Vote markets, democracy and relational egalitarianism

Abstract

This paper expounds and defends a relational egalitarian account of the moral wrongfulness of vote markets according to which such markets are incompatible with our relating to one another as equals qua people with views on what we should collectively decide. Two features of this account are especially interesting. First, it shows why vote markets are objectionable even in cases where standard objections to them, such as the complaint that they result in inequality in opportunity for political influence across rich and poor people, are inapplicable. Second, it specifies the sense in which, politically speaking, we should relate as equals, and in doing provides a richer version of recent relational egalitarian accounts of the ideal of democracy.

1. Introduction

Vote markets are not without their supporters. ¹ Still, a large majority of the philosophers who have examined the issues regard it as pro tanto wrong to legalize vote markets: they think there is an overridable (and typically a weighty) moral reason against the legalization of such markets (Sandel 2003: 99; cf. Hasen 2000: 1324–1325). ² Why is this? I shall argue that the wrongness of vote markets is best explained, and in some cases can only be explained (see the penultimate paragraph in section 2), by a specification of relational egalitarianism requiring us, in our collective decision procedures, to relate as equals qua people whose views about what we should decide collectively have equal normative authority. ³ This, I argue, gives us a good abductive reason in favour of the relational egalitarianism in question.

The wrongness of legalizing vote markets has been explained in various ways. I will treat the following five explanations as the standard objections. The plutocracy objection appeals to an ideal of equality of opportunity to exert political influence. Specifically, it says that since rich people can better afford to buy votes and poor people have a greater incentive to sell their votes, vote markets will typically result in an unequal (or more unequal) distribution of the opportunities to exercise political power favouring the rich, fostering as a result a more unequal, unjust distribution (Downs 1957; Tobin 1970: 269; Hasen 2000: 1329–1331; Satz 2010: 102; Archer and Wilson 2014: 3; Archer et al. 2019). On the autonomy objection, markets in votes undermine the autonomy of voters by making them ignore ‘the full range of a party’s policies’ when they vote for it as part of a market transaction (Stokes 2007: 81, 96). ⁴ Rather differently, meaning critics allege that votes (like some other goods, such as sex and organs) have a cultural meaning that is violated by commodification, and that following such violation one can no longer express a commitment to one’s political community by voting (Karlan 1994: 1465; Sunstein 1994: 849). The republican objection asserts that voters have a duty to cast their vote for the candidate whom they believe will promote the common good, and adds that vote markets will predictably result in more people voting on the basis of other motivating reasons (Mill 1861 [2005]: 128; Sandel 2000: 118; Satz 2010: 103). Finally, the indifference objection claims that electoral markets erode voter interest in public affairs (Karlan 1994: 1472, 1999: 1714; Sunstein 1994: 849).

These objections are powerful, but I shall not defend this view here. I want instead to present what I call the relational egalitarian objection to vote markets. ⁵ Relational egalitarianism, in the specification I have in mind here, says that, through their political decisions, citizens ought to relate as equals qua holders of beliefs about political policies and candidates (e.g. beliefs about EU exit, or Trump). But, generally, vote markets result in significant numbers of citizens not voting based on their beliefs about policies and candidates. That is the relational egalitarian objection to such markets. I will argue that this objection is sound even in unusual situations where none of the standard objections to the commodification of voting apply. Hence, exploration of the relational respect in which legalizing vote markets is wrong results in a better understanding of the value of relating as equals in the political sphere. The provision of this improved understanding is an important aim of this paper.

Section 2 sets out a small-scale political decision-scenario in which the institution of vote buying appears wrong. Section 3 explains why it is indeed wrong, drawing on Niko Kolodny’s relational egalitarian account of democracy. Section 4 shows why the relational egalitarian objection I have canvassed does not imply that certain intuitively acceptable vote markets are wrong. Section 5 responds to six challenges to the relational egalitarian objection to vote markets.

It does not follow from the relational egalitarian account I propose, nor do I believe, that there are no circumstances in which the legalization of vote markets would be morally desirable all things considered. Plausibly, legalization would be a morally defensible option if an extremist free-marketeer credibly threatened to detonate a doomsday bomb if we did not. On my account, legalizing vote markets might not even always be pro tanto wrong – e.g. it might not be wrong in that sense if, admittedly unrealistically, a vote market were to be legalized, no one would trade votes. ⁶ Since my focus is on the moral wrongness of the institution of vote selling, I only discuss the ethics of standard individual acts of buying and selling votes when this casts light on the moral credentials of the institution of vote markets.

2. Vote buying: A simple interpersonal case

I now turn to a case of collective decision-making in a simple interpersonal situation where votes can be bought. The case is designed to offer intuitive support for the relational egalitarian objection to vote markets that I develop in greater detail in section 3. It involves direct democracy, so the decisions being made in it differ from the collective decisions made in large, indirect democracies. All the same, it has some of the central features of standard political decisions – e.g. that the relevant collective decisions have final de facto authority. The advantages of focusing on a small-scale decision-making are, first, that they do not tempt us to reify political decisions, and second, that what each individual does is likely to make a difference to what the collective decision will be. This makes the unfairness of the practice vote buying more transparent. ⁷ Consider:

