Can Relational Egalitarians Supply Both an Account of Justice and an Account of the Value of Democracy or Must They Choose Which?

1. Introduction

Construed as a theory of justice, relational egalitarianism says that justice requires that people relate as equals (e.g., Anderson 1999; Bidadanure 2016; Fourie et al. 2015; Lippert-Rasmussen 2018; Scheffler 2015).¹ Construed as a theory of what makes democracy valuable, it says that democracy is a necessary, or constituent, part of the value of relating as equals (Anderson 1999; Kolodny 2014; Viehoff 2014). These may be said to represent two strands in the existing literature on relational egalitarianism, with the first strand—such as Anderson (1999) and Scheffler (2003)—focusing on relational egalitarianism as a theory of justice, and the second strand—such as Kolodny (2014) and Viehoff (1999), but also Anderson (1999)—focusing on relational egalitarianism as a theory of democracy.² Typically, relational egalitarians do not explicitly distinguish between these two strands or ways of construing relational egalitarianism. One reason for this is that they assume that justice requires that people relate as equals and that democracy is valuable because it is a necessary, or constituent part, of relating as equals. However, construed in the latter way relational egalitarians could simply say that democracy is a necessary part of relating as equals and that relating as equals is valuable, albeit not a requirement of justice. There would then still be a reason to have democracy, only not a reason of justice.

In this article, we argue that relational egalitarians face a dilemma—henceforth the justice-democracy dilemma—if they want their theory to provide both an account of what justice requires and an account of what makes democracy valuable. The dilemma is that understanding social relations in a way that renders relational egalitarianism plausible as an answer to what makes democracy valuable comes at the price of understanding them in a way that makes it less plausible as a theory of justice, and vice versa.

To establish the existence of the justice-democracy dilemma, we start (§2) by discussing what it means to be socially related, since relational egalitarianism only applies to those who are socially related. In presenting two distinctions, which cut across each other, we identify four different accounts of what it means to be socially related. We then analyze what these accounts imply in terms of whether future people (§3a) or dead people (§3b) should be included in democratic decision-making. The motivation for this analysis reflects the assumption that the boundary problem in democratic theory—the problem of identifying those who should be included in democratic decision-making—must be solved through an appeal to the value of democracy.³ However, not all the four understandings of what it is to be socially related enables relational egalitarianism to provide a plausible solution to the boundary problem. Also, this analysis (§3c) brings to the fore the justice-democracy dilemma.

Section 4 concludes by briefly addressing five suggestions about how relational egalitarians could respond to the justice-democracy-dilemma; that is, they could (i) restrict the scope of their theory to justice, while being silent on the value of democracy; (ii) restrict their theory to be solely a theory of what makes democracy valuable, while being silent on justice; (iii) propose an understanding of what it means to be socially related that is different from ours; (iv) bite the bullet, e.g., by accepting that rational egalitarianism does not condemn certain acts that are widely believed to be unjust or by accepting that dead people should be included in the demos; or (v) argue that relational egalitarianism is not one coherent theory but a disjunct of two different ideas of how social relations ought to be, where what it means to be socially related is judged differently on the two. Although there is no way out of the dilemma which does not come with significant costs for relational egalitarians, option (v) may be their least bad option. Whichever response relational egalitarians adopt to the justice-democracy dilemma, our article shows that they must be more explicit on what it means to be socially related.

Before we move on, we should say a bit about how we understand justice, both when it comes to how we are conceptualizing justice, and what we take justice to apply to.⁴ Let us start with the former. As Peter Vallentyne (2014: 40) notes, “the term ‘justice’ is, unfortunately, used to mean different things by different authors.” He then goes on to distinguish several different meanings. Two of those views are particularly relevant for our purposes. On the first, justice is “understood to mean (something like) moral permissibility of social institutions …. [On this view] the specific content of justice is determined by the objects assessed rather than by the set of moral concerns relevant for the assessment” (40). On the second view, “justice is interpersonal morality understood as the duties that we morally owe each other …. What we owe others is what they can claim from us, and this is that to which they have a right against us” (41). We will understand justice in this second sense, but our argument does not require this understanding. When relevant, we present arguments appealing to cases that bear on justice whether one accepts an institutional or interpersonal understanding thereof.⁵

