For years now, we have been hearing about strike waves, upsurges in worker militancy from a labor movement ostensibly on a comeback. The Economic Policy Institute declared a “surge” in striking workers in 2018 and 2019 a “35-year high”—citing General Motors and Stop and Shop.1 The Washington Post said, “Teacher strikes made 2018 the biggest year for worker protest in a generation.”2 A “Strikewave” website was launched that same year and elsewhere strike trackers sprung up. The headlines have been so resoundingly positive for so long that it’s hard to believe there’s any higher to climb. The most recent example was “Striketober,” the “biggest strike wave in a generation,” (again) according to Robert Reich,3 expected to have a “lasting impact” per an NPR headline,4—“only the beginning” according to The Hill.5
Labor journalists also have an understandable incentive to be the first to accurately diagnose the present moment and declare what course history is on. Those on the left, especially, want to believe that the labor movement is becoming more powerful. Under such conditions, it’s tempting to lump together a number of disparate labor disputes, including striking nurses in Massachusetts, film and television crews, and factory workers at John Deere and Kellogg’s.
No one claimed these unions were deliberately coordinating their activity—instead, “objective conditions” did the heavy lifting of establishing why this was a unified phenomenon and not a coincidence. By broad consensus, the pandemic and an advantageous labor market conspired to wake workers up and suddenly demand more.
That in itself is dubious, but first: was there an uptick in labor action? Yes and no. The Bureau of Labor Statistics (BLS) counted six large strikes in effect in the month of October; April saw an equal number, although with fewer workers involved. BLS only counts work stoppages of 1,000 workers or more, which leaves out countless significant actions such as the 420-member strike at Heaven Hill Distillery in Bardstown, Kentucky. There are attempts to round out strike data, such as Cornell’s School of Industrial and Labor Relations strike tracker; it draws in part from Bloomberg Law’s database. For that matter, much of the “Striketober” buzz was over strike votes that did not turn into picket lines.
Even by BLS’s stingy count, October looked significant, but part of the reason we can speak of an uptick is because the range of figures is so small—between two and six work stoppages per month so far this year (15 total for the year), involving between 1,800 and 18,300 workers. Kim Moody, drawing from those other sources like Cornell’s tracker, has a more generous count of 194 so far this year, but even he acknowledges that these involve a total of 73,320 workers, far fewer than 2019’s 432,484.6 No one disputes that the overall trend since 1945 has been a precipitous decline in strikes; as BLS notes: “In 2020 there were 8 major work stoppages, the third lowest number since 1947.”7 In short, in order to register all of these alleged upswings of the past few years—the kind vaunted in the breathless headlines—you have to dramatically stretch the y-axis.
Strikes didn’t decline because they fell out of favor with the working class, or because gutless bureaucrats replaced daring communists and socialists in union leadership, although that is a story you will hear a lot. Instead, they have been severely hemmed in over the course of decades. In the first place, workers can generally only strike after their contract has expired (or before they have one), because nearly all contracts contain no-strike clauses, i.e. a commitment not to disturb production during the life of the agreement. Of course, workers can theoretically defy such things (Southwest Airlines pilots and Rikers Island guards have been engaging in crippling mass sick-outs for months), but they probably won’t have the backing of their union leadership, and certainly not access to its strike fund. Union officials are generally unlikely to want a strike. They will sometimes be pro-strike when they really can’t get an acceptable offer at the bargaining table, but even then they still sometimes push a bad settlement rather than face the risks and challenges of a strike.
Relatedly, it’s important to recognize that there is a difference between actually striking and voting to authorize a strike. Many of the disputes heralded as part of Striketober were actually the latter, which are more common than actual strikes. Authorization votes are often feints by a union leadership to force a better “last, best, and final” offer from employers within a broader context in which employers are able to declare an “impasse” and impose that offer. Unions often take these votes without necessarily intending, or being prepared, to take their membership off the job. The International Alliance of Theatrical Stage Employees (IATSE) vote was exciting because of just how many workers it would have taken out (30,000–40,000), and how overwhelmingly they voted in favor of action (nearly 99 percent, with 90 percent of the membership voting), but members say the union leadership had even signaled to them that they weren’t entirely serious about going out.8 In fact, IATSE has never really gone on strike and the last strike by any crew was in 1945, likely before any current member’s lifetime. It would take a lot of work to organize a strike among IATSE members, especially since maintaining a picket line across such distributed workplaces as film sets could take considerable logistical coordination and collective discipline.