Food: Three individuals, Seller, Upright and Buyer, have been brought together on a camping trip. They are in the wilderness and cannot leave before the end of the summer. They only have resources for making one set of tools – either bows and arrows or fishing rods – and they must decide collectively which set to make. They could rely on a meat diet. They could rely on a fish diet. Each will involve a distinctive set of risks and opportunities. Seller and Upright both prefer the fishing rods option, while Buyer prefers the bows and arrows option. They agree to vote on the matter, with each of them casting one vote. They agree that the option with a majority will be adopted. Moreover, whatever decision they make will be coercively enforced (say, by one from the winning side who will hold the only gun they have in the camp). Buyer predicts that it is unlikely his favoured option will get the most votes. Hence, he offers Seller a small sum of money if he votes for bows and arrows. Seller prefers the package of making bows and arrows and receiving a small sum of money over making fishing rods and receiving no money, so he accepts the offer. As a result, the majority vote for the bows and arrows option.

Intuitively, the fact that Seller, Upright and Buyer have collectively decided in favour of the bows and arrows option does not obligate Upright to accept the decision. Nor does the fact that the decision was made in the way it was made make it legitimate to enforce it if Upright resists it. ⁸ Why is that? ⁹

To account for these intuitions, I need to explain the difference between voting that is based on one’s ranking of the different collective decisions that a given election can result in and voting that is based on one’s ranking of the different voting-related actions available to each individual voter. Typically, the collective decisions a given election can result in are perfectly clear to the voter – they are the answers set out to the question one addresses in voting. For instance, in the UK’s 2016 EU referendum voters were asked which of two collective decisions they preferred: ‘The UK leaves the European Union’ and ‘The UK remains a member of the European Union’. ¹⁰ But while all UK voters were presented with the same set of possible collective decisions, they faced very different sets of voting-related actions. ^{11 , 12} Thus, in London, Sarah’s set had, say, three (salient) available options: ‘Go to the polling station in sunny weather to vote for Brexit’, ‘Go to the polling station in sunny weather to vote against Brexit’ or ‘Stay home and not vote’. And in Scotland, Jon’s set had four available actions: ‘Go to the polling station in torrential rain to vote for Brexit’, ‘Go to the polling station in torrential rain to vote against Brexit’, ‘Stay at Elinor’s place and not vote’ or ‘Return to my flat and not vote’. Roughly, voting-related actions are conjunctions of electoral actions, such as voting for or against a particular collective decision, and other non-electoral actions, such as venturing outside one’s home or receiving payment.

Typically, the items in a set of voting-related actions can be individuated in various ways (e.g. in terms of the exact time of the day at which one votes, whom one goes to the polling station with, what one does before and after voting) only some of which will be relevant from the deliberative perspective of the voter. For present purposes, the exact way in which the voting-related actions are individuated is not important. What is important is that, in a clear enough sense, a voter’s ranking of the different possible collective decisions might come apart from her ranking of the different possible voting-related actions she is presented with. For instance, had there been vote markets in the UK in 2016, Sarah’s highest-ranked collective decision might have been ‘The UK remains in the European Union’ while her highest-ranked voting-related action was ‘Go to the polling station in sunny weather to vote for ‘The UK leaves the European Union’ having been paid £100 to do so’. Here the rankings come apart in the sense that, if Sarah acts on the basis of the second she fails to vote for the collective decision she ranks highest. ^{13 , 14} In a nutshell, she votes contrary to her ranking of the different possible collective decisions.

Returning to Food, the set of collective decisions for the three campers contains two options: making fishing rods and making bows and arrows. ¹⁵ The set of voting-related options facing Seller, given Buyer’s offer to buy his vote, consists of three items: voting for making fish rods without being paid for so voting; voting for making bows and arrows without being paid for so voting; and voting for making bows and arrows and being paid for so voting. These three options do not correspond to the contents of the two possible collective decisions at hand – in other words, no collective decision is being made on whether Seller accepts payment for voting for making bows and arrows. The options in the set of voting-related actions are not, in themselves, matters of common concern. They are certainly not the matter of shared concern that is up for a collective decision (cf. Scheffler 2015: 32). ¹⁶

In selling his vote, clearly, Seller votes in a way that clashes with his ranking of the two collective decisions. For he ranks making fishing rods above making bows and arrows even though he ranks voting for the bows and arrows option with payment for doing so over the other two available voting-related actions. Hence, Seller’s vote does not reflect his ranking of the two possible collective decisions. It reflects his ranking of the available voting-related actions. ¹⁷ Polemically, we might say that his vote is not a political vote, since it does not reflect his stand on what the election is about, i.e. which collective decision should we make? Without dishonesty, Seller can explain: ‘I know I voted for ‘The UK leaves the European Union’, but I would much rather that we do not.’

With the distinction between voting that is based on one’s ranking of different collective decisions and voting that is based on one’s ranking of different voting-related actions in place, we can see why Upright might insist that the collective decision has no authority over him, and that it would not be justified to enforce it against his will. Seller, Upright and Buyer disagree about what they should collectively decide to do. Let us stipulate that none of them has a sound reason to dismiss another’s ranking – say, as inadequately informed. It therefore seems uncontroversial to say that they should resolve their disagreement in a fair manner. But what would this require? As a minimum (and one that is insufficient, as I shall argue in a moment), it will involve the parties to the disagreement treating each other as equals, qua individuals each of whom has a view on how to rank the different collective decisions, by giving everyone an equal opportunity to influence what the outcome will be.

