1.1. The Structure of the Objection

Singer’s article provides a welcome assessment of recent liberal egalitarian considerations concerning corporate governance. In addition to examining the internal consistency of some of these arguments, Singer provides one generic and radical critique according to which all are necessarily inconsistent with the tenets of a political conception of justice. In his words, “attempts to formulate a position on corporate governance using the normative resources offered by Rawls are bound to fail” (2015: 75, emphasis added). This, Singer asserts, may explain why Rawls has been unable to say much about numerous important questions of corporate governance. Furthermore, it disqualifies all liberal egalitarian attempts at discussing these issues.

Singer’s assertion is notably supported by what may be labeled a basic structure objection against the consideration of corporate governance as a matter of justice, in reference to G. A. Cohen (Reference Cohen2008: 124). This objection is straightforward. It contends that the primary subject of justice is society’s basic structure, understood as society’s legally coercive institutions, and that the corporation is not part of this. The conclusion follows that corporate governance, as part of the “internal structure” of corporations (2015: 78), is beyond the reach of a political conception of justice. Thus, no form of corporate governance, including workplace democracy, can be mandated on the grounds of justice.

The basic structure objection follows three premises:

1. (1) The primary subject of justice is the basic structure of society;

2. (2) Only legally coercive institutions are part of the basic structure of society;

3. (3) The corporation is not a legally coercive institution.

To reach the conclusion that:

1. (4) The principles of justice do not apply to corporate governance.

Assuming that (1) is fairly uncontroversial, Singer proceeds to justify (2) and (3). His account of the two remaining premises is outlined below.

1.2. A Legally Coercive View of the Basic Structure

Since the primary subject of justice is the basic structure of society, any assessment of an institution on the grounds of justice requires an account of this structure. Yet Rawls is notorious for having failed to provide an unambiguous definition of society’s basic structure. Singer rightly notes this ambiguity and the several textually supported interpretations of the basic structure it has generated (2015: 76). The basic structure of society is defined in A Theory of Justice as the institutions that structure the terms of social cooperation. Society’s basic structure is indeed “the way in which the major social institutions distribute fundamental rights and duties and determine the division of advantages from social cooperation” (1999: 7; see also Rawls, Reference Rawls2001: 8-9). However, Rawls also claims “the basic structure is the primary subject of justice because its effects are so profound and present from the start” (1999: 6, cited in Singer, Reference Singer2015: 76; see also Rawls, Reference Rawls2001: 55). This suggests an expanded version of the basic structure as including institutions that have a pervasive impact on people’s lives, possibly including non-legally coercive institutions such as customary conventions. Finally, other textual indications may support a third interpretation of the basic structure as “institutions that … subject people to legal coercion” (Singer, Reference Singer2015: 76). Cohen sees, for instance, this interpretation as “mandated” by Rawls’s claim that “the law defines the basic structure within which the pursuit of all other activities takes place” (1999: 207, cited in Cohen, Reference Cohen2008: 133 n39). These perspectives, Singer notes, are identified by Abizadeh’s (Reference Abizadeh2007) “cooperation,” “pervasive impact” and “coercion” theories of the basic structure respectively.

Setting aside cooperation theory due to its vagueness, Singer focuses on the two remaining interpretations and main foci of scholarly debate: coercion theory and pervasive impact theory. For instance, Cohen notoriously assesses both options and argues for the latter. Yet Singer rejects Cohen’s preferred interpretation of the basic structure as both the formal and informal structures that affect people’s lives. For Singer, this expansive construal of the basic structure is inconsistent with the requirements of political liberalism. “Why not set the sights of a theory of justice on all those practices that structurally affect the life-chances and social position of people?” Singer asks (2015: 77). Of three possible reasons for rejecting the pervasive impact theory of the basic structure, and adopting the legally coercive one, he suggests that the most convincing is that of the structure and limits of a political conception of justice:

Rawls offers a political theory of justice that is limited in scope. Since justice applies to the basic structure of society, the restricted interpretation of the basic structure as a society’s legally coercive institutions is preferable. By contrast, the view of the basic structure as those institutions that have an important impact on people’s lives turns justice into a comprehensive doctrine that is necessarily politically illiberal.

