Trends in Theorizing Sortition

Introduction

At one time, the idea of selecting political officials by lot – a practice known as sortition – would have struck most political scientists as quixotic at best. This despite the prevalence of the practice in the Athenian democracy of the fifth and fourth centuries BCE, as well as the contemporary experience of the Anglo-American jury. But over the past fifty years, political scientists have increasingly seen sortition as a potentially useful innovation, one well-suited for addressing the increasing difficulties facing the democratic world. This has led to the development of a variety of institutional proposals for incorporating sortition into democratic politics. These proposals have not remained hypothetical possibilities. Sortition is employed to select citizens taking part in a growing number of real-world democratic experiments – in the Citizens’ Assemblies of British Columbia in 2004 and Ontario in 2007; in the Dutch Burgerforum in 2006; and in Ireland’s Convention on the Constitution (and subsequent Citizens’ Assemblies) in 2012-2014, to name a few examples (Farrell & Stone 2020). ¹

Political scientists have studied these experiments to examine how they function – to ascertain, for example, the quality of deliberation attained by the randomly-selected citizens taking part. ² Parallel to this empirical literature, however, is a theoretical literature that analyzes the normative properties of randomly-selected citizen bodies – especially their democratic credentials – and then uses this analysis to elaborate and defend more ambitious reform proposals based upon sortition (e.g., Van Reybrouck 2016; Hennig 2017; Landemore 2020). This theoretical literature has grown exponentially in recent years, with a symposium appearing in English in Politics in Society in 2018 and another appearing in French in Raisons politiques in 2021, alongside numerous independent studies. This review will examine many of the most prominent contributions to this theoretical literature in recent years, including both of the aforementioned symposia. It will conclude that there is still serious controversy regarding the democratic credentials of sortition, and that this controversy is reflected in the different proposals for further sortition-based reform.

Two points before proceeding. First, random selection can be productively employed to make all sorts of decisions (Stone 2011). Many of these decisions involve allocation problems – admitting students to oversubscribed schools, for example. The selection of citizens to take part in political decision-making is thus only one of many tasks that random selection can perform. This review confines its attention to the recent literature on this specific task only and reserves the use of the term “sortition” for when random selection is used for this task.

Second, a quick word about terminology. There is no universal agreement on a term to use for assemblies of citizens selected via sortition and engaged to perform some public task. The most commonly-used term in the literature is “deliberative mini-public,” although not all the papers discussed here rely upon it. Verret-Hamelin and Vandamme (2021, 25) introduce the term “randomly selected chamber,” or “RSC” for short. This term highlights the importance of sortition to such bodies, especially when they are incorporated into the political process in some way (as opposed to being used, for example, as demonstration experiments by an NGO or academic researcher). As such, this review will employ this term throughout.

Sortition Proposals

At present, RSCs are employed primarily as advisory bodies; at most, they provide input to be taken up by another body (e.g., an elected legislature), usually at the latter’s discretion. Most sortition theorists today want to take RSCs beyond such limited and circumscribed roles. At the same time, however, few sortition theorists envision a democracy based primarily upon sortition. Bouricius (2018) is an exception; he sees little to no benefit in maintaining any elected chamber, and instead favors a multi-body system of RSCs (see also Bouricius 2013). Most other sortition theorists envision the continued existence of at least one elected chamber alongside a strengthened RSC. ³ There are many variations upon this basic theme. The most frequent of these variations would overhaul bicameral elected legislatures entirely, either by replacing one of the two elected chambers with an RSC or adding an RSC to sit beside the other two. The former is the more common proposal; detailed versions of it are made by Gastil and Wright (2018), the focus of the Politics & Society symposium, and by Abizadeh (2021), who would change the selection method for the Canadian Senate from appointment to sortition. Verret-Hamelin and Vandamme (2021) are agnostic between the two options. Courant (2021) catalogs these and similar institutional combinations.

Either of these proposals would effectively give citizens selected by sortition a role equal, or at least comparable, to that of elected politicians. The specific powers to be given to the RSC, however, vary from proposal to proposal. Gastil and Wright (2018, 304) would give both chambers equal powers. Abizadeh (2021, 799) would assign his randomly-selected Canadian Senate much the same task as is currently performed by the appointed one. Sintomer (2018, 352-353) writes that “A sortition chamber should […] have three main tasks: defining the rules of the political game; proposing solutions to highly controversial issues […] and legislating for the long term.” Verret-Hamelin and Vandamme (2021, 25) assign to their RSC the “mission to select law projects for which a second reading seems required (thus delaying the law-making process and making it more deliberative), propose amendments and recommendations, and to initiate new law projects,” with the elected chamber(s) still having the final word. Similarly, Drew (2019, 631-632) imagines bicameral local governments, with an RSC exercising “[w]eak veto powers (the ability to suspend some programmes or the expansion of existing programmes),” and possibly some additional powers. ⁴

Other combinations of election and sortition are possible. Fishkin (2018, 360), for example, rejects the idea of a standing RSC operating alongside an elected one. Instead, he envisions an occasional temporary RSC that would “not do the actual lawmaking but rather filters the agenda […or] provides final approval.” He also, however, considers the possibility of an agenda-setting RSC generating proposals for ratification via referenda. Owen and Smith (2018, 420) also reject the idea of a standing RSC. They fear that “without the application of regular rotation and a limited mandate, the sortition chamber is unlikely to be able to defend its members against asymmetries in social and economic power.” They would create larger but more short-term RSCs, and restrict their function to the scrutiny of proposals; they contend that granting an RSC an agenda-setting role, as elected legislatures currently enjoy, is “utopian” and would require “a whole-scale restructuring of the broader institutional ecology” (ibid., 430).

