Abstract
Authors like Iris Young and Philip Pettit have come up with proposals for theorizing ‘structural injustice’ and social relations marred by ‘domination’. These authors provide conceptual tools for focusing on concrete economic structures and re-focus the debate about justice onto questions of power. In this article we build on their work, but we argue that a positive notion of economic agency is needed as a criterion for what makes economic structures dominating and potentially unjust. We propose a notion of economic agency, which we relate to a more general notion of autonomous agency to create a dual-level account. Domination in the economic sphere happens where individuals are deprived of their economic agency, to the extent that such agency is necessary to lead an autonomous life in general. Using the example of creditor-debtor relations we argue for the usefulness of this theoretical framework. Finally, we define injustices in terms of threshold levels of economic agency, the latter being spelled out in terms of a set of capabilities. This proposal overcomes the artificial boundaries between those theorizing ‘justice’ and those theorizing ‘power’ when analyzing the economic domain.
Introduction
When theorizing about justice, power and domination in today’s world, it is a natural move to consider not only political, but also economic structures. While many mainstream economic theories describe the economic realm as based on voluntary contracts, in contrast to the political realm in which coercive power can be used, other economic and political theories take seriously the possibility that economic circumstances can lead to unjust forms of power and domination. This especially holds for two strands of normative theorizing that have recently received much attention: Iris Young’s theory of structural injustice, and neo-republican theories of (non-)domination as proposed by Philip Pettit and others.
While taking our inspiration from them, we think that the debate needs to be moved further by delivering a substantive normative account for diagnosing economic structures as just or unjust (see the second section). Hence, we argue for a positive notion of economic agency: to prevent agents from being dominated by economic structures, we need to protect their economic agency. To understand which instances of a lack of economic agency constitute unjust domination, we first need to connect economic agency to the general ability for autonomous agency. Hence, an ideal of positive freedom is indispensable, or so we will argue, to make sense of the idea of structural injustice (third section). Once this is in place, we connect it to a notion of economic agency, as the ability to acquire resources for general agency (fourth section). The merits of this theoretical framework are shown by applying it to the case of creditor-debtor relations (fifth section).
To give flesh and bones to this perspective, we argue for understanding economic agency as requiring a set of capabilities, i.e. personal characteristics and social options that empower agents to fulfil certain economic roles. What these are depends on the economic context: renters need specific capabilities to stand up against landlords, low-income households need specific capabilities not to be dominated by the structures of debt markets, etc. (sixth section). Finally, we respond to two objections, namely whether a focus on structural justice, spelled out in terms of economic agency, is necessary and sufficient for an analysis from the perspective of justice (seventh section). A final section concludes, by mentioning two important additions to our approach (eighth section).
By arguing for a specific conception of economic agency, providing a set of criteria for picking out cases in which it is violated, our account moves beyond the use of intuitive examples, such as the domination of industrial workers by capitalists in the late 19th century, or of serfs by their lords in feudal societies. This is important because not all forms of economic inequality and dependence are problematic. For example, if A wishes to possess a special breed of roses for her rose garden and B is the only producer of its seeds, is this dependence unjust? Not all preferences that individuals happen to have, and the forms of dependence that result from them, make for unjust economic structures. Hence, a more substantive account is needed; as we suggest below, it should be cashed out in terms of capabilities. A number of other questions, for example the role of individual responsibility for economic deprivation, also need to be answered in order to identify unjust economic structures. Our account explains where they come into the picture; it thus helps structure the discussion and allows for more nuanced approaches to economic domination.
Structure and non-domination: Towards a theory of economic agency
In this section, we briefly review some of the contemporary literature that has prepared the ground for thinking about economic structures in terms of economic agency. A first important milestone is the work by Iris Marion Young. Her theory of structural injustice revolves around the idea that many injustices cannot be attributed in a clear causal way to the actions of single individuals, so that the latter could be held accountable for them under a ‘liability model’ of responsibility. Instead, she proposes a ‘social connection model’ that attributes responsibility to individuals who contribute to upholding the social structures within which these injustices occur (Young, 2011). 1
Young thus provides a diagnosis of injustices that presents an alternative to theories from the right (conservative-libertarians) and from the left (luck egalitarians), which both emphasize individual responsibility. According to Young, individuals often find themselves in social positions that are unjust, due to complex circumstances beyond their own control. Of course, other theorists – notably John Rawls and his followers – also focused on the ‘basic structure’ of society. But Young’s innovation is to understand ‘structure’ in a broader sense: instead of thinking that injustices are localized in a small set of core institutions, Young (2011: 70) proposes to scrutinize the total array of formal and informal institutions. In doing so, she uses a concept of structure that incorporates insights from sociologists such as Bourdieu, Giddens and others, such as the role of unintended consequences and the reproduction of social positions through the actions of individuals (Young, 2011: 52–64).
