Our new issue, “The Working Class,” is out in print and online now. Subscribe today and start reading.

The Defeat at Amazon and the Union Fights to Come

The union loss in Bessemer, Alabama against Amazon was a crushing defeat. It’s a reflection of a disjunction between “laborism,” the intellectual and activist infrastructure supportive of organized labor, and the labor movement itself.

A sign at the Amazon BHM1 fulfillment center is seen before sunrise on March 29, 2021 in Bessemer, Alabama. (Patrick T. Fallon / AFP via Getty Images)

I was too sentimental this time. In 2014, in these very pages, I noted that a friend had asked me for a prediction as to the vote in the United Auto Workers’ (UAW) effort to organize Volkswagen’s Chattanooga, Tennessee facility. With a lot of luck, leavened by instincts and decades of experience working in and studying the labor movement, I nailed it exactly: a 53-47 union defeat. In that one you could see that the UAW had a shot — they had been working this campaign for a long time and had won a pledge of company neutrality from the German-based company. It turned out to be not quite enough. But okay, a fairly close loss.

So a couple of weeks ago, as we looked to the results at the Amazon warehouse in Bessemer, Alabama, the same friend asked me what I thought would happen this time. I said a 55-45 defeat. I agreed with every labor person I had talked to — nobody thought the union had a chance to win. But my own historiographical awareness of the militant organizing of the Mine Mill and Smelter’s union of mostly black workers during the Depression in the same area in Alabama had me yearning for a modern invocation of that memory today.

The result, however, was much worse than I imagined: a 70-30 rout with a terrible 55 percent turnout, indicating that many workers were either too fearful or too alienated to vote at all. The result was born of a combination of the standard union-busting tactics used in the most anti-union nation in the advanced world, refined to a coercive perfection by an enormously sophisticated and powerful company — and also the manifold mistakes, essentially malfeasance, of the Retail, Wholesale and Department Store Union (RWDSU), the fairly small union running the campaign. (Jane McAlevey smartly walks through many of these mistakes in her postmortem in the Nation.)

The key point McAlevey makes is that once Amazon won the fight to enormously change the composition and size of the bargaining unit at the warehouse, the union had to either sharply revisit its strategy or simply withdraw to fight another day. Companies make these unit size objections all the time in National Labor Relations Board (NLRB) union elections — it’s how they dilute the strength of union activists. In this case, the unit size went from about 1,500 to about 5,800 — a fourfold increase.

Yet the union plowed on and ultimately submitted about 2,000 authorization cards to the NLRB, around 35 percent of the unit. This was not a harbinger of success but of failure.

Whenever I had this discussion with organizers during my years working in labor, we saw between 60 and 70 percent as the minimum percentage for initial interest in unionizing necessary if the union had a chance at winning an NLRB election. (There are other ways to fight a company — long “comprehensive campaigns” that seek to finally get the company management to recognize a union without a board election. But most unions organizing a single site still go through the election process.) This gets back to the tilted terrain of struggle on which the company and union engage each other. Union support will almost always attrit, not accrete, in a time-sensitive board election fight, because union organizing in the United States is not a fair fight and never can be as long as the boss signs workers’ checks and is able to manipulate and intimidate them without fear of severe legal reprisal.

Amazon’s anti-union efforts here have been well documented and were both typical in that pretty much every anti-union company employs them (like captive audience meetings) but also more ubiquitous and sophisticated than most. Workers were surveilled all the time. But, complicating matters, Amazon threw a few carrots in with their sticks which, based on interviews, appealed to some workers — a miserable but not outrageously miserable starting wage of $15.50 (below some other warehouses in the area) and health care coverage.

Given these huge handicaps, the burden is even higher on worker activists and the union to plot out a plan with a chance to win. A union can’t start at 35 percent and hope to get to 51 percent. It has to start at 65 percent and fight to hang on to 51 percent. And it can’t rush through a campaign that it had only gotten off the ground a few months earlier when the company completely undercuts its original assumption that the unit was going to be 1,500 workers, not 5,800.

