Accountability and global governance: challenging the state-centric conception of human rights

Abstract

In this paper I analyze some conceptual difficulties associated with the demand that global institutions be made more democratically accountable. In the absence of a world state, it may seem inconsistent to insist that global institutions be accountable to all those subject to their decisions while also insisting that the members of these institutions, as representatives of states, simultaneously remain accountable to the citizens of their own countries for the special responsibilities they have toward them. This difficulty seems insurmountable in light of the widespread acceptance of a state-centric conception of human rights, according to which states and only states bear primary responsibility for the protection of their citizens' rights. Against this conception, I argue that in light of the current structures of global governance the monistic ascription of human rights obligations to states is no longer plausible. Under current conditions, states are bound to fail in their ability to protect the human rights of their citizens whenever potential violations either stem from transnational regulations or are perpetrated by non-state actors. In order to show the plausibility of an alternative, pluralist conception of human rights obligations I turn to the current debate among scholars of international law regarding the human rights obligations of non-state actors. I document the various ways in which these obligations could be legally entrenched in global financial institutions such as the WTO, the IMF, and the World Bank. These examples indicate feasible methods for strengthening the democratic accountability of these institutions while also respecting the accountability that participating member states owe to their own citizens. I conclude that, once the distinctions between the obligations to respect, protect, and fulfill human rights are taken into account, no conceptual difficulty remains in holding states and non-state actors accountable for their respective human rights obligations.

Cristina Lafont is Professor of Philosophy at Northwestern University. Her current research focuses on conceptions of deliberative democracy and their possible extension beyond national borders. Some of her recent articles include 'The Place of Self-Interest and the Role of Power in Deliberative Democracy,' in The Journal of Political Philosophy 18/1 (2010): 64-100. Co-author with J. Mansbridge, J. Bohman, S. Chambers, D. Estlund, A. Follesdal, A. Fung, B. Manin, and J.L. Marti. 'Can Democracy go Global? Comments on J. Bohman's Democracy Across Borders,' in Ethics & Global Politics, 3/1 (2010): 13-9. 'Religion and the Public Sphere. What are the Deliberative Obligations of Democratic Citizenship?,' in Philosophy & Social Criticism, 35/1-2 (2009): 127-50. 'Alternative Visions of a New Global Order: What should Cosmopolitans hope for?', in Ethics & Global Politics, 1/1-2 (2008): 1-20 [Reprinted in S. Besson and J.L. Martí, eds., Legal Republicanism and Republican Law. National and Post-National Perspectives (Oxford University Press, 2009), 256-77 and in Soziale Welt, 18 (2009): 231-50]. 'Religion in the Public Sphere: Remarks on Habermas's Conception of Public Deliberation in Post-secular Societies,' in Constellations, 14/2 (2007): 236-56. 'Is the Ideal of a Deliberative Democracy Coherent?,' in S. Besson and J.L. Martí, eds., Deliberative Democracy and its Discontents (Ashgate, 2006): 3-26.

Among political philosophers there is virtually universal agreement that democracy is superior to any other alternative political system at the national level. It is surprising then that relatively few theorists recommend democracy as a suitable political system at the global level. However, it would be wrong to infer that there is some specific feature of democracy that those opposed to the ideal of a global democracy reject. What most of these opponents find objectionable is the idea of a single political system with a world government. It goes without saying that they would find it even more objectionable if it were a non-democratic, totalitarian government. In fact, the main fear such opponents have is that a world state, even if formally democratic, would unavoidably become totalitarian due to its extraordinary accumulation of power. However, even if one agrees that a world state is undesirable, there may be some features of democracy that could be applied to global institutions when properly adapted. Transnational democratization may be possible without a global democratic government. Understood in this more modest way, the aspiration for global democracy seems quite widespread in the current literature on global governance. At the very least, the complaint that current structures of global governance lack democratic legitimacy is widespread, and this in turn suggests that some kind of democratization at the transnational level would be desirable. In what follows, I would like to explore this possibility. In particular, I would like to focus on the democratic ideal of political inclusion in order to see whether and, if so, to what extent, it offers an appropriate aim for improving the legitimacy of global institutions in the absence of a world state.

At a minimum, democracy can be defined as the collective authorization of laws and policies by the people subject to them. There are two sides to the requirement of political inclusion implicit in this definition. One side concerns sovereignty, that is, the question of who has decision-making status in a political system. The democratic ideal of political inclusion in decision-making is usually expressed as the ideal of popular sovereignty: in a democracy those subject to the law are, at the same time, its authors. In representative democracies, however, the ideal of popular sovereignty typically amounts to mere electoral sovereignty: citizens delegate most of the actual decision-making on specific laws and policies to their official representatives, but they do have the political power to elect their representatives and in special cases may also directly decide some important laws or policies through popular referenda. The requirement that the citizens have some final say over some important political decisions is the essential difference between democracy and other non-democratic forms of political organization. It finds institutional expression in political rights of participation in decision-making processes that are extended to all adult citizens.

Now, if all citizens had final say over all political decisions to which they were subject, sovereignty would be the only dimension of the democratic ideal of political inclusion. But in representative democracies, in which citizens delegate most political decision-making to others, the democratic requirement of political inclusion imposes normative constraints not only regarding the sovereignty of citizens but also regarding the accountability of their representatives. According to the democratic ideal, the latter must remain accountable to all those who are subject to their political decisions.¹ If one interprets accountability in terms of the principal-agent model, the specific feature of democratic accountability, in contradistinction to other types of accountability,² is that all those subject to the agent's political decisions must be recognized as principals. As with the democratic conception of sovereignty, the key to the democratic conception of accountability is the requirement of political inclusion (i.e. the required congruence between those responsible for political decision-making and all those subject to them). The obvious difference is that whereas democratic sovereignty requires that all decision-takers are included in the decision-making process itself, democratic accountability requires decision-makers to act as representatives for all those subject to their decisions and thus to take the interests of all of them into account in the decision-making process.

Due to these two dimensions of the democratic ideal of inclusion, the term 'democratic accountability' can be understood in different senses. When understood as an answer to the question to whom must representatives be ultimately accountable, the term indicates the democratic requirement that representatives be accountable to all those subject to their decisions. However, it can also be understood as an answer to a different question, namely, by which specific means must accountability be guaranteed. In this second sense, the term can be understood as synonymous with what is usually called 'electoral accountability' and thus as referring to the specific democratic requirement that representatives be held accountable to all those subject to their decisions through the latter's power to remove them from office in elections. If we understand the term in this second sense, 'democratic accountability' entails electoral sovereignty. However, if we understand 'democratic accountability' in the first sense, then electoral accountability is certainly a type of democratic accountability but not the only possible one. In constitutional democracies, there are many different mechanisms to guarantee that representatives remain accountable to those subject to their decisions beyond elections. Since in what follows I am going to address only the first question, I will use the term 'inclusive accountability' instead of 'democratic accountability' to avoid any confusion with electoral accountability.³

Needless to say, both the strong requirement of electoral sovereignty and the weaker requirement of inclusive accountability remain unfulfilled in nearly every currently existing democratic society, as most have large populations of non-citizens who lack any representation in the political process, but are nonetheless subject to its decisions. However, this mismatch between decision-makers and decision-takers at the national level pales in comparison to the same mismatch at the global level which will be the focus of my attention in what follows.

It is clear that the current institutions of global governance fail to meet both requirements of democratic legitimacy for political decision-making. The most glaring deficit is the lack of electoral sovereignty for all those subject to the decisions of global institutions. What is less obvious, however, is whether meeting this requirement at the global level is possible at all in the absence of a world state.⁴ Leaving this question of electoral sovereignty open for now, I would like to focus instead on the requirement of inclusion at the core of the democratic conception of accountability.