Again, the unfortunate reality is that strikes today are highly constrained and circumscribed events that minimize the impact on employers (but not on workers). Because they generally only happen after contracts have expired, employers have a date, known years in advance, to prepare for. And they do prepare: John Deere formed contingency plans that included training management to run production, however poorly; Kellogg’s has hired scabs to cross the picket line. Employers may cut striking workers off from their health insurance, as Kellogg’s has also done. Shockingly, they have the right to permanently replace striking workers (except in the case of an unfair labor practice strike, but all of the examples mentioned have been economic strikes). They may stock up production in advance so that they welcome an idle line or shift production elsewhere.
In other words, strikes are tough. They are less magic weapons that get workers what they want than they are necessary, bloody battles. For the most part, they result in a better contract settlement. Other times, they result in the devastation of the union. As a strike drags on, workers’ resolve may be strengthened or weakened. Sometimes employers will provoke a strike (or initiate a lockout) because they know this comes at a cost to workers, or they figure the workforce isn’t prepared and a strike will just demoralize and divide them. Or they want to draw down the union’s treasury by making them use up their strike funds, which is one reason why union officials of every stripe are very cautious about calling them.
In the case of John Deere, workers were able to secure a more significant raise by rejecting the company’s offer (twice!) and staying out; they also quashed the company’s attempt to introduce a “third tier” of workers, ineligible for pension upon retirement. However, they were unsuccessful in their bigger goal of eliminating the two-tier wage and benefit structure that has been in place since 1997. Some workers today still don’t earn as much as their forebears (in some cases, their literal parents) did then. Nor do they enjoy the pre-1997 hires’ post-retirement healthcare benefits, from a job that is physically punishing.
For their part, the 32,000 workers covered under various unions at Kaiser were successful in fending off the employer’s proposed two-tier structure with their strike threat. IATSE workers secured higher pay when working on productions for streaming services (they had agreed to lower pay when the format was still experimental) but not parity. They sought guaranteed breaks but instead got increased penalty rates for missed meals; they won 10 hours of guaranteed rest between work days rather than the sought 12.
When the left romanticizes strikes, it seems to forget that workers generally don’t like being on strike. (One of the primary union-busting talking points from employers is that the union will “make you go on strike.”) For as invigorating as pickets can be, as a friend once commented, a lot of time on the line is just spent eating hotdogs and watching passing cars, sometimes in the rain or cold. The long hours spent talking to one another and the collective struggle can bond a group of workers, but the betrayal of scabs and a disappointing resolution can also leave a bitter taste forever. Pickets can be rowdy or ugly or dangerous, with scuffles sometimes breaking out and picketers being hit by cars. Sometimes pickets can be whittled down to a handful of workers standing symbolically outside the gates, as happened with John Deere in Iowa.
Having said that, strikes can inspire and embolden other workers, which is why they are contagious. That used to be much more obvious, when one factory would go out and another would follow. Over time, tactics like secondary strikes—striking to support other fellow workers who are on strike at another company—were legislated against or barred in contract language because they were successful. Whether or not workers can respect others’ picket lines varies by jurisdiction and by their own contract language. Every effort has been made by government and business to mitigate contagion and to drive unionism into orderly exchanges of proposals one workplace at a time. Strikes are not meant to spread, which is why those looking for waves are at best tallying patterns in the data. As Bill Haywood of the Industrial Workers of the World (IWW) speechified in 1911: “the A. F. of L. couldn't have a general strike if they wanted to … They have 271,000 different agreements that expire 27,000 different minutes of the year.”9
The IWW repudiated contracts for that reason, seeking to organize all workers into “One Big Union” deploying direct action. It took aim at the practice of dividing workers into multiple craft unions: as William Trautmann complained, also in 1911, “One portion of workers, members of one craft union, remain at work, while others, members of another trade union, are fighting … The result is that no concerted action was possible.”10 A handful of other unions, including the Western Federation of Miners and later the CIO embraced “industrial unionism,” uniting all workers within an industry, but this approach by no means triumphed over craft unionism. For that matter, any certification drive today has to establish what the law calls a “community of interests” to put workers into the same unit. All of this, too, militates against strikes breaking out in waves. Unions today can coordinate with each other, as they did at Kaiser, or line up contract expiration dates, but these strategies do not have the latitude that enabled the general strikes of the 19th century or the 1910s and the mass strikes of the 1930s. If we want to continue to idealize those upsurges, as many do, we have to know why they aren’t materializing today.
