Introduction

The existence of powerful economic actors that negatively and fundamentally affect individual human lives, especially in the absence of effective public regulation, is by no means a new problem.Footnote 1 What is relatively new is policy advocates’ use of ‘human rights’ as the political concept and political project, which they think is well suited to deal with such problems. Over the past 5 to 10 years, NGOs, UN bodies, and policy think-tanks have argued that large, powerful corporations that frequently operate across national borders ought to have duties, similar in some respects to states’ duties, not to violate ‘human rights’ (International Council on Human Rights Policy, 2002; Christian Aid, Amnesty International and Friends of the Earth, 2005; Weissbrodt, Reference Weissbrodt2005; Ruggie, Reference Ruggie2007; United Nations, 2007). They prescribe, for example, that there ought to be legislation or legal interpretation in states that are typically thought of as ‘human rights friendly’, such as the UK. This would allow victims of actions that would typically be considered human rights violations when performed by government agents (for example, disappearances or arrest without due process) to bring tort claims against UK-based companies for those companies’ actions, even when the actions occur outside of the UK. In more specifically international terms, advocates also think that there ought to be the institutionalization of hard accountability rules, as opposed to soft standards, at the international level that aim to affect corporate decision-making directly, such that ‘private’ human rights violations are prevented before they happen. More briefly, advocates argue that the international human rights regime developed in the aftermath of WWII to ascribe human rights duties to states ought to be extended to include transnational corporations (TNCs) (particular non-state actors). This innovation demands analysis.

In fact, these ‘TNC and human rights’ policy arguments have been analyzed comprehensively, though certainly not definitively, from an international-legal perspective (see in particular, Ratner, Reference Ratner2001; Alston, Reference Alston2005; Clapham, Reference Clapham2006; De Schutter, Reference De Schutter2006). Analyses and debates from within an international-legal perspective ask particular kinds of questions, which I call ‘how’ questions: questions about how best to institutionalize international or global principles, so that they become rules that are followed in practice. Although not entirely unsympathetic to this latter project, this paper takes a step back from the legal and institutional perspective, and argues that a move to ‘how’ questions is premature. The idea that non-state actors ought to be ascribed legal human rights obligations first ought to raise ethical and political questions about the justifiability of this practice: questions for international political theory.

When one is looking at political contexts where sovereign states are requiring or directly enabling corporations to commit actions that could prima facie be considered violations of someone’s human rights, then this ‘bad state’ political context makes it difficult to justify the adoption of international or extraterritorial rules that aim to circumvent these states, attempting to regulate corporations’ actions directly. One line of argument taken by advocates of identifying TNCs as potential human rights violators is that a sovereign’s rules governing private action, including TNC action, should not be taken as authoritative when those rules prescribe action that is likely to result in human rights violations. Advocates think that it would therefore be appropriate to create a set of ‘actually authoritative’ rules that tells private moral agents when they are required to ignore sovereigns’ bad rules. Advocates think that one possible and practical way of ascribing human rights duties onto TNCs in a legal manner would be for some states to legislate extra-territorially. TNCs can be a special focus of this, precisely because of their transnationality. However, the types of rules that the ‘TNC and human rights’ advocates prescribe cannot effectively give rise to authoritative legal duties, because conflicting de facto authorities with conflicting sets of rules is bound to confuse, rather than to assist, private decision-making agents when they decide how they ought to act. As this paper will show, this runs contrary to the very point and purpose of the rule of law. Many of the arguments that advocates now put forward are not successful. However, they are unsuccessful in ways that merit analysis, and they are unsuccessful in ways that open up important directions for future research. Although this article uses the specific example of the ‘TNCs and human rights’ project, there are broader critical implications for anyone wishing to take an international-legal perspective on international politics, as well as for anyone wishing to think, across disciplines, about the human rights obligations of non-state actors.Footnote 2

The first section presents an example that will be used at several key points in the paper to test the reasonableness and justifiability of the policy prescriptions on the table. The second section discusses key theoretical concepts, such as sovereignty, authority, and the principles of legality. The third section discusses the way that human rights principles might affect states’ (on one hand) and individuals’ (on the other hand) duties differently, then goes on to discuss the ways in which TNCs can be viewed as relevantly state-like in some ways and relevantly individual-like in others. The fourth section further complicates the analysis of ‘what type of agents TNCs are’, by introducing the issue of their transnationality, and discussing the way that the legal status quo about jurisdiction over corporations handles this transnationality. The fifth section explains that a key difference between human rights discourse and corporate social responsibility (CSR) discourse is that when using human rights language, one is distinguishing, both analytically and in terms of policy prescription, required or prohibited actions (in the case of human rights) from actions that are legitimate or permissible (in the case of CSR). The sixth section then brings together all of the theoretical issues, using the political context of bad states to assess directly the main prescriptions that the ‘TNC and human rights’ advocates put forward.

The example of ‘Yeehaw’ in China

Consider the following example. It is hypothetical, but its key features are closely based on a plaintiff’s allegations in an actual lawsuit filed in US District Court in San Francisco in April 2007 under the Alien Tort Claims Act (Cha and Diaz, Reference Cha and Diaz2007). I have (rather thinly) disguised the identity of the company against which allegations were laid. However, if one takes the plaintiffs at their word, the core of the problem outlined below is a real one, confronted by decision-makers and policy-makers in today’s world.

A global information-technology company called Yeehaw is based in the UK but has operations and subsidiary offices around the globe. The company generates most of its revenue by selling advertisements that appear when internet users use the company’s search engine, which is one of only three internet search engines that has a globally known brand and is widely used. The company offers users free e-mail accounts and user names, and these same user accounts can be used to participate in internet groups organized around common interests and to operate personal blogs. The Chinese government notices that an individual (call him X) with a Yeehaw account is operating a blog under a pseudonym and making anonymous postings to groups containing opinions about multi-party democracy that are illegal to publish in China. Other internet users (call them Y and Z) who are also in China, but who do not have Yeehaw accounts, regularly read and respond to X’s blog and group postings. According to an official Chinese legal process, the Chinese government contacts Yeehaw’s Chinese office (call it Yeehaw China), requiring that the office release the identity and all known personal details of X. It is widely known by all parties involved that if X’s identity is discovered, then X will be arrested by the Chinese authorities. Furthermore, without what liberal-democrats would consider due process or a fair trial, X would be imprisoned for 10 or more years. In recent months, others in a similar situation to X, once arrested, were put under severe psychological pressure – some might say tortured – in order to determine the identities of individuals such as Y and Z. Yeehaw’s regional director for Asia is named Susan. She is a UK national who has been seconded to the Chinese office from her base in London, where she is planning to return (her whole family is there) in about 2 years. Susan, acting on behalf of Yeehaw, needs to decide what to do.