Crucial here is the requirement of treating each other as equals qua individuals with views on how the different possible collective decisions rank. ¹⁸ It might be asked why treating each other as equals qua individuals with views on how the options in their sets of voting-related actions are ranked would not suffice for resolving their disagreement in a fair way. I offer two answers to this question. First, intuitively there is something wrong about the way the collective decision in Food is made. However, if the requirement to resolve disagreement fairly only involved the parties treating each other as equals qua individuals with views on how the options in their sets of voting-related actions rank suffices to resolve their disagreement in a fair way, this intuition would be ungrounded. Essentially, this is an abductive argument for the claim that treating each other as equals qua individuals with views on how the different possible collective decisions matters.

Second, if vote markets exist, then, to the extent that individuals buy or sell votes, they are not really addressing their disagreement when they vote. Their disagreement concerns the ranking of the possible collective decisions. Yet, they vote in the way they do only because of the way in which they rank the different voting-related actions available to them, and these sets vary from one individual to another. For that reason, they are not resolving their disagreement. A fortiori they are not resolving it in a fair way. Of course, they are following a procedure resulting in a decision that is collective in a clear enough sense, but the decision results from votes many – possibly all – of which do not reflect individuals’ views on the ranking of the different collective decisions, which is what the election is about. It is like responding to a question based on one’s assessment of what one’s audience would like to hear. Superficially, one responds to the question. But in deeper sense one does not, because one’s response is not guided by one’s belief about what the correct answer is. It is, of course, in the derogatory sense, a political response.

To support this second point, let me briefly expand on why treating one another as equals qua individuals with views on how the possible collective decisions rank is incompatible with offering to buy other people’s votes. In offering Seller money for his vote, Buyer does not relate to Seller as an equal qua someone who holds a certain view about the ranking of the collective decisions under consideration. True, he does that in the minimal, empirical sense that he acknowledges that Seller’s ranking is different from his own. Absent that, he would have no interest in buying Seller’s vote. However, he does not positively acknowledge the normative force of Seller’s view on the ranking of the available collective decision-outcomes. To do so would require him to want Seller’s view on the ranking of the collective decision to influence the outcome of the election on an equal basis with the rankings of others. Seller’s ranking of the collective decisions, unlike his ranking of the action of voting for a particular collective decision, is not changed because he is offered money for his vote. It is just that he now finds the action of voting for the collective decision that he ranks highest without an economic reward to be less attractive than some other voting-related action. Thus, Buyer treats Seller’s view as something that shapes his incentives to pay Seller, not as something that should influence what the collective decision ends up being on an equal footing with the views of others. Such a view is not exhibited by Buyer when he simply relates to Seller as the owner of a certain means of political influence that is worth more to Buyer than the money Buyer could offer Seller in exchange for the vote.

We can elucidate this distinction further by considering two responses to a colleague’s anger over one’s failure to keep an important promise. In the first response, one acknowledges one’s colleague’s anger only minimally. For example, one sees that the fact that the colleague is angry will affect what she will be willing to do for one in the future. One appreciates that she will be less willing to do one a favour, and that one should probably not ask her to do so. In general, in choosing means to realize one’s goals, one must now factor in the more limited reliance one can place in the colleague. In the second response, one acknowledges the colleague’s anger in a more than minimal way by acknowledging the normative force of the complaint underlying the colleague’s anger. Here, for example, one sees that the complaint gives one a reason to respond in certain ways to the colleague, one of which is apologizing. My contention is that the vote buyer acknowledges the vote seller’s ranking of the different collective decisions in the minimal sense but not in the fuller, normative-force-acknowledging sense.

Admittedly, Buyer respects Seller qua owner of his vote. For example, he makes no attempt to coerce Seller to vote differently and indeed he might think that doing such a thing would be wrong, and not just counterproductive. However, the issue here is not whether market transactions are compatible with any form of respect. It is whether they are compatible with respecting people qua individuals with views on how the available political decisions are to be ranked. Think here of how I might, in rather different senses, both respect and disrespect a person when I offer her money to have sex with me. I respect her qua someone who owns her body, but I do not respect her qua someone whose own views about whom she finds sexually attractive should determine who she has sex with – I might well know that the person thinks poorly of me in this respect. ¹⁹ This is one reason why most of us would be offended, and not feel respected, if we were offered money to have sex. Of course, we might accept that we are being respected in the first sense (in a way that a victim of rape, for example, is not), but that is not the further form of respect that we see ourselves as entitled to – a form of respect that motivates our indignation. I am suggesting something similar is going on in vote markets.

Most importantly, Seller and Buyer do not relate to Upright as an equal qua an individual with views about the relative attractiveness of the different possible political decisions since in concluding their transaction they do not acknowledge that his view should, as a matter of justice, count equally with theirs – in fact, their transaction ensures that it cannot do so. They are asking Upright to be bound by a decision he opposes even though others have treated their possession of an equal say as a tool to obtain private benefits. Given this, Upright can reasonably complain that the way in which the collective decision was made did not involve the other parties to it relating to him as an equal in the relevant respect. As a result, the disagreement was not resolved in a way that was fair to him, so it does not bind him – it would not be legitimate to enforce it against him. This complaint is illuminated and supported when it is tied more explicitly to a relational egalitarian justification for democracy. That is the task to which I turn in the next section.

