What if the “Next Upsurge” Never Comes?

Growing labor's power today will depend on our own dogged efforts against the grain of today's social and political conditions.

Hot Labor Summer. Striketober. There’s been a lot of buzz around the labor movement, particularly among younger workers, since the onset of the Covid-19 pandemic. The buzz is warranted. Since early 2021, there has been a notable uptick in both strike action and union organizing. Popular, high-profile strikes among autoworkers, higher education workers, public school educators, Hollywood actors and screenwriters, and others brought the number of major strikes in 2023 to the highest level in two decades. Starbucks workers have successfully pressured the company to accept a collective bargaining framework for unionized stores. The United Auto Workers (UAW), fresh off their strike win over the Big Three automakers, have announced a potentially historic campaign to organize non-union auto plants in the South. As of this writing, workers at Volkswagen’s plant in Chattanooga, Tennessee and a Mercedes plant in Alabama have filed for union representation elections with the National Labor Relations Board (NLRB). If successful, these campaigns could begin to plant the flag of unionism in a region of the country that has historically been the movement’s Achilles heel.

These are very exciting developments that democratic socialists support in every way possible. At the same time, it is crucial to stay grounded and keep all of this in historical perspective. Last year’s strike action, while important, was an uptick from an extremely low baseline of strike activity. Both the frequency and the overall scale of strikes remain far below pre-1980s levels, and to date they have been confined largely to groups of already organized workers. Moreover, the uptick in workplace collective action has yet to translate into meaningful new gains in union organizing. The absolute number of union members grew by roughly 140,000 in 2023, but the rate of union membership (also known as “union density”) declined to an historic low of 10%. It is surely better for the overall number of union members to rise than to fall. But the rate of union membership growth is currently too slow to keep pace with the growth of the workforce, much less to overtake it and raise union density.

Faced with these realities, it makes sense for organizers to look at the moments in history when labor took great leaps forward for inspiration. The lodestar is the enormous upsurge in unionization that took place in the 1930s and 1940s, when the Congress of Industrial Organizations (CIO) led the charge to organize the mass production industries in the United States. From our vantage point in the present, it can be difficult to truly assimilate the scale of what the CIO – and, let us not forget, the American Federation of Labor (AFL) – accomplished in that epochal time. The numbers are mind-boggling. In his indispensable book The Turbulent Years, historian Irving Bernstein recounts how the number of union members in the US rose from about 2.8 million in 1933 to over 8.4 million in 1941. This tripled labor’s ranks in less than a decade, raising union density from roughly the level it stands at today to nearly 25% of all non-agricultural employment. There was union growth everywhere, but the biggest leaps happened in the mass production industries of the private sector. Union density in manufacturing exploded, from less than 9% in 1930 to over 34% in 1940. While the biggest gains happened in the Northeast, upper Midwest, and Pacific regions, the rate of union membership grew by over 10% in the South during the Depression decade. This was a major breakthrough that finally established the legal right of workers to organize and strike, raised working-class living standards, and realigned the basic patterns of US politics for decades.

It’s become axiomatic that unionization grows not gradually but in spurts of the kind that happened in the 1930s-1940s. In his study of the Depression decade, economist Richard Freeman defined a spurt as “a sharp concentrated episode in union growth,” in which “membership should grow more rapidly in a few contiguous years than would be expected” from random variation. On this basis, he concluded that “unionism generally grows in discontinuous spurts” that, as in the case of the Depression-era upsurge, appear suddenly and unexpectedly. Similarly, Marxist historian Eric Hobsbawm found that the history of working-class formation is marked by “explosions” resulting from “accumulations of inflammable material which only ignite periodically, as it were under compression.”

These insights have informed the strategic thinking of many left labor organizers. But we are now 90 years removed from the upsurge that organized much of the private sector, and overall union density has been in continuous decline since the 1950s. The near disappearance of private sector unionization was only partially mitigated by the growth of public employee unionization, and the judicial imposition of “right-to-work” on the whole public sector after the 2018 Janus v. AFSCME decision is sapping the foundations of labor’s last redoubt. Faced with these harsh realities, socialists and labor organizers need to ask ourselves an unsettling question: what if the next upsurge never comes?

