The Boundary Problem in Workplace Democracy: Who Constitutes the Corporate Demos?

Abstract

This article brings to bear findings from the debate on the boundary problem in democratic theory on discussions of workplace democracy to argue that workplace democrats’ focus on workers is unjustified and that more constituencies will have to be included in any prospective scheme of workplace democracy. It thereby provides a valuable and underdiscussed perspective on workplace democracy that goes beyond the debate’s usual focus on the clarification and justification of workplace democrats’ core claim. It also goes beyond approaches like stakeholder theory in law and economics that determine decision-making rights without taking into account genuinely democratic considerations. My discussion proceeds by considering three principles for inclusion from democratic theory for the specific case of the corporation. I submit that two of them, the all-coerced and the all-subjected principle, are not appropriate for this specific case, because they cannot capture the distinguishing features of the corporation. The all-affected principle however is appropriate but has a very wide range. I further argue that this is not as big of a problem as it first might seem and that this principle is still the most appropriate for defining the demos of the democratic corporation. The article closes by pointing out the consequences of this result for the workplace democracy debate and for the legitimacy of the market as a coordination mechanism.

Introduction

Business corporations are some of the most powerful institutions in the world today, in some cases arguably outclassing states. They play an important role in organizing production and distribution, and increasingly also public debate and social relations. At the same time, the vast majority of corporations is internally profoundly undemocratic. The theoretical debate around workplace democracy has long recognized this fact (Frega, Herzog, and Neuhäuser 2019). Elizabeth Anderson has summarized it in the four features of private government (Anderson 2017): superiors can rule arbitrarily and are unaccountable to their subordinates. Employees have no right to appeal decisions nor do they have any right to be heard. To rectify what is thus taken by many to be a dictatorial status quo, scholars have proposed the introduction of democratic mechanisms in the workplace. Although the details of the proposals vary, they share the idea that, as in national democracies, workers should be given control over who rules them under what conditions via some election mechanism. After all, a widespread conviction is that for a form of rule to be democratic, the governed have to have a say in government.

Although the attention that workplace democracy has received among political philosophers is considerable, the debate focuses on two main aspects: the clarification of the claim of workplace democrats and the justification for that claim. What has been fairly absent from the debate however is any further concern with the subjects of workplace democracy: who should get a say in a democratic workplace? Authors usually quietly assume the subjects to be the workers or employees as opposed to the current holders of decision-making power, management and shareholders. Some authors grant that there might be other constituencies that have legitimate claims to participation (e.g., Ferreras 2017, 142). The only exception to this rule is Robin Archer, who explicitly tests his argument for workplace democracy for whether it applies to other constituencies besides workers (Archer 1994). The otherwise widespread neglect of this question represents a serious omission in the literature, not least because popular justifications for workplace democracy also apply to other constituencies (Moriarty 2010). Relational inequality or domination cannot only arise from the work contract, but for example also from supplier relations or for communities relying on taxes from a corporation. In this essay, I want to contribute to the discussion of nonworker constituencies in workplace democracy by answering explicitly the question who should constitute the demos in the workplace. To do so, I use insights generated in the debate on the boundary problem in general democratic theory (Goodin 2007; Näsström 2007; Abizadeh 2008; Song 2012). This represents an innovation in the literature on workplace democracy as the few discussions of nonworker constituencies have all relied on extending the justification for workplace democracy itself instead of building on the general discussion of the boundary problem in democratic theory. It also represents an interesting practical application of the general findings from the debate on the boundary problem that has so far been almost exclusively focused on state borders. My argument will be explicitly directed at workplace democrats and therefore assume that the general argument for workplace democracy as rehearsed above holds true and not further argue for why democratic accountability for business corporations in general is preferable. I will however refrain from making any further assumptions about the normative grounds of such an argument.

My discussion will proceed as follows. First, I will consider two existing answers to the constitution of the demos in the corporation. This is on the one hand the shareholder primacy doctrine, which stipulates that shareholders have an exclusive right to control the corporation. On the other hand, there has been an extension of this approach as stakeholder theory, which stipulates that people who have a stake in the corporation should also play a role in decision-making. Both of these approaches are however unable to adequately accommodate the specifically democratic concerns outlined previously. Therefore, the second part of my essay will then proceed by applying the general principles we find in democratic theory to the business corporation and reviewing the most plausible result. I will argue that, as in general democratic theory, the all-affected principle yields the most plausible result but gives us a large and unwieldy demos. In my third and final step, I will then consider the implications of this result for the introduction of democracy in the business corporation. Most importantly, strict and consistent application of the all-affected principle to the corporation will give us an argument for abandoning the market mechanism as a device for the organization of production and distribution in favor of a system of democratic planning.

