‘Dams and flows’: boundary formation and dislocation in the financialised firm

Mainstream economic theories of the firm argue that the boundary between firm and market is determined by efficiency-enhancing logics which optimise coordination or bargaining outcomes. Drawing on social anthropological work, this paper critiques these accounts, arguing instead that firms are socially embedded and that firm boundary formation should therefore be understood as an attempt to fix the limits of certain relational rights and obligations that are moral in their conception. Consequently, boundaries are often contested and subject to renegotiation. We employ the parsimonious concepts of ‘dams and flows’ to examine how attempts to curtail the claims of some stakeholders and extend the claims of others at any one historical moment produce boundaries of different kinds. To illustrate this, we first trace the moral arguments used to advance limited liability rights to shareholders during the Companies Act in the mid-nineteenth century, which cut or ‘dammed’ obligations at a particular point and moment, directing new flows of obligation and wealth. We then explore the different moral reasoning of agency theory—the foundation of the financialised firm—which foregrounds the property rights of shareholder principles and obligations of managerial agents to them. We argue that this moral reasoning led to new dams and flows that have changed corporate governance and accounting practice, producing—counterintuitively—a reinvigorated form of managerialism, leaving the firm financially and morally unstable; its boundaries increasingly unable to contain its relational tensions.


Introduction
Mainstream economic theory has traditionally understood firm boundary formation as the result of optimising processes which respond to and resolve transaction cost problems, coordination difficulties or bargaining inefficiencies in markets. This functionalism has been challenged by two main streams within the financialisation literature. One stream has argued that firm boundaries do not emerge ineluctably and optimally to resolve market-based problems; rather firm boundaries are shaped by shareholder preferences and expectations which may produce suboptimal strategic, allocative and social outcomes. This stream argues that investors award premiums to companies that spin-off or outsource low-return parts of their business producing disaggregated corporate structures that generate larger profits from fewer assets (Davis & Kim 2015;Zuckerman 1999). This may allow CEOs to articulate simpler corporate narratives about performance drivers and strategic intent (Froud et al 2006), but can hollow out important skills, resources and innovative capacities (Lazonick 2015). Another stream focuses less on investor perceptions and more on managerial conceptions of controli.e. shared understandings about how best to organise and run corporations for profit-and how a shareholder value conception of the firm encouraged the same processes of corporate disaggregation (Fligstein and Shin 2007).
Our approach to boundary formation follows in this tradition of critique, but reflects the rather different disciplinary backgrounds of the authors. For social anthropologists, there is a long history in the study of boundary-making as part of a social process, as entities emerge through acts of exchange. This work has questioned the taken-for-granted nature of the existence of social groups such as tribes or clans, and has instead explored the ways in which their existence is elicited through perspectives which highlight particular relations, obscuring others. We apply these same insights to firms. Similar themes appear in accounting research. As Llewellyn (1994) notes, accounting does not just enable the 'organization of production', it is implicated in the 'production of the organization' by making boundary decisions about the thresholds of a transaction-i.e. what falls inside and what outside the organisation when it transacts with other entities in its environment, and how those exchanges are represented through internal systems of valuation, reporting practice and accountability which bind the organisation together. Boundary-drawing almost always involves moral questions about rights, responsibilities and obligations and where boundaries lie, classically, for example around how to account for the externalities of an exchange.
Our contribution therefore is to focus on how conceptions of rights and obligations shape the legal form and reporting practices of the financialised corporation, with redistributive effects. Our core argument is as follows: that all market relations are simply one form of social relation. Consequently, the boundaries of the corporation emerge from formal and informal acts designed to fix the limits of relational obligation at a particular moment. Predominantly, though not exclusively, we view boundaries as constructed through socio-legal and accounting processes. We argue that this process of fixing is a dialectical one which can involve choices and often conflict over which relations are deemed to be legally and morally legitimate at particular moments. Boundary drawing, therefore, involves an ongoing struggle to impose a 'line of sight' which subjects firms to renegotiations around the norms, conventions, rules and laws whichfor extended periods-create a relative fixity to those boundaries. In this context, what gives entities their different 'character' is the particular way in which boundary thresholds are established-i.e. how chains of relational rights and obligations are extended or curtailed in order to create particular bodies. This, ultimately, allows us to distinguish firms from other kinds of social entities and to examine changes in the legal form and reporting practices of those entities over time. This approach implies that rather than looking for a singular 'theory of the firm', that our focus should be on the social and technological processes through which particular kinds of firm come into being and evolve. In order to examine how the curtailment or extension of rights and obligations bring different types of entities into being, we use the parsimonious concepts of 'dams' and 'flows'.
This approach departs from existing financialisation work on firm boundaries in two ways. First, we follow Robe's (2011) conceptual distinction between 'the firm' as a grouping of social and technological activities and relations, and 'the corporation' as an individual corporate person and reporting entity as it is legally recognised, where the role of the latter is not merely to represent the economic outcome of the activities of the former, but to organise that representation in time and space in a way that reflects, reinforces and constructs the interests of the dominant social forces that comprise the firm at any one moment. Conceptions of rights and obligations take us, therefore, to the constitutive effects and distributional implications of particular legal forms and reporting practices undertaken by the corporation. This differs subtly but importantly from conceptions of control which focuses more on questions of firm governance and the physical organisation of firm operations to achieve profit-making, or at least conflates the identities of firm and corporation. Second, we take the view that conceptions of how the world ought-to-be can differ from the world that is subsequently structured by those conceptions. The relation is not necessarily performative; it can be 'virtual' producing unintended outcomes (Miller 2002). It is possible, in other words, that a conception of rights and obligations can be advanced by groups who then gain materially from those conceptions, even though the rights and advantages that accrue to those actors are not immediately obvious from those conceptions.
This idea that different conceptions can be accommodated or subverted helps us to understand some of the more puzzling contemporary developments in this era of financialisation, which-if agency theory is to be taken at its own estimation-was supposed to discipline management and increase returns to shareholders. The first is soar-away CEO and CFO pay and the associated rise of top to bottom inequality driven mainly by wages not capital income (Piketty 2018). The shareholder value revolution has been rather more permissive than disciplinary when it comes to senior executive pay. The second, as examples of corporate collapses with accounting irregularities mount up, is the growing sense that we are living in what Jim Chanos has termed, 'the golden age of fraud'-that the shareholder value revolution hascounterintuitively-increased agency problems, producing the incentives, means and contexts for directors to engage in the dark arts of creative accounting and the obfuscation of risk through the corporation's legal form, leading to an explosion of malpractice for personal gain on the blind side of investors.
Our paper attempts to understand these unexpected outcomes of financialisation through a theoretical lens which focuses on the different conceptions of rights and obligations that give form to the corporation and its boundaries. We first revisit mainstream economic theories of firm boundary formation, critiquing the socialdisembeddedness of those accounts. In the following section, we then develop our alternative theory of boundary formation, drawing on the idea that conceptions of rights and obligations, extended or curtailed, constitute an entity through processes of exchange. To clarify this argument, in the next section we apply those ideas to the moral arguments made in support of limited liability in the Companies Acts of the mid-nineteenth century, and the forms of relationality and obligation envisaged and show how a certain set of relational obligations-the claims of commercial debt providers on shareholders-were limited (or dammed) in order to create a new entity-the modern limited liability company (LLC 1 hereafter)-that directed flows of wealth in new directions through the corporate form. A following section then explores how, with the advent of agency theory, notions of shareholder indemnity to incentivise investment were displaced by a morality premised around the inalienable rights of property owners and how this moral conception sought to dam control rights to create new flows that would maximise shareholder returns. We argue that this 'financialisation of the firm' had important implications for corporate governance and financial reporting practice, which counterintuitively increased rather than disciplined the discretionary power of directors. In a final concluding section, we discuss how the shareholder value revolution produced a form of managerialism which fostered modes of relationality and obligation-new dams that direct flowswhich facilitated greater intra-firm inequality and increased the strain on the firm as an economic and moral entity.

