Ensuring Fair Value: A Rawlsian Justification for Lottocratic Selection

Based on Rawlsian ‘justice as fairness,’ this article outlines a novel justification for sortition that uses Guerrero’s lottocracy as a model. I argue that elections allow for undue influence based on what ‘justice as fairness’ determines to be arbitrary characteristics, and by failing to be justifiable to members of minority groups in society. Lottocracy, by contrast, satisfies both of these criteria and better protects the legislature from capture, which prevents the legislature from being used to undermine political justice. 

1. Introduction

Is sortition selection for members of legislatures to be preferred over elected representatives, and if so, on what grounds? When designing democratic political systems, one of the most consequential decisions is the nature of the legislature. Typically, elections play the role of selecting representatives, who make decisions on behalf of constituents. However, the electoral approach to designing the primary legislative body of government often presents problems, such as vulnerability to elite capture, poor responsiveness, sometimes questionable governance quality, and crucially lack of fairness in the selection process (Guerrero 2014; Guerrero 2024; Landa and Pevnick 2021).

Recently, sortition, or lottocratic bodies, in which legislatures are made up of a body of randomly selected citizens who represent society at large, have begun to make a resurgence. Though sortition was used in ancient Greece and the Italian Republics, it had typically been passed over by democratic theorists until relatively recently. Particularly of note is Alexander Guerrero’s framework for the selection of legislative assembly members outlined in Lottocracy: Democracy Without Elections (2024), which I will rely on in this article, with some modification in response to key criticisms.

The lottocratic system outlined by Guerrero calls for the use of a network of single-issue legislatures (SILL) made by randomly selecting citizens from geographic districts of equal population to serve for three years, with a third of each body being selected every year. Following John Rawls, I will consider sortition democratic on the grounds that it is compatible with the idea of a democratic society. A democratic society refers to a society characterized by conditions of equality, rather than a specific political regime (Cohen 2002; Michelman 2002, 397).

Arguments for sortition legislatures have, for the most part, focused on the potential benefits they would have in terms of governance, such as better responsiveness when compared with elected legislatures (Bartels 2008, 261; Guerrero 2014). Proponents of sortition legislatures have also claimed that they help to address one of mass electoral democracies’ major weaknesses: rational voter ignorance, by essentially creating a snapshot of the ideal informed opinions of the electorate (Guerrero 2014, 141; Landa and Pevnick 2021, 52). However, these arguments have been made almost exclusively from a broadly consequentialist conception of democratic institutions (Guerrero 2014). This lack of other arguments has left the case for sortition on shaky ground, given that it is a relatively unproven concept, with only a handful of contemporary examples. The standard consequentialist perspective has also failed to address many of the in-principle concerns of those who oppose sortition, who often focus on the potential lack of accountability, authorization, and suffrage (Lever 2023, 103). This lack of clear non-instrumental normative justification is a significant weakness in the sortition argument, especially as it remains relatively untested. Without such support, there can be little reason to support such an untested system over traditional electoral democracy, and sortition is unlikely to gain any traction.

In this article, I will argue that sortition, as typified by Guerrero’s system of lottocracy, is usually superior in securing the fair value of political liberties concerning the selection of legislative assembly members under Rawls’s framework of ‘justice as fairness.’ Fair value of political liberties is the requirement that, relative to others, one has an equivalent real value of those rights. That is, the equivalent ability to affect the outcome by vote, or by holding office. It is a requirement of substantive rather than formal equality. The normative principles of ‘justice as fairness,’ outlined by Rawls, provide a promising and clear normative justification for lottocratic selection and favor a lottocratic legislature over electoral ones in particular, because lottocracy promises to better secure the fair value of political liberties. While Rawls outlines a system of representative democracy, I believe that lottocracy serves the purposes of ‘justice as fairness’ far better (Rawls 2005). Fair value is better secured by lottocratic legislatures because they are better equipped to resist interference by background injustice in the legislative process and more equitably distribute access to political power, both of which are integral to ensuring the continuation of justice (Rawls 1995, 7-10).

In this article, we will be concerned with the basic structure of a well-ordered society under Rawls’s principles of justice as he defined them in “Justice as Fairness,” and how they evaluate the system outlined for constituting the legislature by Guerrero in his book, making modifications only as necessary to address objections. I begin by outlining the original position and the nature of the inquiry regarding fair value in Section 2. In Section 3, I argue why elections fail to secure the fair value of political liberties, while lottocracy would succeed in doing so in a variety of situations. I address some key objections to this line of argument in Section 4, and make some adjustments to the lottocratic position in Section 5.

