One person, one vote and the importance of baseline

ABSTRACT

‘One person, one vote’ is wedded to the idea of democracy to such an extent that many would hesitate to refer to a system, which deviated from this, as a democracy. In this paper, I show why this assumption is hard to defend. I do so by pointing to the importance of baseline in justifying a system of ‘one person, one vote'. The investigation will show that the reasons underlying the most prominent views on democratic inclusion cannot justify ‘one person, one vote’ robustly. This ultimately means that democrats should be less confident in their judgment that 1P1V is a necessary part of democracy.

KEYWORDS:

• Democracy
• one person, one vote
• baseline

1. Introduction

It is hard to think of what democracy means without almost automatically thinking that ‘one person, one vote’ is part of it. Indeed, ‘one person, one vote’ is wedded to the idea of democracy to such an extent that many would hesitate to refer to a system, which deviated from this, as a democracy.¹ My point in this paper is to show why this assumption – that a necessary part of a democratic system is ‘one person, one vote’—is hard to defend. Others have done so before.² But I will approach this problem from an angle which, to my knowledge, has not been explored before.³ What I want to point to is the importance of baseline in justifying a system of ‘one person, one vote.’ To illustrate, suppose a community is governed by:

Monarchy: The king makes decisions for the people single-handedly.

If Monarchy is the baseline from which we seek to justify a system of one person, one vote, many democratic arguments look appealing. Perhaps a democracy with one person, one vote is needed to give everyone an opportunity to protect their interests. Or perhaps to secure their self-government. Or perhaps because one person, one vote embodies a community with equal relations. To beat monarchy, not much by way of argument is needed. But suppose instead that a community is governed by:

One person, one vote + one (1P1V+1): The people make collective decisions in accordance with a ‘one person, one vote’ rule except that one voter is given an extra vote, say, as compensation because they are the person who is worst off in the community.

What I will argue in this paper is that if 1P1V + 1 is the baseline from which we seek to justify a system of one person, one vote, the reasons underlying the most prominent principles of democratic inclusion cannot explain why it would be democratically objectionable not to implement:

One person, one vote (1P1V): The people make collective decisions in accordance with a ‘one person, one vote’ rule.

Indeed, I will argue that if the baseline is 1P1V + 1, justifying 1P1V by appealing to (i) opportunity for interest protection; (ii) welfare maximization; (iii) self-government; or (iv) equal relations will not work.⁴ These are the reasons underlying the most prominent views on democratic inclusion. Importantly, this shows that these democratic reasons only justify 1P1V given particular baselines. That is, the democratic justifications for ‘one person, one vote’ are not robust. This ultimately means that democrats should be less confident in their judgment that 1P1V is a necessary part of democracy.

The paper proceeds as follows. In the next section, I introduce the boundary problem—the question of determining who should be included in democratic decision-making—and point to three prominent solutions to this problem: the all-affected principle, the all-subjected principle and a principle based upon relational egalitarianism. I analyze these principles and their underlying reasons in Sections 2a-2c to examine whether they can justify 1P1V, assuming a baseline of 1P1V + 1. Section 3 concludes.

2. The boundary problem and a 1P1V + 1 baseline

To examine whether 1P1V can be justified, assuming a baseline of 1P1V + 1, it will be useful to turn to the boundary problem in democratic theory. This is the question of determining who should be included in democratic decision-making (Arrhenius 2005; Beckman 2009; Goodin 2007; Miller 2009; Whelan 1983). Dahl observed, surprisingly, in 1970 that ‘how to decide who legitimately make up the people … is a problem almost totally neglected by all the great political philosophers who write about democracy’ (Dahl 1970, 60). Since then, however, the boundary problem has received more attention. From these debates, two principles have emerged as the most widely supported: the all-affected principle and the all-subjected principle. More recently, however, a third potential solution has arisen based on the now prominent view that democracy is valuable because it is a constituent part of relating as equals (see, e.g. Kolodny 2014b; Viehoff 2014). I will explore whether these three principles entail that it would be democratically objectionable not to implement 1P1V, assuming a baseline of 1P1V + 1.

