Collective responsibility and the scope of justice

Abstract

The paper examines Thomas Nagel's ‘political conception’ of justice that holds that the requirements of socioeconomic justice apply only among those persons who are subject to the authority of the same political institutions. The paper has two aims. The first aim is to clarify the theoretical motivation for Nagel's theory, which it identifies in what it calls the ‘responsibility thesis’, i.e. that inequalities may be considered as unjust only if some agents are responsible for them, and to reconstruct the account of collective responsibility that, together with the responsibility thesis, may support his substantive conclusions. The second aim is to show that Nagel's conclusions must be rejected even if the account of collective responsibility that may support it is correct. This is so, first, because the responsibility thesis cannot be defended, and second, because even if the thesis is assumed to be correct, it does not succeed in restricting the scope of the requirements of justice to fellow citizens of particular nation states. In light of Nagel's own account of the link between legitimate authority and justice, the standards of the latter ought to apply to the state system as a whole, and not to particular states, taken separately.

• responsibility
• global justice
• inequality
• legitimacy
• Thomas Nagel
• state system

Non-associative conceptions of justice hold that the principles of distributive justice apply to the whole of humanity, regardless of the institutional or other relations that may obtain among particular persons.1 Associative conceptions hold, by contrast, that the proper object of evaluation of principles of justice is the position of those persons who stand in some associative relation with one another that is claimed to trigger the requirements of justice. The relevant associative relation may be a scheme of cooperation that distributes important benefits and burdens,2 autonomy-restricting coercive relations,3 or something else. In this paper I will discuss a specific version of the associative conception, one that is sometimes referred to as the ‘political conception’ of justice and that has been outlined by Thomas Nagel.4 The political conception holds that the demands of egalitarian distribution arise only among persons who are subject to the authority of the same state. Unlike autonomy-based accounts, however, it does not identify the relevant aspect of coercive institutions in their autonomy-restricting nature but in the relations of collective responsibility that they are suggested to generate. It holds that coercive rules that are imposed in the name of those who are coerced make the coerced collectively responsible for the effects of the rules by normatively engaging their agency, and therefore such institutions create a new justificatory burden that can be discharged only through an egalitarian distribution of socioeconomic advantages. In this paper I will attempt to reconstruct the account of collective agency and responsibility that may provide support for the political conception. I will then argue that the political conception fails even if the theory of collective responsibility that is suggested to underlie it is correct.

It is important to note at the outset that Nagel's defense of the political conception leaves many interpretive questions open, and therefore a significant part of the discussion consists in developing a plausible rationale for the various claims that he makes in support of his position. Furthermore, in the spirit of criticizing his account in its strongest version, I provide arguments for his position where they are only implied in the text, or when there are gaps in the account. To this extent, the interpretation goes beyond reconstruction and searches for the grounds, not mentioned by Nagel himself, which may provide support for the political conception. That is to say, I will attempt to complete rather than simply reconstruct what I take to be the most promising understanding of the political conception of justice. Then, I will argue that the political conception is not successful even in the supplemented form that I outline here.

Nagel's paper has generated a large volume of critical responses.5 Many of these question Nagel's claim that the relevant associative relations are indeed present only under circumstances of shared citizenship.6 Others reject the relevance of that relation outright.7 The contribution of this paper to the critical literature is twofold. First, it proposes a detailed rationale for the political conception by reconstructing the account of collective responsibility that may support it. In this way, it also explicitly situates the political conception within the growing literature on collective responsibility. Second, by providing this reconstruction of the political conception and criticizing it in that amended form, it makes the criticism relevant for other versions of the political conception, beyond that of Nagel. Thus, the criticism is of more general interest in the global justice debate. Nevertheless, since Nagel's version has attracted the most attention, the presentation will mostly follow his account.

THE POLITICAL CONCEPTION

In Nagel's presentation, the political conception, the content of the requirements of socioeconomic justice is about eliminating or mitigating the unequal distributive effects of morally arbitrary factors, such as the wealth of one's parents, one's sex or native talent, i.e. such factors that are beyond one's control.8 However, in the political conception ‘the presumption against morally arbitrary inequalities is not a principle of universal application’9: rather, arbitrary inequalities are unjust only if they are generated by ‘coercively imposed legal and political institutions’10. Furthermore, only coercive rules that are imposed in the name of those subject to them generate the presumption against morally arbitrary inequalities.11

It seems that Nagel's political conception draws its motivation from the more general idea that not all morally arbitrary inequalities are equally problematic from the point of view of justice; arbitrary inequalities produced by agents and social mechanisms are more troubling than ‘purely’ natural inequalities, even though the individuals suffering natural disadvantages are no more deserving of them than those suffering socially produced inequalities.12 This is sometimes referred to as the ‘social structural view’ of justice.13 The view holds that being the outcome of the operation of agent(s) is a necessary condition for an inequality to count as unjust—natural inequalities are thus not (in themselves) matters of justice. However, the political conception is narrower than the social structural view, because social mechanisms (such as global trade, for instance) that do not occur within shared political institutions coercively imposed in the name of all those subject to them can also produce arbitrary inequalities. These may count as subject to justice on the generic social structural view, but not on the narrower political conception. Therefore, further relevant distinctions between different types of social mechanisms need to be drawn in order to make sense of the political conception. We need to identify the rationale for two aspects of Nagel's political conception: first, that it is only inequalities produced by social mechanisms occurring within coercively imposed rules that create a problem of justification,14 and second, that the rules in question should be imposed in the name of those subject to them to give rise to the problem of justification.15

Nagel provides some guidance to identify the rationale. He writes that coercively enforced rules imposed in ‘our’ name make ‘us’ (i.e. those subject to the coercive rules) responsible for their effects, including for the arbitrary inequalities that they generate or permit,16 and those (and only those) who are made responsible for the effects of the rules in this way ‘have standing to ask’ why they should accept them, i.e. they (and only they) are owed justification for them.17 At other points, he implies that it is only properly collective agents’ acts, with corresponding collective responsibility, that trigger the demands of justice.18 He suggests that the requirements of justice are triggered by a threshold level of ‘collective engagement’ that is currently satisfied only by institutions with strongly centralized decision-making procedures performed in the name of those subject to them.19