2. The Original Position

The original position is a thought experiment of how an agreement could be reached by free and equal citizens for ensuring fair cooperation for the advantage of all (Rawls 1995, 15). Rawls’s (1995) argument from the original position consists of a few fundamental ideas:

1. The idea of society as a fair system of cooperation where everyone has an equal chance at advancement and access to primary goods for developing their two moral powers.

2. A society wherein everyone recognizes and follows a political conception of justice.

3. The idea of a basic structure of society, which includes the fundamental institutions of society.

4. Citizens as free and equal persons.

5. The rationalization of the legitimacy of institutions by members of society.

The citizens are placed behind the veil of ignorance and do not know the following: “race and ethnic group, sex, native endowments, social position, or their comprehensive doctrines.” The veil of ignorance in the original position prevents the citizens’ views from being distorted by the familiar. Under the second step–the “constitutional phase”–however, they are allowed to know the basic structure, history, and social order of their society (Rawls 1995, 15). It is in this constitutional phase that this argument takes place, regarding the institutions implemented to establish and maintain equal liberty as proscribed by justice as fairness (Rawls 1995, 133-134).

Rawls (1995, 18-19) bases justice as fairness upon what he calls the two moral powers:

1. The capacity for a sense of  justice- ability to understand, apply, and act from the principles of political justice.

2. The capacity for a conception of one’s rational good- an “ordered family of final ends.”

Rawls suggests that citizens are endowed equally with the capacity to make reasoned judgements of value based on their conceptions of the good, and society’s conception of political justice (Rawls 1995, 20). This capacity for developing and acting on a conception of political justice is important in justifying lottocracy on Rawlsian grounds, and when properly trained, it is what is necessary for working in the legislature (Guerrero 2024). Essentially, our regard for one another as sufficiently capable of assessing the justice of society suggests that we would regard one another as minimally capable legislators (Cohen 2002, 109). Rawls does not necessarily refer to a political order of electoral democracy. Rather, Rawls refers to a society characterized by conditions of equality, as in a condition where society is conceived of as being made up of equal persons (Cohen 2002; Michelman 2002, 397). Equal treatment of persons is a condition lottocracy satisfies by its conception of everyone as legitimately qualified for public office. All people within the polity are treated as equal citizens in virtue of the capacity for a sense of justice.

3. Fair Value and Elections

With regards to the decision of a political system under the second phase of the original position, Rawls argues, “The justice of a procedure always depends on the justice of its likely outcome” (Rawls 1995, 169). The political system’s creation takes into account, it would seem, the general nature of society, which is information available to the parties at the second phase of the original position in order to render a probabilistic judgment (Rawls 1995, 169). It is not enough then that a system be merely capable of satisfying the demands of justice as fairness. It must include a substantively fair procedure and be the one most likely to result in just and effective legislation over an extended period of time (Rawls 2005, 194). The primary values by which these procedures are to be assessed are: “impartiality and equality, openness (no one and no relevant information is excluded), lack of coercion, and unanimity” (Rawls 1995, 172). These values come from the objective of basic institutions: “Basic institutions must be from the outset put in the hands of citizens generally, and not only of a few, the productive means to be fully cooperating members of a society” (Rawls 2005, xv). These functions are to ensure society is likely to sustain the ideal of a fair system of cooperation among free and equal persons. As a part of the requirements for a system under ‘justice as fairness,’ institutions must support the fair value of political liberties and be resistant to capture. That is, institutions must “not be subject to political bargaining or a calculus of social interests” (Rawls 2005, 4). In this section, I argue that elections fail to support the fair value of political liberties under prevailing circumstances of socio-political inequality and the existence of persistent political minorities (Bagg 2024; Rawls 2005, xv). I will then go on to argue that a system of sortition legislatures, or lottocracy, does support the fair value of political liberties in such a way that it is liable to weakly dominate elections, being as good or better in all relevant respects.

3.1 Fair Value

The fair value of political liberties means that the worth of political liberties must be equal for all, ensuring a fair equality of opportunity to hold office regardless of social or economic standing (Rawls 1995, 149; Rawls 2005, 215). The fair value of political liberties requires substantive rather than merely formal equality in the ability of all to take part in the political process, not just in terms of their ability to influence legislators but to become a legislator themselves: “those similarly endowed and motivated should have roughly the same chance of attaining positions of political authority irrespective of their economic and social class” (Rawls 2005, 197). For an electoral system, fair value would mean that one needs to both have a vote equivalent to all others, but also have a substantively fair chance at election to office – universal suffrage alone then is clearly insufficient. It is insufficient for two reasons. First, universal suffrage only secures the formal equality of political influence but does not by itself manage to secure fair value. Institutional practices such as unequal population districts or gerrymandering can undermine the fair value of the vote in this context. Second, universal suffrage makes no move to secure equality of opportunity for political office (Rawls 2011, 267).