2.1. The all-affected principle

I will start with the all-affected principle since this is, arguably, the most widely endorsed solution to the boundary problem. At its core, the all-affected principle says that those affected by a given (collective) decision ought to be included in the making of that decision (Dahl 1970; Goodin 2007). This formulation is underspecified in various ways. First, there is the question of what it means to be affected. Usually, it is assumed—and I will assume so as well—that being affected is to have one’s interests influenced by a given decision (Arrhenius 2005; Goodin 2007). Second, there is the question of whether, to have a claim for inclusion, one must be actually affected or whether it suffices that one is possibly affected (Goodin 2007; Owen 2012). It does not make a difference to my arguments whether we assume an actuality or a possibility understanding of affectedness. Thus, to focus the discussion, I will assume that a person has a claim to inclusion if, and only if, they are actually affected by a decision where that is to be affected by a choice between a range of options and not only to be affected by an outcome (Owen 2012, 132).⁵

To determine whether 1P1V is a democratic requirement, assuming the all-affected principle, we need to know the reasons underlying the all-affected principle that explain why an affected individual should be included in democratic decision-making because they are affected (Bengtson and Lippert-Rasmussen 2021). We will then be able to determine whether realizing this value requires implementing 1P1V. Three reasons have been proposed in the literature: (a) opportunity for interest protection; (b) self-government; and (iii) welfare maximization. As self-government has also been taken to underlie the all-subjected principle, and since it seems to fit that principle better, I will save the discussion of self-government to the next section in which I discuss the all-subjected principle. But what I say there in relation to self-government will apply to the all-affected principle as well. Thus, in this section, I will focus on the other two reasons, i.e. opportunity for interest protection and welfare maximization.

I will start with welfare maximization. According to Andrić (2017), Whelan appeals to this as the reason underlying the all-affected principle in the following quote:

Utilitarianism holds that the right action (or decision) is that which in the circumstances produces the greatest net increase in happiness or welfare of those affected by it … It [the utilitarian principle] may also serve as a standard for political decision-making, prescribing that public policy aim at the greatest overall welfare of those over whom rule is exercised. The utilitarian argument for political democracy – that those affected should not only be benefited by governmental decisions, but also participate in making them – follows from either or both of two additional points. It may be claimed that each individual is the sole, or best, judge of his own interests, which he defends and promotes with his vote in the democratic process. Or, it may be denied that any ruler or group of rulers other than the people (those affected) as a whole can be depended upon to be motivated by the utilitarian principle: democratic decision-making is the procedure most likely to lead to utilitarian results even in the absence of benevolent motives in individuals. Utilitarianism suitably developed thus may yield the all-affected principle for collective decision-making, and appeal may be made to it in support of this variety of democratic theory (Whelan 1983, 17–18).^6,7

Thus, according to this explanation, the reason affected individuals must be included according to the all-affected principle is that it maximizes welfare. Assuming this underlying justification, and assuming a baseline of 1P1V + 1, the all-affected principle entails that 1P1V is required if it maximizes welfare, or if it at least leads to more welfare than 1P1V + 1.

Whether 1P1V leads to more welfare than 1P1V + 1 is a complex empirical question (to which we simply do not have evidence). Even if there are situations where it would be the case that 1P1V would lead to more welfare than 1P1V + 1, we can easily imagine other situations in which 1P1V + 1 would lead to more welfare than 1P1V. Suppose that in a society, most people do not care about politics. Many of them do not vote and those that do vote do not care much. Suppose, however, that one person in this society derives a large amount of welfare from exercising their right to vote, e.g. because they are extremely interested in politics and because their self-identity is based upon participating in politics. This person would derive a large amount of welfare, let us assume, from being granted an extra vote. It is likely that in this case, 1P1V + 1 would lead to at least as much welfare—and maybe even more—than 1P1V. Due to the small difference between 1P1V and 1P1V + 1, one person deriving a large amount of welfare from an extra vote may make a significant difference. Or suppose, to give another example, that we grant an extra vote to the poorest person in society as a form of compensation. It is hard to see that this would not lead to as much welfare as 1P1V, at least in some situations (Brighouse and Fleurbaey 2010).⁸

Thus, it seems that if we want to justify 1P1V, assuming a baseline of 1P1V + 1, the all-affected principle with its underlying reason being maximizing welfare is not the way to go. This is perhaps not surprising given that most democrats do not believe that democratic inclusion should be justified by appealing to welfare maximization. They believe, and rightly so, that even if a dictatorship would maximize welfare relative to 1P1V, we would still have democratic reasons to prefer 1P1V. Thus, let us look to the other reason taken to underlie the all-affected principle.

