Understanding Responsibility for Justice

Transcription

1 Understanding Responsibility for Justice Answering the challenges raised by Martha Nussbaum Mattias Gunnemyr Draft Abstract In Responsibility for Justice (2011) Iris Marion Young develops a model of responsibility for the injustices produced by social-structural processes; a model called the social connection model of responsibility. Young argues that this model is essentially different from a liability model of responsibility, which subsumes most other contemporary theories of responsibility. The liability models typically seek to single out a blameworthy or liable agent and hence are mainly backwardlooking. Conversely, the social connection model does not single out agents in order to hold them blameworthy. Instead it involves that those responsible have a forward-looking obligation to ameliorate unjust social-structural processes. In the foreword of the very same book Martha Nussbaum questions the distinction between the two models. For instance she argues that it is difficult to maintain the retrospective/prospective portion of the distinction for the simple reason that time marches on. When the relevant time passes, the agent is guilty of not having shouldered her responsibility. In this paper I defend the social connection model against Nussbaum s objections. There are other important arguments in favour of distinguishing between the models that she has overlooked. First, The liability model seeks to identify acts or policies that in a harmful way deviate from a morally or legally accepted set of background conditions, while the social connection model brings into question precisely those background conditions. Second, the social connection model concerns harms that are impossible to trace to specific actions or policies. We can pinpoint persons who participate in these processes, but we cannot measure the contribution of each agent to harm. Therefore, we cannot justly distribute liability for harm. Yet, the social connection model can be improved. Even though participating agents cannot be blamed for the existence of structural injustice, they can be blamed for not taking steps for ameliorating these injustices.

2 Introduction The main question of Iris Marion Young s posthumously released book Responsibility for Justice (2011) is how should we as individuals think about our own responsibility in relation to social injustice? (Young, 2011, p. 15, henceforth referred to only by page number). Her starting point is the anti-sweatshop movement, whose adherents maintained that we as individual consumers have a responsibility towards exploited garment workers in developing countries. We are responsible despite the fact that we do not directly harm these workers, nor do we as individuals have any control over the situation. Young wanted to develop a theory of responsibility that captures this intuition, an aim that resulted in the social connection model for responsibility. This model involves that all agents who in their everyday activities participate in social-structural processes that have an unjust outcome share a responsibility to ameliorate these processes. However, these agents should not be blamed of wrongdoing. The responsibility they share is characteristically forward-looking; they have an obligation to engage in collective action to redress the social-structural processes in which they participate. To denote the kind of harm caused by social-structural processes, she uses the term structural injustice. Young asserts that the models of responsibility usually discussed within philosophy and the social sciences all are different versions of what she calls the liability model. This model involves that the primary purpose of assigning responsibility [...] is to identify the specific culprits or liable parties who should make restitution for the harm (p. 109). 1 When we are looking for someone to hold responsible we are trying to find out who dunnit. The liability model inherently fails to explain responsibility for structural injustice since these harms are impossible to trace to specific acts or policies. There is no salient cause for these harms apart from from the social-structural processes themselves. The social connection model has been criticised on the ground that it fails to assign blame to agents. In the foreword of Responsibility for Justice Martha Nussbaum raises three objections of this kind. First she objects that people get a free pass indefinitely, since no task they have failed to shoulder ever goes onto the debit or guilt side of their ledger, and the new task always lies ahead of them (Nussbaum, 2011, p. xxi). Second, she claims that even though the agents are causing the structural injustice unintentionally they are nonetheless culpably negligent. She argues that here as elsewhere, ignorance of the law, including the moral law, is no excuse (Nussbaum, 2011, p. xxi). Finally, she argues that: So we don t blame an agent for not shouldering the entirety of the social task all by herself, but we blame her for not shouldering the part that she ought to have shouldered, and thus we blame her for her contribution to the bad outcome. (Nussbaum, 2011, p. xxii) 1 See also Young (2011, p. 95). 2 Understanding Responsibility for Justice,