Before I go further into the discussion of the corporate demos, let me delimit the object of my discussion. So far, I have used the workplace of workplace democracy and the business corporation interchangeably. However, they do not denote the same thing. In the workplace democracy debate, the objects of discussion are all the legal institutional backgrounds through which workplaces are organized. Any democratic reform will have to be implemented through the rules of those institutions. Corporations are one specific kind of institution via which workplaces are organized, offering some specific advantages like limited liability when compared to partnerships for example (Ciepley 2013). Here, I will focus on business corporations for two reasons. First, corporations are the most important institutions for the organization of workplaces, both in terms of numbers and for the fact that the most important economic institutions today are organized as business corporations. Second, the focus on business corporations allows me to build on the theoretical body that has already been established regarding the corporation, especially regarding control rights between different factions.

Theories of the Corporation and the Issue of the Demos

What do existing theories have to say about the constitution of the demos of the corporation? I will now briefly consider the two most prominent existing approaches to this question, the shareholder primacy doctrine and its recent extension in stakeholder theory.

In corporate law and economics, there is a clear and prominent answer to who has a legitimate decision-making right in the corporation: the shareholders (Kraakman et al. 2017, 5). This right is conventionally derived from the idea that owning a share means owning a part of the corporation. Therefore, shareholders have a right to decide what is being done with their property. Additionally, proponents argue that having shareholders both control the corporation and have a right to residual profits is the most economically beneficial configuration (Hansmann 1996). For these reasons, shareholders are granted the right to elect the top executives in the corporation as well as further rights, like information rights. Other parties only rarely have these kinds of voting rights, but their interests might be protected via other forms of regulation that, for example, force the corporation to share certain information or appoint “neutral” trustees to the board (Kraakman et al. 2017, sec. 3.3.1). Additionally, other parties’ interests might be protected via distinct areas of law like labor law. In opposition to this shareholder-focused approach, arguments have been raised based on the economic consequences of shareholder primacy (Stout 2012) or the internal inconsistencies of the idea that shareholders own the corporation (Ciepley 2019).

Since the 1970s, stakeholder theory has been developed as an alternative to the shareholder paradigm (Freeman et al. 2010). It aims to shift focus away from the welfare of shareholders as the central criterion for decision-making by considering the ethical and economic impact of corporate decision-making on people who have a stake in the corporation. For this purpose, there is also some discussion of stakeholder involvement in decision-making processes. Such policies are taken to offer effective mechanisms for letting corporate management know what their stakeholders desire and thereby better serving the core purpose of corporations of “creating as much value as possible for stakeholders, without resorting to trade-offs” (Freeman et al. 2010, 28). However, the approach is committed to certain paradigms that make it inadequate for answering the question with which I am concerned. Stakeholder theory, like other more narrowly economic theories of corporate behavior, has no concept of democratic rights. Its core normative guideline is a business corporation’s responsibility toward its stakeholders both in terms of economic fulfillment of their material interest and in terms of ethical responsibilities. But at no point does the theory recognize the core democratic intuition that people have an innate right to having a say in their affairs. Instead, the basic institutional setup of the corporation with its separation between workers and management, the inside of the corporation and outside contractors and customers, is taken for granted. The same holds for management theories advocating social responsibilities more broadly, like ESG or triple bottom line approaches. Although I do not judge the merits of these approaches for their stipulated purposes, they cannot offer an answer to who should be involved in corporate decision-making for democratic reasons. To answer that question I therefore turn instead to democratic theory.

Learning from Democratic Theory: Principles Applied to the Corporation

In democratic theory, there has been some debate on “the boundary problem” (Whelan 1983). At its core, this debate is concerned with the subject of democracy and aims to pay increased attention to the people who are allowed to make decisions instead of just the procedures by which they do so. This is an important issue, because democracy is considered to be legitimate because and insofar as the governing and the governed are identical (Abizadeh 2012, 868). Democracy is the rule of the people over the people. A government that were to exercise power over a group of people without giving them a say could not be considered democratic. A convincing justification of the boundaries of the demos is important because it touches on the core principles of democracy.

Justifying the boundaries of the demos is complicated, however, by the fact that the boundary of the demos cannot be decided democratically. For to do so we would already need a body of people to vote in this decision. But who belongs to this body is precisely the question we want to answer. The determination of the demos is therefore logically prior to actual acts of democratic government. We thus cannot resort to a democratic decision to determine the demos and instead will have to find a general principle that we justify otherwise. This constellation also causes a problem for the question of what kind of democratic mechanism the principles ground. The details of democratic procedures are usually decided by an existing governmental body. For my theoretic inquiry however, I cannot assume that these decisions have been made, because how to assemble such a body is precisely the question I want to answer. One possible way out of this conundrum has recently been outlined by Arrhenius, arguing that Proportionalism could help us to determine democratic rights and think about modes of co-determination at the same time (Arrhenius 2019). In my discussion, I will rely on a simplistic notion of democratic participation denoting a right to vote in a simple majoritarian procedure. A full institutional proposal could take into account different modes of participation and democratic accountability but is beyond the scope of this paper.

