Corporate Social Responsibility as Obligated Internalisation of Social Costs

We propose that corporations should be subject to a legal obligation to identify and internalise their social costs or negative externalities. Our proposal reframes corporate social responsibility (CSR) as obligated internalisation of social costs, and relies on reflexive governance through mandated hybrid fora. We argue that our approach advances theory, as well as practice and policy, by building on and going beyond prior attempts to address social costs, such as prescriptive government regulation, Coasian bargaining and political CSR.


Introduction
Ideally, under the influence of competitive market forces and other incentives, corporations play a key role in the generation and distribution of social costs. As such, the quest to address the social costs arising from business activities is at the heart of the corporate social responsibility (CSR) movement. However, the highly influential neoclassical economic approach to CSR, achieves little in terms of addressing social costs, because externalities are only internalised where internalisation aligns with, or is expected to improve, corporate financial performance (Amaeshi 2010;Orlitzky 2011;Adegbite et al. 2019). The key question, which is often overlooked, is: should corporations abandon the internalisation of their negative externalities if this is not expected to result in improved financial performance? In other words, should corporations ignore the social costs they create in their pursuit of profit unless they can be addressed profitably? If corporations are expected to address them, how should they go about this?
The governance of social costs 1 (i.e. negative externalities) created by corporations is a perennial concern. Since Coase (1960), economists have generally argued that the solution to externalities is either instrumental regulation 2 or bargaining between the creator and the victim (Coasian bargaining) 3 of the externalities. The regulator should choose between these two options based on a cost-benefit analysis. In particular, regulators are required to compare the costs associated with government intervention with the transaction costs confronting parties when they attempt to deal with a particular externality by means of a contract. Many economists assume that regulatory costs (including the costs of producing and enforcing regulation and the distortions 1 Social costs here include environmental and ecological costs or externalities. As such we have used social costs and negative externalities interchangeably. 2 This includes both regulation and taxation, with the State designing and implementing rules and incentives intended to change behaviour in order to achieve the desired goal of mitigating a particular social cost. By instrumental regulation, we mean instrumental regulation of a specific negative externality (hereafter, instrumental regulation or simply, regulation). 3 By Coasian bargaining, we mean bargaining between the creator of the social cost and a specific stakeholder, who might be an individual or an affected group, with a view to resolving the issue to the satisfaction of all concerned (hereafter, Coasian bargaining).
1 3 of economic activity that may arise) will be very high, so the 'cure' of regulation will normally be worse than the 'disease' of externalities, making government intervention undesirable from an efficiency standpoint (Shleifer 2005). 4 This makes many economists sanguine about leaving the correction of some, or even most, externalities to the market, even though market failure led to the externalities in the first place. They then assume that if the parties fail to reach an agreement on a solution to a particular externality, this will be for transaction costs reasons, and as such, leaving the externality where it falls is the most efficient outcome in the circumstances. This approach overlooks the fact that externalities are real impacts on real people and the environment. It also assumes that negative externalities are only worth addressing where many people are affected. However, negative externalities are injuries (Honneth 1997).
Notwithstanding, externalities can also be positive-e.g. where a company contributes positively to the provision and administration of public goods and policies, especially in weak or unstable democracies. Scherer et al. (2013) call this Political Corporate Social Responsibility (PCSR). They see PCSR as a manifestation of corporate citizenship (Matten and Crane 2005), which strongly endorses the corporate provision and administration of citizens' rights in instances where the state is unable to do so. Nonetheless, in such situations, characterised by significant democratic deficit (Scherer et al. 2013), a firm risks losing legitimacy because it does not often have the democratic mandate to provide and administer such public goods and governance; and therein lies the paradox of PCSR, as the extension of corporate citizenship. To address this challenge, Scherer et al. (2013) call for a multi-stakeholder corporate governance system, which will require interested parties to address externalities through the processes of deliberative democracy (Habermas 1996(Habermas , 1998. This emerging approach aptly situates PCSR in the public policy domain (Scherer and Palazzo 2011;Scherer et al. 2014Scherer et al. , 2016, but views it from a non-market strategy perspective (Baron 1995), which "…must not be confused with private behavior 'that is appropriate in the market place' (Elster 1986, p. 111)" (Scherer et al. 2013, p. 479).
In this article, we take the public policy view of PCSR seriously, but unlike Scherer et al. (2013), we approach it from the perspective of correcting market failure. This perspective accepts that "…companies decide on public issues and deliberately cause externalities, meaning they 'affect other people' (Elster 1986, p. 111) who have no contractual relationship with the company and enjoy (in the case of state or regulatory failure) no protection by the state authorities" (Scherer et al. 2013, p. 479), because such externalities are normalised and tolerated as the inevitable outcomes of market failure. In other words, we take the view that addressing negative externalities should be obligatory even if they are borne by what appears to be a negligible number of people and/or things. In focussing on negative externalities as manifestations of market failure (Coase 1960), we argue that the conventional regulatory toolbox for addressing such externalities is incomplete. We submit that prescriptive government regulation or Coasian bargaining only offers a partial governance of externalities in a globalised economy, which is characterised by factually and technologically complex chains of causation (Kim and Davis 2016). This shortcoming is further compounded when managers have to articulate a business case for action even though the social costs created by business actions affect a far wider range of interests, beyond those of shareholders. We thus propose an alternative possible solution: obligated internalisation of social costs (or reflexive governance of negative externalities by corporations), through multi-stakeholder fora.