3 Mattias Gunnemyr Considering these objections, Nussbaum asserts that it is impossible to sustain the conceptual distinction between the social connection model and the liability model. In this paper I will argue that there is are important differences between the two models even though there isn t any clear distinction between them. In addition I will argue that apart from the conceptual differences Nussbaum considers, there are other more central differences that separates the two models. First, the social connection model concerns harms that are not possible to trace to particular actions or policies. Instead, multiple agents pursuing their respective interest produce these harms. The specific contribution to harm of specific actions or policies cannot be disentangled from the complex of intertwined causal chains. Hence we cannot accurately distribute the responsibility (or liability) for harm to agents. Second, the two models take different perspectives. The liability model typically evaluates more immediate interaction between agents, while the social connection model evaluate whether the socially caused positions agents are in afford them fair opportunities. In her objections, Nussbaum does however draw attention to the important idea that there is a relationship between responsibility and blame, something which Young s social connection model fails to explain. I argue that this intuition can be captured by slightly improving certain aspects of the social connection model. Young claims that we are never culpable for participating in socialstructural processes. This assertion is best interpreted as the claim that agents cannot be blamed for the existence of these social-structural processes. However, agents can be blameworthy if they fail to take steps to ameliorate these processes, at least if they have sufficient knowledge about the harms they produce. We can capture Nussbaum s intuition without giving up the social connection model by distinguishing between blaming agents for producing structural injustices, and blaming them for not engaging in redressing unjust social-structural processes. David Zoller (2015) criticises the social connection model and maintains that contributors who knowingly disregard distant harms do incur proper backward-looking guilt. I think Zoller s view is mistaken, partly since it entails that agents are liable for harms they cannot avoid, and partly since it fails to identify the proper response in relation to participating in unjust social-structural processes. Zoller s view entails that the proper response is to refrain from participating in the processes that produce it. However, just to stop participating will do little to help those who suffer from harm. To stop buying clothes manufactured by exploited workers will not improve the situation of these workers. Our responsibility towards those workers instead involve that we have to take steps to ameliorate the unjust processes that produces harm. If we do not engage in such a project, we are blameworthy. Some words about the outline in this paper. In section 1, I explain the social connection model of responsibility in detail. In particular, I consider two central differences between the social connection model and the liability model. In section 2, I answer Nussbaum s objections as well as raise some objections to Zoller s approach. Moreover, I explain the distinction between blaming agents for producing structural injustice and blaming them for not engaging in redressing unjust social-structural processes. In section 3, I will give some further examples of intuitions about responsibility and blame this distinction can explain. Understanding Responsibility for Justice,

4 1. The Social Connection Model of Responsibility In this section I will first explain Young s view on structural injustice by considering her paradigmatic of Sandy who faces homelessness. Next, I will explain Young s account of responsibility for structural injustice. Further, I will explain the liability model and compare it to the social connection model. Finally, I will examine what I take to be the most central differences between the two models; the traceability of harm to agents, and the different levels of perspective on harms. Social-structural processes and structural injustice Young follows William H. Sewell in claiming that the concept of social-structural processes is notoriously difficult to define. 2 Yet, she gives a comprehensive explanation of what these processes involve, eluding on the case of Sandy. Sandy is a single mother who needs to live in a neighbourhood where her children will be safe and have good schools to attend. She also needs to live close enough to her working place in order to facilitate commute. Unfortunately, instead of finding such a home she is forced into homelessness by a combination of factors: condominium conversion where she has been living, high rent combined with the demand of a security deposit and a sex-segregated labour market that makes low-wage service jobs the primary option for women without college degrees (pp ). To begin, social-structural processes are experienced by agents as enabling or blocking possibilities (pp. 55, 133). In the case of Sandy the combination of low income, condominium conversion, high rent in neighbourhoods from which she can commute to work, demand for security deposit etc. result in her and her small family being on the verge of homelessness. None of these factors considered in itself can be regarded as unjust, but together they severely constrains Sandy s possibilities to find a home. In some places Young makes the more general claim that: Social structure, then, refers to the accumulated outcomes of the actions of the masses of individuals enacting their own projects, often uncoordinated with many others. (Young, 2011, p. 62) Thus, social-structural processes are ultimately caused by individuals. This is also the ground for claiming that they are responsible for ameliorating the injustices. I will return to this issue later on, regarding responsibility for unjust social-structural processes. According to Young (pp ), when we consider the outcome of social-structural processes we consider positions in society. Sandy s own past actions is part of the explanation as to why she is in the problematic situation. She could have attended university or college, or stayed with the father 2 In Sewell s (1992, p. 125) words: no formal definition can succeed in fixing the term s meaning: the metaphor of structure continues its essential somewhat mysterious work in the constitutions of social scientific knowledge despite theorists definitional efforts. 4 Understanding Responsibility for Justice,