This is the reason that those affected must be included because they should be granted an opportunity to protect their interests. Goodin (2007, 50) seems to have this reason in mind when he says, ‘protecting people’s interests is thus the most plausible candidate principle for bringing the who and the how of democratic politics into alignment. That principle dictates who should constitute the decision-making group (all affected interests should have a say).’⁹ This is also true of Miller (2009, 216) when he says, ‘Insofar as the [all-affected] principle is meant to reflect the underlying idea that people should have an equal opportunity to advance and protect their interests politically, it seems that in applying it we should try to ensure that each person’s capacity to influence a decision should correspond to how significantly he or she will be affected by its outcome.’ Thus, in short, this rationale specifies that the reason a person should be included when their interests are affected is that they should be given an opportunity to protect their interests, and they can only be granted this opportunity through democratic inclusion.

Can this reason, assuming a baseline of 1P1V + 1, explain why it would be democratically objectionable not to move to 1P1V? For the following reasons, I do not think so. The first reason is that, under 1P1V + 1, everyone actually has an opportunity to protect their interests qua having a vote. So, in that sense, a baseline of 1P1V + 1 seems different from a baseline of monarchy since in that system, everyone clearly does not have an opportunity to protect their interests. The king simply decides. One may object that in an absolute sense, their opportunity is worse under 1P1V + 1 than under 1P1V for which reason 1P1V is preferable from the point of view of the all-affected principle. It is hard to see that this argument suffices. The problem is that being granted a vote is not much of an opportunity in the first place. As Brennan (2012, 10) explains, ‘[i]ndividual votes in fact have vanishingly small instrumental value … since the probability that an individual vote will decide the outcome of the election is vanishingly small’ (see also Brennan 2011 and the references therein; cp. my discussion of self-government in the next section).

This is of course a problem for this rationale in the first place because it makes it unclear that being granted a vote is relevant when people are owed an opportunity to protect their interests. But even if we assume that it is such an opportunity, the difference that one vote makes for others’ opportunity is almost non-existent. And this is the only difference between 1P1V and 1P1V + 1—one vote. It is hard to see that this vote—in terms of the opportunity for interest protection understood in an absolute sense—can carry the democratic weight of justifying why we would have to move to from 1P1V + 1 to 1P1V, especially when we also take into account the response to the following objection. Instead of objecting to the absolute sense in which the opportunity becomes worse, one may instead object to the relative sense in which the opportunity becomes worse. If we have a system of 1P1V + 1, those with one vote (which is everyone except one person) will have a worse opportunity to protect their interests relative to this person than under 1P1V. This is why it would be democratically objective not to switch from 1P1V + 1 to 1P1V. There is the following problem with this response. If we grant the extra vote to the person who is worst off, it may actually be that this is better in terms of everyone having an opportunity to protect their interests. This is the case since their opportunity for interest protection might have been worse in the first place than the opportunity for interest protection of those with one vote. For this reason, moving from 1P1V + 1 to 1P1V might actually lead to more unequal opportunity for interest protection overall.¹⁰ Thus, it seems that, by appealing to the opportunity for interest protection, we cannot explain why it would be necessary, from a democratic point of view, to move from 1P1V + 1 to 1P1V. Let us instead turn to the all-subjected principle.

2.2. The all-subjected principle

As I said, the all-subjected principle is the most popular alternative to the all-affected principle (Abizadeh 2012, 878; Beckman 2008, 351; Dahl 1989, 122; Erman 2014, 539; López-Guerra 2005, 222). Whereas the all-affected principle says that those affected by a given (collective) decision ought to be included in the making of that decision, the all-subjected principle says that those subjected to a given (collective) decision ought to be included in the making of that decision (Goodin 2007, 49). These principles are different. A person may be subjected to a decision without being affected by that decision (in a way that matters for the all-affected principle), e.g. a person may be subject to parking codes even if they do not own a car (Frazer 2014, 387). And a person may be affected by a decision without being subjected to it, e.g. a person who lives in a state that receives foreign aid from another state (cp. Goodin 2007; Miller 2009). As was the case with the all-affected principle, we need to know the reason underlying the all-subjected principle that explains why an individual must be included in democratic decision-making when and because they are subjected. That reason, it has been argued, is self-government. To give an example, López-Guerra (2005, 221) says, ‘if democrats endorse the ideal of self-government, anyone who is subject to the laws of a democratic polity should be included in the citizen body’ (see also Abizadeh 2008, 39–40; Goodin 2016, 369; Miller 2009, 214; Näsström 2011, 120–122).

