2.1. Asymmetries of power

First, many social egalitarians think that asymmetries of power constitute social inequality (Umbers, Reference Umbers2021: 316; Peña-Rangel, Reference Peña-Rangel2022: 146–147; Kolodny, Reference Kolodny2023: 91, 93). This claim should be qualified, but let me first describe the general idea. One’s power over another person is one’s agential capacity to effect outcomes (such as specific states, dispositions, or actions) in that person, in conformity with one’s will or interests, by way of one’s intentional actions (including both performing and forbearing).Footnote ² Where there is an asymmetry of power, either person A has greater power over B and B lacks reciprocal power over A, or A has greater power over others than B has over them. Consider an example. In feudal societies, the lord of the manor has power over peasants, such as the capacity to deprive them of their property or to inflict physical violence on them, in conformity with the lord’s interests or will. The lord exercises this power by, for instance, commanding his subordinates (e.g., armed soldiers) to carry out his preferred course of deprivation or infliction. Suppose a peasant does not have symmetrical power over the lord. In this case, there is an asymmetry of power between the lord and the peasant. According to the view considered here, this fact can constitute the lord’s social superiority to the peasant.

More generally, and with qualifications explained shortly, the claim is this: within an ongoing social relation, if person A has greater power in comparison with person B, that fact presumptively constitutes A’s objectionable social superiority over B, unless that power asymmetry is ‘tempered’ in some appropriate way. The claim is not that ‘A’s greater power in comparison with B constitutes A’s objectionable social superiority to B, period’. That would commit us to an implausibly sweeping problematization of most social relations. Asymmetrical power is everywhere: differences in brute physical strength or financial capacity between spouses and other intimate partners, ‘chain of command’ relations within private firms and bureaucratic institutions, and asymmetrical persuasive abilities among participants in public discourse. Intuitively, not all such relations are objectionable, and some do not even look like ‘social inequality’ in the first place. A pressing task for social egalitarians, then, is to delineate which cases of asymmetrical power count as ‘objectionable social inequality’ and to explain that delineation.

So, the important part of the above claim is that asymmetrical power presumptively constitutes social inequality unless it is tempered. To temper it is to eliminate or mitigate its inferiority-making implication. Niko Kolodny’s account of social inequality (or ‘relations of inferiority’ in his terms) exemplifies this structure. He suggests various ‘tempering factors’ that, if in place, would render an asymmetrical power relation not an instance of social inequality, or else a less objectionable or unobjectionable one (Kolodny, Reference Kolodny2023: chaps. 5 and 8).

Examining all the tempering factors suggested by Kolodny is beyond the scope of this paper. Here, let me review two factors that are particularly relevant to our purpose. The first factor is ‘Escapability’ (Kolodny, Reference Kolodny2023: 100). For instance, although a slave-owner might enjoy greater power over a slave, the slave can still stand as socially equal with the owner if they enjoy an ongoing opportunity to exit the relation. Such escapability may even render the label ‘slavery’ inapplicable. More generally, the easier it is for someone standing in a locally subordinate position to exit the relation at will, the less they seem to stand in a relation of social inferiority. The second factor is ‘Downward Equalization’ (Kolodny, Reference Kolodny2023: 100–101). Asymmetries of power need not constitute objectionable social inequality if they are regulated by higher-order decisions, in the making of which the relevant parties have an equal say. For instance, corporate subordinates may be locally subordinated to their bosses, who regularly possess and wield greater power over them (in task-assignment, performance evaluation, personnel matters, and so on). Yet that relation may not be objectionable, provided that both parties have an effective equal say over the higher-order decisions, such as contracts and labor laws, that determine the terms of their interaction.

Thus, given the appropriate tempering factors, some asymmetrical power relations are unobjectionable, and others are not social inequality at all. However, as Kolodny argues, such tempering manoeuvres are unavailable for some special relations, notably asymmetries of political power. Let political power refer to the power to determine what political decisions (e.g., laws) are to be imposed upon members of society; voting power in referenda is a typical example. The difficulty of tempering asymmetrical political power stems from special features of political decisions (cf. González-Ricoy and Queralt, Reference González-Ricoy and Queralt2018: 618–619; Kolodny, Reference Kolodny2023: 122–123). First, from the perspective of each member of society, the subjection to political decisions is virtually inescapable; the cost of exiting their reach (i.e., the state’s jurisdiction) would be unbearably high. Second, political decisions are final; they can regulate virtually any non-political relation or decision and are not themselves regulated by higher-order decisions. Both Escapability and Downward Equalization fail to obtain. So, any asymmetry in people’s opportunities to influence political decisions may particularly undermine the egalitarian quality of their social relations: to the extent person A enjoys greater influence over political decisions than person B, B is objectionably positioned as socially inferior to A.

