THE AMBIGUITY OF “CAPITAL” OWNERSHIP

Radical republicanism’s noncontingency thesis does not simply affirm that domination—understood as unilateral, nonreciprocal, or arbitrary control by some over the labor of others—is a necessary feature of capitalism. The thesis also affirms that:

1. i. Capitalists necessarily control the labor of others, and thus dominate, by the mere fact of owning the means of production, regardless of their dispositions—precisely as slave masters necessarily dominate, individually, by the mere fact of legally owning slaves.

2. ii. Domination under capitalism is noncontingent in a way other wrongs, including exploitation (and distributive injustice), are not. While whether capitalists de facto exploit—extract surplus value—depends on their subjective, extractive dispositions, whether they dominate does not.

Vrousalis (Reference Vrousalis2026, 2) argues that my rejection of the noncontingency thesis, so understood, rests on the following two claims:

1. 1. Capital’s power over labor is contingent on many features other than ownership of productive assets, including the capitalization process.

2. 2. Capital’s power over labor is contingent on some form of capitalists’ control over the capitalization process.

Vrousalis accepts (1) but rejects (2). Just as a mafioso need not control the system of cultural and social beliefs that supports his power to dominate those subject to his whim, the capitalist need not control the process of capitalization that confers monetary value to his means of production in order to dominate workers (Vrousalis Reference Vrousalis2026, 3).

I certainly agree that an agent, in order to dominate, need not control the structures and processes that sustain his power to dominate. Indeed, I do not endorse (2). I do not claim that, in order for an individual owner O to have the power to control worker W’s labor, O must unilaterally control the valuation of O’s means of production. To the contrary, O could be conferred that power as a result of processes O does not control.

What I do claim is that at T₀, the mere fact that O owns “the means of production”—land, machines, or accumulated savings—does not ipso facto provide O with the actual power to control W’s labor at T₁, because the capitalization process could take place in a way that confers no value to those means, and O lacks the power to unilaterally direct the process of capitalization so as to empower himself. If O did have the power to unilaterally control the process of capitalization, that control, jointly with ownership, would be sufficient (though not necessary) to provide O with control over labor at T₁. We could then say that, if O had that power over the process of valuation at T₀, O would also have the actual power, already at T₀, to control labor at T₁. But insofar as O lacks that power, his individual power to control labor is merely potential at T₀, and its actualization remains contingent on the future-oriented expectations of investors as they aggregate throughout the capitalization process between T₀ and T₁. Ownership of “the means of production”—if understood as accumulated past savings or objects such as factories, land, and machines—is thus insufficient for domination. No one, therefore, necessarily dominates by the mere fact of owning such means. The first part of the noncontingency thesis would thus seem to fail.

Radical republicans might, however, respond that I have misunderstood what they mean by “means of production.” They may argue that this vexed term refers not to potentially productive objects (accumulated savings, land, or machines that could be used to produce other goods) but to assets the value of which is already valorized or “self-valorizing.” On this second reading, individual owners necessarily dominate by the mere fact of owning assets that are actually productive because already valorized—their valorized character is what makes them capital. Ownership of capital, so understood, would be both necessary and sufficient for domination.

There are three problems with this rejoinder. First, there is little textual evidence for the second reading. In the radical republican literature (e.g., Bryan Reference Bryan2023; Cicerchia Reference Cicerchia2022; Gourevitch Reference Gourevitch2013; Vrousalis Reference Vrousalis2020), one notices the absence of any sustained discussion of (i) the conceptualization of “the means of production,” (ii) their relation to the process of valuation, and (iii) what is meant by, and included in, the concept of capital. It is therefore difficult to determine what radical republicans mean when they say that by owning “the means of production” capitalists dominate. At a minimum, my critique should then be read as an invitation to clarify the concept of “means of production,” and its relation, through the process of valuation or valorization, to that of capital.

