Close this search box.


The union organizing drive at Amazon’s giant warehouse in Bessemer, Alabama, inspired great hopes, just as its defeat in April induced much despair. But despair can be paralyzing. If the socialist left is to be better prepared to support workers in the next great conflict at Amazon, it is incumbent that we learn from this defeat, the better to arm our movement with the understanding and the capacity to move beyond it.

Defeats are always replete with lessons for the future. As the great Rosa Luxemburg put it in January 1919, on the eve of the counter-revolution in Germany that would take her life, it is from defeats that “we draw historical experience, understanding, power and idealism.”1Rosa Luxemburg, “Order Reigns in Berlin,” Gessamelte Werke 4 (Berlin, Dietz Verlag, 1919),  It is in this spirit that Spectre has invited a distinguished group of union activists and organizers along with scholars of working class and radical movements to offer their assessments of this recent struggle.

Our commentators have different takes on the events and on the lessons to be drawn. And this is as it should be. Yet they converge on one crucial point: that the labor movement will need to invent new forms of struggle, organization, and solidarity if it is advance against corporate giants like Amazon.

Many of our commentators agree further that accepting the pro-employer “rules of the game” as set out by the National Labor Relations Board (NLRB) is a losing strategy.2NLRB rules are so extremely pro-employer that many of the tactics used by Amazon in Bessemer—such as convening mandatory workplace meetings to denounce the union drive, and sending anti-union email messages and texts to workers—would be illegal in many capitalist states. It is interesting in this regard that Teamster organizers assisting workers at Amazon in Iowa have floated the idea of waging a recognition strike that bypasses the NLRB voting process and instead mobilizes workers in strike action to win union rights.3Tyler Jett, “Could an Amazon Union Form in Iowa? The Teamsters Say They’re Organizing Employees,” Des Moines Register, February 26, 2021. It is not at all clear that high-ranking Teamster officials support the recognition strike approach.

One advantage of such an approach is that it puts direct mass action by workers at the center of union organizing, rather than a relatively passive mail-in ballot process accompanied by a PR campaign. The latter strategy, pursued by the Retail Wholesale Department Store Union (RWDSU) in Bessemer, has been a serial failure in the US for decades now. Such a strategy leaves workers isolated and atomized, and more easily intimidated by employer threats and retaliation. And this returns us to Rosa Luxemburg.

Writing about the insurgent mass strikes in Russia in 1905, Luxemburg argued that they turned upside-down the bureaucratic model of union organizing predominant in Germany at the time. German union officials, said Luxemburg, believed that organization had to precede action. Yet the general strikes in Russia proved “exactly the opposite,” she insisted. Out of “the fire and glow of the mass strike” emerged “young, powerful, buoyant trade unions.” 4Rosa Luxemburg, The Mass Strike, the Political Party and the Trade Unions (New York: Hsaper Torchbook, 1971), 36. Mass action came first. Dynamic union organizations followed. Several of our commentators suggest that this is the scenario that will be required at Amazon—and that it will have to involve insurgent strike action at multiple sites.

A return to this strategy—which built the industrial unions in the 1930s and the public sector unions in the 1960s—would require the US union officialdom to break with nearly four decades of attempting to be “partners” of employers in making their bosses more competitive and profitable, and with the nearly century-old alliance with the Democratic Party. It is highly unlikely the current leadership of the US labor movement will make such a radical change in course. Again, as in the 1930s and 1960s, the leadership for a new labor movement will come from below— from rank-and-file workers in workplaces.

It is equally clear that antiracist organizing will have to figure decisively in union struggles at Amazon. Indeed, antiracism can be a crucial bridge between the workplace and the community—a bridge to the organized solidarity that will be necessary if Amazon workers are to break through in their struggle against this corporate behemoth. Now, let’s turn to our commentators on this struggle for working class and racial justice.

Reason to be Optimistic

Robin D.G. Kelley

The RWDSU might have lost the election at the Bessemer warehouse, but I think there is still reason to be optimistic, as well as critical lessons to learn. First, the Bessemer campaign has already sparked talk of a renewed labor movement, as well as requests from other Amazon workers to launch union drives in Baltimore, New Orleans, Portland, Denver, and here in Southern California.

