THE LIBERAL AND RADICAL DEFENSES OF LABOR UNIONS

Are labor unions compatible with liberalism? While many contemporary liberals working in the tradition of egalitarian liberalism would answer yes, Mark Reiff (Reference Reiff2020) shows that unions are compatible with almost all strains of liberalism, ranging from Nozickean libertarianism to Rawlsian egalitarianism. I take Reiff as exemplary of a more general, liberal model of thinking about labor unions, one that tracks many of the ways unions are institutionalized in European and American law: namely, as manifestations of free association and an effort to balance unequal bargaining power (see also Lindblom Reference Lindblom2019). Reiff’s central argument is that labor unions are necessary given the existence of firms, which means that workers must enter an employment relationship rather than act as independent contractors. Given the existence of such firms, workers will naturally have weaker bargaining power in specifying their employment contracts. Unions are thus forms of voluntary associations that correct for the power imbalance between workers and firms. Unions both equalize the bargaining power of workers in the context of wages and protect workers’ rights in the context of underspecified labor contracts. In both cases, the justification for unions rests on the failure of capitalist labor markets to live up to their purported ideal of equal exchange between workers and capitalists. Market failures are then compensated for by creating strategic assurance mechanisms—in this case, labor unions—that can overcome those failures through collective negotiation and monitoring. Such mechanisms, Reiff argues, will give workers a voice that prevents them from being dominated in the republican sense. Without a union, Reiff writes, workers “lack an effective voice” and so “are effectively disempowered and subject to the kinds of abuse and even domination we do not expect free men and women to have to put up with” (Reiff Reference Reiff2020, 120).

There are two limitations with Reiff’s liberal approach. First, Reiff’s view focuses on the labor contract as the object of union activity. By framing unions as a response to the failure of markets to live up to their own ideals, the liberal view accepts the image of work relations as basically voluntary contractual relationships and delimits the scope of unions accordingly. The problem, here, is to ensure that the workplace lives up to its ideal of free and equal exchange and that workers’ rights are protected. The liberal thus accepts the rationale for the capitalist firm as a mechanism to overcome another market failure—transaction costs—and so the need for a hierarchical structure of control over production and investment within the firm. Unions exist to counterbalance the capitalist firm, not transform it.

The second, related limitation is the focus on republican freedom. Without getting into the entire debate about republican freedom, I want to note that republican debates about the workplace have largely focused on managerial discretion and the power of managers over workers (Frye Reference Frye2024; Gourevitch Reference Gourevitch2014).Footnote ² The point of a capitalist labor relationship is to create an open-ended power of command over workers, given the impossibility of using contracts to specify each discrete task of workers, and this leaves workers vulnerable to the unchecked powers of their managers. As many have noted, this perspective has a harder time accommodating the structural power of capitalists over workers, one based in different relationships of ownership vis-à-vis the means of production. But I think this problem is tied to a larger issue: by restricting freedom to a problem of avoiding power-over, republican freedom loses sight of the much broader question of the positive power-to that workers realize through the collective organization. To be sure, in the history of the labor movement, workers mobilized republican concerns about the power employers exercised over their workers. But they also developed a more expansive vision of freedom, one that focuses on egalitarian cooperation in the workplace and providing workers with the substantive resources for self-development, both at work and in their leisure time. As a result, republican freedom largely sidesteps the larger question of how we should democratically organize productive activity in society, and the role that cooperation versus strategic, exit-based mechanisms should play therein (Klein Reference Klein2017). Thus, it is better to view unions as contributing to a demanding vision of relational equality, one that entails democratic decision-making over productive activity and investment, although there are several points where neo-republicanism and relational egalitarianism substantively converge.Footnote ³

A second prominent perspective on labor unions I call the radical perspective. It points to the tension between liberal principles, and especially notions of property rights, and the actual practice of unions, which often have been born of wildcat strikes and other aggressive rank-and-file tactics. In particular, the radical view starts from strikes rather than collective bargaining. The radical view emphasizes the fact that strikes, to be effective, require the exercise of coercion and the infringement of certain liberal rights, such as the right of employers to hire replacement workers. This is reflected, for example, in the fact that unions often demand that employers not hire replacement workers. In such an instance, workers go beyond just withdrawing their labor to dictating how firm owners dispose of their property. Strikes also impose costs on innocent bystanders and have, some argue, a potentially extortionate logic (Christensen Reference Christensen2026; Dobos Reference Dobos2022; Stehr Reference Stehr2025).