This account has been presented as a non-instrumental reason for democratic political procedure, broadly construed: if anyone has influence over political decisions, everyone else should be guaranteed equal influence (which is compatible not only with universal suffrage but also with decision-making by lot, in which no one has positive influence over selection). Social equality demands equal political power because its absence constitutes objectionable social inequality (Kolodny, Reference Kolodny2023: 137–138, 323–325).

2.2. Disparities of consideration

Social egalitarians are concerned not only with asymmetrical power but also with patterned inegalitarian attitudes among members of society. Among such attitudes, I will focus on consideration for persons, which Kolodny (Reference Kolodny2014: 297) roughly characterizes as ‘those responses that … we feel, either every person is owed equally simply in virtue of being a person, or are problematic for anyone to give anyone else’ (see also Cox, Reference Cox2022: 100–102; Zuehl, Reference Zuehl2024: 573, 579–580).Footnote ³ It encompasses a variety of respectful attitudes or expressive behaviours, such as ‘evaluation’ or ‘valuing’ of the worth of a person (Van Wietmarschen, Reference Van Wietmarschen2022: 923–927), judgements about the ‘fundamental moral importance’ of a person’s interests (Viehoff, Reference Viehoff, Sobel, Vallentyne and Wall2019: 18), or simply the willingness to serve a person’s interests.Footnote ⁴ Specific forms of behaviour expressing such attitudes also matter, such as ‘bowing and scraping, fawning and toadying’ (Walzer, Reference Walzer1983: xiii; Cox, Reference Cox2022: 100), or ‘bending the knee [and] kissing the ring’ (Zuehl, Reference Zuehl2024: 579). The idea, then, is that disparities of such kinds of consideration constitute social inequality: within an ongoing social relation, when A receives such favourable attitudes and behaviours in a patterned way from B or others more than B does from A or others, A thereby stands as socially superior to B. Indeed, Ryan Cox contends that finding oneself receiving lesser consideration is the core of our ‘experience of subordination, or inferiority’ and that such experience is central to the very phenomenon of social inequality (Cox, Reference Cox2022: 101, original emphasis).

To be sure, disparities of consideration are ubiquitous, so once again social egalitarians must explain what factor marks out instances of unequal consideration that instantiate objectionable social inequality. Social egalitarians can answer this question by distinguishing between what Kolodny (Reference Kolodny2023: 104–105) calls ‘merited’ and ‘unmerited’ disparities of consideration. Merited disparities are based on reasonably accurate assessments of the relevant individuals’ qualities, achievements, or situations that are relevant to the weight of their interests and claims within a certain context. Unmerited disparities, by contrast, lack such assessments. In less pronounced instances of unmerited disparities, those who pay greater consideration to person A over others might do so ‘out of pure self-interest, simply to curry [A’s] favor’ or even as ‘unthinking habit’ (Kolodny, Reference Kolodny2023: 104–105). In more pronounced instances, those who pay greater consideration to A over others might do so because they believe A has a certain trait that makes A fundamentally morally superior to others. Viehoff (Reference Viehoff, Sobel, Vallentyne and Wall2019: 17) illustrates this by comparing two cases: a medical doctor has the capacity to make other drivers on the road scramble out of the way (e.g., by signalling a special flashlight attached to the doctor’s car) because this practice is generally believed to enhance the efficiency of essential medical service; in contrast, a manor lord has the capacity to do the same, but because this practice is societally believed to confirm the fundamental moral superiority of the lord over others. The doctor receives greater consideration in a merited way; the lord in an unmerited way.