Second, the Bessemer workers helped embolden 1,100 mine workers at Warrior Met in nearby Tuscaloosa County to go on strike for a better contract. Since the mine came under new ownership in 2016, workers have suffered a succession of severe cuts in pay and benefits while CEO Walter J. Scheller III raked in $4 million a year. One of the most draconian measures is the company’s “four strike” policy, which automatically fires workers who miss four days of work over a fifteen-week period. As I write, the strike is still in progress, but I am convinced that the strikers have enjoyed more support and solidarity than expected because of the focus on Alabama.

Third, because the Amazon struggle was a classic David and Goliath story—mostly Black and female workforce in the deep South going against the biggest company and richest man in the world—it got a lot of media attention. There is currently a real national discussion of, and robust support for, the Protecting the Right to Organize Act (PRO Act), more than there would have been had the world’s eyes not focused on Bessemer. The legislation is significant and is a step toward strengthening labor law and federal institutions once designed to protect the rights of workers. These institutions, notably the NLRB, have been eviscerated—and not just under Trump. Indeed, the kind of union-busting measures that Amazon deployed—forcing employees to attend mandatory anti-union meetings and other forms of intimidation, for example, would be illegal under the PRO Act.

What are some key lessons of this struggle?

It is true that Amazon spent more than $10,000 a day on an outside “consulting” firm to fight the union, and spared no expense to intimidate workers and make union organizing at the workplace logistically impossible. But neoliberals had won the ideological war on unions even before Amazon was founded in 1994. Not only has union membership steadily declined since the 1970s strike wave and the Reagan-era war on organized labor, but we’ve seen neoliberalism’s unrelenting assault on the very idea of unions and labor solidarity. The idea that we are all individual human capital whose fate depends on taking “personal responsibility” forecloses the option of taking action as a group with shared interests.

Clearly, a two- or three-month campaign at a single plant will not do. We need to be prepared… for a long struggle.

“Neoliberal reason,” as Wendy Brown and others call it, has prevailed for a very long time, and is hard to undo. For two generations at least, we’ve been taught that unions are a drain on society; they take dues from hardworking Americans and pay union bosses bloated salaries; they kill jobs with their demand for high wages and undermine businesses and government budgets with the excessive pension packages; they are the takers while capitalists are the makers—and the makers should decide what to pay workers. Personal responsibility means we are to blame for our poverty and misfortune. Free enterprise makes us free, moral, and responsible. How many Bessemer Amazon workers voted against the union or did not vote at all because they came to believe that unions harm more than they help, or that loyalty to the company will bring rewards, is impossible to know.

Nevertheless, neoliberal reason may be cracking among precarious workers. And support for organized labor has ebbed and flowed over the past forty years. Remember that Amazon started in Bellevue, Washington, the year of NAFTA’s passage and a year before the ensconced old-school leadership of the AFL–CIO was overtaken by the election of John Sweeney (president), Richard Trumka (secretary-treasurer), and Linda Chavez-Thompson (executive vice-president). They ran on an agenda of organizing Black, Latinx, Asian-Pacific, and women workers irrespective of race and ethnicity. Sweeney came from a dynamic underdog union movement, Justice for Janitors, whose social composition and progressive democratic vision resembled that of the RWDSU’s campaign to organize poultry workers. In many cities, the janitors’ cause became a civil rights movement—and a cultural crusade. Justice for Janitors waged several successful “crusades” in Pittsburgh, Denver, San Diego, Los Angeles, and Washington, D.C., in solidarity with Latinx leaders and local church officials, and they won wage increases and better conditions, but it took a very long time. And it took significant solidarity from allied social justice movements outside of organized labor.

Clearly, a two- or three-month campaign at a single plant will not do. We need to be prepared not only for a long struggle ahead but for a broad movement that extends far beyond the confines of union halls, shop floors, and warehouses.

Robin D.G. Kelley is Professor of History at UCLA and contributing editor for the Boston Review. He is the author of several books, including Freedom Dreams: The Black Radical Imagination and Hammer and Hoe: Alabama Communists During the Great Depression.

Organizing Begins in Relationships

Barbara Madeloni

The workers and organizers at Bessemer were asked to carry some heavy weight for the labor movement. Somehow their organizing drive was supposed to answer to all of the big questions facing labor: How are we going to take on one of the most powerful corporations in the world? How do we ignite organizing in the South? How will the Black liberation movement and the labor movement come together? Where is our entry point to worker power within the supply chain?

