Making the state responsible: A proxy account of legal organizations and private agents acting for the state

Under which conditions can we make a state responsible for an action? For example, is the United States (and not only Bush and his Cabinet) responsible for declaring war against Iraq? And is there any justification to make citizens contribute collectively to the reparation or compensation of the damages produced by the action? That is, should United States citizens shoulder the burdens and pay restitution to Iraq for destroying civil infrastructure during the military campaign? After reviewing some theories, I develop a framework to answer these kinds of questions.

While the idea of state responsibility has long been under discussion within both political philosophy and international legal theory, no agreement has yet been reached regarding what it is and how it can impact on ordinary citizens. The first problem lies in the difficulty of determining what states are and whether they can perform actions. For example, whereas some prominent theorists take states to be (corporate) agents, with the capacity to form and act upon intentions (e.g., Collins, 2019; List & Pettit, 2011), some others resist extending the notion of agency to anything other than individuals (e.g., Gilpin, 1984; Miller, 2002).

Additionally, there remains a controversy as to when and to what extent it is possible to distribute collective responsibility amongst citizens for the wrongful actions of their state. For instance, whereas Lawford-Smith (2019) argues that (unlike officials) citizens are not culpable for state action, and so they cannot be punished for its bad consequences, Pasternak (2021) and Stilz (2011a) hold that they can indeed be responsible if they are intentional participants or their state satisfies a democratic authorisation principle, respectively.

Although these views have revealed important aspects of state responsibility, they are not exempt from criticism. With the purpose of introducing a more compelling alternative, I discuss in this paper some crucial challenges to these accounts and then suggest another way to move forward, viz., to analyze state agency, state action, and state responsibility in terms of proxy agency, proxy action, and proxy responsibility.

I structure the paper as follows. I begin in Section 2 by motivating the analysis of three major approaches to state responsibility: Lawford-Smith's, Pasternak's, and Stilz's. Then, in Sections 3–5, 3–5, I discuss each of them in some detail and show that, despite their best efforts, they all fail at providing an adequate account of state responsibility. In particular, I argue that by focusing only on the culpability of officials, Lawford-Smith creates a normative gap between citizens and their state; that by appealing to an “intentional participation” condition, Pasternak (falling short of her own goals) blocks out the mechanism for attributing responsibility to citizens; and that by introducing a “democratic authorisation principle,” Stilz excludes nondemocratic states from a general theory of state responsibility and credits democratic states with a prima facie right to claim legitimacy for any action they perform.

To overcome these problems, I propose in Section 6 a fresh alternative. By utilizing key resources from contemporary social ontology and collective action theory, I develop a general view that takes states to be institutional agents acting through proxies. Here, I explain the various elements involved in proxy agency, emphasizing the participatory relationship that grounds the collective responsibility of state members for their contributions to the state action. I conclude in Section 7 by summarizing the key elements of this account.

In the last decade, the research on state responsibility has grown significantly. Various accounts have emerged not only from thinking of states as corporate agents but also from seeing important connections with collective responsibility (e.g., Gilbert, 2006; Isaacs, 2006 and Pettit, 2007).

Amongst those views that have proposed interesting strategies to deal with some philosophical problems associated to state responsibility, three approaches have come to occupy a stellar position. First, Holy Lawford-Smith's (2019) culpability theory, justifies a proportional distribution of responsibility over government officials based on their level of blameworthiness. Second, Pasternak's (2021) participatory theory, suggests a liability approach to the responsibility of state members that participate intentionally in the state action. And third, Stilz's (2011a) democratic theory, which grounds the distribution of responsibility over citizens in the legitimacy of the state.

While they all make valuable contributions to the topic, none of these views are entirely satisfactory. As I show in the next three sections, they all face powerful challenges that undermine their projects and keep us in the search for an alternative. To appreciate this, I presently discuss each of them in some detail.

In her (2019), Lawford-Smith has developed an account of state responsibility that follows a particular understanding of state agency and state action. Her project begins by presenting two general categories of state, viz., the citizen-inclusive and the citizen-exclusive (2019: p. 2). The former, as its name indicates, includes citizens within the state membership. The ideal of a democratic state, for example, would count as a citizen-inclusive state, in which citizens are meant to decide what the state does by revealing through their voting what they want (2019: p. 4). On the contrary, the citizen-exclusive model considers as state members only those who are ‘implicated’ in the state action, viz., government officials (2019: p. 5).