1.3. The Exit Argument

Once the definition of the basic structure is settled, it is possible to ask whether corporate governance forms part of this structure. Again, Rawls is famously ambivalent about the status of corporate governance. Singer (Reference Singer2015: 78-79) recalls that Rawls envisages business firms as associations (2001: 92, 164), the “internal life” (2001: 164) and “rules” of which sit apart from the basic structure (2005: 268-69, cited in Singer, Reference Singer2015: 82; 1999: 126). Yet Rawls hints that justice may address issues of corporate governance (2001: 178). The ambiguity is that, as part of the internal rules of associations, corporate governance operates against the background of the basic structure, without being part of it, whereas by being part of the subject of justice corporate governance must belong to this structure. Of note, Rawls’s approach distinguishes between firms as economic organizations or legal entities, and their internal rules, as partly expressed in the framework of corporate law. Citations excluded, in what follows, “corporation” refers to any given corporation that results from an actual process of incorporation and “Corporation” to the corporate form, defined as the rules and procedures structuring corporations, defining roles, rights and duties, and including but not limited to governance rules. Corporate rules can be conventional as well as legal, and for the latter, be entrenched in corporate law, parts of labor and other relevant bodies of law (depending on legal systems).

In the absence of clear indications from Rawls regarding corporate governance, Singer suggests that we ask whether corporations form part of the basic structure. If “the corporation can be considered part of the basic structure … we can ask whether the principles of justice would have implications for the internal structure of the firm” (Singer, Reference Singer2015: 78). If it cannot, some other regulation might be adopted, yet not define corporate governance directly. Singer then spells out the argument that Rawlsians would have to make to demonstrate that the corporation is part of the basic structure: a claim that “the business corporation … should be considered part of the basic structure … would have to make a case not only about the importance of the corporation and its impact on people’s lives but also about its nature as a legally coercive institution” (2015: 79, last emphasis added). Here, Singer specifies the coercive theory of the basic structure as relating to institutions that have both a pervasive impact and are legally coercive. Setting aside the question of impact, he focuses on legal coercion.

The proposed criterion for determining whether an institution is legally coercive turns on the possibility of exit. This is directly inspired by Rawls’s reflections on the governance of religious associations (Rawls, Reference Rawls2001: 20, 93, 164):

The legally coercive nature of an institution depends on whether its “authority” can be “exited”. Rawls’s considerations on associations and the church suggest, in turn, this exit is unavailable if the rules of the institution are imposed on the individual rather than chosen voluntarily. Furthermore, legal coercion is demanding in one respect: lack of exit needs to be “universally true,” that is, true for all individuals (Singer, Reference Singer2015: 80). Thus, an institution is legally coercive only if no individual can exit its authority, except by leaving the state’s territory. The church, for example, is not legally coercive because individuals can exit its authority and decide not to live by its rules. Rawls remarks that “we impose any such [religious] doctrine on ourselves” (Rawls, Reference Rawls2001: 93); we accept ecclesial authority voluntarily (Rawls, Reference Rawls2001: 182, n3) and ceasing to accept it is no legal offense (Rawls, Reference Rawls2001: 93). Government rules are by contrast coercive because “the power of the government cannot be evaded except by leaving the state’s territory” (Rawls, Reference Rawls2001: 93). Singer ranks “constitutional doctrine[s]” and “tax regime[s]” among legally coercive institutions (Singer, Reference Singer2015: 80), presumably because citizens cannot exit their authority and must abide by their rules. While Singer does not draw this distinction himself, we may note that these examples point to a legal interpretation of exit availability (absence of rule enforcement by the state), but that some also consider non-intentional restrictions on exit as relevant for public justification and justice ^{Footnote 3} .