In his unjustly forgotten book The Transition from Capitalism to Socialism, John D. Stephens identifies the two main drivers of previous leaps in union organization – and concludes that they are not likely to occur again. The first is the political and social impact of the world wars. “During and immediately after the two world wars trade union membership increased rapidly, particularly in the defeated powers.The disastrous consequences of nuclear war,” Stephens concludes, “rules out this factor for the future.” The second is the “combination of severe economic crisis and the accession of leftist parties to power (crisis alone causes membership to deteriorate). This has occurred only once,” Stephens observes, “so it might have been an idiosyncratic unrepeatable event.”

There are two additional developments that, in my view, could forestall the development of a new upsurge on the order of the 1930s. The first is a product of the 1930s upsurge itself: the establishment of the National Labor Relations Board (NLRB) and other institutions that recognize, in however constrained a fashion, labor’s right to organize and serve to mediate conflicts between workers and employers. The second is the dramatic growth of service sector employment at the expense of industrial and manufacturing employment.

To be clear, my argument is not that unions and other labor organizations are doomed to die a slow, painful death under the conditions of contemporary global capitalism. My argument is that the main drivers of the one big leap in working-class organization are absent, raising the possibility of a relatively modest best-case scenario for unions in the US: growth without a 1930s-style explosion in the level of union organization. If this is indeed the case, it raises challenging questions about the place of organized labor in democratic socialist strategy and progressive politics in general today.

Labor’s Wars

If you look at a line graph tracing the level of US union density over time, you might come to the perverse conclusion that World War III is the best thing that could happen to the labor movement. That’s because the two biggest spikes in union membership in this country happened just before and during the two world wars. Nineteenth-century unions tended to be unstable organizations that rose and fell along with the success or failure of strikes. Workers lacked a clear legal right to organize, the unions that did exist were quite precarious, and the courts were deeply hostile to strikes as well as legislation to improve working conditions. Workers waged many bitter struggles for rights, recognition, and power, but the labor movement struggled to raise the level of union density above a minimal level.

World War I’s impact on the subsequent development of the US labor movement was profound, and profoundly contradictory. The war decimated the Socialist Party, which never recovered from the twin blows of government repression and the Communist split in 1919. At the same time, wartime mobilization generated a relatively hospitable environment for union organizing under the aegis of “industrial democracy.” In Labor’s Great War, historian Joseph McCartin argues that wartime conditions “eroded managerial control and ignited explosive labor militancy. This in turn triggered an abrupt, vast extension of federal government power expressed by both the repression of radicals and an unprecedented commitment to the democratization of the workplace.”  No less a labor radical than Willam Z. Foster, who helped lead the drive to organize Chicago packinghouses, established the Trade Union Education League, and served as chairman of the Communist Party for many years, refrained from opposing the war to focus on organizing amid wartime labor shortages and National War Labor Board-sponsored shop committees. By 1919, union density had popped from its prewar level of roughly 10% to nearly 20%, easily the highest level in US history to that point.

After the armistice, federal labor policies that boosted unions were dismantled and employers waged a bitter counteroffensive amid a sharp postwar recession. Labor’s wartime organizing gains were reversed, and by the end of the 1920s union density slumped back to its prewar level of about 10%. The onset of the Great Depression in late 1929 triggered popular unrest, but it did not immediately translate into an upsurge in strike action or union organizing. It took until 1934 for the wave to come, and it was a big one. In The Turbulent Years, Bernstein reports that nearly 2000 strikes involving roughly 1.5 million workers swept the country, including four mass strikes – auto parts workers in Toledo, Ohio; truck drivers in Minneapolis; West Coast longshoremen; and textile workers in New England and the South – that resembled revolutions in miniature. The wave kept rolling through 1937, when the legendary UAW sit-down strike against General Motors opened the path to organizing the auto industry. In a few short years union density more than doubled, to nearly 25% of the entire workforce by the late 1930s.

Despite this great leap forward, labor’s gains were by no means assured. Many newly minted union locals evaporated as quickly as they appeared, and major citadels of employer recalcitrance still remained – namely the steel industry, which continued to resist unionization after the UAW’s breakthrough in auto. In Labor’s War at Home, historian Nelson Lichtenstein notes that the sharp recession that hit in late 1937 “put an abrupt halt to most organizing work between 1938 and 1940.” Republicans and Southern Democrats handed Roosevelt an important defeat in the 1938 midterm elections, which brought the New Deal’s reformist momentum to a screeching halt. “A year before Pearl Harbor,” Lichtenstein reminds us, “the CIO remained a tentative and incomplete structure” – a fragile juggernaut, in the words of labor historians and podcasters.