When it comes to the second issue—what we take justice to apply to—it is common to distinguish between an institutionalist and a non-institutionalist view. According to the former, justice only applies to the basic institutions of society (Rawls famously defended this view). According to the latter, justice applies both to the basic institutions of society and to individual acts (or, in relational egalitarian terms, to institutional-individual relations and inter-individual relations).⁶ Relational egalitarians are divided over this issue.⁷ Relational egalitarians such as Pogge (2004) and Schemmel (2012; 2021) are institutionalists. Other relational egalitarians—such as Anderson (1999), Kolodny (2014), and Scheffler (2015)—are non-institutionalists. Fortunately, for our purposes, we do not have to take sides in this debate. As with the previous distinction, our arguments below apply to both institutionalist and non-institutionalist versions of relational egalitarianism. When relevant, we present arguments appealing to cases that bear on justice whether one thinks that justice pertains to institutional-individual or to inter-individual relations.

2. Being Socially Related: Two Distinctions

Relational egalitarianism focuses on social relations—people must relate as equals. Usually, they motivate their view by pointing to inegalitarian relationships that we find objectionable, for example, “the servant is subordinate to the lord of the manor, the slave subordinate to the master … the plebian is lower than the patrician, the untouchable lower than the Brahmin … the paradigms [these examples of inegalitarian relationships] provoke in us a sense of unease” (Kolodny 2014: 292). The theory has gained traction in recent years and is now prominent both as an account of justice and as an account of what makes democracy valuable.⁸

As seems obvious, relational egalitarianism only speaks to groups of people who are socially related. After all, it is only among those who are socially related that inegalitarian or egalitarian social relationships may exist (e.g., Kolodny 2014: 293). Also, relational egalitarians point to this feature of their theory as an advantage over distributive theories of justice, because the latter, unlike the former, implausibly (so relational egalitarians claim) apply to distributions across individuals who are not socially related—for example, 14^th century French people and present-day French citizens (see Anderson 1999: 313). Accordingly, relational egalitarians must tell us what it means to be socially related on their view—a task that has so far mostly been neglected,⁹ as they have tended to assume an already bounded community of people who must then relate as equals. In this article, we take some steps in the direction of fulfilling the task of analyzing what it means to be socially related in such a way that the ideal of relational equality applies to people who are in these relations.¹⁰

In determining what it means to be socially related on relational egalitarianism, we start by distinguishing a moralized from a non-moralized view of social relations. A moralized view determines what it means to be socially related in line with the concerns of relational egalitarians. For instance, relational egalitarians oppose racism.^11,12 A moralized view would then specify that X and Y are socially related if and only if one, or both, can treat the other in a racist way. Relational egalitarians are not only concerned with racism. For instance, they also object to domination,¹³ exploitation,¹⁴ and paternalism.¹⁵ On a moralized understanding, we would also include these concerns, for example, X and Y are socially related if and only if X is able to dominate/exploit/paternalize Y and/or Y is able to dominate/exploit/paternalize X. The resulting general concept of being socially related would thus be as follows: X and Y are socially related if and only if X treats (or is able to treat) Y in one of the ways identified by relational egalitarians to be morally objectionable and/or Y treats (or is able to treat) X in one of the ways identified by relational egalitarians to be morally objectionable.

On a non-moralized view, we do not define social relations in a way which is tailored to the concerns of relational egalitarians. Instead, we try to come up with a lexical account of what we mean when we say that two people are socially related. This is what one of the authors of the present article intended to do in previous work with the following definition: “X and Y are socially related [if and]¹⁶ only if (i) X is socially related to Y and Y is socially related to X; and (ii) X can causally affect Y and Y can causally affect X” (Lippert-Rasmussen 2018: 126). In motivating the notion, readers were asked to suppose there are people on Venus. The fact that we, as Earthlings, cannot causally affect Venusians means that we are not socially related to Venusians. Suppose, however, that we could and do communicate with Venusians. In that case, we may want to say that we are socially related to Venusians, in the same way that we want to say that Beth is socially related to her cousin, Adam, whom she can and does communicate with via Facebook although she lives in Europe and he lives in the US (and they have never met in person). Thus, what may be driving our taxonomical disposition to say that we are not socially related to Venusians in the aforementioned case is that we cannot communicate¹⁷ or interact with them (and what is driving our view that Beth is socially related to her cousin in the US, although they have never met physically, is that they often communicate with each other). This is why Lippert-Rasmussen (2018: 128) adds a third necessary condition for X and Y being socially related, namely that X and Y must be able to adjust their conduct in light of each other’s conduct and communicate. It thus seems that a plausible non-moralized notion of what it means to be socially related takes ability to communicate and/or¹⁸ interact, or actual communication and/or interaction, to be necessary (especially) and sufficient for being socially related. We will assume so in what follows. But note also that this—that some form of interaction is required for relevant social relations—is in line with what other relational egalitarians have emphasized. Take, for example, Kolodny. He presents the following case:

