2. What Is Democracy?

First, I will explain the understanding of democracy that I will be working under in this article. In a democracy, the people collectively rule, as equals. Adam Lovett and Jake Zuehl recently defined democracy “as a political system with two components: first, citizens all have equal political power and, second, this power collectively largely determines the state's decisions” (Lovett & Zuehl, Reference Lovett and Zuehl2022, p. 483). Lovett and Zuehl's definition is given stipulatively, as is often the case with definitions of democracy, but there are good reasons to accept two-part definitions of this kind in comparison to the alternatives on offer by democratic theorists. Some, such as Joseph Schumpeter and other proponents of ‘leadership democracy,’ have defined democracy simply as a political system where competitive elections determine who is in a position of political leadership (Schumpeter, Reference Schumpeter1976, p. 269). While this definition makes it very easy to distinguish democracies from non-democracies, it has the implication that there is nothing ‘undemocratic’ about restrictions on suffrage based on race, sex, or property requirements. For many democratic theorists, describing democracy in this way will not capture what is valuable or morally significant about democracy. It is also ill-suited to describing smaller-scale decision-making procedures that involve members directly voting on issues rather than electing representatives as democratic.

Some have defined democracy simply as a decision-making procedure in which the parties to the decision are in some respect equal. For example, Thomas Christiano claims that democracy is “a method of collective decision making characterized by a kind of equality among the participants at an essential stage of the decision-making process” (Christiano & Bajaj, Reference Christiano, Bajaj, Zalta and Nodelman2024). This solves both problems I have described with Schumpeter's account in that it requires political equality and does not require that democratic decision making necessarily involves political representatives. Others, however, such as David Estlund, have argued that descriptions of this kind entail that democracy is compatible with decision-making procedures in which the parties are equal because no one has power over the outcome — such as flipping a coin to determine what shall be done (Estlund, Reference Estlund2009, p. 6). Two-part definitions, like Lovett and Zuehl's, require not only individual equality as Christiano and others do, but collective power over outcomes as well.

My argument is concerned with the relationship between individuals and collectives in a democracy, and so looking at Lovett and Zuehl's two-part definition through that lens, we can arrive at two points, which together define democracy for my purposes in this article. In a democracy (1) no citizen individually is politically superior to any other. I will refer to this condition as ‘democratic equality.’ This is a point of contrast between democracy and other systems of government, such as monarchy: in a monarchy, the monarch is politically superior to all other citizens. And (2) in virtue of collectively determining political outcomes, the public collectively is politically superior to any individual citizen. This is again a point of contrast with monarchy, in which the monarch is superior to the citizens collectively. The problem which I shall articulate in the following section is that when the politically superior collective public is embodied in an individual citizen, that individual citizen is politically superior to their fellow citizens, and, per (1), this superior status is at odds with democracy.

It should, of course, be noted that equality is not the only important consideration in determining what type of government we ought to live under. There may be important reasons to prefer non-democratic forms of government. As Schumpeter points out in arguing for his leadership conception of democracy, a democracy that causes famine and persecutes religious minorities is a worse government than an autocracy which does neither (Schumpeter, Reference Schumpeter1976, pp. 241–241). Epistemic democrats widely agree that, whatever other legitimacy requirements there are for government, if an otherwise politically legitimate democratic government is too incompetent at governing, then it is not legitimate. Furthermore, there is more to democracy itself than equality: as we have seen above, only attending to equality is the mistake that Estlund charges Christiano and other democratic theorists of making. Neither is there any problem with inequalities per se — it may be both unnecessary and unhelpful to eliminate many forms of inequality.Footnote ³ However, this does not show that my allegation that elected representatives are the superiors of their constituents is without consequence. Certain specific forms of inequality, such as political inequality, are objectionable in a democracy. The forms of superior status that I allege the institution of elected political representatives creates are, as I shall show, undemocratic, and insofar as we are committed to democracy, then we have a powerful reason to want to solve this problem.

3. The Avatar of the People Problem

In this section, I will outline the avatar of the people problem for elected political representatives. The problem is that because elected political representatives stand-in for their constituencies collectively, political representatives are superior to their constituents individually, and so the superior status of elected representatives is undemocratic. In this section, I shall explain why we should think that politicians are the political and social superiors of their constituents. My explanation begins with a case study that brings the avatar of the people problem into sharp relief. With the case for this critique of elected representatives established, I will be able to make my case for legislative mini-publics.