A core challenge that Rawls identifies as an impediment to substantive political equality is: “The inability to take advantage of one’s rights and opportunities as a result of poverty and ignorance… I shall think of these things as affecting the worth of liberty, the value to individuals of the rights that the first principle defines” (Rawls 2005, 179). He claims that when background inequality, but not necessarily injustice, is able to influence politics to give some people unequal influence, it can give them the ability to capture the political system. Unequal influence in politics violates the principles of justice as it violates the equal nature of all citizens (Rawls 2005, 198).

3.2 The Failure of Elections

Elections under the current conditions of society in countries like the United States are broadly unequal, and Rawls uses the state of public financing in U.S. elections as an example of a failure to secure the fair value of political liberties by allowing undue influence from background inequality (Rawls 1995, 158). This results in elected representatives, as it stands, being manifestly unrepresentative in terms of their demographic characteristics, which brings into question the equality of the election process. In the U.S., elected representatives are disproportionately white, male, educated, and likely to be lawyers or businesspeople (Guerrero 2024, 264). Consequently, the U.S. seems to fail to secure equality of opportunity with regard to political office (Guerrero 2024, 105). Only those with significant resources can have a realistic chance of gaining office, or of influencing those in office (Bartels 2008).

Guerrero argues that unequal influence is not simply the result of historical contingencies but rather an unavoidable part of electoral democracy: the capacity of an electoral representative system “is limited by features that are endemic to the form of government, not peripheral or incidental to it “ (Guerrero 2024, 3). Capture in electoral systems plays out in three primary ways: the first is by reducing the accountability of those in power to constituents, the second by controlling the pool of viable candidates, and the third by actually incentivizing representatives to change behavior (Guerrero 2024, 62). When elections are not well structured and have low accountability, representatives turn into agents for those in power. But capture of electoral institutions is liable to occur regardless, as widespread voter ignorance and structural differences between voters and representatives will ensure low accountability (Landa and Pevnick 2021). Low accountability will allow entrenched influence to erode the background justice of the election because citizens do not have the resources to adequately monitor their representatives and their campaigns: “the suggestion so far on this horn of the dilemma is that ignorance results in an absence of meaningful accountability, and in the absence of meaningful accountability, we should expect to see high levels of political capture” (Guerrero 2024, 64).

Rawls claims that the accumulation of resources in the hands of a few is a serious concern in a society, even when the individual transactions are perfectly in line with justice (Rawls 2011, 266). This problem is not merely an unfortunate current condition, but an inevitable result of the asymmetric time available between individual citizens and their representatives (Landa and Pevnick 2021). Another avenue of capture is the ability to groom candidates for office, something that would be difficult to prevent and could undermine an electoral system even with fair equality of opportunity (Guerrero 2024, 63). Differential access to those in office and resources to influence elections is a very real reason to think that the powerful can manipulate the selection of officials to ensure this happens even when background conditions are otherwise fair (Guerrero 2024, 289).

A third concern has to do with the difficulty of distributing natural endowments and social standing. Even with the equal social basis of self-respect secured, relative social position could still play a significant role in elections. Factors such as endorsement by existing politicians or administration officials, as well as private persons of high standing, allow the influence of the powerful to be greater than it should be. Elections hinge very clearly on capital of all kinds. Access to that capital will not be evenly distributed even in an ideal society, absent abhorrent levels of social control which are liable to be antithetical to the concept of justice as fairness. Candidates from persistent minorities are liable to be especially vulnerable to this problem of lower levels of social capital. Under elections, we will always have to worry about those with high social capital influencing elections or running for office themselves. Many elements that play a huge role in political equality, such as the distribution of wealth, the influence of money in politics, and truly equal education, could go a long way to improving political equality under elections by improving the distribution of different types of relevant capital. However, these would be insufficient to ensure adequate fair value. Social standing, the media, and natural capacities would still potentially play a significant role in elections, undermining the fair value of political liberties.