These remarks suggest that a state of affairs is an appropriate object of evaluation by the principles of justice only if some agent is morally responsible for the fact that it exists.20 Let us call this the responsibility thesis. The social structural view is not committed to the responsibility thesis because not all socially produced inequalities are such that some agent is morally responsible for them in the sense that they (the inequalities) could be the proper basis of morally appraising it (the agent).21 This can be the case if individuals cannot avoid contributing to the aggregate outcome, or cannot possibly calculate the combined effects of their individual acts. When this is the case, it is possible that none of the agents contributing to the aggregate outcome are individually responsible for it in the sense that the aggregate outcome could provide the basis of morally appraising them (individually). Nagel's suggestion, on the present construal, is that such inequalities are not appropriate objects of evaluation by justice unless the individuals contributing to them are constituents of a collective agent to whom responsibility can be attributed for the outcome. The intuition supporting the responsibility thesis might be this: to hold a state of affairs to be unjust is to suggest that some agents are obligated to eliminate or mitigate it. But it is appropriate to hold agents to such task-responsibility only if they are responsible for the state of affairs in the backward-looking sense as well. Therefore, there is no injustice without some agent being responsible22 for it; thus, the social structural view, which logically admits some inequalities for which no one is responsible as unjust, is too broad.

The responsibility thesis thus holds that only those socially produced inequalities can be unjust where the contributing individuals constitute a collective agent as a separate center of responsibility attribution. This raises the need for an account specifying the conditions under which assertions of collective agency and responsibility are warranted. Nagel's remarks about collective agency suggest that he explores just such an account. A critical part of Nagel's political conception is that having centralized decision-making and enforcement is a necessary condition of a group of several individual agents to constitute a collective agent and to be collectively responsible for some outcomes. His point is that various agents that are not acting within a centrally imposed set of rules cannot be attributed responsibility for the aggregate effects of their otherwise permissible actions. To be sure, each of them is responsible for their own acts, and may be blameworthy if they act in individually impermissible ways. But the crucial point is that in the absence of centrally imposed rules, there is no such agent that could be responsible for the overall distributive effects, including seemingly unfair inequalities, of otherwise separately permissible actions. And the implication is that since no one is responsible for these inequalities, they cannot be unjust. In other words, the account implies a connection between centralized decision procedures, the existence of collective agents that are the proper object of collective responsibility attribution, and the possibility of injustice. The task now is to clarify this connection. As noted above, Nagel does not himself provide the details of a theory of collective agency. For this reason, the following discussion of collective agency and responsibility cannot be ascribed to him: I discuss it to strengthen and illuminate the various points that he makes.23

COLLECTIVE RESPONSIBILITY AND CENTRALIZED DECISION-MAKING

The claim that links collective agency to centralized decision-making may draw support from an influential account of collective agency and responsibility developed by Philip Pettit and others. This view holds that for an entity to qualify as an agent, and therefore to be the proper object of responsibility attribution, it must be able to evaluate and choose from the alternatives available to it, and must be able normally to follow through on its choices. Furthermore, it must be able to choose and act consistently over time.24 Several individuals constitute a distinct collective agent only if there are established procedures of evaluating options and making decisions. Through standing decision procedures the collective takes on agential powers of its own that are distinct from those of its individual members. And by having effective enforcement powers, it acquires the capacity to execute its choices by being able to compel its constituent members.25

It is important to see why on this account of collective responsibility it is possible to speak of collective agents as distinct from their individual constituent members, and what the normative significance of this possibility is. It may be helpful to start with the second question, to see why it might be normatively desirable to be able to identify collective agents with their corresponding collective responsibility. There are many cases discussed in accounts of collective responsibility in which a corporation consisting of several individuals is culpable of causing some significant harm to others, but it is also the case that none of its individual members acted in a way that would merit attributing individual responsibility to him or her for the harm. In such cases, absent the possibility of attributing responsibility to the group as a whole, one is faced with the dilemma of having to choose between two unattractive alternatives: either to declare that no one is responsible for the harm and therefore no one is liable to compensate for it, or to attribute responsibility—and thus liability—to some individual(s) alone even though their actions, taken separately, do not warrant this. In such cases there is a ‘responsibility shortfall’26: the sum of individual responsibilities does not add up to the level corresponding to the size of the harm suffered by the victims of the group's acts. By contrast, if the individuals constitute a distinct, collective agent, then it may be attributed responsibility for the harm, and the dilemma can be avoided. Likewise, it is often the case that many individuals’ permissible actions generate morally arbitrary inequalities in such a manner that it would be mistaken and wrong to attribute individual responsibility to any one of them for the unfairness. In cases like this, it would be morally desirable to be able to assert that the individuals and other contributing entities, taken together as a group, are responsible for the resulting inequalities.

To be sure, the fact that in such cases it would be morally desirable to attribute responsibility to the group as a whole does not by itself warrant the appropriateness of such responsibility attribution. The present reconstruction of the political conception sees it exactly as an attempt that enlists a theory of collective responsibility to clarify the conditions under which responsibility for arbitrary inequalities arising as a consequence of many agents’ acting permissibly can be appropriately attributed to the group as a whole. And the point is precisely that when those conditions do not obtain, no one can be held responsible for the inequalities and they are therefore not unjust. In other words, in cases like that Nagel is ready to embrace the first horn of the dilemma described in the previous paragraph.27 The suggestion is that one of the necessary conditions of such collective responsibility attribution is that the group agent is in some genuine sense distinct from its individual members, and that this in turn depends on there being centralized decision procedures governing group decisions. Unless there is such a distinct collective agent, we are stuck with the responsibility shortfall.