Part of the democratic tradition in the United States are informal ‘town hall meetings’ attended by elected representatives and open to their constituents, at which elected representatives answer questions and respond to comments from their constituents. In the summer of 2017, and in the wake of attempting to repeal popular legislation (the Affordable Care Act), many members of Congress from the Republican party eschewed scheduling the expected town hall meetings. Those who did hold these meetings confronted very angry constituents demanding that their representatives change their conduct. In terms of motivating their representatives to switch their position on the legislation in question, these pleas were ineffective. As an example of an interaction at one such meeting, representative Tom ReedFootnote ⁴ faced a town hall meeting full of his constituents who were expressing their displeasure with his attempt to repeal the Affordable Care Act. Reed asserted that the objections of his constituents in the room were unpersuasive because he acts for the constituency as a whole. His view represents what the people of his district want, not the views of the irate constituents in the room. Reed was briefly stymied when a constituent commented that, according to a recent poll, only 17% of his constituents agreed with him that the Act should be repealed. He responded to this by claiming that his actions were taken for the sake of his constituents as a whole (to promote good outcomes in healthcare), even if these were not actions his constituents would want him to take. Senator Mitch McConnellFootnote ⁵ did not hold a town hall meeting during this time, but did move through a crowd of approximately 1,000 protesting constituents on his way to address local business leaders in Lawrenceburg, Kentucky. Once inside, he said to the assembled business leaders that he was proud of the demonstrators for expressing their views, but that they had “had their shot,” and “Winners make policy and the losers go home” (Lartey, Reference Lartey2021).

One way to interpret these stories — indeed the way that Reed asks us to interpret them — is about a disagreement between a small group of thousands of individual citizens, and political representatives who stand-in for millions of citizens. Reed and McConnell politically represented approximately 3 million citizens at the time.Footnote ⁶ Naturally, in a democracy, if citizens are equal, then when thousands are in conflict with millions, all else being equal, the millions should get their way.

But this interpretation is uncompelling. Intuitively, there seems to be something defective about the way that these political representatives are relating to these people they espouse to represent. Here is an alternative interpretation: while I am sure no offence was intended by the remarks, they have the effect of rubbing their constituent's noses in the superiority of their political representatives, in both power and in status. Reed and McConnell, in those stories, assert that how these constituents feel about the Affordable Care Act does not really matter. Instead, because Reed and McConnell stand-in for the collective people, how they feel about it does matter. Reed first argues that his political judgement is more important than the judgement of everyone else in the room, and then doubles-down, arguing that his political judgement is more important than the judgement of all of his constituents. Reed is not the equal of any mere citizen. Instead, he acts like the political guardian of the constituency. Who are these protestors to tell him how he should protect the interests of his ward? McConnell likewise asserts that he isn't interested in speaking with his protesting constituents because they lost and he won the power to make policy (Lartey, Reference Lartey2021). McConnell has been invited to address the business leaders of Lawrenceburg, and the protesting citizens of Lawrenceburg have not been, because what McConnell has to say is politically important, and what the protestors have to say is not.

As I shall demonstrate in the remainder of this section, the problem is not the conduct of these representatives (though it might reasonably be construed as rude). The conduct merely draws attention to the superior status of representatives to their constituents. This superiority itself is the problem. As I shall show, in respect to this superior political status, Reed is no different from any other congressman. The issue is likewise not that the protestors did not get their way. Instead, the issue is that the protestors confront Reed as a political and social superior rather than a political and social equal.

As I have noted above, in a democracy these two things should be simultaneously true (1) the public collectively is politically superior to any individual citizen, and (2) no citizen individually is politically superior to any other. The avatar of the people problem is that, by having an individual representative stand-in for the public as a whole, they are made politically superior to their fellow citizens. The solution, as I shall argue in Section 4, is to have a deliberative mini-public stand in for the public collectively — by representing the public in this way, no citizen individually will be the political superior of any other.

3a. Political Representation

To see why Reed and McConnell are in a certain respect the superiors of their constituents and why this is democratically problematic, we will have to investigate what political representation is supposed to mean in a democracy.

In a conventional representative democracy, the people are understood by democratic theorists to rule through their elected political representatives (for example, see Manin (Reference Manin1997); Pitkin (Reference Pitkin1972); Urbinati (Reference Urbinati2006). These political representatives are generally understood to represent their constituents collectively. However, colloquially, we also say that each individual constituent has an elected representative: the same person as the representative of the collective political constituency to which they belong. So, for example, Arif Virani is the elected representative for the Parkdale-High Park riding (constituency) in Canada,Footnote ⁷ and because I live in the Parkdale-High Park riding, he is my political representative in particular.