Once the union knew it had only 35 percent preliminary interest to consider unionization in that much larger unit of 5,800 — because that is all that signing an authorization card signifies — then, as McAlevey wrote, it should not have gone forward with the campaign at this time at all. This campaign developed rapidly, but it had not ripened to fruition. By moving to an election anyway, the campaign became, as an experienced union organizer and expert wrote to me, a “burn-the-shop-down” effort.

Laborism Without Labor

The issues here go well beyond the strategic analysis made by this relatively small union at this one warehouse. Such defeats are demoralizing for shop floor activists. But in this case, the damage is greater than that suffered in Bessemer alone, because Amazon itself and other enormous nonunion companies are so significant — the larger the adversary, the more important victory or defeat becomes.

Yes, at any given workplace, a union can keep fighting and eventually win, like the United Food and Commercial Workers International Union’s (UFCW) remarkable sixteen-year campaign at a Smithfield, North Carolina poultry plant. Smithfield is a large company with about $13 billion in annual revenue and forty thousand workers. But Smithfield is no Amazon, which has almost $400 billion in annual revenue and about eight hundred thousand workers in the United States alone, second only to Walmart among private-sector employers.

Similarly, even the UAW’s defeat in Tennessee against Volkswagen and then in Canton, Mississippi against Nissan in 2017 did not have the negative weight of the Bessemer loss. As big a deal as those fights were, the CIO’s heroic age of organizing auto, steel, and mining has long passed — the US economy is no longer dominated by those sectors but by health care and education. There are, for example, half the US autoworkers today than there were in 1940 when the population was only about 40 percent what it is now. There are more teachers today in West Virginia than there are mine workers.

And while health care and education are indeed the primary sectoral drivers of the US economy today, Amazon has succeeded Walmart and become the emblematic corporation of capitalist modernity. Amazon — with its seamless combination of labor exploitation, efficient, virtual consumption, and ruthless management of the global supply chain — grows ever more pervasive in our workplaces, communities, and homes.

To paraphrase what W. H. Auden famously wrote about Sigmund Freud, Amazon is no more a corporation now but a whole climate of opinion. Jeff Bezos, its founder and CEO, is the Andrew Carnegie of our time, the richest man in the world. A 2018 survey listed Amazon as the second-most trusted institution in the country (after the military).

In the United States in 2021, no union fight could have had more weight on the level of production and cultural reproduction than one against Amazon. But, as it happens, the weight was mostly on the level of cultural reproduction, not production. And that is a problem.

The coverage — from the mainstream outlets like the New York Times to left media — has been extraordinary, not merely centered in Bessemer but about Amazon writ large. But this is what I have called “laborism without labor.” The intellectual labor infrastructure of youngish urban leftists working in media and academia and education is ahead of the overall labor movement itself. Organizing in those sections is robust. The elite institutions of high liberalism, like the New Yorker and Harvard University, are among the most militant workplaces in America.

This is a good thing, of course. But it’s not a mass base. Although union fights take place every day in this large country in disparate workplaces, the decline in union density continues even as labor scholarship and advocacy in books, articles, and social media grow more compelling.

Finding the Right Leverage

And regardless of the real failures of the union in Bessemer, Amazon is probably just too big and too powerful to be organized one warehouse at a time, even one as large as the Bessemer facility. Amazon has 110 such warehouses in the United States alone in forty-two states and another seventy-five abroad. While winning in Bessemer would have certainly inspired worker activists elsewhere, Bessemer’s workforce represented somewhat less than 1 percent of all Amazon workers in the United States. The struggle might have spread like wildfire and forced a reckoning at the top of the company, but the sheer size of the company makes that hard to bet on.

Think about the 1937 Flint auto plant sit-down strike, which resulted in the UAW’s organizing General Motors. GM had about two hundred thousand blue-collar workers then and about one hundred facilities (about the same number as there are Amazon warehouses today). At the time, GM employed about the same percentage of the total US workforce that Amazon does today, roughly 0.5 percent. But the Flint plant alone employed almost fifty thousand workers, about 25 percent of the workforce in the entire company.