2.3. Power and consideration interact

Asymmetries of power and disparities of consideration often interact to give rise to social inequality. On the one hand, power asymmetries may causally depend on, or be enhanced by, disparities of consideration. Consider, again, the greater power of the lord over the peasants. The lord often does not literally wield his own physical force to deprive the peasants of their property or to inflict physical violence on them. The lord’s power quite often depends on his tendency to be obeyed or complied with when he issues commands to his armed subordinates. Such a tendency is called ‘de facto authority’ (Kolodny, Reference Kolodny2023: 91–93; see also Anderson, Reference Anderson and Estlund2012: 42–43), and Zuehl insightfully observes that it is also a subspecies of consideration for persons, which he calls ‘practical deference’ (Zuehl, Reference Zuehl2024: 578). That is, the armed subordinates, upon hearing the action-guiding commands issued by the lord, practically defer to the lord in concert, carrying out the lord’s preferred course of deprivation or infliction. They thereby make the lord powerful vis-à-vis the peasants. This is one of the ways in which greater consideration often serves as a resource for greater power.

On the other hand, greater power may tend to bring about societal patterns of consideration, where those with greater power tend to be regarded as more excellent, more deserving of practical deference or obedience, worthy of receiving greater respect or consideration for their interests, or otherwise treated in ways that express these attitudes, compared with those with lesser power. Cox’s ‘extraction theory’ of the bad of asymmetrical power articulates this idea: power asymmetries ‘are problematic … because, in certain circumstances, they can be used to extract greater consideration for those with greater power’ (Cox, Reference Cox2022: 99–100, original emphases). To the extent that person A has greater power over person B, or greater power over others than B has over them, A will be able to extract or attract greater consideration from B or others, thereby enjoying superior social standing.

This understanding of the relationship between asymmetries of power and disparities of consideration can be applied to political inequality. Consider horizontal inequality of political power, such as unequal votes. There are two ways such inequality is associated with objectionable social inequality. First, unequal political power may itself reflect or express some unmerited disparities of consideration. The racially biased implementation of the U.S. voter literacy tests in the Jim-Crow era illustrates the point: the effective disenfranchisement of African Americans reflected the general, categorical disregard for their interests in comparison with the interests of Whites (which can be seen as an instance of ‘hierarchy of standing’, in the terms of Anderson, Reference Anderson and Estlund2012: 43). Second, as expected by the extraction theory, unequal political power will often give rise to unmerited disparities of consideration. It does so by providing incentives for others to attend more favourably to the will and interests of those with greater political power than to those of others. Sean Ingham provides a relevant illustration. When voter A enjoys greater political power than voter B (such as extra votes), and if such inequality is publicly known to exist between other A-like voters and B-like voters as well, then the inequality ‘would be expected to shape political parties’ incentives to give consideration to the preferences of’ the A-like voters, rather than those of the B-like voters (Ingham, Reference Ingham2022: 699). Thus, on this account, unequal political power causes social inequality by bringing about disparities of consideration. This instance of social inequality is objectionable if the disparities so caused are unmerited – for instance, if the A-like voters lack impartial regard for the public interest, tend to favour their own partial interests, or are willing to impose harmful laws and policies on B-like voters.Footnote ⁵

2.4. Taking stock

I have reviewed two potential dimensions of social relations that, unless tempered, can constitute social inequality: asymmetries of power and disparities of consideration. The general idea of social egalitarianism that each person has a claim against being involved in relations of social inequality can be specified with the following two moral claims:

• • Claim against Lesser Power: one has a claim, derived from the concern for equal social standing, that one be not subjected to anyone else’s greater inescapable and final power over one.

• • Claim against Lesser Consideration: one has a claim, derived from the concern for equal social standing, that one be not involved in unmerited disparities of consideration, or subjected to practices that reflect such disparities, in which one receives lesser consideration than someone else.

These claims will impose practical demands and constraints on how the state should design and implement its basic social institutions, including the political procedure. As Anderson (Reference Anderson1999: 289), Kolodny (Reference Kolodny2023: 101–102), and Wodak (Reference Wodak2024: 386) stress, the concern for social equality has a deontic point: to impose demands on agents (the state, in particular) to temper relations of social inequality, albeit within the practicable range, and avoid creating new such relations. The Claims against Lesser Power and Consideration specify what those demands are; these claims will inform our subsequent comparisons between electoral democracy and lottocracy.