Political representation is often described by democratic theorists as an agent-principal relationship, where power is delegated from the principal to their agent, and the agent uses those powers according to the principal's direction and for their sake.Footnote ⁸ Niko Kolodny, for example, defines the principal-agent relationship in this way:

I will tentatively accept this agent-principal conception of representation because, as I shall argue below, whether our political representatives can be said to be our agents in the above sense, as individuals, has moral significance independent of whether it is necessary for political representation.

In a conventional representative democracy, Kolodny's description of the political representation relationship is a good description of the relationship that exists between a collective constituency and their elected representative.Footnote ⁹

Importantly, however, it is not a good description of the relationship between an individual constituent and their elected representative. In some instances, elected representatives act directly on behalf of particular constituents as individuals, by providing help with lost passports, family immigration issues, issues with business licencing, and other such things.Footnote ¹⁰ However, when we consider the ways they act as the agents of their constituents collectively as legislators, it is evident that they do not represent their constituents individually in this same way.

Looking back to Kolodny's six features of the principal-agent relationship, we can see important disanalogies between the representative as they relate to their collective constituents and their individual constituents. Concerning (i), it is plausible that the entitlement to vote in parliament is a power that belongs to the constituency and is delegated to their elected representative, but it is certainly not a power that belongs to an individual constituent. Likewise, concerning (v) and (vi), the constituency collectively can replace their elected representative and control the selection of their representative, but an individual constituent has no such power. Arguably, as a result of these facts, representatives are not answerable to individual constituents (ii), nor are they obliged to exercise their powers on behalf of the interests of any individual constituent (iii), nor are they obliged to act on the expressed will of any individual constituent (iv). Because individual constituents are in political conflict with one another, often a representative cannot act on the expressed will of some without failing to act on the expressed will of others.

Relatedly, if we try to understand elected politicians as representing each of their constituents individually in the use of their legislative powers, then on the agent-principal model of political representation, they have intolerable conflicts of interest. Constituents are frequently politically at odds with one another. Having one agent politically represent both sides of these conflicts would, in an important sense, be akin to having the defence attorney and the prosecutor in a criminal trial be the same person, or akin to companies negotiating a business deal being represented by the same negotiator.

For these reasons, Virani is not my political representative in this sense at all. He represents a group to which I belong, but he does not represent me. We should reject the idea that political representatives are the agents of their constituents as individuals in a representative democracy. Individual citizens qua individual citizens go without political representation, at least according to the common representative-as-agent conception.

But perhaps the representative-as-agent conception isn't true to how the word is commonly used.Footnote ¹¹ My purpose, however, is not to make a claim about the meaning of the concept ‘political representative.’ Rather, giving up on the claim that political representatives are agents of their constituents as individuals has independent significance. This is because doing so creates the avatar of the people problem.

As I have said above, the avatar of the people problem, roughly described, consists in the idea that because political representatives speak and act on behalf of many citizens without being their agents, they are politically superior to ordinary citizens in a way that cuts against the political equality of citizens. As I have earlier described, in a representative democracy, equal citizens rule collectively through their representatives. But, as I shall argue, the fact that representatives act on behalf of collectives makes them more than equal citizens — politically speaking — and this is undemocratic. That they stand-in for people whom they are not the agents of creates two kinds of democratically inappropriate inequality between representatives and their constituents, which give rise to the unhealthy dynamics displayed in my above examples with Reed and McConnell.

I will call the first of these objectionable forms of inequality ‘standing inequality.’ This inequality describes the fact that in virtue of standing-in for many citizens, a representative is more politically important than their fellow citizens. The second, which relational egalitarians call ‘social inequality,’ arises from this standing inequality and makes constituents the social inferiors of their political representatives. People who want to effectuate policy change, such as lobbyists, are inclined to befriend and impress elected politicians, and furnish them with expensive gifts. They are, of course, not inclined to treat me the same way. The reason is plain to see: my representative's political beliefs, desires, and actions are far more important than mine. I am only one citizen, whereas my representative stands-in for the people. In a democracy, one person and the collective people are not equal, and so because of this standing inequality, I cannot relate to my political representative as a political equal in the democratically relevant sense.

Lovett, building on Kolodny's view, has recently argued that so long as representatives are ultimately controlled by the people collectively, their power is not incompatible with the political equality of citizens. It is not the asymmetrical power itself that is a problem, but the fact that this power can be independently exercised. He claims: “Insofar as how representatives exercise their power is under popular control, that power doesn't pose a threat to equality” (Lovett, Reference Lovett2021, p. 182). Political representatives need not be the superiors of the public for the same reason that lawyers aren't the superiors of their clients: because they cannot independently exercise the powers of those they represent. My argument poses a problem for both Kolodny's and Lovett's accounts of why the power of representatives is acceptable: even if they are in an important respect controlled by the people collectively, because they are the ‘avatars of the people,’ they are not the political or social equals of their constituents individually, and this undermines the political equality of citizens. Putting elected representatives on par with their constituents collectively does not put them on par with their constituents individually.