Second, and relatedly, Vrousalis is, among radical republicans, the one who has done the most to unpack the economic-theoretical underpinnings of radical republican critiques. Yet what we find in his original account is an economic model that treats the capitalization process as a dimension of expectations formation that can be nearly assumed away altogether. In this account, expectations explain only momentary fluctuations in underlying prices, and so the process of capitalization affects the valuation of capital goods only trivially. I do not endorse this account. As Vrousalis rightly notices, I follow post-Keynesian economic theory (see, e.g., Beckert Reference Beckert2016; Levy Reference Levy2017; Nitzan and Bichler Reference Nitzan and Bichler2009; Orléan Reference Orléan2014) in treating capitalization not as trivial but rather as constitutive of capital. The way investors’ subjective expectations about a radically uncertain future form and aggregate is a constitutive part of the valuation process, and it is this process that transforms potentially productive “means of production” into actual capital. Capital is thus an asset imbued with monetary value through a process of capitalization. Therefore, insofar as they do treat capitalization as trivial, radical republicans are not entitled, in my view, to say that the ownership of “capital” is necessary and sufficient for domination, because what they refer to as capital ignores capitalization, and thus is not “capital” in the relevant sense. This is not a semantic disagreement. It is a theoretical and substantive disagreement about the nature of capital, which remains an undertheorized concept in political philosophy.

Perhaps radical republicans could endorse the view that capitalization is a constitutive element of capital. Indeed, Vrousalis seems to believe they could do so without any consequences for their theory. They could then claim that all those who own capital—assets to which value is attributed according to investors’ subjective expectations—necessarily dominate, because they necessarily end up unilaterally controlling labor, regardless of their own subjective dispositions.

But, first, notice that one may not need to own capital, so understood, to control labor. All one may need is the power to control the deployment of capital, or to be the beneficiary of the process of valuation (many powerful capitalists nowadays are debtors). But let’s leave the ownership/control split aside. Is it really true that those who own capitalized assets necessarily end up controlling labor? Consider a world in which the process of capitalization led to the high valuation of labor-substituting technology—say, AI instruments—rather than industry, land, or services. In such a world, capitalization might not result in the owners of the means of production having an interest in directly controlling the terms of labor, but rather in massive, permanent unemployment. Capitalists’ profits would mostly derive from reinvestment, financial speculation, asset appreciation, and products made by robots and sold to consumers, perhaps endowed by their states with a universal basic income, rather than from the direct exploitation of labor.Footnote ³

Radical republicans might (rightly) reply that fully automated capitalism is impossible and that capitalist reproduction necessarily presupposes the power to control labor, because without some extraction of surplus value the process cannot sustain itself. Domination over labor is thus necessary and ineliminable because exploitation is ineliminable.

Fair enough. But if this is correct, then exploitation has causal primacy over domination: domination necessarily exists if, and only if, exploitation exists. But then in what sense is domination, as Vrousalis instead argues, less contingent than exploitation? Vrousalis’s response (Reference Vrousalis2020) is that exploitation is contingent in a way domination is not, because only the former depends on the capitalist’s subjective, extractive dispositions. Yet, as just noted, there must be at least some exploitation for capitalism to exist. So, the noncontingency thesis cannot plausibly be read as the claim that the mere existence of domination is less contingent than the mere existence of exploitation—both are equally noncontingent. What is uniquely contingent on capitalists’ extractive dispositions must instead be the total amount of exploitation and the identity of exploiters—who, exactly, extracts—not the mere existence of exploitation as such. Logically, then, the noncontingency thesis should be reinterpreted as the claim that the total amount of domination and the identity of dominators are, unlike the amount of exploitation and the identity of exploiters, noncontingent.