Capture by the powerful is inherently problematic in a system designed to care for the interests of the worse off, and therefore inherently problematic in any system of Rawlsian design. Officials captured by the elite may have distorted motivations and have incentives to engage in rent seeking or broader forms of exploitation of their political position. Capture and corruption of officials is a clear danger to the background conditions. As a significant portion of the process of determining the constitution of a just society is to determine what is likely to perform the best, the fact that so many clear avenues of capture exist in electoral systems is concerning. Elections are, in contemporary circumstances, liable to be effectively a tool of the political elite for choosing representatives they favor.

There may well be conditions under which elections are substantially less unfair, but they are not in modern conditions and real modern states, which are relevant to the determination of a constitution in accordance with the principles of justice. That elections ensure representatives so often serve the interests of those in power is alarming (Bartels 2008; Guerrero 2024). Poor representation is an inherent structural problem in an electoral system caused by elections (Guerrero 2024, 113). So many of the channels by which capture occurs, and thus background injustice rears its head, occur within the bounds of free speech and association, such as endorsements and publicity. Even extensive public finance reform would not address elements such as media presentation of candidates (Guerrero 2024, 122). Rawls limits restrictions on free speech to the procedural, not interfering with content (Rawls 1995, 149). Regulating media influence would therefore be almost impossible without infringing on rights to free speech in ways not allowed by the principles of justice. The principles of justice as fairness are explicitly designed to prevent “the use of natural endowments and the contingencies of social circumstance as counters in a quest for political and economic advantage,” and yet this is exactly what elections allow for (Rawls 2005, 14).

Under ‘justice as fairness,’ disparities in the value of political liberties can be established, but this can only be done when it increases the value of the liberties of those whose liberties will be, in the end, lesser (Rawls 2005, 200). Rawls argues that the curtailing of fair value must be justified to those who would possess lesser political liberty. For Rawls, the basic liberties–to the maximal extent possible–are a core part of what actors in the original position desire because these ensure their ability to pursue the production of their various life plans (Rawls 2005, 203). However, any reduction in the fair value of political liberties cannot be on account of some nebulous advantage to the common good, or the interests of society’s members; trading equality for more competent rules, for example, is not permissible.

3.3 Elections, Minorities, and Public Reasons

I see no way in which those on the losing end of electoral democracies’ inequalities in political liberties could be compensated with greater liberties in other respects in ways not also compatible with lottocracy. In an electoral democracy, anyone who is positioned as a part of a permanent minority effectively has no sense of control or influence over who is elected, especially if no political actors are sympathetic to their interests. Guerrero (2024, 380-381) outlines conditions that would potentially prevent a permanent political minority from consenting to a political system:

1. The views of the minority can always be ignored without the majority incurring political costs.

2. The views of the minority will be given little to no consideration.

3. The minority will lose all conflicts with the majority.

4. The members of the minority have basically no political power.

5. The members of the minority have even weaker power than they would with respect to their proportion in the population.

6. The interests of the minority will not be advanced/realized, and their preferences will not be satisfied.

Not all of these problems will occur in all circumstances in electoral democracies – countervailing systems can be put in place (but these often prove objectionable to the majority), and it is not uncommon for permanent political minorities to be decisive when the margins of victory are slim. The combative nature of elections, however, encourages partisanship due to the regular political skirmishes between clearly outlined segments of society (Guerrero 2024). In such circumstances, especially if the margins of victory are slim or one lives in jurisdictions where one’s views are significantly outnumbered, majoritarian electoral institutions will fundamentally disempower members of the minority, effectively reducing the value of their political rights by ensuring they will never be able to hold office (Guerrero 2024, 382). It seems that no justification can be given for anyone in a minority community like this to consent to such an institutional arrangement.

In the original position, given the characteristics of the parties and the precepts of ‘justice as fairness,’ people would not consent to a system of government where they could not fairly exercise their right to hold office and whereby the preferences, values, and incentives of their officials were systematically unlikely to be aligned with their own, especially as majoritarian electoral systems threaten to seriously jeopardize the ability of some groups to participate in politics as officials. Elections furthermore curtail the fair value of political liberties so long as the common historical conditions of inequality and social stratification stand where the parties in the original position are attempting to establish a just society. Elections allow the undue influence of contingent factors like natural endowments and background inequality, either through social standing, the influence of the media, or economic means. These problems are endemic to electoral democracies as we know them and cannot all simply be swept aside by modest reform without causing other significant concerns, such as curtailing free speech.