5 Mattias Gunnemyr of her children. Hence, to some extent she could herself be blamed for ending up in this situation. This fact does not influence the judgement of whether the social-structural processes are unjust. Sandy is not the only one who is vulnerable to homelessness. When thinking about structural injustices we are not only considering the case of Sandy, but all cases of vulnerability to homelessness that involve the same morally relevant facts. Whether a position is unjust or not depends on how it relates to other positions in within the same social-structural processes. Structural injustice, then, exists when social processes put large groups of persons under systematic threat of domination or deprivation of the means to develop and exercise their capacities, at the same time that these processes enable others to dominate or to have a wide range of opportunities for developing and exercising capacities available to them. 3 (Young, 2011, p. 52) Young is not the first take structures in society as the main target for evaluation of justice. Rawls A Theory of Justice (1971) is probably the most well-know theory of this kind. Young agrees with Rawls that structure is the proper target for evaluation of justice. However, she argues that there are room for development. First, she follows Cohen (1997) and objects that Rawls theory of justice cannot explain why individuals should be motivated to care about justice. The pursuit of justice does not only involve considering institutions from a theoretical perspective, it has to involve a personal ethos of justice that motivates agents to pursue justice. Second, also drawing on Cohen s critique of Rawls, Young thinks that Rawls concept of basic structure is ambiguous. In some passages basic structure refer exclusively to the coercive apparatus of society, but in other passages Rawls includes for example relations and families. Young develops Cohen s argument and claims that Rawls is unavoidably ambiguous since he is looking for some subset of societal institutions. When examining different injustices, different parts of society become visible. 4 Third, Young objects that Rawls conception of justice involves that the scope of those who have obligations of justice to one another is a single relatively closed society governed by a single constitution (p. 135). 5 Pogge (2002) objects to Rawls theory on similar grounds, claiming that there already are harmful international institutions that ought to be evaluated in the light of justice. 3 In her earlier work Justice and the Politics of Difference Young (1990, p. 41) suggested an definition of socialstructural processes. The definition is giving in the context of explaining structural oppression. She writes: In this extended structural sense oppression refers to the vast and deep injustices some groups suffer as a consequence of often unconscious assumptions and reactions of well-meaning people in ordinary interactions, media and cultural stereotypes, and structural features of bureaucratic hierarchies and market mechanisms in short, the normal processes of everyday life. Here she refers to groups rather than to persons in a certain position. 4 Risse (2011) argues that Rawls idea of basic structures can capture the examples of injustice Young is considering. This seems partly right if you like Risse (and unlike Cohen) interprets Rawls as claiming that all social institutions are included in the basic structure of society. Yet, Rawls explicitly and repeatedly claims that the boarders of society are relevant for the scope of justice. Thus, it seems to be some differences between the concept of social-structural processes and the concept of basic structure. 5 Rawls also elaborates this claim in The law of peoples; with The idea of public reason revisited (1999). Understanding Responsibility for Justice,

6 Using social-structural processes instead of basic structure as the target for evaluation of injustice has advantages in all these areas. First, it can explain why individual agents have responsibility for structural injustice. I will explain this relation further while considering responsibility for structural injustice. Second, it takes all social relations as possible targets for responsibility, avoiding the delimiting problem of choosing one subset of society as the proper target for evaluation. Third, it explains responsibility for global injustices since national boarders do not delimit social-structural processes. Responsibility for structural injustice Young claims that all who participate in, or contribute to, or rely on social-structural processes are responsible for the outcome of these processes. The social connection model of responsibility says that individuals bear responsibility for structural injustice because they contribute by their actions to the processes that produce unjust outcomes. Our responsibility derives from belonging together with others in a system of interdependent processes of cooperation and competition through which we seek benefits and aim to realize projects. (Young, 2011, pp ) The relation between agency and structural injustice Young has in mind is not entirely clear-cut. McKeown (2015) argues that Young makes use of two different relations. First, you are responsible for structural injustice if you are a part of the processes that causes it. Second, you have a responsibility for others (and the structural injustices they may suffer) if you depend on them in your everyday life. 6 These two relations can be illustrated by Young s example concerning responsibility for the structural injustice involved in the global garment industry. When I for instance by a shirt manufactured by workers working under slave-like conditions, I am responsible for those workers. Yet, the responsibility cannot be understood correctly only considering what my act of buying the shirt will cause in the future. The causal relation is reversed. The manufacturing of the shirt is a part of the processes that make it possible for me to by them. In addition, since the money I pay is a causal part of the processes that most probably will result in workers being similarly exploited in the future, I contribute to reproducing this harm. Thus. by buying the shirt I am connected both to the workers already exploited and to the workers that are likely to be exploited in the future. My act becomes one small part in the unjust social-structural processes. McKeown (2015, p. 20) concludes that by virtue of acting within unjust structures we cannot help but reproduce them or avoid dependency on the oppression of others. This idea is what I want to capture by referring to participation in or contribution to social-structural processes. Yet, even if you are responsible in the social connection sense, you are not responsible in the sense that you are liable. As Young puts it: The point is not to compensate for the past, but for all who 6 This account of responsibility is inspired by Onora Oneill s (1996) theory of the scope of ethical consideration. 6 Understanding Responsibility for Justice,

