Property as power: A theory of representation

Inequalities in wealth are a major concern in societal debates today (Atkinson, 2015; Piketty, 2014). Economic analyses abound showing that wealthy individuals and corporations can achieve commanding heights in the economy. They dominate more vulnerable parties, such as workers, consumers, members of less developed countries, the natural environment, and democracy itself (Christophers, 2019; Eeckhout, 2021; Mazzucato, 2018). How to respond to private owners imposing widespread negative effects on non-owners? Liberal-democratic theory and practice responds to this problem of ownership power by subjecting owners to the power of the state. The state apparatus is called upon to regulate owners in the public interest. Through state regulation imposed on owners, the interests of non-owners are represented indirectly. I call this "the regulation model." The main claim of this paper is that liberal-democratic theory and practice should end its exclusive reliance on the regulation model and expand its toolkit of representation with an "incorporation model." Under this model, non-owner interests are represented directly, by those who are owners, when they make decisions about how to use their assets. In most cases, this requires that property is incorporated, that is, transferred to a corporate legal structure, in which owners are held to account by non-owners. Representation is a key notion in political philosophy, expressing the demand that political rulers should rule in the interest of citizens, who can hold them to account for doing so. The present paper offers a systematic application of this concept to the powers granted with ownership. After an introduction of the regulation and incorporation models (Section 2), the first step is to identify the non-owners whose interests merit representation. Here I argue for a representation imperative: when the purposes determined by owners lead to structural effects on third parties, then these have a claim to be represented. This is an application of the all-affected principle, similar to the principle's application in the context of the boundary problem (Section 3). In the second step, I argue for an incorporation imperative: where non-owners are affected and state regulation is unable to represent their interests, then an incorporation structure should be imposed, through which non-owner interests are represented (Section 4). As a third step, I argue for a power-sharing imperative: to organize accountability, owners and non-owners should share power within the incorporated structure. Owners should not operate as independent trustees who can decide how to represent non-owners without any checks and balances; but nor should they be mere delegates of non-owners (Section 5). In Section 6, I respond to an egalitarian objection to my proposal, that is, that redistribution, not representation, is needed to tame ownership power. Finally, I reflect on some further questions about the relation between the incorporation and regulation models (Section 7). In this section, I contrast two models for taming ownership power; the regulation model and the incorporation model. The regulation model is the standard model in liberal-democratic theory and practice. I argue that it has its shortcomings, which leads to a regulation gap. The incorporation model can help representing the interests of non-owners whenever there is such a gap. The core of "liberal-democratic theory," as I use the term in this paper, is that a legitimate political relation is a democratic-representative one. This has become the standard view in political thought and practice.1 I will follow Urbinati and Warren in their definition of democratic representation. Democratic representers "(a) must be authorized to act; (b), they must act in a way that promotes the interests of the represented; and (c) people must have the means to hold their representatives accountable for their actions" (Urbinati & Warren, 2008, p. 393). Under this definition, representation is summarized by the first two conditions. Representation becomes democratic through the addition of the final condition. In the political context of the relation between rulers (persons holding political power, like government officials and parliamentarians) and citizens, representation is a solution to the exercise of arbitrary power. Without representation, rulers can negatively affect the lives of their citizens without the latter being able to prevent this or object to it. The solution in liberal-democratic theory is to establish a relation of representation. Rulers are to rule on behalf of their citizens. This changes their merely causal power into a normative power, based on law. The relation between rulers and citizens becomes a reciprocal one. Liberal-democratic theory also has implications for thinking about property. Owners receive their normative powers from the state, which legislates and enforces property laws. These powers enable owners to "rule over" the objects over which they have an ownership claim. However, given that this is an enforceable claim, ownership power also shapes the relation between an owner and all others, who are, with respect to the same object, in the position of non-owners. Non-owners have a duty to respect the owner's right over their things. State empowerment of owners vis-à-vis non-owners hence also gives owners causal power over non-owners.2 The owner of the only water well in a desert area can determine whether the thirsty person lives or dies. Owners of crucial resources have potentially vast powers over the fate of non-owners. Through the management of resources, owners can shape the lives of countless citizens. This is potentially as much an exercise of arbitrary power, as the arbitrary power of political rulers (tyrants and dictators). Should it also be turned into a normative power, by being subjected to effective demands of representation by citizens? In the