This is an example that considers the actions of a company that has pieces of itself in two different sovereign states. One of these sovereign states requires non-state actors to perform actions that result in human rights violations, and the state is willing to punish a non-state actor for not contributing to a likely human rights violation, such as arrest without due process. This is the paradigm of the ‘bad state’. It is important to note at the outset that the bad-state paradigm has wide application, despite being under-explored in both theory and practice with respect to TNCs. For example, in the UK, the police can hold terrorist suspects for 28 days without charging them with any offence, and the government has recently tried to extend this period to 42 days (BBC, 2008; see also Abbott, Reference Abbott2008). Authorities in countries such as the USA and the UK could, and probably do, ask internet companies and telecommunications companies for the personal details of suspected terrorists. In the UK, this has the added implication that the suspect could be held for 28 days without being charged with any offence. If a company were faced with such a request, in such a context, what should it do?Footnote 3 To dispel any doubts about the relevance of the bad-state paradigm, consider that we now live in a world in which private security companies (Abrahamsen and Williams, Reference Abrahamsen and Williams2007), private military companies (Singer, Reference Singer2003; Kinsey, Reference Kinsey2007; Pattison, Reference Pattison2008), and private jails (Clapham, Reference Clapham2006: 494–499) all exist. What if a multinational private security company, headquartered in a country where certain kinds of discrimination are illegal on human rights grounds, were required to enforce rules that discriminated against one racial, ethnic or gender group – for example, by preventing certain people from voting or from driving – because of government-sanctioned social practices in a particular host state? What if a transnational private military company, headquartered in the USA, received a request from the American government to capture an individual in Iraq and to transport that individual to Syria for torture?

All states that have some degree of empirical control over their territories, and that have de facto authority over the non-state actors within those territories, have the potential to be bad states. The ‘bad state’ is an ideal type rather than reflective of any one actual state or group of states. Real-world situations will approximate the bad-state paradigm to a greater or lesser degree, depending upon the context. The logic of the bad state needs to be developed sharply and separately from other logics, for example, the logic of the weak state – in which a corporation can do what it wants, irrespective of the host state’s authority – so that those logics can then be fitted together properly, and to the appropriate degree, in real-world situations.Footnote 4 The internet company in China, which is based on a real legal case rather than on the countless hypothetical ones that one could generate, and which also exhibits to a high degree the issues that come into focus when looking at the bad-state paradigm, will remain the focus of this paper, but the broader implications are important to keep in mind as the argument develops.

Sovereignty, authority and the principles of legality

There is a prima facie case that Yeehaw does the right thing if it simply defers to and complies with the Chinese government’s request, and I shall consider the prima facie case before discussing what might be wrong with it. Yeehaw does the right thing by handing over X’s personal details if it is true that moral agents in general – and private-sphere economic actors such as companies in particular – ought to follow the positive law (that is, ought to abide by their legal duties) in whichever sovereign state they happen to be physically located. There might be good normative reasons for emphasizing the unique right of sovereigns to define the legal rights and duties that apply to private-sphere agents who are resident within a sovereign’s territory. Presenting this idea in its best light requires first remembering the international political and economic context of the 1970s. During the colonial era, companies in host states had typically operated as extensions of the home state’s economic, political and legal system (Donnelly, Reference Donnelly1998: 3–13). A substantial matter of interest for many people concerned with the ethical dimensions of globalization in the context of the aftermath of decolonization was to ‘rein in’ (Ratner, Reference Ratner2001: 457) the power of multinational corporations, which, despite the colonial era having formally come to an end, were still seen to represent the interests of their (usually Western) home states. A forceful argument was put forward that companies operating in this context ought to be non-political private-sphere economic actors; they ought to comply with, and not to interfere with, host-state social, political and economic decisions (Ratner, Reference Ratner2001: 454–457; Zerk, Reference Zerk2006: 9–10).

In international relations (IR) thought, the institution of sovereignty is often conceptualized as bound up with the normative principle of political authority. The authority of the domestic is compared and contrasted with the anarchy of the international. Bull (Reference Bull2002/1977), for example, has referred to this idea as the ‘internal sovereignty’ of the state; Krasner (Reference Krasner1999) talks about ‘domestic sovereignty’; Jackson and Rosberg (Reference Jackson and Rosberg1982) refer to political authority as one of the ‘empirical’ characteristics that juridical sovereigns are ideally (or idealistically) taken to possess; and Strange (Reference Strange1996) thinks that this ‘authority’ aspect of sovereignty is the very thing that has been going away since at least the 1970s, and is certainly gone since the end of the Cold War (see also Karp, Reference Karp2008). However, in IR thought, especially since the 1990s, the principle of political authority has been raised as part of a particular normative discussion that identifies states and supra-state bodies as agents to which such principles apply: the discussion about whether and when external (outside) intervention in the affairs of sovereign states would be legitimate. The question of how legitimate political authority might apply to transnational residents (insiders) of sovereign states, helping them to decide whether and when a sovereign’s requirements are legitimate, and therefore what duties they really have as a result of the sovereign’s sovereignty, has been under-explored in specifically international political thought.Footnote 5 This article’s first step is to begin to fill this gap in theoretical IR literature. The paper will first consider how the principle of political authority matters with respect to residents in general, before moving on, in the next section, to complicate the picture by introducing the idea of the transnationality, as well as the corporate (as opposed to individual) nature, of some of those residents.

Anarchists argue that no law or rule can be authoritative: that autonomous moral agents would always be better off if they could decide for themselves what they ought to do, and always worse off if required to defer to the moral reasoning of an external decision-maker. There are sophisticated arguments that one can present in defence of this position (Wolff, Reference Wolff1970), requiring anyone who believes that political authority is justified at least some of the time to explain whether and why this is so. Part of the job of political morality is to define whether, why, and in which cases, authoritative decision-making is legitimate. Doing this job well means recognizing that in many decision-making contexts, authority has no place: that often (if not necessarily always), a decision-making agent ought only to exercise her own judgment, and ought not to defer to any authority other than herself when taking decisions about what to do.

Indeed, for anarchists and non-anarchists alike, at least some duties ought to be discretionary for private moral agents. If I pass someone who asks me for money on the street, my own values, beliefs and legitimate ends are relevant to an analysis of whether I have a duty of charity in that instance. Perhaps, for example, the only money that I have with me is money that I am about to use to buy a plane ticket to visit my gravely ill mother, and in this case I do not have a duty to give the beggar money. By contrast, an investment banker walking just behind me, without a gravely ill mother, who is about to spend $200 on cocktails in an upscale bar, might very plausibly have a duty of charity in that instance, depending upon the ‘strong evaluations’ (Taylor, Reference Taylor1985) and reasons for action involved in his decision. Agent-relative factors, such as how each agent has formulated his or her own set of (legitimate or illegitimate) beliefs about appropriate values and reasons for action, are often relevant to our decisions about what we ought to do. For epistemic reasons, among others, these are not the sorts of decisions that can be legitimately taken in advance by external authorities through the creation of a set of agent-blind rules that apply equally to all those present within a particular social context.