Several authors have objected that in practice we seem to have solved the conundrums raised in the boundary problem, because we are able to decide things democratically. Like Joseph Schumpeter, for example, they argue that there is nothing for us to say about the boundaries of the demos that every demos chooses for itself (Schumpeter [1942] 1950, 244–45; Näsström 2007). But this way forward seems profoundly unattractive, because it forces us to call boundary configurations democratic that are in opposition to other values we assume to be important to democracy, especially the value of equality (Song 2012, 42). With no criteria to evaluate the definition of a demos, we would be forced to accept discrimination along the lines of race or gender, for example, as democratic. Similarly, the contingent historical configuration of demoi along the lines of state borders has been the target of criticisms within the philosophy of migration (Abizadeh 2008). To raise these criticisms we need a general principle on which we can base the evaluation of the configuration of a demos.

In the literature, different principles are considered for defining the demos. These are the all-affected principle, the all-subjected principle, and the all-coerced principle. Respectively, they suggest that being affected, subjected, or coerced by state government gives one a right to a say in the decision-making procedures that lead to these measures. ¹ In the following I will apply these principles to the business corporation. ² Just as with the state, we have to ask who stands in the correct relation to fulfill the respective criterion. And, as I have established previously, we will have to find such a criterion without forestalling any democratic decision. When we try to apply the principles to the corporation, we can however assume that the institutional characteristics of the corporation remain relatively fixed. The core decisions that business corporations take are about production and distribution and the infrastructure that is needed to set these up. In the application of the principles, I will therefore consider who stands in the respective relation to a corporation when it takes a decision regarding this core set of tasks.

The All-Coerced Principle

Let me start out with the most restrictive principle. As Arash Abizadeh has defined it, the all-coerced principle argues that everyone who is the object of an action that “directly and pre-emptively deprives a person of some options that she would otherwise have had” or who is threatened by “sanctions should they carry out proscribed actions” should have a say (Abizadeh 2008, 40). The first group ³ that comes to mind that fulfills these criteria are workers. ⁴ As Elizabeth Anderson shows, the extent to which employers have arbitrary power over many aspects of their employees’ lives is considerable (Anderson 2017, chap. 2). Employers have the power to regulate how employees spend their time. In some jurisdictions, notably the United States, they can even fire workers over political speech expressed in their free time. During working hours, these powers are extended and employers are able to control their workers’ actions down to the second. Although workers are therefore the classic candidate for democratic rights under the all-coerced principle, other groups could also count as coerced if one were to employ a wider definition of direct and preemptive deprivation. Think of customers and competitors, for example: customers lose the ability to purchase something if prices increase. Direct competitors might lose the ability to sell a product when a corporation brings a similar offering into the market at a lower price.

For other groups, it seems much less plausible that they could count as coerced. Suppliers, for example, might suffer from lock-in effects toward a corporation, but it is difficult to see how this would constitute direct and preemptive coercion. It evidently puts the supplier in a vulnerable position and makes them more dependent on the corporation, but the corporation does not seem to preemptively deprive the supplier of any options. It is similarly difficult to conceptualize the effect of emissions via their contribution to climate change as direct coercion, especially where it is unclear what precisely it is that a specific corporation causes with their emissions as opposed to the things that the sum of emissions causes. So, the all-coerced principle excludes cases of the abuse of vulnerable positions and cases of substantial, but noncoercive, effects. This limitation might be counterintuitive for some, but it is a straightforward result of the normative core of the all-coerced principle. This core consists of a commitment to the value of autonomy and the conviction that any infringement upon it must be rectified by giving democratic rights to those coerced (Abizadeh 2008, 39).