Before moving on, let me make a methodological reservation. I have reviewed two distinct thoughts about how political and social inequalities may be related: unequal political power may directly constitute social inequality (§2.1), or it may indirectly give rise to social inequality by causing or expressing disparities of consideration (§2.3). Determining which is the precise relation is beyond this paper’s scope. To proceed on a sufficiently weak assumption, I will regard both the constitutive and the causal claims as potential perspectives on how political inequality might be an affront to social equality (as suggested, for instance, by Peña-Rangel, Reference Peña-Rangel2022: 145). To be sure, especially the first thought is criticized in the recent literature. Some argue that disparities of consideration are the conceptual core of social inequality and that unequal political power is only causally related to them (Cox, Reference Cox2022; Van Wietmarschen, Reference Van Wietmarschen2022; Zuehl, Reference Zuehl2024). Defending the constitutive claim against their objections is a topic for another paper. If no defence is forthcoming, my subsequent arguments, where they depend on the constitutive claim, may be less forceful than I expect them to be.Footnote ⁶

From the next section, I will explore the implications of the accounts just reviewed for the comparative justification between lottocracy and electoral democracy. I will start with the basic problem both systems encounter: the vertical inequality of political power that inheres in any representative political system (§3). I will then discuss more specific problems of biased composition of officials and argue that lottocracy is prima facie more effective in addressing them (§4).

4.1. Biased composition and social inequality

Let us begin with the diagnostic point. Given our discussion in Section 2, whether the biased composition of officials gives rise to objectionable social inequality depends on whether it violates some individuals’ claims against potential constituents of social inequality, namely unequal inescapable and final power and unmerited disparities of consideration. Among these constituents, I first discuss how the biased composition of officials (or the state’s allowing it to continue) matters in terms of disparities of consideration, violating individuals’ Claim against Lesser Consideration. I argue that the biased composition of officials violates that claim by reflecting and expressing disparities of consideration, and that this is particularly problematic for electoral-democratic settings in two ways. The first, narrower problem is biased voting. The other, more general problem is structural inequality of opportunities for entering electoral competition.

First, election involves occasions for voter biases to be reflected in the selection of officials. Free voting allows each voter’s choice to be arbitrary (Ingham, Reference Ingham2022: 694): it does not restrict the reasons on which voters form preferences and cast their votes. In electoral processes, such arbitrariness affects the selection of officials. By allowing voters to form preferences for candidates and vote arbitrarily, election ‘intrinsically permit[s]’ voter discrimination against candidates based on their demographic characteristics (Abizadeh, Reference Abizadeh2021b: 797). Of course, empirical findings are mixed about whether and how often voters discriminate against candidates, and the results depend on the specific demographic characteristics studied. Some studies deny that voters discriminate outright against female candidates (Dolan, Reference Dolan2014)Footnote ¹² or candidates from the working class (Carnes and Lupu, Reference Carnes and Lupu2016). (This is why I said the problem of biased voting is ‘narrower’ than structurally unequal opportunities for electoral competition, which I discuss below.) Nevertheless, other studies show subtler forms of biased voting, for instance penalties against female candidates who are unmarried and without children (Teele et al., Reference Teele, Kalla and Rosenbluth2018), against lesbian, gay, and transgender candidates (Magni and Reynolds, Reference Magni and Reynolds2021), against candidates from minority ethnic and religious groups (Martin and Blinder, Reference Martin and Blinder2021), and so on. So, voter biases do seem to affect the relative electability of candidates to some degree, depending on their demographic characteristics.

Second, and more generally, the biased composition of officials may itself express the state’s disparities of consideration as well. This is because, by implementing formal electoral processes, the state allows electoral campaign competitions to occur, in which socioeconomic advantages and disadvantages of candidates are converted into inequalities in the prospects for electoral success. One typical example is the relative lack of capacity among working-class individuals to bear their financial and networking burdens in electoral competitions. Candidates with more advantaged economic standing and backgrounds (e.g., greater wealth, more affluent family of origin, and more prestigious occupational history) tend to run and re-run more frequently and successfully (Carnes and Lupu, Reference Carnes and Lupu2023; Carnes et al., Reference Carnes, Ferrer, Golden, Lillywhite, Lupu and Nazrullaeva2025). For another example, women tend to face thicker structural barriers, stemming from the gendered inequalities of opportunity in various socioeconomic contexts, that impede their effective entry into electoral competition. As Thomsen and King (Reference Thomsen and King2020) argue, women tend to lack access to ‘pipelines to power,’ such as lower-level public offices in municipalities, legal professions (e.g., lawyers and court judges), and media coverage (e.g., being named in newspapers as a promising candidate for the upper house). These positions and careers are ‘pipelines’ that often form pools of potential candidates for national-level representative offices, but because these pipelines tend to be male-dominated, women as ambitious as men in seeking political office often fail to find ways to enter electoral competition. Similar background inequalities of opportunity will also negatively affect the electoral prospects of racial/ethnic minorities, sexual minorities, religious minorities, and other disadvantaged social groups.