The debate between anarchists and non-anarchists is about whether this category of some duties actually exhausts all duties. Philosophers who believe that deferring to political authority is appropriate at least some of the time need to provide an account of why decision-making discretion can ever justifiably be removed from individual moral agents and placed in the hands of political institutions. That is, they need to provide an account of how discretion over how to weigh up the values and reasons that apply to a certain category of duties, such as the duty to pay taxes, the duty not to assault others, and the duty to drive 100 km/h (60 mph) rather than 140 km/h (85 mph) on the highway, rests appropriately with an external decision-maker, rather than with those individuals who actually have to take decisions and actions about, for example, how best to spend the money they earn and how to drive.

Raz (Reference Raz1986: 23–105) provides one prominent account of authority. He explains that sometimes decision-makers are not best situated to assess how to act on the reasons that apply to themselves and their decisions. In these cases, and only in these cases, they ought to defer to external authorities that are so situated, rather than deciding for themselves what they ought to do. Sometimes our instincts, emotions, beliefs and rational decisions lead us to the right actions. If I am hungry, I look for something to eat (and I probably ought to do so); if I want to become an academic, I try my best to complete a Ph.D. (and I probably ought to do so); if I see a person who has had an accident on the road, I stop to see if I can help (and I probably ought to do so). However, sometimes our own decision-making apparatus can lead us in the wrong direction. I do not want my driving to cause needless deaths. I might also believe that 140 km/h is a perfectly safe speed to drive at, because my previous experience tells me so. Or, I might not believe that 140 km/h is a safe speed to drive at, but I might nevertheless drive at this speed from time to time, because I do not, each and every moment that I drive, reflect upon my evaluation that I do not want my driving to cause needless deaths. In any case, I have neither the time, the resources, the skills nor the will to investigate whether the speed that I believe to be safe is actually safe. Therefore, according to a Razian account of authority, I ought to defer to the authority of an external body that does have the time, resources, and so on, to look into these matters. I ought to follow whatever speed limit that body judges to be safe.

Raz’s account of authority, though important, is neither uncontroversial nor the only one from which to draw (see the essays in Raz, Reference Raz1990). The account just given of his ideas, though necessarily brief and non-comprehensive, should still suffice to identify an important general feature of political authority as it applies to residents within a social context, rather than as it applies to states and/or supra-state actors (particular outsiders). Political authority is legitimate to the extent, and only to the extent, that one can provide an account of why, when deciding what to do in given decision-making contexts, a private moral agent ought to circumvent, rather than to take into account, his or her own desires, instincts, beliefs, and other legitimate values and ends: deferring to an external agent’s judgment about which decision one ought to take. It is worth noting that this is just as much an account of why some rules are not authoritative as it is an account of why some rules are. For example, I ought not to obey an external body’s judgment if it says that all red cars must drive on the right side of the road and all blue cars must drive on the left side. An anarchic situation of everyone deciding for themselves what to do would be preferable to living in a society in which everyone followed this sort of legal rule.

Waldron (Reference Waldron2008) adds another dimension to these ideas about whether, when and why an agent resident within a social context ought to defer to the authority of political institutions. He aims to provide a middle way through an old debate between Fuller (Reference Fuller1969) and Hart (Reference Hart1997/Reference Hart1961; 1965) by drawing a sharp distinction between (on one hand) legal systems, within which the ‘principles of legality’ are instantiated, and (on the other hand) other forms of non-legal political rule, for example, rule by decree. He argues that part of the law’s point and purpose is to provide the residents within a social context the benefits of the rule of law, allowing them better to plan their lives and better to predict the consequences and the implications of their actions. Individuals, and companies for that matter, who live within a legal system can predict that contracts will be honoured. They can attain property and predict with a reasonable degree of certainty that their neighbours will not rob them the very next day. They can know what taxes they will have to pay at the end of the year. They can know which actions have been criminalized, the performance of which is likely to result in serious consequences.

Waldron argues that societies, for example, in which rules are made after the fact, in which rules are decided upon privately and not communicated publicly, or in which punishment is arbitrary – that is, societies which are not governed in accordance with the principles of legality – are not societies with ‘legal systems’-proper at all. The point is that if the sovereign’s rules confuse, rather than assist, private moral agents, when they decide how they ought to act, then these rules do not have a legal character at all. Therefore, they are not even candidates for next-level discussions about, for example, the rules’ authority or validity as far as residents are concerned. The principles of legality are not necessarily the thin end of the wedge that can be used to drive a fuller range of moral considerations into the question of whether rules are or are not legal.Footnote 6 The claim, rather, is that there is a specific category of moral principles that help to decide whether rules have a legal character at all. If rules have a legal character, then further discussions about the moral justification of those rules become possible and relevant. If not, then those latter discussions become moot, at least within a law-based framework.

The account of the distinction between discretionary and non-discretionary duties can now be summed up as follows. With respect to discretionary duties, such as duties of charity, there is a good prima facie case that moral agents ought to be at liberty to rely upon their own judgment when taking decisions about whether, when and how to act on these duties. Non-discretionary duties are moral duties of a different kind, and they raise two further questions. Firstly, they can correctly be classified as legal duties only when they are institutionalized in accordance with the principles of legality (and they then might need to meet further tests, for example, of validity or integrity, which are not of concern in this paper). Secondly, one needs to ask which non-discretionary duties ought, in principle, to be institutionalized. Kant, for example, famously cites the duty not to commit suicide as a non-discretionary duty that, despite being non-discretionary, ought not to be legalized (2008/1785, Section 2).

What light can these ideas – about whether, when and why private moral agents resident within a sovereign state ought to follow a sovereign’s rules – shed on an analysis of what Yeehaw ought to do (an agent-centred question), and on an analysis of which rules purporting authoritatively to govern Yeehaw’s actions would be justified (a structure-centred question)? To answer these questions requires firstly a look at the ways in which corporations are like, and not like, individual persons who sometimes live within a social context (or a legal system) within which legitimate political authority exists. Secondly, the answer requires a look at the contemporary state of jurisprudence, according to which parents and subsidiaries of TNCs are regarded as separate decision-making agents residing in separate sovereign jurisdictions. The next two sections will deal with these questions in turn. Following this, I shall tie these tools and issues together by returning to an analysis of what Yeehaw ought to do, and of whether and how extraterritorial or international-level rules governing its human rights-relevant actions can be justified, given that TNCs sometimes operate in states that exercise de facto authority illegitimately.