To be self-governing is to have control over your own life. A self-governing person is the author of their life and is not merely ‘subjected to the will of another’ (Raz 1986, 154–155).¹¹ For instance, a slave is not self-governing since they do not have any control over whether to be a slave and is merely subjected to the will of their master. May an appeal to self-government explain why it would be democratically objectionable not to move from 1P1V + 1 to 1P1V? I do not think so. According to Viehoff (2014, 350–351), democratic inclusion makes the individual less self-governing:

[O]nce a democratic decision has authority over me, my ability to give shape to my life in light of my own judgment is limited. Furthermore, this loss in autonomy almost certainly far outweighs the purported gain. When the authority of democratic decisions is extended, I gain a very small share of control over the lives of very many other people, and specifically over that aspect of their lives that our collective decision now regulates. Yet I also lose a large share of control over the corresponding aspect of my own life. Autonomy is a matter of self-rule, of giving shape to one’s own life in accordance with one’s own judgment. Since the control I gain is mostly control over others rather than myself, it barely advances my autonomy. By contrast, since the control I lose is largely control over myself, my autonomy is actively set back.

A similar argument is also put forward by Christiano (1996, 19, 24-25) who argues that individual self-government and democracy are incompatible.¹² The problem is that if democratic inclusion does not increase the individual’s self-government,¹³ it cannot explain why we, for self-government reasons, would have to move from 1P1V + 1 to 1P1V.

Even if we were to assume, for the sake of argument, that democratic inclusion could contribute minimally to an individual’s self-government, that argument does not seem appealing if the baseline is 1P1V + 1 instead of, say, monarchy. If monarchy is the baseline, one may try to appeal to the fact that the individual gains a small amount of self-government by being granted the vote. In that sense, self-government may, let us assume, play some role if that is the baseline. But if the baseline is 1P1V + 1, it simply does not seem plausible to say that we have to implement 1P1V in order to secure that the individuals in the society are self-governing. It simply does not seem to make a (relevant) difference to their self-government that one voter has an additional vote.

Viehoff’s argument in the quote above also reminds us that we have to distinguish individual and collective self-government. Even though democratic inclusion does not make a difference to an individual’s self-government, it might make a difference to the collective’s self-government. The relevant collective, consisting of, say, the adult population, is simply more self-governing if there is democracy as opposed to, say, monarchy. In that sense, it may be that democratic inclusion can be justified by appealing to collective self-government (cp. Bengtson and Lippert-Rasmussen 2021, 581–582).¹⁴ But that will not make a difference in this case since under both 1P1V + 1 and 1P1V, the collective is self-governing.¹⁵

I think this discussion nicely illustrates that the appeal of self-government as a reason in favor of 1P1V is highly dependent on the baseline from which we evaluate it. Even if it has some appeal if the baseline is monarchy or similar forms of government, an appeal to self-government cannot explain why it would be objectionable, democratically speaking, not to move from 1P1V + 1 to 1P1V.

2.3. The equal relations principle

We have seen that the reasons underlying the all-affected principle and the all-subjected principle—opportunity for interest protection, self-government and welfare maximization—cannot explain why it would be democratically objectionable not to move from 1P1V + 1 to 1P1V. But these principles are not the only solutions to the boundary problem. Some have argued that we must find a solution to the boundary problem by appealing to the value of democracy (López-Guerra 2005, 221; Miller 2009; Song 2012). As Miller (2009, 204) explains, ‘[T]he domain problem [the boundary problem] cannot be solved by appeal to democratic procedure. But this does not mean that it cannot be solved by appeal to democratic theory, understood to mean the underlying values, such as political equality, that justify procedures like majority voting.’ This is actually what we have already been doing, by looking to the reasons, or values, underlying the all-affected and the all-subjected principles. But there is a prominent, recent argument to which we have not yet looked. This is an argument put forward by relational egalitarians, i.e. those who believe that justice requires equal relations.¹⁶ According to relational egalitarianism, democracy, understood as including 1P1V as a necessary part, is valuable because of its contribution to equal relations (Anderson 1999; Kolodny 2014b; Umbers 2021; Viehoff 2014). I say ‘contribution’ because there are in fact three different relational egalitarian arguments for why democracy is valuable:¹⁷

The causal argument: 1P1V is causally necessary for equal relations.¹⁸

The expressive argument: 1P1V is, for expressive reasons, necessary for equal relations.