It is this notion of structure, and this focus on social positions, that we take to be particularly important for the present purposes. However, Young’s work remains somewhat open-ended when it comes to defining a criterion for assessing when exactly individuals are victims of injustices. She comes closest to providing such a criterion in the following passage: 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. (Young, 2011: 52)
The concept of structural injustice suggests that injustice needs to be cashed out in terms of the power relations between agents in different positions created by social structures. The crucial question hence becomes: which kinds of power differentials are unjust? Much of the recent literature here uses the concept of ‘domination’: when A is in a position of greater power than B and this gives them the opportunity to dominate B, this is unjust. Philip Pettit’s prominent account, in his 1997 book Republicanism, defines the dominating power of an agent over another as follows: ‘1. They have the capacity to interfere 2. on an arbitrary basis 3. in certain choices that the other is in a position to make’ (Pettit, 1997: 52). Interference must be intentional (although domination need not), and includes forms of direct coercion and manipulation. Moreover, it must worsen the position of the other, so Pettit excludes bribes and rewards as forms of interference. Arbitrariness means that the decision to actually interfere is up to the dominating agent. Non-arbitrariness, by contrast, refers to situations in which the dominating agent is ‘forced to track the interests and ideas of the person suffering the interference’ (Pettit, 1997: 55).
Pettit’s paradigm case of domination is the master-slave relationship, where the master can (but need not, as they may be benevolent) interfere with the slave’s life as they wish. A similar neo-republican position is defended by Lovett, who conceptualizes justice as minimizing domination (Lovett, 2010). More recently, the neo-republican ideal of non-domination has also been applied to economic life, leading to widely divergent policy conclusions, including workplace democracy (Anderson, 2015; Gourevitch, 2013), a basic income (Lovett, 2009; Pettit, 2007), market competition (Taylor, 2013, 2017) and market regulation (Dagger, 2006; Pettit, 2006; Rahman, 2017).
Some debate has revolved around the question of whether neo-republican approaches remain too focused on what Rahman calls ‘dyadic domination’, i.e. domination between two agents, one of whom intentionally dominates the other. But in economic life domination frequently occurs as a consequence of structural economic forces, which cannot easily be attributed to individual agents, i.e. ‘structural domination’ (Rahman, 2017: 47–49). Similarly, Alex Gourevitch argues that we need to recognize a separate category of structural domination within republicanism, where subjects are ‘dominated by a number of agents, but not any single, given agent in particular’ (Gourevitch, 2013: 596). Michael Thompson adds a focus on psychological processes: the subjects of structural domination often internalize their subordination, which makes it all the easier for the dominating party to extract a surplus from them (Thompson, 2013, 2018). 3
Acknowledging structural domination does not imply postulating some (metaphysically weird) agency by powerful structures themselves. Rather, it means acknowledging the way in which structures can enable agents to dominate others; structures cannot enable in the absence of agents (Lovett, 2010: 47–49). 4 This leads to new questions, however, about how exactly to understand the notion of ‘arbitrariness’: in a procedural or in a substantial sense. Pettit understands it in a procedural sense, as having certain controls in place, so that dominating agents do not rule over subjects at their own whim. But which controls are adequate depends on what ‘interests and ideas of the person’ (Pettit, 1997: 55) the controls track. For Pettit, this criterion in his earlier work refers to the ‘welfare and world-view of the public’ (Pettit, 1997: 56) and must be decided by recourse to public debate and political decision-making. In his later work, he often refers to the need to identify a set of basic liberties, which may come closer to the view we defend here (Pettit, 2008). 5
A procedural interpretation of domination can be pushed into a dilemma. Either the procedure is purely procedural, or the procedure de facto tracks substantive ‘interests and ideas of the person’. In the former case, republican theories run the risk of excluding cases of domination one wants to include, e.g. a law-making process that leads to the implementation of slavery (Ahlstrom-Vij, 2012). In the latter case, a republican theory would need to spell out the actual interests of the dominated subjects in more detail. However, here the procedural interpretation would collapse into a substantive one (as Pettit’s later work can in fact be understood). 6 In the neo-republican literature, this issue is largely evaded by concentrating on more-or-less intuitive examples, like the domination of 19th century industrial workers, as mentioned earlier. 7
All in all, recent work on domination makes a convincing case for putting forward a notion of structural domination for diagnosing structural injustices in the economy. What it does not do, however, is to provide a substantive account to flesh out which power differentials within the economy are cases of domination and hence unjust. In the following sections, we offer the outlines of such an account.