TNCs, states and individuals

International political theorists should be familiar with the main critique of the idea that absolute political authority (the right of sovereigns to do whatever they want inside their borders) is a normative principle that always justifiably goes hand-in-glove with the institution of sovereignty. As the events of World War II demonstrated so vividly, sovereigns do not always exercise their sovereignty legitimately and responsibly (Arendt, Reference Arendt1968: Ch. 9; Hobsbawm, Reference Hobsbawm1994: Chs. 1–5). The principles of legality are not always instantiated in domestic political systems, and therefore neither are the benefits to moral agents that would come with living in a society governed in a robust sense by the rule of law. We live in a world of non-ideal, rather than ideal, sovereigns. Just as agents need clear principles and rules to tell them what they ought to do when sovereignty is working the way (or closely to the way) that normative theory says it ought to, agents also need principles and rules to help them to identify when – and to help them to define what they ought to do when – sovereignty has backfired: when the concept’s descriptive (institutional) sense does not match up with its normative (principled or prescriptive) sense (Reus-Smit, Reference Reus-Smit2001; Karp, Reference Karp2008).

Rawls (Reference Rawls1999: 89–120) has given us two categories to think about ways that sovereignty’s descriptive and normative senses can be mismatched in the way that I have just laid out. He discusses ‘outlaw states’ and ‘burdened societies’: the former lacking the willingness to abide by internationally accepted principles, and the latter lacking the capacity to do so. These categories are useful, but they are embedded in Rawls’ particular theory of international law, institutions and rules. Rawls’ theory is concerned to examine the duties of ideal states given their co-existence with non-ideal ones.Footnote 7 Framing the issue in this way, as Rawls and many others have done, leads one to ask when sovereign states, as the agents in question for whom rules and principles are needed, can legitimately intervene, militarily or otherwise, in the affairs of other sovereign states. One of Rawls’ answers, which is also the answer in the Responsibility to Protect policy document (International Commission on Intervention and State Sovereignty, 2001), is that when non-ideal sovereigns do not protect ‘human rights’, then other sovereigns (closer-to-ideal ones) are permitted, and perhaps even required, to disregard the non-interference norm usually attached to the institution of sovereignty and to interfere. If this analysis is correct (and I need make no claim here that it is), then the difficult work lies in defining with more specificity what human rights are, such that their violation requires or permits some sovereign states to disregard the sovereignty of others.

One needs to take an important step back from these issues, to consider what is missing. It important to note that sovereigns in general, and ideal (or closer-to-ideal) sovereigns in particular, are not the only sorts of agents that can have duties that come into focus in situations of mismatched sovereignty. Another important angle to this question about the circumstances under which disregarding the sovereign’s authority is warranted – an angle that thinks outside of sovereign states as the only possible duty-bearers – is to look at the duties of individual residents in circumstances where states’ rules are bad. For example, consider the duties of the ‘Grudge Informer’ (Fuller, Reference Fuller1969: 245–253).

A Grudge Informer could be a person who has harboured a long-standing grudge against his neighbour and who takes advantage of a bad state’s, for example, Nazi Germany’s, laws by reporting to the secret police that the neighbour is hiding Jews. The Grudge Informer has fulfilled the legal duties that the sovereign laid down for him, but so what? Precisely because one is looking at the context of a bad state, the question of whether the Grudge Informer ought to be punished by a more ideal sovereign for his actions – either by a new regime that later takes over the governance of the same society (this is Fuller’s question), or by an external legal system such as the Nuremberg tribunals or such as another sovereign state (see Arendt, Reference Arendt1963: 120–134 and 213–256) – is highly complex. One can make a plausible case that the Grudge Informer had a duty to ignore his legal duties in that circumstance, and that asking whether one’s actions are likely to result in ‘human rights violations’ is one factor that can help to define the threshold of whether, when and why individuals ought to ignore, or ought not to accept, the legal duties that the sovereign has laid down for the moral agents that reside in the sovereign’s territory. Pogge (Reference Pogge2002: 52–70), for example, has argued that all moral agents have duties not to provide material and/or political support to any institution that is human rights violating. The human rights regime’s purpose, according to this line of thought, is to specify a category of rules that should never be taken as authoritative by any agent; put another way, its purpose is to specify a category of sovereign rules that can never have a truly ‘legal’ character.

The question of whether and when to ignore the sovereign’s authority raises problems for all moral agents. However, particular problems are raised for moral agents that are neither individuals nor states. When looking at corporations, it is unclear a priori whether one should start with the presumption that they are state-like, and thus ‘move down’ from the first (international-interventionism) discussion of whether and when sovereign states can legitimately override the sovereignty of others, or whether one should start with the presumption they are individual-like, and thus ‘move up’ from the second (Grudge-Informer-type) discussion about when moral agents are permitted or required to ignore the sovereign’s rules even though the rules seem prima facie to be authoritative. This question is central.

The answer ends up being both. Just as light can appear as a wave or a particle, depending upon how one looks at it, TNCs have theoretically relevant dualities that become more or less apparent, depending on what one is looking for, and also depending upon the different contexts in which they operate. As I shall now outline, they can be viewed: as agents with or without capacity to stand up to the state, as single or multiple decision-makers, and as political insiders or political outsiders.

TNCs sometimes have a greater capacity than individuals to intervene in sovereign states’ affairs and decisions outright: to take positive action rather than simply not to act. With a greater degree of safety to themselves and their own legitimate functions, than, for example, the individual Grudge Informer can afford, TNCs’ actions can go beyond simply ignoring prima facie legal duties for principled reasons. The individual Grudge Informer’s duty, if he has one, not to report his neighbour to the Nazis, cannot be smoothly extended into a more robust duty to protest against the Nazis in the streets. This latter action might very well get him shot. Of course, corporations are non-physical entities that cannot get shot (though their managers and directors might), but more significantly, corporations often have power vis-à-vis their host states: power that individuals, because of the natural facts about them, often lack. In this ‘capacity’ respect, TNCs are, or at least can be, more similar to states than to individuals (see Goodin, Reference Goodin1988; Miller, Reference Miller2001: 460–462; O’Neill, Reference O’Neill2001; Wenar, Reference Wenar2007). The relevance of this ‘capacity’ aspect comes mainly into focus when one views corporations as having political and/or economic capacities greater than those of the host states in which they operate. That is, it comes into focus in the context of a weak state. The facts of the matter about internet companies in China bring the other, ‘bad state’ option into focus: corporations are sometimes weaker than the states in which they operate.