In a second step, Lawford-Smith argues that only citizen-exclusive states can meet the requirements of genuine collective agency and so be attributed with collective responsibility for collective action (2019: p. 97).¹ Unlike individual agency, she says, collective agency requires that all individual members participate intentionally in the performance of collective action (2019: pp. 103–104); and in contrast to individual responsibility, collective responsibility tracks the culpability of group members from their participation in the collective action (2019: p. 99). So, in the case of states, only government officials who, by acting in their voluntarily roles, contribute to the collective action of government (understood as the realization of legislative, executive, and judicial activities), satisfy the conditions of collective agency and so can be attributed with collective responsibility (2019: pp. 68, 79).

As a result, then, this kind of responsibility corresponds only to the state members that are directly implicated in the state action because they are culpable for what the state does (2019: p. 96). As state members, that is, they do not only contribute causally to the state action, but they help constitute the agency of the state (idem). Thus, the responsibility at issue here is a form of blame responsibility, whose attribution to individual members is justified in terms of their culpability (2019: p. 139). Moreover, since culpability comes in degrees, distributing the corresponding responsibility follows a proportional order: The more culpable one is for an action, the more blame and punishment one is to take for its consequences.

So, in applying this model to states, we have that government officials can be charged with more-or-less responsibility according to their level of participation in, and corresponding degree of culpability for, the state action. In Lawford-Smith's words, “those who occupy roles highest in government hierarchy are more culpable for government wrongdoing, and should be distributed the greatest shares of blame and punishment” (2019: p. 172).

Yet, by focusing only on government officials, this account encounters a serious deficit. In some cases, the proportional distribution of responsibility is not enough to compensate for the damage produced by the state action. That is, it does not generate sufficient resources for repairing or mitigating the consequences of the wrongdoing. So, as Lawford-Smith admits, her view may produce “responsibility shortfalls” (2019: p. 136). For example, if a state is found guilty of an unjust war and is thereby obligated to financially compensate the state victims, it may have to increase taxes and reduce access to public services in order to secure the funds it needs to fulfill this obligation. In such cases, says Lawford-Smith, although non-culpable for the state action (2019: p. 109), citizens can be made responsible for it and then be charged with a compensatory burden to help remedy the damage (2019: p. 138).

In locating this kind of remedial responsibility within her overall project, Lawford-Smith appeals to reasons other than culpability for the inclusion of citizens in the nonproportional distribution. She considers, in particular, the responsibility that follows from (a) the citizens' associative obligations, that is, the idea that “there are sometimes associative responsibilities for citizens to bear external responsibility for what their states do” (2019: p. 127); (b) the citizens' benefit from the state injustices, that is, “the fact of benefitting may justify citizens' responsibility to disgorge the benefits of their state's injustices” (2019: p. 130); (c) the citizens' (underserved) privilege within the group, that is, “citizens who are privileged might have responsibilities in virtue of that privilege to play a part in the state's remedying its wrongdoing” (2019: p. 134); and (d) the citizens' capacity to help, that is, “that citizens can take responsibility for the harms their states do, so if no one else is going to, then they should” (2019: p. 135).

Importantly, though, since these reasons are only subsidiary, Lawford-Smith contends that they are consistent with both taking the state to be a collective (moral) agent only if ordinary citizens are excluded from the state membership and with the proportional distribution of blame-responsibility (2019: p. 172). Thus, to put it differently, only when there are responsibility shortfalls, there may be some justification for calling citizens to contribute to discharge the state obligations based on their remedial responsibilities; otherwise, only government officials, who are culpable for the state wrongdoing, can be held responsible for it.²

Unlike Lawford-Smith, Pasternak does not leave citizens out of the state. In fact, her view, as developed in her (Pasternak, 2013, 2014, and 2021), focuses on their internal attitudes; particularly, on their participatory intentions. The main difference, then, is that Pasternak adopts the perspective of the participants in state actions (2021: p. 8), while Lawford-Smith adopts the perspective of collective agency. Yet, although Pasternak's approach is more tempting than Lawford-Smith's in this respect, it remains subject to important objections. Before getting to this, let me introduce its main elements.