Firm-market theories of firm boundaries
Understandings of firm boundary formation are conventionally set within a neoclassical economic problematic-that actors are rational and seek to maximise their utility, that information is perfect and that transactions are frictionless. Underlying most theories of firm boundary formation, therefore, whether explicit or implicit, is a theory of the market. This is perhaps most explicit in Coase (1937) where the existence of the market is treated as an a priori. Firms are assumed to emerge from existing market relations, like 'islands of conscious power in this ocean of unconscious co-ordination, like lumps of butter coagulating in a pail of buttermilk' (p. 388).
Coase defines firm boundary thresholds as determined by forms of coordination: markets consist of socially disembedded actors coordinated by price, but firms are coordinated by hierarchy-the direction of activity through authority or fiat (p. 393)-leading to the 'supersession of the price mechanism' (p. 389). But this poses a conundrum for economic thought: why does so much activity take place within firm boundaries if the market is such an efficient allocator of resources? Coase relaxes some of the central axioms of classical economic thought to argue that firms are present where there are latent costs or inefficiencies in particular market transactions. And for Coase, the boundary between firm and market is determined by Marshallian efficiency logics: hierarchical control may be an optimal form of exchange when the marginal costs of negotiating and writing enforceable contracts for each market transaction are high-for example in an employment contract when contracts are incomplete and frequently re-negotiated over the long term. Firms, in other words, make economic activity more efficient when there are market inefficiencies.
This idea was extended by Williamson, retaining the same transaction cost focus, but adding the effects of uncertainty, frequency and asset specificity on firm organisation (Williamson 1981;. The boundaries of the firm for Williamson are driven by attempts to solve 'hold-up' and 'moral hazard' problems that are produced when one economic agent refuses to invest in the production of an asset when it is efficient for them to do so because the asset in question is dependent on the presence of another owned by a counterparty to make profit. In such circumstances, it can hand the counterparty too much ex post bargaining power producing 'hold-up' problems which diminishes the incentives to invest. For Williamson, firms resolve this incentive problem through vertical integration, and hence, the boundaries between firm and market reflect the optimal institutional arrangement for each transaction: either hierarchy to resolve the incentive problems of hold up and moral hazard or the market to access inputs that are uniform or nonspecific. Again, this boundary is seen to follow a marginalist logic and achieve Marshallian efficiencies (Holmstrom & Roberts 1998).
Unlike the transaction cost approach, the property rights theory of the firm (see Grossman and Hart, 1986;Hart and Moore, 1990) does not explicitly rely on the construct of an impersonal market to contrast differing modes of governance and coordination. For Grossman and Hart (1986), a firm is simply defined, and its boundaries delimited, by the non-human assets it owns. For Grossman and Hart (1986), the residual right to control particular assets confers ex ante bargaining power to their owner under conditions of incomplete contracting: the ownership of nonhuman assets can be levered to influence the terms of new agreements, and hence, the future payoffs from investing in the new or ongoing relationship. Property rights theory does, therefore, invoke the market implicitly through the argument that the purpose of asset ownership is to allow a firm to design more efficient organisational structures in situations where markets function poorly because of externalities such as excessive hold-ups. The logics of efficiency thus determine where boundaries are formed, and thus which organisational forms prosper in a context of market competition.
There are some similarities between the resource-based and property rights theory of the firm, although resource-based perspectives focus less on bilateral ex ante bargaining power and more on the role of innovation as a source of sustainable competitive advantage and rents in final markets. Resource-based accounts focus on resource heterogeneity and immobility across firms (Foss and Knudson 2013;Wernerfelt 1984) and argue that firm boundaries are determined by the ownership of bundles of organisational and entrepreneurial resources (Penrose 1959), where those boundaries reflect management's strategic view on which non-tradeable assetsskills, human capital, core competences or capabilities enshrined in organisational routines-are likely to generate long-lived rents (Barney 1996). Ownership of the difficult-to-imitate skills, knowledge and assets required to produce unanticipated goods that generate rents shape the scale and scope of firm boundaries (Prahalad and Hamel 1990).
These theories of firm boundary formation, alongside the nexus of contracts view discussed below, form the building blocks which are the basis for a variety of iterations within economics and beyond. 2 For example Coase's arguments about transaction costs were developed to explain the rise of multinational enterprises (Caves 1996;Hennart 1991). Grossman and Hart's property rights perspectives were developed to consider how power accrues to actors with access to broader critical resources (Rajan & Zingales 1998), including political resources (Zingales 2017). Resource-based perspectives were extended to explore the efficiency of knowledge utilisation (Grant 1996), and to distinguish resources from capabilities (Helfat & Peteraf, 2003), and then organisational from dynamic capabilities (Teece et al., 1997).
These different theories tend to presume that the impersonal market is the natural ontological basis of all human economic or even social activity, where the firm's existence is a kind of puzzle to be solved. They also describe an almost auto-generative process of firm boundary formation, most plainly in the 'buttermilk' quote above, and approach tautology when arguing that firm boundaries form on a Marshallian basis, to improve efficiency and resolve latent market problems. Such an approach naturalises 'the market' as a starting point in a manner that, as many scholars have observed, overlooks the large variations in forms of market exchange and their socially embedded character (e.g. Geertz 1963, 1978, Ho 2009, Polayni 1944, Zaloom 2006).