The constitutive argument: 1P1V is a constitutive part of equal relations.

If any of these arguments succeed, they would be able to explain why it would be objectionable, from a democratic point of view, not to move from 1P1V + 1 to 1P1V.

Let us start with the causal argument. The idea here is that if we grant some people one vote and other people more votes, this difference in political power causes unequal relations between the inferiors (those with one vote) and the superiors (those with more votes). There is something to this argument. Suppose we, in a racist society in which white people are considered morally superior to black people, grant white people more votes than black people. This may cause (further) relational inequalities between white people and black people—and that would be a reason not to deviate from 1P1V.

The problem with this argument, as Bengtson (2020) has argued, is that 1P1V may also cause unequal relations. Consider a democracy with a white majority and a black minority. A sufficient number of whites always vote together for which reason the black minority always end up losing. The black minority is a persistent minority. They will only have it their way insofar as white people agree with them, and not because it is their will (cp. Pettit 1997). In this situation, we may say that 1P1V causes relational inequalities between white and black people. As Bengtson (2020, 1058) says, ‘we can easily expect it to be hard for members of the minority and the majority to relate as social equals [in such a situation]. Whereas members of the minority may regard themselves as inferiors—after all, they constantly lose out to the majority—members of the majority may regard themselves as superiors in that they solely decide on the democratic decisions of the polity.’

Moreover, we may compare 1P1V with a 1P1V + 1 system in which we grant an additional vote to the person with the lowest standing in the community to boost their standing. In terms of causal effects, it is hard to see that 1P1V is necessarily more conducive to equal relations than 1P1V + 1. Perhaps the causal argument is able to justify why, if we had to choose between a monarchy and 1P1V, we would have to choose 1P1V for relational egalitarian reasons since that system causes less unequal relations than a monarchy. But the causal argument cannot explain why, if we had 1P1V + 1 as above, 1P1V is causally necessary for equal relations.

Let us turn to the next relational egalitarian argument for 1P1V, i.e. the expressive argument. This argument points to the fact that for (some) relational egalitarians, the justness of an act/practice is not solely determined by the consequences of the act/practice; it is also determined by what the act/practice expresses (Anderson 1999; Schemmel 2021; Voigt 2018). To see what relational egalitarians have in mind, consider the following example presented by Scanlon (2008, 53): ‘if the only family in the neighborhood that was not invited to the block party is also the only black family in the neighborhood, then that family may reasonably take this as a sign of prejudice, even if they were left out by mistake, or because only people with children were invited and they have none.’ What this act—of not inviting the black family—expresses is not determined by the reasons for which the white families did not invite the black family. The white families may have left them out by mistake or because only families with children were invited and the black family have none, as Scanlon explains, but the act, nevertheless, expresses that the black family is not welcome, that they are inferior to the white families.

Lippert-Rasmussen (2018, 77) presents a similar example: ‘if I fly a Confederate flag over an official building in southern US states, I might express racist attitudes, even if I am not in any way racist’ (Lippert-Rasmussen 2018, 77). Similarly, whether this act expresses racist attitudes is not determined by the reasons for which this person flew a Confederate flag over an official building—in this case, that he simply did not understand the meaning of the flag, as Lippert-Rasmussen assumes—but the context in which the act takes place, i.e. a community with a racist history in which the flag is a racist symbol. Indeed, what an act expresses is determined by how others may reasonably interpret the act in the given social context. One must interpret the act ‘in light of the community’s other practices, its history and shared meanings’ (Anderson and Pildes 2000, 1524–1525; Hellman 2008; Lippert-Rasmussen 2018, 77).

The expressive argument points to the fact that if we deviate from 1P1V, the voting system expresses unequal relations between those who have less voting power (the inferiors) and those with more voting power (the superiors). This will be the case even if the system is not based on any ill intentions. But it will clearly also be the case if the system is based on ill intentions. In that case, the expressive effect of the system may add to the wrongness of the system. Suppose a racist society in which white people are considered morally superior to black people deviates from 1P1V by granting white people more votes than black people. In addition to this voting system being implemented due to ill intentions, it also expresses that black people are morally inferior to white people; that the interests of white people are more important than the interests of black people.