3b. Standing Inequality

The idea of standing inequality is that because elected politicians stand-in for the collective people, whereas no mere citizen can do this, elected politicians are politically superior to their constituents.Footnote ¹² As I have defined democracy in Section 2, part of what it means for a government to be democratic is that its citizens are politically equal. Indeed, democratic theorists often claim that in a representative democracy, citizens are equal partners in the governance of their state.Footnote ¹³ I argue that this is not true, because elected representatives, as avatars of the collective people, essentially monopolize that power.

Elected representatives are agents of their constituency, collectively. But the kind of agent they are is more like a trustee or guardian than an employee, or lawyer.Footnote ¹⁴ In a sense, the collective people needs this kind of representation because it literally cannot speak for or act for itself. Unlike an individual constituent, it neither has a body to speak or act through nor a mind with which to make decisions. The collective constituency requires interpretation, and so someone or something capable of making decisions must be empowered to decide how the collective shall use its powers. In a representative democracy, the elected representative does this. Like a guardian, an elected representative has a dual role as they relate to their principal/ward: they both exercise their ward's powers on their ward's behalf, and are tasked with interpreting what is in their ward's best interest and acting on that interpretation. Because they interpret and act for a constituency that can't act for itself, they have very limited accountability to the constituency collectively. In contrast to a trustee, a normal employee can be directed, and must obey or be dismissed. Even powerful employees, like CEOs, can typically be relieved of their powers and responsibilities by Boards of Directors with very limited notice. Because they decide how the constituency acts, there is no one to ‘obey’ but themselves. An elected representative realistically cannot be dismissed between election periods and is not obliged to act on the direction of any or all of their individual constituents. In any event, as far as I know, there is no jurisdiction in which a representative's vote in the legislature is invalid if it is determined that the representative acted against the orders of any or all of their constituents.

Guardians and political representatives may choose to take their wards’/constituents’ wishes or preferences into account when making their decisions, either because they feel morally obliged to, or because they would like to maintain a good relationship with their wards/constituents going forward. However, there is no sense in which their power is limited by the wishes or directions of any or all of their individual constituents, or their constituents collectively. Hence Reed's reply to his irate constituents: that they wished he would act differently is irrelevant because he acted for their sake. Whether a representative chooses to behave as Reed did, representatives have the power to do so.Footnote ¹⁵ This power to unilaterally interpret what is in the constituency's collective interest and use the constituency's vote in the legislature on the basis of that interpretation characterizes the relationship between an elected legislator and their constituents as individuals.

Hence, the elected representative is the avatar of the people: they alone among the members of their electoral constituency have the political power to speak in place of and act in place of the constituency collectively. They as individuals decide what the collective constituency's good is and how to pursue it in a way that no individual constituent or group of collective constituents can. At election time, constituents vote to decide who their avatar is (the person who will stand-in for that electorate politically), but between elections their power is embodied by the avatar.Footnote ¹⁶

As I have defined democracy in Section 2, in a democracy, the people collectively rule as equals. This standing inequality that I have described is a problem for democracy because it creates a political inequality between elected representatives and mere citizens. In a democracy, each person equally is supposed to be politically equal, and the people collectively are supposed to be superior to any individual citizen. Mere citizens do not rule, but by standing in for the ruling collective, the elected representative does. Insofar as one is committed to having a democratic government, then this standing inequality is a problem.

This entitlement to speak and act for the ruling collective explains why elected politicians such as Reed and McConnell can rebuff complaints from their constituents by saying that the politician speaks with more authority than the complaining constituent about what the collective constituency's interests are. This is because the politician represents the constituents collectively and the complaining constituents do not (Lartey, Reference Lartey2021). Individual citizens are not ‘the people.’ ‘The people’ are never in the room with the politician to contradict the politician about their will — they cannot be because ‘the people’ are not persons.

Some democratic theorists, including Bernard Manin, have developed the argument that elected legislatures are aristocratic because they represent the rule of the meritorious few over the comparatively unmeritorious many, rather than the rule of the people.Footnote ¹⁷ I am sympathetic to this argument, but it is not the argument I am advancing here. My argument is instead concerned with the substitution of the collective people for a ‘representative’ person, and the ways in which this makes that representative the political and social superior of their individual constituents. Standing inequality arises from the fact that elected representatives stand-in for the people collectively, whereas their constituents do not. The effect of this is that representatives are the political superiors of ordinary citizens — they determine how the collective exercises its political powers, and ordinary citizens do not. This inequality in political power is incompatible with the political equality of citizens. This standing inequality should be of concern to anyone who is committed to an understanding of democracy on which citizens are political equals — which is to say, on which citizens rule collectively, and have equal status as members of that collective, as I have defined democracy in Section 2. As I shall argue in Section 4, we could eliminate this standing inequality by replacing the institution of electing individual persons as our political representatives with the creation of mini-publics to act as political representatives of the public. Although the mini-public collectively would be politically superior to each citizen individually, no individual citizen would be the political superior of any other.