Note, finally, that none of the objectionable features the standard objections to vote markets rely on need be present in Food. All three parties might have equal opportunities for buying and selling their votes. Indeed, Seller might even be richer and Buyer poorer than Upright (the plutocracy objection). Seller could be fully aware of what he is doing and able to reflect critically and rationally on his preferences etc. (the autonomy objection). It is possible that none of the parties to the camping trip comes to reasonably see voting on matters of common concern as not embodying a way of expressing one’s full membership of the small camping community as a result of Seller’s sale of his vote (the meaning objection). Buyer could reasonably believe that the bows and arrows option serves the common good of Seller, Upright and Buyer, and Seller could believe it is for the common good that he gets paid for voting because it cheers him up (the republican objection). It could be the case that none of the three becomes less concerned with matters of common concern as a result of Seller’s vote sale (the motivational objection).

Let us take stock of the argument so far. There is something intuitively wrong about vote-buying in Food. This is so even if none of the standard objections to vote markets apply, including the standard equality of opportunity for political influence-based objection (the Plutocracy objection). This fact, which is the point of departure of my argument, can be explained by the following set of claims: (1) that there is a distinction between voting on the basis of one’s ranking of possible collective decisions and one’s ranking of different voting-related actions; (2) that there is an election-related requirement to treat others as equals qua individuals with views on how the different collective decisions rank; and (3) that vote buying violates this requirement and that legalizing vote markets renders acts of vote buying much more common. The fact that (1), (2) and (3) together explain the intuitive objection to Food is an abductive reason to accept them. Although abductive arguments are not deductively valid, they can give us significant reason to accept the explanatory claims figuring in them. I have contended that our explanandum – why is it wrong to legalize vote markets? – cannot be fully explained by any of the standard objections.

3. The relational egalitarian account of democracy

In this section, I present Niko Kolodny’s relational egalitarian account of the value of democracy. I shall try to show that the account is incomplete because it leaves open the respect in which we are to relate as social equals (Lippert-Rasmussen 2018: 61–93). Once the missing details are supplied in a way that enables the account to underwrite a desirable democratic ideal, we see that this version of relational egalitarianism meshes well with the complaint against vote markets articulated in the previous section. If I am correct, two theoretical benefits emerge. First, the most plausible version of the ideal of relational equality in relation to democracy becomes clearer. Second, the objection to vote markets articulated in the previous section is strengthened by being tied to an attractive general account of democratic authority and legitimacy.

Kolodny notes that most of us see democracy as a desirable way of making binding collective decisions. In his view, the justification for democracy is elusive because theorists assume they must ‘search for a more involved, hidden explanation’, whereas in fact the justification ‘lies more or less in plain view’ (Kolodny 2014: 288). Essentially, that justification consists of two claims. First, relations of social equality are valuable, and ‘we have reason to avoid relations of social superiority and inferiority for their own sake, not simply as a symbol of, or means to, something else … To put the thought positively: insofar as we are to have ongoing social relations with other moral equals, we have reason to relate to them as social equals’ (Kolodny 2014: 299). Second, democracy in the form of ‘equal opportunity for influence over political decisions to which [individuals] are subject … is a particularly important component of relations of social equality among individuals’. It is an important component because ‘political decisions characteristically involve force, for example, through commands ultimately backed by threats of force’ (Kolodny 2014: 307).

This justification is appealing. First, the ideal of relating as equals certainly is widely shared, and it is plausible to hold that equal opportunity to influence political decisions is an important component of it. Second, because the justification takes democracy to be non-instrumentally valuable, it can explain why, in cases where the ends that democracy is often thought to promote are actually better served by non-democratic decision procedures, there would still be something bad about deciding matters non-democratically.

Still, Kolodny’s justification of democracy leaves open what we are to say about vote markets where they preserve equal opportunities for influence over political decisions – e.g. because everyone has equal opportunities to sell and buy votes. Acknowledging the possibility of such vote markets, Kolodny suggests they are ‘inappropriate … even when we stipulate that equality is not an issue … There may be sufficient objection to reject a market in votes before we reach a concern for social equality’ (Kolodny 2014: 334n51). ²⁰ There might indeed be such objections (the standard objections are such). However, on the view proposed here, unlike the one Kolodny seems drawn to, these objections can also be grounded in the very concern about social equality to which Kolodny appeals. Or so I shall now argue. ²¹

Suppose that, when you say to a group of people that they should relate to others as equals, they say, ‘Equals in what way?’ If you say, ‘I don’t know. Just relate as if you are equals!’ it will be unclear what, exactly, you are asking the group to do. You need to explain two things. First, relating as equals always involves the parties to the relation enjoying equality in certain (as I shall say) dimensions. Obviously, some kinds of relationship are paradigmatically unequal – say, the relation between master and slave. And some relationships are models of equality, where we can take for granted that partners to the dialogue will have an implicit grasp of the dimensions in which the parties are equals – say, moral personhood.

The second thing you must explain is this. The question ‘Equal in what dimension?’ differs from the question of how one relates as equals in a particular dimension (though answers the former might have direct implications for how one should answer the latter). We can agree with Kantians, say, that most grown up human beings are equals in their capacity to be moral agents (a dimension question), yet disagree over what human beings need to do to respect this dimension of equality (a how question). Should they give equal weight to each other’s interests? Implement equality of de facto authority? Promote equality of opportunity for welfare?