World War II offered the conditions under which labor organized the last major industrial holdouts and consolidated the gains it made in the 1930s. During the war years, Lichtenstein writes, “unions completed the organization of basic industry, more than doubled their dues-paying membership, and fortified themselves with such an array of contractual and administrative defenses” that the problem of organizational security was solved. Wartime labor policies that were piloted during World War I, like the National War Labor Board, were rebooted and expanded and joined by new federal agencies like the War Production Board and the Office of Price Administration, both of which had extensive union representation. The AFL and CIO agreed to a no-strike pledge for the duration of the war (many unofficial strikes nevertheless happened), and the 1943 Smith-Connally Act granted the federal government the power to seize industries where the threat of strikes could disrupt war production. At the same time, Washington’s dire need for uninterrupted production allowed unions to win concessions that resulted in further growth and organizational security. By the end of the 1940s, Lichtenstein writes, the labor movement “had organized at least 80 percent of all workers in basic industry. As a proportion of the nonfarm workforce, total union membership was more than 35 percent in 1954, and for the first time, American unions enrolled a section of the working class that approached northern European levels.”

War is hell – there is nothing “good” about it. The point here is that the dynamics of union growth were closely bound up with the relationship between foreign war and domestic politics. In the twentieth century, total wars between great powers required mass mobilization on an unprecedented scale, which in turn required governments to leverage both coercion and consent. In her landmark book Forces of Labor, sociologist Beverly Silver demonstrates how workers’ bargaining power visàvis states and employers increased dramatically with the massive demand for labor in factories and on the front lines. In this context, workers were able to win economic and political concessions that were not possible in the absence of great power warfare. The great power wars of the early twentieth century depended on the active consent and participation of the masses, who demanded more rights, security, and democracy in return for being hurled into a blazing slaughterhouse. By contrast, the interventions and counterterrorism operations of recent decades have relied on a small and isolated segment of the population, forestalling the kinds of democratizing pressures that built up and exploded around the two world wars. As Silver concludes, “one of the most powerful processes underlying the expansion of workers’ and democratic rights is being reversed, raising the question of whether this reversal is facilitating a major contraction in workers’ and democratic rights.” Labor’s record of largely unmitigated organizational decline since the 1950s suggests that this may indeed be happening.

Depression Lessons

It is hard to imagine the level of human suffering the Great Depression brought to this country. When Franklin Roosevelt took office in early 1933, the economy was essentially stalled. One-quarter of the US workforce was unemployed, poverty and hunger stalked millions, and massive shantytowns popped up in cities and towns around the country. Private relief was completely unfit for the scale of need, and government intervention was piecemeal and threadbare. There have been numerous crises and recessions since the 1930s, some of them quite severe. But none of them have approached the scale of the Great Depression, which remains a singular event in the history of global capitalism.

Crisis alone does not necessarily spur working people to organize in self-defense. The already low level of union density in the US dropped further after the 1929 Wall Street crash, and did not begin to rise again until well into the 1930s. It was the combination of miserable conditions and the accession of Franklin Roosevelt and the New Deal to power in 1933, a development that raised workers’ expectations concerning the role of government in their lives, that encouraged workers to strike and organize on a truly massive scale. As noted above, John D. Stephens speculates that the conditions that drove the 1930s upsurge “might have been an idiosyncratic unrepeatable event.” Many economists thought the apparatus of economic regulation and welfare spending that was built during the New Deal and World War II effectively put an end to the business cycle. They were, of course, wrong. Capitalism continues to suffer recurrent crises and slumps that are an essential feature of the system. But politicians and economists seem to have learned an important lesson from the Great Depression – a worldwide crash of that terrible magnitude simply cannot be allowed to happen again.