Nexus of contracts
Standing somewhat apart from these other theories, there is the 'nexus of contracts' view of the firm. Although influenced by similar neo-classical concerns, this perspective starts from the premise that the distinction between firm and market is much less stark. Alchain and Demsetz (1972) signal this difference in the opening page of their highly cited American Economic Review article, which should be read as a critique of Coase: 'It is common to see the firm characterized by the power to settle issues by fiat, by authority, or by disciplinary action superior to that available in the conventional market. This is delusion. The firm does not own all its inputs. It has no power of fiat, no authority, no disciplinary action any different in the slightest degree from ordinary market contracting between any two people' (p. 777) They explicitly reject the Coasian idea that activities within the firm are governed by authority, and emphasise the role of contracts as a mechanism for voluntary exchange. For example they argue there is little difference between an employer firing an employee and a consumer effectively firing a grocer by stopping purchases-both involve a termination of a contract and the with-holding of future payments (p. 777). The defining feature of a firm, for these authors, is the presence of a 'centralized contractual agent in a team productive process' (p. 778) which incurs a cost of 'metering' (i.e. measuring and controlling output). Jensen and Meckling (1976) extend this view to all economic relations, arguing that both firm and market are simply a series of contractual relations. The firm is, in their view therefore, 'a legal fiction'-an artificial construct under law which allows certain organisations to be treated as individuals and which acts as a nexus for those contracts. Firm boundaries are simply an artefact of legal incorporation. Indeed Jensen and Meckling assert that: 'it makes little or no sense to try to distinguish those things which are "inside" the firm (or any other organization) from those things that are "outside" of it. There is in a very real sense only a multitude of complex relationships (i.e., contracts)

between the legal fiction (the firm) and the owners of labor, material and capital inputs and the consumers of output' (p. 311)
A series of claims follow as a consequence. First, they claim that the firm has no objective function or purpose; it is simply the locus of a process within which the conflicting objectives of internal and external individuals are brought into equilibrium within a framework of contractual relations (p. 311). However, this does not imply that firm boundaries are driven by any inherent efficiency or rentgenerating logic. They develop Alchian and Demsetz, 1972' concept of metering to consider the agency costs of monitoring senior managers in the average public limited company, and conclude with a puzzle about the inefficiencies of contracting in that corporate form, where they accept that suboptimal firm relations can endure for long periods (p. 330).
Jensen and Meckling's idea that the firm is a 'legal fiction' is an intriguing one from a social scientific perspective. It is, in a sense, a constructivist account-viewing the firm as an artefact of law. Consequently, this approach does not assign any superordinate purpose or function to the firm, nor does it presume that firm boundaries emerge from some ineluctable logic of efficiency-the firm is simply a legal convenience which allows them to transact as if they were individual people in a market. They also acknowledge the capacity for divergent interests and inefficient forms of contracting. Yet the market is still held as an a priori and market relations are reified to the exclusion of all other forms of relationality (this is explicit in Jensen and Meckling 1976: 307). Nevertheless their approach opens up a space for alternatives which acknowledges the social-constructionist nature of firm boundary formation.

A social anthropology of corporate boundary formation
Our approach follows this constructionist logic, but departs from the nexus of contracts view by understanding market relations as just one form of social relation (Granovetter 1985). Those social relations are subject to discursive framing that shapes how actors understand themselves in relation to others, giving rise to certain modes of social action and exchange (Miller 2002). Within this context, market, firm and the boundary between the two should be understood as conceptual abstractions of certain types of social relations, whose essentialism and efficiency should not be assumed a priori (Power 2018).
This approach has a long pedigree in anthropological analyses of the emergence of entities in the process of exchange. In Melanesian anthropology, for example there is a large body of literature that questions the taken-for-granted nature of the existence of social groups (such as tribes or clans) and explores the ways in which their existence is elicited in particular contexts by advancing a perspective which highlights particular relations, obscuring others (e.g. Wagner 1974;Strathern 1988). This process of elicitation commonly occurs through moments of exchange. For example, in anthropological studies of the custom of turguvuai among Tolai people of East New Britain Province, Papua New Guinea, two matrilineages jointly gave gifts to other lineages at exchange ceremonies, and over time that process of gift exchange led those two matrilineages to be considered as a single matrilineage or vunatarai. Hence, it was not simply the case that vunatarai were a pre-existing group that entered into exchange. Their very boundaries and the nature of their existence were shaped by these exchanges-by virtue of their gift-exchange relations-hence the nature of the gifts given, shaped the boundaries of the entity (Leaver and Martin 2016).
Anthropological studies therefore show that different kinds of social groupings are constructed through the organisation of different kinds of relations and transactions, and that those relations reflect and embody particular expectations and norms of moral conduct. It should not therefore be assumed that any kind of entitywhether clan or multinational corporation-expresses an essentialist relational logic (e.g. Kirsch 2014: 209). It is the interplay of different kinds of relational moralities and the resolution of the debate over which kinds of obligations should be prioritised at particular instances that require empirical investigation in order to understand the particular entified forms that arise from them.
Companies provide an interesting example of these issues because of the role of law as a formal and codified expression of those normative or moral expectations in defining where relations and obligations begin and end, and thus where firm boundaries lie. Companies can be conceived of as both a firm, i.e. a grouping of social and technological activities and relations, and as a corporation-an individual corporate person as they are legally recognised in the UK and USA (see Robe 2011). Whereas much of the economic work on firm boundaries (excepting the nexus of contracts approach) assumes the former, or conflates the two, understanding boundaries through the lens of moral or even legal rights and obligations allows us to consider firm boundaries in the context of the latter. The limited liability company, for example is an entity created through legal ritual which cuts particular relational obligations-explicitly the limitation of debt obligations-at a particular point based on a particular moral view about where the obligations of shareholders begin and end. Importantly it is this act of legal limitation that brings the corporation into being as a separate person in the eyes of the law, much as a particular instance of public gift exchange creates a perspective that can 'elicit' the clan as an entity of limited or long duration. The notion that firms automatically emerge from the market to resolve contracting inefficiencies like 'lumps of butter in a pail of buttermilk' is therefore ahistorical and misunderstands the active process of entification and boundary formation.
Legal judgements that enforce the priority of particular claims and relational obligations over others not only express a particular perspective but can establish selffulfilling logics that bring into being new entities in the image of those conceptions. Again, the growth of the limited liability company would be a good example of this. However, it is also possible that particular conceptions of rights and obligations are articulated by actors instrumentally who then assimilate new realities into ongoing patterns of relationality and exchange (Miller 2002). For example it is entirely possible that the moral arguments made for the limited liability company actually bring that corporation into being through legal fiat, without that new legal form generating the forms of relationality and sense of moral good that were anticipated.
Our conceptual start point, therefore, is that firms are the outcome of the extension and curtailment of particular relational obligations, where the question of what relations and obligations are extended or curtailed reflects particular moral reasonings which may be formally (though not exclusively) codified in legal form. This moral basis of this process of legal codification is often obscured. The prioritisation of some claims over others is often represented as inevitable, unequivocal, universal, legally unchallengeable and part of the 'natural order of things'. This appearance of naturalism may endure for long periods, displacing alternatives. However, there are also informal norms, habits and customs which structure these boundaries. 3 Those other informal, less codified forms of moral reasoning may include appeals to governance norms, human nature or universal principles of economic rationality in the context of regulation, corporate governance rules or even economic theorisation. All of these arguments can be seen as attempts to resolve the ambiguity over the nature of the firm, by establishing a singular line of sight of one particular order of relational obligations, displacing others. Hence, whilst the legal enforcement of contractual obligations requires a fixity and consistency, these informal features mean firms do not remain homogeneous entities over time-their form, purpose and operation will reflect the different, cross-cutting social relations which underlie them and the different moral reasoning summoned to justify them. Any firm therefore marks a particular instance of a process in which boundaries are formed through the attempt to limit or extend particular relational obligations based on moral reasoning, just as dams direct flows.
The firm relations that take precedence at any particular moment are never permanently fixed-they require ongoing maintenance. This may involve attempts to impose a unitary moral logic to explain the salience of one form of relationality and obligation over another as a condition of the creation and maintenance of entities like the firm. As we will later argue, the moral reasoning of agency theory as a justification for corporate governance reform was one such attempt. However, contests over the moral basis of particular forms of relationality do arise and, at pivotal moments, the outcome of those contests leads to changes in the temporary settlements over which forms of relationality and obligation are deemed legitimate. This is therefore an unstable process which can produce unintended consequences that run counter in many respects to the intentions of their authors.
One might think of the example of a platform company like Uber whose attempts to draw firm boundaries have been contested. Uber argues that it is a technological intermediary rather than a traditional transport employer-a definition which fixes the boundaries of the company's obligations narrowly so that Uber's drivers are classified as self-employed rather than company 'workers'. Uber thus pushes the financial obligations of pensions, holidays, sick pay and so on outside its legal boundaries and which-until relatively recently-had not been accounted for. This boundary formation was justified using moral reasoning-that Uber was part of the 'sharing economy' and that drivers were 'micro-entrepreneurs' (Botsman, 2015), company 'ride-sharers' or 'partners' (Uber Technologies Inc., 2019: 10), and that Uber simply matched drivers with passengers through their GPS algorithmic technology. The narrow drawing of this boundary line generated moral criticism-that Uber's moral obligation, and thus its boundaries, should be extended to commit the company to treat its drivers as workers and thus take the liabilities associated with these worker obligations (Leszczynski 2020). This became a legal matter in October 2016 when an employment tribunal found that Uber's drivers should be considered workers and were thus entitled to compensation for lost benefits from the company (Booth 2020). The battle to place particular relations within or outside the firm and corporation is therefore one based upon divergent moral visions and, in this case, has the potential to have profound effects in terms of the future economic organisation of our society given the legal precedents set.
As the Uber example shows, there can be more than one way to envisage the flows of relational obligation embodied by a firm. We cannot, as economic theories do, assume that there is always a single underlying type of relational obligation that the firm exists to maximise or minimise. It may be that in some circumstances a firm will be used to minimise transaction costs, as Coase and Williamson suggest, or to maximise bargaining power, or even to hoard intellectual property. These different purposes of the firm are reflective of a particular type of relational obligation that the firm exists to regulate, but the extent to which they determine the existence of firms and the location of their boundaries is a matter for empirical study not a priori assumption. Instead of claiming that any of these principles are necessarily fundamental to the firm's existence and nature, we propose the following three starting points to analysis.
1. We cannot assume a singular underlying purpose of 'the firm'. 2. Any claim that there is such a singular underlying purpose of 'the firm' is a statement of intent rather than a statement of fact. 3. Any claim that the boundaries of the firm lie in a particular place or that particular obligations and relations lie inside and others lie outside of the firm is similarly a statement of intent rather than a statement of fact.
To try to explain how this formulation helps us to think about firm boundary formation, and the creation of dams and flows of rights and obligation, we will now use this approach to first examine the creation of the modern limited liability company through the Companies Acts, and second by examining how the moral conceptions of agency theory shaped corporate governance rules and accounting practices, but failed to bring about the kind of relational outcomes anticipated.