Autonomous agency as a criterion for non-domination
Our aim in this section is to introduce a normative benchmark for identifying cases of domination, i.e. unjust power relations, within the economy. In doing so, we want our account to fulfil two desiderata.
The first desideratum is to provide an account that focuses directly on power relations within the economy, not exclusively on distributive outcomes. 8 Even if, counterfactually, the powerful used their power in benevolent ways, thus ensuring outcomes that are just according to one’s favoured theory of distributive justice, the fact that this outcome contingently depended on personal favours by benefactors would still constitute an injustice. Unequal positions of power, not only the distributions that result from the use of unequal power, are an independent normative problem. This is an insight from neo-republicanism we wish to salvage. However, this does not imply that a separate evaluation according to principles of distributive justice is impossible or unnecessary. Rather, our proposal is compatible with (but not committed to) the claim that economic life should be analyzed through a double lens (see also the final section, below).
Second, we want to suggest an account which covers the economic domain as a whole. Most neo-republican texts focus on capitalist-worker relations. While this is indeed a central case, domination can also happen in other markets, for example in financial markets or markets for health care. An adequate normative account should also be able to address non-market forms of economic interaction. For example, domination can also happen within hierarchies, i.e. the internal relations within firms, in informal, non-monetized economic transactions between caretakers and caregivers, or between providers and customers of public services. These cases should also fall within the purview of an account of structural justice.
With these desiderata in hand, we propose to define structural economic injustice with the help of three propositions. The first one is the most general: (1) An agent is dominated if one or more other agents are in a position to impose their will on them in a way that deprives them of their general capacity for autonomous agency (hereafter for shorthand, ‘general agency’).
9
Obviously, there are several conceptions of autonomous agency on offer in the literature. While one of us has taken a stance on these debates elsewhere, here we will remain agnostic about the exact nature of autonomous agency (Claassen, 2017a, 2018). We accept and use the core of the idea, as formulated by John Christman and Joel Anderson when they argued that the debate about autonomy has centred on the precise nature of two conditions: competence and authenticity (Anderson and Christman, 2005: 3). On the one hand, if individuals are to be self-authors, they need certain personal competences – rational, moral, emotional and social – in order to make decisions and execute these in their lives. On the other hand, they need to make decisions that are authentically their own; this requires a certain relation to the self in which individuals are not, to use Christman’s term, ‘alienated’ from their own desires and inclinations. There is a strong family resemblance between the many theories of autonomous agency specifying these two conditions, and whichever one prefers can be used to fill in what follows. Some such notion of autonomous agency is required to make sense of what exactly is harmed when individuals in a social structure are dominated by others.
Ideals of autonomous agency have been subjected to several standard objections, which we all think can be answered. We note the four most important ones with a brief reply, to avoid misunderstandings of our view.
First, autonomous agency should not be taken to be an ‘asocial’, ‘atomist’ or ‘disembedded’ notion. Agency is deeply related to social structures. When an autonomous agent occupies a position in a social structure, however, they are not thereby reduced to a mere machine on which impersonal forces are at work, but empowered to act and thereby influence others in the social structure. To be an agent is not to live in a social vacuum, but rather to participate in a social setting in which the role one occupies, the rights and duties one has, all help define one’s leeway of action. The normative question is how much agency should be available to every agent in the structure; how much (dis-)empowerment every agent should have. This question itself can only be asked once it is first recognized that agency and structure are mutually compatible terms which together are necessary to understand the question.
Second, we should not think of this ideal of agency as overly individualistic, in the sense of being unable to address domination by collective agents, such as corporations (Anderson, 2017; Ciepley, 2013). Collectives are an important source of domination, and can also be subjects of domination. We here focus on individual agency, and do not address the question of what would be needed to extend our framework to collectives, but we do think that such an extension is possible and desirable.