The letdown, even if expected, was hard. We wanted the win for the workers, and we wanted the win for labor—for all workers—for ourselves. But it is a mistake to ride too high on one organizing drive, one strike, as if in the moment we can know what it will mean in the months or years ahead. While many organizers know this, too many got swept up in the corporate media’s framing this as a do-or-die fight. It wasn’t.

Sober assessments of what it will take to shift the balance of power, especially in the face of the tentacles of corporate power such as Amazon’s are necessary. Nothing undoes the fact that we need solid, face-to-face worker-led organizing drives—and these take time and a willingness to learn from mistakes. How are we taking that up in the spaces we occupy?

There are factors, however, that support a promising future for workers. The rank-and-file uprisings of the 1970s were part of broader fights for justice taking place at the same time. In our moment, we see an increasing willingness to name capitalism as the problem and socialism as the answer. We have active struggles for climate justice, Black liberation, housing rights, Medicare for All, and an end to student debt. Worker self-organizing is bubbling up from educators to other Amazon workers.

Nothing undoes the fact that we need solid, face-to-face worker-led organizing drives— and these take time and a willingness to learn from mistakes.

The organizing drive exposed Amazon’s outrages to a broader audience, from their captive audience meetings, to their lies and illegal maneuvers, to the conditions under which workers labor—validating the experiences of workers in those warehouses. Broad public support for unionizing doesn’t win recognition; we need power to do that. But it might make a person more willing to be part of building that power.

When we say that organizing begins in relationships, we are saying that it begins in workers having a different kind of experience of themselves and others. One-to-one conversations change how workers know themselves and what is possible. While one side of the ledger from Bessemer says workers experienced what is not possible, the other side showed what is possible. It is our work as organizers to learn from one side and build toward the other.

Barbara Madeloni is Education Coordinator at Labor Notes and a former president of the Massachusetts Teachers Association.

The Problem isn’t a Lack of Interest by Workers

Kim Moody

The defeat suffered by the Amazon workers in Bessemer, Alabama, should be a wake-up call to organized labor and workers everywhere. To be sure, the last word in the struggle to organize Amazon and other capitalist predators has not been uttered despite this setback. But the fact that only a fifth or less of the 5,876 workers eligible voted for the union needs a deeper explanation than just the massive campaign of lies, intimidation, false promises, and fear. Amazon and its hired guns ran against the Bessemer workers and the RWDSU. Despite the courage of those who did work for and vote for the union, the overall turnout at Bessemer was significantly less than the average turnout of NLRB elections in recent years, even during the pandemic in 2020 and early 2021.

Part of this defeat was almost certainly due to the RWDSU’s apparent choice to run this campaign more like a conventional US election with its celebrity and political endorsements and appearances (from Bernie to Biden) along with the usual media and digital dependency, rather than a direct class conflict with on-the-ground worker leadership, organization, and escalating action. Another piece of the explanation for so devastating a defeat is, as RWDSU president Stuart Appelbaum himself put it, “Our system is broken.” While it’s hardly “our” system, the whole NLRB process and apparatus is, indeed, broken and has been for ages and everybody knows it. Yet, despite that realization the RWDSU plans to appeal for another election. The snooze button pushed again?

The bigger factor in all of this is that Amazon is so big, so rich, so rapidly growing, so much the trendsetter of employment conditions, and one of the few places where working class people in towns like Bessemer (and they are legion) can find any job that pays more than minimum wage. As of this April, Amazon has eight hundred and twenty-four facilities in its US fulfilment network—more than twice what it had two years ago—with three hundred and twenty-five more facilities planned, plus some thirty-nine data centers in its highly profitable Amazon Web Services. It has by various accounts close to one million workers in the US, up from half a million in 2019 with more hiring going on.

Amazon’s size and presence across the country increases its ability to instill fear and claim, as it did in Bessemer, that these jobs were the workers’ only realistic hope for a future. By the testimony of several anti-union workers, this had an impact. Amazon will never be organized one warehouse (much less one NLRB election) at a time—even with the advantages the PRO Act might grant should it ever become law. We call the hope for legislative salvation from above so deeply engrained in most of the labor leadership the “Waiting for FDR” syndrome. Another sleep-inducing malady.

