Historian Nate Holdren questions the assumptions about why strikes have risen and fallen over the last century. Image © Emé Bentancur.
This fall, commentators and news media speculated over-optimistically that we may be in a strike wave, sometimes called “Striketober.” In an article in the Brooklyn-based art magazine The Brooklyn Rail, Marxist theorist Jason Smith expressed some healthy skepticism about these claims. I share Smith’s view that we were not in a strike wave, but I have some differences with his reasoning and with his choice of historical reference points. Specifically, the decline of the labor movement is not explained by a decline in profit rates, but rather by institutional transformations that have hemmed in the labor movement’s tactics and ultimately changed what kind of labor movement we have.
Smith argues that people have overstated the extent of strike activity and labor movement militancy in 2021, and that we shouldn’t expect big increases in strikes, though he seems to hold out hope for a kind of 1990s-style social movement unionism if the labor movement can figure out how to follow the lead of conflicts happening outside waged workplaces, above all struggles over police murdering African Americans.
In one sense this is pretty uncontroversial stuff in a lot of labor movement circles – this is the stuff of social unionism, bargaining for the common good, etc. To be fair, as a communist revolutionary, Smith imagines scenarios playing out that would be rejected by most union officials and pro-union politicians. The question is what role workplace struggle plays in this. What Smith seems to share with these other perspectives is a relative minimizing of the waged workplace as a site of political conflict and workplace organizing as a political activity. Workplace-based politics are at most secondary to activity in other social locations and can only be secondary because objective conditions so dictate.
Whatever one thinks of the integration of broad social struggles with the labor movement (my own view is that it can only be a positive thing), Smith presents a rather dim view of what can arise from workers’ struggles in and around wage workplaces or from labor movement organizations. This puts him at odds with, for example, the Marxist sociologist Simon Clarke, who has argued that, “relatively more marginalised groups and strata” are likely to practice “forms of opposition [which] remain fragmented, isolated and ephemeral unless they are integrated into a broader labour movement, the only secure base of which has proved to be the trade union organisation that develops out of the struggle over the terms and conditions of wage labour.”
Do strikes depend on economic growth?
Much of Smith’s article is devoted to pointing out that there were very large numbers of strikes and people on strike in some years in the 1940s and 1950s, far fewer in the years since, especially from the early 80s to the present. The core of Smith’s argument is that labor militancy relies on how the economy is doing as a whole. The 1940s through the 1970s, as a time of relative economic boom, particularly in the years just after the Second World War, was “the high point of the US labor movement,” he argues. “[I]t is only in such conditions,” Smith says, where “profit rates are elevated, production is augmented, and the demand for labor rises” that “workers can make demands for a larger share of the profits produced in the private economy.” The labor movement and union militancy can’t grow without economic growth, and since economic growth is sluggish nowadays, his argument goes, unions and strikes can’t really grow. Furthermore, since Smith seems to think it’s unlikely that capitalism can ever again deliver high profit rates, presumably for him labor movement and strike growth is unlikely indefinitely.
Smith’s argument makes some sense intuitively. A reader might say “after all, if capitalists don’t have profits they don’t have money to pay workers.” That’s a mistake, though, because the rate of profit isn’t the amount of profit. Say I invest a hundred dollars in a small record label and I make a hundred dollars in profit, while Jeff Bezos invests a hundred million dollars and gets back a million dollars in profit. I would have a far higher rate of profit than Bezos, but he’s the billionaire, not me, and can afford to pay a lot more wages than I can. Smaller profit rates on very large sums of money still mean very large sums of profit. Unless the rate of profit falls to where it becomes negative, i.e., profit stops being made altogether, the amount of profit can always be divided up so that workers get a bigger share. Marx argued this point against views similar to Smith’s within the First International. Marx argued in effect that, if the employer is profiting, that profit can be divided such that workers get a bigger share than they’re currently getting. While Marxists today often stress the importance of the rate of profit falling, if that rate is positive then profit still exists; a drop in the rate of profit is not a loss of money, it’s the continued of money flow to capitalists, just not as fast as it used to flow or could flow in the future.