Although there is widespread agreement on the accountability deficits within current institutions of global governance (such as the UN, the IMF, WTO, or the World Bank), there is considerably less consensus regarding the appropriate remedies. In fact, proposals for applying domestic criteria of democratic accountability to global institutions are contested not only by those who question their feasibility⁵ but also by those who question their desirability due to the potential trade-offs in other positive qualities and valuable goals pursued by the institutions involved.⁶ However, inclusion is an aspect of accountability whose desirability at the global level seems uncontroversial. The normative principle that decision-makers should be accountable to all decision-takers, to use Held's terminology,⁷ seems as valid at the transnational level as it does at the national level. Now, what seems questionable, and has indeed been questioned, is whether inclusive accountability can be achieved at the global level without a world state. The reasons here are not so much empirical, but rather conceptual. Leaving empirical questions of feasibility aside, from a purely normative point of view it seems that if we want to stick to the division of political space by states, we must stick to the corresponding division of principals by states. If this is the case, it would just be a conceptual confusion to insist on maintaining the normative significance of states while at the same time demanding inclusive accountability at the global level. We cannot have our cake and eat it too.

In what follows, I would like to argue for a specific way out of this apparent dilemma. My ambitions, however, are pretty modest. I do not aim to provide solutions to the difficult questions of implementation that accompany proposals for institutional reform. My argument is not an attempt to address all the accountability deficits of current global institutions.⁸ My aim is simply to show that there is nothing conceptually confused or impossible in making global institutions accountable to all decision-takers while the members of these institutions, as representatives of states, simultaneously remain accountable to the citizens of their own countries for the special responsibilities they have toward them.

INCLUSIVE ACCOUNTABILITY WITHOUT A WORLD STATE?

As mentioned earlier, one of the main difficulties with the emergent structures of global governance is the mismatch between decision-makers and decision-takers.⁹ This generates a very special kind of deficit in accountability. The problem at issue here is not so much that whenever decision-takers do not have any final say in political decision-making processes decision-makers can easily avoid accountability. The specific problem at the global level is that decision-makers are not even supposed to be accountable to all decision-takers in the first place. Powerful countries can impose global economic regulations that may have devastating consequences for many decision-takers and this occurs not so much because the delegates of those countries in international financial institutions are avoiding accountability but rather because they make their decisions in the name of such accountability.¹⁰ Since delegates are supposed to be accountable to the citizens of their own countries regardless of whether the decision-making is at the domestic or transnational level, they see themselves as under the obligation to protect and promote the interests and rights of their own citizens and not those of all decision-takers. Thus, strengthening the accountability of delegates to the citizens of their own countries can only exacerbate the problem rather than solve it. In the absence of any institutional features designed to correct the fragmentation among multiple (and highly unequal) principals, there is no reason at all to expect decision-making in global institutions to be accountable to all decision-takers.¹¹

In light of this structural mismatch, many authors draw the conclusion that achieving inclusive accountability at the global level would only be possible within a world state that, as a single agent, would be accountable to all world citizens as a single principal. As is often the case, agreement on this conclusion is open to the usual modus ponens/modus tollens argumentative alternative, so that some authors offer it in support of the need of a world state for transnational democracy,¹² whereas others see it as a direct refutation of cosmopolitan claims to global justice that play down the normative significance of states. Nagel's article 'The Problem of Global Justice' offers a clear example of the latter strategy. He sharply characterizes the structural difficulty at issue here as follows:

I believe that the newer forms of international governance share with the old a markedly indirect relation to individual citizens and this is morally significant. All these networks bring together representatives not of individuals, but of state functions and institutions. Those institutions are responsible to their own citizens and may have a significant role to play in the support of social justice for those citizens. But a global or regional network does not have a similar responsibility of social justice for the combined citizenry of all the states involved, a responsibility that if it existed would have to be exercised collectively by the representatives of the member states.¹³

Nagel's counterfactual claim highlights the normative dilemma: either states and their representatives have the primary responsibility of protecting the interests and rights of their own citizens or a world state would be needed whose representatives would have the collective responsibility of protecting the interests and rights of all world citizens equally. What is not possible is to preserve the ascription of primary responsibilities to states while simultaneously expecting inclusive accountability of global institutions whose members represent states rather than world citizens. The problem here is not the usual discrepancy between normative expectations and the realities of power politics, but a direct conflict between normative expectations themselves. As already mentioned, the objection is not empirical, but conceptual.

Now, since the alternative of a world state is not very attractive, it seems that the most promising route forward is to examine the precise content and scope involved in ascribing primary responsibility to states for the protection of their citizens' interests and rights. This ascription is at the core of the dominant conception of human rights obligations, according to which states bear the primary responsibility for protecting the human rights of their own citizens. In spite of this conception's prevalence, the normative credentials of the state-centric conception of human rights obligations are worth examining, for the fragmentation of principals that makes the expectation of inclusive accountability without a world state seem incoherent is a direct consequence of the ascription of responsibilities characteristic of that conception. To anticipate the line of argument that I will develop in what follows, I agree with two of the claims that lead to the normative dilemma pointed out by Nagel. I agree that some assumptions of the state-centric conception of human rights are incompatible with the ascription of inclusive accountability to institutions of global governance, and I also agree that the formation of a world state is the only way to make inclusive accountability at the global level compatible with the state-centric conception of human rights obligations. But these two claims do not amount to a genuine dilemma, since there is a third (and more attractive) alternative open. Instead of having to bite the bullet of a world state or renounce to the demand of inclusive accountability at the global level, it seems best to reject the problematic assumptions of the state-centric conception of human rights.

CHALLENGING THE STATE-CENTRIC CONCEPTION OF HUMAN RIGHTS¹⁴

In his book The Idea of Human Rights, Charles Beitz offers a paradigmatic statement of the state-centric conception of human rights that he defends. As he explains, 'the central idea of international human rights is that states are responsible for satisfying certain conditions in their treatment of their own people and that failures or prospective failures to do so may justify some form of remedial or preventive action by the world community or those acting as its agents.'¹⁵ Consequently, human rights practice consists 'of a set of norms for the regulation of the conduct of governments and a range of actions open to various agents for which a government's failure to abide by these norms supplies reasons.'¹⁶ Thus, according to the state-centric conception of human rights, states bear the primary responsibility for protecting and promoting the rights and interests of their own citizens. The international community bears some responsibility in the protection of human rights, but, in contradistinction to states, this responsibility is secondary (or residual¹⁷ ), and this is in two senses. First, the international community's responsibility is secondary in the sense that it is activated only if and when states are unwilling or unable to protect the rights of their own citizens. Second, the responsibility of the international community is secondary in the sense that it is not supposed to replace the protective function of states. The international community through the different institutions that act as its agents may provide (temporary) assistance to states, but none of these institutions are supposed to provide the kinds of protections, entitlements, and services that states provide to their own citizens. Whereas states are directly subject to human rights obligations, the protective function of the international community consists in holding states accountable for the treatment of their own citizens. This function is exercised by a variety of international and transnational agents and through a variety of measures. These measures include monitoring states' compliance with international standards of human rights, offering economic and other incentives for compliance (e.g. aid conditionality or preferential treatment in economic relations), using coercive measures such as threats of economic or diplomatic sanctions and, in cases of gross human rights violations, the use of military intervention.