Reflexive governance is an inventive approach to regulation (Osuji 2015). It refers to the processes, procedures and relationships for mutual transformations of corporations and society. For example, when regulations require multistakeholder consultations to be undertaken before certain business decisions, which may have environmental impacts or market monopoly implications are made, an underlying assumption is that the process will facilitate a decision that has considered the interests of the firm and the stakeholders like employees, consumers, local residents and rival firms. These stakeholders are in a better position than the firm or a public regulator to articulate their interests. While the company may technically comply with the regulatory requirement by engaging in a series of bilateral consultations with stakeholders, the resulting decision may not be transformative of society since the different segments of society have not actually engaged with each other. Reflexive governance addresses this by demanding that consultations are undertaken in a collective process that allows all stakeholders to be present and listen to, reflect on and respond to one another's views. The process will assist in arriving at decisions that can positively transform society despite the disparate nature of the interests of the stakeholders who represent different segments of society.
Reflexive governance further recognises that social costs are complex and that any attempt to address them requires collaborative efforts. In other words, single actors such as corporations, civil societies, or governments cannot deal with these challenges in isolation. Reflexive governance builds on Teubner's (1993) reflexive law theory, which supports the organisational and procedural norms necessary for decentralised self-regulation of social subsystems, including corporations. By stressing the importance of 'relationships' and 'mutual transformations of firms and society' as governance outcomes, reflexive governance allows firms to be influenced by, and act on the basis of, facts, norms and values developed in collaboration with other social actors through a process of discussion. Since the key factor for the relevant facts, norms and values is 'mutual transformation', a self-centred and inward-looking business case approach to social costs is inadequate. Rather, the requirement of mutuality creates space for, and indeed requires, an approach to corporate internalisation of social costs, which is oriented towards society and is an instrument of public policy.
Viewing CSR through the lens of reflexive governance is very useful. For example, Scherer et al. (2013) call for multiple stakeholder fora to address the democratic deficits inherent in the contemporary practice of CSR. Our approach, which is grounded in solving market failures by addressing negative externalities, as opposed to the role of companies in creating positive externalities, aims to extend Scherer et al. (2013)'s PCSR. However, our contribution goes a step further by proposing a legal obligation to engage with stakeholders to identify social costs and elaborate acceptable solutions to them. Based on this understanding of reflexive governance and its link to political CSR, we argue that corporations should be mandated to establish 'hybrid fora', which would bring together creators of, and those affected by, externalities in order to trace those social costs and identify mutually acceptable solutions to them. The article proceeds with a critical account of the limitations of (1) instrumental regulation; (2) market-based solutions (Coasian bargaining); and (3) political CSR as approaches to dealing with the problems of social costs. Thereafter, it presents our approach of governing externalities through obligated internalisation of social costs through hybrid fora and the corporate decision-making process. Finally, it highlights some of the contributions and implications of this approach.

Governance of Externalities Through Instrumental Regulation
Economists accept that economic activities sometimes produce externalities, or 'external diseconomies' for outsiders or third parties. They also accept that the creation of negative externalities is a market failure, in the sense that where social cost exceeds private cost, the market does not necessarily achieve an efficient allocation of resources. Addressing externalities through regulation and taxation is however problematic. This is because they are both blunt instruments of social steering which produce distortions or second order effects. The practical impossibility of anticipating those second order effects means the regulator cannot determine a rate of taxation that will match the environmental cost and produce the common good; it will always be too high or too low. Nevertheless, under political pressure to justify their interventions, regulators often ignore this uncertainty and produce clear cost-benefit analyses of the impact they expect proposed regulatory interventions to have.
Furthermore, instrumental regulation faces a trilemma (Picciotto 2017) when faced with the problem of governing externalities. First, the law may be irrelevant in situations where corporations produce externalities across borders, and outside the reach of (national) regulation. Second, it may be difficult for the regulator to identify social costs before they occur. For example, if a regulator does not know the nature of the externality, let alone the harm it causes or the extent of its effects, adopting the precautionary principle will result in regulation that is under-or over-inclusive, distorting social and economic arrangements. This means regulation may be entirely lacking. Vatn and Bromley (1997) emphasise that externalities are novelties that are mostly recognised after they have been produced. This is compounded by the fact that there are often 'large time spans between when a physical act (e.g. emission) takes place, and when one becomes aware of the external effects it creates' (Vatn and Bromley 1997, p. 137). Hence, it is very difficult, if not impossible, to govern unpredictable externalities, such as the sociotechnical externalities which result from new technology, by means of ex -ante regulation.
If the law tries to avoid the first two aspects of the trilemma, it is likely to produce disintegrating effects on itself. As production processes become more complex and supply chains become longer, regulation has to abandon general principles applicable to entire industries or the economy as a whole in favour of an approach, which differentiates between individual firms, and even individual activities. This third aspect of the trilemma greatly increases regulatory costs. A good example is the response to the 2007-2008 global financial crisis, which focused on crisis management at the expense of underlying principles. In particular, government interventions, following the crisis, rescued the economic status quo with the disastrous effect that 'the complexity gap between the legal system and its societal environment was almost completely levelled' (Renner 2011, p. 100). Given the complexities of governing externalities through regulation, we now examine the option of bargaining between those concerned by an externality. Coase (1960) argued that addressing social costs through instrumental regulation would not necessarily increase social wealth for a number of reasons. First, the cost of government intervention is often high. Second, government intervention distorts the incentives of actors within the economic system and so changes their behaviour. Given these effects of intervention, Coase offers regulators an alternative to taxation or regulation: they can do nothing, and leave the particular social cost to possible negotiation between the creator and victim. As long as the transaction costs are lower than the gains from trade, self-interested economic agents should trade. Thus, Coasian bargaining refers to a process in which those concerned by an externality bargain to (re)allocate the right to create (or to be free from) the externality (Coase 1960). Presumably, since this is a market transaction, the right will be allocated to the party which values it most highly. If the right is not reassigned, Coasian economists argue that it is either because it is already efficiently allocated, or because transaction costs meant that it was not worth transferring the right.