There is, however, an additional problem with the all-coerced principle. This problem is independent from one’s precise definition of coercion and one’s intuition about whether workers, customers, or competitors can plausibly count as coerced. In its original formulation, the principle cites single, concrete actions as the determinant of enfranchisement. But, as we have established above, we cannot determine which single, concrete actions a democratic body will pursue before we have determined a demos, because that is a task of precisely that demos. If we want to avoid this undue predetermination, we will have to shift our view toward hypothetical coercive actions or threats. But how can we grasp all the actions that a corporation might undertake that might be coercive? For the state, some have proposed to capture this sense of hypothetical coercion by using an all-subjected principle. Such a principle replaces concrete coercion with subjection to rules, because that likely includes a coercive threat for the case of deviation. For this methodological reason I will next consider the all-subjected principle as a more suitable reformulation of the all-coerced principle. ⁵

The All-Subjected Principle

In its original form, the all-subjected principle stipulates that “all persons subject to the laws of [a] state” (Dahl 1989, 122) should be part of its demos. Being subject to a law can be most plausibly specified as being legally liable in the case of disobeying that law (Goodin 2016, 370). As the original formulation is obviously meant to apply to states, I will make some adjustments for the case of the business corporation. The most plausible reading is to understand subjection in the case of the corporation as subjection to corporate rules and directives, because these come closest in character to the laws of a state. Corporate rules and directives are binding wherever there is a legal contract between some party and the corporation. Unlike in the case of the state, people without an explicit contract with the firm are not bound to obey by any commands of the corporate authority. Additionally, the contract has to be sufficiently incomplete to allow the corporation to subject me to its rules. A sales contract for example is so specific that it does not give the corporation any further room to subject its customer. The only legally enforceable claim the corporation has against them is for payment of the price agreed upon. A labor contract on the other hand is incomplete and gives the corporation ample room to direct me and subject me to their laws. This is necessarily so because of the nature of the work relationship. Whereas a sale is a short-term, one-off transaction, the work relationship is long term and littered with contingencies that are very hard to foresee and specify in the contract in advance (Hart 2016). ⁶

Just as with the all-coerced principle, it is thus clear that workers fulfill the condition of the principle and will have to be part of the demos. The work contract is the paradigmatic example of an incomplete contract that, in conjunction with respective labor law regulation, gives ample room for arbitrary direction to the employer. But in what sense could other stakeholders or the general public be subject to corporate authority? One case where subjection applies is that of the supplier. Supplier contracts can contain significant degrees of freedom for the customer corporation. Consider the following example. Suppliers for big car manufacturers often exclusively supply one of those corporations. This is mainly due to the high degree of specialization that is required. Suppliers thus have to make large investments and are locked into one customer, because a change would require them to invest again and render their former investment useless. The customer corporations on the other hand have much higher market power due to their size and can readily find another supplier. As a result, the supplier’s position resembles that of a department of the bigger corporation. Their contracts are incomplete, and they have to be able to rapidly adjust quantity and specifications of their product under the credible threat of termination because of the market power of their customers. The situation of supplier firms in these cases is not unlike the situation of gig workers and other dependent freelancers. Formally, they are not contracted as parts of the corporation, but the corporation’s control over them is as big as it would be were they fully integrated. Corporations simply avoid fully integrating them, because that would involve additional legal and financial duties.

What about other constituencies? For local communities, there is no sense in which they could be subject to the rules of a corporate authority as there are no incomplete contracts between the two. There might be rental contracts or the like, but the influence that a business corporation can gain over a community does not stem from contract. It stems from its economic power. Competitors are not contracted to the corporation either, they are only connected to it via the market mechanism and the effect that the corporation’s actions have on demand, prices, etc. Customers do enter into contracts with corporations, but they too lack the incompleteness needed to become subjects.

The all-subjected principle excludes these cases, because it is blind to a specific type of relation (Goodin 2007, 49). Many rules have profound effects on people who are not formally bound by them. Imagine living in close proximity to a steel factory. If that steel factory decides to increase ventilation by working with all gates permanently open, this seriously affects the local community surrounding it in terms of noise. The same goes for any decision that a corporation makes regarding emissions. The all-subjected principle risks being underinclusive for cases in which spillover effects arising from corporate conduct are so severe that it seems intuitively implausible not to include those affected by such spillover effects in decision-making just because they are not bound by an explicit rule, even though they are profoundly affected. One could argue that doing so would constitute letting ourselves be blinded by formal legal configurations.

Proponents of the all-subjected principle might reply that this is simply the straightforward application of the principle that ties democratic rights to subjection and not to affectedness, however significantly one might be affected. For corporations specifically, however, subjection with its focus on formal authority fails to capture what are widely considered problematic instances of corporate power. By their nature and unlike states, business corporations cannot exercise formal authority over anyone who does not stand in a formal contractual relationship with them. Although states will, depending on their laws, hold anyone accountable who commits a crime on their territory, against their citizens, or tries to impair their national security (Goodin 2016, 375–82), there is no such catch-all clause for corporate rules. They do not need formal authority to exercise power. Although no German car manufacturer has formal power over any part of the German government, they still succeed in their lobbying attempts to block stronger emission regulation on EU level. Although Apple has no formal authority over the Irish government, they were nevertheless able to secure an enormous tax break that allows them to only pay marginal taxes on their entire EU business. Although large banks did not have formal authority over the design of the global economy, their behavior brought about the global financial crisis and thereby profoundly affected the lives of millions of people.