I start by mentioning what its goal is. In the introduction to her (2021), Pasternak says that she wants to provide an account of state responsibility that “carves the middle ground” between two positions, viz., the proportional distribution amongst officials (à la Lawford-Smith) and the nonproportional distribution (at currency in international law) that sets the burden indiscriminately on citizens altogether (2021: p. 6). In short, the account consists in grounding the citizens' obligation to share remedial responsibilities with the state (particularly, reparations and compensations) in their intentions to participate in their states (idem).

A special promise of Pasternak's account is that, although context-sensitive, it justifies the nonproportional distribution of state remedial responsibilities amongst citizens in both democratic and nondemocratic states (idem). While this is meant to be an advantage over Stilz's democratic theory of state responsibility (see Section 5 below), it is also intended to “justify the real-world practice of letting the burden fall on citizens at large” (2021: p. 9).

A key factor in providing this justification is the existence of “intentional citizens,” that is, citizens with genuine participatory intentions. As Pasternak says, “the intentional-citizenship-based justification of the distributive effect applies only when citizens' participatory intentions are genuine” (idem). Although I say more about this below, it is worth noting already that her characterization of “genuine participatory intentions” is rather negative; that is, it involves the citizens' not feeling alienated from their state or otherwise not being compelled to abandon its territory. But in any case, the justification relies “on citizens' internal attitudes” (2021: p. 10).

Having the appropriate attitudes, which Pasternak believes can be demonstrated by national surveys (2021: p. 14), lends the state support to transfer the burdens of reparations and compensations to their citizens via, for example, increasing taxes or limiting access to public services (2021: p. 17). Indeed, she argues, this follows from taking the states to be (institutional) corporate agents (2021: pp. 19, 22–23) since they can discharge their remedial responsibilities only through their citizens (2021: p. 28). The crucial question is whether they should do it according to a proportional or a non-proportional distribution scheme.

Her objection to the proportional alternative (as mentioned above in Section 3.1) is twofold. First, it leads to “responsibility shortfalls” (i.e., “the state cannot raise enough resources to address its remedial obligations to the victims”); and second, it requires a “complex and costly fact-finding process” (which may impact on its capacity to deliver on its obligations) (2021: p. 44). So, in looking for a better alternative, she opts for a nonproportional distribution model, according to which “citizens can be expected to accept a nonproportional share of the burden of their state's responsibilities if it turns out that a [proportional] distribution is infeasible or very costly” (2021: p. 46). I emphasize later the neglected role that this conditional if plays in Pasternak's account (see Section 4.1), but for now let me simply consider the notion of “participatory intention” that underlies this alleged expectation.

Building on Kutz's account of “participation in collective action,” Pasternak aims to explain the way in which citizens act together in their state (2021: p. 48). And by elaborating on Kutz's “complicity principle,” which constitutes the normative basis for making participants in collective harms accountable, she goes on to say that “when taking part in a collective act we ipso facto commit ourselves to accept a potential share of the consequences of the shared activity” (2021: p. 61). Thus, according to the participatory view, participating in a collective wrongdoing implies the acceptance of collective liability, which can be distributed non-proportionally (i.e., not according to the individual contribution to, or corresponding level of culpability for, the collective act, but based on the participatory intentions of the individuals).

Crucially, this model of distribution is not a default, but only conditional. As Pasternak says, “intentional participation in collective wrongdoing provides, under some circumstances, a membership-based justification for the nonproportional distribution of the group's remedial responsibilities” (2021: p. 62, my italics). Those circumstances, as she clarifies later (2021: p. 66), consist in the satisfaction of two conditions, viz., that members are intentional participants and that their participatory intentions are genuine. Let me briefly elaborate on each.