Governance of Externalities Through 'Coasian Bargaining'
While this view has garnered support from economists, we argue that Coase's approach is counter-intuitive in law because he 'not only rides roughshod over notions of corrective justice, he also undermines fundamental notions of causation' (Ogus 2006, p. 8). The governance of externalities requires a much broader inquiry into how to allocate society's resources in order to increase aggregate social wealth. For instance, the assumption that the government can make a socially adequate choice between intervening with instrumental regulation and doing nothing requires some heroic epistemological assumptions. In order to compare the costs and benefits of regulation versus doing nothing, the regulator must try to anticipate how the parties will respond to regulation and how those responses will affect total social wealth. However, any attempt to predict the impact of a regulation on something as complex as a large, modern economy or the physical environment is very problematic (Vatn and Bromley 1997;Hiedenpää and Bromley 2016).
There are two other important weaknesses of 'Coasian bargaining'. First, there is an assumption that 'an initial distribution of private ownership rights is exogenously given' (Aoki 2001, p. 36). This means that the parties can bargain, knowing what their legal rights are in relation to a contested action. However, this does not correspond with legal scholarship, which suggests that rights are, very rarely (if ever), clearly allocated ex ante (Simpson 1996, pp. 86-87;Milhaupt and Pistor 2008;Morgan 2005). In short, if the law is not clear, rational calculation becomes impossible and 'Coasian bargaining' becomes indeterminate, with the outcome more like 'mutual accommodation' than a means of optimal resource allocation.
The second weakness is the assumption that transaction costs are the only barrier for wealth-enhancing bargains to reallocate property rights in relation to a particular social cost. If this is not the case, a regulator who carries out a cost-benefit analysis will not identify all the relevant costs and might not select the most appropriate governance structure for the externality. Moreover, it is essential for a regulator who is examining a situation in which no agreement is reached to understand why this is the case, as this will influence the regulator's decision on whether regulation can be justified on efficiency grounds. In addition, dispersed victims will encounter further obstacles to collective action, such as free rider problems, and will have to bear the costs of setting up a governance structure to coordinate their dealings with the corporation. Even if they manage to coordinate their efforts, the absence of default structures has strong distributional effects, with affected groups having to bear the costs of establishing an appropriate structure on a case-by-case basis. Thus, if Coasian bargaining is to be a realistic means of governing externalities in medium to large numbers scenarios, the law should arguably provide institutional support for coalitions of the affected, by means of default rules.
While the concept of transaction costs has undergone considerable expansion in the literature, there are other, behavioural and sociological barriers to wealth-enhancing reallocations of resources, which are rarely brought within the category of transaction costs. These are difficult to evaluate and greatly reduce the likelihood that the parties to an externality will strike a bargain. For instance, where one agent acts or takes decisions that harm the well-being of another, this will give rise to acrimony, especially among geographical neighbours. Some people will therefore refuse to bargain for 'psychological or sociological' reasons. 5 As Vatn and Bromley (1994) put it, the moral dimension intrudes into the presumed clarity of economic choice. This means that people may value something highly, but be unwilling to pay for it because they do not consider it appropriate to pay for something to which they believe they have an entitlement. These effects are well-documented in the endowment effect literature (Korobkin 2014), which suggests that people value rights they possess more highly than rights they might acquire. This effect presumably extends to 5 People rarely want to talk to, let alone bargain with, someone who has been impinging on their quiet enjoyment because this creates antagonism, and they are disinclined 'to think of the rights at stake in these cases as readily commensurable with cash' (Farnsworth 1999, p. 384). So, for example, people rarely if ever bargain around nuisance injunctions, and they are highly unlikely to be willing to pay a polluter so that they can have the clean water they believe they are entitled to. Individuals who imagine with some conviction that, say, their drinking water should be uncontaminated, will be expected to be unimpressed, if not irate, about having to pay to prevent it from being even more contaminated' (Vatn and Bromley 1994, p. 141). According to Mishan (1993), a more 'cynical view' would compare the argument that the 'victim' of the externality should pay the creator with a 'protection racket', according to which both the victim who is left unbeaten and the gang who are paid for not administering a beating are better off. Regardless of the effect on aggregate social wealth, a reluctance to pay not to be beaten is to be expected (Mishan 1993). the rights people believe they possess because they enjoyed access to particular resources in the past. Coase's model therefore contradicts the idea that people's willingness to bargain depends to a considerable extent on the perceived fairness or morality of the other party.
Also, and relatedly, Vatn and Bromley (1994) emphasise that individuals may be unwilling to pay to preserve the environment because they consider it incongruous to treat environmental 'goods and services' in the same way as commodities. The authors refer to survey evidence, in which people say that species diversity is very important for non-instrumental reasons, which shows that they value it highly. However, people are unwilling to pay to preserve species diversity. This may be because it is inappropriate to choose between a moral principle and ordinary consumption goods. Also, people view preserving species diversity, along with other public goods like a clean environment, as a public matter (Lévêque 1999). It may also be because there is no institutional framework that helps them understand how these things should be valued. As such, one cannot assume that the absence of 'environmental' bargains is an indication that social wealth is maximised.
Furthermore, the parties may not even agree on the nature, existence and extent of the externality. Bargaining can only occur when there are relevant facts available for all parties concerned. Like transaction costs, the costs of constructing facts are highly relevant to any analysis regarding the form of governance that will be the most cost effective in dealing with a particular externality (Johnston 2011). For example, the costs associated with a commission of inquiry show that governments face high costs of fact production. Since facts have to be constructed before the most appropriate governance can be identified, it follows that Coasian bargaining is unsuitable to prevent irreversible harms. In the case of a complex system like the environment, the consequences of relying on ex post governance may be very serious. An apparently small externality may result in catastrophic consequences where a particular species or other aspect of the environment is functionally transparent in the sense that its contribution to the complex system as a whole is not known until it is removed (Vatn and Bromley 1994). If the regulator refrains from constructing the facts, leaving the matter to those concerned, any bargaining process will also have to confront both transaction costs and the costs of fact production. At present, and as economists recognise, the only institutional structure of fact production available to the parties is litigation, which gives rise to very high costs (Shleifer 2010).
