Social equality and the conditional justifiability of political inequality

Social egalitarian non-instrumentalism

Let us begin by reviewing SENI. Kolodny (2014: 287) articulates its central thesis: “democracy is a particularly important constituent of a society in which people are related to one another as social equals, as opposed to social inferiors or superiors.” In his terms, democracy is a political procedure that at least meets an Equality Constraint: “if a procedure gives anyone a say, it should give everyone an equal say” (Kolodny, 2014: 291). Giving everyone an equal say means giving everyone an equal opportunity to influence political decisions (Kolodny, 2014: 309–310). Typical examples are the equal right to vote and to run for office. We can also say that a procedure has the property of political equality (of a procedural and distributive kind) to the extent that it comes close to fulfilling the Equality Constraint. ⁵ This characterization of democracy, while hardly sufficient, seems to be a common denominator among social egalitarian democrats; I will follow it in this article. The following three paragraphs reconstruct Kolodny's argument for democracy, as it seems to be the most detailed explanation of why political equality constitutes social equality.

Kolodny's argument for democracy is part of his general theory of the regulation of social hierarchy. This theory explains how hierarchical relationships characterized by asymmetries of power, de facto authority, and consideration can be avoided or moderated. ⁶ Hierarchical relations prevail in modern societies and take root in virtually every individual's life, and they would be relations of objectionable social hierarchy unless certain factors moderate them. The first factor is the standing opportunity to escape hierarchical relations or avoid being subject to the decisions made in them (Kolodny, 2014: 304). The second factor, which is required especially when the first factor is absent, is that the individuals involved in hierarchical relations nevertheless stand in a higher-order, non-hierarchical relation in which they have an equal opportunity to influence decisions to set terms for how to interact in these hierarchical relations (Kolodny, 2014: 305–306). For instance, a relationship between a slave and an owner would look less like an objectionable social hierarchy if either factor were present. Indeed, it would not even look like a slave–owner relationship, but more like an employment relationship. Hierarchies in many other relationships, such as family, workplace, school, or civil society, could be moderated in a similar way.

There is, however, a relationship in which both factors are absent. Let us call it a relationship of political membership, although this is not Kolodny's term. This relation and the decisions that apply to its members (i.e. the political decisions) are virtually inescapable: the cost of escaping a state's jurisdiction is unbearably high. They are also final: this relation is the relation of the highest order within a society, from which every other relation and decision could be regulated, and which could not itself be regulated from the standpoint of a higher relation (Kolodny, 2014: 305–306; González-Ricoy and Queralt, 2018: 619).

There are two interrelated reasons why the relationship of political membership thus characterized should guarantee each member equal influence over political decisions. First, in the absence of the two moderating factors, any asymmetry of influence in this relationship (e.g. unequal votes) would be an unmoderated asymmetry in power, authority, or consideration. Political equality is the only way to eliminate this asymmetry. Second, if the political membership relation is hierarchical, then non-political relations and decisions cannot be regulated from a higher “standpoint of equality” (Kolodny, 2014: 306). Political inequality would undermine the second moderating factor and turn hierarchical and inescapable relations in non-political domains into objectionable social hierarchies. Therefore, objectionable social hierarchies are inevitable when political equality is absent or inadequate.

In short: according to Kolodny, inequality of inescapable and final power, authority, and consideration between individuals renders their relationship objectionably socially hierarchical, where some individuals stand “above” and others “below.” Political inequality is a paradigmatic example of such inequality. This account has three attractive features. First, it plausibly explains why we, pro tanto, and not just prima facie, should pursue political equality as a property of the political procedure. ⁷ Even if there are countervailing reasons against pursuing it, political equality remains valuable, and its loss is at least regrettable. Second, the account also seems to establish that at least an ideal political procedure for social egalitarians must be democratic (Kolodny, 2014: 308).

Third, and most importantly, Kolodny's account meets a moral constraint, often raised by political instrumentalists, on reasons appropriate to justify individuals’ share in political power. As Viehoff (2017: 290) interprets it, the constraint “bars justifications that rest on goods which have among their constituent components one person's ruling over another.” This constraint has motivated the idea that there is no non-instrumental value in anyone's opportunity for political influence as such, which is a share in political power over everyone, including other people (Arneson, 2003: 124–125; Arneson, 2009: 200; Brennan, 2016: 127; Gaus, 1996: 250–251; Mill, 1991: 354). Kolodny's account meets the constraint since it holds that the non-derivative good to which political equality constitutively contributes is a social relation in which no individual rules or stands over another.