Corporations are not legally coercive, the argument goes, because their authority can be exited. While the argument applies to all “individuals who contract with the corporation” (Singer, Reference Singer2015: 79), the discussion nonetheless focuses on employees who represent the party that may be intuitively thought of as being coerced ^{Footnote 4} . We may indeed think that customers can buy products from other shops, investors can divest, or management can leave without legal (or other relevant) restrictions. As Singer rightly notes, however, it may be difficult for employees to quit their jobs. In less than perfect markets, employees may be locked into their firms because of scarce employers or non-transferable skills ^{Footnote 5} . Yet due to the demand for universal application, this situation does not qualify as an instance of legal coercion: even if some employees might not be able to leave the corporation that they hold a contract with, usually a number of them can. Crucially for Singer, this distinguishes corporations from states: “while … for some individuals … the ability to exit [the corporation] is not freely or realistically available, this is by no means universally true in the way that parallel obligations under a tax regime or a constitutional doctrine would be” (2015: 80, emphasis added). The corporation does not coerce its employees in the way that a “tax regime” or “constitutional doctrine” coerces citizens. Some employees can leave, even if not all of them, while all citizens must respect fundamental liberties and pay their taxes. Exit is neither legally nor otherwise universally restricted in corporations ^{Footnote 6} .

Call this the exit argument for rejecting the view of the corporation as legally coercive, which we may now summarize as follows:

The exit argument rests on two premises

1. (3.1) An institution is legally coercive only if its authority cannot be exited;

2. (3.2) The authority of the corporation can be exited.

Leading to the conclusion that

1. (3) The corporation is not a legally coercive institution

I now turn to a critical assessment of the basic structure objection, notably the way in which its third premise plays out in the overall objection.

2.1. The Subject Exerting Coercion: Corporate Governance Rather than Firms

To assess the exit argument, we must first clarify the meaning of “institution” in (3.1). Rawls defines institutions as “a public system of rules,” where “public” means that the rules must be known as if people have agreed on them, and people also know that these rules are known to others (1999: 48). Rawls’s illustration of institutions includes non-controversial political and socio-economic institutions of the liberal state, but also, importantly, private conventions: “as examples of institutions … we may think of games and rituals, trials and parliaments, markets and systems of property” (1999: 48, emphasis added). Rawls states that an institution can be thought of “as an abstract object, that is, as a possible form of conduct expressed by a system of rules” (1999: 48), but he also emphasizes that the realization of these rules depends on real people acting upon them. This is a stylized and uncontroversial account of institutions, yet crucially, it draws a distinction between institutions as rules and particular organizations or persons acting upon such rules.

Let us now turn to the meaning of “corporation” in the second premise of the exit argument (3.2). The vocabulary of a “contract with the corporation” (Singer, Reference Singer2015: 79), of quitting “a job” or exiting a “firm” refers to the corporation as one particular economic organization or legal entity, rather than the legal and conventional rules that make up the Corporation. Thus what is established in conclusion (3) of the exit argument is that corporations, yet not their governance structures, are not legally coercive. Turning to the basic structure objection, this version of premise (3) of the objection does not allow us to grant its conclusion (4), according to which corporate governance is not part of the basic structure, and hence not a subject for justice. What is required for this conclusion to follow is premise (3), according to which the Corporation (hence corporate governance) is not legally coercive.

This observation should suffice for the limited argument this commentary attempts to make. The subject of the exit argument is ill-focused for the basic structure objection to work: it should focus on the governance rules that form part of the Corporation rather than on corporations. One may, however, respond, and this was perhaps Rawls’s implicit assumption, that corporations stand in effect for their structuring rules so that if the corporation can be exited, so can the Corporation and associated corporate governance. But this does not follow. In the next section, I argue that even a revised version of the exit argument that focuses on corporate governance would fail to establish that it is not legally coercive, on both interpretations of restriction on exit.

2.2. Non-intentional Coercion, Corporate Governance and Representative Employees

In the revised version of the exit argument, we must ascertain whether corporate governance as defined in corporate law (and parts of labor law) can be “exited,” or whether it establishes its authority over individuals. First, suppose that all economic organizations of a given society are hierarchical corporations. ^{Footnote 8} Furthermore, suppose that employees willing to exit their firms can find similar jobs without incurring any cost, effectively making an exit option from firms available to all workers. While it is true that employees can escape the authority of one particular employer, they will necessarily face the authority of another one in another firm. What they cannot evade, except perhaps by leaving the state’s territory, is the hierarchical structure of the Corporation. Here, exit from the Corporation does not follow from exit from corporations, and corporate governance rules appear coercive because they are inescapable.