What is the significance of central decision-making for the existence of a distinct group agent? The answer can be grasped if we consider the ‘debunking view’ that suggests that group agency is always completely reducible to individual agency. The group as such never has attitudes and choices that are distinct from those of its members; they are simply a logical function—such as majority or something similar—of the relevant attitudes or choices of its constituent individual agents.28 If the debunking view is right, then the group's responsibility can never exceed the sum of the individual responsibility of its members, and therefore the ‘responsibility gap’ cannot be closed: we are left with the dilemma discussed above. However, contrary to the debunking view, groups with a certain structure do indeed qualify as distinct agents some of whose attitudes are not any function of the corresponding attitudes of their members. Pettit illustrates this through the so-called ‘discursive dilemma’, which shows that groups that have to make consistent decisions over time about logically related issues must necessarily structure themselves exactly in a manner that at least some of their choices are not functionally related to the choices of their members. This is so because it is possible under any aggregation rule that even if all members of a group hold judgments about logically related questions that are individually consistent, when aggregated, the individually consistent judgments produce collectively inconsistent ones.29 Therefore, for a group to be able to make consistent decisions on interrelated issues over time (a necessary condition of qualifying as an agent), it must have centralized decision procedures that enable individual officials holding certain positions within the group to make autonomous decisions that are not simply aggregations of the judgments of the individuals constituting the group.30 And as a result of such independence from the judgments of the constituent members, at least some of the group's decisions will not be any function of its members’ judgments. Therefore, contrary to the debunking view, groups with such structure qualify as distinct agents, as well as separate objects of responsibility attribution.31

We are now in a position to identify a rationale for the claim that only groups with centrally imposed and enforced decision procedures can be responsible for the inequalities generated within them by their members’ permissible actions. Only groups with such structures can qualify as distinct agents, and can close the responsibility gap, if any, that is left by its individual members. If the responsibility thesis is true, then the existence of such an agent is a necessary condition for an arbitrary inequality to constitute an injustice, at least in those cases when none of the individuals constituting the group is individually responsible for it.

Let's sum up the takeaway from the discussion in this section. The starting point was the responsibility thesis claiming that an inequality is a proper object of evaluation by the principles of justice only if there is an agent that is morally responsible for creating it. However, there are socially produced inequalities such that none of the individuals contributing to them are individually morally responsible for the outcome. Such outcomes are proper objects of evaluation by justice (according to the responsibility thesis) only if the contributing individuals constitute a collective agent. And finally, separate agents constitute a collective agent only if they are subject to shared centralized decision-making and enforcement mechanisms. Therefore, a state of affairs produced by many agents acting permissibly is an object of evaluation by justice only if they are under shared centrally imposed coercive rules.

NORMATIVE ENGAGEMENT

However, as it has been noted above, in Nagel's view the presence of centrally imposed and enforced rules is only necessary but not sufficient to trigger the requirements of justice. The rules have to be imposed in the name of those subject to them for the justificatory burden to arise with respect to their effects. It is plausible to understand the ‘in their name’ condition as stating that the group makes the claim that it acts with authority: it does not merely force its members to act in certain ways but claims to have a right to do so and thus expects them to comply not only out of fear but on moral grounds.32 When the state acts with authority, its citizens have moral reasons to comply with its decisions and thus their will is ‘normative engaged’33. Nagel makes two related claims in support of this further condition. First, rules enforced in this way (with claimed authority) make those individuals subject to them also responsible for their effects: ‘[t]he society makes us responsible for its acts […] Insofar as those institutions admit arbitrary inequalities […], we are responsible for them’34. And second, only because they are made responsible in this way for the effects of coercively imposed rules do these individuals ‘have standing’ to ask for a justification.35 In other words, if the group agent acts with authority, its responsibility for the actions and their effects distributes over its members, and therefore they are owed justification for them.

Nagel does not clarify whether he refers to responsibility in the attributability sense or in the forward-looking sense when he says that we are ‘made responsible’ for something in this manner. I will return to this ambiguity later, because depending on how we resolve it, we get two different rationales for the restriction of the justificatory burden to those in whose name institutions are imposed.

Thus, the political conception endorses two separate restrictions regarding the application of the requirements of justice. First, it holds that inequalities are proper objects of assessment by the requirements of justice only if they occur within shared political institutions. And second, it holds that the distributive effects of the operation of institutions ought to be justified only to those people who share in their responsibility, i.e. who are their constituent members. Jointly these restrictions imply that only inequalities among people who are subject to shared political institutions imposed in their name may be unjust.

To recap, Nagel's version of the political conception may be understood as having the following structure:

1. Inequalities can be unjust only if some agent is morally responsible for them.

2. When an inequality is caused by several separate agents acting in separately permissible ways, the inequality can be unjust only if the several agents constitute a distinct collective agent that can be attributed responsibility for it.

3. Only groups with centralized decision-making and enforcement can constitute such collective agents.

4. People are owed justification for the distributive effects of a collective agent's acts only if they share its responsibility.36

5. People share a collective agent's responsibility only if it acts ‘in their name’, i.e. with authority over them.

6. Therefore, only people living under shared coercive political institutions operating in their name are owed justification for distributive effects.

I submit that this account provides a plausible reconstruction of the political conception, whether or not. Though the starting point has been Nagel's account, the reconstruction reaches beyond that and offers a rationale for the political conception that may be appealing to a broader circle of theorists who find various elements of this approach convincing. I am not suggesting that this is the only plausible reconstruction of the political conception, but only that it is plausible enough to merit scrutiny. In this reconstruction, 1) states the responsibility thesis in general, and 2) specifies it for cases where many people's separate acts contribute to an inequality. I take 3) to be a plausible claim about collective responsibility and its distribution, and therefore I will accept them for the purposes of this paper. But I will argue that the responsibility thesis is untenable, and therefore 6) fails. I will also present arguments against 4) and 5). These arguments undermine Nagel's view even if the argument against the responsibility thesis fails.

NO INJUSTICE WITHOUT RESPONSIBILITY?

At the heart of the political conception is the responsibility thesis, which is then specified, on the basis of a theory of collective responsibility, for situations where many agents contribute to an outcome. It notes the absence of backward-looking responsibility in the case of some arbitrary inequalities, and infers from this fact the absence of forward-looking (or task-) responsibility to correct them.37 But what may motivate the responsibility thesis itself? The idea may go as follows. To say that an inequality is unjust is to imply that some agent is morally obligated to correct it (task–responsibility). But if no agent is responsible for the fact that it exists (blame–responsibility), then none can be obliged to correct it. Therefore, if no agent can be attributed responsibility for the fact that it exists, then the inequality cannot be unjust.