3.4 Fair Value and Lottocracy

Having argued that elections fail to adequately secure the fair value of the ability to hold office, as well as failing to be justifiable to all parties in the second stage of the original position, we move on to establishing the virtue of lottocracy in these areas. Despite democracies’ deficiency in these areas, it is not immediately clear that lottocracy would be superior. Guerrero’s lottocratic system assumes conditions roughly akin to those of the contemporary United States or other Western democracies (Guerrero 2024, 149). I follow in assuming that this is the historical condition of society that the actors in the original position, considering the core institutions of their society, are faced with.

In Political Liberalism, Rawls describes fair value as guaranteeing that the worth of the political liberties is approximately equal for all, providing them with a fair opportunity to hold office and to influence the outcome of political decisions (Rawls 2011, 327). In A Theory of Justice, he writes, “those similarly endowed and motivated should have roughly the same chance of attaining positions of political authority irrespective of their economic and social class,” and that citizens should have a fair chance to influence the political agenda (Rawls 2005, 197-198). So if lottocracy provides all with a fair opportunity to hold office, and to influence the outcomes of political decisions, it should secure the fair value of political liberties. It is important to be careful not to compare the idealized idea of a lottocratic system with the reality of electoral politics. Lottocracy provokes serious concerns, including the possibility of bureaucratic and expert capture. Regarding fair value specifically, though, by virtue of its formal structure, lottocracy is strictly equal between all citizens, as all have exactly the same chance to hold office, being selected by purely random means.

That citizens having a perfectly fair chance to influence the political agenda through the fair value of the ability to hold office best secures the fair value of one’s political liberties, rather than a more participatory condition of being able to express one’s views, is not immediately clear. Ultimately, it must be an exercise in determining what provides individuals with the greatest chance to exercise substantive influence over the political agenda, irrespective of factors like their social class. Even in a lottocratic system, the media, advocacy groups, and more could still give some groups an outsized means by which to influence the political process. Ultimately, on this dimension, the difference between elections and lottocracy is liable to be a matter of degree. There are a number of reasons, however, to think lottocracy will fare far better than the electoral system in reducing institutional capture. First, in a lottocracy, capture would likely be more difficult as the candidates are chosen randomly with no way to manipulate their selection, and there is historical evidence from ancient states to suggest it is efficacious in preventing capture in this way (Lever 2023, 101). Second, the lack of reelection campaigns will make corruption and contact monitoring of representatives easier with regard to influence and corruption (Guerrero 2014, 164). With regards to concerns about expert capture, Guerrero provides some options, including a database of relevant experts maintained in cooperation with professional associations from which experts would be randomly selected (Guerrero 2024, 186-211). A database like this may go some way towards preventing capture, but there are still major concerns regarding things like the capture of educational institutions or professional associations. Crucially, however, these problems are not unique to lottocracy, as elected officials heavily rely on such experts as well (Guerrero 2024). It may be supposed that perhaps elected officials, by virtue of greater competence, could better push back on captured experts. Perhaps this is true, but evidence from the British Columbia Citizens’ Assembly showed that normal citizens could, when given access to relevant knowledge and expertise, conduct high-level and serious discussions of political issues, weighing expert testimony (Ferejohn 2008, 196).

The influence of the media is a problem that is shared between the two systems. It is no mark against lottocracy alone that one could use the media to inculcate opinions across the populace that would grant them undue influence – it would be that way in any society where something like that transpired. But if such a thing is even possible, it would be expensive, extremely difficult, and liable to be an extremely costly and precarious undertaking (Guerrero 2014, 164-165). It seems that in terms of these criteria, lottocracy, as described, is liable to perform as well or better than electoral democracy. Even in non-ideal conditions, whereby capture is attempted often and forcefully, the avenues of capture are substantially limited in scope, easier to monitor, and more expensive to access. The previously mentioned avenues of capture and influence are likely easier to exploit under elections that allow elite actors to manipulate elections. I believe that the more limited avenues of capture is enough to conclude that this system better embodies the values of the procedure needed to ensure that everyone has a substantively fair chance to input into political decision-making (Rawls 1995, 169-172).

Regarding the consent of the individual actors on account of minority political power, lottocracy significantly improves the chances of members of the minority controlling policy – that is, it significantly increases the chances for minorities of all kinds to attain office. In elections, the majority can simply dominate the minority perpetually. This is in contrast with lottocracy, where the minority has a proportional chance of being selected, and so their views have a considerably higher chance of effective consideration (Guerrero 2024, 290-291; 385). Guerrero also points out how elections create a combative mentality regarding politics, potentially causing the formation of distinct and repressed minorities. Without such regular combative contests, and given most people’s relatively thin political commitments, it’s unlikely disagreements would be so vehement (Guerrero 2024, 393). Potentially, with lottocracy, the use of the minority as a political scapegoat would be reduced, while greater descriptive representation could bring their moral-political problems more sharply into focus (Guerrero 2024, 392). Over the long run, random selection and proportional representation may reduce political frictions while providing minorities with better opportunities for their interests to be represented. Public justification is a core requirement of justice as fairness, and it seems that lottocracy has the likelihood of being able to satisfy it in this capacity (Rawls 1995, 145).