The problem isn’t a lack of interest by its workers. We have seen actions by groups of Amazon workers all over the US that point to a desire to organize and fight the intolerable conditions this giant economic predator imposes. The RWDSU itself reported that it received more than one thousand calls from Amazon workers expressing interest in unionizing. The Teamsters have also reported interest among Amazon (contract) drivers and warehouse workers in Iowa. Outside of these pleas for help, the response from most of organized labor’s officialdom has been nearly invisible. The apparent chasm of power this inertia demonstrates was likely another factor in low turnout and a low union vote.

What are Amazon’s vulnerabilities?

Amazon may seem impregnable, but as I have written elsewhere, it is vulnerable to collective worker action, and it said as much in its 2018 Securities and Exchange Commission report (SEC 10K) where “labor disputes” come ahead of terrorism and “Acts of God” as potential sources of disruption.55. Kim Moody, “Amazon: Context, Structure, and Vulnerability” in The Cost of Shipping: Amazon in the Global Ecnomy, Jake Alimahomid-Wilson and Ellen Reese, eds. (London: Pluto Books, 2020), 21–34; United States Securities and Exchange Commission, Form 10-K,, Inc, 2018 (Washington, DC: United States Securities and Exchange Commission, 2018), 8.

This vulnerability stems from its apparent strengths as one of the most sophisticated, tightly organized logistics systems in the world and the resulting speed at which it moves goods and money in its largely successful competition with other retailers. The combination of speed throughout its network as well as in delivery to its customers, and intensified competition as it grows, can give significant leverage to workers willing and able to strike or occupy major strategic fulfilment centers, in-or-out-going sortation centers, or other key facilities.

Part of this defeat was almost certainly due to the RWDSU’s apparent choice to run this campaign more like a conventional US election.

This is because in addition to stopping the flow of goods in one or more locations, direct action will back up transportation links and affect other facilities in this tightly organized system. Stop the turnover and the cash stops, inventory grows, and fixed costs pile up in the area affected. If the actions last for a while, they will also tend to create an imbalance up and down the “just-in-time” system, at least regionally.

At the same time, because Amazon has a lot of cash and so many duplicate facilities, it will be necessary to strike and/or occupy at least several key choke points to have a major impact on the whole system. Because of this, a serious effort to organize Amazon will have to be nation-wide. It will need to be based on the workers in face-to-face organizing and action (with masks and distance if needs be), not on PR, celebrities, or digital substitutes. It would be most effective if it focused on the major metropolitan areas where Amazon’s facilities and customers are concentrated. This is also where the bulk of America’s union members and social movement activists can be mobilized to support the direct actions of Amazon workers, particularly by blocking truck traffic in and out of these facilities, while running interference against the police.

While there are no simple models or precedents for unionizing the corporate giants of a period, the organization of General Motors in 1936–37, the ruthless corporate Goliath of its day, offers some hints. First, the GM workers had to build a new union, the United Auto Workers. Second, the GM activists explicitly rejected using the procedures of the then-new NLRB because they figured GM would ignore whatever the outcome and tie things up in court.

Instead, they turned to well timed, accelerating direct action in the form of sit-down strikes, that is, plant occupations. These were supported by mass picketing and demonstrations outside the plants that stopped police interference and gave the workers an increased sense of power. It was these actions that forced workers on the fence (and there were lots of them) and those who feared losing their jobs (a very real problem in 1937) to take sides. Finally, they chose the right plant in the GM system, Chevy 4, that made all the engines for GM’s best-selling products and brought GM to a halt. There were no votes on whether or not to go union, no supporting celebrities or politicians (certainly not FDR), and scarcely any favorable media coverage. It was workers and their supporters engaged in organized strategic direct action in multiple forms that carried the day. Their victory inspired hundreds of similar successful struggles that breathed life into the labor movement. History doesn’t repeat itself, but there are certainly some lessons here worth considering.

Kim Moody is a founder of Labor Notes in the US and the author of several books on labor and politics. He is currently a Visiting Scholar at the University of Westminster in London, and a member of the University and College Union and the National Union of Journalists.


We’re Not Giving Up

Frances, Amazon worker

What issues fueled the organizing drive?