3c. Social Inequality

The second form of inequality that arises from the status of politicians as avatars of the people is social inequality: this status makes elected politicians the social superiors of their constituents.

Relational egalitarians ordinarily define ‘social inequality’ as a form of hierarchy, with social superiors and inferiors. Superiors in this hierarchy are treated differently and better. Kolodny describes this difference in treatment as a difference in ‘consideration’ as attracting a certain kind of response from others:

With this understanding of social hierarchy in hand, in extant representative democracies, it is plain to see that elected politicians are the social superiors of their constituents. To pick one example, consider political fundraising dinners — dinners that espouse to celebrate those who donate money to a political cause, but that are commonly understood to be opportunities to speak with and ingratiate oneself with politicians in exchange for money. Private persons are happy to provide the prime minister with lavish gifts just for the opportunity to speak with him. At the time of this article's writing, quite understandably, no one is willing to do the same for me. Likewise, journalists are very concerned to gain access to elected politicians because they are desperate to understand what those politicians think, particularly about political issues. Friendly journalists can be exceedingly courteous, to the point of obsequiousness, to maintain access to important politicians. Again, at the time of this article's writing, no one is willing to go to great lengths to solicit political opinions from me. This is exactly the sort of ingratiating or favour-currying behaviour that relational egalitarians understand to be characteristic of relations of social superiority and inferiority.

I would argue that this is not a contingent feature of elected-representative systems, but a necessary one that arises from the avatar of the people problem. In a democracy, the people rule through their elected representatives. Roughly speaking, because my member of parliament speaks and acts politically for approximately 100,000 constituents, journalists and those seeking to persuasively influence political outcomes would appropriately be approximately 100,000 times as concerned with their political opinions than with mine.

Kolodny claims that these differences in consideration arise from, among other things, differences in powerFootnote ¹⁸ or de facto authority,Footnote ¹⁹ which is precisely what we have here. As avatars of the people, elected politicians have the authority to determine what is best for the collective of their constituents, politically, and act for them on the basis of that interpretation. This is precisely why they receive this special consideration. They are treated as though they politically matter more than mere equal citizens because they do matter more. As long as there is standing inequality, there will be attendant social inequality. The solution to this problem, as I shall argue in Section 4, is to make it so no individual person stands-in for the public collectively, by replacing elected individuals as political representatives with mini-publics.

3d. Inequality and Democracy

That elected representatives are the avatars of the people — because they as individuals stand-in for their constituents collectively — creates these two objectionable forms of inequality. In my view, the above explains why the conduct of Reed and McConnell seems so egregious: they are very transparently treating their constituents as their political and social inferiors. They are asserting that they do not have to listen to or make themselves accountable to protesting constituents. Their justification is that in a democracy the people collectively rule, notwithstanding the objections of any individual persons, and because as representatives they embody the people collectively, they get to rule notwithstanding the objections of the individual protestors. When Reed and McConnell justify themselves in this way, they reveal that they do not face the protestors as equal citizens, but as their superiors. The protestors may make appeals, and Reed and McConnell may choose to listen to them or ignore them — they will not decide together with them because that decision belongs to the representative. Because the elected representative has won the election and the protestors have not, the “Winners [will] make policy and the losers [will] go home” (Lartey, Reference Lartey2021). It is rare for politicians to make this relationship so plain to their constituents, but irrespective of how a politician acts, they do not face their constituents as political equals. This relationship of political superiority and inferiority is not democratic.

Some relational egalitarians argue that democracy is normatively important because it is a necessary component of a society without relations of social superiority and inferiority (Kolodny, Reference Kolodny2014b; Viehoff, Reference Viehoff, Sobel, Vallentyne and Wall2019). According to these relational egalitarians, a society of this kind is normatively important intrinsically. It is the way that we ought to relate to each other.Footnote ²⁰ Given that hierarchical social relations ought to be abolished, and they only could be under a democratic government, we ought to have a democratic government. They may be correct that this is why democracy is important.Footnote ²¹ However, even if they are incorrect, in the case of political representation, the social inequality on the surface that exists between elected representatives and their constituents arises from a standing inequality that should concern any democratic theorist. The necessity of the maintenance of the equal political status of citizens is baked-in to the concept of democracy itself. As I have demonstrated above, the institution of elected political representatives is incompatible with the political equality of citizens, and as I have said in Section 2, the political equality of citizens is an important constitutive element of democracy. This presents us with an important (if not necessarily decisive) reason to abolish the institution of elected political representatives and replace it with an alternative. I shall consider an alternative form of political representation that does not face this problem in Section 4: representation by mini-publics.