Besides, the disparity in resources between a corporation and a private citizen cannot but have an influence on the outcome of any negotiations or litigation between them. Victims of externalities created by large corporations are likely to be risk-averse concerning starting litigation due to the high costs involved and the stress from legal proceedings. This is because they have a personal interest at stake (in comparison with the corporation's managers and shareholders) and because they are more likely to be directly involved in the proceedings. All of these will be well understood by corporations (or at least their lawyers), and will therefore influence any negotiations which take place between them. The corporation will certainly put the complainants to proof, which has the effect of transferring all the fact construction costs onto them. Corporations may also be able to use their superior resources to increase those costs by hiring all the available local experts, and using procedural rules strategically to overwhelm the complainants. If affected individuals seek to pool resources and form a group to advance their complaint, the corporation can respond to the threat or initiation of litigation with 'divide and conquer' strategies. Finally, the corporation can supplement these legal strategies with a public relations campaign, claiming that the issue is one that should be dealt with by government, while retaining lobbyists to persuade the government in question not to intervene (Brueckner and Ross 2010;Parchomovsky and Siegelman 2004).
Together, these arguments suggest that-in legal practice rather than economic theory-Coasian bargaining is, and will always be, a rare occurrence. 6 Whether these obstacles to bargaining are subsumed into the transaction costs category, or whether they are viewed as a separate category of costs, they suggest that bargaining will be a rare-and expensive-way of dealing with corporate externalities. The economist's answer is that if these costs are high, making bargaining unlikely, and if government regulation is also costly, then social costs should be left where they fall. This might be acceptable if externalities were a rare outcome of economic activity, allowing economists simply to assume that the social gains from particular economic activities outweigh the social costs. However, once we move away from neoclassical economic models with their operating presumption that market failure is an exceptional case, there is growing acceptance that externalities are pervasive. This suggests that, before we can be sanguine about leaving social costs where they fall, we ought to attempt to identify them and quantify their economic impact. Given the significant challenges confronting Coasian bargaining, we now proceed to highlight the recent developments and challenges in governing externalities through political CSR, and then leverage these to advance our alternative model of obligated corporate internalisation of social costs.

Governance of Externalities Through Political CSR
Despite the increase in governmental involvement in CSR regulation across the world, 7 the role of government in the business and society debate remains understudied in the literature (Dentchev et al. 2017;Osuji 2015;Adegbite 2012;Scherer et al. 2016;Ehrnström-Fuentes 2016), especially in conceptual and theoretical terms (Gond et al. 2011). The general assumption of a strict division of labour that views governments as political actors and firms as economic actors (Sundaram and Inkpen 2004) has been displaced by globalisation (Scherer and Smid 2000;Marcus and Fremeth 2009;Margolis and Walsh 2003;Scherer and Palazzo 2011;Kaul et al. 2003). In that regard, the tension between 'new problems and received solutions', coupled with the power shift arising from the loss of governmental influence over multinational corporations, suggests the need for a new type of governance. As globalisation has given rise to a series of interdependencies between economic and social actors, especially in terms of risks, competition and opportunities (Beck 2000;Held et al. 1999), this new type of governance will need to address those interdependencies and their consequences.
These developments set the stage for the emergence of political CSR, which can be defined as businesses engaging in activities that are traditionally understood as governmental activities (such as provision of education, infrastructure, goods and services), in order to contribute to public policies and governance (Scherer and Palazzo 2011;Scherer et al. 2014). Political CSR has broadened the scope of governance and regulation and blurred the roles of public and private actors. It has, for example, created interest in 'the phenomenon of "new" and "experimentalist" governance and its relationship to law' (De Burca 2010, p. 227), as well as the concept of 'relational state' that implies shared co-regulatory responsibility of public and private actors (Gunningham 2012;Ho 2013;Lozano et al. 2008;Osuji 2015). As such, Brammer et al. (2012, p. 7) argue that we should seek to place CSR explicitly within a wider field of economic governance carried out by the market, state regulation and other methods. As a form of governmental involvement, political CSR therefore focuses on collective deliberations and concern for (global) public goods (Scherer et al. 2014) such as the European Commission's 'Multi-Stakeholder Forum' (Scherer et al. 2013) and 'CSR Alliance' (Buhmann 2011). In developing countries, political CSR may also lead to corporations filling institutional voids created by government neglect (Amaeshi et al. 2016;Adegbite and Nakajima 2011). Scherer and Palazzo (2011) further posit a thematic characterisation for political CSR that takes into consideration a number of factors including the self-regulatory role of CSR, which provides a new institutional role for private actors in a regulatory context. Scherer et al. (2013) argued that political CSR offers an opportunity to democratise corporate governance to compensate for the democratic deficit inherent in corporate political activity and corporate citizenship theories and practices. However, their suggestion is based on soft law and self-regulation. The transition from hard to soft law also changes the way private actors are regulated, because the process is devoid of governmental influence and sanctions (Shelton 2000). Nevertheless, the contribution of self-regulatory engagement to resolving political and social issues may be ambiguous (Scherer et al. 2009), especially due to the lack of governmental control and enforcement mechanisms. It is therefore important to evaluate private regulation in terms of competing self-regulatory initiatives and measure the level of co-operation between firms and civil society actors (Scherer and Palazzo 2011).