Another issue that has been the subject of considerable debate is whether and when it is appropriate to single out individuals within corporate entities for special responsibilities – for example, in Nazi Germany, the individual train conductor (MacIntyre, Reference MacIntyre1999), the individual officer (Arendt, Reference Arendt1963), or indeed, the individual Grudge Informer – and whether and when, on the other hand, it is appropriate to hold a corporate entity as a whole to account for its actions and decisions. A key implication for ‘TNC and human rights’ issues is that different kinds of international law become more or less relevant as baseline paradigms, depending on which of those two perspectives one takes. Individual persons can have their liberty removed when they commit serious crimes; for example, they can be put in jail, whereas socially constructed collective agents such as corporations and states cannot (Searle, Reference Searle1995; Karp, Reference Karp2009). One often needs to think outside of criminal-law categories when looking at how corporate entities can be held accountable for any seriously wrong actions. If they are to be punished, they need to be punished differently. In terms of their corporate, socially constructed nature, TNCs are therefore more similar to states than to individuals, but in some respects, both companies and states exhibit an individual-like capacity for moral agency, as other research has argued in more detail (French, Reference French1984; Erskine, Reference Erskine2003; Wendt, Reference Wendt2004; Pettit, Reference Pettit2007). International human rights law (unlike domestic or international criminal law), designed as it is with states in mind as the paradigmatic violators, is a framework that has already worked through the implications of holding to account corporate entities rather than individual persons. If one’s focus is on holding individual persons within a TNC to account, then international criminal law, though surely not perfect, seems at first glance like an appropriate paradigm with which to begin. If, however, one’s focus is on holding the TNC as a whole to account as a corporate agent, then the use of ‘human rights’, rather than criminal law, looks more plausible as a starting point, because it is already designed to hold to account corporate entities (states being the original paradigm for this), rather than individuals within those corporate entities, for breaches of duties.

These ‘capacity’ and ‘corporate nature’ issues have not been answered exhaustively, but discussions of them are well under way. The major under-analyzed gap, and here is where the primary focus will now turn, is that TNCs are also private moral agents that are subject to one or more sovereign states’ jurisdiction in ideal circumstances. When sovereignty is working in a normatively robust sense (Karp, Reference Karp2008), corporations have, and probably ought to have, legal duties in the first place, which arise from being resident in one or more sovereign jurisdictions. When states go to war with other states or otherwise interfere in each others’ affairs, they are overriding other states’ sovereignty in general, or as a whole, but, as non-residents (non-insiders) in the other state’s territory, they do not override any particular conflicting non-discretionary duties when they disregard the sovereignty of others. TNCs, by contrast, have particular non-discretionary duties that go along with being within the sovereign’s sphere of legitimate authority (when and where it is legitimate), for example: the duty to pay taxes, the duty to honour a contract to build munitions for use in war or the duty to share the personal details of potential terrorists with public authorities. In these respects, TNCs are more like individuals than like states, since the latter cannot ever be resident or present in another state’s territory and therefore cannot be subject to the non-discretionary duties that residence or presence in a territory can confer, sometimes legitimately, for reasons explored in the previous section. Whether the weak-state paradigm or the bad-state paradigm comes into focus depends upon the facts of the matter of particular political contexts, and the Yeehaw example brings the bad-state perspective sharply into focus.

In addition to the issues that arise because TNCs are neither individuals nor states, but are individual-like in some theoretically relevant ways and state-like in others, problems also arise from a different angle for moral agents that happen to be transnational: agents that exist in more than one sovereign jurisdiction at one time. To introduce the issue of transnationality, and to link the issue of transnationality to the ways in which corporations can be viewed, simultaneously, as political insiders and as political outsiders, I shall describe a paradigmatic legal case that illustrates how TNCs’ parent entities and each of the TNCs’ subsidiaries around the world are treated as separate decision-making agents residing in separate sovereign jurisdictions.

‘Business and human rights’ vs. corporate social responsibility

The legal status quo clashes with the facts of the matter about Yeehaw’s transnationality. Susan will probably call the head office in the UK for input before deciding what to do. Advocates think that this opens up a particular rule-making space – or, at the risk of completely mangling metaphors, creates an authority vacuum – that the human rights regime can fill, by providing a set of actually authoritative rules that TNCs ought to follow in political contexts where states’ rules are bad (see Hopgood, Reference Hopgood2006). The status quo principles of jurisdiction, and the legal facts of the matter about corporate structure, suggest that one ought to treat only the piece of the TNC that exists within one’s own jurisdiction as an insider and the rest of the TNC as an outsider. By contrast, advocates imply that the TNC as a whole, including all of its subsidiaries, ought to be treated as a political insider by those who make the rules. Advocates (explicitly or implicitly) aim to break the theoretical links, which emerged in the modern age, between political authority, territoriality/geographic space, and ‘insider’ political identity (Kobrin, Reference Kobrin1999; Adamson and Demetriou, Reference Adamson and Demetriou2007). The paper will now turn to an analysis and assessment of the particular policy prescriptions on the table in light of the transnational character of TNCs as the relevant decision-making agents for whom, advocates think, human rights principles, rather than the sovereign’s rules, ought to guide action.

I shall clearly state the principled argument that the paper now considers, keeping in mind that the paper is assessing principles that are meant to be action-guiding for residents within a social context, rather than action-guiding for sovereign states (see Meckled-Garcia, Reference Meckled-Garcia2008: 247–252). The argument is: There is a certain category of actions that, if a sovereign state’s law demands these actions to be performed, then moral agents have a duty to ignore the legal rules – defined as those duties laid down by the sovereign in accordance with the principles of legality – when determining what they ought to do. This category consists, at least in part, of actions called ‘human rights violations’. Any sovereign that authorizes human rights violations to occur within its borders is not acting with legitimate authority. In such circumstances, the non-discretionary character of what appear at first glance to be legal duties can no longer be justified, nor can the requirement that agents circumvent other sorts of considerations such as their beliefs, values, desires, and legitimate ends, when deciding what to do.

At this point, one sees a picture forming according to which Yeehaw China is permitted not to hand over X’s personal data to the Chinese authorities. In the absence of a legitimate, authoritative, non-discretionary duty – that is, a duty to obey China’s sovereign laws – which would require Yeehaw China to circumvent a moral analysis of the legitimate values and ends at stake, Yeehaw China is permitted to do ‘the right thing’. This, presumably, fleshes out as not releasing any of X’s personal details.

However, there are still at least two unresolved questions, which suggest that ending the analysis here would be premature. The first question is: If ‘human rights duties’ are the ones meant to be action-guiding for TNCs such as Yeehaw in these circumstances, are the duties meant to be discretionary or are they meant to be non-discretionary? The idea that TNCs ought to be constituted as more responsible (to a variety of stakeholders beyond shareholders), or more virtuous, decision-making agents is the very idea behind CSR discourse. If and only if a moral discussion is about discretionary duties, then the debate can legitimately be about questions such as how best to weigh up corporations’ legitimate ends, balancing, for example, responsibilities to shareholders with wider social responsibilities. The CSR paradigm is one that thinks in terms of discretionary duties: changing the corporations’ conceptions of their own legitimate values and ends, accepting that these play, and ought to play, a role in their decision-making (Zerk, Reference Zerk2006: 15–44). However, the ‘business and human rights’ discourse is presenting a very different picture: a picture according to which corporate social obligations are non-discretionary at least some of the time. If and only if the moral discussion is about non-discretionary duties, then the constitution and reconstitution of corporations’ legitimate values and ends (including the balancing of shareholder interests) would be quite irrelevant, so long as TNCs follow the right rules (International Council on Human Rights Policy, 2002). The second question is: If the duties under discussion are indeed non-discretionary, then would the creation of either extraterritorial law or international rules be an appropriate and justifiable way to institutionalize principled prescriptions in today’s international context?

