The Oligarchic Tendencies of Representative Government and Lottocracy

A once standard presumption in political science that representative government is responsive to the views of its citizens has been increasingly challenged by evidence of its capture by wealthy elites. An influential body of research on US representative government has found that the preferences of most citizens have almost no effect on public policy outcomes, especially when they conflict with the preferences of the wealthiest (Bartels Reference Bartels2008, chap. 9; Gilens and Page Reference Gilens and Page2014; Winters and Page Reference Winters and Page2009). This might be considered a quirk of the United States’ extremely high levels of economic inequality and lax approach to the influence of money in politics. But a recent wave of similar studies of various European social democracies (Elsässer, Hense, and Schäfer Reference Elsässer, Hense and Schäfer2021; Lupu and Tirado Castro Reference Lupu and Castro2023; Mathisen Reference Mathisen2023; Mathisen et al. Reference Mathisen, Schakel, Hense, Elsässer, Persson, Pontusson, Lupu and Pontusson2024; Persson Reference Persson2024; Schakel Reference Schakel2021), as well as an initial global comparative investigation (Lupu and Warner Reference Lupu and Warner2022), have come to alarmingly similar conclusions, suggesting that the problem lies deeper than just the United States’ particular instantiation of representative government.Footnote ²

One explanation may lie in the structural features of representative government itself. As Bernard Manin (Reference Manin1997, chaps. 4–5) influentially argued in his defining study of representative government, two foundational features bias it toward oligarchy. First, elections (in comparison with selection by lot) have a strong inherent tendency to select wealthy public officials (or those willing to appeal to the wealthy for support); this tendency has been widely confirmed in the empirical literature on class and descriptive representation (Carnes and Lupu Reference Carnes, Lupu, Pontusson and Lupu2024). Second, elected politicians enjoy a significant degree of independence from the citizens who elected them. Although elected representatives are subjected to control through regular elections and the pressure of public opinion, they are not strictly obligated to follow the wishes of the electorate through stronger accountability mechanisms, such as the right to recall representatives or binding instructions (Manin Reference Manin1997, 163–67). Representative government thus has built-in tendencies to select elected representatives who are themselves wealthy or have the support of the wealthy and who enjoy significant independence from the citizens who elected them. Elected politicians consequently tend to have the disposition/incentive and the ability to act in favor of the interests and preferences of the wealthy. These two oligarchical features were for long periods of the twentieth century curbed by mass parties and trade unions, whose role in popular mobilization, education, and recruitment to political leadership helped bridge the void between ordinary citizens and elected politicians (Mair Reference Mair2013; O’Neill and White Reference O’Neill, White, Collins, Lester and Mantouvalou2018). The decline over the last 40 years of these extraconstitutional institutions has unsurprisingly led to a reassertion of representative government’s underlying oligarchical features.

The oligarchic travails of representative government have driven a search for institutional democratic innovations, with a particular focus on sortition. Sortition’s anti-oligarchic promise derives from breaking the electoral link between wealth and public office, because random selection ensures that offices are distributed across all classes of society and randomly selected officials do not need to appeal to the support of the wealthy to win election or reelection (McCormick Reference McCormick2006, 148–49). The first generation of sortition proposals envisaged public authorities drawing on the nonbinding deliberative discussions of randomly selected citizens to improve the creation and assessment of public policy (Sintomer Reference Sintomer2023, 193–206). But these proposals’ ad-hoc, nonbinding nature and subordination to existing electoral political authorities limit their anti-oligarchic potential (Møller Mulvad and Popp-Madsen Reference Møller Mulvad and Popp-Madsen2021, 87–89; White Reference White, Leipold, Nabulsi and White2020, 92–93).

Attention has consequently shifted to institutionalized proposals in which elected assemblies are checked by permanent and empowered randomly selected bodies. These include bicameral proposals with a second legislative chamber selected by lot (Abizadeh Reference Abizadeh2021; Barnett and Carty Reference Barnett and Carty2008; Gastil and Wright Reference Gastil, Wright, Gastil and Wright2019; Zakaras Reference Zakaras2010) and proposals for an additional tribunate power (Hamilton Reference Hamilton2018, 489; McCormick Reference McCormick2011, 183–88; Prinz and Westphal Reference Prinz and Westphal2024; Vergara Reference Vergara2020, chap. 9). These promise to limit the oligarchic tendencies of representative government while maintaining the advantages of existing electoral systems (Gastil and Wright Reference Gastil, Wright, Gastil and Wright2019, 32–33). But a more extensive set of proposals argues for the necessity of wholly (or nearly wholly) removing elections from the political system and replacing them with selection by lot. Such lottocratic proposals include suggestions for legislation to be carried out by an all-purpose central national assembly selected by lot (Landemore Reference Landemore2020) or by a series of lottery-selected assemblies dedicated to single issues (Guerrero Reference Guerrero2014, Reference Guerrero2024) or to a particular legislative function (Bouricius Reference Bouricius2013; van Reybrouck Reference Reybrouck2016, 138–50).