Rabbit Hunters:

Suppose that, in a state of nature, several people collaborate in producing some means. Then some of them run off with an unfair share of the fruits of their labors, never to encounter the others again. (2014: 293)

Commenting on this example, he says,

There is a disparity of means (snared rabbits, say) and a disparity that results from a failure of equal concern for people’s independent claims to them (given equal contributions, the rabbits should have been split equally). Nevertheless, because the thieves and their victims do not continue to live together, because the disparity is not, as it were, woven into the fabric of ongoing social relations, there is no structure of hierarchy or subordination between them. (2014: 293)

And he adds: “Since there is no further interaction between them [in Rabbit Hunters], the theft does not produce any disparity in power or authority over their victims” (Kolodny 2014: 299). There is no objection from a relational egalitarian point of view in this case—although there is from a distributive point of view—precisely because “there is no further interaction between them” and thus no relevant social relations between the parties. In short, he takes temporally extended interaction to be required for there to be social relations of a kind that should concern relational egalitarians. And Schemmel seems to make a similar point—emphasizing interaction—when he says,

it is natural to think of requirements to set up significant social and political relations on an egalitarian footing as triggered primarily, or even exclusively, by patterns of already existing, non-trivial, social interactions. That is, it is natural to demand relational content of social justice on the basis of a relational grounding of it (2021: 294).

In our terminology, they seem to have a non-moralized view of social relations in mind.

The distinction between a moralized and a non-moralized view of social relations cuts across another distinction, namely the distinction between a One-way View and a Two-way View. According to the One-way View, it is necessary and sufficient for the existence of a social relation between two parties that one of them satisfies that which we identified as the necessary and sufficient requirements in relation to the former distinction (between a moralized and a non-moralized view). To illustrate, on the moralized view for X and Y to be socially related, it suffices that X is able to treat Y in a racist manner (even though Y is not able to treat X in a racist manner). Or, on the non-moralized view, X and Y are socially related if and only if X can communicate with Y or Y can communicate with X. This is not sufficient on a Two-way View. On this view, it is necessary and sufficient that both parties satisfy that which we identified as the necessary and sufficient requirements in relation to the former distinction. To continue with the racism example, on the moralized view for X and Y to be socially related, X must be able to treat Y in a racist manner and Y must be able to treat X in a racist manner (or in some other manner relational egalitarians find unjust).¹⁹ Or, on the non-moralized view, X and Y are socially related if and only if X can communicate with Y and Y can communicate with X.

The One-way View and the Two-way View are clearly different. Consider the following example:

Racist Nurse:

A black person is in the hospital after a car accident. He has become paralyzed because of the accident—literally, he cannot move and he is unable to communicate. The only thing he can “do” is to lie down. His nurse is a white racist who steps into his room and says aggressively: “We would be better off without black people like you!”

We are aware that there are different views on what racism amounts to. On one view, racism is a matter of having certain non-cognitive mental states and dispositions, or, as one prominent defender of this view puts it (Garcia 1996; 1997; 1999): racism is “in the heart” of the racist. Racism, on this view, “is something that essentially involves not our beliefs and their rationality or irrationality, but our wants, intentions, likes, and dislikes and their distance from the moral virtues” (Garcia 1996: 6). On another view, racism is (if not only, then at least also) a matter of having certain cognitive mental states and dispositions. Shelby (2002), a leading proponent of this view, argues that beliefs are essential to racism. He suggests that “we view racism as fundamentally a type of ideology [where] ideologies are widely accepted illusory systems of belief that function to establish or reinforce structures of social oppression” (415). On a third view, racism is not just a matter of what goes on in the heads of racists. Rather, racism is crucially tied to hierarchical social structures such that, were these structures to be eliminated, racism would cease to exist (e.g., Haslanger 2012; 2015; Mills 1997; Young 2009; 2011). And, of course, there are other views. Now, it is far beyond the scope of this paper to settle how we should understand racism. Taking this into account, we will simply assume that Racist Nurse involves racism in whatever sense the reader favors.