7 Mattias Gunnemyr contribute to processes producing unjust outcomes to work to transform those processes (p. 109). This entails that the social connection model of responsibility is more forward- than backward-looking. Mere participation in social-structural processes tells us very little about the nature and degree of responsibility. Instead participation can be considered as a threshold condition for responsibility; if you participate you have responsibilities. No participation without implication. Even though we cannot assess blame for structural injustices, there are parameters for reasoning about who is responsible for ameliorating the unjust social-structural processes. Young does not claim to give a complete account of these parameters but suggests that the agent s power, privilege, interest and collective ability probably are important parameters. The more power you have to ameliorate unjust structures, the more responsible you are for doing this. Similarly, the more relatively privileged you are as a result of the same structures that produce victims of injustice, the more responsible you are remedying these structures. Privilege does often coincide with power, but not always. For instance middle-class clothing consumers stand in a privileged position within the structures of the global apparel industry. Yet, individually they have very little power to actually change these structures. Their special responsibility to redress injustice in part stems from the fact that they have the possibility to change their habits or make extra efforts without suffering serious deprivation (pp ). The victims of structural injustice have strong reasons to ameliorate these injustices, not because they are blameworthy for their predicament, but because they have an interest in changing these structures. They cannot wait for some knight in a shining armour to come and rescue them: It is they who know the most about the harms they suffer, and thus it is up to them, though not them alone, to broadcast their situation and call it injustice (p. 146). 7 Finally, agents are more responsible in the social connection sense if they have a collective ability. That is what they have if they can draw on the resources of already organized entities and use them in new ways for trying to promote change (p. 147). For example unions, church groups, stockholder organisations and NGOs can exercise significant power. The liability model of responsibility Young contrasts the social connection model of responsibility against the liability model of responsibility. The liability model is not a model, but encompass several different models that share one basic property: that the the primary purpose of assigning responsibility [...] is to 7 Lukes (Hayward & Lukes, 2008, p. 12) also argues (independently of Young, as it seems) that those who are subject to structural constraints have means to address structural harms. As a collective they have power to challenge the structures. Thus, also the responsibility those who have most interest in changing the structures have stems from the fact that they have some kind of power to ameliorate these structures. McKeown (2015) makes a similar point, saying that some of those who participate in social-structural processes that produces harms and have an interest in ameliorating these harms can be excused from not addressing these harms. For instance it seems unplausible to blame children who are subject of sexual exploitation for not working for ameliorating the unjust social-structural processes that facilitated their exploitation. Understanding Responsibility for Justice,

8 identify the specific culprits or liable parties who should make restitution for the harm (p. 109). In a standard case of liability an agent has intentionally caused some harm and is therefore liable for this harm. If you intentionally break a vase at your neighbour s party, you are liable for breaking the vase. The liability models share certain features. First, they explain harms that are caused by individual actions or policies, often in terms of causation and intentions. Second, it concerns direct interaction between agents. Third, the harm for which we seek to identify a culprit is a particular event in the past. An event that has ended. A robbery that has taken place, or an oil tanker has spilled its contents on the beach. In this sense the liability model is more backward- than forward looking. Forth, a moral harm or a crime is generally understood as something that stands out from or break the ongoing normal flow of everyday morally accepted events. This assumes a set of background conditions that we consider morally or legally acceptable. Fifth, the purpose of the liability model is to identify agents to sanction, punish, or exact compensation or redress for harm. Finally, the liable wrongdoer is commonly described as guilty, blameworthy or at fault (pp ). 8 The social connection model differs from the liability model on all these points. The Liability Model Explains harms that are caused by individual actions or policies Concerns direct interactions between agents Mostly backward-looking (concerns singular events in the past) Evaluates acts or policies against morally or legally accepted background conditions The purpose is to find agents to be punished, sanctioned or made to compensate for harm Liable agents are commonly described as guilty, blameworthy or at fault. The Social Connection Model Explains harms that are caused by a complex of actions and policies, where the contribution of each action or policy is impossible to accurately measure Concerns social-structural processes Mostly forward-looking (concerns ongoing processes) Evaluates background conditions themselves The purpose is to find agents that have to engage in ameliorating unjust social-structural processes Agents who fails to shoulder responsibility are described as criticisable, but not blameworthy, guilty or at fault Nussbaum s arguments against the social connection model concerns the conceptual differences that the liability model is backward-looking and seeks to attribute blame to agents, while the social connection model is forward-looking and involves that blame cannot justifiably be attributed to agents. In the last section of this paper I will discuss Nussbaum s arguments, but 8 I follow Young in using the notion liability. Perhaps accountable would be a more appropriate notion since this notion also includes acts and policies for which an agent is praiseworthy. It seems unnecessary to delimit the discussion to responsibility for harms. To use accountable instead of liable would bring the question of whether particular agents can be praised for unforeseen and unintended benefits brought about by social-structural processes. 8 Understanding Responsibility for Justice,