The growing prominence of these lottocratic proposals has sparked a more recent backlash against lotteries and a defense of elections and representative government (Grandjean Reference Grandjean2024; Lafont and Urbinati Reference Lafont and Urbinati2024; Landa and Pevnick Reference Landa and Pevnick2025; Lever Reference Lever2024; Umbers Reference Umbers2021). One of the more worrying counterarguments to emerge from these critiques has been that lottocracy might be just as or even more vulnerable to oligarchic capture as representative government (Bagg Reference Bagg2024a; Landa and Pevnick Reference Landa and Pevnick2021). Two features of lottocracy make this plausible. First, randomly selected citizen-legislators would likely be subject to the same attempts at oligarchic capture as elected politicians and would potentially be even less able to resist them. Although lottocratic assemblies would automatically consist of a cross section of society that largely excludes the wealthiest (thus blocking this oligarchic tendency of representative government), citizen legislators would likely come under sustained oligarchic attempts at capture after they were selected.Footnote ³ Once citizen legislators are empowered to pass binding legislation, it is likely that they would be bombarded by corporate lobbyists attempting to set the legislative agenda, and we would expect the emergence of sophisticated inducements to citizen legislators to act in favor of wealthy and corporate interests (Landa and Pevnick Reference Landa and Pevnick2021, 53–54). Citizen legislators might be even more susceptible to these inducements because they would not be subject to the main accountability mechanism that (imperfectly) constrains elected politicians: the threat of deselection at the next election. Attempts to block this oligarchic capture of citizen legislators through short terms of office might, in contrast, limit their ability to gain the expertise to engage with complex legislative tasks and hence make them more reliant on the support of civil service administrators, opening them up to the converse problem of bureaucratic capture (Landa and Pevnick Reference Landa and Pevnick2021, 58, 64–66). This is not an argument that ordinary citizens are incapable of gaining suitable legislative expertise (something that seems eminently possible), but that the time and training it takes them to do so open them up to avenues of corruption and capture (Bagg Reference Bagg2024a, 96).

The second feature of lottocracy that makes it potentially vulnerable to oligarchy is the absence of political parties from its assemblies and their potential disappearance from a lottocratic society altogether (Guerrero Reference Guerrero2024, 245, 252; Landemore Reference Landemore2020, 145–49). This feature connects to a broader “antipolitical imaginary” that animates some defenders of lotteries, which sees random selection as a way to bypass the venal, partisan politicking of elites and replace it with the nonconflictual deliberation of a sociologically undivided people, an imaginary that results in a “a radical rejection of political parties and the idea that one must transcend politics” (Sintomer Reference Sintomer2023, 242).

Yet political parties can play a crucial role in limiting oligarchy (Bagg and Bhatia Reference Bagg and Bhatia2022). In addition to their role in popular political recruitment, parties and their surrounding institutional ecology (party institutes, think tanks, etc.) can provide elected representatives with the necessary knowledge and expertise to reduce their reliance on the bureaucracy and corporate lobbyists when formulating legislation. For another, political parties can—when connected to a mass base—provide part of the organizational strength required to resist oligarchic reactions to legislative encroachments on their interests. It is hard, for instance, to imagine that a program of large-scale social transformation that threatens oligarchical interests, such as the creation of the welfare state, would be possible without a mass party to formulate a cohesive program and have the necessary organizational strength to implement it.

Because of the potential susceptibility of citizen legislators to corruption and the absence of political parties, it is unclear whether a lottocratic alternative would be less susceptible to oligarchic capture than representative government. Although it may be possible to somewhat allay these concerns by introducing firewalls and high levels of financial compensation to limit the corruption of citizen legislators (Guerrero Reference Guerrero2024, 290–92), the mixed record of such measures in representative government should leave us at least skeptical on this front. Bagg (Reference Bagg2024a) has argued that the broader lesson might be that the real democratic and anti-oligarchic promise of sortition has been eclipsed by inflated expectations. Rather than thinking of sortition as a replacement for representation, he argues that it is best used as a device for anticorruption, in which “the appropriate mandate for randomly selected citizens is not legislation but oversight: that is, screening particular decisions for signs of capture or corruption” (103).

Shifting to thinking of sortition as an effective mechanism for control and oversight, rather than initiation and direction, might also remind us that sortition proposals have tended to focus on only one of representative government’s foundational oligarchic features—the tendency for elections to select wealthy public officials: as a result, the relative independence of elected representatives from popular control has slipped from critical attention (Landa and Pevnick Reference Landa and Pevnick2021, 52; Vandamme Reference Vandamme2024). Yet, the latter is just as essential to explaining representative government’s oligarchic tendencies. Countering these tendencies requires examining not only the social composition of public officials but also the powers they enjoy. Ideally, what is thus needed is institutional reform that both incorporates sortition’s anti-oligarchical promise of filling some public offices through random selection while also limiting the independence of those officials selected via election by making them properly accountable to the electorate.