As such, despite the benefits of political CSR, it has been criticised as a product of globalisation (Whelan 2012) that over-stresses the role of businesses in global regulation and excludes local and regional governance . As a form of global governance, the conflicting standards and moral dilemmas arising from political CSR can cause local-global tension among multiple stakeholders (Pless and Maak 2011;Child and Rodrigues 2011;Jones and Fleming 2003). In this regard, Scherer et al. (2016) considered the new role of businesses in moving away from compliance with moral standards to a more politically inclined activity that does not fit with the current view that CSR should be confined to business case scenarios. Their comprehensive articulation defines political CSR as: 'those responsible business activities that turn corporations into political actors, by engaging in public deliberations, collective decisions, and the provision of public goods or the restrictions of public bad in cases where public authorities are unable or unwilling to fulfil this role. This includes but is not limited to, corporate contributions to different areas of governance, such as public health, education, public infrastructure, the enforcement of social and environmental standards along supply chains or the fight against global warming, corruption, discrimination or inequality. These corporate engagements are responsible because they are directed to the effective resolution of public issues in a legitimate manner, often with the (explicit) aim of contributing to society or enhancing social welfare, and are thus not limited to economic motivations' (Scherer et al. 2016, p. 276).
The heterogeneity of this definition leaves it open to political and economic adaptations for CSR. Given that the literature on the dynamics of capitalism posits that CSR is predominantly determined by individual countries (Kang and Moon 2012), the division of labour between governments, businesses and civil societies is characterised by institutionally different approaches among nations Hall and Soskice 2001), thus requiring new institutional adaptations for political CSR (Amaeshi et al. 2016). However, the concept of political CSR has provoked further thoughts on the evolving role of corporations in society. While political CSR advocates for new measures to be taken to integrate organisations into society-wide processes of political governance Palazzo 2007, 2011;Scherer et al. 2006Scherer et al. , 2016, it also raises the question of whether businesses should be subject to greater democratic control (Matten and Crane 2005;Crane et al. 2009).
For example, Hussain and Moriarty (2016) argue that the present political CSR literature considers corporations as supervising authorities capable of holding others accountable in contrast with their position as functionaries and agents who must be held accountable (see also Fooks et al. 2013;Scherer et al. 2013). In addressing the need for greater democratic accountability, Hussain and Moriarty (2016) focused on Palazzo and Scherer's (2006) model of multi-stakeholder governance to identify a fundamental democratic deficit of political CSR arising from the 'misconception' of corporations as political organisations capable of social deliberation. They argue that the current model of democratic accountability, which allows corporations to become involved in policy-making and governance structures in the same manner as political organisations and private citizens, falls short of deliberative democracy where issues of public concern are regulated by 'free, unforced, rational deliberation of citizens' (Hussain and Moriarty 2016, p. 2). They therefore propose a concept of a politically representative organisation (PRO) to refer to 'an organization that can legitimately represent a group of citizens in social deliberation' (Hussain and Moriarty 2016, p. 9). They suggest a new form of democratic accountability that views corporations as functionaries or agents to whom parties are responsible, thus limiting political CSR to collaborations with political NGOs and other groups that meet PRO standards (Hussain and Moriarty 2016).
Notwithstanding the variation of opinions in the extant literature, studies show that political CSR, including the interactions of private actors with local and regional governments to engage in activities that enhance social welfare, generally tends to reduce negative externalities (Marti and Scherer 2016;Wood and Wright 2015;Bell and Hindmoor 2009;Levi-Faur 2005). Given the prospects as well as challenges of political CSR in governing externalities, and as the previous sections have shown that instrumental regulation and Coasian bargaining cannot provide either sufficient or efficient governance of externalities, we proceed to extend the debate on political CSR by broadening its conceptualisation to include externality internalisation as an obligation.

Reflexive Governance Through Obligated Corporate Internalisation of Externalities
There is an emerging paradigm that CSR, as the governance of externalities, can be obligated with a variety of legal provisions ranging from reporting to prescriptive rules (Osuji 2011(Osuji , 2012(Osuji , 2015. A purely prescriptive approach does not sit well with fundamental assumptions about CSR because legal compulsion of 'social obligations' is in itself a contradiction (Osuji 2015). Moreover, it requires the regulator to overcome the many difficulties we identified earlier in this article, such as identifying and addressing potential negative impacts of business decisions on the environment and assorted stakeholders. Both public regulators and firms may lack sufficient information to identify the nature and breadth of potential negative impacts or to articulate the interests of different stakeholders like employees, consumers, residents and rival firms. Even when regulation requires the firm to consult different stakeholders, the bilateral and disconnected form of the consultations makes them less likely to have a transformative positive impact on society than a collective process involving all the stakeholders that represent the different segments of society. Nonetheless, there are legal provisions on issues such as the definition, justifications, reporting and even enforcement of CSR (Osuji and Obibuaku 2016). We therefore consider that legal provisions intended to drive reflexive governance should impose a binding obligation on corporations to identify and internalise social costs.
Reflexive governance is an inventive approach to corporate regulation and refers to the processes, procedures and relationships for mutual transformations of corporations and society. As noted earlier, reflexive governance demands that consultations are undertaken in a collective process that allows all stakeholders to be present and listen to, reflect on and respond to one another's views, in order to arrive at decisions that can positively transform society despite the disparate nature of the interests of the stakeholders who represent different segments of society. The UK National Institute for Health and Care Excellence (NICE) arguably provides a useful practical example of obligated reflexive governance. 8 The procedures designed and implemented by NICE have enhanced the regulation of complex areas through the emergence of guidelines on evidence-based best practices that relevant professionals and practitioners are compelled to be aware of and apply; knowing that failure to do so can attract legal and professional disciplinary consequences. In contrast, the 2015 Volkswagen's emission scandal highlights the absence of robust and transparent guidelines for ensuring that car pollution tests match real driving conditions. The public regulators lacked sufficient information and ability to issue such ex ante guidelines and the car manufacturers were unwilling to do so either unilaterally or as an industry. Things may have been different if car manufacturers were required by law to participate in a collaborative multi-stakeholder forum including car users, researchers and technical experts. By bringing relevant expertise, experience and evidence into the process, this would potentially bridge the gap between laboratory tests and real world conditions, and introduce an element of reflexivity into the overarching emissions control framework of public regulation.