The two biggest recessions after World War II were the early 1980s slump and the Great Recession of 2007-2008. Both downturns were severe and working people suffered, but these did not come anywhere close to matching the scale of the Great Depression. “Automatic stabilizers” like unemployment insurance, food stamps (SNAP), and Medicaid did not exist a century ago, and while they are inadequate they prevent the roof from completely caving in on millions of people during recessionary periods. Additionally, policymakers have shown themselves to be willing to intervene massively when crisis strikes. The 2007-2008 crash could have rivaled the Great Depression, but it did not because of a globally coordinated response that prevented the crisis from spiraling completely out of control. Domestic stimulus spending could and should have been higher, millions lost their jobs, and many homeowners saw their mortgages go up in smoke. Nevertheless, the unemployment rate peaked at 10% in 2009, far below the 25% unemployment rate that ravaged the country in the 1930s.

The economic impact of the Covid pandemic, as bad as it was, could have been far worse. Unemployment exploded from just 3.5% in February 2020 to 15% in April. The public health side of the pandemic response was disastrous, but the economic intervention was massive and highly successful. Unemployment insurance, according to one study, “was the most important single element of the fiscal response to the pandemic” by providing benefits to over 50 million workers and injecting nearly $1 trillion into the economy by August 2021. At the same time, policymakers sponsored a massive expansion of welfare spending that had a dramatically positive impact on the lives of working people. Families received three rounds of stimulus checks, average federal SNAP benefits were doubled, and an enormous expansion of the federal child tax credit cut childhood poverty in half in a single year. Unemployment fell back to its pre-pandemic level by early 2022, and it has remained at a low level ever since.

Crisis is still endemic to capitalism, but policymakers seem to have learned that mass unemployment is so politically destabilizing that it must be avoided at all costs. Future generations could well forget the lessons of the Depression, and rising tensions between great powers could hamper the international coordination that is crucial to preventing localized crises from going global. But if the policy response to recent economic crises is any guide, the net damage they cause today may be high enough to inspire some degree of workplace organizing and political turbulence, but not nearly enough to facilitate an upsurge on the order of the 1930s.

Solvents of Revolution

The National War Labor Board’s dissolution in 1919 ended the federal government’s brief experiment in regulating and institutionalizing class struggle. That year also saw a massive explosion in strike action, with mixed results for organized labor. The biggest losses were in steel, where employers smashed a major organizing drive, and in the public sector, where the state of Massachusetts so thoroughly suppressed the Boston police strike that it set the cause of public employee unionism back for nearly half a century.

Labor’s political fortunes began to change in 1926, when Congress passed the Railway Labor Act to facilitate collective bargaining and mediation in order to avoid strikes in the railroad (and later, airline) industry. The 1932 Norris-LaGuardia Act, passed before Roosevelt’s election and the 1930s strike wave, banned “yellow dog” contracts and restricted the use of federal court injunctions against unions during strikes. The law was an important advance for labor, as it committed the federal government to a policy in favor of protecting most workers’ right to organize and strike. It ensured that a more favorable legal and political environment would be in place when the big strike wave hit in 1934, and provided a foundation on which subsequent reforms were built.

The landmark reform, of course, was the 1935 National Labor Relations Act (NLRA), also known as the Wagner Act. The Wagner Act had roots in the Norris-LaGuardia Act and the National Industrial Recovery Act, which the Supreme Court ruled unconstitutional in 1935. The Wagner Act guaranteed private sector workers the right to organize, bargain, and engage in “concerted protected action” including strikes. It designated the NLRB to implement the law, including the administration of representation elections. Putting a federal government agency in charge of running the process of union representation meant that workers didn’t necessarily have to win their unions through bitter and risky recognition strikes – the main way that workers won their unions before the Wagner Act. In his study of spurts in union growth, Richard Freeman demonstrates the decline of recognition strikes and the rise of NLRB elections after the Supreme Court affirmed the constitutionality of the NLRA in 1937. According to Freeman’s statistics, there were 2200 recognition strikes in 1937 alone, a staggering number that entailed the unionization of nearly one million workers. 73 percent of all workers unionized that year were unionized through recognition strikes. That percentage dropped to 24 in 1938, and down further to 3 percent in 1944. Meanwhile, representation elections became the method by which most workers became union members. From 1941 through 1944 at least one million workers unionized through representation elections each year.