The One-way View implies that in this case, the black patient and the white nurse are socially related since the white nurse treats the black patient in a racist manner. This is not the case on the Two-way View since the black patient is not able to treat the white nurse in any way disfavored by relational egalitarians. What this shows is that if relational egalitarians were to base their understanding of social relations on a Two-way View, they would be unable to object to this instance of racism qua relational egalitarians because the relationship would not fall within the scope of relational egalitarianism. If relational egalitarians want to be able to object to Racist Nurse qua relational egalitarians, they must support a One-way View, and not a Two-way View, of being socially related.²⁰

One may say that these two distinctions—between moralized/non-moralized and one-way/two-way—fail to capture the complexity in our concept of what a social relation is. There are many kinds of social relations that are qualitatively distinct. We stand in cooperative relations, religious relations, familial relations, and so on. However, we must be wary here to distinguish between the type of relation, on the one hand, and whether a social relation is in place, on the other hand. Clearly, once social relations are in place, such relations can, qualitatively speaking, take many different shapes. Familial relations are different from friendship relations. And religious relations are different from cooperative relations. And so on. But that speaks to the type of social relation, e.g., whether it is of a religious or familial character. We are interested in the question, not of the quality of the social relation, but when there is a social relation to begin with. We take our question to be, at least in some respects, prior to the quality question, although there is, to be sure, some overlap between the two. However, there might still be other ways of defining social relations than those we have so far considered. We explore some of these later, when discussing possible ways out of the dilemma.²¹ But we also believe that the understandings of social relations we have proposed are a natural starting point, both because of what we typically have in mind when we say of two people that they are socially related and the commitments of relational egalitarians.

For now, it is important to notice that the distinction between a moralized and a non-moralized view and the distinction between a One-way View and a Two-way View cut across each other in a way which gives us the following quadripartite taxonomy of what it means to be socially related:²²

That we may distinguish between these four accounts of what it means to be socially related, which may in turn underlie four different accounts of relational egalitarianism, is interesting for several reasons. First, it is interesting because it shows that relational egalitarians must be clear on which account of being socially related that they prefer. There are large differences between these views, as we will see. Second, in laying out their view more fully, relational egalitarians may be vulnerable to novel objections which were hidden by the lack of clarity regarding what it means to be socially related, as we will also see. A third reason, which is connected to the second, is that who are socially related in the relevant sense matters for who should be included in democratic decision-making. Let us start by turning our attention to this third point.

3. Democratic Inclusion on the Relational Egalitarian View of the Value of Democracy

To see why these four views of what it is to be socially related matters to democratic inclusion, we must start with the relational egalitarian view of the value of democracy. According to relational egalitarians, democracy is not, or at least not only, valuable due to the outcomes it produces.²³ Democracy is valuable because it is a necessary, or a constituent, part of relating as equals. We may illustrate this through Kolodny’s prominent relational egalitarian view of the value of democracy.²⁴ He argues that democracy is a particularly important constituent of relating as equals because democratic decisions (i) cannot usually be escaped at will; (ii) characteristically involve the threat of force against non-compliers; and (iii) have final de facto authority (Kolodny 2014: 304–307). They have final de facto authority, first, because political decision-making cannot be moderated by a higher court of appeal, and, second, because political decisions have final authority over nonpolitical decisions (306).²⁵

According to the relational egalitarian view of the value of democracy, social relations between people and democracy go hand in hand (Lippert-Rasmussen & Bengtson 2021). If people are socially related and do not make democratic decisions together, they do not relate as social equals, as explained by Kolodny’s account. But, as we have seen, relational egalitarians may choose between different accounts of what it means to be socially related. These lead to different answers as to who should be included in democratic decision-making. Hence, which account of being socially related relational egalitarians subscribe to may matter as to whether the relational egalitarian view of the value of democracy is promising as a solution to the boundary problem in democratic theory.