These examples illustrate the core difficulty with business corporations: while most exercises of state power are captured by the subjection relation because of the basic formal relation between states and individuals, corporations have no such basic relation but nevertheless exercise power over large numbers of people. To the extent then that we want the basic principle for determining the demos to track important differences in the degrees to which power is exercised over others, the all-subjected principle fails for the case of the corporation. Now, if we do not intend such a principle to track severe exercises of power, the all-subjected principle might be an adequate solution. But, without delving further into this debate, I suggest that tracking exercises of power is more in line with the general thrust of both the workplace democracy debate and the general public debate about corporations than the alternative. Both debates seem chiefly concerned with reigning in unchecked exercises of corporate power.

This constellation also distinguishes business corporations from citizens. Among individuals we mostly lack formal decision-making power over one another but are able to significantly affect each other. However, there are two important differences between individual citizens and corporations. First, individuals are subjects with autonomy rights. Their actions are exercises of these rights and, up until they significantly infringe upon others’ rights, protected thusly. For collective agents like corporations, however, several authors have convincingly argued that they do not have such autonomy rights (e.g., List and Pettit 2011, chap. 8). Corporate acts thus become normatively problematic much earlier than individual acts, because there is no initial autonomy right to justify them. Second, corporations are in a much better position to exercise influence because of their structure. Limited liability and entity shielding lower capital costs for corporations, making it much easier to acquire capital and thereby economic power. As legal persons, corporations also live much longer lives than individual citizens, making it possible to keep and expand capital and influence over centuries without death getting in the way.

Therefore, although it might be justified to limit democratic rights to cases of formal subjection in other cases, the absence of a basic formal relation compared to the state and the absence of autonomy rights with much larger material means when compared to individuals warrant a turn away from the all-subjected principle for the case of the business corporation. To be clear, I do not reject the principle because it does not recommend including some constituency that I intuitively think should be included. My concern is with the fact that the subjection relation fails to capture what is threatening about corporations specifically. It uses a distinction between subjection and other kinds of affectedness that might be important in the case of the state but is much less meaningful with the corporation. To use that somewhat arbitrary distinction as the baseline for determining the extent of the demos is therefore suspect, independent from the substantial question of who fulfills this relation.

In search of a principle that is better suited to accommodate what is distinctive about the corporation, I will now turn to the most expansive principle for the determination of the demos, the all-affected principle. ⁷

The All-Affected Principle

The all-affected principle was the first to be proposed in the modern debate of the boundary problem (Whelan 1983). It stipulates that all those who are affected by a political decision should have a say in making it. For the institution of the business corporation, we will thus have to consider who will be affected by the kinds of decisions that corporations usually take, decisions on production and distribution but also internal structuring and external activities like marketing. It is clear that a great many people are affected by such decisions. Corporations have a profound impact on everybody working for them. They also have profound impacts on their local community, offering employment and tax revenue but also altering neighborhoods and exerting political influence. Moreover, corporations affect all their customers and competitors via their decisions on which products they offer at which price. This is especially true under the conditions of globalized markets where economic relations often extend across the globe. On top of that, corporate decisions about energy sourcing affect the environment and thereby affect every person on the planet.

What the all-affected principle thus has to recommend is a profound extension of the corporate demos all over the globe. Even though the principle should arguably include a clause that limits it to people that are affected in some relevant sense so that not all minuscule effects trigger democratic rights, the demos it proposes would still be enormous. After all, as I have stipulated previously, it is the whole range of potential decisions that a corporation might take that we have to consider for the determination of the demos. For bigger business corporations that can be said to relevantly affect people all over the world via their emissions, for example, the demos will have to be extended to include all people worldwide. This result is precisely what the other two principles tried to avoid. By only granting democratic rights for cases in which people are affected to the extent that it constitutes coercion or subjection, they aimed to limit the demos to a more manageable scope. However, although they might succeed in doing so for the case of the state, I argued that both failed to grasp the specifics of the corporation.

Corporate democracy would thus have to include vast amounts of people. This seems profoundly impractical and therefore a result not more plausible than the other two. How could we ever be able to have global democratic decision-making fora for every corporation? Here, however, we have to distinguish between democracy as an ideal and democracy as a decision-making procedure (Arrhenius 2019). Democracy can describe an ideal about the institutional setup of decision-making to which we can try to approximate, just as utilitarianism or Kantianism describe ideals for ethical decision-making to which we can approximate. It might well be practically impossible to always follow these ideals, but we appreciate the guideline they provide. Alternatively, democracy can describe an actual institutional setup for democratic decision-making, just as monarchy or aristocracy. All three principles I have considered aim to explicate the normative ideal, not an institutional setup. Their results will surely affect the institutional setup we derive from the ideal, but we will have to find a suitable trade-off between feasibility and adherence to the principle. The practical difficulties the unbounded demos raises are thus no direct argument against the plausibility of the ideal result, they are considerations we have to take into account when translating the ideal into practice.