It’s also not clear that workers always know whether they can win or not – people rarely have perfect information – nor is the future predetermined. Nor is it even clear that workers must always believe they can win in order to strike. As Charles Tilly and Edward Shorter put it in a classic study of strikes in France, “the outcomes of French strikes have been so dismal over the years, and their growing lack of success so palpable, that one wonders why the workers bothered. Indeed, anyone who insisted that the true purpose of strikes was to achieve the stated grievance would absolutely be at a loss to explain why in the 1960s workers still struck at all, so few of their demands were ever met.”
The idea that economic growth is necessary to make union and strike activity possible is especially hard to sustain given the history of the 1930s. 1933 and 1934 were major strike wave years and occurred amid the Great Depression. In those years, about 2.5 million people struck (Table 1 here). The rest of the 30s were a relative highpoint in strike activity as well, and a period of exceptionally intense union growth as well (Figure 8.1 here). The 1920s, on the other hand, were a banner decade for capitalists yet a time of union decline in the United States. Somewhat similarly, as World War II got underway, fostering economic growth, the rate of growth in unionization slowed in the US and while strikes remained frequent for a few years, they didn’t increase as quickly as they had during the Depression.
Given the historical record, it seems straightforwardly mistaken to say that strike increases and union growth require economic growth or high rates of profit. I will note as well that even after the economic crisis of the early 1970s, workers in the US participated in National Labor Relations Board union recognition (“RC”) elections in massive numbers throughout the 1970s, as the historian Lane Windham has shown, and workers and unions’ win-rates in these elections remained high throughout the decade, as NLRB records show. Furthermore, while union density shrunk after its peak in 1945, the absolute number of union members grew in the 30s, 40s, 50s, 60s and the 70s, reaching its historic peak in 1979. This trend of membership growth by decade reversed in the 1980s, with the total number of union members falling each decade since (table 8A.2 here) and the number of union elections filed falling as well. Windham’s book has a good account of why this turn happened, though I would stress that this turn was made possible by longer-term trends related to the effects of labor law on the labor movement. I elaborate on this below.
Waves and tides
In thinking about these matters, I find it helpful to think in terms of metaphors of waves and tides, where a wave is a year with a big jump and a tide is a group of important years. As Smith mentions, the scholars Charles Tilly and Edward Shorter provisionally defined a strike wave year as a year when the number of strikes in a country is fifty percent higher than the country’s average of the previous five years. (This wasn’t their only definition of strike wave, just one that helped them guide their research: definitions are tools to think with and they’re only as good as the thinking they facilitate.) If we use this definition, between 1921 and 1947 there were only two strike waves: 1933 and 1934 (Table 1) and between 1947 and 2020 there were only three: 1967, 2018, and 2019.
In the 1920s, the tide of union and strike activity began to roll out. There were about 2385 work stoppages in 1920, and over a million workers were involved. The numbers of both work stoppages and workers involved fell steadily over the decade until by 1930 there were only 637 stoppages involving 183,000 workers. In 1933 the tide began to roll back in, with almost triple the number of 1930s work stoppages, 1695 of them, and over six times the number of workers on strike as in 1930, with 1.17 million workers involved. The tide stayed high, growing steadily higher, through the late 1940s. There are some ups and downs in some specific years but the five-year average for strikes went up every year from 1933-1947 (Table 1 here). In terms of historical time, the turning of the tide between 1931-1933 was sudden and intense, as shown by the 1933-1934 strike waves.
Smith suggests that the labor movement of the years just after World War II should be taken as a historical baseline for comparison, calling it “the classical labor movement,” a classical era that figures in Smith’s article mostly as a symbol of something now impossible. It’s worth noting that the 1940s and early 1950s were still high points in work stoppages in the US, but they were a different kind of high point than a strike wave: they were years at high tide, and not years of sudden relative increase.