The boundary problem in democratic theory is concerned with the question of who should be included in democratic decision-making.²⁶ As it has been forcefully argued recently, the solution to the boundary problem derives from the value of democracy. That is, we must bound the demos in accordance with what makes democracy valuable in the first place.²⁷ If the demos is bounded in accordance with the value of democracy, we cannot, out of a concern for democracy, complain against this demarcation.²⁸

Why should we demand of a view of the value of democracy that it can provide a plausible solution to the boundary problem? To be clear, we do not want to suggest that this is the only requirement in relation to which we should evaluate a view of the value of democracy. But it clearly seems to be one such requirement. As López-Guerra says,

However democratic the procedures, if the demos in question is composed only of white men, the clergy, the rich, or the “best” people in society to the exclusion of others who are equally bound by the decisions (full adult members of the state but not of the demos), then we are certainly not in the presence of a democracy. Varying restrictions on membership in the citizen body while holding political institutions constant could, for example, produce either an aristocracy or an oligarchy. What makes the difference is who governs or, alternatively, who holds ultimate power over those who govern. (2005: 218)

If this is true, then demanding that a view of the value of democracy provides a plausible solution to the boundary problem is a warranted requirement in the sense that, if it did not, it would no longer be a democratic view in the relevant sense, or at least it would not be a plausible theory of democracy. Suppose that a view of the value of democracy implied that only people with an IQ above 100 should be part of the demos. In that case, we may indeed be sceptical that this was a view of the value of democracy, as opposed to, say, the value of epistocracy. Or, less strongly, we may indeed be sceptical that this was a plausible view of the value of democracy. What distinguishes democracy, as López-Guerra explains, is not the decision-making institutions as such. As he says, “nothing precludes a military junta from adopting fair voting mechanisms and establishing deliberation rights among its members” (López-Guerra 2005: 218). What makes a democracy (valuable) is (also) a matter of how the decision-making body is composed. It suffices for our purposes that the demos composition entailed by a view of the value of democracy is at least part of the basis on which we evaluate a view of the value of democracy. If this assumption is granted, then a view of the value of democracy must also deliver a plausible answer to the boundary problem. Thus, a view of democracy’s value which led to an implausible demarcation of the demos need not fail to specify a democratic decision-procedure. However, it would be an implausible view, for that reason, of democracy’s value.

As we said, those who are relevantly socially related should make democratic decisions together on the relational egalitarian view of the value of democracy. What we want to do now is therefore analyse what these different accounts of what it means to be socially related entail for who should be included in democratic decision-making on the relational egalitarian view of the value of democracy.

We will primarily be concerned with what the accounts entail in terms of whether future people and dead people should be included in democratic decision-making. We focus on future people because there is a central disagreement about whether a plausible answer to the boundary problem entails inclusion of future people.²⁹ Hence, analysing what the relational egalitarian view entails with regard to whether future people should be included places the view in relation to the two most prominent solutions to the boundary problem, namely the all-affected principle and the all-subjected principle.³⁰

We focus on dead people because it seems a considered moral judgment that dead people should not be included in democratic decision-making.^31,32 One may object to this by pointing out that dead people have interests in virtue of which they can be harmed or wronged—depending on your view of posthumous harm (Bradley 2009; Wilkinson 2011)—so they should be included to have their interests protected. That dead people have interests is insufficient to establish that it is plausible to claim that they should be included in democratic decision-making. Animals also have interests, but it would still be implausible to maintain that animals should be included in democratic decision-making in the way that competent human beings are. This does not mean, however, that people may ignore the interests of animals when making decisions democratically—the interests of animals ought to be considered by whoever form the demos even though animals should not be included. That may also be the case for dead people: Dead people should not be included, but this does not imply that contemporary people may rightfully ignore the relevant interests of dead people, if they have any such interests, when making decisions (Saunders 2011: 286).

An objector may push further and say that even if a solution to the boundary problem were to imply inclusion of dead people, it is not clear why this should lead us to conclude that the view would be implausible as a theory of democracy. After all, not getting the right answer on dead people does not seem to have the same paradigmatically inegalitarian implications as some of the cases that seem to cut against the egalitarian core of democracy, such as distributing political power based on race, religion, wealth, etc.³³ But if we look to history and democratic theory, democratic inclusion of dead people is taken to cut against the egalitarian core of democracy. When it comes to history, Thomas Jefferson (1789) famously said,

The question whether one generation of men has a right to bind another, seems never to have been started either on this or our side of the water. Yet it is a question of such consequences as not only to merit decision, but place also, among the fundamental principles of every government. The course of reflection in which we are immersed here on the elementary principles of society has presented this question to my mind; & that no such obligation can be so transmitted I think very capable of proof. I set out on this ground, which I suppose to be self-evident, ‘that the earth belongs in usufruct to the living’: that the dead have neither powers nor rights over it.³⁴

And when it comes to democratic theory, Otsuka (2003: 133) says,

The complaint that the American Constitution is a dead hand from the past is a familiar one among some present-day democratic theorists [among which Otsuka includes Dahl 1989]. Their complaint is directed against the entrenchment of these laws against repeal by anything less than a super-majority of the (democratically elected representatives of the) living, where this super-majority greatly exceeds fifty per cent plus one. The dead are able from their graves to thwart the will of a simple majority of the living, which seems an offence to democracy.