Let us take the first question first. In order for an analysis of the human rights duties of TNCs to have the type of consistency and clarity that the principles of legality require – specifically, so that law-abiding agents know in advance and with transparency what they will and will not be punished for – one wants an analysis according which the non-violation of human rights by non-state agents such as TNCs is unambiguously required, and not just permitted. If Yeehaw China is permitted to ignore the sovereign’s rules, and therefore permitted to decide what it ought to do according to its own set of legitimate ends (whatever these are), then the company might decide that it ought not to hand over X’s personal details, but it might also very well use its new-found discretion to decide that it ought to hand over X’s personal details.

In this latter case, maybe Yeehaw China has made the wrong moral choice: a choice that a virtuous agent would not have made. However, to focus on this is to miss the point. The point and purpose of ‘human rights’ as a contemporary political concept is similar, at least in one respect, to the set of legal duties that arises in situations of ideal sovereignty. Would-be torturers and would-be Grudge Informers are not supposed to be thinking about and weighing up agent-relative factors such as the benefits that their actions could bring to their gravely ill mothers. One wants human rights duties to be non-discretionary; that is, one wants to define human rights duties such that agents’ own private sets of values and legitimate ends are circumvented, rather than weighed up, in their overall considerations of what they ought to do. This is one aspect of the political project of ‘human rights’ that distinguishes human rights duties from moral duties in general (Arendt, Reference Arendt1963: 120–134, 232–256; Freeman, Reference Freeman2002: 32–42; Nickel, Reference Nickel2007: 7–14, 70–82). In principle, the CSR discussion needs to be circumvented here, rather than settled in one way or the other, because the human rights discussion is about non-discretionary rather than discretionary duties.

There is, therefore, a significant problem with stopping the analysis by concluding that Yeehaw China is permitted to ignore China’s requirement to hand over X’s personal details. The concept of human rights suggests that human rights duties – whatever they are – ought to require and not just to permit certain forms of action or abstentions from action. To the extent that any discretion exists over what human rights duties one has, and over what constitutes a violation of these duties, the discretion should not ultimately rest with private moral agents who are themselves, in other circumstances, permitted to weigh up their legitimate, sometimes conflicting, values and ends in making moral decisions about what to do.

It should now be clear that human rights duties are similar to the legal duties that arise in situations where legitimate political authority is exercised in the sense that discretion over this category of obligations does not rest with the decision-making agent. However, human rights duties differ from legal duties in a key sense. Discretion does not rest with the duty-bearer, but it clearly cannot rest with the sovereign (the default public authority) either, as the latter is the paradigmatic potential human rights violator (Nickel, Reference Nickel2007: 38–41; compare Donnelly, Reference Donnelly2003: 33–37 and Arendt, Reference Arendt1968: Ch. 9). Different types of rights theorists have tried to account for the non-discretionary character of human rights duties in different ways. ‘Choice’ or ‘control’ theories of rights, for example, hold that the discretion rests with the right-holder and not with the duty-bearer (Hart, Reference Hart1982: Ch. 7). ‘Interest’ theories of rights hold that discretion over rights-based duties has to be fleshed out against objective facts of the matter about human beings (Raz, Reference Raz1986: 165–192); discretion is thus removed from epistemically fallible humans, and the ‘decision-maker’ becomes external reality (the moral and empirical facts of the matter about what’s important to human beings) itself. ‘Institutionalist’ theories of rights argue that human rights are whatever they have been defined as, accepted as or agreed as, by particular types of social institutions (Bellamy, Reference Bellamy2001: 247); the question then becomes about the legitimacy and character of the institutionalization process, and perhaps about the ‘secondary rules’ or ‘constitutive rules’ according to which institutions can be created and changed (Hart, Reference Bellamy1997/1961; Bellamy, Reference Bellamy2001: 246). Without entering into this (often intractable) discussion, I wish to note that the striking feature about it is how much debate has occurred about how to specify, in more precise detail, a feature of human rights that all sides accept as a starting-point: that human rights give rise to duties over which duty-bearers themselves do not have discretion according to the values and ends that they hold as private, autonomous moral agents.

Therefore, whereas CSR language identifies as central the legitimate values and ends that corporations take themselves to have, particularly when they perform actions that might have social and/or environmental impacts, the emerging ‘TNC and human rights’ paradigm is something quite different. The use of the human rights concept points to a different, narrower, range of actions that TNCs might perform or fail to perform: a range of actions over which TNCs’ conceptions of their legitimate values and ends are not central – indeed they are quite irrelevant – so long as the authority of human rights ideas and/or institutions is accepted and followed (Hopgood, Reference Hopgood2009).

This brings us directly back to the problem of Yeehaw and the policy prescriptions that have been put forward in an attempt to resolve it. Advocates think they have arrived at particular answers to the debate about how to ascribe non-discretionary human rights duties onto corporations in a justifiable manner. The principal policy prescription on the table is that TNCs’ home states ought to hold parents accountable for subsidiaries’ actions (overhauling the status quo rules about corporate residence/presence and sovereign jurisdiction). A second policy prescription on the table is to create and/or to expand existing international human rights mechanisms, which are currently designed to hold states to account for a broad range of human rights violations – and, as a more recent development, to hold individuals to account by way of the International Criminal Court for a relatively narrower range of actions – using these to identify TNCs as unique kinds of violators, taking into account, that they are neither states nor individuals (Ratner, Reference Ratner2001: 489–522). Both types of prescriptions are put forward according to the idea that the legitimate regulation of transnational actors is part and parcel of the emergence of a ‘global rule of law’ (Teitel, Reference Teitel2005), which explicitly aims to detach the ideas explored in pages 92–97, about authority and the principles of legality, from the status-quo institutions examined in pages 102–104: state sovereignty and territorial jurisdiction. This move would be in line with a great deal of ‘globalization’ discourse, which argues that authority is shifting to levels above and below the state (Strange, Reference Strange1996; see also Krasner, Reference Krasner1999; Karp, Reference Karp2008).

The ‘bad-state’ context

The extraterritoriality prescription – the idea that UK institutions should hold Yeehaw accountable, within the UK, for actions that occur outside of the UK – is an attempt to export the USA’s Alien Tort Claims Act (ATCA) paradigm to other legal jurisdictions: especially to Europe. In the United States, the ATCA represents the domestic institutionalization of international-law principles, allowing American courts to justify extraterritorial jurisdiction over tort claims where the wrong committed is a fundamental violation of international standards, such as piracy or torture. The ATCA came about in the 18th century, when pirates operated outside of anyone’s sovereign jurisdiction, so that victims could bring pirates to a court of law somewhere to claim compensation for damages. This was preferable, in principle, to allowing victims to take the law into their own hands, outside of legal institutions. The ATCA has been resuscitated in the late 20th century and applied to ‘TNC and human rights’ cases (Shamir, Reference Shamir2004; Steinhardt, Reference Steinhardt2005: 196–198; Teitel, Reference Teitel2005; Clapham, Reference Clapham2006: 252–63, 443–450).