Traditionally, radical democrats have turned to measures such as binding instructions (known as an imperative mandate) and the right to recall to create this greater accountability. But binding instructions are often thought to be too crude a tool to be workable because they seemingly tie representatives to pre-election decisions, regardless of changing circumstances or the need to compromise with other factions and parties in parliament (Manin Reference Manin1997, 166).Footnote ⁴ Thus, even though there are good reasons to limit the independence of representatives, it is equally undesirable to entirely remove it and leave representatives with no flexibility whatsoever. In contrast, the right to recall does not face this objection because it functions as a post hoc accountability mechanism. But the evidence from its limited implementation suggests that its effectiveness remains unproven because recalls are rarely initiated and, when implemented badly, can be misused by sore losers hoping to overturn an electoral defeat (rather than any particular failing on the part of the representative) or by wealthy special interests who are able to sponsor a costly recall campaign (Vandamme Reference Vandamme2020; Welp and Whitehead Reference Welp, Whitehead, Welp and Whitehead2020). It is thus not clear whether a straightforward power to recall is itself sufficiently robust against oligarchic capture and is able to deliver the accountability needed to appropriately limit the independence of representatives. These kinds of considerations lead many to conclude that “it is unthinkable that citizens can actively hold accountable those who represent them. ‘Accountability’ directly to citizens requires a level of citizen participation in monitoring their representatives that is impossible in large-scale complex democracies” (Castiglione and Warren Reference Castiglione, Warren, Disch, van de Sande and Urbinati2019, 29). But that conclusion has been reached too hastily, especially given the as-yet underexplored possibility that sortition might be able to provide that accountability.

In sum, when considering how sortition might address the oligarchic features of representative government several paired desiderata emerge:

• • Reform should incorporate the anti-oligarchic potential of sortition but maintain the advantages of elections.

• • Reform should be institutionalized by being a permanent and empowered component of the constitution but at the same time be itself resilient to oligarchic capture.

• • Reform should limit the independence of elected representatives while still providing them with sufficient flexibility.

Constituency juries show promise in meetings these desiderata and are thereby meant to expand the existing toolbox of anti-oligarchic constitutionalism, not displace other promising, empowered proposals that combine sortition and elections. Whereas existing proposals, such as a second chamber selected by lot, introduce a horizontal check on the whole elected chamber, constituency juries explore the overlooked possibility of each elected member of the legislature being checked vertically by their respective constituency.

In comparison with bicameral proposals, constituency juries are likely to involve greater expense and logistical complexity (because of the number of juries and jurors); are potentially more likely to be biased toward local concerns (compared to the national focus of a single sortition assembly); and are especially suited to majoritarian electoral systems (rather than proportional ones). But they are also potentially better able to restrict the independence of representatives (by subjecting each representative to close monitoring and regular calling to account) and more resilient to oligarchic capture (because of the large number of jurors, their shorter terms of office, and the possibility of making their service mandatory). I consider these relative (dis)advantages in the following discussion, but further consideration and ultimately experimentation, perhaps involving a combination of models, would be required to come to a firmer judgment.Footnote ⁵ My aim here is to place constituency juries on the table as one possibility for our collective reconsideration of the institutions of representative government.

General Model of Constituency Juries

Constituency juries, as I envisage them, are defined by three core features corresponding to their (1) purpose, (2) composition, and (3) constitutional powers. Constituency juries

1. (1) hold the constituency’s elected representative accountable

2. (2) consist of randomly selected constituents, and

3. (3) are empowered to

   1. (a) sanction their elected representative, and

   2. (b) require the elected representative to regularly and publicly inform and explain/justify their actions to the jury (figure 1).Footnote ⁶

Figure 1 General model of constituency juries

These three features constitute the general model of constituency juries, which would modify representative government so that, for each elected representative, a corresponding constituency jury would be established and filled through random selection from the constituency that elected the representative.Footnote ⁷ In a legislature of 500 elected representatives there would thus be 500 corresponding constituency juries. The Member of Parliament for Bristol would, for instance, be held accountable by the constituency jury of Bristol and so on for each constituency. Each elected representative would be required to regularly and publicly meet with their constituency jury to give an account of their actions in the legislature, and the constituency jurors would have the power to sanction their elected representative if they found that account wanting. Constituency juries would thus carry out a supervisory function, in which they provide ongoing oversight of elected representatives.

The three features of the general model are the guiding institutional principles that would need to be part of any specific institutionalization of constituency juries for them to meet the desiderata outlined in the previous section. Their realization would require decisions on several more detailed institutional design questions that are explored later in the article. But it is helpful to separate the general model of constituency juries from any particular realization of them. Questions of the best institutional design are likely to invite significant disagreement, especially when judgments remain speculative before experimenting with their operation. By separating the general model of constituency juries from particular institutional designs, lines of disagreement can be clarified, and those sympathetic to the former need not be put off by their opposition to specific institutional choices.