But there are further objections to the demos that the all-affected principle recommends. A second one centers around possible undue decision-making power. The all-affected principle demands that we define the demos by considering all possible decisions a corporation could ever make and then give all those a voice who would be affected by that decision. If in the operation of the corporation we then take a vote on a specific decision, it might be the case that some individuals get a say whose interests are not affected by that specific decision. Their ability to co-determine would then amount to an undue exercise of power over some of the other constituencies. As an example, think of a car manufacturer taking a democratic decision on shift schedules in a Hungarian factory. Why should the non-Hungarian workers be able to determine this for the Hungarian workers?

What could such an undue determination look like (Goodin 2007)? An undue overpowering of the Hungarian workers would happen when a considerable amount of those not affected voted for one option that is not the preference of those affected. But, assuming somewhat rational decision-making, one would speculate that those not affected by the decision would either abstain from voting altogether or distribute their vote randomly over all options. After all, if they had any preference for any of the options, this would constitute an interest and they would thus be affected by the outcome and therefore legitimate decision-makers. Random decision-making or abstaining from voting would however be unlikely to drown out anyone. Therefore, the voters that are not Hungarian workers are either justified in co-determining the shift schedule because they have an interest, or they will most likely not distort the vote. The only case where the wide allocation of voting rights might be problematic is when the number of those actually affected is so small that they might be drowned out by the variance inherent to the random allocation of votes by the majority. I thus submit that this complaint has almost no practical instantiation.

Regarding the exercise of unjustified power, one also has to keep in mind that an unbounded demos is not unbounded in its decision-making. Basic democratic rights still hold, and they limit the decisions that the corporate demos can make. Additionally, the initially unbounded demos might decide to limit its own powers. It is entirely compatible with the all-affected principle to, for example, introduce further conditions that one has to fulfill to be able to vote on a specific decision. It merely demands that such limitations must be decided upon by all those affected. A corporation could, for example, limit decisions on internal procedures in one specific country to all employees in that country, as long as this decision is made democratically and not simply assumed from the outset.

Another institutional setup that would avoid overinclusiveness complaints is to have a separate demos for each separate decision by redetermining it based on the scope of that decision. For the previous example, we would thus limit the franchise to Hungarian workers only. Although this approach can alleviate some of the practical problems, it also has an important limitation. We will still have to somehow determine what decisions are to be taken and will have to do so in conjunction with anybody that could possibly be affected by any decision that could be taken. While some of the decisions could thus be taken with a more limited demos, the overall agenda setting will still have to be undertaken by the unbounded demos.

Overall, then, the objections about the size of the demos either misunderstand the objective of the principle or they rely upon a scenario that is not realizable in practice. At the same time, the all-affected principle avoids the kinds of objections that I raised against the other two principles. It does not rely on the consequences of single actions, and it is able to capture the problematic instances of noncontractual exercises of power that most worry us when it comes to corporations. Thus, my discussion shows that in the realm of the ideal norm, the all-affected principle can deliver a convincing result for the corporate demos.

However, this leaves open a wide array of practical questions about how to realize such a demos in a decision-making structure. On these practical questions, let me discuss two strategies that Goodin has proposed for limiting the demos of states to avoid the practical inconveniences that an unbounded demos brings with it (Goodin 2007, 62). The first strategy suggests limiting the decision-making power of existing states to their existing demos so that no democratic right arises for those that are not currently members of the demos. Whatever the prospect of this strategy for the state, it is clearly insufficient for the business corporation. There is no current demos to which activities could be limited. Even if we agreed on some definition of who is already a member of the corporation (Ciepley 2017), it would be impossible to limit the activities of the corporation to them. By their very nature as commercial entities, corporations have to extend their influence over others through the market and otherwise. The only solution here would be the abolishment of the corporation as a legal institution and exclusive reliance on natural persons for the organization of exchange. As mentioned previously, their activities would not trigger the same standards of justification as they have less means at their disposal and enjoy some baseline autonomy over their actions.