Despite these attractive features, it is worth emphasizing that the constitutive relationship between democracy and social equality is not a conceptual entailment relationship. Rather, democracy is structurally necessary to avoid or moderate social hierarchies in modern societies in which individuals are regularly exposed to asymmetric power, authority, and consideration. Under other historical conditions, democracy might be unnecessary to realize social equality (Ziliotti, 2020: 422–423). Perhaps everyone might be willing to forgo, for the right reasons, the exercise of greater power, authority, or differential consideration. Or perhaps everyone might have an adequate opportunity to exit any hierarchical relationship. Although these possibilities do not affect the claim that our contingent historical conditions require political equality, they illustrate an important point: even if the level of political equality is held constant, social equality could be advanced or undermined in other ways. To see this, we need to see how the political procedure can also causally contribute to social equality, which is the main concern of SEI.

The instrumentalist challenge

Social egalitarian instrumentalism (SEI) starts from the premise that the social standing of an individual is not only affected by her relative political power. It is also affected, for example, by how non-political goods and opportunities are distributed, such as civil liberties, income and wealth, and positions of authority and responsibility. ⁸ Social egalitarianism thus has implications for how such goods and opportunities should be distributed to advance social equality to a greater extent (Schemmel, 2011). Those who accept Kolodny's theory of regulating social hierarchies may agree that there should be an adequate or perhaps equal distribution of opportunities to escape hierarchical relationships. This distributional requirement may mean, for example, that individuals should be provided with sufficient goods and opportunities to secure socio-economic independence and avoid poverty, not only in terms of subsistence goods but also in terms of status and agency (Ci, 2013; Scanlon, 2018: ch. 3). Similarly, the ideal of social equality may plausibly imply the wrongness of public racial segregation and other discriminatory policies (cf. Anderson, 2010; Moreau, 2020) and economic systems such as a laissez-faire market system that do not regulate the inequality-producing effects of natural and social contingencies (Scheffler, 2015: 38). These are distributional matters that are independent of political procedures.

However, the implementation of socially egalitarian distributional arrangements usually requires political decisions, that is, laws and other rules that define and regulate institutions such as welfare, the economy, property, education, employment, and marriage. This seems to be true, at least under the normal circumstances of modern societies, in which patterns of behavior that are desirable from a social egalitarian standpoint will not spontaneously emerge. So, the relevant question is: which political procedure tends more to produce political decisions that prevent objectionable social hierarchies from arising in various domains of individuals’ lives? It is an instrumental, epistemic question whether and to what extent alternative political procedures causally contribute to the realization of social equality via the political decisions they produce. ⁹

Taking this thought further, SEI even questions social egalitarians’ commitment to democracy. Arneson (2019: 881) writes: “[t]he various possible regimes would likely score higher along some but not all dimensions, so a weighing is needed to establish which regime, all things considered, scores higher” on a given standard of social equality. A regime or procedure that is all-things-considered justified may not necessarily be democratic. Capturing this idea, what I call the instrumentalist challenge says that if, under some circumstances, the epistemic value of some non-democratic procedure is great enough to offset the damage its political inequality does to social equality, then perhaps the non-democratic procedure should be implemented to advance social equality more effectively. Taken to an extreme, one might say that even a benevolent autocracy could in principle be justified; its proponents would only have to take pains to demonstrate that it would “score higher” on a given standard of social equality.

The instrumentalist challenge and the underlying SEI framework are worthy of detailed discussion. This is because they press us to revise our understanding of the relationship between social equality and epistocracy, suggesting the latter's potential justifiability from the social egalitarian standpoint. Epistocracy is a political procedure characterized by the de jure unequal distribution of political power, sensitive to the unequal measurable competence of individuals. ¹⁰ Some such procedures are advocated in the literature (e.g. Bell, 2015; Brennan, 2016; Jones, 2020) from the perspective of political instrumentalism, according to which the political procedure should be evaluated primarily as a causally efficacious means of achieving moral ends such as the fulfillment of fundamental rights. ¹¹ If instrumentalism is described in such general terms, without particular regard to the value of social equality, then social egalitarians might simply respond by invoking the standard view that even if epistocracy instrumentally outperforms any democratic alternative, its compromise of political equality is still incompatible with the moral requirement to secure everyone's equal social standing (cf. Peña-Rangel, 2022: 142). However, SEI may have the potential to offer a particular challenge that casts doubt on this standard view—the challenge that the very value of social equality may be best realized through epistocracy rather than democracy. At least, Arneson (2019: 880–3) and Wall (2007: 426) suggest such a challenge. It must be problematic for social egalitarians if this specific epistocratic doubt about their standard view remains unaddressed.