When discussing the general model and the specific institutional design, I assume a majoritarian electoral system with territorial single-member constituencies, such as Britain, France, India or the United States.Footnote ⁸ In such systems, constituency juries could be straightforwardly formed in each of the constituencies that elect representatives to those countries’ respective lower houses: the House of Commons, the National Assembly, the Lok Sabha, and the House of Representatives.Footnote ⁹ Constituency juries could be adapted relatively straightforwardly for mixed electoral systems, such as in Mexico or Germany, where the legislature comprises both proportionally assigned members and members directly elected in single constituencies. Proportional systems, such as Brazil or the Netherlands, would require greater rethinking.

The justification for focusing on majoritarian systems springs partly from the relative ease in introducing constituency juries to systems with single-member constituencies but also because majoritarian systems are still the most widespread form of electoral system in terms of population. Of the countries that are commonly categorized as democracies, most use a proportional or mixed electoral system, but it is easily forgotten, in part because of the neglect of Indian democracy, that the total population living in majoritarian systems (~2.5 billion) outstrips that of proportional (~1.1 billion) and mixed systems (~0.75 billion).Footnote ¹⁰ Although proportional systems have independent advantages, it is unlikely that majoritarian systems will disappear, and it is consequently incumbent on us to consider how they might be reformed to achieve greater accountability. Beginning with majoritarian systems also provides a plausible starting point for subsequent consideration of how constituency juries might be adapted to mixed electoral systems.

I turn now to elaborating and defending the three features of the general model in reverse order: starting with constitutional powers, then turning to composition, and then finally to purpose.

Constituency Juries and the Constituency

The preceding discussion of the purpose, composition, and constitutional powers of constituency juries was primarily focused on their relationship with their representative. I now turn to specifically consider the relationship between the constituency jury and the constituency. This matters because of a concern that there may be discrepancies between juries and the constituency and because juries might either disempower the constituency or, conversely, overly empower the constituency’s local interests relative to the national common good.

Random selection would mean that the jury is drawn from a cross section of the constituency’s population and its associated views and preferences. But because it would be impractical to make juries so large that they completely cancel out sampling errors, they will usually produce only a rough correspondence with the constituency, rather than a perfect descriptive representation. Descriptive representation of the constituency would be achieved over time through repeated selections from the constituency, and the full set of constituency juries would, given the number of constituencies in a legislature, be quite likely to be descriptively representative of the whole nation. Yet at any one time the views of a single constituency jury might diverge, potentially significantly, from those of its particular constituency. It might thus come to decisions at odds with the views of the wider constituency. That possibility should constrain, as I discuss in the next section, the choice of sanctions that are made available to the constituency jury and inform the design of the precise mechanisms for their application.

Random selection from the constituency provides, as I have argued, some democratic legitimacy for the powers assigned to constituency juries. But given their ability to sanction representatives and to require them to regularly give an account of themselves, there is a genuine democratic concern that constituency juries would displace or disempower the wider constituency. Although this article is premised on the assumption that the existing setup of representative government fails to effectively empower the constituency in regard to its representative, constituency juries would undermine their own aims if they were to reduce what limited control constituents currently have. The democratic justification for instituting constituency juries consequently lies in whether they do, in fact, enhance the accountability of representatives to their constituency.

Constituency juries would in essence introduce an additional agent that helps hold one agent (the elected representative) accountable to another (the constituency). Yet they would not remove the constituency’s own primary accountability mechanism: its ability to vote out the representative at the next election. Just as constituency juries do not supplant the role of elected representatives in making legislation, neither do they supplant the role of citizens and their power to choose their representatives. Concurrently, representatives would continue to be incentivized to consider the views and interests of the constituency to win reelection. Constituency juries are thus premised on the continued operation of representative government’s existing electoral accountability mechanism, rather than its removal as in lottocratic models.

What constituency juries instead do is extend the mechanisms that keep the representative accountable to their constituents by providing a form of “extra-electoral accountability” (Arlen Reference Arlen2022, 202; cf. McCormick Reference McCormick2011, vii). Representatives would face the additional prospect of being immediately (or imminently) sanctioned, instead of only being subject to the potential threat of being thrown out of office at the next election, potentially four or five years in the future. Elected representatives would thus be actively held to account during their mandate, rather than only at the end of the electoral cycle. This would allow for a more fine-grained assessment of issues than provided by the blunt tool of elections. Representative government forces voters to make a single all-encompassing package judgment on the past performance and likely future actions of the elected representative. In contrast, constituency juries enable elected representatives to be held to account for single decisions on specific issues as they emerge during the course of their electoral mandate. Constituents would still be able to make a package judgment through their power to reelect or vote out the representative at the next election, but that judgment would itself be informed by the deliberative communication provided by the public sessions of the constituency jury. The constituency’s ongoing consideration of the elected representative’s regular public justifications would, in turn, feed into the representative’s own reconsideration of their actions through the normal public channels of democratic discourse, including petitioning, campaigning, media pressure, and opinion polling. Constituency juries thus provide a space for ongoing, recursive deliberation and communication between constituents and their elected representatives.