To see what is wrong with the responsibility thesis, one must consider different answers to the question whether arbitrary inequalities are morally problematic in themselves and whether they provide pro tanto moral reasons for action. Let us assume for the moment that they are. If so, then it does not follow that non-responsible individuals can have no task–responsibility regarding the inequalities to which they contribute. One may plausibly suggest that in cases of a ‘responsibility shortfall’ the individuals contributing to the problematic outcome have a duty to create the collective agent that could be the proper object of responsibility attribution for the inequalities. Consider the case of arbitrary inequalities generated by the global economy, and let us assume that there is no such collective agent that is responsible for them in the backward-looking sense. Let us further assume that many of these inequalities would exist even if every participant in global trade acted in individually permissible ways. In that case global trade is an instance of morally problematic inequality without (backward-looking) responsibility. If the inequalities in question are morally problematic, and there are feasible ways of organizing global trade that would predictably mitigate them, then it is reasonable to suggest that participants of global trade have reasons of justice to organize themselves in those ways by creating the necessary political institutions. In other words, absence of backward-looking responsibility for arbitrary inequalities is consistent with the existence of task responsibility to establish institutions that could mitigate such responsibilities, if the inequalities are problematic. The fact that no currently existing agent has obligations to directly address the inequalities does not rule out the existence of a duty to create a collective agent that has such obligations.

Indeed, this is exactly the kind of duty that many theorists, including Nagel, endorse in other contexts. They recognize the duty to leave the state of nature and create such legal institutions that make the enjoyment of certain pre-existing rights possible.38 They recognize that individuals may not be able to honor the duties imposed on them by the prepolitical rights of others without appropriate institutions,39 and therefore recognize a duty to create such institutions when they are absent (i.e. to leave the state of nature). But if individuals cannot avoid violating the rights of others without such institutions, then they cannot be responsible for those violations either. They stand in exactly the same moral relation to these rights violations as they do to the inequalities discussed above. However, while in the case of rights violations the political conception acknowledges the arising duty to create political institutions that enable individuals to respect the rights of others, it does not recognize a parallel duty to create institutions that would enable them not to impose arbitrary inequalities on others.40

Now, adherents of the political conception may object that there is no inconsistency here, because the relevant difference between right violations and arbitrary inequalities is exactly that rights violations are always objectionable, while arbitrary inequalities are not inherently so. To say that we might have a duty to create political institutions to mitigate arbitrary inequalities is therefore question-begging, because it assumes that such inequalities are problematic. So let us now assume that arbitrary inequalities are not in themselves morally problematic. But if so, then it is mysterious why they become objectionable—and the basis of assigning blame—if some agent is responsible for them. To recall, to say that an agent has backward-looking responsibility for a certain state of affairs is simply to say that it is appropriate to take that state of affairs as the basis of morally appraising the agent. But, as Scanlon notes, ‘[n]othing is implied about what this appraisal should be’41: blame, praise or neither. Whether it should be one or the other is to be decided independently, on the basis of the morally relevant features of the state of affairs that is the basis of moral appraisal.42 Therefore, if arbitrariness is not an independent wrong-making property of inequalities, then such inequalities cannot form the basis of assigning blame to the agent that is responsible for creating them. Thus, they cannot be unjust within the state either, where there is a collective agent that is responsible for them. On the other hand, if arbitrary inequalities are morally problematic in themselves, then those who (blamelessly) contribute to creating them seem to have at least a pro tanto forward-looking responsibility to create institutions that enable them to mitigate them if institutions are necessary for that goal. Whether they have such forward-looking responsibility depends, at least in the first instance, on whether arbitrary inequalities are problematic in themselves.43 But they cannot be made unjust by the fact that some agent is responsible for them if there is nothing problematic about them in the first place. The presence or absence of backward-looking responsibility plays no role in determining whether and when arbitrary inequalities are morally problematic. If they are not in themselves problematic (unfair, etc.), then they cannot be made wrong just because some agent is responsible for them. If they are in themselves problematic, then the absence of backward-looking responsibility does not rule out the existence of task-responsibility to correct them.44

One might suggest that the normative significance of the distinction between doing and allowing, of contributing to arbitrary inequalities and simply letting them exist explains the relevance of backward-looking responsibility for inequalities. However, the distinction can do any work only if it is allowed that arbitrary inequalities are in themselves problematic, because otherwise one cannot explain what is wrong with contributing to them. (The distinction does not have moral significance if the thing done or allowed is morally neutral.) Once that is conceded, it might be suggested that contributing to arbitrary inequalities is more objectionable than simply letting such inequalities stand. This might be taken to imply that while arbitrary inequalities are always bad, they are not always unjust but only under the circumstances specified in the political conception. But there is no textual support for reading Nagel this way. If he endorsed this view, he would have to acknowledge at least a pro tanto duty to mitigate arbitrary inequalities globally, but there is no sign that he does so. In his view ‘[s]ocio-economic justice is […] fully associative’45. For this reason, he cannot be taken to rely on the doing/allowing distinction. In any case, that distinction would support only the much weaker claim that while some global inequalities are morally problematic, we currently lack the global institutions that could address this unfairness directly. But that is not at all what the political conception claims; it holds that inequalities are not morally problematic outside the state.

I therefore conclude that the responsibility thesis cannot be sustained, and thus the political conception, to the extent that it depends on that thesis, does not succeed.46 It faces what appears to be a fatal dilemma. It either cannot explain why arbitrary inequalities are problematic among people subject to shared political institutions, or it must accept that they are problematic in other contexts as well.

‘IN OUR NAME’

As noted above, Nagel claims that justification is not owed for the effects of state policies to those (such as would-be immigrants) who simply suffer the impact of the rules but who are not morally expected to uphold them. They are simply coerced.47 Given that outsiders’ agency is not normatively engaged with these rules, responsibility for their effects does not distribute to them. And therefore they have no standing to demand justification for their effects.48 This claim holds independent interest even if my previous arguments to defeat the argument from collective responsibility all fail. Let us assume, for the sake of argument, that Nagel is right to hold that arbitrary inequalities can be unjust only if there is an agent who is responsible for creating them. It would still not follow that only those have any ground to complain against the effects of the collective agent's actions whose agency is normatively implicated in what the collective agent does. It might still be coherently objected that everyone whose lives are significantly affected by the collective agent's actions are owed justification. Nagel's reasoning makes two assumptions here. The first is that only those are owed justification for coercive rules who are morally expected to uphold them and thus share responsibility for their effects. The second is that outsiders are not morally expected to comply with border rules but are merely forced to do so. I examine them in turn.