The problem with the expressivist argument is that expressivist arguments are contingent arguments (cp. Brennan and Jaworski 2015, 1057; Mogensen 2019, 91) but it assumes a necessary connection between deviating from 1P1V and expressing unequal relations. In some contexts, 1P1V may also express unequal relations. Consider again a democracy with a white majority and a black minority in which a sufficient number of whites always vote together for which reason the black minority always end up losing. The black minority is in this case a persistent minority. Black voters may reasonably take 1P1V in this situation to express that black people are inferior to white people; that the system expresses that the interests of white people are more important than the interests of black people (even if this is not in any case the intention behind implementing 1P1V). In this case, implementing affirmative action in the political domain by giving extra voting power to black people compared to white people may express less unequal relations than 1P1V (cp. Bengtson 2022a).

Also, the baseline from which we are determining whether 1P1V is, expressively speaking, necessary for equal relations is 1P1V + 1 in which, let us suppose, we are granting an additional vote to the person with the lowest standing in the community. In this situation, it seems implausible to maintain that 1P1V is better, from the point of view of relational egalitarianism, than 1P1V + 1 in terms of what the systems express.

The final relational egalitarian argument for 1P1V is the constitutive argument, i.e. that 1P1V is a constitutive part of equal relations. Kolodny (2014b, 287; see also Viehoff 2014) puts forward this argument, saying: ‘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.’¹⁹ Democracy is a particularly important constituent of relating as equals, Kolodny argues, because democratic decisions (a) cannot usually be escaped at will; (b) characteristically involve the threat of force against non-compliers; and (c) have final de facto authority (Kolodny 2014b, 304–307). They have final de facto authority in the sense that, first, political decision-making cannot be moderated by a higher court of appeal, and, second, political decisions have final authority over nonpolitical decisions (Kolodny 2014b, 306). Peña-Rangel (2022) more recently argues in favor of the constitutive argument. His central claim is that ‘part of what it means to relate to one another as equals is to have equal political power [understood as one person, one vote]’ (Peña-Rangel 2022, 24). This is the claim that is at the core of the constitutive argument: that 1P1V is a constitutive part of equal relations.

Can the constitutive argument explain why it is democratically objectionable not to implement 1P1V, assuming a baseline of 1P1V + 1? Let us assume that 1P1V + 1 grants an additional vote to the person with the lowest standing in the community. There are the following problems with the constitutive argument. First, it seems that 1P1V + 1 is at least as constitutive of equal relations as 1P1V. To grant an additional vote to the person with the lowest standing to boost their standing vis-á-vis people with higher standing does not seem constitutive of unequal relations (to a larger extent than 1P1V). To simply maintain that it does—that 1P1V simply is a constitutive part of equal relations—is to beg the question against those who do not already believe that 1P1V is a constitutive part of equal relations. Second, even if we assume that 1P1V is constitutive of equal relations such that 1P1V + 1 would constitute unequal relations, it does not necessarily follow that these unequal relations are objectionable. To see why this is the case, consider the pervasiveness problem, facing relational egalitarians, which Scheffler (2005, 17–18) lays out as follows:

[relational] equality is in some ways a puzzling value and a difficult one to interpret. After all, differences of rank, power, and status are endemic to human social life. Almost all human organizations and institutions recognize hierarchies of authority, for example, and most social roles confer distinctions of status which in turn structure human relationships, such as the relationships of doctors to patients, teachers to students, parents to children, attorneys to clients, employers to employees, and so on. If there is any value at all in such relationships, then at least one of the following things must be true. Either some relationships can be valuable despite having a fundamentally inegalitarian character or else it is not necessary, in order for a relationship to qualify as having an egalitarian character, that it should be altogether unmarked by distinctions of rank or status … Both points are almost certainly true. This means that, in order to understand the value of equality, one needs to investigate the specific respects in which egalitarian relationships must be free from regimentation by considerations of rank or status.