Next, some dimensions in which we can relate as equals are more important than others. It is plausible to say that relating as equals in relation to the dimension of moral agency is valuable. It would be silly to suggest that there is something amiss to the extent that we do not relate as equals when it comes to the dimension of ability to play Scrabble. Moreover, it seems that the dimensions in which we should relate as equals, and how much weight these dimensions have, on our most attractive understanding of the ideal of social equality, varies across contexts (cf. Scheffler 2015).

With these general points on the table, I want to suggest that one dimension in which we should relate as equals in the political sphere is as individuals with views about what our collective decisions should be – a theme I introduced in section 2. ²² Elizabeth Anderson – arguably the most prominent, contemporary relational egalitarian – appears to adopt this suggestion: ‘Democracy is here understood as collective self-determination by means of open discussion among equals, in accordance with rules acceptable to all. To stand as an equal before others in discussion means that one is entitled to participate, that others recognize an obligation to listen respectfully and respond to one’s arguments’ (Anderson 1999: 313). The arguments here are arguments for or against the different political decisions at stake and not for or against individual voting-related actions, since the relevant discussion concerns ‘possible object[s] of collective willing’ (Anderson 1999: 330). By definition, the fact that I am paid a small sum of money in a private market transaction for voting for a particular political decision is not an object of collective willing. The fact that I will benefit financially from voting in favour of a particular collective decision probably would be a reason for me to vote in favour of it. But it is not a fact that can constitute a reason for (or against) the collective decision, which I then vote for, and therefore it is not something my fellow citizens can have an obligation to listen, and respond, to respectfully. As a fact, it is irrelevant to the issue at hand, which is what we should will collectively. Accordingly, vote buying is incompatible with Anderson’s ideal of democratic equality – a particular version of the ideal of relational equality. ²³

What we have done is to specify the ideal of relating as social equals in a political context in more detail by bringing in an account of the wrongness of vote markets that is internal to the ideal of relational equality and thus to our account of the value of democracy. To the extent vote buying takes place, the significance of collectively binding democratic decisions as a constituent part of valuable relations as equals is reduced. ²⁴ This relational account is attractive, because to know what we value when we value people relating as political equals we need to have in mind one dimension or other of that relation; and if our specification is such that the ideal of relating as equals is entirely compatible with vote markets, it follows that the ideal is quite different from what people think they value when they value relating as equals in a political sphere.

I want to elaborate the last point in two respects. First, one might concede that it happens to be the case that it is in our capacity as people with views about the best ranking of the different possible collective decisions that we – and relational egalitarians – think we should respect each other in the political sphere. But one might then ask why – from any perspective, relational egalitarian or otherwise – it is not in some other capacity, such as our capacity as owners of votes, that we should respect each other? That is a good, and a deep question, which can be asked in relation to other spheres of respect too. ²⁵ One important line of thought here is that if we think that respect in relation to collective decisions should be conceived along the lines of respecting people as property owners, it might become hard to see why we should be concerned to relate as equals in the political sphere in the first place (in the light, for example, of the fact that people vary considerably in terms of their property-related interests). Arguably, other things being equal, people who will die soon have less sizeable, or weighty, property-related interests in the collective decisions to be made than people who will be around for longer. Hence, if we were to respect people in their capacity as owners of property in the political sphere, it would seem natural to conclude that the weight of a person’s vote should depend on the number of years she has left to live. Few subscribe to this view. Setting special cases, such as people who are severely demented, aside, everyone has (or can acquire) views about the ranking of different collective decisions. To have such a view is to have an opinion on an either-or issue, not a scalar one. In this way, the egalitarian shape of political respect seems tied to the capacity in which we should relate as equals (Rawls 1999: 441–445; Carter 2011: 548–550). ²⁶

Second, on what we might call the default view of relating as equals the fact that two parties do not relate as unequals entails that they relate as equals. This view is false. One way to see this is to consider two agnostics about the equal moral worth of all individuals. They do not treat each other as unequals – they avoid doing that, because each wants to avoid treating the other as inferior just in case it turns out that individuals possess equal moral worth. Nor, however, do they affirm their equal standing, e.g. they do not defend the view that everyone are equals if this issue comes up in a public discussion – they feel, understandably given their agnosticism, that this would be disingenuous.

There is a similar reason why egalitarian vote markets do not involve relating to one another as equals qua individuals with views on what the collective decisions should be. Suppose everyone has equal opportunities to influence political decisions. Specifically, everyone has an equal opportunity to buy the votes of others and everyone, say, acknowledges that just as they might sometimes win a vote as a result of buying votes, sometimes they might lose a vote following the vote buying of others, and they are happy to play by those rules. Here it might be true that the voters involved do not relate as unequals qua individuals with views on what the collective decisions should be. No one thinks that others should heed his beliefs about desirable collective decisions more than their own. However, because the default view of social equality is false, the equality of opportunity to exert political influence (e.g. by vote buying) does not per se ensure that the voters relate as political equals in the respect we care about when we care about democracy. We might say that in the scenario of egalitarian vote buying markets envisaged the voters do not relate as unequals qua individuals with views on what their collective decisions should be. For it is not as if anyone relates to others as if his own views on what the collective decisions should be somehow counts for more than that of the others. However, as the agnostics example given above brings out, this does not guarantee that the voters relate as equals qua individuals with views on what the collective decisions should be, and the existence of vote markets is likely to ensure that a significant proportion of votes will not relate to others in this way. ²⁷