4. Representation by Deliberative Mini-Publics

The alternative I would like to consider is representation by mini-publics, rather than representation by individual persons. This will enable us to have political representation without necessitating the two forms of inequality that representation by elected representatives creates. The avatar of the people problem arises from the fact that one person is empowered to speak and act on behalf of many. By putting a body in that position rather than a person, I will argue that we can resolve this problem.

As I mentioned at the beginning of this article, a mini-public is a deliberative assembly composed of hundreds of randomly selected citizens. How exactly a legislative system that uses mini-publics as political representatives of the public rather than elected representatives would work is not important to the resolution of the avatar of the people problem. But, to illustrate the idea, some scholars — such as Terril Bouricius and Alexander Guerrero — have proposed legislative systems where many deliberative mini-publics, each with hundreds of participants, are created to consider a single legislative issue, after which they disband (Bouricius, Reference Bouricius2013; Guerrero, Reference Guerrero2014). Under such a system, additional mini-publics will be required for other functions, such as setting the agenda for the single-issue-legislature mini-publics (Bouricius, Reference Bouricius2013). As an alternative, more participatory structure to the one proposed by Bouricius and Guerrero, we might instead simply have a randomly selected council set the agenda for referendums, and have all actual legislation enacted by referendums instead of having the legislation enacted by mini-publics.Footnote ²² My argument is not about the all-thing-considered best way of organizing such a legislature; rather it is that sometimes political representation is indispensable (as is the case at the agenda-setting phase of legislation),Footnote ²³ and in these circumstances, in order to resolve the avatar of the people problem, deliberative mini-publics should displace elected political representatives.

Aside from the argument I make for legislative mini-publics in this article, there have been several other arguments made both for and against the use of mini-publics to displace elected legislatures. Some epistemic democrats, such as Hélène Landemore, have argued that mini-publics, owing to their cognitive diversity, are superior to elected legislators at deliberating (Landemore, Reference Landemore2012). Other proponents of the idea, like Guerrero and Dan Van Reybrouck, stress the inherent conflicts of interest that politicians have in acting on behalf of ordinary citizens and the inability of ordinary people, with the limited accountability mechanisms available under electoral democracy, to discipline their behaviour.Footnote ²⁴ Others, such as Samuel Bagg (Reference Bagg2024), Yves Sintomer (Reference Sintomer2023), and Peter Stone (Reference Stone2011), have stressed the comparative incorruptibility of the lottery mechanism and ordinary citizens compared to the multitude of avenues for the capture of politicians by special interests that electoral democracy represents. On the anti-mini-public side, some, like Christina Lafont (Reference Lafont2019), argue that empowering deliberative mini-publics is an unacceptable ‘shortcut’ to democracy that attempts to sidestep the democratic necessity of having the whole public learn and deliberate about political issues. Others have raised the concern that quickly rotating ordinary citizens into and out of positions of legislative power would destroy the institutional memory of the legislature, make the content of the law unstable by creating situations where an incoming mini-public might want to un-do the work just done by an outgoing mini-public, or make budgetary planning impossible.Footnote ²⁵ These kinds of reasons are precisely why John Gastil and Erik Wright (Reference Gastil and Wright2019, Chapter 1) advocate for a bicameral legislature, with one chamber for randomly selected citizens and one with elected members, and for longer terms for the randomly selected citizens put into office, comparable to the terms of elected politicians. Others, such as Bouricius (Reference Bouricius2018), have attempted to rebut these criticisms of the feasibility of legislative mini-publics, and argued against the desirability of a bicameral legislature of this kind.

However, the success of my argument does not turn on the outcome of any of these debates. I aim to demonstrate that the avatar of the people problem gives us an important reason to prefer a system of legislative mini-publics to a system of elected legislatures. Perhaps, for example, epistemic democrats (for example, Goodin & Spiekermann, Reference Goodin and Spiekermann2018) who argue that if a democratic structure is too incompetent its use cannot be legitimate are correct, and perhaps (although it seems unlikely) legislative mini-publics are too incompetent to be legitimate. If so, this would be a pity because, according to the argument I make in this article, legislative mini-publics are substantively more democratic than elected legislatures because they do not create these two problematic forms of inequality and so do not face the avatar of the people problem.