Does the Claim against Lesser Power also play a diagnostic role here? Some might doubt it, suggesting that unequal political power is diagnostically irrelevant in this context for two reasons. First, as we saw (§3), representative systems as such establish inescapable and final political inequality, which already seems to violate non-officials’ Claim against Lesser Power (absent rather idealistic, egalitarian popular control). It is then unclear how that claim shows the biased composition of officials to be any more objectionable for social egalitarians than a representative system as such. Second, informal inequalities of political influence, such as the differential rent-seeking capacities of individuals or corporate entities relative to their financial or organizational power, may violate the Claim against Lesser Power of those lacking such power; yet this problem would haunt both electoral democracy and lottocracy (Kolodny, Reference Kolodny2023: 383–384). Although there have been attempts to compare election and sortition in terms of their effectiveness in alleviating political capture (Guerrero, Reference Guerrero2024: 62–64, 287–288; see also Abizadeh, Reference Abizadeh2021b: 797–798; Hutton Ferris, Reference Hutton Ferris2025: 388–391; Ingham et al., Reference Ingham, Landa and Pevnick2025: 639–640), at present we seem to lack robust formal, experimental, or empirical base to settle the dispute.

I suggest, however, that if we combine the Claims against Lesser Power and Consideration, we can provide a distinctive explanation of why the biased composition of officials in electoral democracy is particularly problematic for social equality. Granted, under both electoral democracy and lottocracy, officials hold positions that give them greater political power to influence the laws and policies binding on citizens at large. Hence, officials might potentially stand as socially superior to non-officials. However, I think it also matters for social equality which factors determine individuals’ relative prospects of obtaining positions that potentially make their occupants socially superior. Voting behaviours that express voters’ discriminatory or otherwise biased views of candidates based on their demographic characteristics, and background inequalities of opportunity between social groups, both amount to disparities of consideration that eventually affect which demographic characteristics help one obtain greater political power, where that power potentially makes officials socially superior to non-officials. Thus, when the selection of officials is biased, individuals who bear disadvantaged demographic characteristics tend to have their Claim against Lesser Power violated (due to their lower electability and greater tendency to be subordinated to officials) because they tend to have their Claim against Lesser Consideration violated (due to having their demographic characteristics tracked for lesser consideration). It is one thing that someone inevitably occupies more politically powerful, and thus potentially socially superior, positions; it is another thing that the ‘someone’ tends to display particular demographic characteristics and, for that reason, tends to enjoy a higher expectation of occupying those positions.Footnote ¹³

4.2. Lottocracy’s advantage in avoiding disparities of consideration

Lottocracy is prima facie more effective in avoiding the above problems by limiting the ways for disparities of consideration to be reflected and expressed in the composition of officials. First, since sortition disallows electoral voting, lottocracy intrinsically prevents voters’ arbitrary reasons from influencing the selection outcomes, which suffices to prevent voter discrimination of whatever sort from biasing the composition of officials. Second, since lottocracy makes electoral competition unnecessary, it relieves individuals of concerns about whether they can effectively enter electoral competitions in the first place.

Of course, these two expected effects of lottocracy have often been observed in the literature, so this article’s contribution is not merely in pointing them out. Instead, I am offering an explanation of why these effects are significant advances for social equality. Because lottocracy more directly prevents disparities of consideration from biasing the composition of officials, it is better positioned than electoral democracy to satisfy individuals’ Claim against Lesser Consideration. Given the deontic interpretation of what social equality demands (§2.4), when such claims of individuals are recognizably violated in the political, social, and economic processes in society, the state or its government bears a duty to take measures to prevent the violation within the practicable range. Implementing lottocracy over electoral democracy can be an effective way for the government to discharge this duty.