This article aims to explore the best response from the perspective of SENI to the instrumentalist challenge so conceived. That is, I ask whether social egalitarians’ non-instrumental commitment to democracy can be sustained in the face of this challenge. In doing so, I stipulate, for the sake of argument, that some epistocratic institutions might significantly outperform democracy in causally advancing social equality, at least under certain circumstances. Of course, this stipulation is not to deny the importance of asking whether there is any realistic circumstance in which epistocracy is reasonably expected to outperform democracy in causally advancing social equality. One might well try to defend democracy over epistocracy within the SEI framework, building on the instrumental or epistemic defenses of democracy in the literature (e.g. Estlund, 2008; Goodin and Spiekermann 2018; Ingham and Wiens, 2021; Kogelmann, 2023; Landemore, 2013; Talisse, 2022; Vandamme, 2020). Yet, there is also an increasing number of scholars who argue for the instrumental superiority of epistocracy over democracy (e.g. Bell, 2015; Brennan, 2016; Jones, 2020) and those who defend some forms of epistocracy against critiques (e.g. Brennan, 2022; Lippert-Rasmussen, 2012). It is worth asking what if the advocates of epistocracy are right even on social egalitarian standards; if they are, then does there remain no sense in which we ought to pursue democracy, as far as our concern is with social equality? Let me turn to this question in the next section.

Restrictive social egalitarian non-instrumentalism

Restrictive SENI gives an uncompromising answer to the instrumentalist challenge by advocating the requirement of Unconditional Democracy. However, as I argued in §3, Unconditional Democracy does not immediately follow from the Procedural Principle; a further argument is needed to support it. I will examine two anticipated arguments and contend that both are problematic.

First, one might argue that political equality enjoys categorical priority over any other feature of political procedures, even the tendency to produce decisions that guard against objectionable social hierarchies in non-political domains. The idea is that political equality is not a mere pro tanto procedural value that can be squarely compared with the other values but rather a constraint. ¹⁹ According to this view, Permissive SENI is wrong because it mistakenly takes a teleological form by measuring the various ways in which political procedures contribute to social equality on a single scale. Some theorists seem to raise such a challenge when they invoke a relational conception of normativity to explain the value of democracy. For instance, Rostbøll (2015: 269) maintains that democracy “establishes the right way for citizens to relate to each other” rather than “promotes some good.”

In one way, however, Permissive SENI does accommodate the idea that those committed to the value of social equality must accept a non-teleological constraint against political inequality. As I highlighted in §3, Permissive SENI holds that any political procedure that does not satisfy the Procedural Principle is objectionable from the social egalitarian standpoint. It thus maintains that non-democratic procedures are objectionable without exception, even if they are justified to achieve greater social equality. It only resorts to the Sub-Full Max Principle to guide our choice among procedures that all count as objectionable; and, as I said at the end of §3, even if a non-democratic procedure is justifiably implemented by appealing to that principle, people or society, while implementing that procedure, still ought to seek ways to live up to the Procedural Principle. This theoretical structure does not make Permissive SENI teleological even if it allows for the conditional justifiability of non-democratic procedures. ²⁰

Admittedly, if Permissive SENI states that some institutions that are objectionable from the social egalitarian standpoint can still be justifiably implemented, one might wonder how permissive the view is. For instance, in a society where a group, Smug, dominates another group, Meek, and both groups internalize the view that Smug is superior to Meek, is it permissible for the state to conduct a contemptuous publicity campaign against Smug if it is likely to boost Meek's self-confidence and thus improve the way Smug relates to Meek (Peña-Rangel, 2022: 149)? For another example, if the state could help limit the outsized political and economic power of wealthy business executives by kidnapping them and restricting their freedom of movement and speech, would it be permissible for the state to do so? One might think that, for expressive or other deontological reasons, there should be unconditionally prohibitive constraints against such policies and that Restrictive SENI's constraint against political inequality is one among them.