4. Objections

There are a number of objections to the use of lottocratic selection to replace elections in the literature (Rawls 2011, 294-295). However, to be effective against the fair value justification, objections must show that the implementation of lottocracy for the purposes of selecting members of the legislature must damage some other basic liberties, as appeals to the public good are insufficient (Rawls 2011, 295).

The first concern is that the kind of representation offered through lottocracy cannot bind the representative or hold them accountable. They argue that the lottocratic model expects citizens to defer to the assembly’s decisions and takes away accountability. Elections, by contrast, do allow for a clear connection between representatives and the people (Lafont and Urbinati 2024; Landa and Pevnick 2021, 98; Rummens and Geenens 2023, 4). Guerrero’s response is that accountability is not a factor in the democratic legitimacy of lottocracy on account of the lack of a principal-agent relationship between the parties (Guerrero 2024; Hurrelmann, Schneider, and Steffek 2007, 47). However, elections do not do well on accountability either – we have evidence that elections fail decisively to produce accountability and are unlikely to do so in most prevailing conditions (Bartels 2008; Landa and Pevnick 2021, 52). In particular, there are good reasons to think the epistemic task is too daunting for even the most informed citizens.

Furthermore, the accountability critique is based on a limited but familiar idea of accountability as sanction, common in modern representative democracies, but this is far from the only form of accountability. There are several systems, like regular public town halls with representatives, recall votes, and the more representative nature of the legislature, that make their ability to potentially give an account of and thus, in some sense, be accountable to the interests of those they represent potentially greater than current elected representatives (Guerrero 2024; Lee 2024, 4). Accountability need not necessarily take the form of a sanction by electoral outcome. Many of the aforementioned methods of accountability are explicit elements of lottocracy (Guerrero 2024, 164-165).

The lottocratic system does, however, lack any form of accountability by sanction. Guerrero considers the use of a recall vote, which I believe would address these accountability concerns (Guerrero 2024, 167). The inclusion of a recall vote institutes a form of accountability by sanction in the hands of the citizens and provides them with a mechanism for direct political action. Another option is for lottocratic assembly members to undergo an approval vote upon entering office and/or at certain points in their careers. This would introduce points of accountability for representatives who would then need to convince enough of their constituents that they ought to keep their jobs. This would encourage them to consider the views of their constituents, and justify any decisions made to them, in much the same way politicians are thought to be incentivized to do under elections.

The second primary objection to the use of the lottocratic system is the lack of authorization for the representatives. The charge is that authorization is vital to democratic governance, and lottocracy’s failure to provide it is a real concern and strikes at the existence of certain forms of political liberties. In The Lottocratic Mentality, the authors draw the distinction between voting as consent and voting as a true power (Lafont and Urbinati 2024, 127). The fundamental issue here is that representative elections do not authorize lawmaking except through representatives. With little meaningful accountability over those representatives, it is difficult to see how the citizens have the real power to authorize lawmaking. If Lafont and Urbinati are right, it may be that by removing voting, the people lose access to political liberties in some fundamental way, not merely a curtailing of these liberties along a specific dimension but a fundamental absence (Lafont and Urbinati 2024, 128). It seems, however, that both of these functions of elections can be disunited. One moment of authorization occurs when voting on representatives or policy through a direct method like a referendum (Guerrero 2024, 334). The authorization of policy, as in Landemore’s view, is undesirable according to Guerrero due to the widespread voter ignorance that lottocracy hopes to offset. A better alternative would be to authorize the candidates in some way (Guerrero 2024; Lafont and Urbinati 2024, 100-101). A second moment of authorization could involve a form of approval voting, whereby candidates are required to be approved by their district in a vote. In a society with deep social divisions, this may be a poor solution. Entrenched discrimination against religious and ethnic minorities, as well as women, could cause these votes to be abused. In a society with a healthy political culture and a unifying sense of political justice that everyone follows, a formal procedure of consent-giving should be viable.