To see how the above argument based on social equality is interesting, observe that it is different from the argument that lottocracy realizes a fair distribution of access to legislative office. If the justification for lottocracy were simply that it satisfies an ideally fair distribution, it would encounter an important objection by Umbers (Reference Umbers2021). A fair, lottery-based distribution of a good is justified when (1) there are many persons, each having an independent claim to the good, (2) each person’s claim is roughly of equal weight, and (3) there is no way to distribute the good equally among everyone with this claim (Umbers, Reference Umbers2021: 314). Umbers argues that, from a social egalitarian perspective, it is dubious that condition (1) is satisfied when the good in question is the access to political office. This good is inherently in tension with social equality between political officials and those coercively subjected to their greater political power (Umbers, Reference Umbers2021: 317). Although some lottocrats (e.g., Jakobi, Reference Jakobi2019: 50; Abizadeh, Reference Abizadeh2021b: 797) seem to assume that individuals have an independent, positive claim to occupy such office, no one in fact has such a claim, so condition (1) fails. Therefore, Umbers concludes, the lottery-based fair distribution of access to political office is especially hard to justify if we have a concern for social equality.

My argument avoids Umbers’s objection. I am not insisting that the biased composition of officials is unfair, nor that the virtue of lottocracy lies in its fairness in distributing the access to political office. Rather, my argument emphasizes the following idea. Given that the access to political office must be distributed somehow, it is an affront to social equality if the distribution tracks people’s demographic characteristics in ways that reflect or express unmerited disparities of consideration. This argument does not deny that the good being distributed (i.e., the access to political office with greater political power) might inherently be in tension with social equality, nor does it insist that each person has an independent claim to that good.Footnote ¹⁴ Indeed, acknowledging this inherent problem with political office strengthens my case. If anyone must hold a position that potentially violates others’ Claim against Lesser Power and thereby makes its occupant socially superior to those others, no individual’s opportunity to access that position should be relatively worsened due to unmerited disparities of consideration that track their demographic characteristics.

In other words, the concern for social equality does not only count against establishing an inequality of inescapable and final power among individuals (which any representative system does); the same concern also serves as a second-order constraint on how the state should attempt to distribute positions of social inferiority and superiority among people with different demographic characteristics in non-ideal circumstances where distributing those positions unequally is inevitable. It is an affront to the Claim against Lesser Consideration that people with advantaged demographic characteristics, such as being richer, male, of a majority racial/ethnic group, of faith in a majority religion, and so on, enjoy disproportionately higher prospects of occupying potentially socially superior positions. The point of my argument is that when the state implements electoral democracy, the state tends to contribute to maintaining such disproportionality and thus violates the constraint; the state can at least avoid making such a contribution by implementing lottocracy instead.

4.3. Anticipated objections

Having thus clarified the point of my argument, I anticipate two important objections to it. One objection doubts lottocracy’s capacity to mitigate selection bias, and the other purports to defend election by qualifying it with quotas and other countermajoritarian measures. I will address these objections below.

The first objection is that insofar as lottocracy is to respect a liberal constraint, it will fail to solve the problem of selection bias. If lottocracy is to be compatible with individual freedom, it must grant people randomly selected for office an option to decline the offer. But this option creates room for biased demographic compositions of the lottocratic legislature through attrition. That is, some invitees, especially those from disadvantaged social groups, will tend to be less willing to accept the invitation to serve in political office, due to the fear of being denigrated, greater responsibilities for unpaid labour and caring burdens, greater susceptibility to chronic illness resulting from lesser access to medical care, and so on. Indeed, Lever and Destri (Reference Lever, Destri and Grandjean2024: 96–97) suggest that such worries can be supported as an extrapolation from similar observations in analogous contexts such as jury trials and deliberative mini-publics (e.g., Jacquet, Reference Jacquet2017; Collins and Gialopsos, Reference Collins and Gialopsos2021).

In response, I grant that lottocracy will not fully eliminate the factors giving rise to differential willingness to serve, so some failure of descriptive representation is to be expected. Even so, the prima facie advantage of lottocracy over electoral democracy stands. The factors described in the previous paragraph affect willingness to serve under both electoral democracy and lottocracy, although what they affect differs: under electoral democracy, willingness to run for office; under lottocracy, willingness to accept the randomly issued invitation to serve. Either way, the problem of differential willingness to serve in office is, as such, not a decisive consideration in favour of election over sortition.Footnote ¹⁵ The particular problem with election is that it exacerbates the biased composition of officials that reflects and expresses disparities of consideration. For instance, voters tend to disapprove of female candidates who do not have traditional household profiles such as being married with children; they tend to favour female candidates who do. Thus, ‘female candidates without families—who could better afford the time to invest in political careers—may face greater degrees of voter disapproval, while women in more demanding family roles have less time for political work’ (Teele et al., Reference Teele, Kalla and Rosenbluth2018: 526). Election thus undermines the electability of even those willing to run and to serve if elected. Because lottocracy avoids this problem more effectively by undermining the effect of biased voting, it is prima facie better positioned to satisfy the Claim against Lesser Consideration held by those potentially willing to serve in office.