liberal-democratic model, the only way to ensure that ownership power is not exercised arbitrarily, is by subjecting it to regulation (see the left-hand side of Figure 1).3 Non-owners are deemed to mobilize their representative relation to political rulers (the state) to press their demands. Owners can do the same. The democratic state then arbitrates these demands and formulates the public interest in question. Where non-owners' claims against owners are deemed to outweigh owners' claims, the state regulates the owners' exercise of their property rights. For example, states subject owners to taxation laws, laws enabling expropriation, regulations of access (e.g., to the beach), zoning laws, intellectual property rights law, etc. These measures impose duties on owners to take the interests of non-owners into account. In the liberal-democratic model, owners have these duties to the state, which acts on behalf of the non-owners. The latter are not themselves representatives. Representation of non-owners' interests hence happens indirectly, via the state.4 Kantian property theory is an example of this model, where an omnilateral will represents the interests of all, and property is regulated by the state through the omnilateral will (Kant, 1996; Ripstein, 2017).5 Arguably, however, the support for this division of labor between public and private actors is much broader, and philosophically deeply baked into liberalism.6 The regulation model is only one of two ways of dealing with the problem of ownership power. This paper wants to highlight an alternative, which I will call the incorporation model. This model is equally legitimate from the point of view of liberal-democratic theory's basic assumptions about the legitimacy of power, as stated above, since it only requires that power must be subjected to a demand of representation but should be flexible as to how this must be done. Now my argument in the following is not that the incorporation model should replace the regulation model, but that it should supplement it, enhancing the toolkit of liberal-democratic theory and practice. To motivate this project, we need to remind ourselves of the shortcomings of the regulation model. Waheed Hussain, in his recent investigation of the legitimacy of the market mechanism in liberal democracies, highlights several shortcomings of state regulation (Hussain, 2023, pp. 124–133). One shortcoming is epistemic: regulators often have great difficulty knowing what happens on the ground. They are put at an epistemic distance with respect to the production processes and actors they are deemed to regulate. Collecting and processing information is costly and time-consuming. The regulatees can make strategic use of this distance, and actively withhold information for the regulators. A second shortcoming has to do with the pace of innovation in markets. Regulators always lag behind the latest developments and their regulatory frameworks therefore do not cover the newest production processes and techniques. Moreover, regulatees can respond to regulations by exploiting loopholes which then take precious time to close, after which the cat-and-mouse game enters a new cycle. Both of these shortcomings are located at the "regulation" arrow in Figure 1. Hussain mentions a third shortcoming, which he refers to as "obstacles to political mobilization" (ibid., p. 128). Ordinary citizens often have great difficulties in influencing political actors and building coalitions to change state policies in their interests. This shortcoming is located at the "representation" arrow in Figure 1. But this problem should be seen in conjunction with another problem, that is, state capture. In his recent discussion of state capture, Samuel Bagg argues on the basis of a wealth of empirical literature that concentrated private interests are able to set political agendas to favor their own interests, to the detriment of more dispersed, poorer constituencies. In its most extreme manifestations, this leads to corruption and the establishment of façade-democracies with are better characterized as oligarchies (Bagg, 2024, pp. 83–88). Hence it is not only that wider groups of citizens (or non-owners) face obstacles in organizing collective action but also that elites (owners of concentrated wealth) actively hinder this, through their privileged access to the state's regulatory capacity. Given these shortcomings in regulating ownership power indirectly, via the state, it seems that there will always be a "regulation gap," a residue of untamed ownership powers, exercised arbitrarily over non-owners without effective normative control through state regulation.7 The incorporation model provides an alternative which can fill this gap. Its basic idea is that the relation of representation between affected non-owners and owners can also be organized directly, at the level of the exercise of ownership rights. In this model, owners become representers, rulers subject to demands of representation vis-à-vis non-owners. The dynamic of decision-making (about how to use certain assets/resources) and accountability for these decisions, then takes place at a "lower" level. The role of the state is here more in the background. It can enable, facilitate or mandate such a representative structure; but it doesn't itself formulate the substantive duties to which owners' decisions become subject, as it would when it regulates (see the right-hand side of Figure 1). The attractions of the incorporation model are the flip side of the shortcomings of the regulation model. Since owners are more aware of relevant information than external bodies (such as state agencies) the problem of epistemic distance is largely absent. The problem of temporal lag between innovation and regulation is also absent when representation is built into the owners' decisions directly; for owners must take into account the repercussions on non-owners at the same time when they decide to introduce