And in their recent discussion of the value of democracy, Lovett and Zuehl (2022: 470) similarly point to “the familiar democratic concern with government ‘by the dead hand of the past.’ It is a common and plausible idea that entrenched constitutional constraints are, in some way, anti-democratic.” Thus, when we look to history and democratic theory, democratic inclusion of the dead has been considered anti-democratic.³⁵

Note also that our argument does not require that getting the wrong answer to the issue of dead people has exactly the same paradigmatically inegalitarian implications as getting the wrong answer when it comes to, say, democratic inclusion of women. The implications of a wrong answer to the question of democratic inclusion of dead people may still be inegalitarian enough for such an answer to be implausible. As the above illustrates, this seems to be the case to many.

Thus, we take it to be a considered judgment that dead people should not be included in democratic decision-making. It follows that if all four accounts of being socially related were to imply that dead people should be included in democratic decision-making, this would show that the relational egalitarian view of the value of democracy cannot deliver a plausible solution to the question of who should be included in democratic decision-making. As we will see, analysing what the four views of what it means to be socially related entail for whether dead people and future people should be included in democratic decision-making is sufficient to show that relational egalitarianism suffers from the justice-democracy dilemma; and it is also (almost) sufficient to determine which account relational egalitarians should choose if they want relational egalitarianism as a theory of what makes democracy valuable to deliver a plausible solution to the boundary problem.

4. How May Relational Egalitarians Deal with the Dilemma?

According to relational egalitarians, those who are socially related must relate as equals. This means that relational egalitarians must determine what it means to be socially related. In answering this question, we proposed our quadripartite taxonomy in §2. We then argued that which one we choose makes a difference to who should be included in democratic decision-making, given that this question must be determined by what makes democracy valuable and given that relational egalitarianism is a plausible view on what makes democracy valuable.

In analyzing what these four accounts imply for whether dead people and future people should be included in democratic decision-making, we argued that given the assumption that it is a considered moral judgment that dead people should not be included in democratic decision-making, relational egalitarians are left with the following options: (i) if a plausible solution to the boundary problem must entail inclusion of future people, relational egalitarians should choose the Non-moralized One-way View; (ii) if a plausible solution to the boundary problem must not entail inclusion of future people, relational egalitarians should choose either the Non-moralized Two-way View or the Moralized Two-way View. These accounts have similar implications when it comes to democratic inclusion, but the justice aspect may count as a tiebreaker in which case relational egalitarians should choose the Moralized Two-way View. This investigation interestingly showed that there is a fundamental dilemma for relational egalitarianism: Making relational egalitarianism plausible as an answer to what makes democracy valuable—and thus as a solution to the boundary problem—comes at the price of making relational egalitarianism less plausible as a theory of justice, and vice versa. Or, if you are wary of justice talk, what we have shown is that relational egalitarian complaints against racism and sexism cannot be understood in the same terms as relational egalitarian complaints against non-democracy. There is no unified relational egalitarian theory that makes sense of both complaints.⁴⁸

How, if at all, may relational egalitarians deal with this dilemma?⁴⁹ There seems to be at least five options (some of which we have already briefly exhibited). Relational egalitarians could

• (i) 

  restrict their theory to be solely a theory of justice;

• (ii) 

  restrict their theory to be solely a theory of what makes democracy valuable;

• (iii) 

  propose a different understanding of what it means to be socially related;

• (iv) 

  bite the bullet (for example, by accepting that dead people should be included in the demos or by accepting that relational egalitarianism is compatible with certain racist acts, which, however, might be objectionable on grounds other than those captured by relational egalitarianism); or

• (v) 

  argue that relational egalitarianism is a disjunct of two different ideas of which social relations there ought to exist and where these ideas are premised on being socially related in different senses (i.e., what it takes for individuals to stand in social relations such that relations of justice obtain between them is different from what it takes for individuals to stand in social relations such that they should be bound only by democratically made collective decisions).