At this point we know that 2021 did not end up being a year of massive increase in strikes (which is not to say that nothing happened). But still, to compare 2021 and the late 1940s seems mistaken to me, since the events of the late 40s were the result of about fifteen years of prior developments. 2022 will certainly not replay 1946, nor will 2023. But it’s still possible that this or next year could be the beginning of another ten to fifteen year process leading to a new high tide. I won’t be holding my breath, but given the history of the 1930s (the growth of the United Mine Workers in the 1890s is another example of explosive and unexpected growth tied to strikes), we can’t rule out in advance the possibility of a future change in the tide of class conflict.
Having said that, there are massive obstacles to such a turning of the tide, in that there has been a host of legislation specifically designed to prevent such a change, as I argue below. We also now have a labor movement unfortunately reshaped in part by that legislation, a reshaping that was already substantially underway by the early 1940s. (Marianne Garneau covers the effects of these changes on the labor movement effectively in a recent essay.) While these are significant obstacles to a turning of the tide, workers could still defeat these obstacles. The future is substantially unwritten and depends significantly on what large numbers of workers do or don’t do.
I assume this won’t convince people who believe, as Smith does, that there has been “an epochal and structural shift in the composition of capital and the allocation of labor” away from manufacturing with its “often enormous workplaces employing thousands of workers.” Certainly things have changed a great deal since the 1940s, but here too looking prior to the 1940s is illuminating. Labor economist Leo Wolman wrote in 1936 that “one of the most significant facts concerning union membership in the United States” was what “unions in manufacturing industries should represent so slight a fraction of total membership.” Unions “in manufacturing have consistently remained the weakest segment” of the US labor movement, he wrote. He added that unionized manufacturing workers had tended to decline as a proportion of all union members in the US, more or less continually from 1897 to 1929 (Wolman, page 89-90). Economist Gerald Friedman has similarly estimated that union density in manufacturing in the US was 7.3% (table 4, here), almost identical to the unionization rate in manufacturing in 2021. Nor does there seem to have been an epochal transformation in the size manufacturing workplaces. In 1929, seventy six percent of manufacturing establishments employed less than 500 people, while in 2019, it was seventy two percent (Quarterly Census of Employment and Wages, table 14). I completely understand why someone might look at the US today where union density in manufacturing is about 7% and where about three quarters of manufacturing employees work in facilities of five hundred people or fewer, and think that unionization isn’t going to increase. But the thing is, very similar conditions obtained in 1929. Further, no one in 1929 could have imagined the tremendous changes that would come over the following decade.
Another reason why the strike wave and tide of the 1930s is worth considering, and why it’s a mistake to only focus on the period after the Second World War as we think about the present, is that strikes in the 1930s brought about increases in unionization. Union density in 1933 was 10.95%, growing to 27.63% by 1939, meaning density in 1939 was 250% what it was 1933. Union density hit its highest point in US history in 1945, at 34.23%. 1945 density was about 125% of density in 1939 (Table 8A.2 here). To put this in perspective, union density in 2021 was 10.3%. If union density were to rise at the growth rate of unions from 1939-1945, by 2026, union density would be 13.4%. If it were to rise at the rate from 1933-1939, by 2026 union density would be 27.2%. (As readers of Organizing Work will likely know, from 1945 to 1955 union density fell in the US, though it fluctuated as it did so, and the larger trend ever since has been that it continues to fall.)
One thing this draws out is the variability in the relationship between strikes and density. The two increased together in the 1930s – I want to underline that point: strikes built unions, or rather, the process of unionization was a process of workers taking collective action – but the two became far less linked after World War Two. That delinking is a direct result of policy changes which go much further in explaining the decline in unionization and in strikes over the mid-to-late twentieth century than does Smith’s appeal to profit rates.
Institutional context and political vision
To understand these matters, I think we have to ask questions that don’t fit well into the wave and tide metaphors, which can have the unfortunate effect of making it seem like unions and class struggle are an easily identifiable substance like water, when the reality is more complex than that.
