenly appear one by one. Perhaps a restaurant beside a bridge, a coffee shop around the corner near a station, a harbor, all of which I haven’t seen for years. It’s impossible to concentrate on any one scene. How can you contemplate anything serious in a cycle of a minute and twenty seconds? You just have to accept fragments of memories as they race by. Even your unconscious mind starts to mimic the line.²⁵

Despite the promises of dignity and workplace equality – NUMMI’s promoters made a great deal of the fact that, where GM-Fremont had separate cafeterias for labor and management, the new facility had only one – many workers found life at NUMMI degrading and dispiriting. The new contract “provided for only two categories of workers – assemblers and technicians.”²⁶ The work itself “was regimented in its minutest gestures.” The dream of a more humane assembly line, symbolized by the andon cord, was quickly dashed: “You can stop the line, but you’ll have hell to pay for doing it. Management makes no secret of the fact that they hate to see you stop the line,” explained a team leader.²⁷ Some new policies appeared particularly draconian. A points-based attendance policy counted all absences – no matter the reason – towards a non-negotiable tally. Earn too many points, and you would be fired. “You have to schedule your illnesses in advance,” one worker complained to a reporter.²⁸

Opinion was deeply divided on the new regime, as Fremont workers carried their decades-long traditions of shop-floor debate into NUMMI. Some, particularly those who had been chosen to make one of the trips to Japan, welcomed the new system with open arms. Others were simply chastened by the experience of the plant’s closure, feeling themselves lucky to be working at all: “We got a second chance here . . . Many people don’t get a second chance.”²⁹ Local 2244 president Tony DeJesus, who had led a 1977 wildcat strike, “abandoned his militant attitude” because “we would have had three or four hundred people on the street.”³⁰ Others rebelled, organizing the dissident, reform-oriented People’s Caucus, which demanded “union leaders who are worker advocates rather than management-labor counselors.”³¹

In late 1990, local UAW leadership outraged NUMMI workers by striking a deal with management to lower the rate of pay for new hires outside of normal contract negotiations. The People’s Caucus swept into a majority on a wave of plant anger. At the same time, workers at Mazda’s plant in Flat Rock, Michigan – one of several across the country that consciously followed NUMMI’s lead into the team concept – also voted in a militant leadership slate. But divisions remained, and subsequent union elections saw the plant split almost evenly down the middle, as the People’s Caucus was edged out of majority. Despite intermittent challenges by rank-and-file reformers, no major shifts in the basic labor–management relationship occurred during the life of the Toyota–GM joint venture. During contract negotiations in 1994, NUMMI saw the only strike in its history. Work stopped in the early morning before most workers had even arrived for their shifts. The union and the company reached a tentative agreement within 90 minutes.

The single significant counter-record to company propaganda about life at NUMMI is found in the pages of the Barking Dog, a monthly newsletter edited by worker Caroline Lund from 1998 until the mid-2000s.³² Lund named it after an earlier Fremont newsletter produced by a Black rank-and-filer in the early 1970s, one of many circulated throughout the plant during the heady GM-Fremont years. The Barking Dog published the thoughts of (usually anonymous) NUMMI workers on their jobs, their bosses, their union, and their lives. It also shared stories of labor struggles in other parts of the country and the world, as well as socialist commentary by Lund on world affairs (the Dog was rabidly anti-war). These excerpts, each by a different worker, give a sense of the newsletter’s tone and content over the years:

When I was a new hire working on the line, I recall team members saying, no matter how bad the pain is, don’t ever go to Medical within your probation period at least. The reason for this is that you might get fired and there is nothing you can do . . . It is part of NUMMI practice that you risk your well-being for NUMMI.

I just heard from a G/L [Group Leader] that we will be running FASTER. Another second will be taken away this December – from 83 seconds down to 82, without us knowing why . . . They know they are making us work harder, faster, and unsafe, but yet they say “all accidents are preventable.” You can’t prevent the accident if you are being driven into it.

We all know cutting jobs creates problems and overload of work. It will just travel its way down the line for more time needed for repairs. Running the line faster will increase accidents and injuries. All jobs that can no longer get kaizen just have to work faster as the line moves faster . . . I can’t believe our union has not fought to keep these jobs from being cut.³³

The newsletter was widely read in the plant and frequently withstood attacks from the company and Local 2244 leadership, who were the targets of regular criticism in its pages. On one occasion, a union official unsuccessfully attempted to have Lund censured for using the word “black” rather than “African-American” in the Dog; on another, a fake Barking Dog was circulated in the plant containing defamatory statements about other union members. As class conflict at NUMMI rarely exploded into public view, the Barking Dog provides important evidence that the rebellious, autonomous shop-floor culture of GM-Fremont did not simply disappear under the Toyota Way.