Using reflexive law to anchor CSR as a form of reflexive governance helps us to extend Scherer and Palazzo's (2007) theoretical exposition of Habermasian ideal discourse. For example, while they relied mainly on a political philosophy theoretical framework to advocate a new role for the firm as a political actor in a globalising society, we rely on the theory of reflexive law to suggest other ways in which firms and other actors can contribute to this process. Where Scherer and Palazzo's framework is limited in terms of precise regulatory intervention and the attendant implications for firms which do or do not play the role of political actors, our obligated approach helps to regulate corporations, 'that otherwise would be impossible to regulate' (Febbrajo 1992, p. 30). Our reflexive CSR model also lessens the tension between the positivist (instrumental/normative) and post-positivist (relative) CSR framing and is particularly appropriate for complex, functionally differentiated societies where prescriptive interventions in the legal system create interference or 'irritations' in other social subsystems such as the economy or corporations (Teubner 1993). Our framing of obligated internalisation of social costs, as complementary to hard law and/or Coasian bargaining, advances the concept of the firm as a political actor, while recognising the democratic gap (Hussain and Moriarty 2016). It also draws attention to the firm's limitations in terms of societal legitimacy, potentially allowing the firm through self-regulation to achieve the objectives of political CSR, relying more on internal governance, and less on other actors, such as regulators, who are external to the firm.
We argue that the theory of reflexive law provides a powerful justification for using the corporate decision-making process to govern externalities. In particular, reflexive law avoids the regulatory trilemma, discussed earlier, by understanding and working with the autonomy of corporate decision-making processes, but also steering them so that they are more likely to identify and take account of the effects they have on their environment. The examples given above draw on reflexive governance techniques. No doubt, stakeholder advisory boards, human rights due diligence groups, and a wide array of multi-stakeholder memoranda of understanding, agreements and other soft laws have been developed to make firms across different countries accountable for their social costs. Transnational soft laws, country-specific hard laws and various combinations of the two certainly have a role to play in terms of promoting increased nonshareholder democratic control over corporate action, but on their own they are insufficient. Legal obligations grounded in reflexive law can complement these mechanisms and make internalisation of externalities more likely.
In autopoietic systems theory, which forms the basis for the theory of reflexive law, corporations are understood as the law's reconstruction and personification of the organisations which it observes in its environment and which meets specified criteria laid down by the law itself (Teubner 1993). Organisations are therefore social subsystems, which consist of linked decisions, and those decisions determine what the corporation selects as relevant from its environment, as well as how it responds to those selections. Firms are 'autopoietic' because they themselves produce the decisions of which they consist according to their own logic and without direct input from their environment. In other words, firms have qualified autonomy from their environment, which they construct within their own internal communications, according to their own procedures. Within organisations, each decision forms the premise for the decisions which follow it, which means that it serves as 'a normative point of reference' to be 'taken into account in the process of generating, recognising, and connecting operations as decisions to prior decisions' (Achterbergh and Vriens 2010, p. 157). Those decisions in turn motivate actions in the organisation's physical environment, and those actions may produce externalities because those affected have not consented to them. Reflexive law thus seeks to influence the way in which decisions are made, rather than prescribing ways of acting in pursuit of specific goals.
One decision which exercises a strong normative influence over subsequent organisational decisions is its goal. Corporate law is highly permissive with regard to goals, and only requires that decisions are made in 'the interests of the corporation', giving management a broad margin of discretion under the 'business judgement' rule (Sjåfjell et al. 2015). However, under market pressures, many corporations have adopted the goal of producing shareholder value, commonly expressed in terms of return on equity or the current share price. As systems of recursively linked decision premises, corporations tend to continue to do things which worked in the past, unless and until a decision to do things differently can be justified and gains acceptance among decision-makers. The main insight that systems theory contributes to corporate governance regulation is that corporations observe law in their environment and reconstruct its demands. Corporations then decide whether and how to comply with the law. Since law can threaten sanctions (which corporations reconstruct as a financial cost), these will be taken into account in making decisions, but a regulator can never be sure how a corporation will respond to a particular law. Where law is instrumental, corporations may make decisions to act in ways, which frustrate the regulatory goal, even if they formally comply with the regulation in question. The Basel II Revised International Capital Framework for Banks is a good example of this, with financial corporations complying with the rules about capital, but doing it in a way that moves assets and liabilities off balance sheet, thus frustrating the primary aim of the regulation, which was to control risk-taking (Tahir et al. 2017). They did this because the rules threatened their goal and other decisions linked to it-namely producing shareholder value.
Reflexive law therefore abandons instrumental regulation in favour of procedural regulation, and aims at steering the decision-making process without attempting to impose particular outcomes on it. It is a means by which a regulator can steer corporations towards greater internalisation of their externalities without producing second order effects, or distortions. In order to achieve this, we suggest two procedural norms. First, corporations might be required to consult with those who consider themselves affected by the corporation's decision-making so that they learn about their effects on their environment and identify means of internalising those effects. Second, decision-makers such as directors and managers might be required to take decisions in the interests of the corporation, but subject to an obligation to internalise any externalities of which they become aware of in the course of consulting affected groups. Procedural norms such as these would bypass the regulatory trilemma, and, by bringing the corporation and affected groups together for dialogue at a relatively early stage, would also sidestep many of the psychological and sociological barriers to 'Coasian bargaining' discussed earlier. In this way, it would mark out a middle ground between bargaining and regulation, a form of 'regulated self-regulation'.