Recognition strikes have not entirely disappeared, but they have become extremely rare. According to the Cornell University Labor Action Tracker, there were just 13 recognition strikes in 2023 involving 476 workers. In his study, Freeman argues that “the key condition for growth spurts rather than the gradual growth of unionization is confrontation over the union institution.” So long as government policy did not accept the basic legitimacy of the union institution, Freeman’s key condition was present. There hasn’t been an organizing upsurge in the private sector since the achievement and consolidation of the NLRA. And there hasn’t been one in the public sector since the public employee organizing spurt of the 1960s and 1970s led to the establishment of union representation and collective bargaining systems in many states. Many employers have never accepted the fundamental legitimacy of unions and collective bargaining. They fight them tooth and nail, and the legal-institutional protections for union organizing through the NLRB process have become far too weak. But workers won the NLRB in the 1930s, and it’s still here. Whatever its weaknesses and limitations, workers will typically use it to organize instead of riskier efforts like recognition strikes. Mark Beissinger, a political scientist who studies revolutions, argues that “democracy is the great solvent of revolution” because it offers the possibility of achieving your demands without risking your life or liberty. In that sense, the recognition and institutionalization of unions is analogous to the ways in which capitalist democracy renders revolutionary politics unviable. So long as this modicum of workplace democracy exists, it will tend to channel conflict into a less disruptive government-sponsored process that forestalls the kind of upsurge that helped give birth to it in the first place.

Pervasive Fragmentation

The 1930s upsurge generated union growth across the entire workforce. That growth, however, was not evenly distributed by sector or occupation. The biggest leaps in union density were concentrated in industrial and manufacturing employment, which accounted for a much larger share of the workforce in the 1930s-1940s than they do today. According to Freeman’s estimates, union density in manufacturing nearly quadrupled from 1935 (10.6%) to 1947 (40.1%) bringing millions of workers in the mass production industries into the labor movement, primarily through the CIO. Notable leaps also occurred in transportation, communications, and utilities as well as building and construction trades, where the AFL maintained its predominant position. Union density in private services, which accounted for roughly half of total employment at the time (and over 70 percent today) tripled, but from an extremely low level. According to Freeman, union density in this sector grew from 2.8% in 1935 to 9.2% in 1947. That is a higher unionization rate than what exists in the entire private sector today (just 6%), but the fact that density in private services remained quite low amid the biggest unionizing upsurge in US history is sobering. Even government employees, who were excluded from the rights and protections afforded by the Wagner Act, had a higher unionization rate in this period according to Freeman’s estimates.

Today, roughly 85 percent of all US employees work in private service providing occupations (about 70%) or in the public sector (about 15%). The share of workers employed in manufacturing has dropped from roughly 40 percent during the 1940s to just about 8 percent; mining and logging, two industries at the vanguard of radical labor militancy in the twentieth century, dropped from around 3 percent to less than half a percentage point; construction stands at its rough historical average of about 5 percent. According to Census Bureau data, 43 percent of all employed civilians are now in management, professional, and related occupations, constituting the single largest occupational group by far. The next largest group, sales and office occupations, constitutes 19 percent. This means a significant majority of everyone working in the US today is employed in some sort of professional or white-collar occupation.

In The Transition from Capitalism to Socialism, Stephens speculates “whether the economic structure of large countries” like the United States “does result in some strong structural restraint to labor organizing.” Neither Stephens nor I suggest that organizing white-collar and service workers on a large scale is impossible. As Stephens sensibly notes, “labor organization can be increased through voluntary human action.” The industrial working class, however, has historically been the vanguard of union organizing in the US and similar countries. Their power stemmed significantly from their strategic location in the heart of the production process, but not just there. Historical patterns of industrialization also tended to concentrate workers from the same or similar workplaces together in urban neighborhoods, thereby enhancing their capacity to form trade unions, political parties, and other forms of organization. They were able to combine disruptive potential with organizational capacity in ways that other sections of the working class have typically not been able to. That unique combination powered many of the victories that working people in the US were able to win in the twentieth century.

Stephens finds that “the level of white-collar organization is the single most important source of variation in labor organization” between the countries he observes in his study, which includes the US. But what accounts for differences in the level of white-collar organization? According to Stephens, the answer is to be found in the political realm. “The most important cause of variations in the growth of organization in the post-war period,” he concludes, is “incumbency of socialist parties” that use public policy to encourage unionization. The US, of course, has never had a socialist party capable of winning power on a scale that could fundamentally alter the balance of power between workers and employers. At times, however, the alliance between unions and liberal-labor factions of the Democratic Party has served as a partial equivalent of the labor and socialist parties found in many other countries. The task facing the labor movement and democratic socialists is to build, together with other democratic and progressive social movements, a social democratic faction in the Democratic Party dedicated to protecting and promoting labor organizing as strongly as possible. The Squad is serving this purpose in embryo, but it is still too small to help shape the agenda of the party as a whole.