9 Mattias Gunnemyr first I want to explain what I take to be the most central differences between the two models. First I will consider that the liability model concern direct interactions between agents while the social connection concerns social-structural processes, and second I will consider that according to the liability model harms are traceable to agents, while the social connection model concerns harms that are impossible to trace to agents. Two different perspectives The social connection model is not meant to replace the liability model, but to complement it. The concept of responsibility includes both responsibility in the social connection sense and responsibility in the liability sense; while liability obviously is restricted to the liability model of responsibility. Young claims that: It is important to evaluate both whether people treat one another with respect in their direct interactions with one another, and whether the socially caused positions they are in afford them fair opportunities, especially when compared with the positions of others. (Young, 2011, p. 72) We could say that the two models concern different objects. The liability model describes responsibilities between agents at a micro level, whereas the social connection model describes responsibilities for the outcome of social-structural processes at a macro level. This discussion is closely related to the issue of choosing the right target for evaluation of justice. Even though Young agrees with Cohen (1997) that Rawls theory of justice cannot explain individual motivation to care about justice, she agrees with Rawls that structure is the proper target for evaluation of justice. In fact, she criticises Cohen for taking a too individualistic perspective on responsibility for justice. 9 Without a macro perspective on social-structural processes, Young argue, social-structural processes become invisible. Young evokes another argument for preserving a dual perspective on responsibility. She considers Murphy s (1998) claim that each rich First Worlder ought to promote justice directly rather than engage in the Quixotic task of promoting just international institutions (Murphy, 1998, p. 281). Put differently, the primary duty is to promote just outcomes, not just institutions. Young argues that it is hard too see what it means for individuals too promote justice directly as opposed to through institutions. For instance Young argues that Murphy cannot avoid referring to institutional means of promoting justice, namely, giving money to humanitarian aid organizations (p. 69). Yet, it is not as simple as saying that structural injustices are found at a macro level and other harms are found in the immediate interaction between individuals. The liability model is applicable on many cases of harm at a macro-level. For instance, Young argues that the wrong Sandy is suffering is different from that many African-Americans and Jews suffered in the United States because laws forbade them from owning or renting in certain districts. There is no collective agent who has enforced laws or policies with the intent to put Sandy in the situation she is in. Similarly, we can compare the 9 Jessica Payson (2012) elaborates on a similar argument. Understanding Responsibility for Justice,

10 example of Sandy with the homelessness caused by president Robert Mugabe s decision to destroy the shantytowns of Harare, Zimbabwe, which left at least 200,000 people homeless. In this case the harm can be explained with reference to a policy enacted by president Mugabe and the harmful actions that enforced this policy. In both these cases the liability model is applicable since the salient cause of harm is the policies or acts of agents. Yet, both are cases of harm on a societal scale. We have to find some further qualification for distinguishing between the two models. We can consider another way to distinguish between the two models. Young claims that the liability model seeks to identify acts or policies that in a harmful way deviates from a morally or legally accepted set of background conditions, while the social connection model brings into question precisely those background conditions. However, there is no straightforward distinction between background conditions and what is counted as a salient cause of harm. Consider for instance Hart and Honoré s (1985, pp ) example that an Indian peasant might identify drought as the salient cause for the great famine in India , but the World Food Authority might instead identify the Indian government s failure to build up food reserves as the salient cause, and the drought is just a background condition. Since what counts as a background condition or a cause depends on context, this distinction cannot be used to clearly distinguish between the two models of responsibility. Yet, Young s argument can be improved. Young is slightly wrong when she claims that the important difference between the two models is whether we look at causes or background conditions. The important difference is rather if we seek to find a subset of causes for harm that stands out as salient or if we consider all causes indiscriminately. To say that the social connection model involves looking at all causes all social-structural processes without singling out some causes as salient fits well with the idea that no one can be blamed for causing harm. Blaming presumes that some causes stands out as salient. This idea also fits well the idea that the social connection model concerns ongoing injustices. We evaluate ongoing social-structural processes, and if they turn out to be unjust, we have the responsibility to ameliorate them. We could specify Young s point as claiming that all agents whose acts are part of the complex of causes for harm are responsible for ameliorating this harm. Moreover, they are responsible for doing this regardless of if there is an act or policy that stands out as the salient cause or not. So, the two models concern two different perspectives on causal relations. The question is then why we should adopt the structural perspective. In short Young s idea is that without such perspective, we have no tool to assign responsibility for ameliorating harms for which no one can be held liable. Hayward takes a similar position to Young, and argues that it is important to identify significant, inegalitarian, and remediable social constraints on human freedom (Hayward & Lukes, 2008, p. 10), constraints that are not always caused by liable agents. The origin of structural injustice is impossible to trace to particular agents The second central difference between the two models I want to focus on is that the liability model seeks to trace harms to particular agents, while the social connection model is most 10 Understanding Responsibility for Justice,