The history of the limited liability company as a moral and relational entity
In this section, we take the establishment of the limited company in the UK as an illustration of the ways in which corporate entities that are designed to create new flows of wealth, are brought into being through the damming of relational obligations-the potential claims that creditors could make upon investors-in order to release new flows of wealth to shareholders. The establishment of the modern limited liability company in the UK in the late nineteenth century marked a revolution in the nature of the firm. Limited liability privileges in seventeenth century Britain had only been granted by the crown to a select group of joint stock firms like the East India Company. These privileges were gradually extended to all private firms through a series of amendments to the Companies Acts between 1844 and 1862, most notably the 1844 Joint Stock Companies Act and the 1855 Limited Liability Act. At the heart of the legal discussion about limited liability was a moral argument about the optimal relational form of the firm necessary to achieve certain social ends. In the House of Lords, the motion to accept the Limited Liability Bill into Law was moved by Lord Stanley on the basis that it was necessary to allow for 'undertakings of great importance' and that most of the UK's historically unprecedented economic and industrial growth in the previous half-century had been due to exemptions being granted and a strategic blind-eye being turned to the official legal doctrine of unlimited liability (Stanley HL Deb 7 August 1855).
The extension of limited liability was justified as being of benefit not just to investors but to the 'country' so that it might 'boast' of 'the greatest works ever constructed in any part of the world' (ibid). This claim to a higher moral and wider social purpose for the limitation of liability was countered by others, such as the former Secretary of State for War, Earl Grey who argued that it was morally wrong to remove the debt obligations from those who stood to benefit from the transactions entered into by their corporate surrogate, which would instead bear risk on the shareholders' behalf, or as Grey (1855) put it, to '…depart from the old-established maxim that all the partners are individually liable for the whole of the debts of the concern'. Grey (ibid) protested that the Bill, was being rushed through the legislature without due scrutiny because of the emergency of the UK's entanglement in the war in Crimea. As one of the Bill's supporters, Earl Granville (ibid) acknowledged that the need to 'free commerce from restrictions' took on additional resonance under these circumstances. But for Grey (ibid), this did not override the greater concern that, '… if the Bill passes precipitately and in an imperfect shape, you may find, when too late, that great mischief has been done which it will be most difficult hereafter to repair, and that vested interests may have been created which will be found very inconvenient in regard to future legislation'.
Grey agreed with Stanley and Granville that the 'great concerns' that benefited the nation were of importance, but raised two counter-arguments. First, he turned Stanley's point back against him, observing that, '… really useful undertakings have no difficulty under the existing law in finding the needful capital' (Grey ibid). Instead the Bill risked creating, '… bubble Companies and bubble schemes, which no capitalists will take up, and the projectors of which wish to trade, under cover of limited liability, upon the small sums advanced by other people, and thereby to escape from personal risk' (Grey ibid).
Grey argued that the limited liability company removed responsibility for debts from individuals. He doubted whether they would add to the wealth of the nation through production and economic growth and emphasised instead the risks that bubble companies would bring moral disgrace upon the country: '… in legalizing the principle of limited liability the greatest caution is requisite in framing the details of the measure, in order to prevent the door from being opened to practices detrimental to the high mercantile honour and credit of this country' (Grey ibid).
Yet it was Granville's argument that won through-that the limitation of debt liability for shareholders would encourage passive investment by small rentiers; that equity investment would be particularly valuable in the case of large scale, long gestation capital intensive projects where low risk was in itself not enough to attract private investors; and that society would benefit from an increase of the national economy's capital stock which was the superordinate aim of the two main British acts (Bowman et al 2015).
The rearrangement of debt obligations to allow the creation of the company as an artificial person who carried debt obligations on behalf of its owners was therefore a moral as much as a technical question. The formation of the modern limited liability company arising from this and other Acts, most notably the 1862 Companies Act effectively allowed for the creation of a dam intended to redirect previous flows of obligation by shielding the shareholders from the corporation's creditors; effectively limiting creditor claims to the estate of the corporation rather than those of shareholders.
Over the 170 years since limited liability has been established in UK law, the acceptance of limited liability and the moral arguments underlying it have become naturalised across the spectrum of mainstream political opinion, but the moral challenge of Grey that the firm would become a dumping ground for debt never fully disappeared. It was resurrected in 1897 Salomon vs. Salomon and Company which became a foundational legal dispute concerning the boundaries of the limited company in the UK. The case, which has shaped all subsequent legal judgement on the matter in the UK and most of its former colonies, was in essence a simple dispute over whether or not the company was a separate entity from its founder Aaron Salomon. Salomon was an East London shoemaker who established a company-Salomon and Co-in 1892 under the provisions of the 1862 Companies Act in order to purchase his own shoe making business. The new company was established with 20,007 shares; 20,001 of which were owned by Mr Salomon himself, the other six each owned by separate close relatives of Mr Salomon, thus providing the minimum of seven shareholders as stipulated by the provisions of the Act. The debt-laden company went into liquidation shortly afterwards and unsecured creditors launched claims against Mr Salomon on the basis that there was no real distinction between himself and the company. The High Court and the Court of Appeal both ruled against Mr Salomon, on the basis that, as one law lord at the final judgement put it, that, … the appellant [Mr Salomon] was in truth the company, the other members being either his trustees or mere "dummies," and consequently that the appellant carried on what was truly his own business under cover of the name of the company, which was nothing more than an alias for Aron Salomon (Watson [1897] Salomon vs. Salomon and Co.). This decision was overturned by the House of Lords whose judgement was based on a position that it was not their job to adjudicate the intentions of the legislators in 1862. As long as the technical provisions of the 1862 Act had been enacted then the company was to be judged as a separate entity in reality.
The Salomon case is more than a simple example of how the 1862 Act was open to moral hazard problems. It shows how a moral evaluation about the salience of particular relationships became part of legal reasoning, foregrounding some relations, backgrounding others-leading to the formation of different boundaries of rights and obligation. Hence, Salomon's pre-existing kinship relationship with the other directors of the company Salomon and Co. was deemed salient and morally dubious in the first two legal battles-the other directors are deemed to be puppets or shams by dint of their familial relationship to Aaron Salomon. The earlier judgements were premised on the idea that business relations should be sufficiently separated from kinship relations and that if this separation was not maintained they should not be viewed as legitimate business relations, nor the corporation a legitimate entity. From this perspective, the corporation does not emerge naturally from relations governed by market principles-it is seen to be the outcome of different moralities of relational obligation, and it is the dispute over the extent to which one is kept sufficiently separate from the other that determines whether the corporation is deemed 'real' and separate or not (see also Yanagisako 2003).
However, this kinship relation is considered irrelevant to the ruling in the House of Lords appeal case, thus fixing obligations at a different point, maintaining a boundary between Salomon & Co and Aaron Salomon, even after the company was liquidated. In this sense, kinship ties were necessary to bring the firm named 'Salomon and Co.' into being, allowing it to engage in transactional relationships. But those relations were backgrounded in the appeal ruling that determined the company was a discrete legal entity separate from Aaron Salomon, damming claims of relational obligation (commercial debt) from creditors and releasing flows of wealth along different relational paths.
It was only through the legal precedents set by Salomon versus Salomon and other disputes that limited liability privileges were extended to corporate subsidiaries, so that it became possible to tier subsidiaries upon subsidiaries (Blumberg 1985, p. 608) with each recognised as separate legal persons in the eyes of the law. In an age of financialisation, this was to have a profound effect on the organisation of corporations, the dams and flows constructed within its boundaries and the extent to which those relational obligations are understood as part of a shared moral purpose of the firm.