By the time the 2008 financial crisis shook the auto industry, culminating in the bankruptcy and bailout of General Motors, the NUMMI experiment had outlived its purpose. GM had absorbed NUMMI’s lessons, and by then only used the facility to produce the unimpressively selling Pontiac Vibe. The company washed its hands of NUMMI during its Chapter 11 restructuring in 2009, leaving Toyota holding the bag in Fremont. Like GM, Toyota had achieved what it wanted from the joint venture long ago: a foothold in domestic US manufacturing that led to a proliferation of non-union plants in the South and the Midwest.

As the economy spiraled out of control and GM abandoned ship, members of UAW Local 2244 became concerned about the future of their jobs. The ensuing crisis laid bare the yawning rift between the rank and file and union leadership, who were wholly unprepared to address a sudden crisis of such magnitude. An emergency meeting ended in chaos after a union official, confronted by numerous hecklers, told one worker to “Shut the fuck up, motherfucker!” and another to “Shut the fuck up, you little queer!”³⁴

Picketing outside their plant, workers pleaded with Toyota not to leave Fremont. While they were never the harmonious worker bees depicted in the orientalist Chevy Nova ads that announced NUMMI to the world (“Imported from . . . America”), after two-and-a-half decades of the joint venture, many workers had developed a strong “NUMMI identity.” Taking great pride in the plant’s reputation for unusually high quality, many identified more strongly as NUMMI workers than as UAW members, seeing their strong on-the-job comradery – a legacy that stretched back even beyond GM-Fremont to the days of wildcat strikes in Oakland’s auto plants – as a product of Toyota’s team concept.

In a call-and-response chant, they appealed to their status as exceptional, high-quality producers and team players:

Everywhere we go (Everywhere we go)
People wanna know (People wanna know)
Who we are (Who we are)
So we tell them (So we tell them)
We are NUMMI (We are NUMMI)
Mighty, mighty NUMMI (Mighty, mighty NUMMI)
25 years (25 years)
Of good quality (Of good quality)
That’s who we are (That’s who we are)
Quality, quality NUMMI (Quality, quality NUMMI)
We’re number one (We’re number one)
With a good audit (With a good audit)³⁵

Asked whether he thought Toyota would prefer to run a non-union workforce, one picketer replied, “I hope not, because I don’t think the union has never really been a problem for Toyota. Especially here . . . We always worked towards the same goals, making a quality car.”³⁶

Shared goals notwithstanding, Toyota soon elected to close its only American union plant. As production wound down, anger was trained on union leadership, both local and international, far more than on auto executives. On April Fools’ Day, 2010, the last Toyota rolled off the Fremont line. Over six thousand people woke up on April 2nd to find themselves out of work. Unlike the UAW’s standard Big Three contract, the NUMMI “team” contract did not afford Local 2244 members the right to transfer to other UAW shops elsewhere in the country. Facing a potentially astronomical environmental clean-up bill on the site, which had been leaking solvents and other toxic chemicals into the soil since the 1960s, Toyota went looking for a buyer to take responsibility for the facility. At the height of the Great Recession, few were in the market for a massive auto plant in the high-wage labor market of Northern California.

Few, that is, besides Tesla Motors. In most respects, Tesla made an unlikely occupant for the sprawling NUMMI complex. The tiny Palo Alto company, which assembled fully electric roadsters in the service bay of a former San Mateo Chevrolet dealership, was more of a tech firm than an auto manufacturer. Its founders were former executives of software and computing outfits, and its highly ambitious CEO Elon Musk (who bought himself “retroactive founder status” when he joined the company) had recently made a considerable fortune from eBay’s $1.5 billion acquisition of PayPal, of which he was a major shareholder. Toyota sold NUMMI to Tesla Motors for a pittance of $42 million. In addition, the world’s biggest car company made a $50 million investment in the new and unproven Tesla – in effect, paying Tesla $8 million to take the factory off their hands.

When GM-Fremont closed in 1982, UAW Local 1364 remained a lifeline for its members in the difficult days before NUMMI opened. Even the UAW international, which many Local 1364 members accused of selling them out, fought hard in public and private to keep Fremont Assembly a union shop. But in 2010, the future union status of the factory’s workforce does not seem to have been a topic of serious public concern. It may have simply been taken for granted that a boldly future-oriented company would have no time for anachronisms like labor unions – in 2009, American approval of unions polled at an all-time low of 48 percent. When the plant began Tesla production in October 2010, non-union workers manned its assembly line for the first time in its 47-year history.