For citizens to count as intentional participants and so inclusive authors of state actions, they simply have to do their part in the realization of the corporate agency of the state. Though one can do this, for example, by voting, paying taxes, or engaging in public debates (2021: p. 115), one need only “contribute to the maintenance of the corporate agency of the group, knowing that it will generate policies that one will not necessarily agree with” (2021: p. 71). This is important because it shows that someone can participate in a state action regardless of whether she supports it.

The second condition, related to the genuineness of the participatory intentions, “requires that group members choose not to leave their group even though leaving the group is not unreasonably costly” (2021: p. 75). So, when applied to state members, the idea is that “citizens are genuine participants in their state if they could leave their state without incurring unreasonable costs but choose not to; or if, even if the costs are unreasonable, that fact is not what motivates them to stay. Instead, they participate in the state for their own reasons” (2021: pp. 75–76).

It is crucial to add here that Pasternak herself concedes that “most people in the world cannot be characterised as intentional citizens in the sense that they could leave their country if they wanted to” (2021: p. 96). However, she believes that the intentional citizenship model can still be useful if the condition of genuine participation only considers whether people feel “deeply alienated” or see their membership as “forced against their will” (idem). And this, which Pasternak takes to be an empirical issue, is something that we can decide by looking at “cross-national attitude surveys,” “which ask participants about their levels of attachment to their country” (idem). However, while she recognizes that “These surveys do not directly track the intentional dispositions that characterize intentional citizenship”, she is certain that “the factors that these questions aim to track have a close enough conceptual affinity with the constituent elements of intentional citizenship, and that we can therefore use these questions to gauge the prevalence of attitudes of acceptance in real-world states” (2021: p. 97).

Thus, by assuming that this conceptual affinity demonstrates that in “real-world states” citizens are intentional citizens and that national surveys reveal that people recognize their citizenship status “in terms of the roles and obligations that they have as citizens” (2021: p. 100), Pasternak concludes that the intentional citizenship approach can justify the non-proportional distribution of liability amongst citizens, and thus guarantee that the state meets its remedial obligations (2021: pp. 152–153).

By asking whether citizens are liable for what their state does in their name, Stilz (2011a) advances a theory of state responsibility that includes a principle of democratic authorization to justify the distribution of task responsibility amongst citizens. Thus, unlike Lawford-Smith and Pasternak, Stilz does not build upon an ad hoc characterization of states (as excluding citizens from their membership) nor on empirical findings about citizens' internal attitudes (e.g., participatory intentions); instead, she focuses on certain conditions of state legitimacy (2011a: p. 191). While this makes her view more resistant than the other two alternatives, it is nevertheless subject to important objections. As before, I start by presenting the main elements of the approach and then move to making some critical comments.

Like Pasternak, Stilz is concerned with citizens' liability and the distribution of task responsibility, rather than with culpability and the assignment of blame-responsibility (2011a: pp. 194–195). The difference, as she puts it, is that the former “involves assigning duties to people to repair a particular situation, even when they did not cause the outcome and cannot be blamed for it” (2011a: p.195). For example, a non-contributing group member (e.g., an employee) may not be blameworthy for what the group did but may still be assigned with certain tasks to contribute to the reparation (idem).

In this sense, then, Stilz's account of state responsibility aims to justify making citizens contribute to the reparations of their state's wrongdoing, regardless of whether they are also accountable for participating in it. Her strategy is to argue that democratic legal states (i.e., states that meet certain criteria of legitimacy) “can morally require their citizens to do their part in discharging state responsibility” (2011a: p. 191). In those states, she says, membership alone can justify holding citizens liable for what their state does (without further investigation regarding, e.g., their participatory intentions); however, in nondemocratic states, the same principle does not hold (idem). In a nondemocratic state, that is, citizens cannot be assigned with task-responsibility simply because they are citizens (as there is no guarantee that they have any reason to approve or support their state).

Building on this, Stilz proposes the democratic authorization account, which has as its goal to allow “democratic legal states to distribute responsibility to their citizens” (idem). In developing this account, she starts by considering states as “incorporated groups” (i.e., groups that possess an “internal constitution” based on which its members can deliberate about their intentions and control their acts, thus also satisfying conditions of moral personhood) (2011a: pp. 192, 195). As it happens with companies, she says, by incorporating into a group, the members of a state accept the risk that comes with enjoying certain benefits, viz., to “own-up to” the state's liability (2011a: p. 194).