The language of union density can have a similarly obfuscating effect. Unions are political organizations, having explicit political visions and engaging in daily practices that are themselves deeply political, both in their internal operations and in their relation to employers, and unions differ from each other in their politics. We can see important changes in union politics over time. Talking in terms of wave, tide, density can flatten out these differences and ideologically smuggle in specific political visions of unionism at the expense of other such visions, as I have argued before here at Organizing Work.
Concrete instances of class struggle occur in specific times and places in capitalism, where capitalist social relations are institutionalized in specific ways. Those institutions are always, as the Marxist sociologist Simon Clarke has put it, institutionalized forms of class domination. In his work on the UK, Clarke has examined the “history of class struggle in and against the institutional forms of the capitalist mode of production” (Clarke, page 16). I’ll mention as an aside that Clarke is very good on the conceptual and political limitations of those versions of Marxism that, like Smith seems to, lean heavily on the supposed law of the tendency for profit rates to decline in capitalism over time). As Clarke stresses, actually existing class struggle always takes place in specific institutional contexts. Those contexts shape class struggle in its specific traits, and struggle in turn causes institutions to be reorganized. Marx’s discussion of conflict over and change in the laws regulating the length of the working day in 19th century England are illuminating here as well.
With the above in mind, I want to suggest that we can see three general phases in the history of the labor movement in the US, with each phase involving a somewhat different approach to how to institutionalize class domination and a variety of visions and practices in the labor movement in relation to those institutions. Early on, the labor movement fought for the right to govern society. That right was contested by some of the official powers-that-be. A wide range of institutions existed for, on the one hand, either facilitating or preventing unions from exercising governing authority over some aspects of social life and, on the other hand, for managing disputes between employers and unions. These institutions included, on the friendlier end, arbitration and mediation bodies and forms of negotiation or collective bargaining, and, on the more brutal end, courts, police, and private security forces. These institutions were unstable and relatively decentralized. Furthermore, the labor movement itself was internally tremendously contentious, including a very wide range of visions of how workers should govern, and why, and how to achieve that goal, and how similar or different society would become under their governance.
In response to both the strike waves of 1933 and 1934 and the global economic depression, the US government overhauled these institutions, imposing greater uniformity and centralization under the federal government. The National Labor Relations Act was a key piece of this overhaul. I won’t go into the details but the NLRA, often called the Wagner Act after its sponsor Senator Robert Wagner, was influenced by the National Industrial Recovery Act of 1933. The Wagner Act was also heavily informed by a series of prior experiments in industrial relations in the 1910s in Wagner’s home state of New York, as demonstrated by the historian Richard Greenwald. Those experiments were a direct response to massive strikes in the garment industry in 1909 and 1910, and were a training ground in industrial relations for Frances Perkins, who later became Secretary of Labor in 1933. The NLRA in effect took the lessons in conflict management that Wagner and Perkins and other state personnel learned in the 1910s in New York and made them into national policy.
The period from the passage of NIRA in 1933 to the passage of the NLRA in 1935 to the Supreme Court upholding the NLRA as constitutional in 1937, is an approximate marker for the opening of a second phase, where the labor movement’s right to govern was partially recognized by official institutions. Furthermore, that partial recognition has been institutionalized in a way that has encouraged the labor movement to lose much of its commitment to governing. As Simon Clarke has argued, the development of industrial relations was an especially important mechanism “through which the capitalist state sought to confine the aspirations of the working class within the limits of its liberal form, and in and against which the class struggle has subsequently developed” (Clarke, page 19).
The NLRA marked a greater consensus in the state about the importance of industrial relations as a means to bring about a similar confinement of aspirations in the US. This confinement came in part by constructing what Clarke called “institutionalized forms of class collaboration” (Clarke, page 20). It’s important to note that class collaboration can still be quite conflict-prone, and some of the time conflict within a narrow scope is a way to organize and reproduce domination.