Third, such an ideal of agency should not be understood in a utopian sense: it does not take perfect abilities to participate in social life. Rather, we think about it as a gradual concept (one can have more or less of the bundle of capacities constituting agency), which for normative purposes of diagnosing structural injustices is in need of some threshold (Feinberg, 1986: 29). All agents are influenced by social pressures and expectations (see also Herzog 2018: 23–28), and sheer luck continues to influence whether or not agents reach their goals, but in this mix of internal and external forces, only actions that have come about by a sufficient degree of someone’s own agency can be described as autonomous.
Fourth, such an ideal is not to be thought off as problematically perfectionist. A purely neutral view of politics is, in our eyes, impossible, since a polity that aims to remain neutral between the different conceptions of the good that individuals choose is itself non-neutral in giving itself this mission: to preserve the freedom or autonomy of citizens to do so. Of course this mission is not perfectionist in the way of, say, theocratic visions, which aim to dictate the scripts of how to live for all citizens in a society. It is a mildly or moderately ‘liberal perfectionist’ conception of politics. One can try to remain neutral between different conceptions of the good, but if one does so, this is not for ‘neutralist reasons’, but out of a respect for some ideal of autonomous agency. 11
Finally, we need to clarify the link between the concept of domination, as introduced in the first proposition, and the concept of (in)justice. As we see it, not every case of domination is one of injustice. The prisoner is dominated by their society in that the latter has deprived them of an important area for the exercise of their capacities for autonomous agency (without reducing it to zero, their agency is severely confined). Whether the prisoner is unjustly dominated is a separate question, for which we need a theory of just allocations of agency. In the fifth section we will provide part of such a theory when talking about threshold levels of agency; but we note that this is not a complete picture (since that would also require, for example, a theory of punishment for the case of the prisoners deprived of their agency). Here we only note that the two concepts of domination and injustice are not identical. Alternatively, one could keep domination and injustice together, and say that the prisoner is not dominated; but this way of speaking is less in tune with the established use of language and with the phenomenology of being held captive. The result would be the same: one would then need a theory for delineating justifiable and unjustifiable ways in which individuals can be deprived of their agency. We address this question in the sixth section.
Based on this concept of autonomous agency, we can now introduce the second and third propositions, which will provide a specification of the first one for economic contexts.
Introducing economic agency
In this section, we outline a concept of economic agency which can be used to diagnose structural injustices in the economic sphere. The second proposition links economic to general agency: (2) An agent is economically dominated if one or more other agents are in a position to impose their will on them in a way that deprives the former of economic agency: i.e. agency in acquiring the resources necessary for their general agency.
We do not imply that every resource deficit translates one-on-one into an agency deficit. For example, a theft of some of our belongings by definition deprives us of some of our resources, while it may only in some cases deprive us of our agency to such an extent that we are dominated. For example, if the good that is stolen is the bike we need to be able to get to work and we lack the means to buy a new one, we may become dependent on the neighbour who can give us a ride but can also refuse to do so, and this dependence can amount to a form of domination. Therefore, contextual analysis is necessary to show how resources and agency relate to each other in specific situations.
With this specification for the economic domain we leave open the possibility that there are also other (i.e. non-economic) forms of agency that are necessary conditions for general agency. For example, having agency in family affairs may also be a necessary condition for general agency. Moreover, we also leave open the possibility that there are non-economic forms of agency that are necessary conditions for acquiring economic resources (see also the conclusion, on political agency). While any definition that marks off one ‘sphere’ or ‘domain’ in society may be challenged – because the social reality often does not allow neat boundaries between these domains, e.g. the political and the economic are related in countless ways 12 – we think that such a definition is nonetheless helpful for theorizing.
Economic agency may remain a highly abstract and elusive notion if we do not attempt to spell out what its specific preconditions are. So what are these forms of economic agency? For answering this question, we introduce a third proposition: (3) Economic agency consists of the ability to acquire resources, by 1) producing these resources oneself, 2) acquiring these resources indirectly by a) earning income and b) gaining access to these resources by using one’s income, 3) acquiring these resources indirectly without actively earning an income, through systems of reciprocal gift exchange or entitlements to receive resources from the political community.
The definition given above may not be exhaustive, but it includes the main mechanisms of acquiring resources, i.e. through autarkic self-provisioning, through market-based production and consumption or through informal or public systems of provisioning. The definition does not by itself express any preference for some way or other of acquiring resources, but merely states that for whatever social system of provisioning is in place, individuals should be granted economic agency within it (and to be sure, the avoidance of domination is also an important consideration when one thinks about the choice between different forms of provision, but not necessarily the only one). In this way, we do not preclude difficult questions about which economic systems are best. However, it may turn out that some economic systems have a stronger tendency to lead to domination than others, as we will also see in the case of credit markets, in the next section.