11 Mattias Gunnemyr relevant in cases where harms are impossible to trace to particular agents or policies. Harms produced by social-structural processes have this property. Actually, it is not entirely right to say as Young does that harms are not possible trace to particular acts or policies. They are possible to trace, at least in theory. When I buy a shirt made by exploited workers, it is obvious that there is a connection between act and harm. In fact, that it is possible to see the connection between my action and harm is essential to Young s concept that we are responsible for the unjust socialstructural processes we participate in. Young is better interpreted as saying that harms are possible (and often easy) to trace to a set of acts and policies. Yet, it is not possible to decide the size of contribution to harm of each act or policy. Therefore it is impossible to distribute liability and blame to specific agents. To illustrate this, we can turn back to the example of Sandy. Sandy is the victim of injustice, yet the other involved agents cannot be held liable for her situation. They are decent, respectful and they act within the law and according to accepted norms and rules. Some of them even go beyond what can be morally expected of them in order to help her. For instance one agent at an apartment finding service listens to her situation and preferences and goes out of his way to find Sandy an apartment. We can also imagine that the landlord who sold the building from which Sandy is displaced sold the building in order to not increase the rents in other buildings. Even though the landlord plays an important role in causing Sandy s eviction, the landlord cannot be held liable since he or she did what was best for the entire group of tenants (p. 46). Sandy is in this situation because of a convergence of many different factors, none of which constitute injustice considered in itself. There is no liable agent that has caused her situation. The liability model of responsibility depends on identifying individual or collective agents whose actions can be causally linked to a harm and who do not have valid excuses for those actions. Structural injustice, however, is the outcome, often unintended, of a multitude of routine and deliberate actions within institutions. (Young, 2011, pp ) 10 Still, in most real cases at least some part of harm is traceable to an agent. To the extent that there is an event (or agent) that stands out from the complex of intertwined causal chains as the salient cause of harm, the liability model is possible to apply. Barry and Ferracioli (2013) question the argument that structural injustice is impossible to trace to acts or policies. They argue that the injustice that Sandy and many others like her suffer are primarily the result of bad choices on the part of political leaders and those who support them. They argue that these leaders for instance could have invested more in social housing or passed legislation that precludes investors from keeping investment properties unoccupied. However, these arguments involve counterfactual dependence for liability. On Barry and Ferracioli s argument the political leaders are liable, not for causing injustice, but for not addressing these 10 See also Young (2011, pp. 44, 96-97, 100, 106, 109, 175, 185). Understanding Responsibility for Justice,

12 injustices while they had the opportunity. Yet, it would require an extraordinary situation to demand that politicians and their adherents personally compensate for injustices they have failed to address. Instead, it seems plausible that responsibility for redressing the harms maps onto individuals in some other way. 2. Defending the social connection model In her foreword to Responsibility to Justice, Nussbaum locates and puts together Young s different arguments for the social connection model. According to Nussbaum, Young uses two conceptual and five pragmatic arguments. Concerning the conceptual arguments Nussbaum writes: The first conceptual argument says that guilt requires blameworthiness, but agents often participate in structural injustice without blameworthiness. The second conceptual argument says that guilt is backward-looking, whereas responsibility is forward-looking (Nussbaum, 2011, p. xx) A bit later in the same foreword Nussbaum also adds a third conceptual point: that guilt is individualistic and responsibility is shared. Even though she initially was thoroughly convinced by Young s account (Nussbaum, 2011, p. xx) Nussbaum ends up questioning all these conceptual arguments. I will address Nussbaum s arguments one by one. First though I would like to say something about the general structure of Nussbaum s argument. All arguments are variations of the same general objection; that it is impossible to maintain Young s conceptual distinction between the social connection model and the liability model. There is no clear line between responsibility for structural injustice and the responsibility an agent has for what s/he has done. I will consider each of Nussbaum s objections in detail, arguing that there are important differences even though there is no sharp line between the models. Yet, the main argument in favour of distinguishing between the models lies elsewhere. Nussbaum has not considered the two most central differences between the models. The first central difference is that the liability model seeks to identify acts or policies that in a harmful way deviate from a morally or legally accepted set of background conditions, while the social connection model brings into question precisely those background conditions. The second central difference is that the social connection model concerns harms that are impossible to trace to specific actions or policies. We can pinpoint persons who participate in these processes, but we cannot measure the contribution of each agent to harm. Therefore, we cannot justly distribute liability for harm. However, as I will explain, Nussbaum s objection rests on an important intuition about responsibility; namely that there is some connection between responsibility and blame. The retrospective/prospective distinction Nussbaum s first objection concerns what she calls Young s second conceptual argument that the liability model is backward-looking while the social connection model is forward-looking. 12 Understanding Responsibility for Justice,