After the NLRA, the labor movement was gradually accepted to a more widespread degree but as a junior partner. Over time, the movement’s officials increasingly accepted that junior status, at least implicitly, as part of accepting the new institutional mechanisms of dispute resolution. Furthermore, and more importantly, those mechanisms compelled unions to act specifically as junior partners regardless of the political views of union leaders. (By the way: what do you call a socialist employer? Boss. What do you call a socialist elected to the Senate? Senator.)
I want to stress that these developments were conflict-prone throughout. They were repeatedly contested in a series of conflicts over the details of this emerging new institutional order, as well as conflicts within the labor movement about how best to proceed. Furthermore, none of these changes were a matter of whether or not conflicts would happen. Rather, each change was a shift in how conflicts would happen and what the stakes of conflict would be. But to make a long story short, over time, partly in response to its new social environment and partly due to how conflicts inside the labor movement played out, the labor movement became more narrow in political vision and in practice. The mid-1940s marked the transition from the second to the third phase: the labor movement, a newly acceptable junior partner, began its long slow decline: the tide began going back out. Lane Windham’s book gives a good account of how and why that change accelerated in the 1970s and 1980s.
“To diminish the causes of labor disputes”
Forcing the tide of workplace struggle to go back out and minimizing strike waves was the whole point of the institutional reorganization wrought in the 1930s, as the historian Christopher Tomlins has demonstrated in his book The State and the Unions. As Tomlins has written more recently, “the labor relations system that prevails in the United States is not a consequence of law’s failure but, rather, the intended outcome.” Joe Burns has demonstrated in his book Reviving the Strike that this tide cannot be reversed while remaining within that system: unions have to violate existing labor law if the tide is going to turn again for the better. (I’ll note that Burns has a new book coming out, reviewed here at Organizing Work. I’ve not read the book yet, but I look forward to doing so.)
The point of the National Labor Relations Act is expressed in its full original name: “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.” The Findings and Policies section that opens the Act states the problem the Act was written to solve, namely that employers’ refusal “to accept the procedure of collective bargaining led to strikes and other forms of industrial strife,” which in turn obstructed commerce. The point of the Act, then, was to aid commerce by preventing strikes.
Marianne Garneau covers the effects of these changes effectively in another essay in the Brooklyn Rail expressing further doubts about the claims that we’re in a strike wave. As she writes, strikes “have been severely hemmed in over the course of decades.” She adds the important point that striking often, well, sucks. Workers’ income drops and while there’s a lot to be won, there’s a lot to be potentially lost as well. As Garneau demonstrates, strikes have loomed larger in the left’s imagination of unions as the law has stripped unions of past effective tactics, including by hemming in strikes. Thus, she argues, emphasis on strikes this fall reflects an overly limited political vision of the labor movement, its possible futures, and best courses of action in the present. (Garneau’s article also includes a good discussion of the mistakes involved in overemphasizing macro-level economic factors in explaining workers’ collective behavior; in my view Smith’s article is an example of the kind of overemphasis Garneau criticizes.)
Here too the tide and wave metaphor has big limitations: strikes change over time partly in response to changes in the law regarding who can strike over what and in what ways, so that a strike in 1934, 1944, and 1954 aren’t necessarily the same set of practices. There isn’t a single, stable referent to “strike” across times and places. Indeed, much of what Joe Burns means by “revive” in Reviving the Strike is not only that unions need to strike more often, but that they need to strike differently. Furthermore, the loss of other tactics and narrowing of vision means that the strike has taken on a new meaning as other options have become less available, and depending on whether the point of unionism is understood as governing society vs. enforcing a limited and relatively fixed set of citizenship rights set by other people.
Less abstractly, consider the relationship between recognition strikes and union elections. From 1934-1939, there were an average of 980 recognition strikes per year, with over 2.3 million workers involved in a recognition strike in that time. From 1940-1945, the average was 690 recognition strikes per year and a million fewer workers were involved in a recognition strike. (Table 8.6 here.) Bear in mind that strikes were still increasing in the 1940s, yet recognition strikes were shrinking in absolute terms and as a proportion of strikes. This means that what kind of strike workers were engaged in – what the dominant form was for the strike as a social phenomenon – was changing.