An illustration: Creditor-debtor relations
To show the relevance of a focus on agency, and the specific content of the three propositions introduced in the previous section, we will now briefly consider the case of creditor-debtor relations.
The merits of credit markets can obviously be analyzed from an economic perspective, focusing on the efficiency of credit as a tool for matching the preferences for saving of some with the preferences for immediate investment or consumption of others. In such a perspective, it may be recognized that both creditors and debtors get into a position of dependency towards each other, given the uncertainty involved about the future ability of debtors to pay interest and pay back the principal. From this perspective, however, this dependence is relevant only to the extent that it has distributive consequences: it leads creditors to think about the right interest to ask for, and policy makers to create institutional mechanisms (e.g. bankruptcy laws) to resolve problematic cases. With the right interest rate and the right institutional mechanisms in place, the practice of creditor-debtor contracts enhances economic efficiency. In contrast, from the perspective of structural injustice, informed by a focus on the (lack of) agency of those in structural positions of domination, other issues and concerns come to the fore (Herzog, 2017a; for critical perspectives on financial markets in general, see also Herzog, 2017b).
First of all, debtors may borrow money because they cannot pay for their current needs or wants out of their current savings. Inequalities of wealth thus play a role at the very origin of credit markets. Such inequalities may or may not be unjust in themselves. What we can say is that it renders the economic agency of those on both sides of credit markets unequal. Where such inequalities are gross, debtors are often at the mercy of their creditors; in modern credit markets, this is usually not the kind of personal dependence that characterized many historical debt relations, but rather a structural dependence in the sense that the debtors have no way of controlling the conditions that they are offered or of holding creditors accountable.
Societies can enhance or diminish the economic agency of debtors, by making available options for bankruptcy or restructurings of their loans, by regulating interest rates, by enhancing their financial literacy, etc. Societies can also enhance or diminish the economic agency of creditors, by making available to them options to put their debtors into prison or forced labour, disallow debt restructuring, harsh bankruptcy laws, etc. The occurrence of dominating relations between creditors and debtors is a matter of the economic agency (i.e. the capacities and the options) available to actors on each side, as our perspective emphasizes. It is not the underlying inequality of wealth per se which determines whether there is domination, but the institutions that determine the economic agency of debtors and creditors.
Second, our focus on autonomous economic agency also brings into focus what was called the ‘authenticity conditions’ of autonomy (see the explanation of proposition (1) above). Thus, as Alex Gourevitch argued, two students both going to college, one from a wealthy background, the other with the prospect of a $100,000 student debt, will make different educational choices. The poorer student: will be guided toward choosing degrees and career paths that promise better earning potential. To pursue his education with the same intellectual freedom and experimental attitude as the other student would require the indebted student to be much less risk-averse than his colleague. (Gourevitch, 2012: 144)
Third, one may wonder whether all debt is equally problematic from the point of view of structural injustice. Is the case of a person getting into debt to finance their private indulgences equally unjust as a person getting into debt over financing their family’s house, health care and education? Here the agency perspective suggests differentiating between cases. The reason lies in the connection we have drawn between general and economic agency. In our proposition (2) domination was restricted to cases where economic agency was necessary for achieving general agency. Therefore, our perspective can restrict domination in credit markets to cases where people try to finance the goods and services necessary for general agency (as arguably housing, education and health care are), and exclude cases where debt contracts relate to non-necessary goods.
Fourth, debt is an alternative source of income to finance these necessary goods compared to receiving them as public goods or through labour-earned income. Thus the analysis of debt also raises the question of the comparative merits of each of these economic mechanisms, and the forms of economic agency involved. Empirically, dependency on debt has increased as the state has rolled back its direct provisioning of public goods and as wages have stagnated. Normatively, this raises the question whether it is just to become dependent on debt for these goods, and if not, which mechanism(s) would be (more) just. Thus Gourevitch makes a case for the unconditionality of public provisioning as best able to provide these essential goods in a way which does not undermine the ‘value of the freedom’ that these goods are meant to serve (Gourevitch, 2012: 147). If that value lies, as we suggest, in general autonomous agency, then this is the standard to which credit markets should be held, in comparison to alternative economic mechanisms. We will not do that comparative analysis here, but note that the disjunctive form of proposition (3) makes this question visible.