Although the state-centric conception of human rights obligations is widely accepted, it has some problematic consequences. One striking feature of this conception is its remarkable silence regarding the obligations of non-state actors (from individuals¹⁸ to multinational corporations¹⁹ or international financial institutions like the WTO, IMF, or the World Bank). If states bear primary responsibility for protecting the human rights of their own citizens and the secondary responsibility of the international community consists in holding states accountable for the treatment of their own citizens, it seems that non-state actors have no responsibility to protect human rights and, consequently, that the international community has no responsibility to hold such actors accountable for the impact of their own actions or decisions on the protection of human rights. However, under current conditions of globalization, it is becoming increasingly clear that decisions on global economic regulations by non-state actors such as the WTO, the IMF, or the World Bank can have a tremendous impact on the protection of human rights worldwide. Now, if this is the case, isn't it implausible to claim that these institutions have no human rights obligations? Even worse, how can the international community hold states accountable for the consequences of global regulations that are not in their hands to determine? Shouldn't the international community hold those actors accountable whose decisions and actions hamper the protection of human rights, whether or not they happen to be states, instead of holding states accountable for the consequences of decisions or actions that are beyond their effective control?

ACCOUNTABILITY GERRYMANDERED

It is true that, according to the state-centric conception, if states are unable to protect the human rights of their own citizens due to external factors, they won't be held accountable in the same way that they would be if they were able but unwilling to provide those protections. They won't be subject to coercive sanctions and may receive international assistance instead. But the fact that states are not held accountable for regulations and decisions beyond their control can hardly make the state-centric conception normatively compelling if those who are responsible for such decisions cannot be held accountable either. The impossibility of closing this 'accountability gap' within the state-centric conception is a major shortcoming in its own right, but it also leads to implausible results regarding the secondary responsibilities of the international community. If, ex hypothesi, a global economic regulation is impairing the protection of the human rights of citizens in certain states, providing assistance to these states seems like the wrong remedial measure for the international community to take. The appropriate response would be to address the problem at its root and to change the problematic regulation instead. But this intuitively obvious response cannot be accommodated within the state-centric view. Let's see why.

Given that the state-centric conception acknowledges that the secondary responsibility of the international community regarding human rights protections is spread through a variety of international and transnational institutions (UN human rights agencies, states, NGOs, etc.), it might seem like it can unproblematically account for the fact that different remedial measures may be needed to properly discharge this responsibility. In The Idea of Human Rights, Beitz offers an example of this argumentative strategy. Although he explicitly defends a state-centric conception of human rights, he provides a list of six 'paradigms of implementation' or 'enforcement' of human rights through which various agents might seek to prevent or remedy the failures of states to protect human rights, one of which includes the possibility of changing a global regulation. He calls this type of action 'external adaptation' and explains its difference from the other five paradigm types of implementation (accountability, inducement, assistance, domestic contestation, and compulsion) as follows:

The first five paradigms consist of mechanisms that aim to influence the behavior of domestic agents ... But it may be that the obstacles to a government's compliance are to be found in the policies of other states, multinational actors, or regimes rather than in its own lack of resources, capacity, or will. Consider, for example, trade policies that discriminate against agricultural products or intellectual property rules enforced in international law that increase the costs of pharmaceuticals. If it were true that without adaptations in these 'external' policies, a government would not be in a position to ensure the satisfaction of its people's human rights, then, although it may seem to be a linguistic stretch, reform of the policies might plausibly be considered a means of 'implementation'. (It may seem less of a stretch if one recalls the declaration's recognition of the need for 'an international order ... in which human rights can be fully realized.) In any case, it is clear that external adaptations are among the forms of action for which justifications are sometimes sought in considerations about human rights.'²⁰

It is not surprising that Beitz considers the inclusion of this form of action to be a linguistic stretch. In light of his description of human rights practice as 'consisting of a set of norms for the regulation of the conduct of governments,' the inclusion of this type of action is indeed highly anomalous. As he himself indicates, one striking difference between this type of action and the other five is that it is not aimed at the affected states (i.e. their governments), but at non-state actors. Now, within the state-centric conception, it is obvious why measures aimed to influence the behavior of states in their treatment of their own citizens are appropriate ways for the international community to discharge its (secondary) human rights obligations, since, according to this conception, the international community's obligation is to hold states accountable for their responsibility to protect the human rights of their own citizens. But how can a measure aimed at influencing the behavior of an actor who does not have the responsibility to protect human rights in the first place be justified as a way of implementing human rights protections? Moreover, how could the non-state actor undertake reform of the regulations at issue in the name of implementing human rights without thereby acknowledging its own obligation to protect human rights?²¹

Beitz is certainly right in reporting that considerations of human rights are in fact often used in the international arena to justify the need for this type of action. But that does not mean that he can justify this peculiar feature of the emergent discursive practice of human rights within the limits imposed by the state-centric conception. An appeal to human rights protections can only be a valid reason to reform global economic regulations if the institutions in charge of implementing that reform themselves have human rights obligations. The normative plausibility of these appeals speaks in favor of abandoning the monistic ascription of human rights obligations entailed in the state-centric conception.

In fact, some of Beitz's own arguments seem to directly undermine the plausibility of the state-centric view. The general aim of the book is to offer a compelling interpretation of human rights as an emergent political and discursive global practice. Beitz insists that the practical conception of human rights he defends offers a normative and not a merely descriptive model of an emergent practice.²² This allows him to counter the objection that the practical approach, by aiming to reconstruct the features of a given practice, must give too much authority to the status quo. Against this objection, he indicates that a normative model has the resources to criticize the practice it aims to reconstruct, for example when 'the practice's norms are ill-suited to advance its aims.'²³ Now, Beitz also acknowledges that, in light of the currently existing structures of global governance, states are bound to fail in their ability to protect the human rights of their citizens, whenever the potential violations either stem from transnational regulations or are perpetrated by non-state actors. If this is the case, then sticking to the norm that only states are primarily responsible for the protection of human rights appears to serve no other purpose apart from releasing non-state actors of any obligation to protect human rights, while simultaneously acknowledging that they may be uniquely suited to fulfill this function in some cases. This situation seems to fit Beitz's description of a 'practice's norm that is ill-suited to advance its aims.' If so, the practical approach should have the resources to adopt a critical stance toward the state-centric norm precisely to the extent that it is ill-suited to advance the practice's own aim of securing the protection of human rights worldwide. Seen in this light, and contrary to Beitz's own claims, it appears that the practical conception offers a strong justification for a pluralist norm (and against an exclusively state-centric norm) as the appropriate guide for human rights practice under current conditions of globalization.²⁴ Beitz's argument against this alternative is that the model he proposes should be descriptively accurate of current practice and thus should not be changed unless the practice itself changes.²⁵ However, this seems to undermine any action-guiding function of the model vis-à-vis current practice. In so doing, it opens the practical conception of human rights to the objection that it gives 'too much authority to the status quo by taking an existing practice as given,'²⁶ something that Beitz's arguments explicitly aim to refute.

In my view, there is indeed a very good reason for not granting too much authority to the status quo by taking the existing state-centric norm as given. A conception of human rights that aims to give an account of the meaningfulness of human rights practice can hardly be plausible if it turns out that, in light of that very conception, the practice's aim of securing human rights protections worldwide is doomed from the start under contemporary circumstances.²⁷ This negative conclusion, however, seems unavoidable unless the 'accountability gap' that arises from the monistic ascription of human rights obligations to states can be effectively closed. Everyone agrees that the goal of securing human rights worldwide cannot be achieved unless global financial regulations do not in fact hamper such protection. In light of this situation, it is simply not enough for defenders of the state-centric view to indicate, as Beitz does, that there are legal means by which global economic regulations that drastically hamper the protection of human rights could be modified. What needs to be shown is that there is some (legally effective) way to assure that such regulations will indeed be modified. This, in turn, requires identifying actors who have both the obligation and the effective legal capacity to undertake such a modification. However, this task seems impossible to achieve within the framework of the state-centric conception, since the actors who have the obligation (individual states) do not have the legal capacity to do so and the actors who have the legal capacity (the WTO, IMF, or World Bank) do not have the obligation. It is true that, according to the state-centric view, a state's failure to protect human rights may give appropriately placed and capable outside agents (such as non-state actors) pro tanto reasons to act. However, as Beitz himself acknowledges, 'this means that, in the general case, a human rights failure in one society will not require action by outside agents' (117). This is the difficulty that, in my view, the state-centric conception of human rights cannot overcome even in its most favorable interpretation.²⁸