Why is it that only those are owed justification in whose name institutions are imposed?49 As mentioned above, Nagel claims that those in whose name institutions are imposed ‘are made responsible’ for their effects, but he does not clarify whether they are made responsible in the backward-looking sense or in the task–responsibility sense. At one point he writes: ‘[w]hat is objectionable is that we should be fellow participants in a collective enterprise of coercively imposed legal and political institutions that generates […] arbitrary inequalities’.50 These lines read naturally to suggest that if we participate (through upholding political institutions) in producing arbitrary inequalities, we become blameworthy (even if we disagree with the arrangements that produce them). The thought appears to be that, because we share in the blame that accrues to the wrongful acts of the coercive institutions that are imposed in our name, whether or not we approve of them, we have special standing to demand that the state does not commit wrongful acts.51 But this reading runs into a similar problem to the one identified in the previous section. It fails to explain why participating in the production of arbitrary inequalities through upholding coercive rules that sustain them is objectionable if these inequalities are not independently problematic. If arbitrary inequalities are not objectionable, then it is not obvious why it should be objectionable to participate in producing them. The fact that we share (backward-looking) responsibility for them is significant only if what we are made responsible for is independently problematic. Therefore, this reading cannot explain the significance of the ‘normative engagement’ of our will in the scope restriction of the requirements of justice.

The problem may be restated in the following manner. This reading of the ‘in our name’ restriction seems to appeal to our moral claim as agents against being made complicit in the production of injustice. The idea is that as citizens who are morally expected to uphold coercive institutions we have a special interest against being made ‘fellow participants’ in injustice in addition to our interest as patients against suffering injustice, and this additional interest grounds the claim for special justification that is owed to us. The claim against being made complicit in injustice is certainly a legitimate one. However, it can hardly do any independent work for the political conception, because it presupposes that what we are made responsible for is unjust. That might be so either because certain inequalities are independently problematic, or because they are made so on account of the responsibility thesis. The first alternative assumes that the political conception is false, while the second alternative fails because of its dependence on the responsibility thesis that was shown to be mistaken. Additionally, the agent-claim against being made complicit in injustice strikes me as derivative of the more fundamental patient-claim against suffering injustice. Therefore, it could not make a significant contribution to the scope restriction of justice even if the responsibility thesis were true, because then that thesis would be sufficient on its own to explain the restriction.52

Alternatively, we might read ‘being made responsible’ by institutions imposed ‘in our name’ in the forward-looking sense, since it might be plausible that members of the group may be liable to share the burdens of correcting wrongs committed by the group agent.53 The thought now may be construed along the following lines: if the state as a collective agent is properly seen as acting in the name of its citizens, i.e. legitimately, and if it is responsible for some morally objectionable situation, then the costs of correcting or eliminating the problematic situation are legitimately imposed on its citizens. Therefore, citizens have special standing to demand that the collective agent avoid creating such situations. To the extent that morally arbitrary inequalities are objectionable, citizens have special standing to demand that the state does not sustain such inequalities. On this reading, the special justification regarding coercive rules that is owed to citizens is explained with reference not to their claim against being made complicit in wrongs but to their interest in not being compelled to share the burdens of correcting the wrongs committed in their name.

At this point in the discussion, the problem with this way of justifying the restriction of the presumption against arbitrary inequalities to those who are citizens of a shared state should be obvious. Citizens’ interest in not being compelled to bear the burdens of correcting wrongs committed by their state ‘in their name’ can only support their demand against inequalities if the inequalities are indeed wrong. By contrast, if they are not independently problematic, then the state has no responsibility to correct them and citizens cannot legitimately be compelled to bear the costs of correction. Therefore, whether citizens’ interest in not being compelled to bear the costs of correcting their state's wrongs justifies their special standing to demand that it refrain from sustaining inequalities depends on whether those inequalities are wrong. Once again, if they are wrong, that may be because they are independently problematic or because they are made so on account of the responsibility thesis. In the first case, the political conception is assumed to be false, whereas the second alternative is untenable because of its dependence on the responsibility thesis. In either case, the ‘in our name’ condition does little independent work.54

I conclude therefore that on neither of the two plausible readings does the ‘in our name’ condition contribute to justifying the scope restriction of the presumption against arbitrary inequalities to those subject to the authority of the state. It either assumes the independent wrongness of such inequalities, in which case the political conception is defeated for that reason, or it is dependent on the responsibility thesis and fails for that reason.

However, let us assume for the sake of argument that one of the readings succeeds, and that the special normative engagement of agency has this kind of significance regarding the scope of justice. I now turn to Nagel's claim that the agency of noncitizens—would-be immigrants—is not similarly normatively engaged, because they are simply coerced, but the coercion is not imposed in their name. As an observation about the actual attitudes displayed by current national governments towards potential immigrants, this looks mistaken. Both representatives of nation states and theorists who defend the practice of nation states to unilaterally determine border policy claim that states have the authority to do so; they do not see states as engaging in sheer coercion but as practicing a rightful exercise of authority.55 And it is not just current attitudes that suggest so: the practice of border control is a fundamental aspect of the interstate system that assigns this legal right to each state, and each state's right is recognized at least in principle by all other states. Another way of putting this is to say that individual states are participating in a single ‘unified and distributed scheme of coercion’ that is necessary to sustain the overall system of nation states.56 And the practice of distributed coercion can be justified only if the state system as a whole is morally justifiable. In other words, every person in the world is subject to the interstate system of coercive border control, and everyone is owed a justification for it; they are owed justification not necessarily for the coercive border control as practiced by each state, one by one, but for the global scheme of distributed border coercion that makes it a legal right of each state to practice border coercion and for the overall system of nation states of which it is a necessary constituent part.

If this analysis is correct, then every person in the world has moral reasons to comply with and uphold border controls insofar that the state system as a whole is justified. And then, by parity of reasoning, their agency is normative engaged in the system as a whole, and they are thus made responsible for its effects (in one of the two senses). But then, analogously, they are owed justification for these effects. The implication seems to be that each individual state's practice of coercive border control is justifiable only if the overall distributive effects of the whole interstate system are just, according to some standard of justice. Therefore, the restriction of the scope of justice is unsuccessful even if we assume, for the sake of argument, that Nagel is right about the implications of the normative engagement of agency for arbitrary inequalities.57