The creation of this deliberative space at the constituency level sets constituency juries apart from existing sortition proposals, including bicameral ones. Although a national sortition assembly could hold sessions to inform and elicit feedback from the public (Gastil and Wright Reference Gastil, Wright, Gastil and Wright2019, 30–31), constituency juries have the unique setup of requiring every representative to attend regular sessions with a cross section of their constituents and submitting to their questioning. Constituents would be able to indirectly (as observers), and sometimes directly (as jurors), hear explanations and justifications from their representative for their actions in the legislature. Compared to other sortition proposals, constituency juries would bring representatives closer (literally so, when attending a jury session) to the constituents who elected them. Because these deliberative jury sessions are also backed up by potential sanctions, representatives would also be more tightly constrained to act in line with the views of their constituents relative to other sortition proposals.

However, this feature will potentially make constituency juries more responsive to the local concerns of the constituency compared to the national concerns of randomly selected citizen legislators from across the whole country or indeed compared with elected representatives in existing representative governments. Although this may be an inherent defect when compared with national sortition assemblies, it is arguably outweighed by how this feature would provide a more targeted constraint on the independence of the representative. Furthermore, although constituency juries may sometimes lead to individual elected representatives ignoring the national common good by defending an issue of local importance, this would still (as in existing representative government) be counteracted by the legislature’s other elected representatives unattached to that specific locality.

Moreover, this article started from the finding that representative governments are not especially successful in promoting the common good, given their oligarchical proclivity to promote the good of the wealthy. By making each elected representative more accountable to the people of each constituency, constituency juries will make the overall political system more responsive to the common good of all its citizens, rather than just its wealthiest. If experiments with constituency jury were, however, to consistently show them leading to local particularism, an institutional amendment could be considered in which a small number of jurors from other constituencies are included in each jury. It would also be an argument for considering constituency juries within a mixed electoral system, in which the directly elected representatives from a constituency are balanced by proportionally elected representatives from across the nation.

In summary, although constituency juries may lead to a degree of localism, they would also provide constituents with a more active, fine-grained, deliberative, recursive, and anti-oligarchic form of accountability than currently offered by representative government.

Institutional Design of Constituency Juries

In this final section I propose one way in which the general model of constituency juries could be institutionalized. This design is not intended as a definitive blueprint. I introduce it here to show that the general model can be operationalized and realistically incorporated into representative governments. I also hope that, by specifying and justifying the design choices for constituency juries, my proposal will stimulate further debate on the possible institutionalized forms they could take, including by those convinced by the general model but perhaps less persuaded by some of the institutional choices.

In the specific institutional design that I propose, constituency juries

1. (1) hold the constituency’s elected representative accountable,

2. (2) consist of around 50 randomly selected constituents who serve a mandatory three-month term of office with one-third rotated after each jury session, and

3. (3) are empowered to

   1. (a) sanction their elected representative by censuring them and/or triggering a recall election,

   2. (b) require the elected representative to publicly inform and explain/justify their actions at monthly sessions with the jury after a week of internal deliberations (figure 2)

Figure 2 Specific institutional design of constituency juries

I defend this institutional design by showing how it might be more successful in fulfilling the desiderata outlined in the first section than alternative design choices.

Monthly sessions with the elected representative would ensure that constituency juries have sufficiently regular opportunities to provide ongoing scrutiny of their conduct. Votes on specific pieces of legislation would be closely preceded or followed by a constituency jury session in which jurors could express their agreement or displeasure at the representative’s (prospective) vote and the representative could, in turn, give timely updates and explanations for their (impending) actions. Having less frequent sessions, such as every 6 or 12 months (Maloy Reference Maloy2008, 188) or twice during the representative’s mandate (Sintomer Reference Sintomer2023, 270), would risk specific legislation going unscrutinized and governments shifting the timing of votes on unpopular legislation to just after constituency jury sessions to avoid scrutiny. That would reproduce a core problem of representative government, in which the independence of elected representatives allows them to ride out discontent from unpopular decisions until the next election when sufficient time has passed to deprioritize an issue or demobilize its opponents.

The monthly sessions with the elected representatives would be preceded by a week or so of preparatory deliberative meetings of the constituency jurors. If the session were, for instance, to take place on a Friday, jurors could be assembled to deliberate from Monday to Thursday. These internal meetings would allow jurors to receive basic training about their role (e.g., the powers and rules of procedure of constituency juries), learn about the elected representative’s conduct (e.g., their votes on recent and upcoming legislation), and prepare for the sessions (e.g., formulating potential questions, deciding on the agenda, and the possible use of sanctions). Jurors’ three-month term of office would mean that they would participate in three sets of these internal preparatory meetings and three sessions with the elected representative.