Alex Gourevitch (Reference Gourevitch2018) argues that such coercive means are justified as a method of resisting oppression. Against the liberal view, he challenges the overall legitimacy of the firm as a site of economic production. Actual capitalist societies, characterized by class divisions grounded in the right to private property, are systems of oppression. Gourevitch focuses on the disruptive, often extra-legal tactics that workers must use to succeed in strikes, especially in those segments of the economy where strikes are most necessary.

To be sure, Gourevitch does not set out to provide a general theory of labor unions. Yet there seems to be at least an implicit image behind his account—one of rank-and-file labor radicals organizing direct strike action, challenging captured, reformist union leadership that has become dependent on the bureaucracy of grievances and collective bargaining. Yet in emphasizing the disruptive, extra-legal aspects of union activities such as strikes, Gourevitch’s view risks giving away too much to the status quo. His view presupposes that a strict understanding of private property rights is a necessary component of capitalist social relations. From this perspective, a strike always involves an illegitimate restriction on the rights of capitalist property owners to hire new workers to replace old workers.

Yet it is also possible to see a strike as a clash between two different normative orders—one, a private law paradigm that understands the work relationship through the lens of voluntary contract, and the other, an everyday moral economy that contains notions like a right to employment and a fair wage. Legal restrictions on capitalists hiring replacement workers embody a recognition that the firm is socially embedded in a broader set of relationships, such that strike breaking is not a legitimate reason for an employer to alter an employment relationship. Put differently, capitalist property relations are not being broken through an extra-legal act of resistance to oppression. Rather, the normative meaning of property itself is being contested. Striking workers are contesting the claim of capitalists to exclusive control over the productive resources legally codified in the firm, insisting that workers themselves have a just claim to a negotiated portion of profit. Firing workers to break a strike, from this perspective, is an illegitimate claim to unilateral control over the firm’s property.

In many cases, the argument I develop below overlaps with the radical view (and indeed, also covers aspects of the liberal view). But there are some core differences with important implications. As I will discuss in the next section, the radical view presents the capitalist firm as inevitably structured by class domination and so argues that unions and strikes are justified to the extent that they point to a fundamental rupture with capitalist economic institutions. This places our analysis of labor unions in a bit of a straitjacket, insofar as it leads to the conclusion that unions, and union activities, are only legitimate to the extent to which they are anti-capitalist. From the democratic perspective, there are already democratic normative principles embedded within capitalist social relationships, even if they stand in a tense and sometimes contradictory relationship with other values. Labor unions then engage in contestation over the appropriate structure of economic production by drawing on resources already present, even if implicitly, within capitalist societies.

While this may seem like a small distinction, it has concrete implications for both union strategy and our normative analysis of unions—for example, it enables a broader discussion of the relationship between different union structures and democracy without boiling it all down to the question of whether unions are anti-capitalist. Ultimately, the distinction between the radical and the democratic view may hinge on our understanding of capitalism and structural domination under capitalism. If one accepts that capitalism is an order of structural domination and the firm is inherently a vehicle of class power, then many of the activities I point to as valuable elements of unions will appear as so many devices to seduce workers into accepting their own subordination by softening the most overt manifestations of structural class power. This debate can hardly be settled here. At the very least, my democratic view may bring to light aspects of union activity—and normative dimensions of both work and workplace democracy—that are obscured if we begin only from class domination.

LABOR UNIONS AND THE CAPITALIST FIRM

Existing views of labor unions, then, emphasize either their liberal function in ensuring equality of bargaining power in specifying labor contracts or their radical function in disrupting capitalist relationships through direct action such as strikes. In the background for their analysis of unions stand two contrasting visions of capitalism, and of the relationship between capitalism and democracy. For the liberal vision, firms are economically efficient responses to transaction costs. According to the famous Coasian theory of the firm, firms exist to overcome the transaction costs involved in negotiating contractual exchange (Coase Reference Coase1937). Firms thus operate through hierarchy and command, rather than exchange and contract, up to the point that the marginal costs of further expansion in size match the marginal benefits of reducing transaction costs. Thus, the existence of firms is welfare and efficiency-enhancing to the point that they successfully minimize transaction costs. Unions can then be justified on two grounds: first, as already seen, liberal principles of free association and nondomination insofar as unions balance the inherent power of otherwise necessary firms. And second, they can be justified on efficiency grounds, insofar as unions provide workers with an assurance mechanism to enable them to invest in firm-specific skills (Hall and Soskice Reference Hall and Soskice2001).