Let us consider the implications for the Yeehaw situation if the UK were to have law or legal interpretation similar to the ATCA, allowing UK courts to justify jurisdiction over extraterritorial actions in cases of human rights violations where those violations occurred in ‘bad states’ in which there is no domestic accountability mechanism. What if Yeehaw China decided to release X’s personal details to the Chinese authorities, resulting in his arrest without due process and also in his torture? X could eventually hire lawyers in the UK (or, NGOs could hire lawyers to take up X’s cause), and these lawyers could sue Yeehaw’s UK headquarters for an amount that threatens it with bankruptcy (holding the corporation as a collective entity accountable), as well as sue personally Susan, the regional director, who was responsible for taking the decision to release X’s details to the Chinese authorities (holding individual decision-making agents within the corporation accountable). Western countries have here a method according to which they could ‘get’ Yeehaw and ‘get’ Susan. Practically speaking, someone can be held accountable for unpunished abuses that occurred abroad. The UK could perhaps justify this legally on the grounds that issues of limited liability, jurisdiction and ‘piercing the corporate veil’, issues that come into play in baseline private-law cases such as Adams v. Cape (Reference Adams v.1990), are only relevant when there exists a full-fledged legal system that exercises legitimate authority. ‘Reasonable decision-making agents should have known’, the argument can go, ‘that the normal rules about jurisdiction do not apply when bad states exercise their right of exclusive jurisdiction over a territory illegitimately, in a way that is detrimental rather than helpful to the community’s and individuals’ pursuits of their legitimate ends’.

In the Yeehaw case, China’s (hypothetical) rules are not authoritative in the normative sense, whereas the UK’s (hypothetical) rules are. The former half of this claim, at least, is sound to the extent that one accepts that torture and/or imprisoning individuals without due process and/or violating individual privacy are ways of exercising sovereignty illegitimately (Reus-Smit, Reference Reus-Smit2001). At first glance – one often says this when one wants to show that the first glance is not the best one – this ‘TNC and human rights’ policy prescription appears to be a solution to the problem of how to institutionalize a set of public, in-advance, in line with the principles of legality, human rights rules that ascribe non-discretionary duties onto non-state actors. Linking this to the original impetus behind the ATCA, metaphorical pirates (TNCs) cannot use the fact that they are operating outside of any true legal system (defined in Waldron’s robust sense) to shield them from legal accountability in jurisdictions where the law does exist.

Alternatives to the extraterritorial application of human rights standards include either the creation of new international institutions, or the use of existing ones such as: the ICC (Ratner, Reference Ratner2001: 500–506) or arbitration tribunals that preside over international investment agreements (Mann et al., Reference Mann, von Moltke, Peterson and Cosbey2006; Muchlinski, Reference Muchlinski2007). These alternatives aim to create the same type of clear rules and consequences, letting moral agents know in advance and with transparency what the consequences to them could be if they engage in actions that are human rights violating. Bracketing off potential objections against the internationalist option (Helms, Reference Helms2000/01; Rivkin and Casey, Reference Rivkin and Casey2000/01) – as debates about the strength of these objections are not relevant to the analysis that now follows – in theory, the internationalist option aims to have the same effect as the extraterritoriality option. The intention is to create an ‘actually authoritative’ set of human rights rules for non-state actors, to let them know in advance and with transparency when they have a non-discretionary duty correlative to individuals’ human rights: a duty to ignore the bad state’s rules and to be guided by a different authority beyond or outside of the state level (Hopgood, Reference Hopgood2009).

In terms of both policy options, extraterritorial and international, it is not true that the institutionalist answer – the idea that one ought to institutionalize rules so that they are publicly available – actually solves the rule of law problem. The latter problem is how to ascribe non-discretionary human rights duties to TNCs (the kind that make sense in principle) while at the same time doing so in a way that respects the fundamental principles of legality that make the institutionalization of non-discretionary duties justifiable. Let us put ourselves in the position of Yeehaw’s regional director, Susan, who has just been told by one of the mid-level managers working under her that the Chinese government has requested X’s personal details. Despite the claim that China is not exercising its sovereignty legitimately in a normative sense, one still has the descriptive fact of China’s sovereignty to deal with and to account for in a complete analysis (Karp, Reference Karp2008). Regardless of whether legal theorists call a legal system in which human rights are not respected a full-fledged legal system, and regardless of whether the rules that exist in such a system ought or ought not to be labelled as truly authoritative rules that justifiably give rise to non-discretionary duties, the fact remains that Susan can be arrested by the Chinese government for failing to comply with its request.

Now use the example of Susan, the individual, as an analogy to help one to think about Yeehaw, the corporate entity that is meant to be subject to China’s laws while in China and the UK’s laws while in the UK. Yeehaw China, the corporate entity, can be shut down for failing to comply with China’s request. Furthermore, because China is a strong, sovereign state with a significant GDP and a large population, the state can probably afford to shut Yeehaw China down without a very great medium- or long-term impact to the Chinese economy or Chinese society. That the UK, or the ‘international community’, has institutionalized a different set of clear rules, according to which if Yeehaw China does release X’s personal details, then Yeehaw China (via going after its parent company) can be forced into bankruptcy, and according to which Susan can be arrested as soon as she crosses the UK border to return from her 2-year regional directorship in China, creates a ‘damned if you do, damned if you don’t’ scenario: a tragic range of choices, within which any possible course of action has severe negative consequences for the decision-making agent (Brown, Reference Brown2007). It is hard to see how this scenario is in line with the principles of legality, which are supposed to clarify, rather than to confuse, decision-making, by giving moral agents accurate information about the potential consequences to themselves for performing or not performing specific required actions (Waldron, Reference Waldron2008). The policy prescriptions have an international-political, and perhaps a moral, character, but it would be a misunderstanding to view the successful implementation of these policies as an advancement of the ‘global rule of law’ (Teitel, Reference Teitel2005) and to discuss the principles that might justify their implementation from within a legal framework.

From an explicitly normative perspective, one could put forward a plausible argument that, all other things being equal, all moral agents ought not to go to foreign countries and to cause avoidable or unnecessary harm to others who live there. This is not the core problem in the context of a bad state. Rather, the core problem is the way that three distinct political ideas, ‘human rights’, ‘legitimate authority’ and the ‘rule of law’, can and cannot be combined in order to create and to justify the legal and institutional policy prescriptions that advocates are putting forward. Human rights duties are non-discretionary. The entire idea behind non-discretionary duties is that judgment about what is ‘avoidable’ or ‘unnecessary’ is and ought to be removed from particular decision-making agents and placed in the hands of an external authority. The rule of law problem occurs when different de facto authorities come to different judgments about how best to interpret human rights standards (see Çali and Wyss, Reference Çali and Wyss2008). Advocates assume uncritically, in situations of conflicts of judgment between different de facto authorities, that there is a normative hierarchy of institutional authority: that Western governments’ political institutions (and the international-level institutions over which they currently wield a large amount of control) will judge the moral questions correctly, whereas non-Western governments at the domestic level will get the moral questions wrong. There is no rule of law problem if and only if one accepts the incorrect and troubling idea that legitimate authority is hierarchical in this precise way (Carr, Reference Carr2001/1939; Barkawi and Laffey, Reference Barkawi and Laffey2006: 333–344).