Another option would be to vote on the confirmation of policy priorities. For example, imagine every time the new legislature is sworn in, people go to the polls to select the priorities of the body. One could imagine a system where citizens were perhaps given a lengthy list of potential topics and asked to rank them. Institutional rules within the legislature would then incentivize or require addressing specifically these issues that were ranked highest. Deviations would likely have to be justified through public reasons to the wider public, especially if legislatures face being stripped of their office and the departure is substantial. Such a method of authorization may sidestep both the concerns about the mere authorization of candidates and the authorization of policy directly. Instead, by determining the priorities of the legislature and authorizing the focus of policymaking, we would both more directly determine policy, while leaving the particulars to the better epistemically positioned legislators.

A third objection to lottocracy is that the distribution of power is problematic, given that while the chances of being selected are equal, they are so low as to be negligible, ejecting suffrage in favor of mere descriptive representation (Lafont and Urbinati, 2024; Rummens and Geenens, 2023). The claim is that a lottery, while being a substantively fair procedure, fails to provide real equality because it disenfranchises most of the population, a concern Guerrero himself notes (Guerrero 2024, 315). The contention appears to be over some kind of ‘actual say’ in the running of government or selection of representatives (Guerrero 2024, 315). That ‘actual say,’ however, need not be with regard to who holds political office.

As we have established, elections substantially reduce the fair value of political liberties by restricting the ability of people to obtain public office fairly, especially those in a minority. As such, the best strategy in keeping with justice as fairness would be to provide another avenue for this ‘actual say’ in governance. One option following from my previous suggestion could be to have annual referendums regarding the agendas of the various SILLs. Voters with a list of agenda items rank them using some kind of ranked choice scheme which determines legislative priorities. This gives the people one of the most fundamental powers in politics, the power to control the agenda and the shape of political action. It seems as if this kind of approach, or one founded on referendums, would provide substantially more actual say in the shaping of policy by citizens. As it stands, citizens have little actual say in electoral politics, and policy is often made one or more steps from their vote, by their candidates, electoral leadership, parties, or interest groups. That they lack any input is a potent criticism to be sure, but it by no means follows that the appropriate mechanism for providing that ‘actual say’ is elections.

Ultimately, the most forceful objections against the use of lottocratic institutions to select legislative assembly members are the lack of accountability, the lack of a clear moment of authorization by the people, and that the lack of active, equal political power for all citizens cannot be addressed within the lottocratic system. Through changes such as the introduction of town halls and recall votes, avenues of accountability to the people can be added to the system. Similarly, through the addition of a procedure of authorization by the people of a district for their representatives, we can provide a mechanism for acquiring consent from the people. Finally, by disaggregating the notion of voting from elections, we can identify new and better ways for citizens to exercise direct and equal collective control over the institutions of government. There need be no trade-offs between ensuring authorization, accountability, and suffrage and adopting a lottocratic system to select citizens for political office.

5. Conclusion

The exact nature of the system that emerges will have its own unique flavor in accordance with the moral-political problems that need solving in the society in which it exists. Lottocracy may not be optimal for all circumstances, but in the conditions that pertain in Western democracies (and elsewhere around the globe) it best satisfies the principles of justice with regard to the selection of citizens for public office (Guerrero 2024, 149). There are specific principled reasons from the Rawlsian tradition to adopt legislative institutions of lottocracy.

Elections are unable to effectively secure the fair value of political liberties under justice as fairness with regard to the selection of candidates for office. The fair value of political rights will only be substantively maintained if one has an equivalent real value of those rights – that is, the real equivalent ability to affect the outcome by vote, or by holding office. The influence of the powerful, of socioeconomic status, natural endowments, and the media, substantively alters the outcomes of elections and the value of political liberties. In addition, the sidelining of minority citizens of all kinds prevents their justification of the system, and invalidates it under the original position. By contrast, the use of sortition in Guerrero’s system of lottocracy can ensure the fair value of the right to hold public office. It cannot do so absolutely, but by removing elections and replacing them with random selection, many factors that make access to political office unequal and violate the principle of fair value are removed.

According to Lafont and Urbinati, “lottery is a fair procedure for distributing a good only when it is not possible to provide the good in question to everyone that has a proper claim to it” (Lafont and Urbinati 2024, 175). This is precisely what lottocracy does – the good of political office cannot be properly provided to all who have a claim. Public office is a good that everyone is owed equal access to by virtue of their role as citizens under justice as fairness, even if they have no desire to claim it. Since only a few can hold these offices and all have an equal entitlement, a lottery is a perfectly legitimate means for allocating them.