As explained, we believe that (i) and (ii) are not appealing. In isolation, relational egalitarianism provides a plausible account of what justice requires and of what makes democracy valuable. This means that we are left with (iii), (iv) and (v). Perhaps relational egalitarians can find another understanding of social relations which avoids the dilemma. We have already seen that a threshold view cannot do the job. Part of the reason why relational egalitarians end up in this trade-off is because the non-moralized understanding of what it means to be socially related involves communication. Perhaps relational egalitarians can find a non-moralized understanding that does not require communication.

One option here might be a subjectedness view according to which those who are subject to the same (state) institutions enforcing the law are relevantly socially related. For instance, if (but not only if) X and Y live in the same state, subject to those state institutions, they are socially related. One problem with this option is that it is not clear that dead people are not subject to state institutions. When Steve Jobs was granted new patents after his death, he may have been said to be subjected to state institutions, at least in some sense (Bengtson 2020). Another problem is that, if we assume this state view of social relations, giving aid to one country instead of another country because of racism towards the latter country’s citizens would not be unjust because the relations would not fall within the scope of relational egalitarianism.⁵⁰ But what if we assumed an understanding of subjectedness which entailed that everyone globally were socially related?

First, it is not clear that everyone globally is subjected to the same institutions enforcing the law in the relevant sense. And, again, it is not clear that Steve Jobs is not subjected in this sense, and so it is not clear that this would avoid the dilemma.⁵¹

Second, even if we set that aside, it has the surprising, and for most relational egalitarians unwelcome, implication that relational egalitarian justice would require global democracy. And, in any case, it would indeed be a surprising upshot if they had to accept global democracy to escape the dilemma.⁵² Instead of another non-moralized view, perhaps relational egalitarians can come up with an alternative moralized understanding of what it means to be socially related. But it seems that a plausible moralized view must at least in some sense include the concerns of relational egalitarians; and then it seems that relational egalitarians cannot avoid the dilemma.

Option (iv) would be to bite the bullet.⁵³ They might say, for example, that while most sexist etc. acts are objectionable from the point of view of relational egalitarianism, some outlier cases, e.g., such as those involved in Sexist Son, are not. However, such acts may be objectionable on grounds other than those captured by relational egalitarianism, in effect committing to a pluralist view of justice, or at least of what we owe to each other.⁵⁴ Alternatively, relational egalitarians could say that although such sexist etc. acts are not unjust, those treating others in a sexist manner can still be blamed for their character, e.g., because their reasoning is flawed. But then that would mean that some instances of sexism etc., which, as explained, relational egalitarians take to be a paradigmatic relational injustice, would not be unjust on relational egalitarian grounds. So that would in essence be for relational egalitarians to, at least partly, give up on a central motivation for their view. And, in any case, if some relational egalitarians had a view along one of the lines suggested in this paragraph in mind, this would then be an important point about the restricted scope of their view which relational egalitarians should have mentioned, i.e., that their view was only meant to capture some cases of racism and sexism.

Perhaps (v) is, after all, the most promising way for relational egalitarians to deal with the justice-democracy dilemma. It would be to say that whereas we might have suspected that relational egalitarianism is one coherent theory, it is a disjunct of two different ideas of how social relations ought to be—one focused on justice, the other focused on democracy—where what it means to be socially related such that one is bounded by these two ideals differs. However, this would conflict with the aim of (at least some) relational egalitarians, for example, Anderson’s account of relational egalitarianism as a matter of democratic equality. Hence, this option does not come as cheap as it might initially seem. But it might still be their best option.

However, if they want to tackle the dilemma, relational egalitarians, going forward, must reflect seriously—when laying out their theory of justice and/or their theory of what makes democracy valuable—on what it means to be socially related. If they do not, we will be unable to judge whether relational egalitarianism in general is plausible as a theory of justice or as a theory of what makes democracy valuable. This is unfortunate given the prominent position of relational egalitarianism in both discussions of justice and discussions of democracy. Perhaps this indication of a future line of research which relational egalitarians ought to embark on is the most important upshot of this article.

Acknowledgments

A previous version of this paper was presented at the Democracy and Intergenerational Justice reading group, Nova University of Lisbon (online). We thank the participants for useful comments. A special thanks to Devon Cass and Andre Santos Campos for the invitation. For helpful written comments on previous versions, we thank Andreas T. Schmidt, several anonymous reviewers and an Ergo area editor. Both authors would like to thank the Danish National Research Foundation (DNRF144) for funding. Andreas would like to thank Independent Research Fund Denmark (1027-00002B) for funding.