An alternative setup would be for jurors to only serve for a single session with the elected representatives and meet directly with them without prior internal deliberative meetings. The justification for this alternative would be to protect jurors from internal and external influence and corruption. Serving for only a single session would make it very difficult for outside lobbyists to target jurors. Blocking the opportunity for internal deliberation would ensure that the jury is not influenced by discussions dominated by those jurors with the socioeconomic characteristics associated with greater authority and knowledge. That would mirror the practice of ancient Athens’s popular courts where randomly selected citizens would judge cases without discussing them among themselves so that they were not swayed by the rhetorical gifts of aristocratic jurors (Cammack Reference Cammack2022, 2003–5). This pared-down alternative setup might thus have some anti-oligarchic advantages.

Yet these considerations are, I believe, outweighed by the need for jurors to acquire sufficient training, knowledge, and confidence to adequately hold the elected representative accountable. Without such preparation, it is likely that representatives would find it relatively easy to direct discussions toward their concerns, explain away criticism, and simply provide a justification of the status quo. That elected representatives would be able to do this would not be surprising because they are, after all, professional politicians, skilled and trained in justifying their actions and policies and explaining why there are no feasible alternatives. For constituency jurors to effectively challenge representatives on their conduct and, if necessary, impose sanctions, they would need the time and opportunity to inform themselves about the elected representative’s positions (and especially alternatives to them) and develop the confidence to disagree. Without a prior opportunity to deliberate, jurors would be wholly reliant on the knowledge they themselves bring to the session and what the representative tells them. Jurors who simply showed up on the day of their session with the representative could easily end up deferring to the seemingly greater wisdom and experience of the elected official. Furthermore, if jurors only served for a single session, it would be difficult to provide consistency over time, with issues easily repeated or forgotten from one session to the next.

There is a valid concern that these internal deliberations would be dominated by elite jurors. That could be mitigated through the selection of chair positions and speaking time through lot and the use of professional support staff to provide information and moderation. Concerns that the latter might in turn open the way to bureaucratic capture would be assuaged by jurors’ overlapping terms of office, so that more experienced jurors would be able to help with the training and integration of new jurors (Gastil and Wright Reference Gastil, Wright, Gastil and Wright2019, 20). Bureaucratic capture could also be addressed by having any official information on legislation provided to jurors checked by the political parties represented in the legislature and supplemented by their own party position papers on recently passed or impending legislation. Having jurors attend multiple sessions might expose them to external lobbying and oligarchic capture, but those elements would be tempered by jurors’ relatively short three-month terms of office and the consequent costs involved in targeting and influencing a continually rotating set of new jurors.

Making service in the constituency jury mandatory is justified by the need to make its composition as descriptively representative of the constituency as possible and the comparatively low burden that service would place on jurors. If jury service was voluntary, jurors would likely be disproportionally drawn from those with the time, resources, and interest to engage in politics, biasing selection toward the privileged and better educated, as has, in fact, been observed in real-world uses of sortition (Blome and Hartlapp Reference Blome and Hartlapp2023, 489–90; Fabre et al. Reference Fabre, Apouey, Douenne, Fourniau, Giraudet, Laslier and Tournus2021, 7; James Reference James, Warren and Pearse2008, 109–13): this would in turn risk the institution’s anti-oligarchic promise. Sortition proposals often reluctantly resort to voluntary service because of the high burden they place on participants (Gastil and Wright Reference Gastil, Wright, Gastil and Wright2019, 19–20; Landa and Pevnick Reference Landa and Pevnick2021, 57, n25; Vandamme and Verret-Hamelin Reference Vandamme and Verret-Hamelin2017, 11–13; Zakaras Reference Zakaras2010, 469, n28). Citizen-legislators in a lottocratic assembly or second chamber selected by lot, for instance, would be required to move to the capital city for months or years and bear the brunt of national public scrutiny. In contrast, disruption to the lives of constituency jurors would be comparatively low, because they could continue to live in the constituency and would only be required (in this design) to serve for three weeks spread across three months. The demands placed on constituency jurors would thus be similar to the widely supported existing duty to serve on legal juries. In sum, constituency juries have a significant anti-oligarchic advantage over other sortition proposals in that mandatory service, with its associated descriptive representation benefits, is a viable option.

Mandatory service would need to be paired, of course, with provisions to ease the burden for selected jurors as much as possible. It is critical for jurors not only to receive adequate compensation for their time but also to be robustly protected against employer retaliation by ensuring that they do not lose their jobs during or after their service; they should also receive free care for any dependents, so that constituency juries do not unfairly disadvantage exactly those less privileged jurors whom they aim to politically incorporate. Mandatory service would have the ambiguous implication of requiring nonvoting constituents to serve in the jury, hence leading to a potential discrepancy between the jury and the constituency’s voting population. But because voting is correlated with class, mandatory service would mean that representatives are likely to face a jury that is less wealthy than the constituency’s voting population and more closely matches the constituency’s whole population, hence contributing to the institution’s anti-oligarchic potential.