Another proposal to improve the practicality of the all-affected principle is the compensation of those that are affected but cannot be granted a voice. For any infringement upon their interests, individuals would be granted monetary compensation. Such a scheme could be relatively straightforwardly applied to the business corporation. However, it raises difficult questions about the monetary value of any particular infringement. For example, if an EU citizen has an interest in strong regulation on car emissions, how much compensation are they owed by German car manufacturers who have successfully lobbied against such regulation? If my favorite breakfast cereal is taken off the shelves by its manufacturers, should they compensate me for that? Furthermore, the scheme raises some basic questions about the possibility of compensation for rights infringements. If we favor democratization for noninstrumental reasons—that is, if we believe that those affected have a right to power and voice over their fate—what is the monetary value of that right? Nevertheless, in a full scheme of democratic governance, the compensation of minor affected interests might be a practical improvement upon the democratic legitimacy of corporate conduct by somewhat offsetting the harm done. Another possibility for the practical realization of this principle is to rely on regulation where democracy becomes impractical. This could, for example, take the form of the scheme mentioned previously where the threshold for democracy is subjection in a contractual relationship and regulation is supposed to protect those in other relations with the corporation.

These additional arguments support the plausibility of the application of the all-affected principle by showing ways to reign in its wide scope. Let me now move on to the wider consequences of my result for debates around corporate reform. In the next and final section, I will point out what I take to be the two main points that my argument raises. These are, first, the requirement of an extension of the workplace democracy debate beyond workers, and, second, the proposed suspension of the market mechanism.

Implications

Extending the Workplace Democracy Debate

The first central implication of my argument is that the workplace democracy debate should not unduly focus on workers when debating the democratic organization of workplaces. Any attempt at organizing work more democratically will necessarily involve making decisions that have significant effects on many other constituencies. For these decisions to be democratic, we have to involve these constituencies.

I do not doubt that the work relation gives workers an extensive right to democratic participation. However, workers would only be the exclusive decision-makers in a democratic corporation if they enjoyed a special claim to have a say and other constituencies could not bring forward any legitimate claims (Moriarty 2010). For example, if we grounded a call for workplace democracy exclusively on a right to meaningful work that is enforceable only when workplaces are democratized, workplace democrats could be content with a democratic decision-making scheme that only included workers. Similarly, if the right to a democratic say is grounded exclusively on the alienating effects of labor under capitalism, workers would be the only constituency we would have to involve. Insofar as the considerations for workplace democracy rely on the reasons familiar from general democratic theory, however, other constituencies also have a right to democratic participation, because they are also affected by corporate decision-making. In the workplace democracy debate, nothing suggests an exclusive focus on any special status of workers. Rather, arguments mostly proceed from a mixture of different normative grounds and take workers to be the paradigmatic constituency that should be enfranchised (Frega, Herzog, and Neuhäuser 2019). Their case is simply the strongest and most obvious, because their lives are intricately bound up with the business corporation in a number of ways. They have important relations with their colleagues, their work forms an important part of their identity, and they might wholly depend on the wages from their jobs. Here, it is the sum of these forms of affectedness that grounds their democratic rights, not their special status as workers. In principle, any other group that had a similar degree of affectedness would have the same rights to democratic self-determination. Let me elaborate upon this point along the lines of one way to normatively ground workplace democracy.

The most prevalent normative grounds for arguments about workplace democracy are relational-egalitarian accounts that share some convictions about how relationships between persons should be structured—namely in an egalitarian manner. In these accounts, democratic rights are taken to be an antidote to the unequal relationships that develop in traditional, hierarchical corporate structures. Although the precise structure of these arguments differs, all would have to justify an exclusive focus on workers in a similar manner. They would have to show that workers face normatively deficient kinds of relationships in hierarchical firms and that they are the only constituency that faces this kind of relationship. As I have outlined earlier, the extent to which other constituencies stand in unequal relationships with corporate officials is considerable. For example, compare two cases: on the one hand, a laboratory technician at a large chemical corporation. They have a superior that assigns them projects while they have little say in what projects these will be. However, in the execution of their projects, the technician is largely free to organize their work. On the other hand, think of someone running a small metal machining operation that supplies parts to some company. Similar to the laboratory technician, they receive assignments but then can themselves determine how they will fulfil this task. In contrast to the lab technician, however, the supplier is under much greater economic pressure from their customer. As is often the case with supplier contracts, the client reserves the right to change or cancel orders on short notice. Although the labor contract for the technicians gives them some economic security and protection against the power of their employer, the supplier does not enjoy these privileges. To justify an exclusive focus on workers, relational egalitarian workplace democrats would have to show why the arguably less unequal relationship between the lab technician and their employer grounds self-determination rights while the more unequal relationship between the supplier and their customer corporation does not. Similar examples can be constructed for the other normative dimensions grounding workplace democracy.

Thus, arguments for workplace democracy should take into account other constituencies, as workers are not the only persons in unequal relationships with the business corporation. My argument has provided a further elaboration of this claim. Independent from the reasons we have for favoring democracy in the workplace, there are considerations internal to democracy that favor the inclusion of additional constituencies as well.