As recognition strikes declined, recognition elections increased. There were 163 union elections in 1936 and 74,000 workers who voted in favor of unionization in those elections, while in 1945 there were 5253 elections and 773,000 workers voting for unionization in those elections, down from over a million every year from 1941-1944 (Table 8.6). That is to say, the recognition election gradually replaced the recognition strike. Bringing about that relative replacement was a success in terms of the overriding goal behind the National Labor Relations Act, namely to reduce strikes. That relative replacement was a direct result of actions by the National Labor Relations Board and courts’ interpretations of the NLRA. Indeed, the NLRB congratulated itself for this result in a 2015 document celebrating the 80th anniversary of the NLRA’s passage: “the secret ballot took the place of the recognition strike” (page 25 here). In a similar spirit, the Labor-Management Reporting and Disclosure Act of 1959, made it an unfair labor practice for workers to picket an employer for recognition for more than thirty days without filing a certification election.
The NLRA encouraged the replacement of recognition via the collective action of a strike with recognition via the less collective, more state-mediated, and more employer-friendly process of the certification election. This should give us pause, since as Matt Dimick has written, collective action is “the means by which workers build solidarity and develop class consciousness.” As such, Dimick argues, replacing collective action with “other means of reaching working-class objectives may, whether intentionally or not, undermine working-class interests.” By limiting some activities and offering (modest) protections to others, labor law has encouraged workers and unions to pursue other means than collective action for recognition and negotiation, as well as signaling to employers that they could fight non-election recognition efforts more aggressively. (Tomlins, The State and the Unions, 267-270, 292.) I don’t mean to romanticize striking, but I do want to point out that these changes in the law encouraged changes in what unions were and what they did as social actors. These were not simply changes in unions’ external environment, but changes that remade what unions tended to do and be. I want to again encourage Organizing Work readers to read Charles Romney’s Rights Delayed, emphasizing how the NLRA and NLRB influenced unions for the worse in the canning industry.
In addition to bringing about the relative replacement of recognition strikes with certification elections, the NLRA changed the meaning of recognition. As Christopher Tomlins has shown, prior to the NLRA, when an employer recognized a union, that entailed conceding some of its demands – in important respects, recognition was a matter of employers concluding negotiations by conceding to workers’ demands – while under the NLRA recognition meant agreeing to begin negotiations. And the negotiations in question were, by legislative design, less prone to “burdening or obstructing” commerce. The NLRB also reduced the scope of what unions could bargain over by reducing the ability of unions to demand that only union members be hired. (This change is often attributed to the Taft-Hartley Act of 1947 but this was in fact a prior board decision that Taft-Hartley codified, and to be fair, amplified (here too see Tomlins). Generally speaking Taft-Hartley is overemphasized by pro-labor liberals and social democrats, an overemphasis that promotes myths about the National Labor Relations Act as benevolent legislation.) These changes in the law created a social environment that was far more amenable to some approaches to unionization as a practice and some labor movement political visions than others, and that was the legislators’ intent in making those changes. As Charles Romney detailed in his excellent book, the National Labor Relations Act and National Labor Relations Board fostered the success of some kinds of unions over others and in doing so helped pull the labor movement further to the right in a variety of ways.
Rather than understand capital accumulation as a condition for union success – such that a supposed epochal transformation of capitalism rules out the labor movement successfully reinventing itself – we should instead see unions as political and as having a complex relationship to capital accumulation depending on their politics and practices. Some unions’ visions have sought to promote accumulation and get a share, others have sought to climb more fully into the driver’s seat of how capitalist enterprises operate, while still others have sought to end capitalism altogether. And some have advocated one of those visions which practically pursuing another. Of course, most are a grab bag of all of these elements and more. Furthermore, state institutions like labor law are a significant influence on these visions and practices. These institutions shape the playing field in ways that advantage some versions of unionism over others. In general, the approaches to unionization that are most encouraged by prevailing institutions are the approaches with the least possibility of turning the tide of workers’ struggle.