There have always been deep disagreements within the labor movement on both strategy and principle. But compared to a hundred years ago, today’s union landscape is a monoculture: the scope of disagreement simply isn’t nearly as broad. From the early 20th century especially, government and employers conspired to create conditions in which the tamest version of labor unionism was incentivized and rewarded, and all other forms were edged out. That, and not the purging of communists, was what drew down militancy. As historian Nate Holdren explains, summarizing Charles Romney’s book Rights Delayed, the National Labor Relations Act:
Introduced complicated legal procedures to unionization. These procedures were slow in ways that hurt workers—employers could stand to wait in a way that workers without a paycheck could not. He details, for example, years-long delays in responding to unfair labor practice charges, already in the early 1940s. These procedures required legal professionals as well, which made them less than democratic—lawyers and other experts are generally not a force for collective decision-making; technocracy and democracy are at best an unstable combination—and were also very expensive. As a result, unions that were more progressive, democratic, and militant saw fewer benefits from the National Labor Relations Act (NLRA) and were in some ways hamstrung by it. More conservative, top-down, and conciliatory unions, on the other hand, did better: This history is important because it means the NLRA did not create an environment that was generally good for unions, so much as an environment that was good for specific kinds of unions.11
What other forms of unionism used to exist? Pretty much everything taken for granted today (paid staff, dues check-off, contracts) was once merely one practice among many, if not outright controversial, and that includes many of the structures and practices that prevent strikes from rolling into waves. Perhaps most significantly, workers in nearly all unions much more frequently took action directly on the work floor in immediate response to what they did not like. “It was common practice in the auto shops for negotiations on the shop level to consist of a steward, surrounded by all the men in a department, arguing with the foreman. No one worked until the grievance was settled,” wrote Martin Glaberman.12 Historian Nelson Lichtenstein reminds us that:
At Armour’s main Chicago plant, packinghouse union stewards developed a tactic they called “whistle bargaining.” Each steward wore a whistle around his or her neck. Whenever a supervisor declined to discuss a grievance, the steward gave a blast on the whistle and the department halted work. When the issue was resolved, the steward whistled twice and production started once again.13
As I have argued elsewhere,14 there was an evolution in the labor movement away from worker militancy on the floor until all-out strikes were the only tactic left—before these too were severely weakened. In the years following the passage of the National Labor Relations Act,15 and particularly after World War II, workers were bribed to trade away shop-level disruption for much higher wages and benefits.16 “The more ‘victories’ they recorded, the bigger and more technical the contracts became,” says Glaberman. “A contract is a compromise. That establishes that, no matter what union gains are recorded, the rights of the company to manage production are also recorded.”17 Grassroots forms of worker action were replaced by orderly bargaining and grievance handling, away from the shop floor, and increasingly dependent on professional staff. Likewise, decision-making was moved from the ranks to union leadership. With that came a tremendous diminishing of the agency of the working class.
More of that kind of ground-level activity seems to have kicked off during the pandemic, with workers engaging in direct bargaining with employers over things like health and safety and PPE, hazard pay, and so on. Some have suggested it is more common in non-unionized workplaces—Bloomberg Law analyst Robert Combs has pointed out that it’s actually “easier for non-union workers to walk off the job, as long as they do it together.”18 The United Electrical Workers and Democratic Socialists of America (DSA) set up the “Emergency Workplace Organizing Committee” (EWOC) to coach non-unionized workers through the process of coming together as a workforce to make demands. That’s because, contrary to “Striketober” commentary, abysmal conditions are not enough to make workers spontaneously fight and win. One EWOC organizer I spoke to estimated that probably 80 percent of workers who actually get to the point of forming a committee and taking action do win. The problem is how few of them get to that point. Collective action is not automatic or natural.