Well, the first one of course was just simple wages because of the demand from the job, but it’s more than that. It’s the physical conditions. There are very short breaks. Like ten-minute breaks, fifteen-minute breaks. We have only thirty minutes for lunch, and that includes the amount of time that it takes for us to get to the cafeteria. Our breaks include the time that we take walking to there. We have to try to hold it when we have to use the bathroom. We have to hold it for as long as we can because they keep track of everything that we’re doing at all times. They are tracking everything that we’re doing and it’s dehumanizing. They treat us like we’re just their revenue. We’re human beings. And so, we’re basically just out here trying to get all the rights that we can, the ones that we deserve.

How did the antiracist uprising of the spring and summer shape the organizing drive?

I’ve noticed that a lot of the other people who are out here are people from different organizations that have been out during the summer. And I think they mainly picked up on it because the majority of Amazon’s workers are African-American. And so, they saw it not just as the struggle of us not wanting to be killed, but also as the fact we have to deal with discrimination from all different ways. And that includes in the workplace. And so, everybody just noticed that it was these people who were suffering through the workplace, and they decided to step up and stand with us.

What is your assessment of the union’s strategy and tactics?

I feel as though they have pretty strong method for getting things done. We tried the legal way—the nice way so to say—just to make sure that we could go through the easier steps to start the revolution. We’re not giving up even if we’ve already lost the vote. I think that we have a very determined group of people who are out here focused on the goal. They’re not just focused on any possible downfalls. They are focused on true results and they’re going to do everything that they can to get there.

Frances is an Amazon worker in Bessemer, Alabama.

We Must Nurture Shop Floor Organization

Nantina Vgontzas

The Bessemer campaign has advanced public discourse about labor abuses at Amazon, the grievances of its vast workforce, and the tools needed for building power within the company. These run from the adoption of tactical models focused on building organizing committees that have mass shop coverage, to the passage of labor law reforms like the PRO Act, to the need for greater investment, scale, and coordination among an array of movement forces that can circumvent the redundancies built into the company’s logistics network. Having studied and supported trans- national organizing at Amazon for several years, I will argue here that these three points converge on the strategic question of shop floor control.

Notably, much of the reporting on Bessemer focused on this question of control. In the New York Times, David Streitfeld wrote, “At its heart, the conflict is about control…The company needs to lower labor costs and increase productivity, which requires measuring and tweaking every moment of a worker’s existence.” At the Bessemer facility, one of the key grievances that spurred workers to reach out to RWDSU was the discipline they faced for “time off task,” which includes any intermittent break taken while working on the line—any stretch, any water break, any bathroom break, even any time a robot malfunctions and doesn’t bring the products to a station in time. But it wasn’t simply that managers wrote workers up for this. It was how they did it. When Amazon opened the warehouse at the start of the pandemic, workers didn’t interact much with managers on the floor; the company has automated managerial tasks in its newer fulfillment centers to the point that workers can be assigned stations and monitored remotely. But after the several-month grace period typically afforded when the company is getting operations off the ground, workers began to be called into rooms where managers grilled them for their time off task. This was just as the summer’s protests against policing were picking up, when white managers took particular and sick pleasure in exercising power over a mostly Black workforce.

The company fears this interracial, cross-class, international solidarity precisely because it imagines a radical alternative to corporate power.

Indeed, one of the key themes to have emerged from the organizing at Amazon during the pandemic is its policing of racialized workers seeking to protect themselves and their communities. In Queens, where workers organizing as Amazonians United temporarily shut down a warehouse after they exposed the company’s concealment of Covid–19 exposure, the company targeted one of the walkout leaders, a Black abolitionist who was pulled aside and interrogated by a manager who identified himself as a former FBI agent. The manager worked for Global Security Operations Center, the same group running Amazon’s Covid–19 containment efforts. Later, workers found out that the agent was brought in to dismantle the warehouse’s existing team of managers respon-sible for enforcing Covid–19 safety measures; corporate had deemed the team too sympathetic to their efforts. In Minnesota as well, worker leaders were fired when a predominantly Somali workforce (organizing with the Awood Center) protested the company’s refusal to shut down the warehouse despite knowing its infection rate exceeded the community rate. Here, too, time off task was used to enforce discipline.

What patterns of worker resistance do you see at Amazon?