Solidarity and the spirit of organization
The labor movement’s decline is not primarily due to trends in capital accumulation and profit rates but rather to the institutional transformations that have hemmed in the strike and other labor movement tactics. By doing so, those institutions have exerted an important influence on the labor movement’s overall vision and which forces in the movement predominate. As Marx put it “unions work well as centers of resistance against the encroachments of capital. They fail partially from an injudicious use of their power. They fail generally from limiting themselves to a guerilla war against the effects of the existing system, instead of simultaneously trying to change it, instead of using their organized forces as a lever for the final emancipation of the working class, that is to say the ultimate abolition of the wages system.” Developments like the National Labor Relations Act foster that self-limitation and “injudicious” use of collective power. Fully becoming a lever for emancipation will require unions to break out of those institutional limitations, a nut that has not yet been cracked.
The ability of the working class to eventually crack that nut is not so much dependent on trends in profit rates as Smith seems to think. The issues instead are a matter of what Anton Pannekeok called working class virtue, meaning the qualities cultivated through organizing. As Pannekoek put it, “united action of the whole class is not possible, if it is not sustained by a strong moral force.” As such, a key element of workers power, he wrote, is “solidarity, the spirit of unity, organization. Solidarity is the bond that unites the will of all the separate individuals into one common will, thus achieving one mighty organized action.” I suggested above that we see labor history in three phases, the third being the slow decline – now very advanced – of the labor movement as junior partner within a model of economic citizenship. That decline is not only quantitative, a matter of density alone, but also qualitative, a matter of political narrowing. This is part of why starting from the late 1940s as a reference point is mistaken. That starting point came after a substantial narrowing of political vision and practice had already occurred in the labor movement, in relation to the changed institutional structure wrought by the policy changes of the 1930s.
In my view, as we think and struggle through our present and future, it’s a good thing to draw on the past to do some of that thinking. Within the phases I suggested, I think the value of the second and third phases serve primarily critical purposes and offer us warnings. They help us to think about the ways in which institutionalized forms of class domination can become very malleable in the face of intense class struggle and so become subject to re-organization. One effect of reorganizing those institutions of domination is to defuse struggle and shape the vision and practices of present and future struggles. There are some other critical lessons and warnings in the first phase, in the growth of the United Mine Workers of America and later the CIO, as object lessons in the growth of conservative unionism (even if sometimes militant), for instance. To the limited degree there are positive lessons in the past, the sequence of events leading to the foundation of the Western Federation of Miners and then the IWW remain fruitful. One important point to bear in mind within each phase and even more so across each phase is that unions as a social practice and a political vision change over time and to an important extent differ at any given moment. As such, being simply pro-union and pro-strike has little more concrete political content to it than being simply pro-political party, pro-voting, pro-riot, etc.
For the time being, the slow decline seems likely to continue to further advance in all of its often genuinely lethal brutality, until workers organize to break out of the current institutional frameworks limiting the labor movement and “confin[ing] the aspirations of the working class” within the limits of capitalist society, to borrow Simon Clarke’s words again. In my view the working class’s break out is not likely to come from efforts like the PRO Act (a current favorite among people the Internationale referred to as condescending saviors), nor from macro-level economic changes, nor from social movement protests in the streets – however righteous and important in their own right. Instead such a change will come from lots of small-scale organizing of the kind this site seeks to highlight and enrich, until that organizing reaches a larger scale. “The habit of solidarity,” Eric Hobsbawm wrote, “takes time to learn.” This kind of organizing is activity in which workers act collectively against the immediate constraints placed on them in their workplaces and in the rest of their lives, and in doing so cultivate – specifically in a practical sense – what Pannekoek called “the spirit of unity, organization.”
Nate Holdren lives in Iowa. He wrote the book Injury Impoverished. He writes occasionally for Organizing Work and Legal Form and keeps an infrequently updated blog.