innovations. Finally, one would expect collective action by non-owners to be easier when they get direct standing vis-à-vis owners, since their interests are then less dispersed. If representation works well at the level of ownership, then it would even turn the owner into a true representative of the non-owners' interests, hence change the owners' self-centered motivation and identity, which leads to state capture under the regulation model.8 Of course, representation at the level of ownership may also have shortcomings of its own. The aim of this paper is therefore not to show that the incorporation model is superior to the regulation model. Rather the aim is to question the current state of liberal-democratic property theory, with its exclusive reliance on the regulation model, and offer a picture in which both models are equally plausible from point of view of the demand for representation that is at the heart of the liberal-democratic tradition.9 If my argument is successful, then representation at the level of ownership gains the same intuitive appeal as provincial or municipal democracy, or democratic elections for fire brigades or water boards. The fact that the state is organized democratically, does not take away the soundness of also organizing representation relations at these lower-level organizations that rule particular geographical areas or execute certain functional tasks. In the private realm, where owners accumulate vast powers over non-owners, a similar logic may apply. Both models can supplement each other so that the current representation gap resulting from regulatory failures is filled. To develop the case for including the incorporation model in a theory of representation, I proceed in three steps. The first step, in this section, is a determination of which non-owner interests ought to be represented. This step is common to the regulation model and the incorporation model: for the operation of both models a determination of relevant non-owner interests is needed. The next two steps (in Sections 4 and 5) will be particular to the incorporation model. For this first step, it is helpful to apply the theoretical tools developed in the context of discussions of the boundary problem. Normally the boundary problem refers to the problematic legitimation of the state. The state can only be legitimately authorized by its citizens, but prior to the act of authorization, how to non-arbitrarily determine who should belong to the authorizing citizenry? Pointing to the actual citizen body on a given territory at a given point in time seems arbitrary, since this excludes outsiders (refugees, animals, future generations, etc.) who however may be coerced, subjected or affected by state actions without having had a say. We need a more principled determination about who should be included (Näsström, 2007; Whelan, 1983). The property-theoretic analogue to the boundary problem is: which non-owners interests need inclusion (either indirectly, through the regulation model, or directly, through the incorporation model)? To answer this question, we now need to apply one of the alternative principles from the boundary problem literature: the all-coerced, all-subjected, and all-affected principles (Abizadeh, 2012; Goodin, 2007; Näsström, 2011). The all-affected principle has been favored by many on the grounds that the other two principles have problematically underinclusive tendencies. This was argued recently by Philipp Stehr in a plea for adopting the all-affected principle with respect to decision-making in business corporations. While only workers are subjected, suppliers, future generations, consumers, etc. may be relevantly affected by the decisions of corporate owners (Stehr, 2022). I agree, and propose that this is true for property generally, not just for property held in business corporations. Let's define the representation imperative for property as follows: if and to the extent that an owner gives property a purpose which structurally affects the interests of others, they have a claim to have their interests represented. Instead of the old rallying cry "no taxation without representation," we would arrive at "no property without representation," or more precisely: "no property without a test for checking whether representation of non-owner interests is necessary." Under such an imperative, a lot of property will not have important effects on non-owning others. As the formulation of "a test for checking" makes clear, we are not envisaging a situation in which each and every exercise of a property right becomes subject to a demand to include representation mechanisms towards non-owners. This raises the question where to put the boundary line. The wording of the representation imperative above, with its reference to "structurally" affecting the interests of non-owners, is meant to exclude incidental and trivial effects on others. Intuitively, many consumption items, from furniture to food, from pots and pans to pencils, would be exempted, while property in the means of production would be the main target. This is also the dividing line that theorists have often taken—think of Rawls' restriction of the right to property under his first principle of justice to "personal property" (Rawls, 1999, p. 53).10 We can indeed take this as a heuristic, but we should remember that on principle, this dividing line is too coarse grained. A rich person using their private jet in acts of consumption may structurally affect the neighbors' opportunities for undisturbed living, causing severe nuisance. Some means of production (think of the carpenter's tools) won't structurally affect the lives of others. Rather than a distinction between types of property which draws the boundary in the ontology of goods, then, a normative theory would be needed to spell out which non-owner interests are so weighty that when they are "structurally affected" a requirement for representation would kick in. This is not just a feature of