The important question here, of course, is whether there is a plausible alternative to the status quo. Is it plausible to claim that non-state actors such as the WTO have an obligation to protect human rights? Wouldn't this require that the WTO cease to be a voluntary association designed to facilitate trade among its members and that it become a human rights organization instead? Now, in order to show how an alternative to the state-centric conception's monistic ascription of human rights obligations does not also need to lead to such implausible consequences, it is important to pay attention to the ambiguity within the notion of 'protecting' human rights. If, following what has become standard terminology, we distinguish between the duties to respect, protect and fulfill human rights,²⁹ it is clear that speaking of the obligation to 'protect' human rights can have very different meanings. These different meanings depend on whether obligations are interpreted in the narrower sense of (merely) respecting human rights or in the more expansive sense of (actively) fulfilling human rights. Whereas in the second, more expansive sense it is indeed very plausible to claim that states bear the primary responsibility in providing the protections, entitlements, and services necessary for fulfilling (i.e. promoting and enforcing) the human rights of their citizens, it does not seem plausible at all to claim that states are the only actors that bear primary responsibility in respecting the human rights of their citizens. The obligation of respecting human rights in the sense of not contributing to their violation seems to be a universal obligation and thus one that binds states just as much as non-state actors.

In this context, it is important to resist assimilating the distinction between these two senses of 'protecting' human rights to the distinction between 'acts' and 'omissions', according to which the 'fulfillment' of human rights requires positive action whereas 'respect' requires only self-restraint. There may be contexts in which this distinction is useful, but our present context is not one of them. Inaction may be an appropriate way to discharge the obligation of respecting human rights in some contexts by some non-state actors, but surely not by all of them. A multinational corporation may decide to cease involvement in a country with a high record of human rights violations in order to discharge its obligation of respecting human rights. However, this is not a live option for international financial institutions that are in charge of regulating different sectors of the global economic order (such as the WTO, the IMF, or the World Bank). So long as their mission is to implement global economic regulations and policies, they have no option but to actively choose among alternatives and implement some regulation or other. As far as these institutions are concerned, the relevant difference between promoting and respecting human rights is not the difference between action and omission. It is the difference between taking the fulfillment and enforcement of human rights worldwide as their own goal (i.e. becoming a human rights organization) and accepting the obligation to ensure that the regulations they implement in the pursuit of their respective goals (e.g. trade liberalization, financial stability, economic growth, etc.) do not hamper the protection of human rights worldwide.³⁰ In light of this distinction, it seems clear that the question of whether or not these institutions ought to make the goal of actively promoting and enforcing human rights part of their legal mandate or whether this function ought to be left to states and human rights institutions, has no bearing on the quite different question of whether they are bound by international human rights law to respect human rights by making sure that the regulations they implement (in pursuit of their own specific goals) do not have a negative impact on the protection of human rights worldwide. Whereas the former question is complex and its appropriate answer is therefore highly contested, the positive answer to the latter question seems hardly questionable from a normative point of view.³¹ Once the distinction is drawn, it is hard to imagine that any of these institutions could openly defend an interpretation of their legal mandate as consisting in the single-minded pursuit of their specific economic goals in total disregard of their impact on the most basic human rights of the world's population.

In fact, if one looks at recent official statements from these institutions regarding the question of their human rights obligations, one almost invariably finds recognition of the combination of claims just mentioned, namely, a rejection of the obligation to directly promote or enforce human rights as a part of their legal mandate (thus becoming a human rights organization) and an acknowledgment of the obligation to ensure that the policies and regulations they enforce do not hamper the enjoyment of human rights worldwide. An example of this position can be found in a letter written by Sergio Pereira Leite, an assistant director of the European office of the IMF, with the title 'The International Monetary Fund and Human Rights,' which was published in Le Monde in September of 2001 and can be found in the official IMF website.³² On the rejection of making the implementation of human rights part of the IMF mandate, he states:

Since 1999, the IMF has stressed the central role of poverty reduction in the Fund's strategy for low-income countries ... It is important to remember, however, that the ownership of the poverty reduction strategy needs to remain with the country. While human rights advocates should be given every opportunity to participate in the PRSP [poverty reduction strategy paper] consultations, they should not expect the IMF to impose human rights conditions on its assistance to member countries. The IMF simply does not have the expertise required to make judgments in this area. (My italics)

But, after stating the IMF's lack of mandate and expertise in the area of promoting and enforcing human rights, we also find recognition of the obligation to make sure that the IMF's own policies do not have a negative impact on the enjoyment of human rights³³ :

The International Monetary Fund recognizes that it should be aware of any adverse side effects of policies it proposes. In those cases, it is often necessary to introduce appropriate safety nets to help alleviate adverse social consequences. This is fully accepted by the IMF ... The IMF should endeavor to be open to criticism and undertakes to change its policies when results are disappointing. But it should be recognized that the IMF was created to promote international monetary cooperation and orderly balance of payment adjustment. (My italics)

Similarly, in the vast and complex legal literature on trade and human rights, there is disagreement among legal scholars on the issue of whether or not the WTO is under an obligation to protect human rights. However, if one pays attention to the difference between the two quite different senses of the obligation to 'protect' human rights mentioned before, the space for disagreement narrows down pretty quickly. Interpreted in the expansive sense, there is indeed disagreement on whether the protection of human rights is part of the mandate of the WTO. On the one hand, many legal scholars answer this question in the negative. They do so mostly by pointing out that the goal of protecting human rights is not mentioned in any of the foundational documents of the WTO as part of its mandate. On the other hand, however, some legal scholars argue that the goal of protecting human rights can be derived from some of the specific goals that are stated in the Preamble of the WTO such as the goal to raise the standard of living and of sustainable development.³⁴ Be that as it may, what seems clear is that this interesting debate has no direct bearing on the quite different question of whether the WTO is under an obligation to protect human rights in the narrower sense of being bound to respect international human rights law.³⁵ Regarding this question, there is widespread agreement among legal scholars on the affirmative answer,³⁶ although there is plenty of disagreement on the precise scope and implications of this obligation as well as on the variety of legal and technical means for its implementation.

Needless to say, disagreement on the latter questions is a clear indication of the limited value of agreement on the former question under current circumstances. The abstract obligation to respect international human rights law remains vague and ineffective as long as it is not translated into clearly specified and legally entrenched responsibilities. Thus, proposals as to how this goal could be achieved abound. The less ambitious proposals ascribe the main obligations to the Dispute Settlement institutions and analyze the different legal tools at their disposal when judging cases brought up by specific member states (those could range from the relatively minor concession of exceptions to specific countries to the more consequential demand that the regulation at issue be revised by the crafters to avoid conflict with any relevant human rights law).³⁷ The more ambitious proposals include, in addition, the obligation to assess the impact of specific trade policies on the ability to protect human rights at the early negotiating and crafting stages and not only at the implementation stage.³⁸ This obligation could be discharged by establishing a 'Trade and Human Rights Division' within the WTO Secretariat that should provide technical assistance in conducting human rights impact assessments .³⁹