In reply: Permissive SENI does not set up an unconditional prohibition against such policies; yet, this view is compatible with introducing some cautious criteria for the justifiability of such extreme policies. Although constructing such criteria is beyond the purpose of this article, let me briefly illustrate. One attractive criterion, for instance, is whether the effects of the policies are tractable: can the policies be adequately focused on just equalizing Smug's and Meek's confidence or on just limiting the business executives’ political and economic power? On the one hand, if the policies also have other harmful effects, such as inducing violence against Smug or limiting the business executives’ engagement in various valuable and innocuous activities and relationships outside the political and economic domains, then the negative impacts of the policies on Smug's or the executives’ social standing will tend to be intractable. It will often be advisable for those choosing between alternative policies to search for more proportionate ways to achieve social equality (assuming that there are some better ways than contemptuous publicity campaigns or kidnapping). On the other hand, if the policies are well-focused on equalizing people's social standing and their effects are tractable, then they may be justified; they may even express the state's admirable intention to pursue an egalitarian society. It is worth stressing again, though, that according to Permissive SENI, such policies as illustrated here may not cease to be objectionable as deviations from full social equality.

Let me turn to a second anticipated argument to connect Unconditional Democracy and the Procedural Principle. Some advocates of Restrictive SENI might argue that Permissive SENI fails to take seriously the idea that political equality constitutes social equality. On a strict reading of the term “constitute,” A constitutes B if and only if B would be unintelligible without A. Therefore, if political equality constitutes social equality, then social equality without political equality is unintelligible. This reading of the constitutive relation seems expressed in Beerbohm's (2012: 39) analogy in his explanation of democracy's value: “[i]f you betrayed a friend to make several new ones, you would be affronting the value at stake.” On this reading, Permissive SENI could appear incoherent, since it claims that political equality constitutes social equality but also that the latter is intelligible and more or less realizable even without political equality.

However, if Restrictive SENI relied on this strict reading of “constitute,” it would fail to evaluate the existing democracies’ standard practices adequately. For instance, consider how social egalitarian theorists often approach the issue of children's disenfranchisement. Is it compatible with full social equality? A possible line of argument for compatibility might hold that since everyone ages, children and adults can at least enjoy an equal diachronic opportunity for political influence and that it suffices for their full social equality. A problem with this argument, however, is that some cases of synchronic inequality appear problematic even when diachronic equality is guaranteed. For example, it seems problematic for society to permit parents to impose corporal punishment or child labor on children even if those children, after growing up, will enjoy the same permission. One (if not the only) significant element of the problem seems to be what Bidadanure (2021: 100) calls synchronic relational inequality between generations, which is not fully addressed by merely guaranteeing diachronic equality. From this perspective, Bidadanure (2021: 110, 224) suggests that children's disenfranchisement can also be criticized for creating a social inequality between them and adults, even if all children are given the right to vote when they come of age.

However, though the synchronic equality view seems on the right track, it nevertheless leaves an important ambiguity. ²¹ On one interpretation, it suggests that the disenfranchisement of children is one of the factors that pro tanto establish their social inferiority to adults, factors that would no longer exist if full social equality were realized. On the other interpretation, we cannot even intelligibly speak of more or less social equality between children and adults if children are disenfranchised, for there could be no such relationship. The difference between these interpretations bears on the difference between Permissive and Restrictive SENI. Permissive SENI needs only the first interpretation. In contrast, if Restrictive SENI endorses the strict reading of the term “constitute,” it has to accept the second.

The second interpretation, however, is implausible. Although children have regularly been disenfranchised in modern democratic societies, we can intelligibly speak of the improvements in their social standing in non-political domains. Children have gained improved access to education, legal protection from corporal punishment, and the like, which they could not expect when their parents treated them as labor forces or regarded them as objects of unconstrained domestic discipline. In many societies, the enfranchised adults have been competent enough to address these problems and achieve some significant improvements in children's social standing vis-à-vis adults. However, if Restrictive SENI endorses the strict reading of “constitute,” it fails to acknowledge this relative achievement.