The influence of agency theory
It is possible to observe more contemporary examples where new moral reasonings have altered boundary-forming relations and obligations informally without necessarily being codified in legal form. The effects of agency theory represent one such case where ideas about the rights of shareholders and the obligations of directors and managers created new dams and flows.
However, it should first be acknowledged that the role of agency theory as a driver of the shareholder value revolution is contested. It is argued that many US conglomerates adopted a more explicit financial orientation as early as the 1960s (Fligstein 1993;Fligstein and Markovitz 1993) which became embedded in a variety of shareholder-oriented metrics, developed by management consultants for a section of the corporate establishment already committed to increasing returns to investors (Knafo & Dutta 2020). As Knafo and Dutta (2020) argue, it would therefore be wrong to assume that agency theory brought about the financialisation of the firm, since this investor-focused orientation was evident before Jensen and Meckling's seminal contribution.
But the influence of agency theory should not be entirely discounted-it provided an explanation of and justification for shareholder prioritisation at a critical time of flux and industrial change, as the long boom ended and international competition began to encroach on previously sheltered domestic markets. Agency theory was therefore at heart a wider moral injunction as to the purpose of the firm (see Fourcade and Healy 2007) which sought to define which actors' rights should be prioritised, what activities were deemed legitimate for management to undertake, and which governance arrangements were necessary to run corporations efficiently for all stakeholders (Erturk et al 2007).
Agency theory advanced the inalienable rights of property owners-that shareholders stood at the apex of a hierarchy of organisational claims as property rights holders and suppliers of risk capital (Shleifer and Vishny, 1997). The board (and management generally) were thus understood to be 'agents' serving the interests of those owners (Eisenhardt, 1989). The agency problem, in this moral formulation, was that individuals act in their own narrowly defined self-interest, often with guile and deceit, and that shareholders' and managers' interests diverged producing a basic dilemma: managers sought to maximise their own perquisites, shirk, satisfice, or build larger, less efficient combines to enhance personal status. This imposed a cost on shareholders who wanted to increase returns on their invested capital, from which society would ultimately benefit through improved allocative efficiencies (Fama & Jensen 1983;Jensen 1994). Agency problems would be reduced, according to these authors, through stock-based remuneration for executives which align their interests with shareholders, adjustments to board composition and their monitoring effectiveness, and the creation of a more active market for corporate control.
Agency theory, in this sense, like the example of limited liability, should be understood as an attempt to construct a new line of sight on the firm, foregrounding some relations, backgrounding others, to produce new understandings about which rights are legitimate and thus where boundaries of obligation are drawn. Whereas limited liability focused on the relation between shareholders and creditors, where the drawing of a legal boundary cut shareholder obligations to the firm at a particular point to produce new relational flows of income and wealth; agency theory focused instead on contractual boundaries and the damming of managerial control rights to direct a flow of obligations from directors to shareholders. The allocative effects of curtailing or 'damming' control rights would, according to Jensen (1986), direct flows to shareholders in the form of the disgorged cashflows of the corporation, where investors would then be free to allocate their capital elsewhere in the market, creating liquidity and stimulating wider economic growth (Jensen 1986). This form of moral reasoning, although not legally codified, should be understood as an attempt to resolve the ambiguity of the firm through governance norms rather than law. Agency theory has no formal statutory basis and in fact mis-states the precise legal status of shareholders and directors in relation to the corporation in both US and UK Company Law (Lan & Heracleous 2010). As legal scholars have argued, agency theory positions shareholders as 'owners' of the firm, but their ownership rights are more accurately defined as being limited to the equity paper issued by the corporation; creditors have a prior ownership claim over corporate assets (Robe 2011;Stout 2012). It also positions directors as agents of the shareholder, rather than as fiduciaries to the corporation, which is the formal understanding of the directorial role in UK and US company law position (Clarke 2014). Finally, directors' obligations are not to pursue shareholder value maximisation, but to comply with the capital maintenance regime through the prudent stewardship of the corporation in the interests of its members and other stakeholders (Lan & Heracleous 2010). In this sense, agency theory's conceptions of rights and obligations do not correspond with legal definitions, but nevertheless legitimised and naturalised the shareholder primacy model which informed a range of new governance rules, policy initiatives, reporting requirements and training programmes.
Before discussing the impact of agency theory on those rules and initiatives, it is worth highlighting the moral contingency of modern agency theory by examining the somewhat different moral understanding of rights and obligations from earlier theorisations on the same agency problem. Berle and Means (1933), for example argued that by handing control rights over the corporation to management, shareholders effectively, 'surrendered the right that the corporation should be operated in their sole interest' and 'released the community from the obligation to protect them to the full extent implied in the doctrine of property rights' (Berle and Means 1933:355). For Berle and Means, this meant shareholder property rights should be sublimated to those of the wider community in order to 'demand that the modern corporation serve not alone the owners or control but all society' (op cit:355-6). This different moral conception posited an alternative hierarchy of relations and obligations, bringing the community interest inside the boundaries of legitimate firm calculation, with quite different allocative implications.
In this sense, it is possible to see parallels with our earlier discussion of limited liability. Both are particular manifestations of the general tendency for firms or corporate entities to be brought into being and maintained in particular states through processes of boundary formation that rely upon the limiting of particular relational obligations, whether that be the obligation of the shareholder to the creditor or of the director to the shareholder. Limited liability is seen as the basis of the corporation having an independent legal existence in the eyes of the UK courts, even though many would look at the Salomon case and agree that in practice the corporation might well be a sham. The modern corporation, as analysed by Berle and Means, had become an 'independent' object in practice due to the separation of legal ownership from actual control (something missing in the Salomon case), or as Rathenau (cited in Berle and Means) might have it, 'the detachment of property from possessor'. Although the legal judgement of the UK law lords of the 1890s and the theoretical judgement of the US academics of the 1930s focus on different relations, both are in essence about the ways in which the firm or corporation is only established as a truly independent entity by the management and curtailment of the particular relational obligations under consideration. But a focus on different kinds of relational obligations as being central to the creation of the firm as an independent entity is what gives the firm a different historical character in different contexts. And it is here that we now look at how Jensen-ian agency theory became embedded in governance reform and the reform of financial reporting.