A lot of the “reporting” on trends in organized labor looks like licking your finger and sticking it in the air, prognosticating whether we are on an uptick of militancy. Explicitly, nearly every “Striketober” article took objective conditions to be responsible: First and foremost, the “labor shortage” gives workers more bargaining power. Second, the pandemic allegedly taught workers their worth, and/or foisted conditions on them that made them fed up. Relatedly, there is allegedly a historical relationship between worker uprisings and just having emerged from a period of social disruption.19 And finally, there is the “accumulation of grievances”—basically, the poor conditions under which workers work. Many tied the strikes to the high “quit rate”: “the same forces making work intolerable for so many—not enough workers and too much work—are simultaneously preparing workers to fight back,” argued Jonah Furman and Gabriel Winant at the Intercept.20 The problem, again, is that strikes in this day and age do not burst forth from worker inspiration, but are cautiously deliberated upon by union membership and leadership in a process that has been years in the making and is not entirely straightforward (a majority of IATSE members voted against the proposed settlement, but the agreement was ratified because of IATSE’s Electoral College-like voting structure).
We should be extremely skeptical of “conditions are ripe” arguments. I’ve never encountered an organizing campaign that was made meaningfully easier by workers being screwed over. Capitalism has already insidiously handed them adaptive tools for that: stew in anger, or better yet, hopelessness; resent other working people, either for making more or working less or trying to steal your job; aim for a promotion or maybe a job elsewhere; shrug that the world is unfair. It’s true that there is an expression, “The boss is a good organizer,” but where that usually helps is after workers have already started organizing and therefore developed some ability to talk to each other and collectively respond to challenges. The boss screwing up can tip the fence sitters or, in a slightly dormant union, it can provoke people back to action. But in a less organized workplace, it just alienates workers, perhaps to the point of quitting. What it doesn’t do is organize the work floor because, believe it or not, that is not a matter of how workers feel as individuals but a matter of them building relationships with each other.
It is true that workers are in a better position when unemployment is low, but even so, working-class power is a matter of organization—labor markets don’t create actions, only organizing does. Employers may raise wages to attract workers at a given moment but they will also lower them once they are less desperate. Likewise, in a workplace, what you can wrest out of an employer has to do with how organized you are, not how healthy profits are.
If you want to talk about objective conditions: objectively, companies that are deep in the black should offer their workers raises that keep pace with inflation during a time of allegedly “very tight” labor markets. And yet we just saw Kaiser proposing that new hires—essential health care workers now recognized as heroes by the public—be paid less than starting wages at McDonald’s. Why would they do that? Because employers understand that their relationship with the union is a power struggle. At every contract negotiation, the company asks for givebacks—regardless of the financial health of the company. Between negotiation sessions, they fight the union by ignoring the contract (Kellogg’s workers have complained that the company wasn’t making good on its commitment to move workers into the higher tier). Everything they ask for is a maneuver for power and they are playing the long game. Two-tier contracts aren’t just about saving the company money, but about driving a wedge between workers and making the union seem powerless. There are other cunning moves. John Deere has structured pay so that workers receive more when they meet or exceed production targets. Money gets placed into an account, and drains back out if workers fall behind. This is a particularly insidious form of the speed-up. The pace of work used to be one of the most fundamental and fought-over issues in the workplace; John Deere has put workers in the position of fighting to keep this form of financial discipline, in a production environment with significant dangers.
The class struggle never ends. When unions certify a workplace, the employer begins their campaign to punish workers for it on day one. They slow-walk negotiations, or try to force workers to accept less than what they currently have. A certification is not a win; it’s just a gauntlet thrown down. The employer will not rest until they render the union’s presence meaningless, and/or legally decertify it. Unions’ ability to persist in winning concessions and protecting workers lies in members’ willingness and ability to continue to mobilize for ongoing struggle—not just during bargaining time, but at all times.