For the radical view, capitalism is an order of illegitimate class domination. The firm is simply the vehicle through which capitalists as a class extract surplus value from workers. The wage labor relationship is inherently exploitative. But this makes the radical attitude toward unions ambivalent. On the one hand, unions are useful vehicles to resist capitalist exploitation. On the other hand, though, unions often fight to get their workers a larger piece of the capitalist pie. For this reason, the radical view emphasizes strikes over other union activity. Only during a strike do workers potentially break with the logic of wage labor, delegitimizing capitalist wage relationships in general.

In contrast to both the liberal and radical views, my democratic interpretation of labor unions begins from a specific analysis of capitalism as a political order. It sees the firm first and foremost as a legal and political entity designed as one possible mode of resolving coordination problems. However, as Sanjukta Paul (Reference Paul2023) argues, once constituted by legal rules, firms become political entities pursuing legal and political power to protect their market coordination rights, as much as economic entities competing with other firms to maximize efficiency.

But while a capitalist economy is distinctive in that core productive activities are organized through profit-seeking firms, the legal and political constitution of firms means that the normative logic of the market retains a connection, however tenuous, to other sources of legitimacy. The “exception” to the principle of market competition—as well as the typically extensive flow of various government subsidies, tax breaks, and other concessions that flow to larger firms—places firms under pressure to appeal to normative legitimacy narratives other than pure profit-maximization, narratives that can become operationally embedded within their hierarchical systems. Or, at the very least, it means trade unions can appeal to competing, existing legitimacy narratives to a firm’s orientation toward profit maximization.

From this political perspective, then, we should view firms as one legally constituted way of allocating power over productive activities. Even as firms compete with each other, the extent and nature of such competition is constituted by—and often curtailed by—the legal allocation of coordination rights to firms. Unlike the liberal view, the political analysis of the firm does not view capitalism as a market order that sometimes falls short of realizing its ideal of free exchange. Rather, the political interpretation approaches capitalism as a specific mode of organizing power over economic activities. Firms are vested with control over decisions about production and investment and, therefore, the power to direct labor. The idea of the “economy” as a distinctive domain, one constituted through market exchanges, is what disguises the structure of power within capitalism. This appearance, as well as the idea of distinctive labor rights, masks the reality of hierarchical structures of production. This picture is further reinforced by institutional mechanisms that instead represent unions as mechanisms for wage bargaining and overcoming market failures. However, I argue that we should regard these as efforts to tame the democratizing effects and character of trade unions, rather than the purpose of trade unions.

To this extent, then, the radical and the democratic view converge. But while the radical view looks to unions primarily as sites of resistance to oppression, therefore highlighting the disruptive features of strike actions, the democratic view contends that the full range of union activities can constitute a partial democratization of production and investment decisions. To do so, unions can mobilize precisely the normative principles and legitimation narratives that compete with profit-maximization in justifying capitalist social relations. In particular, unions mobilize a democratic counterlogic. It can be the case that, given the relative power of multinational corporations, profit-maximization will be the dominant organizing principle of capitalist society and that such societies still retain latent pockets of alternative normative principles, however weak they may be. But what are those normative principles?

THE DEMOCRATIC VISION OF LABOR UNIONS

Let us now turn to those normative principles. In the following sections, I develop the three normative pillars of the democratic view of labor unions, ascending from the level of the individual employment relationship to the level of the firm and then finally the economy as a whole. At the level of the employment relationship, the democratic view advances a normative vision of employment as a valued social role rather than a market-based contractual arrangement. Unions, then, are normatively necessary to ensure that workers both gain social recognition for their work and have agency in the construction of the social meaning of their jobs. At the level of the firm, unions advance a vision of the workplace as a cooperative venture rather than a hierarchical institution defined by shareholders’ property rights. Then, finally, at the level of the economy as a whole, the democratic vision of labor unions supports a model of federated industrial democracy, where firm-level cooperation is coordinated at the level of sectors by union federations. Unlike existing models of workplace democracy, this view adapts democratic principles to the distinctive functional demands of economic production, rather than importing a model of democracy designed for the political realm.