So reinterpreted, however, the thesis is false. The overall amount of domination, as well as the identity of dominators, is as contingent on subjective dispositions as the overall amount of exploitation and the identity of exploiters. Although it is true—and I accept as much in my article—that the process of capitalization will necessarily end up providing some owners of potentially productive assets with the actual power to control labor, the total amount of domination—how many of those owners, and which ones, will actually end up controlling labor at T₁—depends on how investors’ subjective preferences aggregate throughout the capitalization process between T₀ and T₁, that is, on which and how many of those assets are transformed into capital. The only difference between exploitation and domination, in this respect, is that the former is contingent on the owners’ own dispositions, while the latter is contingent on the dispositions of investors. The second part of the noncontingency thesis is thus rejected.

At this point one might ask: if Vrousalis and I both agree that, under capitalism, some agents will necessarily end up with the actual power to control the labor of others, where exactly does the disagreement with radical republicanism lie?

First, depending on which accounts of capital, of valuation, and their relation radical republicans adopt, we may differ on the source of capitalist power. Does this power exclusively derive from legal ownership of previously accumulated savings, or also from control without ownership; from inherently productive objects, or from assets that are capitalized by means of future-oriented expectations? The answer to these questions makes a difference, if only to understand the specific temporality of capitalism.

Second, and most importantly, republicans assume that controlling labor necessarily amounts to a form of (republican or Kantian) domination. I reject this claim in the following section.

CAN KANT RESCUE THE REPUBLICAN CRITIQUE OF CAPITALISM?

Vrousalis (Reference Vrousalis2026, 6–7) seems to agree with me that capitalists cannot dominate in the traditional neo-republican sense merely by controlling the terms of labor, since arbitrary interference is insufficient to account for capitalist domination (I defend the stronger view that individual capitalists do not, in fact, interfere). He argues, however, that capital dominates labor in a Kantian sense, which does not require interference. Specifically, he claims (7) that “capitalist private property cannot be reciprocally justified to free and equal people because it would make the poor into servants of capital. And, since servitude entails domination, capitalism is necessarily a form of domination.” He then adds that “no democratically revisable vote could condone capitalist private property… otherwise free and equal humans would be, in Kant’s memorable phrase, ‘throwing away their own freedom.’”

The above passage leaves unclear who the dominator is. On one reading, domination appears to be a feature of the legal system (of property): it pertains to the fact that the system of laws subjects citizens to rules that cannot be hypothetically consented to, thereby violating their equal freedom. Since the law cannot be regarded as emanating from “an omnilateral will,” it remains a merely unilateral imposition. On this reading, it is the state—which Kant understands as a system of legal rules—that dominates, not “capitalism” or “capital.” The capitalist system of private property is just one among the many laws that may fail to be hypothetically justifiable. By imposing such laws, the state dominates by restricting the freedom of all citizens, not only laborers, in a way that contradicts their independence.

Vrousalis could respond, however, that those laws are unjustifiable precisely because they sanction domination at a second stage—the stage at which “capital” gets to control and command, at its discretion, the labor capacity of nonowners. Since control and command require an agent with intentional capacities, “capital” must be understood either as a class, qua group agent, or as a set of individual capitalists. Since Vrousalis (9–10) later rejects the view that capital as a class qualifies as a group agent capable of intentional action, I will proceed with an individualistic account of the agents of control over labor.

The question thus becomes: does individual capitalists’ control over labor violate Kantian freedom? This is a difficult question, and I will not be able to do full justice to it here. I will limit myself to arguing that Vrousalis’s affirmative answer to it remains unpersuasive.

In the Kantian account, domination is understood as a violation of freedom as independence, defined as “independence from being constrained by another’s choice” (Kant Reference Kant and Gregor1996, 6:237). Since work relations under capitalism involve constraints by another’s choice—insofar as owners unilaterally, that is nonreciprocally, control the labor process of nonowners, e.g., whether to offer them jobs and under what conditions—Vrousalis is correct that the capitalist mode of production would seem, prima facie, to violate Kantian freedom (Vrousalis Reference Vrousalis2020, 88).