13 Mattias Gunnemyr Nussbaum argues that is very difficult to maintain the retrospective/prospective portion of the distinction, guilt being appropriate to past acts only, and responsibility to future acts only, for the simple reason that time marches on (Nussbaum, 2011, p. xxi). However, Young does not claim that responsibility in the social connection sense is exclusively forward-looking and that responsibility in the liability sense is exclusively backward-looking. Instead Young claims that both the liability and social connection models refer both to the past and the future. They differ, however, in temporal emphasis and priority (p. 108). The liability model is mainly backward-looking. Its primary purpose is to single out a liable culprit for some harmful act that has happened in the past, but the sanctions that follow can also be futureoriented insofar as they intend to deter the acts they sanction (p. 108). Likewise, the social connection model is primarily forward-looking. This is partly because the unjust structures are still ongoing and needs to be ameliorated, and partly because it is impossible to trace the contribution of each individual to the unjust structures. Nussbaum is right that there is no clear line between the models in this respect, but there is still an important difference in temporal focus. While Nussbaum s objection fails to show that there is no essential difference between the two models, her objection still conveys an important point. She also argue that if A has responsibility R for social ill S, and she fails to take it up, then, when the relevant time passes, she is guilty of not having shouldered her responsibility (Nussbaum, 2011, p. xxi). This argument captures something very important, it points at a distinction between blaming an agent for the existence of structural injustice, and blaming an agent for not doing anything about the injustice. Yet, the argument has actually little to do with time. An agent can never be blamed for the existence of structural injustice, no matter how much time passes. If the agent was not liable for the social ill S at time t, the agent is not more liable for S at t+1. For instance, suppose that I am some third party causally connected to Sandy s vulnerable situation. I am not directly connected to Sandy as the landlord or the apartment-hunting agent are. Instead we can suppose that I am a part of the unjust social-structural process because I have contributed to inflating housing prices in neighbourhoods with desirable schools and since I thought it to be unfair to my children to choose housing in a neighbourhood whose schools are overcrowded or poorly equipped and hence have bid on houses in them (p. 108). In this case it would be unreasonable to say that I should be considered liable, blameworthy or guilty for Sandy s situation. It is impossible to trace the harm to a specific action or policy of mine, I had no intention to cause harm to Sandy, and in addition we can assume that I didn t know of these unfortunate harmful unintended consequences. According to Young, responsibility in the social connection sense does not make us guilty or blameworthy or directly liable for paying compensation to victims of harm (p. 104). However, since I participate in the social-structural processes that make Sandy (and many others) vulnerable to homelessness, I have a responsibility to ameliorate these injustices. Understanding Responsibility for Justice,

14 Agents can participate in structural injustice without doing anything liable Nussbaum s second objection is that agents are culpably negligent even though they have no intention to cause or participate in unjust social-structural processes. Nussbaum qualifies her objection further stating that if it is a general moral truth that citizens ought to monitor the institutions in which they live and be vigilant lest structural injustice occur within them, then I think it follows that they are culpably negligent if they do not shoulder that burden (Nussbaum, 2011, p. xxi). 11 This objection is in one way very similar to the first objection. In the end both these objection boils down to this what is the appropriate response to responsibility in the social connection sense not taken up? Young answer that we should enlighten the agent of the circumstances and then criticize the agent for not taking responsibility. More specifically the agent should be criticized for not taking action, not taking enough action, taking ineffective action, or taking action that is counterproductive (p. 144). 12 In this vocabulary blaming someone and criticizing someone are different things. We blame agents for having caused harm in accordance to a liability model, and we criticize agents for not shouldering their forward-looking responsibility in the social connection sense. A correspondent example when it comes to individual moral responsibility would be a case where you can t be blamed for some harm, but where you can be criticized for not addressing this harm since you are embedded in the causes of it. Imagine that you accidentally fall down some stairs. In the fall you unintentionally hit a stranger quite badly. In such a case you are not blameworthy for the harm induced to the stranger due to your fall. Yet, you would be criticisable if you just walked away without addressing the stranger in some way, perhaps by asking if s/he got hurt or if you can help in some way. In one way this example is misleading. There is a clear line between the agent and the harm, but nonetheless it serves to underline the point that you might be responsible for addressing some harm even if you can t be blamed for it. That is, you cannot be blamed for bringing the harm into existence, but your causal embedment in the situation enables you to be criticised for not addressing the harm in some relevant way. Guilt is individualistic and social responsibility is shared Nussbaum s last objection rejects the conceptual point that guilt is individualistic, while responsibility is often shared (Nussbaum, 2011, p. xxii). This objection is misguided. First, Young states that guilt is not necessarily individualistic. It is true that guilt is necessarily related to an agent but this agent might be a collective one. Second, shared responsibility is different to collective responsibility. While collective responsibility maps to a collective (and might in a 11 Young does not claim that citizens ought to monitor the institutions in which they live simply because they are citizens in that nation. In fact Young does quite fiercely and successfully reject such a view (Young, 2011, pp ). However she is committed to the similar view that the agents has an obligation to monitor social-structural process since he or she unintentionally contribute to these processes. 12 This is a central theme in Responsibility for Justice See also (Young, 2011, pp. 84, 87, , 144, 153, 165) 14 Understanding Responsibility for Justice,