I see each of these principles as ultimately grounded in a conception of democratic equality. However, my goal is to develop these principles in a way that is compatible with a relatively broad range of views of what such democratic equality entails. Notably, each of these dimensions of the democratic vision of labor unions is largely orthogonal to the question of ultimate ownership of productive resources. That is, there are many understandings of socialism, such as market socialism, that would leave no role for unions and discount the importance, for instance, of employment as a valued social role even as they called for the public ownership of productive resources (Lange and Taylor Reference Lange, Taylor and Lippincott1938; Roemer Reference Roemer1994). So, on this view, public or collective ownership is neither necessary nor sufficient, although as an empirical matter, it is more likely these principles would be realized under a system of public or collective ownership. At the same time, it is possible, in principle, to render each of these values compatible with some role for private investment in the distribution of capital, although such a role would need to be much more constrained than at present.

EMPLOYMENT AS A VALUED SOCIAL ROLE

While there is no evidence of direct institutional continuity, many early unions carried forth a cluster of ideals that emerged from the European guild movement (Chase Reference Chase2000). One of the core aspects of the guild ideal was a notion of the dignity of the craftsman (typically, but not always, a man) whose worth was grounded in a notion of skill and service to the community and not just the market value of the goods they produced (Black Reference Black2013). Thus, one normative principle that unions advance is a notion of employment as a valued social role and not just a market relationship. One aspect of this is the idea that the value of a job should be a matter of public esteem and recognition, and that remuneration should track these notions of public esteem, and a union’s ability to mobilize both the solidarity of their members and public opinion in struggles about pay and working conditions. The normative principle of employment as a valued social role also finds expression in principles of seniority, which again divorce remuneration from performance as defined by managers and instead provide workers themselves with a real form of ownership over their job. Of course, in highly unequal societies, the demand to view work as a valued social role tends to be hotly contested. Market ideologies present workers as inherently lazy without hierarchy and discipline, and so capitalists and employers can reframe such demands as efforts to escape market discipline for a privileged elite of workers. However, even in a more democratic society, we can imagine different ways of organizing employment and work, with some market socialism schemes requiring highly flexible labor markets alongside collective ownership of productive resources (Lange and Taylor Reference Lange, Taylor and Lippincott1938; Roemer Reference Roemer1994).

Against these views, the normative ideal of unions implies that, even in a just society, dignity requires a system of public esteem of employment that views a job as part of a valued social role and not just a matter of maximizing efficiency. Dignified work, from the perspective of this union norm, has two components: the first is the nature and quality of the work itself. Unions fight for improved working conditions, resist “speed-ups” and other efforts to intensify the work process, and improve the health and safety of the workplace. The second, though, is the idea that workers, through their collective self-organization, should be able to make and have recognized their own claims about the social value of their work. These two dimensions go together: the dignity of work itself is constitutively tied to the ability of workers to exercise control over the work process and gain recognition through their claims to public esteem. In all, then, the union ideal of employment as a valued social role contains the concerns that philosophers have recently identified as important elements of good work or dignified work (Clark Reference Clark2017; Cholbi Reference Cholbi, Zalta and Nodelman2023; Tyssedal Reference Tyssedal2023). But it extends beyond them in emphasizing the co-constitution between the work process itself, workers’ ability to shape the work process, and workers’ ability to gain social esteem for their work.

THE WORKPLACE AS A COOPERATIVE VENTURE

Closely connected to an ideal of employment as a valued social role, one which requires public recognition and esteem, is that of the workplace as itself a cooperative endeavour. While of course one option is to organize the workplace directly as a workplace cooperative, unions have nonetheless advanced this vision of the workplace even within the context of the capitalist firm. Indeed, as Karl Marx observed, part of the production of surplus value under capitalism was precisely the way in which capitalist firms relied on the cooperation of labor. Marx (Reference Marx1990) writes,

Marx’s point is that capitalists benefit from the increased productivity arising from the cooperation of workers, even as such cooperation does not change either the cost of capital or the cost of labor. But because that cooperation is produced by the transformation of the productive process undertaken by capital, the value appears as a product of capital itself.