The second anticipated objection is that electoral democracy can also address the biased composition of officials by adopting countermajoritarian mechanisms. Critical democratic theorists have discussed various such measures, including ‘quotas in electoral lists, proportional representation, reserved seats, the drawing of boundaries for electoral jurisdictions’ (Young, Reference Young2000: 141; see also Guinier, Reference Guinier1994: chap. 5; Phillips, Reference Phillips1995: chaps. 3–4; Young, Reference Young2000: 148–151). Such measures could help prevent the demographic composition of officials from deviating too much from an unbiased composition. Moreover, one might suggest a further progressive potential of such measures. Purely random selection would make numerical minorities in the general population, such as minority ethnic groups (accounting for, say, 5–20% of the general population), also numerical minorities in the legislature. By contrast, the measures suggested here might increase the proportion of representatives for a social group even more than would be guaranteed by random selection, thus further closing the intergroup power gap within the legislature (Lever, Reference Lever2024: 142–143).Footnote ¹⁶ If so, why should we pursue lottocracy to advance social equality?

In response, I agree that such countermajoritarian measures, duly designed, can mitigate the biased composition of elected officials. Indeed, where some social groups have been persistently excluded or marginalized, the concern for social equality would require an electoral system with quotas, reserved seats, and other mechanisms of group representation if it is compared with an electoral system without them. However, the problem is that these measures primarily aim at correcting the demographic composition of the resulting elected body. They do not fully remove the ways in which electoral competition converts certain demographic characteristics into unequal prospects of obtaining office. Even when quotas are in place, individuals must still pass through processes of candidate emergence, party recruitment, party nomination, campaign financing, media exposure, and voter evaluation. Each of these processes may remain affected by background inequalities of opportunity and unmerited disparities of consideration. For example, gender quotas may increase the number of women in office, but they need not eliminate the gendered pipelines through which some women rather than others become viable candidates (given the findings by Thomsen and King, Reference Thomsen and King2020), nor do they eliminate voter penalties against candidates whose family status, ethnicity, sexuality, religion, disability, or other characteristics make them appear less conventionally credible or electable. Similarly, reserved seats for one protected group may leave untouched inequalities within that group or across other unrecognized groups.

To be sure, some of these concerns might be addressed by using more fine-grained conditions for quotas or reserved seats.Footnote ¹⁷ For instance, we could reserve seats not for ‘women’ but for ‘women with/without marital status, with/without children, more/less affluent women’, and so on. This is also a possible institutional design. However, I doubt that its possibility counts in favour of election over sortition. As the conditions become more fine-grained, devices for special representation will face increasing problems of contestable classification and administrative complexity. By contrast, sortition does not need to specify ex ante every socially salient group whose members might be disadvantaged by electoral competition. It reduces reliance on the electoral-competitive filters that generate such disadvantage in the first place.

Thus, although the availability of electoral reforms might weaken the worry about descriptive representation in electoral democracy, it does not defeat the more specific social egalitarian claim defended here. The relevant complaint is not merely that some groups receive too few seats, but that unequal accessibility to positions of greater political power tends to track demographic characteristics in ways that reflect or express unmerited disparities of consideration. Electoral reforms might reduce this tracking at the level of aggregate outcomes (e.g., using quotas or reserved seats), but they usually leave intact the fact that unequal electability stems from electoral competitions. Sortition more directly prevents advantages in wealth, social networking, personal or ideological fit with a political party, favourable appearance and performance on the media and campaign, and voter approval from functioning as advantages in obtaining office. In other words, sortition functions not as an ex post facto corrective device for the distribution of seats but as an ex ante mechanism to prevent the tracking of demographic characteristics; at the initial selection stage of lottocracy, no one is made less likely to receive the chance to hold office because voters, parties, donors, media actors, or social or economic organizations give them lesser consideration. That is why sortition makes the selection of political officials less of an occasion for social inequality than election does, even if later stages, such as acceptance of the offer and actual service in office, require further institutional safeguards.