Suspending the Market Mechanism

The second implication of my argument is that a democratic organization of corporations demands the suspension of the market mechanism. Customers and competitors are clearly affected by production and pricing decisions, and they will thus, prima facie, have to be included in those decisions. Such inclusion would however undermine the conditions of competition. There would be no market of several independent sellers competing for customers, because they would all be democratic participants in each other’s decision-making for the market. Such conditions would be closer to a planned economy than market competition. ⁸

Here too the merits of democracy have to be weighed against other normative considerations. Regarding markets, this would be the idea that markets are, in some sense, more efficient than forms of democratic planning. There is a long debate whether socialist planning can achieve a grade of efficiency comparable to markets (Levy and Peart 2016). Recent developments in digital technologies have reignited that debate as they have considerably expanded our computational abilities (Daum and Nuss 2021). Against proponents of economic planning, different authors have proposed a variety of market socialist schemes that leave some features of markets intact to improve efficiency (e.g., Malleson 2014, part two). Here, I do not want to delve deeper into this debate. Instead, let me sketch one possible way to justify markets while upholding the democratic considerations presented.

A further proposal for closer approximating the democratic ideal regarding inclusion for states is the establishment of a global federal structure (Goodin 2007, 65). Under such a structure, we could delegate decisions depending on who would be affected by them. If a decision had effects on all people worldwide, we would submit it to a global democratic forum while smaller-scale decisions could be made within nation-states or similar institutions. For economic issues arising from the conduct of business corporations, we could imagine a similar scheme where decisions with wide effects are discussed in a global institution while small-scale decisions are to be made in separated corporations. A global market could be part of such a federal scheme if we decided democratically in the global democratic forum to yield decision-making on the relevant decisions to the small-scale corporations for efficiency reasons. To give an oversimplified illustration, if all affected by the breakfast cereal production, consumer, suppliers, competitors, and workers came together in the global forum and decided to set up a market for breakfast cereals and allow some number of breakfast cereal producers to compete in that market by making individual decisions, this would be a democratically legitimate decision possibly leading to market competition between democratically organized corporations.

This argument does not, however, justify the status quo of economic relations. Markets affect large portions of our lives, but a global democratic institution is nowhere in sight. We thus lack the forum to make the justifying decision that the argument from workplace democracy would require. Dissenting from the conventional top-down view that markets are justified because they are (ideally) organized through democratic states, this bottom-up perspective of democratic theory argues that the absence of a global democratic institution means that neither the rejection nor the acceptance of the market mechanism at its current scale are justified. Given the chronic disagreements about the merits of the market versus planning, it also seems premature to simply assume that if we had such an institution everybody would agree to market organization. Until such a democratic procedure comes about, my argument shows that the power that business corporations exercise over others through global markets is deeply undemocratic, because those affected do not get a say in what is being done to them.

Conclusion

I will now conclude my essay by summarizing the steps of my argument, relating it to another recent approach, and pointing out where further research is needed. My argument started with an assessment of the extent to which democratic considerations have so far influenced the definition of the corporate demos in which I showed that neither the debate in law and economics nor the workplace democracy debate have much to say about this. Consequently, I turned to the discussion of the boundary problem in general democratic theory where I took three principles and applied them to the business corporation. Two delivered underinclusive results for the special case of the corporation while the third recommended a vast unbounded demos. I then argued that overinclusiveness is a much smaller problem than it initially seems and that the all-affected principle is therefore the most plausible choice. Finally, I articulated the two core consequences of my argument. Firstly, the workplace democracy debate has to extend its focus beyond workers. Secondly, a discussion of the boundary of workplace democracy provides an argument for the suspension of the market mechanism.

A similar argument regarding the suspension of the market mechanism has recently been made by Nicholas Vrousalis, arguing that the two most common normative justifications for workplace democracy also demand economic democracy (Vrousalis 2019). My argument proceeds from a different angle and also urges the extension of the workplace democracy debate to further constituencies. Instead of engaging with the substantive arguments for workplace democracy, I consider the demos question and derive an argument for the suspension of the market mechanism from that. My argument does not rely on any further premises about the intricacies of domination and exploitation. It supposes only that our concern for the democratization of corporations is not based on completely different normative grounds than our concern with the democratization of states.

As I have stressed several times in the course of my argument, this is only a pro tanto consideration of the extent of the corporate demos. A realistic proposal for institutional design will have to weigh these considerations against other normative criteria to determine what is a reasonable extent of the corporate demos in practice. Here, further research into the full nexus of normative considerations seems warranted. Another avenue for future research would be a discussion of what steps should be taken toward the realization of these normative considerations. What are the short-term institutional reforms that can bring us further toward democratic corporations?