Labor unions’ normative claims insist on recognizing the cooperation of workers as an independently valuable aspect of the work process. One of the clearest manifestations of this has been the formation of shop steward systems, whereby unions asserted a direct role in the ongoing direction of labor within a workplace. At the peak of their power, for example, shop stewards in American workplaces could stop the work process at will if there were safety concerns and generally played an intimate role in directing the production process. As Nelson Lichtenstein (Reference Lichtenstein2013) writes, “at the shop-floor level, day-to-day conflict over production standards and workplace discipline permeated the structure of work and authority in most factories and mills. The early union contracts were sketchy and ambiguous, and their meaning was worked out in battles large and small” (120). Shop stewards would call for spontaneous work stoppages if there was a dispute over working rates or safety. These forms of “direct shop-floor activity legitimized the union’s presence for thousands of previously hesitant workers, and such job actions established a pattern of union influence and authority unrecognized in the early, sketchily written contracts” (Lichtenstein Reference Lichtenstein2013, 120). Unions, here, drew their legitimacy not just from their ability to raise wages and protect workers’ rights. They also directly challenged the authoritarian control of managers over the production process, thus prefiguring a democratic alternative to capitalist production. As Lichtenstein argues, unions here forged a new vision of industrial citizenship. Workers and union activists sought to transform the capitalist firm such that workers had at least a say, if not actual discretionary control, over the production process.

Through their union activism, then, workers contested the right of managers to direct the labor process and instead insisted that they were best situated to run production. They sought to replace the unconscious cooperation of labor produced by capitalist investment with a conscious, self-directed cooperation in the workplace. As already noted, this union ideal of cooperation in the workplace is separate from the specific ownership structure of the firm. While it may be that in a just society there are reasons to privilege worker-owned cooperatives, there may be reasons to choose alternative ownership structures that give everyone, workers and nonworkers, shares in companies. Under such circumstances, unions will still have a vital role to play in enabling workers to view the work process as a cooperative enterprise that they have the capacity to shape and help direct.

FEDERALIST INDUSTRIAL DEMOCRACY

Related to the above, unions demand a form of democratization of the firm. Yet the union vision of democracy is distinctive from recent calls for workplace democracy (Ferreras Reference Ferreras2017; Frega, Herzog, and Neuhäuser Reference Frega, Herzog and Neuhäuser2019). Many of these recent ideals of workplace democracy translate constitutional ideals drawn directly from liberal political democracy to articulate democracy in the workplace, such as introducing bicameral corporate governance with elected officials, or some model of direct democracy within the workplace.Footnote ⁴ In their ideal form, unions instead advanced a vision of corporatist democracy that combines firm-level representation with a federalized system of industry-wide democratic coordination. First, unions functioned as representatives of the interests and demands of workers at the firm level, partially democratizing the firm by ensuring workers have an ongoing voice. But especially in the tradition of industrial unionism, unions would then organize and represent workers across firms and occupations. This then enabled informal coordination of negotiations, as when unions sought out benchmark contracts with key strategic forms. Or else with formal corporatist arrangements, unions would be an explicit part of the democratic coordination of economic policy, helping coordinate wages with the elected government’s macroeconomic policies.

Unions implicitly advance an ideal of democracy specifically tailored to the distinctive structure of economic production and coordination, which requires a different approach from democracy in the political sphere. Because of functional differentiation within the economy, it makes sense to apply a federalized structure based on the location of different activities within the productive process. Decisions made at the firm level need to be coordinated at the regional and national levels. If investment is incentivized through ownership and profit, then this system creates a conflict dynamic between workers and capitalists that, at least in theory, reconciles the efficiency-enhancing benefits of private investment with a principle of political equality. In practice, this is an unstable equilibrium, as owners of private capital will always have powerful incentives to maximize profit by weakening workers’ agency and power rather than through productivity-enhancing investments. Yet even with an alternative method of organizing investment, such as through public ownership or market socialism, there will be conflicts of interest between workers in different sectors, between workers and consumers, and between workers and the government. In such cases, a system of unionized workers’ representation will remain necessary for democratically resolving such conflicts in a way that respects political equality.