Why do I say that replacing elected representatives with deliberative mini-publics would resolve the avatar of the people problem? Because, when mini-publics are used as political representatives in this way, each mini-public politically represents the public as a whole. By contrast, each randomly selected participant represents no one but themselves. In aggregate, the mini-public acts for the whole public as an elected legislator would. As individuals, the members deliberate with their fellow participants and act for themselves on the basis of their own judgements, deciding how to vote just as a citizen would in an election or referendum. As I shall argue, even though these randomly selected individuals have powers that other citizens do not have, this representative structure does not generate problematic forms of political inequality.

One persistent criticism of the democratic credentials of empowered mini-publics is that if we just one-to-one replaced the members of an elected legislator with random citizens, as some advocates of the political uses of random selection suggest we ought to do with at least some elected officials,Footnote ²⁶ is that as representatives these random citizens would seem to have a much weaker connection to those they represent than would elected officials.Footnote ²⁷ I would agree. One of the prominent strengths of the use of a deliberative mini-public as a legislature is that its members should be expected in aggregate to be a representative sample of the population from which they are randomly selected. Hence the name ‘mini-public’: they resemble the public in miniature.Footnote ²⁸ But any individual member will, of course, not represent the public in miniature, because that is not the sort of thing that one person can do. I think that advocates for lottocracy would be wise to give up on the claim that these individual members politically represent the public. In my view, while the body collectively represents the public collectively, each randomly selected member is a mere citizen, with no more authority to speak or act on behalf of the public than any other citizen. These mini-public members would be selected through a procedure that treats all citizens as equals and gives each an equal chance of sitting in the legislature, rather than through a procedure that pits them against other candidates in a contest to determine who is best suited to speak and act for the rest.

In terms of the social inequality discussed above, I think that it is quite possible to design a legislative model where randomly selected citizen-legislators would not be the social superiors of those who are not selected. Using ‘single-issue-legislatures,’ these citizens would be quickly rotated into and out of office. Under such a system, after considering one issue and voting once, the single-issue-legislature disbands. We could even take steps to keep the contents of their deliberation secret until after the single-issue-legislature has disbanded. With changes like these, there is no threat that those chosen to serve as legislators would become members of a superior social class. I do not think that randomly selected legislators would be subjected to favour-currying, fawning, or ingratiating behaviour from the public any more than jurors are at an important trial. Like jurors, they will have arrived at their powerful position through chance, and their power is quite temporary and circumscribed in scope.Footnote ²⁹ Does this structure generate standing inequality? It is reasonable to observe that the randomly selected citizens on the mini-public have political powers not possessed by their fellow citizens. If the avatar of the people problem is that elected representatives are more powerful than their fellow citizens, then one might think that replacing election with random selection does not solve that problem: the representatives are still more powerful. I have two responses to this objection.

First, standing inequality is not the mere fact of unequal power. Standing inequality, as I have explained above, is inequality that arises from the fact that an individual representative stands-in for many constituents and is therefore the political superior of their constituents. On this definition, there is no standing inequality because these randomly selected members do not act or speak on behalf of anyone except themselves. Instead, it is the body collectively that represents the people collectively.

But second, it is reasonable to wonder: even if there is no standing equality, might the randomly selected members be the political superiors of their fellow citizens through a different mechanism? These randomly selected citizens have very important input into that mini-public's decisions (a vote and participation in deliberation). Other citizens lack this input. In this sense, those selected for the mini-public directly have more power than other citizens over some legislative decisions, even if they don't stand-in for these other citizens. How can the unequal distribution of this political power be compatible with the political equality of citizens? Even if it is not ‘standing inequality,’ it may be another type of problematic political inequality between citizens — possibly, a kind of inequality more objectionable than standing inequality.

I do not think this is the case; the powers that members of the mini-public have do not make them the political superiors of their fellow citizens. I would argue that so long as political powers are distributed in ways that treat citizens as equals (such as by rotation or lottery), temporary asymmetries in the ability to exercise political powers are not incompatible with the political equality of citizens. In a democracy, the people rule collectively as equals. What is required for them to be ‘equals’ in the collective enterprise of ruling does not rule out the possibility of role-rotation that leaves some temporarily with more powers than others. An analogy to equality in other sorts of collective enterprises will be helpful to explain why this is the case.