While lottocratic selection surely possesses some flaws, my arguments have shown that they are able to satisfy the requirements of a democratic regime under Rawlsian justice. With small changes such as the introduction of confirmation votes for representatives and recall votes, the main in-principle objections to the system can be swept aside. Ensuring that all citizens have access to real and equal power to influence the political process at all times surely deserves more consideration. However, unlike what many suppose, this need not come through elections, and by separating these ideas, we can come to a better solution.

Acknowledgement

I would like to acknowledge Professor Jeppe von Platz for his help with this article. He was invaluable in helping me get started with my research on the Rawlsian literature. Without his help, this would have been a substantially more difficult project.

References

Bagg, Samuel Ely. 2024. The Dispersion of Power: A Critical Realist Theory of Democracy. 1st ed. Oxford University Press. https://doi.org/10.1093/oso/9780192848826.001.0001.

Bartels, Larry M. 2008. Unequal Democracy: The Political Economy of the New Gilded Age. Princeton University Press.

Cohen, J. 2002. “For a Democratic Society.” In The Cambridge Companion to Rawls, S. Freeman. Cambridge Companions to Philosophy. Cambridge University Press.

Ferejohn, J. 2008. “Conclusion: The Citizens Assembly Model. In Designing Deliberative Democracy.” In The British Columbia Citizens’ Assembly. Cambridge University Press.

Freeman, Samuel. 2007. Rawls. Routledge Philosophers. Routledge.

Garry, John, James Pow, John Coakley, David Farrell, Brendan O’Leary, and James Tilley. 2022. “The Perception of the Legitimacy of Citizens’ Assemblies in Deeply Divided Places? Evidence of Public and Elite Opinion from Consociational Northern Ireland.” Government and Opposition 57 (3): 532–51. https://doi.org/10.1017/gov.2021.4.

Gilens, Martin, and Benjamin I. Page. 2014. “Testing Theories of American Politics: Elites, Interest Groups, and Average Citizens.” Perspectives on Politics 12 (3): 564–81. https://doi.org/10.1017/S1537592714001595.

Guerrero, Alex. 2024. Lottocracy: Democracy without Elections. Oxford University Press.

Guerrero, Alexander A. 2014. “Against Elections: The Lottocratic Alternative.” Philosophy & Public Affairs 42 (2): 135–78.

Hurrelmann, Achim, Steffen Schneider, and Jens Steffek. 2007. Legitimacy in an Age of Global Politics. Transformations of the State. Palgrave Macmillan.

Lafont, Cristina, and Nadia Urbinati. 2024. The Lottocratic Mentality: Defending Democracy against Lottocracy. Oxford Scholarship Online Political Science. Oxford University Press.

Landa, Dimitri, and Ryan Pevnick. 2021. “Is Random Selection a Cure for the Ills of Electoral Representation?*.” Journal of Political Philosophy 29 (1): 46–72. https://doi.org/10.1111/jopp.12219.

Landa, Dimitri, and Ryan Pevnick. 2025. Representative Democracy: A Justification. 1st ed. Oxford University Press. https://doi.org/10.1093/oso/9780198897538.001.0001.

Lee, S.M. 2024. “Empowered Mini-Publics: A Shortcut or Democratically Legitimate?” Journal of Deliberative Democracy, 1–11.

Lever, Annabelle. 2023. “‘The Circumstances of Democracy’: Why Random Selection Is Not Better Than Elections If We Value Political Equality and Privacy.” Washington University Review of Philosophy 3: 100–114. https://doi.org/10.5840/wurop202339.

Michelman, FI. 2002. “Rawls on Constitutionalism and Constitutional Law.” In The Cambridge Companion to Rawls, S Freeman. Cambridge Companions to Philosophy. Cambridge University Press.

Rawls, John. 1995. “Political Liberalism: Reply to Habermas.” The Journal of Philosophy 92 (3): 132. https://doi.org/10.2307/2940843.

Rawls, John. 2003. Justice as Fairness: A Restatement. 3. printing. Edited by Erin Kelly. Belknap Press of Harvard University Press.

Rawls, John, ed. 2005. A Theory of Justice. Original edition. Harvard University Press.

Rawls, John. 2011. Political Liberalism: Expanded Edition. Columbia Classics in Philosophy. Columbia University Press.

Rummens, Stefan, and Raf Geenens. 2023. “Lottocracy Versus Democracy.” Res Publica, ahead of print, November 22. https://doi.org/10.1007/s11158-023-09648-6.