The pervasiveness problem points out that relational egalitarians must come up with an explanation of which unequal relations are unjust, and which are not, given the plausible assumption that some unequal relations are not unjust, e.g. the parent-child relationship and the university teacher-student relationship. Not any unequal relation is objectionable according to relational egalitarianism. I take it that a plausible solution to the pervasiveness problem would not judge the unequal relations resulting from 1P1V + 1, if such unequal relations were to result, to be objectionable, since the system would target and improve the standing of the person with the most inferior standing in the community. Thus, even if we assume that 1P1V is a constitutive part of equal relations, such that 1P1V + 1 is constitutive of unequal relations, this does not necessarily mean that the unequal relations are objectionable. In fact, if they are unequal, it is in a way which favors the person with the lowest standing in the community. And it seems that a plausible account of relational egalitarianism should not find such relational inequality objectionable. If so, the relational egalitarian argument does not justify 1P1V, assuming a baseline of 1P1V + 1.

Finally, and most importantly, 1P1V is not constitutive of equal relations. If 1P1V is constitutive of equal relations, it would entail that in any society without 1P1V, there would be unequal relations. But that is not the case. And Peña-Rangel (2022) even admits that. Consider a scheme which distributes voting power equally over the course of people’s lives, but which at no point in time distributes voting power equally. One such scheme is an age-differentiated voting scheme ‘which accords more votes to people at certain points in their life, and so distributes power unequally at time t but equally at time t + n’ (Peña-Rangel 2022, 30). Peña-Rangel (2022, 32) concludes with regard to such a system: ‘So while it is possible for an entire-lives distribution of equal power to be compatible with relational equality, this is true exceptionally.’²⁰ But given that this is true, 1P1V cannot be a constitutive part of equal relations since if it were, any society without 1P1V would contain relational inequality, and that is not the case according to Peña-Rangel. And given this, we may question why 1P1V + 1 cannot also constitute equal relations (or less unequal relations) since that system may also secure that people will be equals, or unequals to a lesser degree, at any point in time and/or over their lives as a whole.

What this again shows is that if the baseline is monarchy, or a similar system, the constitutive argument looks promising as a defense of 1P1V. But the constitutive argument is not successful across any potential baseline. If the baseline is 1P1V + 1, the constitutive argument, or any of the other relational egalitarian arguments for that matter, cannot explain why it is democratically objectionable not to implement 1P1V.

3. Conclusion

In this paper, I have argued for the importance of baseline in justifying a democratic system of ‘one person, one vote.’ Although many justifications for ‘one person, one vote’ look promising when compared to a system such as a monarchy, this appeal must not deceive us. We cannot deduce from this that these democratic reasons also explain why it is necessarily democratically objectionable to deviate from a system with ‘one person, one vote.’ Indeed, we have seen that if the baseline is 1P1V + 1, appealing to (i) opportunity for interest protection; (ii) welfare maximization; (iii) self-government; or (iv) equal relations cannot justify 1P1V. This is highly important: it shows that the justifications for ‘one person, one vote’ are not robust across different baselines. And it ultimately also shows that we, as democrats, should be less wedded to ‘one person, one vote’ than is usually assumed.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Independent Research Fund Denmark (‘Danmarks Frie Forskningsfond’): [grant no 1027-00002B].

Notes

1 For instance, Kolodny (2014a, 196) says, Even in political philosophy, which is, as to be expected, more reflective, a commitment to democracy, thus broadly understood, often outstrips any explicit justification. Sometimes it is just assumed that our task is to construct a political philosophy for a liberal democracy, where some principle of one person, one vote is, like the injustice of chattel slavery, a ‘fixed point.’

2 See e.g., Brighouse and Fleurbaey (2010); Guerrero (2014); López-Guerra (2014); Mulligan (2018); Saunders (2010); Wall (2007).

3 Perhaps an exception is Brighouse and Fleurbaey (2010) in the sense that they argue that including people according to their stakes is more conducive for self-government and utility than a system with equal voting weights. But their investigation is still different from mine in that they do not discuss this as a matter of the importance of baseline. Moreover, they do not discuss the, to my mind, most plausible understanding of the value of democracy: the relational egalitarian view that democracy is a constitutive part of equal relations.

4 (i)-(iv) underlie, as I will explain in the next section, the all-affected principle, the all-subjected principle and the relational egalitarian view(s) that democracy is valuable because it causes, expresses and/or constitutes equal relations.

5 I phrase it in this way because, as Owen (2012, 132) points out, it is possible to distinguish between a choice and an outcome version of the all actually affected principle. The choice version says that ‘all whose interests are actually affected by a choice between a range of options should have their interests taken into account in the determination of the option chosen.’ The outcome version says that ‘all whose interests are actually affected by an outcome should have their interests taken into account in the determination of that outcome.’ Owen argues, and I agree, that the outcome version is implausible for the following reason. Suppose a polity is to decide between policy X, Y, Z and W. Suppose they choose Z. Who are actually affected by this? According to the outcome version, only those affected by Z are actually affected. According to the choice version, on the other hand, those affected by X, Y, Z or W are actually affected.