While creditor-debtor relations are a common example in neo-republican literature, we hope to have shown how an analysis of the injustice of these relations requires the introduction of general and economic agency and the interplay between both. We have not offered a full analysis of the conditions of agency in credit markets, but hopefully enough has been said to show how a concern for agency underlies worries about domination.
Specifying injustices: Capabilities and thresholds
We are now in a position to return to the question raised at the end of the third section: which instances of domination form an injustice? In this section we will suggest that this requires us to elaborate both a metric and distributive principle for economic agency.
As to the metric, critics might worry that the notion of economic agency is still too abstract. It can, however, be spelled out in more concrete terms, for example in terms of capabilities (Nussbaum, 2006; Robeyns, 2017; Sen, 2009). 13 If one understands economic agency as a bundle of capabilities, both the internal and the external sense of the term need to be considered. Internal capabilities refer to a person’s skills or capacities. For example, an agent in a financial market needs skills to interpret financial information. Such financial literacy is a key component of agency in this context, and its absence can lead to egregious cases of abuse, especially if a society’s economic structures are organized such that agents have to rely on financial markets, e.g. for saving for their old age. External capabilities refer to options agents have. For example, financial consumers need to have meaningful choice between financial products with different risk structures. Taken together, internal and external capabilities can define a person’s agency in a specific context. We find the capability framework the most plausible way to spell out the demands of agency, as one of us has done both in general (Claassen, 2017a, 2018) and as a way to think about (financial) markets (Claassen, 2017b). Note, however, that those who reject a capability framework might still want to accept our arguments about agency, and spell out the requirements of agency differently.
Our account is both quantitative and qualitative. It is qualitative in the sense that it defines agency by spelling out substantive conditions for being able to acquire resources in specific contexts. Hence, we must analyze the economic practices in which health care, or mortgages, or wage labour are (re)produced, to see what it means to have agency in these contexts. This may change over time, and differs between societies. The identification of domination thus depends on an interpretation of what is required in each of these specific contexts. The proposal remains substantive, rather than merely formal, in the sense that justice is defined through a substantive account of agency, e.g. by a list of capabilities. Nonetheless, societies will develop political procedures for arguing about these substantive conditions, and a fully worked out philosophical theory of economic justice would be an input or proposal in such a political procedure.
A full theory of economic justice would thus consist of 1) an identification of a full list of all necessary capabilities for general agency, and the economic resources necessary for such agency; and 2) an identification of the requirements of economic agency, i.e. of a complete list of the capabilities necessary to acquire these economic resources. While the first step would arguably be somewhat context-independent, in the same sense in which a list of human rights is context-independent, the second step is highly context-dependent, and here a theory of justice would have to position itself: it could be formulated in more exact detail for a specific context, or remain at a more general level so as to capture a greater range of variation in how societies provide for the resources in question.
The proposal is quantitative in that it provides guidance for spelling out, for each context, how much economic agency is required for justice. This is the distribuendum that our account proposes, which needs to be complemented by a distributive principle. The latter could, for example, be sufficientarian, prioritarian or egalitarian. Theorists with different leanings could accept our general framework, and yet disagree on what justice requires in different contexts.
Our own proposal would be a combination of sufficientarian and relative criteria. Below a certain threshold, agency stops being agency in any meaningful sense. In some areas, this threshold is absolute – for example, if individuals cannot cover basic physical needs, they obviously lack agency. This is the case in Young’s example of Sandy, a young mother threatened by homelessness: the ability to access physical shelter is an essential capability, without which individuals are at the risk of being dominated by others (specific individuals or groups of others), which is unjust (Young, 2011: 43). To remedy this injustice, Sandy could receive either a financial subsidy in order to acquire housing in the market, or access to social housing at affordable costs.
In many contexts, however, it is not enough to provide sufficient economic agency for individuals not to be dominated, or it is difficult to even understand what ‘sufficient’ economic agency would even mean in the abstract, without taking into account how agency is socially embedded. Several factors can have an impact on how large the agency of individual agents is. A first, obvious point is that the general price level is decisive for how much their income is worth, hence their agency relative to that of others is influenced not only by what happens to their own position, but also by what happens to others’ positions. Their bargaining position in markets also depends on their own agency relative to that of others.