the incorporation model defended in this paper, but it also applies to the familiar theories of regulation—both types of representation need such a normative theory. In the context of regulation, for example, economic theories of market failure have often been called upon to fulfill this function. They draw attention to certain problems (externalities, information problems) which would justify regulation. Other normative theories (focusing on justice, or exploitation) can draw wider or narrower boundaries. I will not here venture into this rich terrain, since whatever theory is used, it can be plugged in at this point in the argument. The contribution here is to argue that whenever such a successful normative argument applies, we have two options for representation—regulation and incorporation—not just one. Whichever normative theory is used, it would have to be applied on a case-by-case basis to particular ownership structures. Without doing this full analysis, it is not hard to identify some "usual suspects," that is, groups that will often be implicated as affected in the exercise of ownership rights, especially in the context of production. Workers are often exploited, despite formal protections of the labor contract. They may accept working hours and wages under duress, given their precarious bargaining position vis-à-vis the owners of the means of production (Anderson, 2017). Where owners exploit natural resources without adequate compensation, they affect the lives of peoples depending on these resources; from local communities suffering from pollution to future generations affected by climate change (Wright & Nyberg, 2015). Consumers will also sometimes be a structurally affected group, especially where the products they buy harm them, for example by making them addicted (Akerlof & Shiller, 2015), or when they depend on producers for necessities such as live-saving drugs. Each of these—and sometimes other—groups could become the subject of a representative claim towards owners. Note that the application of the boundary problem here implies a rejection of state socialism. In discussions of the political boundary problem, some argue for accepting current national boundaries, and then ask about the inclusion of constituencies one by one for different decisions (or decision areas). But others want to have a global demos make the decisions about state boundaries at the first stage (Agné, 2010; Bartelson, 2008). Applied to property, this latter strategy would mean that the state decides the purposes for which all property is used (central planning), canceling the idea of private property altogether. The theory of representation developed here takes the first strategy. It can accept that the state sets apart under a regime of public property any assets that it deems best owned publicly. The theory is agnostic as to the size of this publicly owned domain relative to all resources in the economy. But it applies the all-affected principle to all other assets, held privately.11 For these assets private owners get the right to private initiative: to set a purpose for the objects of their property rights. For different types of use, different sets of people will be affected by the owner's use of their goods. The application of the all-affected principle accepts these purposes set by private owners, and then seeks to identify the non-owners affected, given these purposes. Finally, note that the representation imperative by itself doesn't yet dictate whether representation should take place through the regulation model or the incorporation model. Consider David Owen's objection to the all-affected problem. In his view, being affected does not establish a right to become a full member of the polity; it only requires "impartial treatment" of one's interests. This right could also be guaranteed when there is a "form of legal governance" in place that requires the affecter to exercise "impartial governance" towards the affected. He specifies: "This could be accomplished through the subjection of the polity to either a legal order or to a second-order polity which has the authority to regulate the actions and choices of the first-order polity in order to secure the impartial treatment of the legitimate interests of all affected." (Owen, 2012, p. 137; similarly Saunders, 2011, p. 290). The analogy to the property problem is striking. Under Owen's arrangement, the second-order polity (e.g., an International Public Organization with coercive powers) enforces representation of the affected in the legal order of the nation-state, just as—in the regulation model—the nation-state represents non-owners by regulating owners' behavior. In both cases this is an alternative to direct representation; in Owen's example for the political context, inclusion as member of the nation-state; in the property context, inclusion of non-owners through incorporation at the level of ownership. Objections such as Owen's, however, do not provide a reason for rejecting the all-affected principle as a principle for the identification of relevant interests. They merely warn us to keep an open eye to what the best way is to organize the representation of these interests. In some cases, the interests of affected third parties are best dealt with through state regulation of owners' activities, while in other cases the representative burden is best discharged at the level of ownership itself. How much of each model is appropriate is a complicated question, which needs an answer finetuned to various types and objects of property. It is unlikely, however, that state regulation will always perfectly promote the legitimate interests of non-owners. To the extent that it does not, and a regulation gap ensues (see Section 2), the incorporation model becomes attractive. What does the representation imperative require in practice? In the previous section, I defined it as a claim by non-owners imposed on owners. This can be understood as a moral claim. But