But Amazon logistics workers haven’t simply contested these abuses. During the pandemic, they advanced a vision of transforming Amazon’s infrastructure from one that causes harm to one that promotes care, and not just for themselves. In Michigan, workers walking out chanted, “Until it’s essential, shut it down,” while in France, workers did precisely that, shutting down operations until essential shipments were prioritized.

By articulating a politics of essential work, logistics workers advanced a perspective first put forth by their climate activist tech coworkers, who argued that a company owning half the global public cloud and domestic retail markets could and should decarbonize its operations and, in moments of disaster, relieve communities lacking basic protections and supplies. Amazon retaliated against these workers, firing many; at least forty have filed charges with the NLRB. The company fears this interracial, cross-class, international solidarity precisely because it imagines a radical alternative to corporate power: a democratically controlled infrastructure that serves social need.

This point might seem abstract, part of a distant future. But the implications for current efforts are rather concrete. A core reason that Amazon logistics workers have captured the imagination of organizers is their potential power in disrupting key economic choke points: The more the company shortens delivery times, the more its operations become vulnerable to strikes and other labor actions. But what I’ve found through my work is that, precisely because of this vulnerability, the company has been automating its network flows so as to minimize the impact of disruptions by quickly redirecting order fulfillment within and between the different types of warehouses constituting Amazon’s network. A few strategically placed workers can no longer shut down an entire warehouse or distribution channel, as they could in the golden days of manufacturing in the twentieth century. Workers can overcome this constraint by building majority participation for strikes within warehouses and coordinating with workers in other warehouses, trucks, and headquarters.

These organizing constraints make the task of building power far more difficult in contemporary logistics than in, say, the auto sector of the 1930s. But they also constitute an opportunity to bring about the informational access, political education, and strategic coordination needed for transforming an increasingly major infrastructure in our lives, rather than simply striking the productivity bargains of the past. An early version of this dynamic emerged in the pandemic, when logistics workers shared information with their tech coworkers, who then exposed other elite workers to the realities of the conditions on the logistics end. Down the road, tech workers could share information about network flows with their warehouse coworkers, who could then coordinate with truck drivers. Throughout this process, these various sets of workers will emphasize the issues animating their communities. Among the most urgent of these is climate change, which links suffering in the communities where warehouses are located with the concerns elite workers have for future generations. It is through these interactions in the workplace and in communities that workers can situate the struggle to gain control over their work within the wider struggle to reimagine care infrastructures.

As a Left that aims to support workers in this effort, it is crucial that we connect their fight to the many others in which we are implanted. It is also important, beyond the PRO Act, to fight for more targeted legislation that both curtails productivity targets and forces company metrics to include environmental impacts. Above all, we must nurture shop floor organization that harnesses such protections and builds militant capacity for the long struggle to abolish Amazon as we know it, and to remake it.

Nantina Vgontzas is a labor activist and postdoctoral researcher at New York University’s AI Now Institute. They are working on a book project about the shop floor politics of Amazon’s global logistics network. 

The Central Question is Fear

Ellen David Friedman

It seems the central question of organizing at this moment is fear. Fear that’s so pervasive, reaching such an unfathomable depth, and with a stickiness and saturation, that it can’t merely be factored in to an organizing strategy but should be a singular focus of organizing. I take as a given the results national polling consistently shows—that a great majority of workers want what a union offers—and then try to understand what we, as organizers, can do to address the fear that interferes.

I’m putting aside the factors of bad law, disproportionate power and resources of employers, and their venality. They—employers and politicians who assist them—are simply doing their job to maximize efficient extraction of surplus value. The question for me is what we do as organizers within this highly constrained reality.

I’ve organized workers in a range of sectors since the 1970s, and watched as this numbing paralysis spread to become both foreground and background. It is, simply, everything in the terrain of the workplace.

By this I mean the fear that prevents a fully tenured teacher, with decades of experience in one district, with a mature collective bargaining agreement, comprehensive “just cause” rights, and a well functioning union, to quake, unable to raise a question in a staff meeting that would suggest disagreement with a supervisor. Fear that prevents a unionized nurse, who labors under terrifyingly short staffing conditions, who is fully aware that violations of professional licensing, accreditation, and insurance standards are the daily norm, and yet who refuses to sign a petition, wear a pin, attend a meeting, file a grievance, or make a complaint.

The goal of organizing is to create conditions for workers to see themselves and one another with respect, to start trusting one another to assess the workplace, to identify shared goals, and to start taking whatever steps they themselves decide to begin to shift the balance of power.