Michel Callon's work sheds some light on how this might work, and highlights some of the advantages of this approach. Callon emphasises that externalities occur because they overflow the 'frames' used by actors and decision-makers, and are therefore not taken into account. The frame is the boundary within which the interactions in question 'take place more or less independently of their surrounding context' (Callon 1998, p. 249). What falls outside the frame is 'bracketed' and removed from consideration by the relevant actors. As corporations adopt the goal of shareholder value, they build up decision-making frames in accordance, and exclude anything that is not in accord, with the goal. 9 Thus, the solution is for decision-makers to expand their frames to include more of the externalities that overflow them. Callon describes the place where decision-makers and affected groups meet-so that this broader process of framing may occur-as a hybrid forum. This is because 'facts and val-ues… become entangled' and specialists and non-specialists have to work together to construct an image of the overflows in question (Callon 1998, pp. 260-263). The forum mixes together scientific construction of facts with decision-making and rule-making, all of which is carried out by a variety of actors with different interests and expectations (Callon and Rip 1992). 10 Whether an overflow is governed by a corporate decision, a contract or regulatory intervention, it has to be traced or mapped first. As highlighted by the focus of political CSR on collective deliberations in public-private 9 Corporate managers and shareholders have 'agreed' that managers should make decisions using a frame which includes effects on shareholders (measured by reference to the share price or return on equity), while effects on third parties will only come within the frame if management considers that they are likely to have consequences for returns to shareholders. This may be the case, for example, where a particular action is illegal or where it is likely to harm the corporation's reputation. Other consequences of corporate activity, such as long term and diffuse effects on the environment or other 'difficult to measure' externalities, will be bracketed outside the corporate frame and will not be taken into account by management. Like Coase, Callon recognises that externalities are 'the rule' rather than an exception. 10 Callon's more recent work has focused on the role of hybrid forums in constructing acceptable solutions to issues of public concern such as deep burial of nuclear waste or the spread of Bovine Spongiform Encehpalitis (BSE) in the United Kingdom (Callon et al. 2009).
representative organisations (Young 2004;Scherer et al. 2014), a hybrid forum helps to trace overflows and identify appropriate and mutually acceptable ways of dealing with them. This does not necessarily entail producing a bespoke way of dealing with a particular, fully specified externality. Rather, it could be a decision premise which tells corporate decision-makers how to act where they identify specific facts in the corporation's environment, or it could involve the creation of norms about how the hybrid forum ought to proceed in relation to a class of externalities. Callon and Rip (1992) note that the role of the hybrid forum (or 'expertise' as they term it) is to establish an acceptable alignment between what one knows (or believes one knows), what the actors want and expect (which may be contradictory) and the procedures to follow in order to establish norms. They argue that norms that emerge from these forums must be 'scientifically plausible', 'socially viable' and 'juridically acceptable'. Where all three criteria are satisfied, the norm will 'stabilise for a certain period an agreement on what one knows, what is socially acceptable and the rules for reaching agreement' (Callon and Rip 1992, p. 153). Like corporate decision premises, these norms are always revisable, and represent an arrangement which is 'by no means perfect, but is acceptable' because it is 'collectively elaborated, constructed, and by which we reconcile our differences, at least for a limited period' (Callon and Rip 1992, p. 154). Although his work is not normative and does not address the governance of externalities through the corporate decision-making process, Callon's approach informs the design of reflexive regulation designed to steer corporations to identify the social costs they create and appropriate means of internalising them.
While this is a fresh and provocative argument, it is beginning to gain some currency with policy makers, with the European Commission now defining CSR as 'the responsibility of enterprises for their impacts on society' (European Commission 2011). This change in approach is potentially of great importance as it appears to extend the United Nations' Guiding Principles on Business and Human Rights beyond human rights concerns to encompass other externalities. However, the proposal that corporations should constitute hybrid fora to guide their decision-making implies the need to develop procedures for the hybrid fora. Such procedures are lacking from the Commission's 2011 communication, which simply refers to 'a process to integrate social, environmental, ethical, human rights and consumer concerns into their business operations and core strategy in close collaboration with their stakeholders' (European Commission 2011). Research into reflexive governance emphasises that the rules governing interactions between stakeholders should be established by the participants and revised in the light of experience. Lenoble and Maesschalck (2006) describe this as a 'pragmatic' approach to reflexive governance. It extends beyond the development of 'shared understandings' and appropriate solutions to give stakeholders 'choice as to how to choose', thereby allowing them to improve and develop the very mechanisms of their participation over time.
While desirable in theory, this proposal, like all proposals for collective action or decision-making, raises some queries. For example, must everyone concerned by a particular activity agree on the facts that are constructed and the proposed solution, or will a majority suffice? What are the ways of establishing the procedural norms of the hybrid forum? There is potential here for hold-up or for more socially or economically powerful groups to dominate proceedings. It may be that these are questions that can be left to the forum itself. It may also be that the law, through the imposition of standards, can prevent unduly long, unruly or unfair participation (Sarra 2011). However, this is not a fatal objection to the suggestion, as corporate law has had to deal with similar problems regarding relations between majority and minority shareholders, and between board and shareholders, and it managed to avoid deadlock while remaining, for the most part, permissive and facilitative for effective corporate governance. We argue that this is achievable in relation to a hybrid forum.