To sum up the argument in this section, in the relational egalitarian ideal of citizens relating to each other as equals the dimension in which citizens should be so related needs to be specified. One such dimension, in voting scenarios, refers to the voters’ individual rankings of different available collective decisions. Given this, relational egalitarianism meshes well with the set of claims (1)–(3) that I argued in section 2 account for the intuition that the practice of vote buying in Food is wrong. Indeed, (2) in that set – that there is an election-related requirement to treat others as equals qua individuals with views on how the different collective decisions are to be ranked – is simply a way of specifying the relational egalitarian ideal. This means that the ideal of relational equality can explain why vote markets are morally wrong. This, in turn, strengthens our belief in the ideal of relational equality, because the ideal gains credence by explaining the unease that most of us feel about achieving political influence through vote buying even in cases where none of the five standard objections apply.

4. Acceptable vote markets

Is the relational egalitarian objection to vote markets too strong because it implies that such markets are wrong in all contexts and in all forms? No. To put this worry to rest, I consider three cases where vote markets do not look wrong.

Many Decisions: Sociable, Indifferent and Gourmet are making collectively binding decisions on a wide range of issues, including what to have for dinner, where to erect the tent, and who should collect firewood, in a situation otherwise similar to that in Food. Each issue is decided by a majority vote, with each of them casting one vote. Before they vote on any of these matters, however, Sociable and Gourmet trade votes. Sociable does not care much about what to have for dinner, but strongly dislikes walking alone in the wood. Gourmet is almost indifferent on the latter, but cares strongly about what is for dinner. Sociable ‘sells’ his dinner vote in return for Gourmet’s firewood vote.

In this case there is a restricted market for votes – ‘restricted’ because it is a barter market where you can only exchange votes for votes. Setting aside worries about permanent minorities, it is arguably permissible for Gourmet and Sociable to trade votes in this way.

The view that restricted vote markets are not wrong coheres with the relational egalitarian objection to vote buying. In such markets, vote buying does not change the fact that individuals cast their votes on the basis of their rankings of the different collective decisions (e.g. Gourmet casts his two dinner votes on the basis of his ranking of the collective decision regarding dinner) and not on the basis of their rankings of their different available voting-related actions. ²⁸ Voters trade votes to ensure that certain collective decisions are taken on the matters that they care most about (cf. Hasen 2000: 1343). ²⁹

A second case:

All Decisions Up for Sale: Many collectively binding decisions need to be made on this Food-like camping trip. On each issue the three parties involved can spend a certain amount of money (including zero) on the collective decision they rank highest, and the decision which has attracted the most money will be the outcome of the ‘election.’ The three friends have an equal amount of money allocated exclusively for ‘voting’.

This an extreme version of vote buying that does away with votes (as an intermediary) and thus vote markets altogether. Vote markets have been reduced to decision markets, as it were. In vote markets you buy votes and then use ‘your’ votes to influence political decisions. In the present scheme you spend money directly to try to influence political decisions. You can use your votes in a way that fits the issues you care most about. This has the advantage that the decisions made will reflect people’s preferences better than they would if decisions were tied together in bundles. Arguably, this scheme of vote buying, if that is what it is, is not pro tanto wrong.

Suppose that this is the case. That would be consistent with the relational egalitarian objection to vote markets, since the scheme involves no incentives to vote, even partly, based on one’s ranking of different voting-related actions rather than solely on the basis of one’s ranking of the different collective decisions. In effect, All Decisions Up for Sale is simply a money-mediated version of Many Decisions.

Consider, finally, vote buying in relation to publicly held companies: ‘[Buying votes] may be illegal in politics, but in the corporate arena, it is a legitimate, if controversial, strategic tool’ (Lan and Leracleous 2005; cf. Hasen 2000: 1348–1354). I shall assume that the legality of buying votes from stockholders reflects the fact that the stockholders, at least, feel the practice is not morally wrong. If we now agree, also, that vote buying in Food is impermissible, we might ask what makes publicly held companies and Food different.

I see two differences. First, a political community (of which Food etc. are supposed to be micro versions) is an egalitarian enterprise. A publicly held company is not. If you own more stock, you have a greater say than other stockholders, and if you want a greater say, assuming that you have the money required, you can simply buy more stock. If you can do that, what is the problem with acquiring a greater say simply by buying the votes of other stockholders? By contrast, all participants in a political community are supposed to have an equal say. ³⁰ If you want a greater say, you cannot literally buy people and thereby acquire their right to vote.

Second, with a public company, people are in it for the money. Things are more complex in a political community. Typically, the people who make up a political community will not be uninterested in promoting their self-interest. But most want to promote their interests on terms that are fair to others. In this, there is a matter of common concern that is not present in the stockholder case, where there is no justifiable expectation that people will decide how to vote, as stockholders, only based on their rankings of the different collective decisions. Relating as equals means different things in different contexts. In the context of a publicly held company there is no expectation that one will relate to others as equals qua individuals with views on the best ranking of the different collective decision-outcomes.

It appears, then, that the relational egalitarian explanation of what is wrong about vote markets does not rule out too much. It can be combined with the view that in some contexts (Many Decisions, All Decisions Up for Sale and publicly held companies were those considered) vote buying is not wrong.