Measuring worker strength and militancy is more complicated than observing occasional, statistical upticks in strikes, especially given what we know about the bureaucratic structuring of unions. The most important things that could happen to rebuild the agency and ultimately the fighting capacity of the working class are things that wouldn’t necessarily make headlines, but happen behind the scenes, and are harder to quantify than strike votes. These would include workers talking to each other and collectively deciding on actions to take—and taking them, as they did in decades long past.
If you tie worker militancy to objective conditions, as much of the Striketober commentary did, then you are left wondering and prognosticating when the next upsurge will be. But such explanations mystify worker agency. This is a perspective you will actually see articulated: Rich Yeselson, a former union staffer turned writer, has explicitly argued that the labor movement should simply do nothing and wait for workers to rise up spontaneously: “Wait for the workers to say they’ve had enough. When they demand in vast numbers collective solutions to their problems, seize upon that energy and institutionalize it. That is how massive union growth occurs.”21
Labor scholars Michael Goldfield and Cody R. Melcher wrote more recently that “the biggest increases in union membership and strikes have not happened incrementally, but in enormous, often unforeseen upsurges … growth often occurs more or less automatically” (my emphasis).22
Kim Moody, quoting Eric Hobsbawm, writes, “‘explosive situations’ are the result of ‘accumulations of inflammable material which only ignite periodically, as it were, under compression.’” Bringing it back to Striketober, Moody notes “the declining conditions of working-class life and the grievances they have spawned … might interact to produce a continued upswing in working-class militancy and activism.”23 And yet, he admits, conditions have been worsening (wages stagnating while productivity went up) for some time.
The problem with tying worker activity to objective conditions is that, though you have a causal hypothesis on your hands that you can put to empirical test, the evidence doesn’t quite fit it.
We have a tendency to erase the deliberate organizing that goes into upsurges and just remember the flashbulb moments. Take the Lawrence textile (“Bread and Roses”) strike, which is told of as though it were a spontaneous walkout when employers shorted workers of pay. As G. DeJunz showed, there were actually decades of underground organizing by individuals who themselves had decades of previous experience organizing in the textile industry in Europe.24 There were also numerous smaller actions prior. There was also a very different legal landscape and organizing practice in 1912, which allowed the strike to spread from one mill to another.
The context today is different. The last several decades have seen a shackling of labor unions to prevent disruption on that scale. On rare occasions, union workers have defied those constraints—Stan Weir’s wonderful “A New Era of Worker Revolt” documented wildcat strikes and radical turns among autoworkers, longshoremen, miners, painters, and airline mechanics in the 1960s. Those members often found themselves at odds with their own union leadership; the unions that dared to be more rowdy were raided, merged, or put into receivership. The system is built to find “reasonable” negotiating partners. A tiny increased frequency in strikes today is not the unshackling we might initially take it to be. To accomplish that requires putting into question some of the cornerstones of organized labor as we know it, including the centrality of contracts.
The current barriers to the spread of strikes are no reason for defeatism but they are reason to radically rethink existing practice. Some have pointed out the institutional memory of striking that gets lost when strikes become as infrequent as they have. Before that, there was an earlier loss, of a different institutional memory—actually a day-to-day practice—of workers collectively defying management and insisting on control of work. An unruly workforce settling grievances and “whistle bargaining” on the spot. This can be revived through worker education and practice to become habit again. That may seem difficult to imagine, but consider the strike: Every strike is an act of pure bravery. Every time a worker trades their wages for paltry strike pay and long days on a picket line, taking on the risk even of being permanently replaced, they show what the working class is capable of.
“Number of striking workers surged in 2018 and 2019,” Economic Policy Institute, https://www.epi.org/press/number-of-striking-workers-surged-in-2018-and-2019-data-show-largest-two-year-average-in-35-years/.
Andrew Van Dam, “Teacher strikes made 2018 the biggest year for worker protest in a generation,” Washington Post, February 14, 2019, https://www.washingtonpost.com/us-policy/2019/02/14/with-teachers-lead-more-workers-went-strike-than-any-year-since/.