So now imagine the fully encapsulating fear that isolates a typical non-union American worker—someone who is almost certainly already burdened by debt, cultivated to “be grateful to have a job,” accommodated to monstrous wealth inequities as if they are normal, coerced to compete ruthlessly against coworkers and the clock, and accustomed to authoritarian-type hierarchies in work and society. This worker will almost certainly not have a meaningful conversation with a union organizer, much less sign a card, much less become a workplace agitator—unless something has been done to thaw the fear that freezes both the emotions and intellect, and which generates well defended habits of obedience, passivity, and suspicion.

What strategies and tactics of organizing might bear fruit?

I now use only one approach in organizing, summed up in this way: The goal of organizing is to create conditions for workers to see themselves and one another with respect, to start trusting one another to assess the workplace, to identify shared goals, and to start taking whatever steps they themselves decide to begin to shift the balance of power.This seems to be the best route to undermine fear and replace it with purposeful resolve. Specifically, avoid building a time line for any campaign (whether new organizing, a contract campaign, a strike build-up, an electoral campaign) without first considering what it will take to:

  • Have sufficient one-on-one conversations over however long it takes to convene an organizing committee of sufficient size to do the bulk of all organizing (that is, don’t build a plan based on staff organizers doing the bulk of the work once an organizing committee is established).
  • Bring the organizing committee members together, however long it takes, for the purpose of assessing the actual balance of power in the workplace and designing the campaign they will carry out.
  • Go as slowly as necessary to build and test relationships of trust, which means lots of low visibility off-site conversations and meetings.
  • Take great care to develop strong consultative channels for discussion, debate, and decision-making within a growing circle of coworkers, and refrain from unilateralism at all costs.
  • Whenever any action is decided upon, spend as much time as possible talking through fears, adjusting accordingly, and then evaluating honestly and openly to learn for the next time.

Ellen David Friedman is chair of the Labor Notes board and a facilitator of the United Caucuses of Rank-and-File Educators.

They are Being Sold Business Unionism and They Don’t Want to Buy It

Joe Burns

It seemed pretty early on there was a disconnect between the reality of the campaign and the expectations of supporters of labor. But the fascination about the election has more to say about the desires of progressives to take down Amazon than the actual desires of the workers on the ground—which is part of the problem here.

The key point to stress is we need to start listening to workers. We have decades of experience that they do not want to vote in NLRB elections for existing unions. Most of the analysis focuses on reasons they made the wrong choice. But we should assume workers involved are rational economic actors and made a determination that the union could not improve their lives. In other words, they rejected the approach of a business union (even a progressive one) scraping by in a NLRB election and then struggling to get a first contract at a single location of one of the largest corporations in the world.

Let’s say the union won the election. Wages are already twice the minimum wage in the area. Would the union have a strategy to dramatically raise wages? Probably not. They would be struggling to lock in a first contract. A key issue was control of the shop floor, the speed-up and the schedules. But most unions gave up that struggle decades ago. I know it is a tough pill to swallow, but how much would have changed with organizing a single shop?

For decades we have been spending millions to convince workers to join a weak and declining labor movement. In Reviving the Strike I went over some of the numbers but Service Employees International Union (SEIU) and other unions spent billions of dollars in the organizing push of the late 1990s and early 2000s. It did not work, and union density declined during the period. Even in core jurisdictions they were unable to get it to budge. At some point, we should listen to the workers telling us they don’t want to vote for unions in these elections.

They are being sold business unionism and they don’t want to buy it—for three decades they have been saying that. So we can try and change the rules on elections, or get rid of elections and have card count, but at the end of the day, it won’t really change the fact that folks have said they don’t want the existing unions.

What is the alternative for building union power?

The alternative analysis asks what sort of union organization it would take to really take on Amazon. It won’t be our business unions—we have decades of experience there. We need a union movement capable of militant tactics, run by workers not paid staff (whose jobs would be threatened by militant tactics) and with a set of ideas willing to break through the status quo.