Thus to organize Flint was to control the commanding heights of the company. Trying to organize Flint was actually a shrewd assessment of what was required to organize GM in total. And it worked: the UAW grew almost twenty-fold within a year.

That can’t be done today when there aren’t flagship facilities so large that winning at them will effectively set the company on its heels and result in victories everywhere.

Such work sites at major nonunion companies like Walmart (with its 1.2 million workers and 4,700 similar-sized stores) or FedEx don’t exist today. Thus, only an overarching national campaign, combined with as much rank-and-file militancy as possible, can result in the organizing of branded, transnational corporations.

The leverage must be ubiquitous, not discretely linked to one warehouse and one work site. It must also be backed by as many workers as possible, so the selection of the sites to organize — and the development of majority support in those units — is critical. (Nelson Lichtenstein discusses one idea for such a leverage strategy based on antitrust pressure here.)

To organize these enormous corporations, something different has to happen — worker activists have to be linked from work site to work site. Simultaneously, an enormous amount of pressure on the CEO from the top — legal, legislative, and regulatory pressure, as well as relentless media scrutiny — has to occur.

And the logic has to be to organize workers. Fight for $15 is itself a kind of remarkable flourishing of laborism — it has benefited millions of workers around the country without creating a single new union member. There is nobody to “blame” for that, but organizing has to be the goal, because organizing generates political and economic power.

Legislation to make organizing something closer to a protected right will also define this larger terrain of struggle. It took the massive, militant 1934 strike wave to enable the passage of the 1935 National Labor Relations Act in the face of a reluctant FDR. Today, labor activists, labor-left policy advocates, and congressional Democrats, along with a surprisingly receptive Biden presidency, may have to figure out a way to find enough Senate votes to enact the Protecting the Right to Organize (PRO) Act without a mass worker’s mobilization.

Oddly, the failure to pass labor law reform in 1977-78 might be a more instructive example than the enactment of the NLRA. At that time, labor still had significant union density in key sectors, as well as a Democratic Senate and president. But Jimmy Carter was not interested in labor law reform and did barely anything on its behalf, the Senate was filled with reactionary Southern Democrats, and the AFL-CIO absurdly hoped to cut deals with those unicorns also known as “corporate liberals.” Despite all of this, organized labor came within a vote of enacting the bill into law.

Two generations on, the circumstances in the Democratic Party and the influence of unions themselves are very different. And the PRO Act is a far superior piece of legislation to the labor law bill of the Carter years, one that would effectively strip companies of much of their union-busting arsenal, and restores the right to choose a bargaining agent to the workers themselves. But the near miss of two generations ago might still be a better model than 1934-35. It might not be possible to organize and strike at such an enormous scale anytime soon, yet there is a laborist cadre within Congress and the Biden administration trying to pass the PRO Act.

Labor always has a better chance against bosses when the state, via the cross-pressured Democratic Party, puts its thumb on the scale on its behalf. Like during the beginning of the New Deal Order and through the war years, that support is needed again. But a much weaker labor movement is demanding it, and the rejection by rural white workers of the Democratic Party makes it harder to build the necessary geographical political coalition.

Victory, Somewhere on the Horizon

In his enduring book Prisoners of the American Dream, Mike Davis wrote of the importance of the “sedimented historical experiences of the working class as they influenced and circumscribed its capacities for development in succeeding periods.” Davis continued: “Each generational defeat of the American labor movement disarmed it in some vital respect before the challenges and battles of the following period.”

Davis was discussing the aggregated trajectory of large labor epochs throughout the nineteenth and twentieth centuries. But discrete political and labor losses symbolized those larger generational defeats within the political economy — whether in 1877, 1886, 1894, 1919, or even the truncated victories of the CIO era and the real but stolid stasis of the postwar, high-water mark in union density.

The Bessemer debacle is a discrete defeat. But to invert Davis for a moment, Bessemer — more than the two auto plant defeats in recent years and perhaps more than the hopeful militancy recently in health care and education — might do more to buttress rather than disarm labor for the necessary fights to come. This time, may defeat lead to wisdom and determination, and may all of labor catch up to the ideological advances of laborism.