The second query concerns how to guarantee that the forum will be a place of mutual learning and dialogue which transcends self-interest. These are challenges that all reflexive law or 'new governance' proposals must confront. It is clear that they will not be addressed through regulations and prescriptive rules because the regulator cannot anticipate the factual context of particular decisions. It is possible that the public nature of proceedings would constrain some of the most intense self-interest seeking. It may be that by embedding CSR considerations in the corporate governance process, the law brings about a change in people's conceptions of the role of corporations in society, and in the governance of externalities. However, we recognise that hybrid fora are not a silver bullet, and that corporations may not engage with them in good faith, in much the same way as many currently treat regulation. However, all of this remains to be seen and answers to these questions will only be identified through experimentation. Irrespective, an important advantage of our proposal to elevate externality internalisation to the status of a corporate goal is that it would provide a decision premise that exercises a normative influence on all corporate decision-making.
Indeed, while reflexive governance raises a number of difficult questions, it offers a number of advantages in terms of governing externalities compared with instrumental regulation and Coasian bargaining. Instrumental regulation requires the regulator to identify in advance the type of externality and to prescribe how corporations should respond to the occurrence of that kind of externality. Without detailed knowledge of the context, this is extremely 1 3 difficult. Reflexive governance has greater capacity both in terms of identifying externalities on an ongoing basis, and in terms of generating mutually acceptable solutions. A hybrid forum might also be a more effective, efficient and a fairer way of constructing facts than interest group lobbying in a political process. It also offers advantages when compared with litigation. Although judges only have to deal with one specific factual context, their adversarial procedures make courts a very expensive method of constructing facts, and give rise to antagonism between the parties. Moreover, courts are limited as to the remedies they can award. As for Coasian bargaining, it is far from clear that the parties to an externality ever get together of their own accord to agree on the facts about a particular externality. If the law were to require companies to publicly consult those who consider themselves affected, there would be a better chance of a mutually acceptable set of facts emerging, and there will be a greater scope for the parties to identify a remedy that satisfies everyone. This argument suggests that a hybrid forum may well be a lower cost means of identifying and addressing externalities than instrumental regulation or Coasian bargaining. At the very least, therefore, regulators should consider our approach as an alternative mechanism for the governance of externalities, especially where it seems likely that there are complex, 'socio-technical' externalities. In other words, reflexive governance helps to explain how hybrid fora, underpinned by procedural norms that accord with the tenets of reflexive law, would be a valuable addition to the regulatory toolbox. Table 1 presents the comparative advantages of our obligated approach in governing externalities, in relation to previously discussed options.

Contributions and Implications
The extant literature does not clarify how social costs/ negative externalities should be addressed. This article has developed an alternative framing of CSR as an obligated act of responsibility on corporations to identify and internalise their negative externalities. We argued that a business case for the internalisation of these externalities (strategic/ instrumental CSR in management scholarship) is irrelevant; for the regulator to identify social costs before they occur Transparency issues Self-interest orientated Assumes clear property rights will allow rational calculation; narrow transaction costs are the only barrier to agreement; and that parties will agree on the nature, existence and extent of an externality ex ante. Not suitable to identify and prevent irreversible harms.
Stakeholders' willingness and ability to participate Competing self-interests Potential dominance of one group Learning may not occur leading to high costs Possible inefficient procedural norms and practices Corporations may not engage in good faith rather, this internalisation approach should be the primary framing of CSR. This argument extended the emergent political CSR discourse and proposes reflexive governance/ hybrid fora as a way to operationalise our obligated approach to governing social costs. The article advances the internalisation reframing of CSR on the one hand and the implementation through hybrid fora on the other hand. This reframing rests on both ethical and self-governance premises. We argued that while our proposition may offer some degree of economic efficiency, firms should primarily internalise all identified social costs. We advance the theory, practice and policy of how to fully address social costs or negative externalities, and respond to the increasing stakeholder criticisms of firms, by arguing that since firms cannot properly neglect their externalities, they should internalise those externalities. We showed that framing CSR in this way potentially allows it to resolve the governance deficit and become a means of efficient governance of social cost provided two conditions are met. First, CSR must be understood as the responsibility of corporations for identifying and internalising the externalities their operations create. This requires corporate decision-makers to change the frames used to bring social costs within the scope of what they take into account. Second, corporations must be steered towards a socially adequate identification and internalisation of those costs by the careful use of reflexive governance. A reflexive approach to CSR would require that corporations engage in dialogue with those who consider themselves affected in order to construct the 'facts' about the externality, and then require corporate decision-makers to internalise that externality in an acceptable manner. This would arguably result in the identification and correction of more externalities. The procedural rights to participate in this process would be enforceable, whether by public or private actors. As shown in Table 1, our obligated approach is an additional method of governing social costs, to be used as an alternative or complement to the other, more conventional methods of governing externalities.
However, an important question remains. Will corporations go beyond what the law requires where their managers cannot advance a 'business case' for doing so? As discussed in this article, this is problematic from the perspective of the wider corporate governance system, which creates powerful incentives to increase short-term shareholder value. Making the (admittedly large) assumption that this barrier can be overcome, this article has argued that CSR can become one possible mechanism by which corporate activities can be governed so as to internalise social costs. In this regard, CSR should be framed and promoted as a necessary condition for carrying on in business and not an option that can be dropped at will-especially when it does not support corporate financial performance. Framing CSR in this way requires corporate management to understand CSR as the process of sustainable governance and resolution of externalities (Crouch 2006). Such an understanding of CSR provides a concrete rationale for stakeholder engagement, namely the identification and efficient correction of externalities and the social costs arising from them. Furthermore, our approach helps to deal with social costs, in a way that meets that allows corporations to make a wider contribution to the public good.

Compliance with Ethical Standards
Conflict of interest All authors declare that they have no conflict of interest.
Ethical Approval All procedures performed in the studies involving human participants were in accordance with the ethical standards of the institutional and/or ethical national research committee and with the 1964 Helsinki declaration and its later amendment or comparable ethical standards.
Informed Consent This article does not contain any studies with human participants or animals performed by any of the authors.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.