CONCLUSION

This paper reconstructed Nagel's political conception of justice as holding that arbitrary inequalities are unjust only if some agent is responsible for them in the attributability sense (the ‘responsibility thesis’), and that in case individuals’ otherwise permissible actions contribute to inequalities this condition is satisfied only if they constitute a collective agent. Furthermore, they constitute such an agent only if they are subject to central decision-making and enforcement. In addition, the presumption against arbitrary inequalities arises only if these institutions are imposed in their name, because then they are made responsible for their effects (in one of two senses). I argued that a plausible theory of collective agency and responsibility might be invoked in support of the conception, but that it fails even if we accept that supplementary theory. The main weakness of the account is its reliance on the responsibility thesis. Arbitrary inequalities are either independently problematic, in which case they are prima facie unjust across states as well, or they are not independently problematic, in which case some agent's being responsible for them cannot make them unjust. I further argued, however, that even if the responsibility thesis is defensible, the restriction of the scope of justice to citizens within individual nation states still fails. This is so because on Nagel's own account the agency of all persons globally is normatively engaged in sustaining the whole state system of which particular states are only constituent parts, and should therefore be regarded as being responsible (in the relevant sense) for the effects of the state system. Therefore, on his own account the state system as a whole must be justified with reference to some global standard of justice. Therefore, the political conception fails even if we accept most of its theoretical premises—but we have strong reasons to reject at least some of them.

ACKNOWLEDGEMENTS

Earlier versions of this paper were presented at the ECPR General Conference in Bordeaux, 2013, September 4–7, and at the Institute of Philosophy in Budapest, 2015, November. I benefited greatly from the audiences at these events. Special thanks are due to my discussants, Laura Valentini and Attila Mraz. The two reviewers of this journal also made very helpful criticisms.

Notes

1 Simon Caney, Justice Beyond Borders: A global political theory (Oxford: Oxford University Press, 2005); Kok-Chor Tan, Justice, Institutions, and Luck: The Site, Ground, and Scope of Equality (Oxford: Oxford University Press, 2012).

2 Different versions of this view are offered by Charles Beitz, Political Theory and International Relations (Princeton, NJ: Princeton University Press, 1979); and Andrea Sangiovanni, ‘Global Justice, Reciprocity, and the State’, Philosophy & Public Affairs 35, no. 1 (2007): 3–39. Beitz and Sangiovanni reach different substantive conclusions because in Beitz's view the relevant cooperative relations obtain globally, while in Sangiovanni's view the benefits reciprocally produced within nation states are normatively special and therefore only they give rise to egalitarian distributive requirements.

3 Michael Blake, ‘Distributive Justice, State Coercion, and Autonomy’, Philosophy & Public Affairs 30, no. 3 (2001): 257–96; and Laura Valentini, ‘Coercion and (Global) Justice’, American Political Science Review 105, no. 1 (2011): 205–20. These accounts are different to the extent that in Blake's view the morally relevant coercion is only the one currently exercised by nation states, while Valentini holds that there are other forms of coercion as well that trigger duties of distributive justice.

4 Thomas Nagel, ‘The Problem of Global Justice’, Philosophy and Public Affairs 33, no. 2 (2005): 113–47, at p. 120.

5 For a non-exhaustive list of some of the most influential criticisms, see Arash Abizadeh, ‘Cooperation, Pervasive Impact, and Coercion: on the Scope (not Site) of Distributive Justice’, Philosophy & Public Affairs 35, no. 4 (2007): 318–58; A. J. Julius, ‘Nagel's Atlas’, Philosophy & Public Affairs 34, no. 2 (2006): 176–92; Joshua Cohen and Charles Sabel, ‘Extra Rempublicam Nulla Justitia?’ Philosophy & Public Affairs 34, no. 2 (2006): 147–75; Andrea Sangiovanni, ‘Global Justice, Reciprocity, and the State’, Philosophy & Public Affairs 35, no. 1 (2007): 3–39; and Miriam Ronzoni, ‘The Global Order: A Case of Background Injustice? A Practice–Dependent Account’, Philosophy & Public Affairs 37, no. 3 (2009): 229–56.

6 Julius, ibid., and Cohen and Sabel, ibid.

7 Abizadeh, ibid.

8 The Problem of Global Justice’, p. 127. Other versions of the political conception need not be committed to this egalitarian understanding of the content of the demands of distributive justice. The political conception is first and foremost an account of the ground and scope of the requirements of justice, and as such it is in principle separable from any account of their contents. Therefore, my criticism is relevant for those adherents of it, too, who reject Nagel's view on content. (I am grateful to an anonymous reviewer for pressing me to clarify this.) Nonetheless, as most participants in the global justice debate on either side accept some form of egalitarianism, I will continue my discussion in this manner for the sake of convenience.

9 Ibid.

10 Ibid., p. 128.

11 Ibid., pp. 128–29.

12 Cf. the discussion in Nagel, ‘Justice and Nature’, Oxford Journal of Legal Studies 17, no. 2 (1997), 303–21.

13 See Dan W. Brock, A. Buchanan, N. Daniels & D. Wikler, From Chance to Choice (Cambridge: Cambridge University Press, 2000), pp. 66–75. More recently, it was suggested by Elizabeth Anderson that ‘there can be no injustice without an agent who is (or was) substantively responsible for it—someone obligated to avoid, correct, or bear the costs of the injustice’ and that acceptance of this thesis is one of the features that distinguishes her favored conception of egalitarianism from luck egalitarianism. See Anderson, ‘The Fundamental Disagreement between Luck Egalitarians and Relational Egalitarians’, Canadian Journal of Philosophy 40, sup1 (2010): 1–23, at p. 5.

14 Justice applies […] only to a form of organization that claims political legitimacy and the right to impose decisions by force, and not to a voluntary association or contract among independent parties concerned to advance their common interests’, Nagel, ‘The Problem of Global Justice’, p. 140.

15 Nagel's view is not that only inequalities created directly by political institutions are matters of justice. His view is that arbitrary inequalities generated by social processes are matters of justice too, but only as long as the processes unfold within shared political institutions. The thought appears to be that the state has an obligation not to sustain or let such inequalities stand within its dominion, as well as not to establish them in the first place. Thus, he writes, ‘even within a state, through economic competition for example, some members […] may impose serious consequences on others without any implication that the others are asked to accept or authorize the actions that have those consequences. Citizens are not expected to treat each other equally in private transactions. But the broader legal framework that makes those actions possible and that legally sustains their results is subject to collective authority and justification and therefore to principles of social justice: not act by act, but for the system as a whole’, ibid., p. 130.