Thus, I reject the strict reading of the term “constitute.” On an alternative reading I endorse, A constitutes B only if B's full realization requires A (at least under the prevailing socio-historical circumstances). Permissive SENI endorses this reading and holds that children's relationship with adults can be more or less socially equal, though not fully socially equal, without universal and equal enfranchisement. The reason full social equality is not realized, as we saw in §2.1, is that an asymmetry in inescapable and final power, authority, and consideration could not be moderated in the way that the other asymmetries could. The asymmetry between disenfranchised children and enfranchised adults is of the inescapable and final kind. My point against Restrictive SENI is that this argument for political equality is compatible with the claim that disenfranchised children can still have more or less equal social relations with adults.

To be sure, some versions of Restrictive SENI might not endorse the strict reading of “constitute,” in which case they might avoid my last objection. However, if so, and the other anticipated arguments for Unconditional Democracy are also problematic, as I have argued, then it is unclear what argument remains for Restrictive over Permissive SENI.

Pure social egalitarian instrumentalism

Permissive SENI must address one final concern: why should political equality be necessary for full social equality? In answering this question, I have thus far followed Kolodny's account summarized in §2.1. To recapitulate, very roughly:

Full social equality requires that there be no asymmetry of inescapable and final power, authority, or consideration (call this condition the Necessity of Final Equality for full social equality.)

However, there might be a pure instrumentalist objection to this view, raised from a social egalitarian standpoint. The objection says that the only necessary condition for a political procedure to constitute full social equality is that the procedure is instrumentally optimal in advancing social equality. This objection, in effect, rejects the first premise in the argument described in the last paragraph and implies that even the ideal political procedure for social egalitarians need not establish political equality. This rejects the Procedural Principle, the central normative principle of Permissive SENI. In this subsection, I argue that we need not abandon Permissive SENI in the face of this objection from Pure SEI by critically examining Viehoff (2019) as an example of the latter position.

Viehoff argues that social equality does not require political equality. It only requires that for every status-based difference in the advantages and disadvantages of individuals, there is a social norm that justifies the difference “by appeal to the interests of everyone around here, where all of these interests are treated as equally significant” (Viehoff, 2019: 17; see also Wall, 2007: 426). The idea is that no one's interest should be treated as fundamentally more important than anyone else's. Call this condition the Sufficiency of Egalitarian Justification for full social equality. ²² Now, political inequality is one form of status-based difference in advantages and disadvantages. For instance, the opportunity for political influence is unequally distributed between those with the “epistocratic voter” status and those without it. Viehoff, applying the Sufficiency of Egalitarian Justification, argues that some social norms can instrumentally justify political inequality without insisting that the interests of epistocratic voters are fundamentally more important than those of non-voters (Viehoff, 2019: 20). ²³ If such an instrumental justification is provided, “then political inequality is not even regrettable insofar as our concern is solely with social status hierarchy” (Viehoff, 2019: 24, my emphasis). ²⁴ Notice that this view differs from Permissive SENI in an important respect. Permissive SENI holds that political equality is not just prima facie valuable as a “baseline” (Viehoff, 2019: 23) but it is also pro tanto valuable; as political equality remains valuable even if outweighed, political inequality is regrettable even if justified, all things considered.

Viehoff attempts to support his position with a thought experiment he calls Necessary Representation: Someone from an egalitarian nomadic tribe happens to meet someone from another egalitarian nomadic tribe and talks with her about the long-standing dispute between the two tribes over their shared hunting fields. Although it is imperative to avoid open conflict, and this serves all tribal members’ interest in peace equally (and, we might add, prevents the emergence of objectionable social hierarchies typical of militarized communities), it has been difficult for the nomadic tribes to meet and negotiate. So, these accidental “representatives,” though unauthorized by the other members, take the opportunity to negotiate rules and bring home the outcomes. To avoid conflict, the tribes recognize the rules as binding on their members. Viehoff (2019: 20) suggests that in this case, the accidental representatives do not seem to stand as socially superior to the other members despite the greater political power they have exercised. The lesson from this thought experiment is that the Sufficiency of Egalitarian Justification adequately captures the condition for full social equality.