However, she notices that, unlike companies, states are nonvoluntary (i.e., we do not normally choose to “join” a state). What is more, “Exit from the state is either very costly (entailing the loss of ties to family, friends, and culture for those who can leave) or nearly impossible (citizens of less-developed countries usually have nowhere else to go)” (2011a: p. 196). So, just appealing to a principle of incorporation will not be enough to secure that citizens of democratic states accept the burden of sharing in task responsibilities.

In developing an alternative, Stilz submits a democratic authorization principle, which is meant to show that “the will of the individual citizen is implicated in the acts of her state even though she did not consent to membership and even though she may disagree with the state's policies” (2011a: p. 197). Thus, to put it another way, this principle claims both that “citizens of democratic legal states have good reasons to affirm their membership […] and that this fact suffices to justify distributing the burdens of political liability to them” (2011a: p. 198).

In closing, Stilz mentions two further consequences that follow from this. First, using the democratic authorization principle to distribute liability amongst citizens “gives them an incentive to exercise their political participation rights to control the state, minimizing harms to outsiders” (2011a: p. 206). Second, if the democratic legal state abuses its power, “even dissenters are liable for repairing harms their state inflicts” (idem). The justification for this is that those dissenters also benefit from having a state that acts on a reasonable interpretation of their rights. So, in the end, inasmuch as the state satisfies the democratic authorization principle, it will count as acting in the name of the citizens, who will then be taken as accepting the risk of having a legitimate government that represents their interests.

As mentioned above, when presenting some challenges to Lawford-Smith's view (see Section 3.1), there are no good reasons to exclude citizens from the state membership; however, if this is so, how can we understand that the state acts in their name? Stilz is right here in calling attention to the notion of proxy. For instance, when a lawyer acts in representation of her client in a courtroom, she acts as a proxy; that is as if her client herself was acting. But are the citizens and the state two separate entities, like the lawyer and her client?

Though I believe that we can understand state responsibility as a special case of proxy responsibility, I do not think that Stilz's idea has captured this properly. First, a proxy is not a guardian (i.e., the proxy need not act in the best interest of the principal). Second, the state does not act as a proxy; rather, the state acts through proxies (e.g., legal organizations and private agents). Third, to understand the capacity of the state to act in the name of its citizens, we need to understand the relationship between the state and its proxies (which, as Stilz rightly observes, has nothing to do with consent but with authorization; though, not the kind of authorization that commits us to a certain moral view about legitimacy). To clarify all these elements, I submit a brief analysis of state agency, state action, and state responsibility in terms of proxy agency, proxy action, and proxy responsibility.

After introducing and rejecting three major approaches to state responsibility (viz., Lawford-Smith's, Pasternak's, and Stilz's), I submitted my own view. Although I did not elaborate at length in all its details, I presented some of its core elements. By utilizing the notion of proxy agency, I explained the capacity of the state (as an institutional group, with official and non-official members) to perform actions through the agency of authorized proxies (legal organizations and private agents). Then, I elucidated the notion of state action in terms of proxy action (i.e., an action performed in accordance with, and within the normative boundaries of, a proxy power). Finally, I claimed that a state is responsible for the actions that it performs through its proxies, and so it is justified in distributing appropriate burdens amongst its members.

While all this is not enough to fully evaluate the proposed account, I believe that there are good reasons to consider it as a serious contender. First, it clarifies state agencies without excluding citizens from their membership. Second, by explaining how principals and proxies are contributing authors of proxy actions, it demystifies how state members (officials and non-officials) participate in state actions (without appealing to subjective attitudes, e.g., participatory intentions). Third, instead of granting the state a prima facie right to claim legitimacy for any action it performs, it provides state members with a mechanism to evaluate whether the state is responsible for such-and-such putative action. And fourth, by focusing on constitutional states (which may have different forms of government), it puts forward a general principle of authorization that does not exclude non-democratic states from a general theory of state responsibility.

This work was supported by the Irish Research Council. Project ID: GOIPD/2022/590.

The author declares that there is no conflict of interest.