5. Challenges

I shall now consider six challenges to the relational egalitarian objection to vote markets. The first sees my use of Food etc. as problematic (cf. Freiman 2014):

(1) Selling and buying votes in Food etc. is morally no different from other scenarios where people vote for reasons that do not pertain specifically to the ranking of the collective decisions – e.g. a scenario where Seller votes for the bows and arrows option to win the approval of Buyer, whom he admires. ³¹

(2) In these other scenarios legalized voting for reasons not pertaining specifically to the issue being voted on is acceptable.

(3) Hence, legalized vote selling in Food etc. is acceptable.

To respond to this, we need to distinguish individual market transactions involving votes from the act of legalizing vote markets. Where the latter is concerned, there are important differences between making vote markets illegal and making it illegal to vote so as to win the approval of others, and so on. The enforcement of the first of these prohibitions is easier and involves less intrusive means than enforcement of the second (which in practice can be achieved through secret voting). Moreover, vote markets are likely to be consequential for electoral outcomes, and in ways that harm specific groups. The fact that some people vote to impress others is more likely to be electorally inconsequential. I suggest these differences justify us in making vote markets illegal while declining to do the same about voting with other, nevertheless improper, motivations. They do so even if, ceteris paribus, voting for other reasons not pertaining to matters of common concern (e.g. to win the personal approval of one’s peers) is morally no different from voting for money. ³² This means that (1) and (2) do not entail (3). Even if two kinds of acts are morally no different per se, there might be reasons to prohibit one and not the other. ³³

It might be replied that while pragmatic concerns about violations of privacy and affecting electoral outcomes are relevant to what the content of the law should be and might even justify treating vote markets differently from approval voting, these concerns do not reflect the relational egalitarian ideal. However, a reply that countered the present challenge to my relational egalitarian account would have to be grounded in that ideal. ³⁴ I wish to indicate two responses to this reply. First, are relational egalitarians disqualified from accepting that all forms of voting that are not guided by one’s ranking of the available collective decisions are wrong and appealing to values other than relational equality to explain why law should treat them differently? My own view is that this is doubtful. Second, plausibly, the banning of a vote market will make a big difference to the degree to which vote buying takes place. The prohibition of approval voting is much less to affect the degree to which approval voting takes place – e.g. because approval voters can almost always vote differently from the way they say they have voted, making it virtually impossible to enforce a prohibition on approval voting (see footnote 5 also). ³⁵ Given this contrast, the ideal of relational equality provides at least some (additional) justification for the notion that the law should treat vote markets and approval voting differently.

A second objection contends that my account relies on a republican understanding of voting on which votes should reflect voters’ views on the common good. Votes should not be ‘means of pursuing their economic interests’ (Freiman 2014: 771; Taylor 2017: 327, 317; cf. Sunstein 1994: 849n49; Thompson 1995: 29; Sandel 2000: 108; Satz 2010: 103). We can note first that this view of voting is controversial. The standard liberal conception of voting allows voting on the basis of self-interest and sees elections merely as mechanisms for aggregating individual preferences into a collective choice.

We should also note that the proposed relational egalitarian objection, though consistent with the republican complaint (see section 1), relies on a view of voter motivations that is far thinner than that generally adopted by republicans. The former makes no claims about the positive basis on which one ranks different collective decisions. It simply says that one should not vote in a way that clashes with one’s ranking of the different collective decisions. I have already noted that this requirement is neutral based on such a ranking, and in particular on whether the ranking is based solely on the degree to which the different items promote one’s self-interest. ³⁶ Hence, the question whether the relational egalitarian objection to vote markets differs from the question whether the liberal or the republican view of voting is correct. Admittedly, the relational egalitarian objection does involve limiting the permissible motivations for voting in the way one does, and this might reasonably be said to be republican in a sense. But the sense is itself limited – much thinner than the usual one and therefore part of a view that is less controversial. For instance, the relational egalitarian objection is consistent with offering people financial incentives to vote. Republicans would have no truck with this, as they require people to vote out of a concern for the common good (cf. Hasen 2000: 1358–1359; Lippert-Rasmussen 2010).

The fact that the relational egalitarian objection does not appeal to the claim that ranking possible collective decisions on a purely self-interested basis is morally wrong does not imply that ranking options in that way is not wrong – on relational egalitarian or other grounds. The point is that my objection to vote markets does not require me to take a position on this matter. Thus, even if, say, my relational egalitarian objection commits me to standard republican objections to vote buying, this does not show that the two objections are not distinct.

A third objection says that my use of Food etc. makes dubious assumptions about what counts as a matter of common concern. In Food, for instance, I assume that Seller’s being paid a small amount of money is not a matter of common concern. But could it not be so, in some sense? If, for instance, the three friends all happen to be welfare egalitarians, then by their own lights they should see Seller’s benefit as a matter of common concern given that, as they see it, their collective decisions are designed to bring about a just distribution of welfare.

The welfare egalitarian version of Food might well show that, given certain assumptions, my example will not work. However, it does not show that there is anything wrong in the spirit of my appeal to Food etc. It is remains inaccurate to say that when Seller sells his vote, he treats it as a ticket to voicing his view on what the relevant collective decision on the matter of common concern should be. Rather, he uses it as an opportunity to realize a personal aim. This is so even if by selling his vote Seller brings about the equal distribution of welfare that the three friends agree is desirable. Hence, my objection still applies (see also footnote 10).