But is this actually the case? One reason to doubt it is that what, for Kant, violates independence is not only unilateral but merely unilateral control over labor.Footnote ⁴ To see why, consider the case of property. A homeowner unilaterally controls what others can do on her property, for she has the right to exclude others from her home. Yet, for Kant, the homeowner’s right to exclude can be omnilaterally authorized—it can be willed compatibly with everyone else’s equal freedom—so that the resulting control, though unilateral, is not merely unilateral, and as such does not violate independence, for it does not subject anyone to the purely private legislative will of another (Ripstein Reference Ripstein2009). The crucial question therefore becomes: can unilateral (nonreciprocal) control over labor be omnilaterally authorized?

Vrousalis is right that not everything can be omnilaterally authorized compatibly with freedom as independence. Most notably, the omnilateral will cannot sanction a system of slavery, or a property system that would leave some entirely dependent on others for access to subsistence. The main reason is that such systems would make some people’s capacity to form and pursue ends dependent on the discretionary will of others, thereby violating their independence. This is why, for Kant, the state has a duty to secure some level of public provision as a necessary condition for the legitimacy of any property system (Weinrib Reference Weinrib2003, 795).

But unlike complete propertylessness or slavery, work relationships under capitalism do not necessarily infringe on the capacity of nonowners to pursue ends as such—their general purposiveness.Footnote ⁵ At most, they undermine the ability to independently set and pursue productive ends, or ends through production (indeed, Vrousalis [Reference Vrousalis2022b] argues that unilateral control over labor undermines “productive purposiveness.”). But it is unclear why independence should protect this specific kind of purposiveness. As free beings, for Kant, we are arguably entitled to set and pursue ends, but not to the free pursuit of any particular end or of ends through specific means or even practices, such as productive practices (Ripstein Reference Ripstein2009). In capitalist societies, where the welfare state is strong enough to guarantee a reasonable work-life balance and to support civil society associations and other forms of nonproductive activities, people may still be able to form and pursue many ends, including the creation of new objects and the provision of services to others, outside of the employment relationship.Footnote ⁶

There is a second reason one might think that unilateral control over labor cannot be omnilaterally authorized: without some control over the labor process, the ability to freely legislate as an independent political subject may be compromised (Kant Reference Kant and Gregor1996, 6:314). Republicans have long argued that economic independence is a precondition of political independence, since complete material dependence on others undermines either one’s independence of judgment or one’s capacity to act freely on that judgment.Footnote ⁷ Both capacities can, however, arguably be secured by providing substantial exit options from any particular employment relationship, together with policies ensuring ballot secrecy and the like.

Vrousalis (Reference Vrousalis2022b) might add that civil personality requires that citizens retain their status as independent agents—agents whose free purposiveness is not under the unilateral control of others. Under capitalism, the argument goes, nonowners cannot be independent, for the exercise of their productive powers is subject to alien control: if you own nothing, the only way you can deploy your productive capacities is by granting others permission to use them. It remains unclear, however, why this undermines one’s status as an independent agent. After all, the only way I can exercise my associative powers is also by giving others permission to associate with me and to benefit from those powers—I give my friends permission to rely on my friendship and on the obligations it generates—yet this seems perfectly compatible with my independence.Footnote ⁸ Furthermore, if my status as a free citizen depends on whether the exercise of my productive powers is subject to alien control, that status would seem to depend a fortiori on whether my subsistence is subject to alien control. But then no one would seem to possess civil personality, for even independent producers remain dependent for their subsistence on giving others permission to purchase their products, and on others’ discretionary willingness to do so. The argument that a capitalist mode of production cannot be omnilaterally authorized because it is inherently incompatible with civil personality thus risks overreaching, and as such it remains unpersuasive.