Yet one might object that liberal societies, notably near-just ones, are likely to allow for various organizational and business charters (including not-for-profits, sole proprietorships, partnerships, different kinds of co-ops, state-owned firms, etc.) alongside corporations. Assuming some effective level of pluralism among business forms in society, it can be argued that some employees are able to escape the corporate form, choosing for instance to work for a partnership, a cooperative, a sole proprietorship or perhaps not work at all. They need not all live by corporate rules, indicating that corporate governance is not, as such, coercive.

We may still resist this conclusion, if we shift the focus of the discussion from a specification of what is to be exited, a firm or an institution, to the question of who is coerced. The discussion thus far has held the coerced to be particular, identifiable individuals. However, the appropriate standpoint for assessing the possibilities of exit is arguably not that of an actual individual, but of a “representative” individual, say the worst-off (Rawls, Reference Rawls1999: 56). Suppose that the worst-off social position includes low-ranked workers in corporations. Suppose that some, but not all, of the worst-off can become self-employed. This is plausible as not all big corporations are likely to disappear, even in a property-owning democracy that spreads capital ownership. Even if the worst-off are not coerced individually they may be seen as coerced collectively, ^{Footnote 9} because a number have to work for a corporation and operate according to corporate rules. The worst-off social position remains subject to the Corporation and its governance rules. In other words, the requirement that some (but not all) can exit might set the bar too low for defining non coercive institutions if the relevant standpoint is that of a social position. This view of coercion may seem stretched, but as a criterion for defining a society’s basic structure, it is congruent with Rawls’s method for assessing the impact of policies. What matters for justice is how different policy options affect social positions, notably the worst-off, not how one particular individual fares under each policy. ^{Footnote 10} In sum, it is possible to conceive of a society in which exit from corporations is legally available for all and actually effective from some and yet in which a social position remains subjected to corporate rules. This argument rests however on a view of coercion as non-intentional restriction on exit, so I offer another argument that retains the more stringent criterion of state-enforced restriction on exit, and need not rely on representative employees being coerced.

2.3. State-enforced Coercion, Corporate Governance and Citizens

Here, we assume the account of legal coercion as an inescapable exposure to non-optional rules backed by state power. A preliminary consideration for the argument to follow is that it is inaccurate to describe corporate governance as simply “bottom-up” contractual arrangements ^{Footnote 11} (Orts, Reference Orts2013; Néron, Reference Néron2015: 97). Many structures of corporate governance are state-granted and -guaranteed in both corporate and labor law (O’Neill, Reference O’Neill2008, Reference O'Neill2009; Orts, Reference Orts2013; Néron, Reference Néron2015). Anderson similarly argues that “capitalist property” cannot “exist without state action” (2015: 61). There are three ways in which state action may appear coercive.

First, some emphasize that state enforcement of the terms of a private contract, whatever they may be, is coercive. For Blake “even private law … is rife with coercion” (2001: 276-77). This is because state involvement turns private promissory commitments into actually enforced ones. Once one enters into a contractual agreement, the implications of contract law cannot be escaped, as those of a game can, simply by leaving the game. At some point in the contracting process, exit is no longer available and “the adjudication of disputes will issue in a coercive transfer of legal rights” (Blake, Reference Blake2001: 277). However, this consideration still focuses on parties who, some may insist, have voluntarily chosen to contract. The suggestion here will be that we shift our evaluation of coercion from employees (and contracting parties) to citizens.

A second consideration is that the state makes corporate and labor law coercive not just for those who have contracted with the corporation but for those who plan to do so, by extending the authority of what would be local norms to the entire political society. As Andersen notes, for capital ownership to be recognized and operate as an asset, it requires “the state to formalize [people’s] property rights” (2015: 62) with a “written record” of property and the “standardization of property-rights” through their integration into the broader legal framework of society. Corporate law and labor law “standardize” across society the various “rights and obligations” attached to capital ownership, corporate governance, and employee status. For instance, anyone operating a corporation has to apply non-optional features in a given country (e.g., a required number of directors). For Anderson, standardization is unavoidable in capitalism because it creates among shareholders and employees the shared expectations that are necessary for the “scaling up” of productive enterprises. State action is arguably coercive as it imposes a standard that all involved, or planning to get involved, in corporations must respect. Yet some may still reject this as relevantly legally coercive, because it only concerns individuals who chose to get involved in corporations.