when the affecting is structural and severe, a legal regime enforcing the claim will be necessary. What would that imply? Here too, let's start from representation theory for the political sphere. As Mónica Brito Vieira and David Runciman argue, the concept of representation is closely linked to the concept of a corporation: "It was only when the concept of representation was brought together with another legal idea, that of the corporation, that its political potential began to be realized." (Brito Vieira & Runciman, 2008, p. 10).12 A set of persons chooses a common representative when they want to form a polity. A representative structure creates a community. These representative communities since the Middle Ages were often referred to as a corporation (universitas). For individuals to become a political group is to incorporate, to become a unity in one corporate body. This body was sometimes referred to as a "fictive person" (persona ficta) or "representative person" (persona repraesentata), that was accepted as a person in law to act on behalf of its members. The concept of the corporate legal person was first used in relation to collectivities such as guilds, universities, towns and monastic orders, but soon came to be used for states (commonwealths) as well (Ciepley, 2017; Tierney, 1982). The acts of representation of the body politic by the governing person or board within the corporation, are the glue that holds such corporations together.13 In the property context, the representation imperative implies that when relevant non-owners are structurally affected by an owner's power, and regulation cannot sufficiently protect their interests, then an incorporated structure should be created for these property-holdings, through which non-owners' interests are represented. Call this the incorporation imperative. For example, when a billionaire would affect the lives of countless others (as does Bill Gates when using his wealth to cure diseases) these others would deserve representation. Bill Gates would have to set up an incorporated structure, such as a non-profit foundation or a business corporation, whose rulers (officials, such as directors) have the duty to represent their interests. In the following, I will work with this notion of incorporation, which implies a formal legal structure. However, we can imagine that for some cases a lighter, informal form of representation would do. For example, individual landowners consulting on a regular basis with neighbors can "incorporate" their interests where relevant, without this leading to the creation of a formal legal structure of incorporation between them.14 Incorporation in the end is about the substantial process of interest representation, whether or not set apart in a legal form.15 A prescription of enforced incorporation may seem a highly provocative consequence of the theory of representation developed here.16 However, this is less of a revolutionary consequence in practice than one might think, given the fact that formal legal incorporation is already standard practice in the economy. Bill and Melinda Gates have already set up the Bill and Melinda Gates Foundation for this purpose. Almost all exercises of property rights that structurally affect the lives of non-owners already happen through incorporated structures: states and other public organizations, foundations, associations, trusts or business corporations. Hence this wealth is already subject to a representative structure.17 The real normative problem—from the perspective of the representation imperative—is that these structures are often only imperfectly representative of the interests of non-owners. Take a charitable foundation. A founder (A) authorizes an officer (B) to represent the foundation (fictive person, C), for the purpose of benefiting a (set of) beneficiaries (D). Or take a business corporation, in which a set of shareholders (A) authorize a board of directors (B) to represent the corporation (fictive person C), so as to make profits. Here too third parties (D) exist such as workers, consumers, the environment and many other "stakeholders" affected by corporate power. In both these cases, the affected parties (D)—in most jurisdictions today around the world—do not have a voice in the representative structure. Note that in these structures, the corporate entity (the fictive person) itself becomes the owner of the assets. The former owner is now either the founder who has given away their money or a shareholder who gets shareholding rights. The officials (directors) are not owners but "asset managers," managing their property on behalf of the shareholders (business corporation) or of the purpose set by the founder in the charter (charitable foundation) (Ciepley, 2020). The representation structures serve their interests, and their interests alone. This explains why Bill and Melinda Gates (and other wealthy individuals) readily use incorporation. It is an extension of their power. It better serves their interests than leaving their wealth unincorporated. In The Code of Capital Katharina Pistor explains how assets are coded by law, i.e. given certain attributes which extend the claims of their holders and protect them against others. How and for whose benefits these assets are coded, are core questions for the "political economy of capitalism." (Pistor, 2019, p. 4) In a chapter on land, she describes how the pre-industrial economy in the UK was revolutionized through the enclosure movement. One of the key legal inventions to protect the assets of landowners was the trust. The trust helped landowners to shield these assets from creditors, and remains a "favorite legal coding device among the wealthy who wish to protect their assets from tax authorities and other creditors" (Pistor, 2019, p. 43).18 Similarly, in a chapter on business corporations, Pistor describes how these are wealth-shielding devices for shareholders: the use of the legal corporate form "can be turne