The union vision of democracy very much emphasizes the intrinsic value of equal standing in decision-making about economic production. Yet there are also epistemic benefits to union democracy. In the absence of market mechanisms, managers of firms face a variety of epistemic challenges both in running the production process and in understanding the interests of workers. To sustain collective action, union leaders develop dense networks among their members that enable them to understand their interests, complaints, and concerns. Unions can thus enable firms to make more epistemically informed decisions about remuneration, safety, and the best way to structure work processes. Union collective deliberation can also better inform workers about their interests and the larger economic constraints that inform their choices and decisions.

THE DEMOCRATIC BENEFITS OF UNIONS: RECIPROCITY AND EGALITARIAN SPILLOVERS

Thus far, we have examined the distinctive normative values that follow from the democratic view of labor unions: employment as a valued social role, the firm as a cooperative venture, and the economy as federated industrial democracy. Each of these ideas presupposes the other—that is, the idea of employment as a valued social role entails a view of the workplace as a cooperative venture, which in turn entails the idea of federated industrial democracy. So far, my argument has focused on what the democratic view of labor unions implies about the organization of productive activity. But part of the significance of labor unions is their contribution to democratic values and institutions more broadly. Here, I turn to these external (in the sense of outside the domain of the economy) benefits of labor unions. Labor unions are distinctive institutions in that they bridge democracy in the economy with democracy in the political realm. Their pursuit of democracy in the economic realm can have broader democratic benefits. Here, I discuss two such benefits: first, unions rely on attitudes of reciprocity and solidarity that we would want to foster in a democratic society; and second, they can enable “egalitarian spillovers,” insofar as unions will provide material anchors for the pursuit of democratic institutions. But labor unions also, importantly, have limitations as vehicles of democracy, which I identify as the problems of exclusion and co-optation. Unions will not solve all of democracy’s current ills, but my argument can help us identify more precisely the conditions under which they can foster democracy more broadly.

In the first place, unions both embody and prefigure attitudes of reciprocity that are foundational to the democratic value of relating as equals. While the power of corporations comes through state-sanctioned ownership rights, the power of unions comes, in the first place, through solidarity and collective action, and only secondarily through state-enforced labor law, which typically exists to protect unions’ capacities for collective action. While property rights give owners or shareholders automatic authoritative control over a company, either directly or via an elected board, both union certification and the exercise of workers’ power via strikes require more demanding forms of reciprocal solidarity among workers. This also means, as Claus Offe and Helmut Wiesenthal (Reference Offe and Wiesenthal1980) argued in their classic article, unions must be structured via a more open-ended dialogical practice where members balance multiple values against each other, whereas firms can engage in monological action.

All of this means that unions structurally require that their members take up attitudes of cooperation toward each other—attitudes that prefigure the sort of widespread forms of reciprocity and mutual recognition that would be characteristic of a fully democratic society. Most prominently, G. A. Cohen (Reference Cohen2009) has argued for the centrality of an ideal of community to socialist justice. Community exists where “people care about, and, where necessary and possible, care for, one another, and, too, care that they care about one another” (20). For Cohen, community requires an attitude of reciprocity that contradicts the attitude individuals take in the market: “reciprocity is the antimarket principle according to which I serve you not because of what I can get in return by doing so but because you need or want my service, and you, for the same reason, serve me” (20). Labor unions both require and cultivate such an attitude of reciprocity among their members. Costly union actions, and most centrally strikes, entail that workers act based on concern for the interests and needs of other workers. Unions prefigure a solidaristic culture within a market society, realizing a democratic society through means that generate the very cultural resources a democratic society requires to be sustained.

They also provide institutional lessons about how to potentially mobilize the beneficial aspects of market competition while preserving an overall culture of solidarity and reciprocity. As Cohen notes, markets can provide important informational and coordination benefits, yet these come along with attitudes that undermine reciprocity. While one potential limitation of unions under capitalism is that their demands are constrained by the requirements of firm profitability, they nonetheless show that markets in consumer and input products can be reconciled with nonmarket negotiation over the pricing of labor. By removing labor from processes of market competition between workers, unions enhance reciprocity and solidarity even as firms may still compete on the market, incentivizing the adoption of productivity-enhancing innovations and coordinating production.