6 This quote, and the other quotes expressing these different reasons underlying the all-affected principle and the all-subjected principle, I borrow from Bengtson and Lippert-Rasmussen (2021).

7 Bauböck also points to utilitarianism as the reason underlying the all-affected principle: ‘AAI [the all-affected principle] is rooted in utilitarian and public choice views of democracy according to which its legitimacy and advantage over alternative forms of political rule lie in its capacity to maximize the satisfaction of political preferences and to resolve collective action dilemmas in the production of public goods’ (Bauböck 2018, 48).

8 Brighouse and Fleurbaey (2010) show that granting voting weights in accordance with social justice such that those who are worst off get a larger weight than those who are better off produces better outcomes than 1P1V.

9 However, he may also be pointing to the utilitarian reason discussed above.

10 Note also that an argument one may put forward in arguing why it would be better, from the point of view of opportunity for interest protection, to move from a monarchy to 1P1V is that the latter grants the relevant collective an opportunity for interest protection (which they do not have if the king single-handedly decides). This argument cannot justify why we should move from 1P1V+1 to 1P1V since the collective does have such an opportunity in the former system.

11 I have adopted this understanding of self-government as that is the one commonly adopted in discussions of democratic inclusion and democracy more generally (see, e.g., Abizadeh 2008, 39; Bengtson and Lippert-Rasmussen 2021, 580; Brennan 2011, 99; Christiano 1996, 19, 24-25; Griffin 2008, 247).

12 Many believe that we should not try to, indeed cannot, justify democratic inclusion by appealing to individual self-government. In addition to Christiano and Viehoff, see e.g., Brennan (2011, 99); Brennan and Lomasky (2006, 242); Griffin (2008, 247); Kolodny (2014a, 208–209).

13 I phrase it in this way since it is not necessary, for my purposes, that democratic inclusion decreases the individual’s self-government, as Christiano and Viehoff argue; it suffices that democratic inclusion does not increase the individual’s self-government.

14 But as Kolodny (2014a, 209) points out, appealing to collective self-government raises the problem that it is not clear why the collective must decide democratically for it to be self-governing.

15 This is also why appealing to collective self-government cannot be used to justify why a particular individual must be included in democratic decision-making. Arguably, the collective would remain self-governing even if that individual were not included (Bengtson and Lippert-Rasmussen 2021, 581–582).

16 Much has been written on relational egalitarianism, see e.g., Anderson 1999; Bengtson 2022b, 2022c; Bidadanure 2016; 2021; Fourie 2012; Fourie, Schuppert, and Walliman-Helmner 2015; Hojlund 2022; Kolodny 2014b; Lippert-Rasmussen 2018; 2019; McTernan 2018; Miller 1998; Moles and Parr 2019; Nath 2020; O’Neill 2008; Satz 2010; Scheffler 2003; 2005; 2015; Schemmel 2012; 2021; Schmidt 2022; Tomlin 2014; Viehoff 2014; 2019; Voigt 2018; Wilson 2019; Wolff 1998; Young 1990).

17 I adopt this tripartite distinction from Peña-Rangel (2022, 23).

18 Strictly speaking, this may only be true assuming that some decisions have to be taken collectively. It may be that, in a system without collective decision-making, people may be able to relate as equals, e.g., in a state of nature, since at least in principle people may have equal power (but see Viehoff 2014). This is also the case for the expressive and constitutive arguments. I set this possibility aside since our baseline is 1P1V+1.

19 Strictly speaking, Kolodny’s argument is compatible with any equal distribution of votes (also one in which everyone receives 0 votes), as he defines democracy as follows: ‘a political decision is democratically made when it is made by a process that gives everyone subject to it equal or both equal and positive, formal or both formal and informal opportunity for informed influence either over it or over decisions that delegate the making of it’ (Kolodny 2014b, 289). So not only 1P1V but also, say, 1P2V is compatible with relational equality according to his argument.

20 It is true exceptionally since it requires that the terms of the relationship cannot be easily altered and there are good exit options (Peña-Rangel 2022, 32). But these are contingent matters.