The extensive privatization and fragmentation of social life under post-industrial capitalism, layered on top of the already existing differentiation of the US working class by race, gender, national origin, citizenship status, religion, and other conditions makes it very hard to construct collective actors capable of effectively challenging capitalist power. In his final book, How to Be an Anti-capitalist in the 21st Century, the Marxist sociologist Erik Olin Wright observes how “the class structure has become more complex in ways that undercut the shared sense of fate and life conditions…there is a pervasive fragmentation of lived experience that makes a common class identity difficult to forge.” This is especially true in a time when most people work in the extremely heterogeneous service sector, which runs the gamut from fast food workers to nurses to software developers. The basic experience of waged and salaried labor may be widely shared, but the specific modalities of that experience are extremely varied. Wright names a few of them: the “level and security of earnings; precariousness of employment; autonomy within work; skill levels and education required within work; opportunities for creativity; and so on.” The upshot is that the working class is, as Wright concludes, “fragmented with divergent interests” from one segment of the class to another. “Many people still experience class as a salient identity, but it does not provide the universalizing basis for solidarity for which progressives once hoped.”

Politics to the Fore

If the main historical drivers of working-class upsurge are absent, and the class structure of contemporary capitalism makes it even more difficult to organize workers than it was before, what are the implications for our political strategy? In Forces of Labor, Silver concludes that “the twentieth-century trend toward increasing workplace bargaining power,” concentrated among manufacturing workers in industries like auto, “is at least partially being reversed in the twenty-first century….As such, we might expect the weight of associational power in the overall power strategies of labor movements to be on the increase.” Silver is probably right about the shift away from workplace bargaining power toward associational power, which includes various forms of collective organization both inside and outside the workplace. In that sense, twenty-first-century labor movements might resemble those of the nineteenth century more than those of the twentieth, with unions engaged in routinized collective bargaining constituting just one expression of organized working-class power among others.

Silver’s insights are reinforced by those of Göran Therborn, one of the leading theorists and scholars of the radical left since the 1970s. In the lead essay of his recent collection Inequality and the Labyrinths of Democracy, Therborn argues that the “dialectic of industrial capitalism, which Marx analyzed and predicted with impressive accuracy, is no longer operating in the Global North and has been stymied in the South. Post-industrial capitalism is no longer producing a growing, ever more concentrated working class.” Since contemporary capitalism does not possess a potentially progressive structural trend, making progress is now “more dependent on political mobilization and leadership” than it was before. Moreover, Therborn argues, in today’s world “the labor movement is only a necessary component of egalitarian politics, no longer sufficient as its natural center. Decisive to any successful egalitarian politics in the post-industrial era is a positive middle class policy of the left.”

Therborn uses “middle class” to describe professional and white-collar employees – many of whom are actually in the working class according to Marxian definitions, others of whom are not. Workers in sociocultural services, disproportionately women with high educational attainment but relatively modest salaries, have become the core left-wing constituency not just in the US but across the rich capitalist democracies. They are also among the most highly unionized workers in the country. Workers in education, training, and library occupations, for example, have the highest unionization rate among all occupational groups. The key challenge for the left is building beyond this core constituency to win over, as Therborn describes them, people “in the bottom half of the population, the losers of neoliberal globalization.” One important way to do this is to build organized labor’s power in the service and retail sectors, where employment has exploded but unions are rare. But this will take time, and we cannot bank on a general social explosion to do much of the organizing work for us. The growth of unions and other workers’ organizations in today’s environment will depend, to a very great extent, on our own dogged efforts to build power against the grain of today’s social and political conditions – not in preparing to take advantage of a “next upsurge” that shows few signs of arrival.

Image: Amazon warehouse workers outside picketing outside of the New York City office of the NLRB on October 25, 2021. Photo by Joe Piette and licensed under the Creative Commons Attribution-Share Alike 2.0 Generic.