Robert Reich, “Is America experiencing an unofficial general strike?” The Guardian, October 13, 2021, https://www.theguardian.com/commentisfree/2021/oct/13/american-workers-general-strike-robert-reich.
Alisa Chang, Alejandra Marquez Janse, Courtney Doring, “’Striketober’ could have lasting impact,” NPR, October 28, 2021, https://www.npr.org/2021/10/28/1050177227/striketober-could-have-lasting-impact-on-labor.
Edward M. Smith, “’Striketober’ has been a long time coming and is only the beginning,” The Hill, October 27, 2021, https://thehill.com/opinion/civil-rights/578562-striketober-has-been-a-long-time-coming-and-is-only-the-beginning.
Kim Moody, “Upticks, Waves, and Social Upsurge,” Spectre, November 15, 2021, https://spectrejournal.com/upticks-waves-and-social-upsurge/.
“Work Stoppages Summary,” Bureau of Labor Statistics, February 19, 2021, https://www.bls.gov/news.release/wkstp.nr0.htm.
Sarah Hughes, “Interview: Why I’m Voting No on the IATSE Hollywood Basic Tentative Agreement,” Labor Notes, November 9, 2021, https://labornotes.org/2021/11/interview-why-im-voting-no-iatse-hollywood-basic-tentative-agreement.
William Haywood, “The General Strike,” in Rebel Voices: An IWW Anthology, edited by Joyce Kornbluh, (Ann Arbor: University of Michigan Press, 1965), 49.
William Trautmann, “Why Strikes Are Lost,” in Rebel Voices: An IWW Anthology, edited by Joyce Kornbluh, (Ann Arbor: University of Michigan Press, 1965), 19.
Nate Holdren, “’Union’ is a verb,” Organizing Work, March 31, 2021, https://organizing.work/2021/03/union-is-a-verb/.
Martin Glaberman, Punching Out and Other Writings, edited by Staughton Lynd (Chicago: Charles H. Kerr, 2002), 8.
Nelson Lichtenstein, State of the Union: A Century of American Labor (Princeton University Press, 2002),60-61.
Marianne Garneau, “Big Strikes and the Sabotage of the Labor Movement,” Organizing Work, November 12, 2020, https://organizing.work/2020/11/big-strikes-and-the-sabotage-of-the-labor-movement/.
An act to diminish the causes of labor disputes burdening or obstructing interstate and foreign commerce, to create a National Labor Relations Board, and for other purposes, July 5, 1935; General Records of the United States Government; Record Group 11; National Archives.
“Treaty of Detroit and Postwar Labor Accord,” The Oxford Encyclopedia of American Business, Labor, and Economic History, edited by Melvyn Dubovsky (Oxford University Press, 2013).
Glaberman, op cit, 14-15.
Robert Combs, “Analysis: Covid-19 Has Workers Striking. Where Are the Unions?” Bloomberg Law, April 14, 2020, https://news.bloomberglaw.com/bloomberg-law-analysis/analysis-covid-19-has-workers-striking-where-are-the-unions.
Harold Myerson, “In Hollywood and America, the Strike Is Back,” American Prospect, October 7, 2021, https://prospect.org/labor/in-hollywood-and-america-the-strike-is-back/.
Jonah Furman and Gabriel Winant, “John Deere Strike Shows the Tight Labor Market is Ready to Pop,” The Intercept, October 17, 2021, https://theintercept.com/2021/10/17/john-deere-strike-labor-market/.
Rich Yeselson, “Fortress Unionism,” Democracy: A Journal of Ideas, No. 29, Summer 2013, https://democracyjournal.org/magazine/29/fortress-unionism/.
Michael Goldfield and Cody R Melcher, “Moments of Ruptire: The 1930s and the Great Depression,” Organizing Upgrade, November 9, 2021, https://www.organizingupgrade.com/moments-of-rupture-the-1930s-and-the-great-depression/.
G. De Junz, “The Intensive Organizing Behind One of History’s Most Famous ‘Spontaneous’ Strikes,” Organizing Work, April 14, 2020, https://organizing.work/2020/04/the-intensive-organizing-behind-one-of-historys-most-famous-spontaneous-strikes/.