The specific institutional design of constituency juries proposes that jurors have two forms of sanction available to them: a motion of censure and the power to trigger a recall election. (Other potential sanctions, not considered or defended here, include the immediate removal, suspension, or disbarment from office, as well as monetary fines or even imprisonment.) These two sanctions give jurors the choice between a lighter and a heavier penalty. A censure motion would describe the representative’s actions and explain why the jury finds them objectionable and deserving of a public reprimand. Such a motion might appear toothless but could lead to significant embarrassment for the representative, especially if given prominence by local media and competing political parties. Provision might even be made for the motion to be automatically distributed to the addresses of all constituents in the constituency. Censure would also provide a useful early warning system for the representative before jurors imposed the heavier sanction of triggering a recall election.

If jurors triggered the procedure, the elected representatives would be forced to campaign to keep their seat, and the constituency’s voters would have the choice to reelect them or elect a new representative. A motion to trigger a recall should require a majority (or perhaps even a supermajority) of jurors to vote in favor in two consecutive jury sessions. Because jurors’ terms of office are overlapping, the latter measure would mean that any successful motion to recall would have been approved by a new injection of jurors. That would mitigate the chances of a recall motion being passed by a jury that is not representative of the wider constituency (which, as we saw earlier, will inevitably occur because of sampling errors). Minimizing that possibility is critical to avoiding juries triggering a recall not because of any failing of the representative but perhaps because that particular random selection of jurors never voted for the representative and their party. Additionally, any decision to trigger a recall election would always be made after the representative had been given a deliberative opportunity to explain themselves. This would help ensure that the sanction of triggering a recall election would be reserved for especially grave cases of democratic misconduct.

Two cases from recent British political history can serve as illustrative examples of such egregious breaches of a representative’s democratic mandate. First, Liberal Democrats voted in 2010 to nearly triple tuition fees for university students from £3,225 to £9,000 a year, despite having been elected just six months earlier on a platform explicitly and vocally committing them to opposing any rise in fees (Atkins Reference Atkins, Thackeray and Toye2020; Butler Reference Butler2021). Second, the Labour government’s decision to join the US invasion of Iraq in 2003, despite the largest protests in the country’s history and polling consistently showing strong public opposition up until just before the invasion (Strong Reference Strong2017, chap. 3). These cases illustrate how the independence accorded to elected politicians in representative government allows them to break pre-election pledges and override public opposition to issues emerging during their term.

If constituency juries with the power to trigger recall elections had existed in 2003 and 2010, it is possible that these decisions would never have been made. National campaigns by students and antiwar groups, paired with the prospect of recall elections being triggered in constituencies across the country might have been enough to convince Labour and Liberal Democratic politicians to pull back from these decisions. British students might today not be mired in tuition-fee debt; more consequentially, Britain might never have joined the catastrophic war in Iraq. This is speculative, of course; the deliberative component of constituency juries would rightly have given MPs the opportunity to explain/justify their votes, and those explanations may well have convinced their juries. In the case of tuition fees, Liberal Democratic MPs might, for instance, have pointed to the need for compromise in their coalition agreement with the Conservatives—a justification that many jurors would likely not have found persuasive, in this case at least, given how decisively the party was punished at the subsequent election (Cutts and Russell Reference Cutts and Russell2015).Footnote ¹³ The point is to illustrate both how constituency juries might operate as a firewall against these kinds of decisions and the importance of having the significant sanctioning power of triggering a recall election to realize that possibility.

At the same time, the power to trigger recall elections does raise the serious concern that, by making constituency juries more powerful, attempts at its oligarchic capture would inevitably emerge. It is easy to imagine corporate interests mobilizing to oppose impending legislation on climate change or corporate taxation, either by directly targeting individual jurors or by running well-financed national media campaigns to shift public opinion (and thus that of jurors as well) in favor of a recall to block said legislation. That such efforts will sometimes succeed is unavoidable. The challenge is to ensure that they happen less often than in existing representative government. The counter-oligarchical mechanisms discussed earlier, such as short, overlapping terms of office, would provide an important initial layer of protection. These would need to be supplemented by strict legal controls on outside contact with jurors during their service to protect them from lobbyists (Abizadeh Reference Abizadeh2021, 802).Footnote ¹⁴ But perhaps most importantly, constituency juries do not displace the need for political mobilizations in defense of anti-oligarchic policies. Campaigns countering the actions of wealthy and corporate interests by trade unions, civil society groups, and competing political parties would still be necessary. These would in turn be able to launch their own national campaigns in favor of recall to block oligarchic policies. Empowering constituency juries with the power to recall provides a better terrain for the pursuit of anti-oligarchic politics but does not rule out the need for politics as such.