There is a third and more decisive respect in which corporate governance appears coercive: it is coercive for all citizens, and not just those contracting or planning to contract with the corporation. The standards set in corporate and labor law establish rights and duties for corporations, shareholders, managers and employees, which relate to important aspects of society such as property rights, social rights and taxation, and which must be respected by all citizens and public agencies. Crucially, the implications of corporate governance are not simply internal to a community of practice, such as playing chess or hide-and-seek might be, but are integrated into society’s wider structure of rights and duties, thereby creating unescapable constraints for third parties. For instance, limited liability makes it impossible to sue a shareholder for more than the value of its shares (O’Neill, Reference O'Neill2009: 176). As Néron rightly emphasizes, this has to be imposed upon third parties by the state; two individuals cannot, on their own, contractually limit the other’s right to sue them (Néron, Reference Néron2015: 97). Thus, the bearing of limited liability on society is strikingly different from that of the various associational rules that Rawls mentions—“apostasy,” “a point of theological doctrine” or “the rules of evidence used by a scientific society” (Rawls, Reference Rawls2001: 93). This is one notable difference between corporations and associations like independent church congregations (see Néron, Reference Néron2015: 96). Unlike other voluntary associational rules, limited liability is intentionally imposed by the state upon citizens of a political society in a way that they cannot exit. This echoes Blake’s consideration that state-backed contract law empowers individuals “to make legal rules determining ownership that all must be compelled to obey” (Blake, Reference Blake2001: 277). Another example is that of somebody’s status as an employee (labor law) or a manager (corporate law): it gives or denies access to social rights such as unemployment benefits that are claimed on society (or third parties in society). The point is that when a state operates rules that create rights and duties not just within but also outside the community of practice, these rules are relevantly coercive for the wider political community. In sum, state-backed corporate law, as well as the relevant parts of labor law, is relevantly coercive once we identify one important standpoint from which to assess legal coercion. Regardless of whether employees and other contracting parties are coerced, citizens of the wider political community are, just as they are coerced by the constitutional protection of the basic liberties of others, while individuals see their actions enabled and protected by these very liberties. Citizens, in other words, are subjected to the authoritative social meaning and implications of state-backed corporate governance in a way they cannot escape ^{Footnote 12} . And if one still wishes to resist the view that citizens are coerced by enabling corporate law, one would also have to explain according to which understanding of coercion it is that enabling basic liberties are part of the legally coercive structure of society, while enabling corporate law is not.

To summarize, if one holds coercion theory to be an appropriate account of the basic structure of society, and exit a relevant criterion for identifying coercive institutions, then corporate governance appears no less coercive for citizens than other less controversial dimensions of the basic structure. While this argument disputes the view that no aspect of corporate governance can be part of the basic structure of society, it does not tell whether and which elements can. As noted earlier, the latter also requires demonstrating that these aspects of corporate governance affect people’s lives significantly.

It is plausible to think that Rawls suspected that aspects of firms’ governance might be significant. He sometimes characterizes the basic structure “as a social and economic regime” (2001: 56) and asks later whether such regimes, “viewed as social systems, complete with their political, economic and social institutions,” are “right and just” (2001: 136). The kinds of regime that he examines are various types of capitalism, state socialism, democratic socialism and property-owning democracy. These differ not just with respect to taxation and redistribution, but also with respect to firms’ governance features (e.g., type of ownership, workers’ participation rights, etc.) indicating that these features could be relevant to justice. This commentary suggests that the basic structure objection does not give us reason to consider that this is illegitimate in political liberalism, nor to see as irrelevant Rawls’s hypothesis that democratic forms of corporate governance might be preferable in terms of political values.