The embeddedness of shareholder primacy in governance reform and financial reporting
Agency theory became embedded in corporate governance principles and arrangements during the 1980s and 1990s, figuring as the control technology to encourage or enforce the pursuit of the shareholder interest (Blair and Stout, 1999;Clarke 2004;Lan and Heracleous, 2010). Those ideas influenced corporate governance thinking at the Business Roundtable and the National Association of Corporate Directors in the USA (Daily et al 2003), and the 1992 Cadbury and 1995 Greenbury Reports on UK corporate governance and director remuneration, where shareholder primacy and board obligations to shareholders are stated matter-of-factly without discussion of company law requirements (Erturk et al 2007). However, agency theory also influenced other dimensions of the firm, specifically in the arena of financial reporting. Like the history of shareholder value and corporate governance described by Knafo and Dutta (2020), the history of shareholder primacy and the move towards market-based fair value reporting also predates agency theory. Its usage dates back to at least the nineteenth century (Richard 2004). Debates also raged within standard setting bodies like the FASB and IASB in the 1970s about whether historic cost modes of valuation should be replaced with fair value in a context of runaway inflation (Georgiou & Jack 2011). Agency theory added a more holistic moral reasoning for the shareholder primacy model advocated by a coalition of actors lobbying for change within those financial reporting rulesetting bodies (Nolke and Perry 2007;Perry and Nolke 2006). These shareholderoriented principles subsequently became embedded in the conceptual frameworks of accounting standard bodies like the IASB and the FASB, advancing a view that the role of accounting was to enhance the 'decision-usefulness' of reporting for the purposes of resource allocation by investors (Whittington 2008), where the act of financial reporting itself was often represented as part of the contractual obligations of directors to shareholders within an agency theory frame (e.g. Watts and Zimmermann 1986). The accounting framework moved from a backwards-looking, transaction-oriented historic cost mode of valuation to forward looking, balance sheet oriented, market-based fair value accounting (Power 2010). Concepts like 'stewardship' which were embedded in accounting rules and which previously connoted managers' broader fiduciary responsibility for the firm, were given a different evaluative accent as the term became increasingly conflated with the concept of accountability to shareholders (see, for example Accounting Standards Board 2007, para. 4.4-4.7; para 5.2). The fair value principles aligned valuations with a capital market view that expectations of discounted future cashflows should become the basis for recording asset prices in the annual report (Richard 2004).
Although these two developments in corporate governance regimes and financial reporting regimes might be thought of as commensurable and supportive, we know that these initiatives failed to produce many of the outcomes expected by agency theorists. New corporate governance regimes were supposed to discipline management and prioritise shareholders. Yet CEO and CFO pay has grown significantly in the post-crisis recovery: mean CEO pay in the FTSE100 grew from £3.69 m in 2009 to £4.70 m by 2019 (High Pay Centre 2020) and from $10.37 m in 2009 to $21.28 m in 2018 in the 350 largest US firms (Economic Policy Institute 2020). Top to bottom inequality has been driven mainly by wages not capital income (Piketty 2018), implying that command positions inside the firm, rather than ownership relations outside, explain rather more about the nature of inequality in our economy. Similarly, the disgorging of cashflows that was supposed to liberate equity capital in search of new investment opportunities has produced no investment boom or productivity miracle (Tori & Onaran 2018). Instead financialisation has produced long periods of stagnation and crisis (Magdoff & Foster 2014). Finally, there is mounting evidence that the shareholder value revolution has-counterintuitively-increased agency problems where creative accounting and obfuscation through the corporation's legal form has led to a worrying increase in fraud and malpractice on the blind side of investors. These outcomes are, we contend, the result of three blindspots in agency theory arguments, which not only underestimate the capacity for management to subvert the intentions of agency theory, but for agency theory itself to produce outcomes which entrench management power further.
First, the relational field constructed by agency theory centres narrowly on property rights and the principal-agent relation; all other firm relations are relegated to contractual exchanges of equal weight and significance, where the firm-as a legal person in the eyes of the law-is just another economic actor. In this sense, all actors are represented as being outside the boundaries of the firm which, as a legal fiction, contracts equivalently with directors, management, creditors, workers, suppliers and so on in a socially disembedded market. Consequently questions of organisational power embedded in ongoing social and organisational relations rarely feature. It is simply presumed that superior contracting, such as the payment of executives in stock options or other shareholder-linked modes of remuneration, resolves those problems. The effects were therefore always more likely to be 'virtual' rather than 'performative' because powerful actors will try to preserve their existing privileges born out of those relations, accommodating those new initiatives into their ongoing relations and exchanges whilst keeping their more challenging effects outside of their everyday interactions (Miller 2002: 231). It is unsurprising that executives embraced the new forms of shareholder linked remuneration and assimilated them into the 'a priori's and reiterated practices' of organisational life (Fleming and Banerjee 2016) because it allowed them to extend their wealth and status.
Second, in addition to this blindspot around organisational power, the agency theory view understands relations as market relations between actors, where the firm appears as a singular actor through its 'fictive' legal construction. This ignores Robe's crucial distinction between the firm as a bundle of physical and organisational practices and the corporation as a legal entity, which reports on the activities of the firm. It is thus possible to have one 'firm' and multiple corporate identities, each transacting with one another within a group structure. The nature of those transactions is not determined by each corporation as a sentient, reflexive transacting being, they are an artefact of director fiat.
Third, and relatedly, at the level of reporting, the production of decision-relevant information for shareholders which pivots the accounting framework towards expectations of future cashflows, discount rates and other events necessarily introduces greater subjectivity into the valuation process. Financial reporting after financialisation has, in other words, not only handed directors greater discretionary power in reporting on the economics of the business they operate, it has greatly enhanced the role of the 'Big 4' valuation intermediaries as both consultants who can work with management on the presentational aspects of financial reporting as well as symbolic verifiers of that value recorded via their auditing function.