16 Nagel, ‘The Problem of Global Justice’, p. 129.

17 Ibid.

18 Ibid., p. 141, where he writes that ‘I doubt that the rules of international trade rise to the level of collective action needed to trigger the demands of justice, even in diluted form’, and in the same paragraph expresses his doubt that a theory of different levels of collective responsibility can be worked out.

19 Ibid., p. 142. The reference to centralized control is at p. 127.

20 See also Anderson, ‘The Fundamental Disagreement’, p. 5 for this view.

21 Some clarification about this notion of responsibility may be useful here. To say that an agent is responsible for a state of affairs in this sense is to say that the state of affairs is an appropriate basis for morally evaluating the agent, and for making various moral consequences for the agent, such as blame, praise, reparations, bearing certain burdens etc. appropriate. Obviously, the fact that the conditions of responsibility attribution are satisfied in a particular case does not, in itself, say anything about what the appropriate moral consequences are in that case. The range of moral consequences, if any, that are made appropriate by the agent's responsibility for a state of affairs depend on what kinds of morally relevant features (if any) the state of affairs in question has. Thus, this notion of responsibility is morally neutral. See Thomas Scanlon, What We Owe to Each Other (Cambridge, Mass: Harvard University Press, 1998), pp. 248–49. This idea of responsibility is backward-looking, and is to be distinguished from forward-looking or ‘task-responsibility’ that is about the obligations that particular agents have towards others.

22 As the text of the pervious footnote makes clear, being responsible for something in the attributability sense does not by itself imply blameworthiness, since the moral appraisal that is warranted by responsibility attribution may be any one of blame, praise, or neutrality. An appraisal of blameworthiness suggests that the state of affairs that is the basis of appraisal has some morally wrong-making property. I will return to this point later.

23 It should be noted that Nagel approvingly refers to Ronald Dworkin as another representative of the political conception (ibid., pp. 120–21), and Dworkin does provide an account of corporate responsibility that is consistent with the one discussed here (see reference in footnote 26 below). This may give some indirect support for the interpretive strategy adopted here.

24 Philip Pettit, ‘Responsibility Incorporated’, Ethics 117 (2007): 171–201.

25 See also Anna Stilz, ‘Collective Responsibility and the State’, Journal of Political Philosophy 19, no. 2 (2011): 191–5. Stilz originally developed her account of collective responsibility in her book Liberal Loyalty (Princeton, NJ: Princeton University Press, 2009), especially pp. 173–208.

26 This is Stilz's term, ‘Collective Responsibility and the State’, p. 193. See also Ronald Dworkin's discussion of corporate responsibility in Law's Empire (Cambridge, Mass: Harvard University Press, 1986), pp. 168–75, and Pettit, ‘Responsibility Incorporated’, p. 194. It is important to stress that in relying on this example, it is not suggested here that the problem of collective responsibility can arise only in the context of remedial responsibility for harm. To the contrary, throughout the paper my focus is whether collective responsibility attribution can be relevant for identifying inequalities as unjust.

27 It should be noted, though, that this response to the ‘responsibility shortfall’ is not the only possible one. If we find that there is an outcome that we have independent reasons to find morally problematic, and which is produced by several agents in such a manner that a ‘responsibility shortfall’ is present, we may conclude that the relevant agents have a duty to constitute a collective agent that is capable of addressing the problem. In fact, as I discuss below, in other contexts Nagel endorses such a strategy. See also Ronzoni, ibid., and Stephanie Collins, ‘Collectives’ Duties and Collectivization Duties’, Australasian Journal of Philosophy 91, no. 2 (2013): 231–48, for similar points.

28 Pettit discusses the debunking view under this description, see op. cit., pp. 180–1.

29 Ibid., pp. 181–2. The problem was originally described under a different name in a specific legal context in L. A. Kornhauser and L. G. Sager, ‘The One and the Many: Adjudication in Collegial Courts’, California Law Review 81 (1993): 1–59. See also Christian List, ‘The Discursive Dilemma and Public Reason’, Ethics 116 (2006): 362–402.

30 List and Pettit argue that it is not only majoritarian aggregation rules that lead to inconsistent results, but any aggregation rule that derives group decisions automatically according to some simple function from the relevant judgments of its members. See Christian List and Philip Pettit, ‘Aggregating Sets of Judgments: An Impossibility Result’, Economics and Philosophy 18 (2002): 89–110.

31 This is not to say that sometimes individual members and especially those occupying important roles in the group could not be separately and individually responsible for their actions. The point is that sometimes this could prove insufficient.

32 ‘The Problem of Global Justice’, p. 128. Construing the ‘in the name’ condition as referring to claiming authority is also supported by the reference to legitimacy in the text quoted in footnote 11.

33 See Stilz, ‘Collective Responsibility and the State’, p. 204. Nagel speaks of the ‘normative engagement’ of those in whose name coercive rules are being imposed. See ‘The Problem of Global Justice’, p. 129.

34 Nagel, ‘The Problem of Global Justice’, p. 129.

35 Ibid.

36 Nagel makes it clear that those (such as would-be immigrants) against whom the rules are simply enforced but not in their name are not owed justification. He suggests this is the case because as would-be immigrants are not asked to accept and uphold these rules, their role is entirely passive, whereas those ‘in whose name’ rules are enforced are expected to actively cooperate—their will or agency is also implicated (pp. 129–30). He seems to imply that coercion without justification is morally problematic only if active engagement is expected of the coerced. I will return to this point below.

37 Stilz makes the distinction between blame-responsibility and task-responsibility, which corresponds to backward-and forward-looking responsibility. She makes it clear that even when a collective agent is blame-responsible for a wrong, it is possible for many of its members to be only task-responsible to correct or compensate for the wrong. See ‘Collective Responsibility and the State’, pp. 194–5. Scanlon's distinction between responsibility as attributability and substantive responsibility is similar; see footnote 18 above and the accompanying text.

38 ‘The Problem of Global Justice’, p. 133.

39 There are several theorists who, inspired by Kant, hold that political institutions are necessary enabling conditions, even among well-motivated and reasonable persons, of honoring the negative rights of others. See for example Japa Pallikkathayil, ‘Deriving Morality from Politics: Rethinking the Formula of Humanity’, Ethics 121, no. 1 (2010): 116–147. Thomas Pogge, World Poverty and Human Rights (Cambridge: Polity, 2002), p. 135; Arthur Ripstein, ‘Authority and Coercion’, Philosophy & Public Affairs 32, no. 1 (2004): 2–35, Anna Stilz, ‘Collective Responsibility and the State’, Journal of Political Philosophy 19, no. 2 (2011): 190–208 at pp. 199–203; and Jeremy Waldron, ‘Kant's Legal Positivism’, Harvard Law Review 109, no. 7 (1996): 1535–66.