15 Mattias Gunnemyr second stage be distributed to its individual members), shared responsibility maps directly to agents. Young does not rely on the conceptual point that Nussbaum ascribes to her, at least not when it comes to explain why we need a social connection model of responsibility. Nevertheless, a fair interpretation of Nussbaum s worries would expose a more pressing objection. She continues by claiming that Young s various criteria (power, privilege, interest, collective ability) give us a good way of thinking about the size of a given agent s share (Nussbaum, 2011, p. xxii). We can decide an agent s given share of responsibility. Hence we could blame her for not shouldering the part that she ought to have shouldered, and thus we blame her for her contribution to the bad outcome (Nussbaum, 2011, p. xxii). Nussbaum is partly right, partly wrong. We can blame the agent (or rather criticize her) for not shouldering her part, but we cannot blame her for her contribution to structural injustice. Young s various criteria do not tell us anything about the size of the agent s contribution to harm. Consider again Barry and Ferracioli (2013) argument that politicians and their adherents are counterfactually responsible for structural injustices. This is exactly the same argument as Nussbaum proposes: They are responsible for not shouldering their part. They had both power and collective ability to do something about the social injustices, but failed to do so. Therefore they are criticisable. This raises the question of what to be criticisable means. Criticisable agents may be blameworthy In Moral Responsibility for Distant Collective Harms (2015) David Zoller argues that agents who has sufficient reason to believe that some specific action is likely to contribute to distant harm are blameworthy if they perform this action. He considers what he calls complicit reformers, which are people who participate in the harmful collective practices they are actively protesting. For instance he considers this example: S regularly advocates against sweatshop labor in her teaching and among her family, and is reasonably well informed about which retailers source most unscrupulously from sweatshops. But today she needs a new pair of pants, and she purchases them from a cheerful clerk at one of the very retailers whom she knows to be particularly unscrupulous. (Zoller, 2015, p. 995) Zoller agrees with Young that such agents are not blameworthy for contributing to distant harm (or their share of it). Still he thinks that complicit reformers such as S are blameworthy. In order to explain this blameworthiness he argues that we need to shift the object of blame. The object of blame is not the agent s share of distant deprivations, but the way a knowledgeable agent consciously disregard the victims moral status (Zoller, 2015, p. 1009). Agents such as S are blameworthy since they very directly perpetuate the moral invisibility and the lack of recognition from which the global poor generally suffer (Zoller, 2015, p. 995). Therefore, S is blameworthy even though her knowledge of the exact causal chain between her purchase and distant harm is vague. In addition she is blameworthy even though she does not intentionally cause this harm. Understanding Responsibility for Justice,

16 This, Zoller argues, gives agents reasons to refrain from performing actions that they know are likely to contribute to distant harm. There is merit to Zoller s arguments. In order to assign blame to agents we have to apply another object of responsibility than existing structural harms. In addition, epistemic conditions are relevant when we assign blame. Yet, Zoller s solution is objectionable for several reasons. First, refraining is often not the correct response to responsibility for distant harms. The victims of distant harms are possibly worse off if the complicit reformers choose to avoid blame by refraining to participate in the social-structural processes. As Young points out, sweatshop workers do not want consumers to refrain form buying their clothes since this would entail that the sweatshop workers would loose their jobs (p. 134). Since the agent s contributions to distant harm is negligible or non-existent, refraining from participating in unjust social-structural processes does little to change the situation of the victims. This problem arises since Zoller uses a property of the agent as object for responsibility. Thus, the agent can distance herself from blame without addressing the harm the victims suffers. Second, Zoller s solution fails to explain our reasons to change unjust social-structural processes. Zoller s suggestion is that we shift object of responsibility from the deprivation that the social-structural processes impose on distant anonymous persons to the agent s maxim. That is, we are to abandon Young s original object of responsibility in favour of the object Zoller suggests. If we accept this solution we can explain why complicit reformers are blameworthy, but we cannot explain our forward-looking responsibilities to ameliorate social-structural processes. To use an idiom, we would throw out the baby with the bathwater. This point can be more elaborated. To participate in the socialstructural processes that results in an unjust outcome for exploited textile workers involves more than buying or not buying clothes. For instance, the companies that sell clothes also pay taxes and VAT. The employees of the company pay income tax. These taxes are usually used to finance welfare services that we probably benefit from. They may for instance finance education, health care, pension, paid parental leave, or subsidise mass transit. To participate in these welfare services would also, according to Young, be a basis for responsibility in the social connection sense. However, Zoller s approach fails to account for this responsibility. Even worse, on some interpretation it would require us to completely refrain from participating in these services. My suggestion is that there are two separate objects of responsibility at stake. We have to distinguish between blaming an agent for the existence of structural injustice, and blaming an agent for not doing anything about the injustice. We cannot be blamed for the existence of structural injustices; not for producing them nor for reproducing them. However, if we participate in unjust social-structural processes this gives us an additional responsibility to ameliorate those processes. If we fail to do so, we may be criticised for not taking action, not taking enough action, taking ineffective action, or taking action that is counterproductive. This is not to say that we are blameworthy. Young does not think that blame is properly used in this context. This is however mistaken. As Zoller suggest, under certain epistemic circumstances we may be blameworthy. If we are sufficiently aware of our responsibility and still fail to shoulder it, we are blameworthy. The complicit reformer is blameworthy when buying clothes she has 16 Understanding Responsibility for Justice,