The relational contradictions of the financialised firm
Each of these three blindspots has opened up space for greater managerialism. 4 Collectively they provide directors with both the incentive (through their shareholder value linked remuneration bonuses) and the means (through their newfound reporting discretion) to take an optimistic view of their own organisation's economic performance (Baker et al 2020). Strategies that involve creative accounting and financial engineering become more appealing and operationalisable as a means of increasing profits and cash or minimising losses (Bratton 2001;Benston 2006). These include revenue recognition techniques which bring income forward; paying dividends out of fair value asset revaluations; factoring and reverse factoring to manage cashflow and postponing goodwill impairments (see Baker et al 2020 for extended discussion of these processes).
By handing management both the incentive and the means to report optimistically, a number of relational tensions have built up which are beginning to pull the boundaries of the corporation in different directions. There are three emerging dislocations.

The dislocation between group and parent boundaries
Directors have increasingly exploited the different reporting identities of the consolidated group and the parent company to pay out higher distributions, often in excess of their net income generated or even their group retained earnings. They have sought, in other words, to exploit the legal affordances granted to them post-Salomon to direct-through fiat-parts of the corporation to transact with each other. Whilst the tax avoidance implications of this practice are well understood (Palan 2006), its implications for distributions are less well recognised. Group reports net out the asset and liability position of the parent company and subsidiary network, and adjust for inter-company transfers-they hence present a picture of the company by examining the consolidated performance of all entities adjudged to be within its boundaries. The parent, however, usually acts as a container for dividends paid up through the subsidiary network, which are its assets. This boundary ambiguity means it is possible to use profit overloads (and loss overloads) within the subsidiary network, through, for example transfer pricing, to pay more dividends to the parent than there is profit generated at a consolidated level, without breaking company law rules on distributable profits. 5 This can lead to a growing disconnect between the reported net asset and retained earnings positions of the two identities, so that they become difficult to reconcile, even though one is notionally part of the other.

The dislocation between firm and corporation boundaries
That is that the reporting becomes increasingly difficult to square with the operations, and relatedly that profit becomes difficult to reconcile with cash balances. After the financialisation of financial reporting, firm balance sheets have become a reflection or refraction of market prices, blurring the boundaries between what is the corporation and what is the market; what is inside and what is outside the firm (Hayoun 2018;Huikku et al 2017). This is particularly the case when proprietary valuation models which act as market proxies are used to value level 3 items where no market prices are available. This effectively brings the market, or at least representations of the market, inside the firm, so that the process of valuation itself becomes recursive and circular-it uses imaginary markets (Power 2010) to produce imaginary prices (Benston 2006). This can lead to the production of accounting outputs that become resilient to changes in company fortunes or external economic conditions, producing their own hermetic reality or 'simulacra'-signs built upon signs (Bougen and Young 2012;Macintosh et al 2000). But this can also create risks. The capacity to hold-out, for example on goodwill impairments for long periods, can create pent-up impairment pressures which cannot continue indefinitely. And if nonimpairment becomes, suddenly, untenable in the eyes of the market, and there is pressure brought to bear on directors and auditors, then those pent-up impairments can be very large when finally recognised, and may create going concern problems.