40 ‘[T]hough the obligations of justice arise as a result of a special relation, there is no obligation to enter into that relation with those to whom we do not yet have it’, Nagel, ‘The Problem of Global Justice’, p. 121.

41 Scanlon, What We Owe to Each Other, p. 248 (emphasis added).

42 See also fn21 above.

43 See also Thomas Christiano, ‘Immigration, Political Community, and Cosmopolitanism’, San Diego Law Review 4, (2008): 944–8.

44 It may still be the case that the presence of backward-looking responsibility gives special, additional reasons for agents for correcting morally problematic inequalities for which they are responsible. But these additional reasons depend on the inequalities being independently problematic. (I thank an anonymous referee for this journal for bringing this point to my attention.) Let me also emphasize at this point that my criticism of the responsibility thesis does not impugn deontological conceptions of justice in general, because it is not committed to the idea that states of affairs (e.g. inequalities) can be unjust non-derivatively. It is only committed to the weaker claim that inequalities can be morally bad non-derivatively, and thus provide reasons for actions to appropriately positioned agents such that a failure to act on those reasons would constitute an injustice.

45 Nagel, ‘The Problem of Global Justice’, p. 127. A revised version of the political conception might hold that arbitrary inequalities are morally bad everywhere yet they are unjust only where the conditions specified by the responsibility thesis obtain. Substantively, this would cede most of the ground to the cosmopolitan, since it would imply the existence of at least pro tanto reasons to create the kind of global collective agent that could mitigate inequalities everywhere.

46 To be specific, it may be coherently held, as some libertarians do, that a state of affairs is a proper object of evaluation by the principles of justice only if there is some agent that is morally responsible for it, i.e. that responsibility is a necessary condition of some state of affairs being unjust. What has been shown to be false is the claim that an otherwise neutral state of affairs can be made unjust by the single fact that some agent is responsible for it, i.e. that responsibility is sufficient for something being wrong. This is not in itself a problem for libertarians, since they take inequalities to pose no problem domestically, either. However, it is fatal for the political conception that regards domestic inequalities as unjust because of the presence of a responsible agent. (I am grateful to reviewer of this journal for pressing me to clarify this point.)

47 Nagel does not deny that would-be immigrants are subject to border coercion. This issue has recently been debated by Arash Abizadeh, who holds that border regimes are coercive and therefore need to be justified to potential immigrants as well, and David Miller, who maintains that border protection is not coercive in the normatively relevant sense and therefore no justification is owed to potential immigrants. See Abizadeh, ‘Democratic Theory and Border Coercion: No Right to Unilaterally Control Your Borders’, Political Theory 36, (2008): 36–65; and David Miller, ‘Why Immigration Controls Are Not Coercive: A Reply to Arash Abizadeh’, Political Theory 38 (2010): 111–20. Nagel agrees with Abizadeh that border protection is coercive, but he sides with Miller in denying that justification is owed to outsiders, given that the regime is not made in their name.

48 Nagel actually oscillates between saying that they are not owed any justification and are simply coerced, and the claim that the fact that border policies respect their prepolitical basic rights is sufficient justification for them. See Nagel, ‘The Problem of Global Justice’, p. 130. But then it is not clear why the latter, lower threshold of justification does not also trigger the ‘normative engagement’ of the agency of the outsiders, with all the suggested implications.

49 While it is clear that in the background of the political conception is the general contractualist idea that that justification is owed to each person who is subject to the effects of coercive rules, Nagel links this thought specifically to the idea that citizens in whose name the rules are made are in some sense responsible for these effects. It is this thought that I explore here. (I am grateful to an anonymous reviewer of this journal for pressing me to clarify this point.)

50 Ibid., p. 128.

51 I am assuming that dissenting individuals may bear backward-looking responsibility in these cases not in their individual, private capacity but as constituent members of the collective, ‘the people’ in whose name the rules are made. I thank Attila Mraz for suggesting this possibility to me.

52 I am grateful for an anonymous reviewer for this journal for pressing me to clarify the relation between this problem and the responsibility thesis.

53 See also Stilz, ‘Collective Responsibility and the State’, pp. 198–9 n17, and p. 205. She reads Nagel's position in a way that is similar to the one offered in this paragraph.

54 It might still be correct that those who bear the burden of correcting some injustice has special standing to demand justification for the actions of the agent whose potential injustices they are morally required to correct. Thus, employees of a corporation may have special standing to demand that it does not commit harms for which they may be liable. But this special standing may pertain only to the strength of the demand for justification rather than to the content of justification. I am grateful to Attila Mraz for proposing this idea to me.

55 See e.g. Miller (Oxford: Oxford University Press, 2007), pp. 201–30; and Michael Walzer, Spheres of Justice (New York: Basic Books, 1983), pp. 31–63. Possibly, Nagel might avail himself to a conception of legitimacy that distinguishes it from authority. On some authors’ understanding, legitimacy is simply the permissibility to use coercion, without a corresponding obligation on the part of the coerced to obey, whereas authority includes such an obligation. See e.g. Allen Buchanan, ‘Political Legitimacy and Democracy’, Ethics 112, no. 4 (2002): 689–719. But Nagel nowhere makes this distinction.

56 Thomas Christiano, ‘Immigration, Political Community, and Cosmopolitanism’, p. 940. Christiano argues for a conclusion similar to mine.

57 The argument so far has assumed that Nagel is right that justification for the effects of institutions is owed only to those whose agency they normatively engage, and that he is mistaken only about whose agency is engaged in this way. But the assumption is in fact highly dubious. We do not normally think, for instance, that only those are owed compensation for harms who were somehow involved in producing them; quite the contrary, it intuitively seems more urgent morally to compensate the harms of those who in no way were responsible for them. Suppose a corporation is responsible for some significant harm, and that among those suffering the harm are some of its own employees. It does not appear to be the case that the corporation's employees ought to enjoy any kind of priority in compensation simply in virtue of the fact that they share responsibility for the harm (though it might be that their claim of justification for the acts of the corporation is stronger, see n54 above).