Vrousalis offers a further reason why a capitalist system of private property over the means of production cannot be authorized. Citing Przeworski on the structural domination of politics by capital, Vrousalis (8) argues that “capitalism makes the implementation of any democratic decision subject to the private choices of the owners of productive assets,” and that this is why, on republican grounds, it cannot be democratically authorized. Several problems beset this move. First, Przeworski (Reference Przeworski1985) would be the first to acknowledge that the degree of political autonomy from capital is largely a contingent matter, and it is therefore simply not true that the implementation of any democratic decision is subject to such domination. Second, if capital dominates by dominating politics, its domination falls on all those subject to political decisions, not just workers. Third, capitalists can dominate politics quite independently of their control over labor at the point of production—in a fully automated world, capital could still dominate politics by threatening to strike or to withhold investment in essential activities. Last but not least, if nonreciprocal influence over politics is where the crux of the Kantian objection rests, the problem could in principle be resolved, without exiting capitalism, by distributing and dispersing private property sufficiently widely. The result would be a property-owning democracy, not a postcapitalist, socialist one.

Because of the above reasons, I remain unpersuaded by Vrousalis’s argument that an omnilateral will cannot authorize a system of unilateral control over labor, provided that regulations secure sufficient exit options from particular employment relationships, adequate leisure time, and the conditions for a property-owning democracy. If the capitalist mode of production can be validly authorized under such conditions, capital does not dominate. The domination thesis is compromised. If capital does dominate—because and when those conditions have not been met—it does so only contingently, in which case the noncontingency thesis is compromised.

DOES “THE CAPITALIST CLASS” DOMINATE?

Vrousalis (8–9) agrees with me that capitalists as a class do not intentionally dominate qua a group agent. He also agrees that capitalism cannot impersonally dominate—the subject of domination cannot be an impersonal structure (Cordelli Reference Cordelli2025). He wants to argue, however, that there is such a thing as agentless class domination: domination by a class, where the class nonetheless does not qualify as a group agent. To hold this middle position, Vrousalis appeals to the intriguing notion of “plural requirements”: requirements that certain pluralities of agents act jointly, though not severally. Vrousalis (10) writes: “In not giving up her profits, none of the capitalists K₁, K₂… does her workers any wrong, assuming enough capitalists are not giving up their profits either [for, at least under certain assumptions, individual capitalists do best by themselves and their workers if they maximize their profits]. And yet, the capitalists collectively do violate a plural requirement not to dominate.” In this way, “class domination involves groups of agents wronging other groups, without necessarily amounting to group agents.”

I must confess I find mysterious the notion of a group of agents doing wrong without itself being an agent, and without any mechanism for either forming shared intentions or coordinating action. Moreover, domination by a plurality of agents is not the same as structural domination—at most it is a structurally enabled form of domination. But let us set both points aside.

Why would capitalists dominate as a group of agents? Because, Vrousalis says, they fail to give up their profits and thus their power over labor. But why does this failure amount to domination? Presumably because having such power over labor itself necessarily amounts to neo-republican or Kantian domination. Yet this is precisely what, I have argued, neither neo-republicans nor Kantians can establish. Let us set this point aside as well.

One arguably distinctive feature of the concept of domination is its inherently inegalitarian character. Both in its neo-republican and Kantian versions, domination implies a status harm that undermines its victims’ ability to see themselves as free persons who can “walk tall, and look others in the eye” (Pettit Reference Pettit2012, 82), or their standing as equal normative authorities (Gädeke Reference Gädeke2020). The kind of status harm that defines domination can, it should be noted, only obtain through subordination to a superior agent—someone whose status is recognized as higher according to an established social hierarchy, or someone who claims or commands higher normative authority. Now, insofar as a group of agents joined together only by plural requirements is not itself an agent, and cannot claim higher social status or higher normative authority as such, there can be no status harm in relation to it. It follows that, without such harm, domination, conceptually speaking, cannot obtain.