As this paper has explained, since Raz (Reference Raz1986) developed his account of legitimate authority, many political philosophers have come to believe that authority does not become legitimate on the basis of who makes the rules. Legitimate authority no longer needs to be viewed as a ‘who’ question; rather, it can be viewed as a property of the rules themselves. The authority of rules is legitimate if and only if the rules – regardless of whose rules they are – meet certain tests, such as Raz’s suggested test that all agents within the scope of the authority will be better off, in terms of their pursuit of their own legitimate ends, if they defer to an external agent’s judgment, rather than deciding for themselves in each and every case what they ought to do. China’s judgment that the advocacy of multi-party democracy is potentially destructive to the fabric of Chinese society (and the UK government’s decision that it might be necessary to detain a terrorist suspect for 28 days without charging her with any offence) meets this test in one way. Human rights advocates’ judgment that arrest without due process is never acceptable meets this test in another way. Therefore, in terms of ‘legitimate authority’, we are right back to where we started. Different authorities often come to different judgments about how to interpret moral standards. One needs a further normative argument, rather than an institutional pushing-through of one idea over another, in order to move toward a principled resolution of this question. To object that there is nothing illegitimate or confusing if Yeehaw decides to go to China, knowing what the political and legal climate is like, and then gets itself into a bind because of conflicting rules,Footnote 8 overlooks the way that the concept of ‘legitimate authority’ is operating at multiple levels. As well, the political and legal climate can quickly change, as authorities introduce new rules, and it is not always easy, or morally and politically desirable, for companies to pick up and leave when they have disagreements with a sovereign state’s new requirements.Footnote 9

There is therefore a rule of law problem, which can only be resolved by concluding, in a way that is actually and legitimately authoritative for moral agents, that one set of rules is legitimate whereas the other is not. One of the jobs of the political concept of ‘human rights’ is to enable this discussion to happen. But the discussion needs to be a normative discussion. The discussion is not adequately resolved if an oppositional set of rules, aimed at guiding the action of political insiders – as distinct from a set of rules that is aimed at guiding the action of bad states – is institutionalized in a different place, without answering the more fundamental question of whether and why the bad state is exercising its de facto authority in an illegitimate way. If some of China’s rules are bad, and if progressively minded people therefore take the view that China is a human rights violator, then why, in principle, should we bring corporations – private-sphere economic actors who, in the aftermath of decolonization we thought, above all else, ought not to view themselves as political agents of their home countries’ policies – into it? Rather than to provide a simple answer to this question, one of the aims of this article was to make the question more complex, by opening up grey areas that need to be explored before going forward with one policy or another. The principles of legality call into question whether, in the complex context of a bad state, it would be legitimate to put Yeehaw into a situation where it will be punished regardless of what it does.

Conclusion

This paper has argued that the types of extraterritorial or international law that are being prescribed by NGOs, think-tanks, UN bodies and other public-policy actors, cannot be justified within a legal framework when dealing with a bad-state political context. To the extent that these policy recommendations can be justified at all, they can only be justified according to (non-legal) moral or political factors, but at the potentially significant cost of recognizing that the creation of such rules would violate the principles of legitimate authority or the principles of legality: principles to which one normally looks when one thinks about fairly imposing non-discretionary duties upon private-sphere moral agents such as individual or corporate persons.

The political context of bad states is one in which agent-centred analyses about what Yeehaw ought to do cannot be neatly separated from an analysis of the social structures that affect the corporation’s reasons for action. An analysis of what Susan, acting on behalf of Yeehaw, ought to do rests on an answer about whether and why Yeehaw ought to follow rules (and whose rules, and which ones) that aim to ascribe non-discretionary duties onto the company. To complicate the decision-making process of moral agents is not necessarily problematic. Often it is a good thing. However, this does become problematic – and therefore advocates’ policy prescriptions become problematic – when juxtaposed against the claim that human rights rules are the ones that are truly authoritative and/or in line with a global rule of law, justifiably regulating transnational actors irrespective of borders. ‘Human rights’ as a political concept stands in an uneasy relationship with ‘authority’ and ‘the rule of law’. Although these tensions are not necessarily intractable, they need to be tackled rather than assumed away, as tends to happen when human rights prescriptions are understood within a purely legal framework. As this paper has shown, this all comes to the fore when one thinks about how global human rights rules purporting authoritatively to govern the conduct of transnational actors would interact with the de facto authority of bad states.

This article made theoretical distinctions that open up important directions for future research. I discussed Rawls’ (Reference Rawls1999: 89–120) identification of two situations where, as I have put it, the descriptive and normative aspects of sovereignty are mismatched: ‘outlaw states’, lacking the willingness to abide by internationally accepted principles, and ‘burdened societies’, lacking the capacity to do so. I then went on to discuss the ways in which corporations can be characterized as state-like in some theoretically relevant ways and individual-like in others. Human rights standards are often understood to apply differently to individuals as compared to states, firstly in terms of how individuals (on one hand) and states (on the other) ought to respond to human rights standards, and secondly in terms of who ought to be held to account for violations, in what way, and why. This article clarified that it makes sense to treat corporations as individual-like, with respect to at least some of these questions, to the extent that they operate in political contexts that resemble the bad-state ideal type.

The political context of a weak state or a ‘burdened society’ is sufficiently important, and sufficiently different, to warrant an entirely separate argument: one which looks at the same reality, but which shifts the focus by considering some of the ways in which TNCs are more state-like than individual-like in certain political contexts, due to their enhanced power, capacity, autonomy and agency, vis-à-vis the host states in which they operate. This is a clear and important direction for future research, and one that ought to be analyzed in light of the conclusions of the present piece of work, in order more fully to make sense of the whole picture.

Legal and policy analysts often treat the ‘powerful corporation in the weak state’ as the major relevant baseline for a discussion about the potential human rights responsibilities of TNCs. Their question is therefore: how can one affect corporations’ behaviour from outside the host-state level in order to cause them not to do serious harm to others. Using this as the baseline is problematic from a practical perspective, because it unjustifiably narrows the range of problems that ought to concern those of us engaged in the ‘TNC and human rights’ research agenda. It is also problematic from a theoretical perspective, because it takes the focus off of a huge body of legal and political theory that addresses the question about which rules (regardless of whether they are ‘human rights rules’ or ‘sovereign rules’) ought to be obeyed by moral agents, which ought to be disregarded, and why.