Workers in the private sector did not quit striking after the 1980s because they did not want to fight. They lost battles because the rules of the game were set against them with the restrictions on stopping production and solidarity. Winning an election or putting faith in the NLRB will not address that basic problem. We need a labor movement capable of violating labor law, conducting militant actions, and bringing capital to its knees. At some point, starting with the left of the labor movement, we need to refocus on the essentials of class struggle unionism. The core ideas of “us versus them” unionism—class-wide demands, the battle for the shop floor, and militant rank-and-file struggle.

Joe Burns is a long-time labor negotiator, an attorney, and the author of several books on labor including the forthcoming book Class Struggle Unionism (Haymarket Books).

It is a Long Struggle and It Continues

Michael Goldfield

How do you assess the campaign run by the RWDSU in Bessemer?

First, their mobilization tactics were not adequate.

While I have no objection to the support of political figures or entertainers—Paul Robeson, Josh White, Pete Seeger, and Zilphia Horton were omnipresent at union events during the 1930s and 1940s—the union did not do a good job of mobilizing their most important allies, what I call associative power. They should have put pressure on largely Black politicians in Bessemer and Birmingham to come out actively in support. Black elected officials in Bessemer, a town over 70 percent Black, who had facilitated Amazon’s arrival, played no visible role in supporting the union.

The successful organizing of catfish farm workers in Mississippi, for instance, emerged as a civil rights struggle, mobilizing community members and forcing political leaders to actively support them. The appeal to Black Lives Matter sympathies, in contrast, seemed somewhat wooden. Contrary to media reports, Alabama is not that unorganized compared to other states in the South. With a bit under 10 percent unionization, compared to 3 or 4 percent in North Carolina and South Carolina, they have many potential allies. They could have set up picket lines outside the facility and encouraged unionized miners, food processing workers (which they did a tiny bit), and others to join them.

And there were not clear sets of public demands put forward by the union. They should have said, if the union is certified, we will ask for $20 or so per hour (the striking miners at Warrior Met Coal company nearby rejected well over this amount), union safety and health committees, longer and more frequent breaks and lunch periods, less monitoring by computer and supervisor, no discussion of output and breaks without a union steward present, and so on—demands that could have been developed at public meetings of workers, not to put in stone the examples that I have given.

The question of control of the workplace, pace of work, monitoring, and so on is ubiquitous across industries, something that needs to be formulated precisely for each type of work.

Finally, while it is necessary to organize extensively inside the facility, even at times on a nonmajority basis, this is not a permanent solution. Large companies can only be forced to bargain extensively, especially over wages and benefits, but general safety conditions as well, when the whole company is organized into a union.

It is a long struggle, and it continues.

Michael Goldfield is a former labor and civil rights agitator, author of numerous books and articles, most recently, The Southern Key: Class, Race, and Radicalism in the 1930s and 1940s.

  1. Rosa Luxemburg, “Order Reigns in Berlin,” Gessamelte Werke 4 (Berlin, Dietz Verlag, 1919),
  2. NLRB rules are so extremely pro-employer that many of the tactics used by Amazon in Bessemer—such as convening mandatory workplace meetings to denounce the union drive, and sending anti-union email messages and texts to workers—would be illegal in many capitalist states.
  3. Tyler Jett, “Could an Amazon Union Form in Iowa? The Teamsters Say They’re Organizing Employees,” Des Moines Register, February 26, 2021. It is not at all clear that high-ranking Teamster officials support the recognition strike approach.
  4. Rosa Luxemburg, The Mass Strike, the Political Party and the Trade Unions (New York: Hsaper Torchbook, 1971), 36.
  5. Kim Moody, “Amazon: Context, Structure, and Vulnerability” in The Cost of Shipping: Amazon in the Global Ecnomy, Jake Alimahomid-Wilson and Ellen Reese, eds.  (London: Pluto Books, 2020), 21–34; United States Securities and Exchange Commission, Form 10-K, com, Inc, 2018
    (Washington, DC: United States Securities and Exchange Commission, 2018), 8.
If the socialist left is to be better prepared to support workers in the next great conflict at Amazon, it is incumbent that we learn from this defeat, the better to arm our movement with the understanding and the capacity to move beyond it.

Please Log In

You must be a subscriber to access articles from the print issue.

Not a subscriber yet?

Create an Account

Please Log In

You must be a subscriber to access articles from the print issue.

Not a subscriber yet?

Create an Account


While logged in, you may access all print issues.

If you’d like to log out, click here:


Support our Work

Gift Subscriptions, Renewals, and More