In addition to oligarchic capture concerns, there is a deeper democratic worry with recall and constituency juries: whether it is legitimate for them to trigger a recall election against a representative elected by the entire constituency. Jurors could be seen as making a decision for the constituency without being accountable to them (Mansbridge Reference Mansbridge, Gastil and Wright2019a) or even subjecting the constituency to a kind of “blind deference” to their decisions (Lafont Reference Lafont2020, chap. 4). These concerns could be partially allayed by making jurors somewhat accountable for their decisions. In addition to being held legally accountably through sanctions for jurors who take bribes or submit to other forms of corruption, one could require jurors to justify their decision to recall through a collective statement to the constituency, thereby upholding the deliberative aspect of accountability. Moreover, although the ability to trigger a recall election represents a considerable power, it does not supplant the democratic rights of constituents: they would still be able to reelect the same representative at the recall election. Constituency juries would effectively be offering constituents the opportunity to select a new representative based on their own conclusion that such a decision may be warranted, but the final decision-making power would still rest with constituents. The triggering of a recall election would represent some cost to the electorate and to the representative’s party supporters in terms of campaigning for a new election. But that would be outweighed by the expected democratic dividends to be gained from the institution of constituency juries.

Conclusion: Prospects and Limits of Constituency Juries

Constituency juries are intended to improve representative government’s resilience to oligarchy by using sortition to make elected representatives more accountable to their constituents. Although some defenders of sortition have made the case for completely replacing elections with sortition and have in turn been criticized by those who want to maintain representative government’s existing electoral setup, this article was grounded on the premise that the most promising anti-oligarchic approach lies in combining election and sortition. Constituency juries expand the existing constitutional toolbox of anti-oligarchic combinations of elections and sortition by proposing a novel way to bring them together. Rather than national-level sortition assemblies that horizontally share power with an elected assembly, constituency juries hold each elected representative vertically accountable to their constituency. They thereby present an especially promising way to directly target the independence of representatives and make them answerable to their constituents, while also being institutionally resilient to attempts at oligarchic capture. This article consequently aims to have broken new ground, and I hope it will stimulate further discussion on using sortition to check oligarchy and enhance accountability at the level of the constituency.

Yet like any institutional innovation, constituency juries may have unintended side effects that make them potentially less attractive than existing representative government and other sortition proposals. One especially prominent concern is that heightened accountability to the constituency may negatively disrupt the operation of political parties by limiting their ability to control and direct their representatives. Given that this article has been guided by an underlying assumption that parties have an important anti-oligarchical function, undermining them would potentially undermine the core aim of constituency juries. There is consequently good reason to integrate political parties into the operation of constituency juries wherever possible. I earlier suggested that parties could be involved in the supervision and provision of information supplied to juries for their internal deliberation. One might also consider giving past or prospective candidates from competing political parties the right to present an opposing view at the constituency jury sessions. If such measures were, however, to prove insufficient and parties were indeed undermined by constituency juries, that might be one reason to limit the sanctioning powers available to them (perhaps by removing the power to trigger a recall election). But we should also bear in mind that even though political parties can play an anti-oligarchic function, that is not always the case (Bagg and Bhatia Reference Bagg and Bhatia2022, 352–57). Parties are often themselves committed to oligarchy (though rarely explicitly so) or else easily captured and turned away from their anti-oligarchic commitments. Through their popular supervisory function, constituency juries present the possibility not only of checking the actions of oligarchical parties but also of keeping a watchful eye on parties ostensibly committed to overcoming oligarchy.

The popular supervisory function of constituency juries is intended to curb the oligarchical tendencies of representative government by targeting the independence of representatives and the elite composition of public office. It is important to stress, however, that constituency juries do not address all drivers of oligarchy, including the international constraints faced by states and the background cultural and political factors influencing the formation of citizens’ preferences. Their ability to combat oligarchy by making elected representatives more responsive to the preferences of the non-elite citizens of the constituency hinges on the prior assumption that those citizens do not share the preferences of elites. Where the preferences of constituents and jurors align with those of wealthy elites, the constituency jury will not prevent elected representatives from acting on the latter’s behalf. Constituency juries are also unlikely to challenge elected representatives on issues where elite and non-elite preferences may differ in the abstract, but where those issues remain unaddressed or suppressed in public debate and absent from party political competition. Constituency juries cannot be expected to magic into the air action on those issues. Their creation thus does not negate the need for social mobilizations and political campaigns that change the preferences of non-elite citizens, bring issues to the forefront of the political agenda, and bind them into a coherent and electorally successful political program. Constituency juries, like any realistic sortition proposal, depend on the prior and concurrent operation of political parties, civil society organizations, and social movements if they are to effectively challenge oligarchy (Bagg Reference Bagg2024b, 207–11; Sintomer Reference Sintomer2023, 246–47).

What constituency juries do promise is that, where there has been effective popular mobilization on an issue, action on that issue is less likely to be blocked by the oligarchical trappings of the political system itself. Thus, they do not so much guarantee anti-oligarchical outcomes as help level the political terrain for anti-oligarchical action (Wright Reference Wright2018). Taking inspiration from a recent wrongheaded call for “10% less democracy” (Jones Reference Jones2020), we could instead hold a modest expectation that the institution of constituency juries might lead to 10% less oligarchy. Taken together with other political reforms and social mobilizations, constituency juries could form part of a more comprehensive challenge to the oligarchical trappings that have dogged representative government.