Unions exist to advance the interests of their members. Yet unions have not existed in isolation. In many cases, they become part of a larger labor movement, allied with left-wing political parties, that connect the workplace struggles of workers within a single firm to a larger egalitarian agenda. The widening of union agendas has particularly occurred when capitalist employment structures have broken down traditional craft unions, which are organized to protect the specialized, individualized skills of their members, in favor of deskilled work processes. Unions are, therefore, a crucial anchor of egalitarian political movements. Unlike more voluntary social movements, they penetrate the domain of work and individual material reproduction. They appeal directly to the material needs of their members. Unions can thereby link the short- and medium-term goals of their members with the longer-term political goals of crafting egalitarian social and political institutions. These larger institutional reforms will affect the balance of power and interests within the workplace—for example, unemployment insurance raises the reserve wage of workers. There is thus a well-established empirical connection between high levels of unionization and economic equality (Ahlquist Reference Ahlquist2017).

These are egalitarian spillover effects, where the pursuit of equality in one area entails a dynamic restructuring of other institutions. If we assume we are not designing democratic institutions from scratch, or according to a fixed blueprint, then a key question becomes which institutional structures are most likely to generate these egalitarian spillover effects, or what I elsewhere (Klein Reference Klein2022) call egalitarian feedback loops. Unions are distinctively situated for generating egalitarian feedback loops—because they appeal to the short- and medium-term interests of their workers and not just abstract moral goals, because they are anchored in the workplace and so material reproduction, and because they internally rely on cooperation and solidarity and so require an egalitarian internal culture.

The strike is the most dramatic manifestation of these forces. In the strike, unions appeal to the short-term interests of their members but draw on solidaristic values that point to the longer-term normative ideals specified above. Indeed, we can see strikes as a crucial mechanism of exercising democratic power, one that combines features of voting and protest. Like protests, strikes are continuous and can be deployed at the discretion of the strikers. Like voting, strikes are not just symbolic—they can activate democratically-legitimated coercion, insofar as strikes block replacement workers from undermining workers’ collective power. It is important, in my view, that we see strikes as manifesting collective democratic power and not just a vehicle for resisting oppression, as Gourevitch argues. Indeed, I worry that Gourevitch concedes too much. A strike is a challenge, to be sure, to a narrow, liberal view of private property, but if the above set of normative arguments is correct, a strike does not violate a firm’s owners’ private property rights. Rather, they embody a different set of normative claims, claims already partially implicit in the organization of capitalist firms, but ones that are most clearly embodied in the self-organization of workers. They represent a moment of conflict between two normative orders, and one where the structures of power backing those two orders are pitted against each other. The question then becomes which will win: the bottom-up solidarity of workers’ cooperation, or the top-down coercive power of private property.

One may worry that this argument downplays the coercive element of strikes. Instead of a clash between two different interpretations of the same institution, we could more straightforwardly say there is a clash between two different rights-claims, with there ultimately being a correct answer as to when strikes can permissibly harm others (Borman Reference Borman2017; Christensen Reference Christensen2026; Dobos Reference Dobos2022; Raekstad and Rossi Reference Raekstad and Rossi2022; Stehr Reference Stehr2025). To some degree, this moves into questions of individual ethics that are beyond the scope of this paper. Individual workers as strikers will confront moral dilemmas in balancing their claims with potential harm to innocent bystanders, and their answers will no doubt be based on background moral commitments (for a discussion, see Christensen Reference Christensen2026). However, my more institutional analysis enables us to theorize the democratic function of strikes in the absence of agreement on larger moral and metaphysical questions regarding individual ethics. In addition, the ethics of strikes will need to contend with questions about the relationship between individual morality and political conduct, the nature of collective action, and nonideal political ethics.

Even if it does not provide a final answer to those questions, my account can further this effort by presenting a more accurate view of the institutional context in which to apply various moral principles. Put differently, our analysis of strikes will go astray if we accept from the outset the characterization of the firm as a domain of private property rights which strikes violate. This makes it seem like there are two competing normative principles at play—one, private property, the other, resistance to oppression—that stand opposed to each other. My point is that the strike advances normative claims that are already implicit in the capitalist firm—namely, that workers have a claim to a say in the production process and a negotiated portion of profit grounded in the workplace as a cooperative venture. The potentially coercive effects of strikes are downstream from workers’ right, grounded in this cooperative venture, to cease cooperating if they are denied voice and an appropriate share of profits. We can then inquire into the appropriate means workers can deploy when they decide to stop cooperating with employers, but our answer to that question will crucially depend on how we characterize the actions of strikers—as a justified violation of property rights, as in the radical view, or as an assertion of the already existing fact of cooperation that is obscured by the legal fiction of shareholder ownership, as on the democratic view.