Furthermore, recall that for Vrousalis the class of capitalists, understood as a plural agent, wrongs noncapitalists, because and to the extent that the capitalists can jointly give up their power to control the labor of others. Yet such power, for Vrousalis, is attached to the “office” of the capitalist, not to the particular individual that occupies it. Yet, like for all other offices, the occupants of the capitalist office are substitutable. Even if most capitalists were to vacate their office voluntarily, as long as the office remains, noncapitalists might seek to replace them (former members of the proletariat might attempt to become capitalists themselves), rendering the former’s action futile. Therefore, in order to end “class domination” it is not sufficient for the capitalists to “give up their profits.” The relevant office must also be abolished by eliminating private ownership as such. All this means that the plural requirement to leave the capitalist office vacant in order to eliminate class domination is highly conditional and must be effectively universal: capitalists have a duty to relinquish their office, and give up their property and profits if and only if most capitalists simultaneously do the same, most noncapitalists refrain from substituting them, and everyone refrains from supporting the institution of private property. In the absence of such conditions, the capitalist class’s actions to vacate the office risk being futile, while potentially coming at high costs not only to the capitalists, but also to the workers. The relevant plural agent who must jointly act to end domination would then seem to be broader than the capitalist class alone, and the latter’s failure to vacate the office, if taken separately, may not amount to wronging. This is a further reason why Vrousalis seems unable to rescue, in analytical terms, the notion of “class domination” as the distinctive wrong of capitalism.

This is, I think, a serious problem not only for Vrousalis but for all those normative critiques of capitalism that center on the notion of class domination. If domination by an impersonal structure is incoherent, if a class cannot be understood as a group agent, and if a plurality of agents cannot dominate in the relevant sense, then the idea of “class domination”—however intuitive and politically attractive—remains philosophically mysterious.Footnote ⁹

NONALIENATION AND THE DEMANDS OF SOCIALISM AS RECONCILIATION

The mode of investment is the engine of capitalism. Investment, including in the form of institutionalized credit, can enable production even in the absence of accumulated savings and, because of the way money can serve as a store of value, speculative investment can obstruct production even when accumulated savings are present (Cordelli Reference Cordelli2025). But is there anything wrong or amiss with being subject to such a mode?

In my view (Cordelli Reference Cordelli2025), when it comes to the mode of investment, the core ill of capitalism is alienation, not domination. The latter is, conceptually, about being subject to the power of an identifiable agent; the former is about the inability to relate to the external world as something we, as citizens, have contributed to author and can affirm as actualizing our joint intentions, rather than as the product of uncontrollable, nonagential, and impersonal forces. Phenomenologically, domination is about ingratiation and a sense of inferiority toward a superior; alienation is about powerlessness and detachment from a world that is and appears as the result of impersonal, uncontrollable forces.

The reason I take capitalism to be necessarily alienating is that living under the governance of investment market forces is like living under the governance of an impersonal computer machine that makes transformative decisions—decisions about the future shape and course of our society—in our place (Cordelli Reference Cordelli2025), by aggregating private choices. The analogy rests on the assumption that massive inequalities of power between investors notwithstanding, there is no agent—individual or collective—to whom investment patterns can be attributed. The investment process is thus impersonal in the sense of being nonagential. If we assume that the machine is not itself an agent, we must also assume that the machine cannot dominate us. Indeed, we could not ingratiate ourselves to it. Further, in the specific case of capitalist investment, I argue, “the machine” is also anarchical and prone to generate irrational outcomes. Living under the rule of an impersonal and irrational machine, I further contend, is alienating, because those subject to it can only relate to the society that derives from that rule as the product of an uncontrollable and impersonal force, rather than as the result of their collective doing and valuing. Moreover, they cannot affirm either the process or the outcome as something that expresses their deliberatively endorsed priorities (Cordelli Reference Cordelli2025). Since, as others have also argued (Stilz Reference Stilz, Sobel, Vallentyne and Wall2016), the citizens’ interest in relating to their own political society in a nonalienated way is a ground of democratic self-determination, democracy cannot authorize capitalism without contradicting itself.Footnote ¹⁰ Capitalism may thus be an illegitimate mode of governance, whether or not it is also unjust.