Take, for example, information as a precondition for economic agency. If the inequality in information between a supplier and a customer is too large, then one can dominate the other. Another factor is positional goods, the value of which depends, at least in part, on their status relative to others (Brighouse and Swift, 2006; Hirsch, 1999). The example of Sandy illustrates this point as well: what counts as a ‘good’ neighbourhood, in which she can bring up her kids without them suffering from various disadvantages, has to do not only with absolute, but also with relative standards. Hence, there are often reasons to move beyond a sufficientarian standard, and to make sure not only that every person is above a certain threshold level of economic agency, but also that the differentials between them do not become too large.
A separate question is whether such a distributive standard needs to be adjusted to account for self-inflicted forms of domination; i.e. whether a concern with individual responsibility for bad, i.e. domination-engendering, choices needs to be built in. While a concern for agency implies that agents are to some extent allowed to make bad choices, the same concern also requires taking into account everyone’s level of agency after such choices. Thus, where free exercises of agency lead to dominating relations, the effects of these choices need to be undone to some extent to ensure sufficientarian levels of future agency (this may lead to moral hazard, but this needs to be accepted).
Thus, at this level, theoretical considerations that are familiar from ‘conventional’ theories of justice, which usually focus on outcomes, make a re-entry. Some readers may wonder how much we have then gained, compared to these conventional theories. However, we hope to have shown that the opposition is a false one. It is highly desirable that theories of justice can go beyond an evaluation of economic outcomes, and also take into account the economic structures generating those outcomes. But a normative evaluation of such structures requires a view on which distributions of economic power are fair. And such a distribution, we have argued, can be evaluated with economic agency as the distribuendum of justice.
Objections and replies
In this section, we address two objections. Both objections concern, albeit in different ways, the relation between a normative evaluation of economic power structures (as proposed in this article) and theories of justice that focus mostly on distributive outcomes, without taking structures, and the processes that are interrelated with them, into account. One can think of the first objection as questioning the necessity of a focus on structural injustice and the second objection as questioning the sufficiency of such a focus.
The first objection questions whether a focus on structural injustices is necessary. Why can’t one focus on unjust outcomes alone? Why bother with the content of the ‘black box’ if the outcomes it produces are just? Should we not be agnostic between two scenarios, both of which are different in terms of how the economy is organized and how much economic agency is available to its participants, if they have the same outcomes for all? And why not simply bring these outcomes about by redistributive taxation? Why should one want to implement laws and institutions that would try to abolish domination in the power relations between creditors and debtors, capitalists and labourers, or any other two economic groups, if one can also simply make everyone (sufficiently) independent from the economic game by old-fashioned redistribution?
In the latter case, one would not need to bother with establishing and enforcing detailed mechanisms of control on economic activity. To be sure, some proponents of welfare entitlements would argue that what they achieve is precisely to change the power relations within the economy, and traditionally more vulnerable parties see their economic agency enhanced. From that perspective, welfare entitlements are considered not only as an ‘outcome’ (giving access to consumption possibilities on a par with income generated on the market), but rather as an ‘input’ into tomorrow’s starting positions for new rounds of economic activity. This is in line with our general proposal, but as clarified below, we would question whether welfare entitlements are always and sufficiently effective for enhancing the economic agency of vulnerable parties.
Against this objection, two responses can be made, one stressing the instrumental value, the other the intrinsic importance of economic agency.
Instrumentally, handing out welfare entitlements might often not be enough. The many disparities in agency within economic life are likely to be too complex to be solved with one magic bullet. A first problem with welfare entitlements is that they may not be high enough to provide general agency for all. But even if they were high enough, economic agency often requires other-than-monetary conditions as well. For example, it may require access to certain information. Now to the extent that ‘money can buy anything’, one may try to argue that these non-monetary conditions can be realized through publicly financed entitlements as well, but this seems unlikely. For example, it might not be sufficient to secure economic agency in contexts in which the legal position of different market participants is highly unequal; here, changes in the legal framework might be required as well.