These arguments focus on the significance of strikes for workers at the level of a firm or industry. But strikes are also a key mechanism for both prefigurative politics and egalitarian spillovers. The collective power embodied in strikes was crucial for democratization, either directly, as in Germany and Belgium, or as an implicit threat, as in the United Kingdom. These egalitarian spillover effects are closely connected to the prefigurative role of unions. Strikes can function as schools of democracy. As the 19th-century French syndicalist theorist Victor Griffuehles declared, “a strike is worth more than all the contents of libraries” (Magraw Reference Magraw1993, 109). Indeed, more broadly, there is extensive empirical support for the idea that unions play a crucial role in shaping democratic citizens, affecting attitudes toward being more inclusive and egalitarian and helping citizens gain valuable political knowledge (Frymer and Gumbach Reference Frymer and Grumbach2021; Macdonald Reference Macdonald2021).

THE PROBLEMS OF EXCLUSION AND CO-OPTATION

Yet these three features of labor unions—their connection to self-interest, their anchor in material reproduction, and their reliance on cooperation and solidarity—leave them vulnerable to problems of exclusion and co-optation. Many early unions were craft unions, whose members were focused on protecting the value of their skills by restricting entrance into their trade. To do so, these unions often relied on ethnic, racial, and gendered differences as grounds for exclusion. Labor unions supported laws restricting female entrance into the workforce. The most extreme manifestation of these exclusionary tendencies occurred in America, where some union members engaged in spontaneous, wildcat strikes to resist racial integration of workplaces. To an extent, unions are like all organizations and will reflect the society in which they operate. Of course, unions have also been powerful forces for inclusion—alongside the racial exclusion of craft unions, the Congress of Industrial Organizations (CIO) was at the forefront of fighting for racial inclusion in the workplace, and unions became important vehicles for feminist movements. Yet I think the exclusionary potential of unions is more persistently grounded in the very features that make them important vehicles for democracy. Unions’ reliance on solidarity and their ability to mobilize members’ material self-interest make them prone to narrowness and short-termism at the expense of broader demands for egalitarian inclusion. For this reason, the structure of the labor movement, unions’ connection to political parties, and unions’ relationship to other social movements will all factor into the democratic and egalitarian potential of union organization.

The other pathology of unions arises from their rootedness in the process of material reproduction. Even as this enables them to sustain organizational capacity among their workers, it ties unions to the economic fate of their erstwhile capitalist opponents. Unions can only demand so much if it threatens the economic viability of their members’ employers. This also ties workers to sectional national interests in the context of global interstate competition. Unions are not just threatened by co-optation; they begin co-opted. The working class is already constituted by capitalist relationships of production, such that their material existence is dependent on capitalist market processes. Unions are a feature of capitalist production as well as a source of counter-market resistance. Yet taken to their logical conclusion, unions are incompatible with the private control over investment and so the capitalist firm. The decoupling of wages from the market ultimately undermines capitalist profitability and so private investment. For this reason, far-sighted members of powerful labor movements have recognized that sustaining nonmarket wage systems requires restructuring investment. At their peak, such as in Sweden at the end of the 1960s economic boom, unions began pushing for a democratization of investment decisions, in that case through the Meidner Plan that would create union-controlled investment funds (Warner Reference Warner2026).

Some could argue that these structural features of unions will inevitably limit them as vehicles for democratization. They are adjacent to worries that, since unions represent workers, they will ignore the interests of the self-employed, the unemployed, or those who cannot conform to the values of the “work ethic,” such as people with disabilities or carers (James Reference James1974; Weeks Reference Weeks2011). But these worries ignore the importance of egalitarian spill-over effects. To be sure, under some conditions, unions will narrowly focus on the interests of their existing members. But those will be the cases where they lack the institutional support to balance the short, medium, and long term. Where unions have those institutional supports, such as through federated structures and unions that encompass skilled, unskilled, and service workers, they will see it as in their interests to push for reforms that benefit, not just their members, but broader segments of society. The embedding of unions in capitalist production is both their source of power and their potential limitation. However, it is one that can be overcome, and indeed the union ideal as I reconstruct it precisely points beyond a narrower vision of firm-based collective bargaining to these more encompassing institutional reforms.