Vrousalis accepts the analogy between capitalism and the computer machine, and agrees that capitalism is illegitimate. He disagrees, however, that this illegitimacy stems from the alienating character of society under the machine. The problem for him is instead about surrendering an inalienable entitlement: “to hand over political decision-making to the machine is [for citizens] to give up their entitlement to be judged on the basis of their own acts” (7). This entitlement, he holds, is unwaivable. But Vrousalis does not tell us why a democratic people cannot alienate the entitlement to be judged on the basis of their own acts, why they would possess such an entitlement to begin with, or why delegating choices to the machine would amount to giving it up. After all, even if the machine could not provide reasons for its decisions, citizens could still judge one another on the basis of that very original choice to delegate.Footnote ¹¹

By contrast, it is, I think, clear why an entitlement not to be alienated cannot be democratically surrendered. The argument, which I have developed elsewhere (Cordelli Reference Cordelli and Zimmermann2026) and to which I can only briefly allude here, runs as follows. To explain why democratic self-determination—rule of all—is justified over rule of none, we need a value that accounts both for why we want our joint intentions to be expressed in the world and for why we want them expressed through our joint authorship (Lovett and Zuehl Reference Lovett and Zuehl2022). The answer points to a more fundamental interest in relating to the world in a particular way. We want the world to express our joint intentions because we want to relate to it as something we can affirm and identify with, rather than as something alien to us. And we want our intentions to shape the world through our own choices and actions because we want to relate to it in an agential capacity—as something we have contributed to bring about—rather than passively, as an external, alien imposition. In short, we want joint authorship because we want to relate to our society in a nonalienated—authorial and affirmative—way.

Importantly, this relational interest in nonalienation also explains what is wrong with being ruled by an impersonal machine—“rule of none”—even if such rule does not subject us to anyone’s particular will or create objectionable relational inequalities between rulers and ruled. Under rule of none, alienation arises from the fact that citizens are passively subject to rules they have played no part in shaping and cannot control or help transform. Further, if the rule is anarchical, they cannot reasonably be expected to affirm its process and results as reflecting their endorsable priorities, even if the outcomes happen coincidentally to converge with them. Citizens cannot therefore be reasonably expected to regard the socio-political world that results from impersonal and anarchical rule as an expression of their own shared agency, and they lack the conditions to affirm it as their own.

The fundamental advantage of an alienation-based framework for understanding the wrong of capitalism, I believe, is to make visible how, even in the absence of identifiable masters—and even in a fully automated capitalism with a universal basic income—so long as the power to invest and to decide about transformative decisions concerning a society’s future remains privatized and exercised through irrational and impersonal market forces, capitalism will remain illegitimate. Capitalism, qua a form of governance, is illegitimate because it is alienating, and it is alienating because, first and foremost, it usurps citizens of the right to collectively deliberate and decide on which future they want to try to actualize. It deprives citizens of the opportunity to take advantage of the open and indeterminate nature of the future to jointly shape their society’s course according to values and priorities that go beyond the fulfillment of private investors’ claims to monetary returns.

Would democratic capitalism, as Vrousalis suggests, suffice to dissolve the problem of social alienation? Certainly not. Nonalienation requires that the direction of investment be determined through citizens’ democratic, ongoing involvement, not by firms acting atomistically.Footnote ¹² More precisely, an authorial relation with the social world requires joint control over investment. An affirmative relation further requires direct agency involvement, needed to develop an attitude of caring for the direction of that world. It also requires that investment be planned ex ante so as to avoid anarchy,Footnote ¹³ and that citizens retain control and involvement over both the making and the implementation of that plan. Efficacious and robust control in turn presupposes the socialization of productive ownership.Footnote ¹⁴ My alienation-based critique thus demands the socialization and democratization of the investment process, over and above the democratization of the workplace and workers’ collective control over the means of production. Socialism as reconciliation (i.e., disalienation) is more demanding, not less, than republican or liberal socialism.Footnote ¹⁵