With the World Bank the situation is similar. Although the official understanding of the Bank's own legal mandate does not include the direct enforcement or implementation of human rights, the Bank recognizes its obligation to respect human rights.⁴⁰ Beyond this abstract recognition, however, there are at least two features of the Bank's current structure and practices that provide an institutional basis for the possibility of entrenching obligations to respect human rights in the workings of the Bank. In contrast to the WTO, it is already part of the Bank's current practice to engage in 'social impact analyses' which are designed to calibrate the social impacts of its own policies and programs at the country level as they affect 'the well-being or welfare of different stakeholder groups, with particular focus on the poor and vulnerable.'⁴¹ Although the implementation record of this type of analysis for most of the Bank's supported programs is poor, its existence does point to the feasibility of incorporating human rights-based impact assessments within the Bank's operational policies that would be sensitive to the wider range of factors relevant for human rights protections in the specific context and circumstances of different countries and would thus provide guidance as to whether and how a specific operation or program can proceed. This is a practice that is also present in the IMF, although to a lesser extent. Another institution within the Bank that shows promise for further entrenching human rights obligations in the Bank's operations is the Inspection Panel established in 1993. Its mandate is to review complaints from any group of private persons alleging that they are suffering or expect to suffer material adverse effects from the failure of the Bank to follow its operational policies and procedures. Obviously, the more the obligation to respect human rights is entrenched in the Bank's operational policies, the greater the potential for this institution to hold the Bank accountable for its human rights impacts. An additional normative significance of the Panel's establishment is that it constitutes the first formal acknowledgement that international organizations are accountable not just to their member states, but also to individuals or private parties that are affected by their operations, independently of either the organization's or the private actor's relationship to a member state.⁴²

ACHIEVING INCLUSIVE ACCOUNTABILITY THROUGH A PLURALIST CONCEPTION OF HUMAN RIGHTS OBLIGATIONS

This brings us back to the initial question of whether it is conceptually sound to expect members of global institutions who are representatives of states to be accountable to all decision-takers while at the same time remaining accountable to the citizens of their own countries for their special responsibilities toward them. It seems to me that, once the question is framed in terms of the human rights obligations of global institutions, the conceptual dilemma loses much of its plausibility. It seems perfectly coherent to claim that members of global institutions have, as representatives of states, the special responsibility of advancing the interests and rights of their own citizens as strongly as possible, so long as they respect the limits imposed by the general obligation they have as agents of global institutions to make sure that their collective decisions do not negatively impact the possibility of protecting human rights worldwide. The analogy at the national level seems sufficient to render this view of plural obligations normatively plausible. Representatives of different regions in a country with a federal political structure may have the special responsibility of promoting the interests and rights of the citizens of the regions they represent as strongly as possible, so long as they respect the limits imposed by their obligation as participants in a national institution of making sure that their collective decisions do not negatively impact the possibility of protecting the constitutional rights of all citizens.

I do not mean to suggest that the analogy between the national and the global case is perfect nor that the institutional solutions established at the national level (such as a supreme court with final authority to interpret the constitution) would be appropriate or desirable at the global level. One important difference between the national and the international case is that at the national level all citizens have the same constitutional rights, whereas at the international level the constitutional rights of citizens in some nation-states can be much more extensive and demanding than the human rights recognized as applicable to all world citizens. For this reason, the proposal I am defending here is compatible with Nagel's claim that the obligations of social justice among the citizens of a particular nation-state are qualitatively different from and greater than the obligations these compatriots have towards citizens of other countries.⁴³ However, in ascribing human rights obligations to global institutions, my proposal is directly incompatible with Nagel's additional claim that the actions and decisions of global institutions do not 'raise to the level of collective action needed to trigger demands for justice, even in diluted form. The relation remains essentially one of bargaining, until a leap has been made to the creation of collectively authorized sovereign authority. On the 'discontinuous' political conception I am defending, international treaties or conventions, such as those that set up the rules of trade ... are 'pure' contracts, and nothing guarantees the justice of their results.'⁴⁴

Nagel's view of the regulations of international institutions such as the WTO as 'pure contracts' (and thus as exempt of any constraints of justice, however minimal) seems motivated by a false dilemma. As we saw at the beginning, Nagel's argument seems to assume that we have only two conceptual choices: either representatives of states have the responsibility of advancing the interests and rights of their own citizens or they have the collective responsibility of equally advancing the interest and rights of the combined citizenry of all states involved.⁴⁵ If these were the only choices, inclusive accountability would indeed seem to be conceptually incompatible with domestic accountability, since exercising the responsibility of equally advancing the interests of all world citizens would leave no space for exercising the responsibility of advancing the specific interests of domestic constituents. In the absence of a world state, representatives trying to meet the demands of inclusive accountability would be subject to the reproach of systematically neglecting the legitimate expectations of their own constituents.

However, once the question of the proper accountability of global institutions is framed in terms of the obligation of respecting human rights, a way out of Nagel's dilemma opens up. For we can see how the global and domestic responsibilities of representatives of member states are significantly different and can thus be discharged simultaneously. To the extent that this is so, it is hard to see how citizens of any country could legitimately expect or demand anything different from their own representatives. Let's take the WTO as an example and let's accept for the sake of the argument the view suggested by Nagel (as it is quite widespread) that this institution is designed merely to facilitate bargaining among mutually self-interested parties seeking their own advantage.⁴⁶ On such an understanding, WTO members are not trying to collectively agree upon the best trade policies for the global public interest, but rather they are trying to negotiate the best deal for themselves. Now, even in such a strategic setting, it is one thing to expect that your representatives advance your interests and rights as strongly as possible and quite another to expect that they advance your interests and rights as strongly as possible even if that involves obvious (and foreseeable) violations of the basic human rights of others.⁴⁷ Since avoiding the latter does not require giving equal weight to the interests of all world citizens beyond a relevant threshold, the obligation of respecting human rights shared by all members of a global institution seems perfectly compatible with pursuing the strategic aim of advancing the interests of those that one represents as strongly as possible while staying within the normative limits established by the prior obligation. To the extent that this plural view of obligations seems both plausible and normatively compelling, it seems that the burden of proof lies on defenders of the state-centric view to provide a normative justification for denying that human rights obligations can and should be legally entrenched in global institutions even in the absence of a world state.

Now, moving from the normative question of plausibility to the practical question of feasibility, there is no denying that the practical difficulties in the implementation of mechanisms to entrench human rights obligations in global institutions like those previously mentioned are immense. Beyond the usual fear that they could be manipulated by the powerful as much as the current mechanisms are, an additional problem is that no coherent set of criteria for evaluating the specific impact of global economic regulations on human rights protections is currently available. Obviously, such criteria won't be available any time soon unless there is prior agreement among members of the international community that they are needed. In any event, the development of such a set of specific criteria would be crucial to coordinate the work of the different international financial institutions and thus achieve greater coherence in global economic policy-making. Certainly, agreement on the specific criteria of basic human rights protections by the international community may be hard to achieve and it may also be contested whenever it is perceived as politicized or tainted by an ideological bias. But, however, imperfect, contested and in need of revision such agreements may be, implementing them would certainly be an improvement over the alternative of leaving the impact of global economic regulations on human rights protections entirely out of the equation in the decision-making process of global institutions. Establishing internal mechanisms of accountability in global institutions to guarantee that their obligations to respect human rights are discharged may prevent only the most obvious cases of gross human rights violations if the criteria agreed upon are minimal or too narrowly construed. But in the absence of any such mechanisms there is no reason to expect that even the most obvious violations will be prevented at all. Even such a modest aspiration seems out of reach unless and until there is widespread agreement on the need and the normative appropriateness of entrenching human rights obligations in global institutions.

ACKNOWLEDGEMENTS

I am grateful to the participants at the Conference on Deliberative Democracy and Global Justice (Frankfurt, Germany) and at the Workshop on Global Democracy (Prato, Italy) for very stimulating questions and comments on previous versions of this essay.