My aim below is to cast doubt on the Sufficiency of Egalitarian Justification for full social equality and support the Necessity of Final Equality. Let me first clarify the argumentative burden. What is the hypothetical case of Necessary Representation supposed to accomplish? For Viehoff's view to be vindicated, the case must not only convince us that some cases of political inequality can be justified, all things considered (the point Permissive SENI accommodates); it must also convince us that such instances of political inequality are beyond reasonable regret for social egalitarians. Only then can the case serve to establish Pure SEI's point that political equality does not constitute even full social equality. However, if political inequality can be reasonably regrettable for social egalitarians, the straightforward explanation is that it compromises, at least in part, something that social egalitarians wish to achieve, namely, full social equality.

With this understanding of argumentative burden in mind, I raise two problems with the Sufficiency of Egalitarian Justification. One problem is that it can be satisfied too easily. Consider Lovett's example (2021: 180–181): ordinary citizens of China often think that the superior political power of Communist Party members is instrumentally justified as a means of advancing the economic development or interests of China's working people, rather than as a premium for recognizing the superior intrinsic worth of Party members. It seems strange, however, to conclude that the superior political power of Party members is beyond reasonable regret for social egalitarians. We can also imagine other cases, such as justifying some form of epistocracy as the most effective political procedure for satisfying a utilitarian, libertarian, luck egalitarian, or other conception of distributive justice, all of which regard the interests or claims of all people as in some sense equally important. However, even if we granted arguendo that some such procedures were indeed societally justified for this reason, we could still question whether the society implementing the procedure would count as an element of “an egalitarian political community” (Scheffler, 2003: 35–36). The Sufficiency of Egalitarian Justification then seems dubious.

Another problem is that some egalitarian-instrumental justifications of political inequality might be successful just because of some regrettable circumstances. Consider Necessary Representation again. Even if the accidental representatives are justified, all things considered, to exercise greater political power, some may reasonably find the case regrettable. The reason, I submit, is that the justifiability of political inequality here seems to be owed merely to a circumstance of extreme emergency, namely the imminence of an outbreak of conflict. I think it is appropriate for agents involved in an emergency circumstance to regret the circumstance itself if and because it prevents them from achieving some value of their concern. In the present case, it seems appropriate for the accidental representatives to wish that they had had a few weeks before the conflict was expected to erupt, and that instead of making hunting rules themselves, they would have had enough time to schedule an intertribal meeting in which every member would have enjoyed a democratic say. ²⁵

Why does regret seem reasonable here? Not because the democratic intertribal meeting would be instrumentally better; if the accidental representatives themselves decide on hunting rules, the tribes can avoid open conflict and the resulting social evils anyway. Instead, an explanation that invokes the Necessity of Final Equality seems plausible. Because hunting rules have profound and inescapable effects on various aspects of each member's life, a member with a robust social-egalitarian commitment against ruling over the other members would be discontent with exercising greater power in setting the rules, even if exercising that power is instrumentally the best course of action under the emergency circumstance.

To be sure, some might reject this intuitive sense of regret. However, as I clarified above, it is the argumentative burden of Pure SEI to show, with Necessary Representation, that the cases of political inequality that are societally justified on egalitarian-instrumental grounds are beyond reasonable regret for social egalitarians. My discussion above casts doubts on whether Pure SEI can meet this burden. Thus, Pure SEI's critical point against Permissive SENI that democracy or political equality is not even a constitutive element of full social equality does not seem to be adequately supported.

Finally, some might object that even if regret is appropriate in Necessary Representation, it can be explained without invoking the value of social equality. Pure SEI might explain, for example, that political inequality in Necessary Representation is regrettable not because it undermines social equality but because it undermines some other values, such as the autonomous self-determination of tribal members. If so, Pure SEI can reject the Necessity of Final Equality without denying that Necessary Representation has a regrettable aspect. In response, I concede that Pure SEI can invoke a value that has nothing to do with social equality to explain the appropriateness of regret. However, the crucial question is whether the concern for social equality should not also be invoked as part of the explanation. It seems that Pure SEI has not presented a compelling reason to reject the explanation that invokes the Necessity of Final Equality. I have suggested above that because of their commitment to social equality, the accidental representatives may regret having to rule over the other members by exercising greater political power. Unless this explanation is demonstrated to be unreasonable, the present objection does not show that Pure SEI is explanatorily superior to Permissive SENI.