The second response points out the intrinsic value of economic agency. In Hirschman’s terminology, the strategy of offering welfare state entitlements puts all cards on the ‘exit’ strategy: individuals receive resources from outside the economic sphere. But we think that ‘voice’ in economic life is important as well, from the perspective of autonomous agency in general. Economic life is an enormous part of one’s existence, and not being able to operate autonomously within this sphere is itself a violation of the principle of autonomous agency, because an important component of general agency is missing. This point is particularly relevant for our working lives, and has in fact been commented upon by a number of authors, e.g. by those republicans opposing basic income in favour of regulations enhancing more just working conditions. Here the idea is that individuals spend so many hours on the job that no compensation in terms of income, however generous, can compensate for working conditions in which they are dominated. 14 We think this point generalizes to other parts of economic life as well, although most likely not to all of them. Depending on how a society organizes its economic sphere, different positions, e.g. as renter or as customer of health services, can be such that it matters intrinsically that one is not dominated within them. Often, this is a matter of legal protection and the availability of counter-power (e.g. unions or consumer associations), rather than of redistributed income.
The second objection questions whether a structural injustice focus is sufficient as a full analysis of justice. It seems, so the objection goes, that economic processes not hampered by structural injustices could still have distributively unjust outcomes. Familiar examples include cases in which economic processes generate small differences in income, which over time accumulate to large inequalities in wealth, or cases in which voluntary transactions with some degree of ‘winner-takes-all’ characteristics lead to single individuals ending up with large amounts of resources, as in Nozick’s Chamberlain example (Nozick, 1974: 161). We can imagine such processes even in societies in which no economic domination were present. The outcome might nonetheless contradict our intuitions about a just society.
There are two avenues for responding to this objection. The first is to accept that one should analyze economic processes from two perspectives: one that focuses on dominating structures, and one that focuses on outcomes in terms of resource distributions. Both domination and distributive outcomes could give rise to complaints of justice, which are often intertwined in practice, but nonetheless based on different theoretical foundations. This would mean that different kinds of theories of justice would lead a peaceful co-existence, except for cases in which they would point into different directions. Such cases would have to be accepted as cases in which trade-offs and compromises between different normative considerations need to be sought. For example, unjust power relations might sometimes appear as a price worth paying in order to bring about more just distributive outcomes in terms of resources, especially in situations in which the distributive concerns are urgent. This is indeed how debates about controversial economic practices, e.g. sweatshops, are sometimes framed in public discourse, e.g. when they are described as oppressive – i.e. containing an unequal distribution of power – but desirable for distributive reasons – i.e. leading towards improvements in terms of distributive justice.
Second, for those who want to hold up the theoretical primacy of theories of justice that focus on distributive outcomes, without taking power structures into account, there is a simple pragmatic reason for why economic agency might nonetheless be an interesting additional focus of theorizing. In practice, low degrees of economic agency are likely to also lead to distributively unjust outcomes – this is the instrumental point discussed earlier. Remedying them by redistributive taxation is not always politically feasible. Hence, a focus on economic agency might be seen as broadening the arsenal of tools for bringing about more just distributions. It aligns theories of distributive justice with the various political struggles that are fought in order to strengthen the economic agency of vulnerable agents, e.g. by NGOs, cooperative movements and unions. This is a valid theoretical possibility. We find it more plausible, however, to read such struggles as expressing not just an exclusive focus on outcomes, but also a quest for economic agency and for an end to economic domination.
Conclusion
In this article we have attempted to lay out the major components of a theory of structural injustice for the economic sphere, which defines unjust domination in that sphere in terms of autonomous agency. In concluding, we would like to emphasize that we do not see this theory as a full picture of what is required to combat injustices in the economic sphere. At least two major components would have to be added.
First, we have related the value of economic agency entirely to its instrumental function in safeguarding conditions for what we called ‘general agency’. While we do see this as a major function of economic activity, it does provide an instrumentalist lens on the economic sphere which needs to be complemented by an intrinsic lens. Individuals engage in economic activity not just to acquire resources for their general agency, but also to exercise their general agency. Individuals expect their working lives to have a quality of their own, as is expressed, for example, in demands for meaningful work (see e.g. Yeoman, 2014). Where economic activity becomes unjust on account of failing to provide meaningful work and other intrinsic qualities, is to be discussed elsewhere.
Second, to combat injustices in the economic sphere, political agency is necessary. Our theory of economic agency therefore needs to be complemented by a theory of political agency. 15 This is all the more urgent since the two overlap to some extent, especially since money can enter politics all too easily. While much political action happens in and around non-economic institutions (parliamentary democracy, public sphere and media, street protests, etc.), fights for economic justice happen partially within the economic sphere itself. We understand political agency (in so far as it serves economic justice) here as agency in order to change the economic rules of the game, and some of that happens as individuals participate in trade unions, work councils in corporations, shareholder meetings, etc. A full theory of just economic agency would need to address such forms of agency as well.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.