NOTES

1. For present purposes, I interpret accountability in terms of the principal-agent model. As Robert Keohane defines it, 'an accountability relationship is one in which an individual, group or other entity makes demands on an agent to report on his or her activities, and has the ability to impose costs on the agent. We can speak of an authorized or institutionalized accountability relationship when the requirement to report, and the right to sanction, are mutually understood and accepted.' See Robert Keohane and Joseph Nye, 'Redefining Accountability for Global Governance', in Governance in a Global Economy: Political Authority in Transition, ed. Miles Kahler and David A. Lake (Princeton, NJ: Princeton University Press, 2003), 139. Given that my concern here is specifically with the scope of inclusion in the accountability structures of global institutions, the principal-agent model is very useful for my analysis. However, I am not committed to the view that the principal-agent model of accountability is or needs to be the only or the most appropriate model to be used when thinking about the emergent structures of global governance. For an interesting analysis of the difficulties in applying traditional principal-agent models of accountability at the global level, see Joshua Cohen and Charles Sabel, 'Global Democracy?', NYU Journal of International Law and Politics 37 (2005), 763-97.

2. For example, hierarchical and supervisory accountability, legal accountability, peer accountability, reputational accountability, and market accountability. For a detailed discussion about different types of accountability, see Keohane and Nye 'Redefining Accountability for Global Governance'.

3. I am thankful to an anonymous reviewer for pressing the need to clarify the precise sense in which I use the term here.

4. Fears that the answer may turn out to be negative leads some authors who clearly endorse electoral sovereignty at the national level to either reject democracy at the global level [see for example, John Rawls, The Law of Peoples (Cambridge, MA: Harvard University Press, 1999); Jürgen Habermas, 'A Political Constitution for the Pluralist World Society?', in Between Naturalism and Religion (Cambridge, MA: MIT Press, 2008), 312-52] or to eliminate electoral sovereignty as a requirement for global democracy [see for example, John Dryzek, Deliberative Global Politics (Cambridge: Polity Press, 2006); J. Bohman, Democracy Across Borders (Cambridge, MA: MIT Press, 2007]. Although I find these reactions unconvincing, I must confess that I do not yet see how electoral sovereignty could be implemented in global institutions in the absence of a world state. I certainly would welcome institutional proposals that show how to democratize current institutions of global governance by granting a final say over some of its decisions to all those subject to them in the absence of a world state. Many of the current proposals in that direction [e.g. the use of cross-national referenda regarding the implementation of contested global regulations, see David Held, Global Covenant. The Social Democratic Alternative to the Washington Consensus (Cambridge: Polity Press, 2004), 113] do give the impression that their proper implementation would require the replication of state political structures at the global level.

5. For example see, Robert Dahl, 'Can International Organizations Be Democratic? A Skeptic's View', in Democracy's Edges, ed. Ian Shapiro and Casiano Hacker-Cordón (Cambridge: Cambridge University Press, 1999), 19-36.

6. The usual dilemmas are not very different from those that arise at the national level. Enhancing publicity and transparency in decision-making by increasing public access to information may impair the institution's ability to reach the most efficient decisions or may lead to polarization and exacerbate instability. Consensus rules of decision-making can lead to inaction and award excessive influence to smaller countries, whereas replacing them by majoritarian rules may marginalize small countries and silence important concerns, etc. For a detailed analysis of some of these dilemmas regarding global financial institutions see, M. Kahler, 'Defining accountability up: the global economic multilaterals' in Global Governance and Public Accountability, ed. David Held and Matthias Koenig-Archibugi (Oxford: Blackwell, 2004), 8-34.

7. I am borrowing Held's terminology to refer to what he calls 'the principle of inclusiveness' or 'the principle of stakeholder-decision-maker equivalence' [see, Held Global Covenant, pp. 13, 103 and David. Held, 'Democratic accountability and political effectiveness from a cosmopolitan perspective', in Global Governance and Public Accountability, ed. D. Held and Matthias. Koenig-Archibugi (Oxford: Blackwell, 2005), 246-8]. However, as already mentioned, in this paper I am exclusively focusing on the dimension of accountability and not addressing the dimension of sovereignty in the principle of inclusion. Held does not explicitly distinguish these two dimensions, so he often uses the principle of inclusion in a way that involves sovereignty, for example when he claims that, according to the principle, decision-takers must 'have a say' in decision-making. If by 'having a say' one means that decision-takers ought to have some decision-making status, then this is a stronger principle of accountability than the one I am referring to here, since it entails a requirement of electoral sovereignty. In the weaker sense that I am using the principle of inclusion here, inclusive accountability does not require that decision-takers 'have a say' in decision-making, but only that their interests and rights are taken into account by decision-makers or, to borrow Cohen and Sabel's rendering of the norm of inclusion, that in the decision-making process their interests and rights are not treated 'as if, beyond the humanitarian minimum owed even in the absence of any cooperation, they count for nothing.' [J. Cohen and C. Sabel, 'Extra Rempublicam Nulle Justitia?', Philosophy & Public Affairs 34/2 (2006): 154].

8. The aim of my argument is pretty limited, among other things, because it focuses exclusively on the dimension of accountability in the principle of inclusion and it does not address any of the important issues related to the dimension of sovereignty (e.g. how to achieve equality of representation among members in global institutions, which rules of decision-making they should follow, etc.).

9. For an excellent analysis and overview of this problem see R. Keohane and J. Nye, 'Redefining Accountability for Global Governance', in Governance in a Global Economy: Political Authority in Transition, ed. M. Kahler and D. Lake (Princeton, NJ: Princeton University Press, 2003), 386-411. There is much that I agree with in this analysis. In particular, I share Keohane's judgment that, among all the accountability gaps present in current institutions of global governance, 'the most serious normative problems arise with respect to what I call external accountability: accountability to people outside the acting entity, whose lives are affected by it.' (141) However, I find his distinction between internal and external accountability problematic in light of the proposal I would like to defend here. According to his distinction, 'internal accountability' holds when the principal and the agent are 'institutionally linked to one another,' either because the principal is providing resources to the agent or because it has directly authorized the agent to act on its behalf. A paradigmatic example of this type of accountability relationship is between citizens and their representatives in global institutions. By contrast, 'external accountability' holds in the absence of any institutional link between the agent and what Keohane calls, for that reason, the 'would-be principals'. It holds just in virtue of the fact that the latter are affected by the decisions of the former. His own example is the relationship between the World Bank and African farmers who are affected by the Bank's policies. The main problem with this distinction is that any proposal that aims to establish an institutional link between those affected by global decisions and the agents who make them cannot be accommodated under either of the two categories. To the extent that such a link would be based merely on impact (and not on authorization or resources), it would fall under Keohane's category of 'external accountability', but to the extent that it would be institutionalized within the agencies at issue, it would also fall under the category of 'internal accountability'. The irony behind this difficulty with Keohane's distinction is that, although he is very aware of the normative importance of making institutions of global governance accountable to all of those who are affected by their decisions, his distinction seems to leave no space for the possibility of establishing institutional mechanisms to reach that goal.

10. For a dramatic illustration of this problem see Note 47.

11. I leave aside the obvious additional problems that arise from the fact that the proximate principals of delegates in global financial institutions are governments and that their political agenda and perception of the national interest often differs from those of their citizens. This is especially (although not only) the case with governments of countries that have no democratic structures of accountability.

12. For example of this line of argument see, Rainer Schmalz-Bruns, 'An den Grenzen der Entstaatlichung. Bemerkungen zu Jürgen Habermas' Modell einer 'Weltinnenpolitik ohne Weltregierung', in Anarchie der kommunikativen Freiheit. J. Habermas und die Theorie der internationalen Politk, ed. Peter Niesen and Benjamin Herborth (Frankfurt: Suhrkamp, 2007), 369-93.