Those Tesla workers who remembered NUMMI soon found themselves in a new and unfamiliar environment, as Musk imported the autocratic, cult-like management techniques of Silicon Valley to the auto industry, with its 80-hour work weeks and fierce demands of personal loyalty. “A maniacal sense of urgency,” he was fond of repeating, “is our operating principle.”³⁷ In an email titled “Ultra Hardcore,” he instructed employees to “prepare yourself for a level of intensity that is greater than anything you have ever experienced before. Revolutionizing industries is not for the faint of heart.”³⁸ In place of the precepts of the TPS, they learned Musk’s “algorithm”:

1. Question every requirement
2. Delete any part of the process you can
3. Simplify and optimize
4. Accelerate the cycle time
5. Automate³⁹

In themselves, the twin pillars of the Musk approach – cost-cutting and speed-up – were nothing new to auto manufacturing. But the means of determining and implementing these goals marked a stark departure from the norm. Rather than the dispassionate statistical quality control, planning, and target setting of Toyota, Musk led by impulse and personal fiat, making “incredibly difficult, possibly unfeasible orders” and expecting “his employees to work around the clock like he does” in “bursts of condensed effort [called] ‘surges’” to meet them, unhesitatingly firing even high-level employees when they balked at, or failed to meet, his demands (in the early days of the Fremont Factory, he fired no fewer than three production-quality chiefs).⁴⁰ A fixation on automation led Musk to hastily install huge numbers of expensive state-of-the-art robots at Fremont. Upon discovering that the robots were actually less efficient than ordinary human labor, Musk walked the floor with a can of orange spray paint, giggling “with childlike humor” as he marked unlucky robots for destruction, after which workers ripped them from the line.⁴¹

Musk’s cartoonish intensity, unpredictability, and contempt for workers combined to create a factory that was singularly unsafe, with an injury rate “30% higher than the rest of the industry.”⁴² Reports emerged of overt, daily racist harassment against Black employees, decades after the organizers of the Black Panther and Brotherhood Caucuses had worked so hard to drive such harassment out of the same factory.⁴³ This sharp disintegration in working conditions was accompanied by a dramatic decline in wages.

As Richard Ortiz, a NUMMI veteran at Tesla, explained,

At NUMMI . . . we were making a lot more money, between $30 and $35 an hour, and at Tesla we’re getting $20 an hour on second shift, that’s with the premium. The hours we had to work were just twelve hours every day, no questions asked, twelve hours, six days a week. I’ve seen guys that were so afraid of missing work that they were throwing up in buckets because they were sick, but they didn’t want to lose their job . . . They’re using them and spitting them out.⁴⁴

In 2017, Tesla fired Ortiz for his involvement in an ultimately unsuccessful union drive that began in Fremont the previous year, a firing the National Labor Relations Board later judged to be illegal. As of this writing, the Fremont Factory remains non-union, as do Tesla’s eight other US factories. Many of Tesla’s facilities now dwarf Fremont in terms of production. Like GM and Toyota before it, the company may soon outgrow the site.

*

In 2009, as Fremont workers picketed to save their jobs, hurled insults at each other in union meetings, or quietly resigned themselves to losing the wages that sustained their families, longtime Toyota manager John Shook wrote a reflection on the plant’s 25-year tenure: “Why is NUMMI closing? Clearly, the dollars and cents don’t add up for either Toyota or the new GM. Neither needs the capacity right now . . . But, also – more importantly – the learning is done.”⁴⁵ Fremont proved the stateside efficacy of “Japanese management,” which has become the new common sense in industries far afield of auto manufacturing, as medical services providers embraced JIT and Trader Joe’s employees received mandatory lessons in kaizen.

Whether Fremont will continue to serve as a bellwether for changes in the granular structure of American class society is an open question. Elon Musk has recently concluded a short-lived experiment in applying the Silicon Valley Führerprinzip and his own fanatical obsession with cost cutting to the federal government. The balance sheet of the DOGE project – made all the more obnoxious for being named after an Obama-era internet meme – is yet to be drawn up. Similarly, whether the “cult of the founder” will survive in the tech industry (or even Tesla, Inc. itself), let alone become standard across the broader economy in the fashion of the TPS, remains to be seen. In any event, Tesla’s Fremont Factory may well be an image of darker days to come.

On the other hand, the revitalization of the UAW under new president Shawn Fain, the first directly elected by the union’s members, suggests the possibility of another path forward. The once-moribund union has undergone an impressive reorientation towards shop-floor militancy and new organizing. Fain’s call for unions across industries to line up contract expirations in preparation for a potential mass strike on May Day 2028 is a far cry from the conciliatory legacy of Bruce Lee and his ilk. But there is still a great distance to travel before even the most progressive American union leaders catch up to the program of Fremont’s Kenny Horston: “Gather the masses in a revolution to throw out the bogus leaders . . . by any means necessary.”⁴⁶