Imagine four members of a housing co-op. They have decided that each of them will do the work of maintaining their garden by rotation — for one week in each month, each of them will do the work. It is clear that through this system, these four people collectively maintain the garden (as equal members of that collective), even though it is the case that in any given week only one of them does the work. Now imagine instead a co-op with 400 members. These co-op members don't want to use a pure rotation system because some weeks of the year involve much more work than others as a result of seasonal changes. So, instead, each week they randomly select some number of them who shall maintain the garden in that week, and remove the names of those selected from the pool until each member of the co-op has taken a turn. Here too it is clear that the members of the co-op collectively maintain the garden, even though it is the case that per the example some will be unlucky and will be given more work than others.

It is clear in this case that they are not made equals in their collective enterprise of maintaining their garden by the fact that they all have equal maintenance duties. Rather, it is because they are treated as indistinct in the distribution of maintenance duties, and the distribution of duties is not final (some are not chosen to maintain the garden forever), that they are equal. A system of rotation or random selection is perfectly compatible with equality for those in the rotation in relation to their shared project because of the way that both rotation and random selection treat the participants as indistinct from one another.Footnote ³⁰

While there are some notable differences between the political case and this housing co-op case, I would posit that as it is with the ongoing distribution of duties in this case, so it is with the ongoing distribution of powers in the political case. In terms of their dissimilarities, unlike the political case, in the co-op, the powers and duties being distributed are not important powers over people's freedoms and lives. Furthermore, in the co-op case, the ultimate goals of the selection process are understood by the co-op members prior to their selection — this could be the case but is not necessarily the case when members are selected for a mini-public.Footnote ³¹ But these differences are not relevant to the analogy I wish do draw: what the co-op example illustrates are the requirements of equal partnership in a shared enterprise. While the co-op's maintenance duties are more clearly defined at the outset and less weighty than the mini-publics, both are shared enterprises, and so the insights from what it takes for members to be treated as equals in the distribution of duties in one case will apply to the other.

Furthermore, note that it is not my position that maintenance duties in a co-op must be shared by equal partners. This is something the members of a co-op are permitted to do, of course, but I don't mean to argue that they ought to. In many spheres of life, the kind of equality between persons that democracy requires is unnecessary, and may even be inappropriate. Equal partnership in a collective enterprise is an element that a democratic legislature can have in common with many other sorts of collective endeavours. Examining these other collective endeavours can be useful for clarifying our thinking about what the collective rule of equals requires. Unlike garden maintenance duties, however, it is the case that the legislature ought to be collectively controlled by equal citizens because it ought to be democratic.

As my co-op example shows, to be equal partners in a collective enterprise, people need to be treated as equals in the distribution of the powers and duties of that collective enterprise. Choosing legislators by random selection treats citizens as equals in the distribution of legislative powers and duties, and so in this way citizens are made equal partners in the collective enterprise of governing their state.

This requirement is, of course, not meant to be exhaustive of the requirements for a democratic legislature. As noted earlier, it is plausible that the distribution of power must be meaningfully ongoing and not a one-time event. In this way, it makes sense to say that a randomly selected emperor-for-life would not be democratic. On Bouricius' (Reference Bouricius2013) model of political random selection for the legislature, randomly selected legislators are given very circumscribed powers for very limited periods of time. A typical randomly selected legislator will have only one of these three sorts of powers: the power to set the agenda for different randomly selected legislatures, the power to draft a proposed law to address an agenda item, or the power to consider and vote on a proposed law given to them by another randomly selected legislature. After the legislature works through its agenda item, it disbands. By distributing power in this way — to random citizens, in small parcels, for short periods of time — we can ensure that no sub-group meaningfully distinct from the broader citizenry ever wields entrenched legislative power. Randomly selected legislators cannot legislate to the benefit of their private interests as distinct from the interests of the broader public because, in aggregate, the interests of these bodies are identical to the interests of ordinary citizens. They are only different in that (temporarily) they are legislators. The quick rotation and narrow scope of their powers ensures that they have no reason to act to advance the interests of legislators as a group because, as soon as they have legislated, they cease to be legislators.

As I have alluded to above, the requirement that citizens be treated as equals in the distribution of powers and duties is not very exacting. It does not point to one unique form of government that is ‘most democratic.’ A great variety of legislative structures (as well as other government structures) are equally capable of clearing this bar. Perhaps we could, for example, have one mini-public perform the same functions as an elected parliament does now, and randomly select new members at the same times that we would have an election under the current system. We could have a randomly selected chamber that does not pass laws, but instead sets the agenda for referendums that the entire public votes upon. We could do as Bouricius, Guerrero, and I suggest and have many different legislative mini-publics with different functions — some to set agendas, some to draft legislation, and some to vote on whether that legislation shall be enacted (see Bouricius, Reference Bouricius2013). This requirement that citizens be treated as equals in the distribution of political powers is a kind of ‘democratic minimum,’ and this minimum rules out elected legislatures.