13. Thomas Nagel, 'The Problem of Global Justice', Philosophy & Public Affairs 33/2 (2005): 139-40. Nagel's argument in this passage tackles two questions at once. One is the question of 'inclusion', that is, the question of whether global institutions have responsibilities toward all world citizens even though their members only have primary responsibilities towards the citizens of the countries they represent. The other is the question of the 'content' of those responsibilities, that is, the question of whether or not the responsibilities of global institutions towards world citizens are as strong as the responsibilities of social justice that national institutions have toward their own citizens. The focus of this paper is on the first question only, but in the last section I briefly indicate how the proposal I defend bears on some of Nagel's claims regarding the second question.

14. As it should become clear in what follows, my aim is not to challenge the primary responsibility of states in protecting the human rights of their citizens. I entirely agree that states should bear primary responsibility. What I aim to challenge within the state-centric conception is the claim that only states have primary responsibility in protecting human rights and therefore non-state actors do not have human rights obligations.

15. Charles Beitz, The Idea of Human Rights (Oxford: Oxford University Press, 2009), 13.

16. Ibid., 44.

17. See R. Goodin, 'Globalizing justice', in Taming Globalization, ed. D. Held (Oxford: Polity Press, 2003), 76f.

18. The possibility of prosecuting leaders of non-state armed rebel groups for human rights violations such as genocide or ethnic cleansing, irrespective of any recognition of official status as agents of a state, is just one example of the disconnect between the state-centric conception of human rights and current human rights practice. On this issue, see Andrew Clapham, Human Rights Obligations of Non-State Actors (Oxford: Oxford University Press, 2006), 271-316.

19. The possibility of prosecuting multinational corporations in US Courts for violations of international human rights law under the Alien Tort Claims Act is another example of current human rights practice that cannot be easily accounted for within the limits of the state-centric conception. For a good overview of the existing international regimes that cover the human rights obligations of corporations see Andrew Clapham, Human Rights Obligations, 195-270. See also, Philip Alston, ed. Non-State Actors and Human Rights (Oxford: Oxford University Press, 2005).

20. Beitz, The Idea of Human Rights, 39-40.

21. It is precisely because of the far reaching implications of accepting human rights obligations that the amendment to the TRIPS Agreement regarding patents for pharmaceuticals recently undertaken by the WTO, to take Beitz's own example, carefully avoids any use of the term 'human rights' in official documents. Instead of using the term 'right to health', less committal expressions such as 'concern for public health' are employed; instead of referring to the WTO members' obligations to protect the human right to health (and to access to medicines) of their citizens, less committal expressions are used like in the following example: 'the Agreement can and should be interpreted and implemented in a manner supportive of WTO member's right to protect public health and, in particular, to promote access to medicines for all' (my emphasis). For more information on this specific reform see http://www.wto.org/english/tratop_e/trips_e/implem_para6_e.htm. For an excellent analysis of the WTO's human rights obligations under international law regarding the TRIPS agreement and the right to health see Holgar Herstermeyer, Human Rights and the WTO. The Case of Patents and Access to Medicines (Oxford: Oxford University Press, 2007).

22. Since my concern here is with the state-centric view of human rights obligations, I do not address the separate and complex issue of whether the practical conception of human rights offers a plausible (or even the best) account of the nature of human rights among the many alternative conceptions currently under discussion. For other versions within the broad family of political conceptions of human rights see Rawls, The Law of Peoples; Thomas Pogge, World Poverty and Human Rights (Cambridge: Polity Press, 2002); Joseph Raz, 'Human Rights Without Foundations', Oxford Legal Studies Research Paper 14/2007 (2007): 1-21. The main alternatives to the practical conception that Beitz discusses are naturalistic conceptions [see, e.g. Maurice Cranston, What Are Human Rights? (London: Bodley Head, 1973); James Griffin, On Human Rights (Oxford: Oxford University Press, 2008)]; and agreement conceptions [see, e.g. Taylor 1999; J. Cohen, 'Minimalism about Human Rights', Journal of Political Philosophy, 12 (2004): 190-213].

23. Beitz, The Idea of Human Rights, 105.

24. As mentioned in a previous note, evaluating the plausibility of the practical conception as an account of human rights is beyond the scope of the present discussion. While I am sympathetic to the practical conception, the main reason I focus on it here is because it may seem the hardest candidate to challenge the state-centric view, since this view is indeed predominant in current human rights practice. However, as I try to show, this is an erroneous impression. For the practical approach is not guided just by the factual features of current practice but by its underlying normative aims. Given that the aim of human rights practice is to secure protection of human rights worldwide and that under current circumstances this aim is seriously undermined unless human rights norms regulate the behavior of a larger range of actors than just states, a pluralist view of human rights obligations seems perfectly justified on the basis of the practical conception of human rights. By comparison, other conceptions of human rights do not pose a similar challenge for defending a pluralist view of human rights obligations, since the claim that human rights consists 'of norms to regulate the behavior of governments' is neither required nor prima facie plausible outside the constraints of the practical conception. If one takes the Universal Declaration of Human Rights (UDHR) at face value, human rights seem to be first and foremost rights that every human being has regardless of any political boundaries. Thus, even if one adopts an institutional conception of human rights, nothing in the UDHR suggests that states are the only institutions with human rights obligations. In fact, as Beitz himself points out, article 28 seems to directly undermine the state-centric assumption by establishing an explicit link between human rights and the international order (Everyone is entitled to a social and international order in which the rights and freedoms set forth in this Declaration can be fully realized.). According to this article, human rights obligations fall on whichever institutions are responsible for the international order at any given historical time, whether or not they happen to be states. For an example of an institutional conception of human rights that is not state-centric, see Pogge, who defines human rights as follows: 'a human right to X entails the demand that, insofar as reasonably possible, any coercive social institutions be so designed that all human beings affected by them have secure access to X' (World Poverty and Human Rights, p. 46, My emphasis).

25. For some evidence against the claim that the current human rights practice is exclusively state-centric see Clapham, Human Rights Obligations; and August Reinisch, 'The Changing International Legal Framework for Dealing with Non-State Actors', in Non-State Actors and Human Rights, ed. Alston (Oxford University Press, 2005), 37-89.

26. Beitz, The Idea of Human Rights, 105.

27. This aim is explicitly identified in the main international human rights documents. Article I of the United Nations Charter identifies 'promoting and encouraging respect for human rights and for fundamental freedoms for all' as one of its main aims. In the Preamble of the Universal Declaration of Human Rights (UDHR) it is stated that 'the General Assembly proclaims this Universal Declaration of Human Rights ... to the end that every individual and every organ of society ... shall strive ... to promote respect for these rights and freedoms and by progressive measures, national and international, to secure their universal and effective recognition and observance' (My emphasis).

28. I am grateful to Charles Beitz for making me see the need to clarify this point.

29. This particular terminology was introduced by Asbjørn Eide, The New International Economic Order and the Promotion of Human Rights. Report on the Right to Adequate Food as a Human Right, UN Doc E/CN.4/Sub.2/1987/23. The conceptualization of the multiple obligations structure applicable to all human rights expressed in this tripartite division was originally proposed with a different wording by Shue [see, Henry Shue, Basic Rights. Subsistence, Affluence and U.S. Foreign Policy (Princeton: Princeton University Press, 1980)].

30. Some authors designate this obligation as a 'duty of vigilance,' i.e. the duty to ensure as far as practicable that the institution's actions do not negatively impact the ability of its members to implement their international human rights obligations. See, Mac Darrow, Between Light and Shadow: The World Bank, the International Moneraty Fund and International Human Rights Law (Oxford: Hart Publishing, 2003), 132, 137.

31. For a comprehensive overview of the vast legal literature on this issue see Sigrun Skogly, The Human Rights Obligations of the World Bank and the International Monetary Fund (London: Cavendish Press, 2001).