The dislocation between boundaries and obligations
In order to insulate parents from such risks, or even to accelerate distributions, subsidiaries can act as fuses or partitions for the corporate parent, maximising returns, minimising internal risks but externalising the costs of contingencies. Crucially this relies on the maintenance of firm boundaries within the wider corporate network. It requires, in other words, a fixity of risk and obligation within the subsidiary network to facilitate the flow of riskless returns to the corporate parent. But this fragmentation carries its own tensions and contradictions as to where obligations lie, which entity is liable and which jurisdictional rules apply when things go wrong. This can be seen in cases where there is a legal challenge as to whether English or US courts have jurisdiction in private equity debt settlements when the corporate entity spanned both jurisdictions (see Leaver and Martin 2016), or when borrowers challenge lenders to prove they own their debt, as with the robo-signing mortgage scandal after 2008 (Holland 2011).
These three boundary dislocations highlight a wider break-between the moral conception of rights and obligations offered by agency theory and the results of its application in the ongoing systems of relationality and exchange that comprise the firm. There is a very real sense that firm and corporation are pulling apart, unable to contain tensions and contradictions within and between their boundaries. These dislocations point not only to an emerging financial instability, but an instability of a profoundly moral kind as the firm is increasingly unable to account for its relational features and their outcomes in the terms used to justify its present form. The tensions of firm boundaries are, in other words, indicative of wider tensions in the social relations of production and accumulation, and the inequalities they currently engender.

Conclusion: moral overflows and renegotiating firm boundaries
Our paper has highlighted how a different line of sight on boundary formation provides an alternative entry point to understanding the problems of the financialised firm. Agency theory's blindspots around organisational power, corporate structure and the subjectivism of financial reporting has-in our view-had a permissive rather than a disciplinary effect on the exercise of discretionary management power. That is because the agency theory focus on principal-agent relations, and the wider foregrounding of contractual relations generally, backgrounds the nature of the corporation as a social, legal and reporting entity. In these blindspots, the tendency for directors to allocate greater effort and resource to corporation-based priorities around law and accounting structures, in Robe's terms, at the expense of firm-based operations has increased (see also Dobbin & Zorn 2005). Consequently, the use of creative accounting and the exploitation of the legal affordances won post-Salomon have served to intensify rather than resolve the ambiguity of the firm; only whilst the Salomon case presented a simple legal choice-the company was either legitimate or a sham depending upon perspective-today's corporation is made up of multiple entities where the corporation can appear as a singular body or as different combinations of multiple bodies depending upon the shifting perspectives of different moral or legal evaluations. This uncertainty complicates the enforcement of responsibility and obligation. Resolving which precise entities within a group hold which obligations in which jurisdictions, and which jurisdictions' laws are consequently salient has become harder, not simpler. This reveals an emerging paradox whereby, … at the micro level, capital seeks an ever more detailed refinement of liability obligations in ever more elaborate corporate structures; but as a direct consequence of this tendency, responsibility has never been so contested, ambiguous and diffuse at the aggregate (Leaver and Martin 2016).
Our approach which understands firm boundaries as the outcome of moral and social contests builds upon and extends classic work on financialisation. That work has highlighted how shareholder-primacy has produced dysfunctional outcomes across various dimensions: employment security and prosperity (Lazonick & O'Sullivan 2000;Lazonick 2012), working conditions (Fligstein & Shin 2007), investment (Agliettta & Breton 2001;Stockhammer 2004;Tori & Onaran 2020), accumulation (Orhangazi 2008;Stockhammer 2006;Tori & Onaran 2018), offshoring (Auvray & Rabinovich 2019), innovation (Stout 2012) and inequality (Dumenil & Levy 2015;Goldstein 2012;Lin and Tomaskovic-Devey, 2013). That evidence suggests that a desire to control labour costs and reduce investment are driven by pro-shareholder governance norms (and executive pay structures) which incentivise short-term distributions, with negative consequences for measures of productivity, accumulation and equality. Our approach shows that the governance norms that drive these outcomes not only affect the organisation of, and investment in, productive resources-in the form of offshoring, an erosion of working conditions and a depleted capital stock, but also shape the reporting mechanisms and legal organisation of the corporation.
That observation, however, does have implications for how we think about the processes of financialisation. If distributable profits can be generated by manipulating what is accounted for within the boundaries of the corporation-levering the subsidiary network to purchase a recently revalued asset or simply using forms of revenue recognition which bring distributable profits forward in time-then allocative pressures are not always zero-sum. If profits can be 'realised' through creative reporting, then other claimants, notably workers or investment, do not necessarily have to lose for shareholders to gain. In such circumstances, distributions are not paid out of operating cashflows, but from debt-financing. As Kliman and Williams (2015) note, debt has the capacity to resolve these zero-sum allocative trade-offsa view supported by recent reports that nearly half of all US share buybacks are funded by borrowed money (Light 2019). Consequently, other measures of financialisation, which focus on the financial sources of accumulation, may in fact reflect quite different processes related to financial engineering and creative reporting within a network of global subsidiaries-as Rabinovich's (2019) findings imply. That is not to say that senior management will not try to cut costs and use creative reporting or financial engineering simultaneously, but rather all are essential if as Dobbin and Zorn (2005) maintain-the emerging approach to running the corporation is to creatively produce the numbers that analysts and institutional investors like. Understanding the relation between corporate legal structure and accounting practice on distributions, investment, productivity and performance might be one fruitful area of future research. This takes us back to the nature of the firm and corporation. The frontiers of the firm are likely to become the battleground upon which future social contests are fought. If boundaries demarcate obligations, we might expect the next phase of financialisation to generate new forms of corporate legal organisation that seek to avoid the costs of those obligations, hoard wealth and make that wealth distributable.
One might look, for example to how the social pressure to bring the costs of climate change onto firm balance sheets has encouraged a splintering of oil and gas corporations who are spinning off their less environmentally sound assets into non-operable joint ventures or affiliates in which they have an equity stake to try to avoid assuming responsibility for the reporting-and thus the costs of reducing-scope 1 and 2 emissions. Or how platform companies who define their moral purpose as intermediating atomised providers and consumers are forcing 'normal' businesses to reimagine their structure and operations along such lines to stay competitive, pushing responsibility for social costs outside of their boundaries and back onto the workforce. Or how precarious, debt-loaded corporations with high dividend aspirations are protecting their income-generating assets from creditors by moving them from restricted to unrestricted subsidiaries where loan covenants do not apply. Each of these cases illustrates how boundaries can be drawn and redrawn to create new dams that direct flows of obligation and wealth.
Whether a moral justification for these practices can be sustained remains to be seen. The longer-term social, economic and ecological outcomes of this detailed boundary management are unlikely to be positive. Agency theory, in this sense, appears to be not only a failed economic project, but a failed moral project which cannot explain how it managed to empower the very agents it was designed to constrain and why its economic consequences have been so insipid. And so now, more than ever, we require a new moral conception of rights and obligations for the firm in order to prepare us for those social contests ahead.
Funding This work was funded by the Independent Social Research Foundation (PERF1) Political Economy Fellowship.
Data availability All data is taken from publicly available sources.
Code availability Not applicable.

Conflict of interest The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.