A fourth objection imagines a version of Food in which everyone has equally good opportunities to sell and buy votes. In cases like this, the objector continues, surely vote markets are no obstacle to collectively binding decisions being made in a fair way. ³⁷ Even if Buyer gets to decide which food they should seek, Upright does not have a justified complaint if he too could have bought Seller’s vote and his opportunities for so doing were as good as Buyer’s.

Here, my sense is that there is something wrong about vote markets even under perfect equality of opportunity for influencing political decisions. Even under such circumstances, people fail to relate to one another as equals qua individuals with views about what the collective decision should be. This reply has considerable force, given, as Kolodny notes, that vote buying is ‘inappropriate’ even in the absence of issues of inequality. However, I can see that some might concede this but wish to argue that under some circumstances vote markets are going to be compatible even with the requirement to relate to one another as equals qua individuals with views about what the collective decision should be (recall my second clarificatory remark in section 1). That takes me to the fifth objection.

This begins with the observation that people could unanimously decide to allow vote markets, and then claim that if they did there would be nothing wrong about legalizing the practice. We can imagine, Brennan-style (Brennan 2011: 148), that while most of us would not feel bound by a collective decision that results from vote buying, we might have felt differently about it and decided to let laws reflect this view of ours. The mere fact that we think about our political association in the way we do does not show that there is anything morally wrong about thinking differently about it. Specifically, there would – so the challenge goes – be nothing wrong about allowing vote buying if everyone freely agrees to be bound by collective decisions involving vote buying and there is perfect equality of opportunity to trade votes. Surely, the Brennan-style challenge continues, there would be nothing wrong about a marriage where the spouses paid each other to have sex at times and in ways of their choosing provided that both parties consented to this. Indeed, introducing vote markets in this way would arguably be one way in which we could relate to others as equals qua holders of views about what should be the object of a specific collective decision, in casu how to make other collective decisions.

In my view, there might well be something wrong with vote buying even though all voters consent to the institution. However, with certain provisos – the obvious one being that people sufficiently understand what they are consenting to and can determine accurately enough the likelihood of the various outcomes that may result – I agree that the Brennan-style challenge looks powerful. ³⁸ The main point I wish to make, however, is that, even if we accept that the institution of vote buying is morally unproblematic if everyone agrees to it, the significance of the present objection is limited in a way that teaches us an insightful lesson. The objection presupposes that it is consensually decided to legalize vote markets. The natural question is: How was this done? Were market transactions in votes permitted when a collective decision was made on the permissibility of vote markets. If ‘yes’, one might reasonably reject being bound by the outcome of that decision. If ‘no’, then the present challenge is consistent with the claim that the permissibility of vote markets is parasitic on the normative force of collective decisions made in the absence of vote markets. ³⁹ In support of this claim, we can note that the situation is normatively asymmetric if some people do not consent to forbidding vote buying. These people are bound by a collective decision not to permit vote markets even if they think that market transactions in votes should have been permitted at that stage. ⁴⁰ The binding force of collective decisions untainted by vote buying is not conditional on prior unanimous consent, unlike the binding force of collective decisions involving vote buying.

The sixth and last objection to my account of what is wrong with vote markets suggests that, at its core, it is not egalitarian. Consider a quasi-Burkean hierarchical society in which people have unequal opportunities to influence political decisions based on some public assessment of their political competence. If you are judged to have a high level of political competence your vote counts for more than those of people deemed politically incompetent. Someone who subscribes to a relational inegalitarian view according to which people ought to relate as unequals (whenever fitting, as is, according to the relational inegalitarians in question, often the case) qua holders of beliefs about what the object of the collective decision should be, could object to vote buying on grounds exactly analogous to those I have set out. In other words, they could say that in selling and buying votes people do not relate to one another as they should. Ex hypothesi, they do not relate as unequals in the relevant dimension whenever that is fitting. An incompetent person can buy the votes of a highly competent person, and when this happens the former does not relate to the latter as a superior qua holder of a view on the ranking of possible collective decisions.

It is true that relational inegalitarians, no less than relational egalitarians, may disapprove of vote markets in the way described. Moreover, the reasons for this are identical in one respect: they are both grounded in a certain view of the proper social relations between individuals. However, we should not forget that in one respect the objections are very different. Relational egalitarians favour relating as equals, and inegalitarians do not. That difference suffices to defeat the present challenge.

6. Conclusion

Vote markets are widely felt to be wrong. The argument presented in this article has proceeded on the basis that in establishing that a particular account of the ideal of democracy explains why legalizing such markets are wrong, we also provide support for that account. I have argued that a version of relational egalitarianism – one specifying the dimension in which one should relate as equals in the political sphere – supports and explains the intuitive unease that most of us have about vote buying in contexts such as Food. In this version, the relational egalitarian objection is that vote markets typically involve voting in a way that clashes with one’s ranking of the set of possible collective decisions and instead voting on the basis of one’s ranking of different private acts. That, it is argued, does not involve relating to others (or oneself) as equals qua holders of views about the ranking of different collective decisions. To the extent that this line of argument is forceful, it supports the relevant relational egalitarian account of the ideal of democracy.