32. The letter can be found at the following address within the IMF website (http://www.imf.org/external/np/vc/2001/090401.htm). A longer version of the letter was published later in the IMF's magazine Finance & Development. This longer version is also available on the IMF website at the following address (http://209.133.61.129/external/pubs/ft/fandd/2001/12/leite.htm). In the longer version, the titles of the sections where each of the passages quoted here can be found are revealing. They read 'What is the IMF's contribution to human rights?' and 'Do IMF-supported programs harm economic, social, and cultural rights?', respectively.

33. The term 'human rights' is not explicitly used in the sentence that describes the IMF's obligations, although when taken in context it is clear that the enjoyment of human rights is the elliptical object of reference for the 'adverse side effects' explicitly mentioned in the sentence. As already mentioned in relation to the WTO in Note 21, the official documents of all three institutions (IMF, WTO, and World Bank) go to great lengths to avoid the explicit use of the term 'human rights' when describing their own obligations and responsibilities.

34. See Gudrun Zagel, 'WTO and Human Rights: Examining Linkages and Suggesting Convergence', Voices of Development Jurist Paper Series 2/2 (2005): 1-37. Even those legal scholars who aim to derive obligations to protect human rights from some of the specific goals of the WTO agree that this institution is not mandated to enforce human rights law, but only WTO law. Their claim is rather that WTO law must be interpreted and implemented in a way consistent with any relevant human rights law. For more on the importance of this distinction see Clapham, Human Rights Obligations, 164ff.

35. In fact, in 2001, the WTO Secretariat acknowledged that the WTO is bound by the customary international law of human rights. See UN Doc. E/CN.4/Sub.2/2001/SR.12, at para. 24, Summary Record of the 12th meeting on 8 August 2001.

36. Unsurprisingly, legal scholars agree that WTO law must be interpreted consistently with international human rights law (this opinion is shared by members of the Legal Affairs Division of the WTO; see G. Marceau, 2002, p. 755) However, scholars follow different paths in order to derive the obligation to respect human rights from specific features of the WTO. For an example see Zagel, ''WTO and Human Rights'', pp. 31-32.

37. For a good overview of the main proposals currently under discussion by legal scholars see Herstermeyer, Human Rights and the WTO, 209-29.

38. See Zagel, 'WTO and Human Rights', 30-1. A similar proposal, although without specific reference to human rights, is made by Stiglitz and Charlton, see Joseph Stiglitz and Andrew Charlton, Fair Trade for All. How Trade can Promote Development (Oxford: Oxford University Press, 2005), 7.

39. See Zagel, 'Human Rights Accountability of the WTO', Human Rights & International Legal Discourse 1/2 (2007): 367.

40. World Bank, 1998.

41. World Bank, 2002 at 2.

42. Daniel Bradlow, 'International Organizations and Private Complaints: The Case of the World Bank Inspection Panel', Virginia Journal of International Law 34/3 (1994): 554; Ngaire Woods and Amrita Narlikar, 'Global Governance and the Limits of Accountability: The WTO, the IMF, and the World Bank', International Science Journal 53/170 (2001): 576-77. For additional examples of accountability mechanisms adopted by other international institutions in response to the increased demands for 'good governance,' see Reinisch, 'The Changing International Legal Framework', 50f.

43. In pointing out the compatibility of my proposal with Nagel's statist view of social justice my intention is not to endorse this view but rather to show that, even if one endorses it, it still fails to provide convincing reasons for rejecting the ascription of human rights obligations to global institutions.

44. Nagel, 'The Problem of Global Justice', 141. For a convincing criticism of this claim see Cohen and Sabel, 'Extra Rempublicam Nulle Justitia?,' 171.

45. This assumption is quite widespread, not just among critics like Nagel, but even among authors who aim to defend the claim that members of global institutions and regional networks ought to be subject to both global and domestic accountability. Anne-Marie Slaughter offers an example. On the one hand, she claims, very plausibly in my view, that the members of government networks 'first, must be accountable to their domestic constituents for their trans-governmental activities to the same extent that they are accountable for their domestic activities. Second, as participants in structures of global governance, they must have a basic operating code that takes account of the rights and interests of all peoples.' (39; my emphasis) However, her account of the latter responsibility seems too strong to avoid the criticism that it leaves no space for a meaningful exercise of the former responsibility. She explains: 'Even if participants in government networks around the world were satisfactorily accountable to their domestic constituents, what duty do they owe to other nations? It may seem an odd question, but if these networks were in fact primary structures of global governance ... then they would have to be subject to global as well as national norms. They would be responsible for collectively formulating and implementing policies in the global public interest' (51; my emphasis). Unfortunately, Slaughter does not explain how these two responsibilities can be simultaneously exercised in the absence of a world state. See Anne-Marie Slaughter, 'Disaggregated Sovereignty: Towards the Public Accountability of Global Government Networks', in Global Governance and Public Accountability, ed. D. Held and M. Koenig-Archibugi (Oxford: Blackwell, 2005), 35-66.

46. In spite of WTO's official commitment to the goals of raising the standard of living and of sustainable development, many critics maintain that the design of the WTO as a voluntary association to facilitate trade among its members makes it simply a marketplace for bargaining in which its members are not trying to collectively agree on the best trade policies for everyone, but rather trying to negotiate the best deal for themselves. In his book The Bottom Billion, Collier offers a colorful characterization of the WTO along these lines: 'It (the WTO) is not a purposive organization but rather a marketplace. The WTO secretariat is there merely to set up the stalls each day, sweep the floor each evening, and regulate the opening hours. What happens is determined by the bargaining ... The present round of trade negotiations was termed a "development round", but such labels really have no possibility of content in an organization designed for bargaining. You might as well label tomorrow's trading on eBay a "development round". Trade negotiations are there to get the best deal for their own country, defined in terms of the least opening of the home market for the maximum opening of others.' (170-1). The legal development of the WTO and its internal institutions suggests that this narrow assessment of its functioning is likely to become less and less accurate over time, but even if one accepts it as an accurate description of how its members and the citizens they represent understand the role of this institution, as I do in the text for the sake of argument, it would still seem entirely unjustified to claim that this institution has no obligation to respect human rights. It should be noted that, in spite of his cynical assessment of the current design of the WTO, Collier does not see any impediment to his ambitious proposal for reform of this institution, which involves 'adding a transfer role to its bargaining role' with the explicit aim of helping the poorest countries at the expense of the strongest economic interests of the richer ones. As he explains, 'by a transfer I mean an unreciprocrated reduction in trade barriers against the bottom billion: a gift, not a deal ... The secretariat of the WTO should be charged with negotiating such a gift as the first phase of each round' (171). As a former director of the World Bank, Collier takes the Bank's own development as a model for the WTO: 'The Bank evolved by adding a transfer role targeted on low-income countries to what was originally a mutual assistance role for richer countries. That is what should happen to the WTO.' (ibid.). See Paul Collier, The Bottom Billion (Oxford University Press, 2007).

47. President Clinton's recent public apology for championing trade policies that destroyed Haiti's rice production offers a clear illustration of the normative plausibility of the conception of inclusive accountability defended here. Speaking to the Senate Foreign Relations Committee on March 10 about his push for dramatic tariff cuts on imported U.S. rice in Haiti, Clinton declared 'It may have been good for some of my farmers in Arkansas, but it has not worked. It was a mistake ... I had to live everyday with the consequences of the loss of capacity to produce a rice crop in Haiti to feed those people because of what I did.'

Ethics & Global Politics eISSN 1654-6369, ISSN 1654-4951

This journal is published under the terms of the Creative Commons Attribution-Noncommercial 3.0 Unported License. Responsible editor: Eva Erman.