The Value of Sortition

What would such combinations achieve? In particular, what would sortition contribute to these new and (allegedly) improved democratic processes? Most theorists of sortition agree that the practice can generate bodies capable of high-quality deliberation. They point to two features of sortition enabling this. On the one hand, sortition at the individual level ensures that participants are not preselected by any interested party. They are also not selected due to their allegiance to any cause (such as a political party or wealthy campaign donor), and so can approach deliberation with an open mind (Sintomer 2018, 352). Impartiality is a virtue of sortition frequently stressed in the literature (e.g., Abizadeh 2021, 792). On the other hand, sortition at the collective level can generate a sample that is descriptively representative of the people as a whole (Abbas & Sintomer 2021; Malleson 2018, 403-404; Sintomer 2018, 353). This ensures that a wide diversity of perspectives take part in the process, and no voice in society is systematically excluded. The recent literature on the “wisdom of crowds” – notably the Hong-Page Theorem (Hong & Page 2004) – stresses the contribution that diversity can make to high-quality decision-making, and so this property is invoked by many proponents of sortition (e.g., Landemore 2020). Gastil and Wright (2018, 307) invoke both features of sortition and conclude that it provides good reason for believing it may realize government for the people better than election (see also Malleson 2018, 409; Vandamme et al. 2018, 386).

Different theorists of sortition put sortition’s contribution to high-quality deliberation to different uses. Both Vandamme (2021) and Wright (2018), for example, take an explicitly Marxist approach, with the latter viewing RSCs as creating the conditions for an understanding of injustice to emerge and the former viewing them as vehicles for overcoming false class consciousness. The most detailed and innovative discussions of the connection between sortition and deliberation, however, appear in papers by Verret-Hamelin and Vandamme (2021) and Southall (2018).

Verret-Hamelin and Vandamme (2021) consider four arguments why an RSC could be expected to improve deliberative quality regarding environmental issues, such as the climate crisis. One of the arguments – the “generational rebalancing argument” (ibid., 26), based upon the underrepresentation of youth in elected legislatures – is plausible to them but not as compelling as the other three arguments. These three arguments are as follows: 1) sortition will increase the inclusion of marginalized voices, including those with “environmentalists’ beliefs;” 2) randomly-selected representatives are “no longer constrained to the short-termism of electoral cycles,” and so can envision more long-term strategies (cf. Sintomer 2018, 352); and 3) non-elected representatives will not be beholden to corporate interests, which “tend to sacrifice the long-term view in favor of short-term earnings” (Verret-Hamelin & Vandamme 2021, 30-37).

Southall (2018), by contrast, sees sortition as a solution not to environmental problems, but to dangerous levels of partisanship. Southall worries about the declining quality of democracy in South Africa, which he links to the country’s “national list PR electoral system.” This system leaves MPs “accountable to their party bosses rather than to the voters,” both because they depend upon those bosses for their nomination and because they lose their seats if expelled from the party (ibid., 252-253). The result is a parliament incapable of checking the increasingly bad behavior of the dominant African National Congress (ANC), given that intraparty dissension is effectively squelched. Reform of the electoral system is badly needed, but the elites benefitting from the status quo are unlikely to support it. An RSC, by contrast, would be able to act impartially, without regard for the demands of the ANC and other parties. For this reason, “there is a strong case for referring the case for electoral reform to the processes of sortition rather than simply leaving it in the hands of the politicians who benefit from the present system” (ibid., 254). ⁵

Other sortition theorists agree with Southall in associating the practice with nonpartisanship. Moreover, they see partisanship as a force capable of undermining the high deliberative quality that RSCs can attain (e.g., Malleson 2018, 412). They differ, however, in the conclusions they draw from this fact. Gastil and Wright (2018, 322) defend the continued existence of an elected chamber in part because of the importance of maintaining a role for political parties. For them, “political parties can play an important role in formulating political programs, educating the public about policy alternatives, formulating broad visions for social change, and mobilizing people for collective action.” Bouricius (2018, 436) and Vandamme (2018, 121-122), by contrast, believes that political parties will continue to have a role in a sortition-dominated system, similar to that of think tanks, activist groups, or civil society organizations. This position is critiqued by Urbinati (2021) and Cordier (2021), both of whom view parties as essential to any society marked by inherent division. A democracy based exclusively upon sortition effectively denies the existence of such divisions, and so is “anti-political” in a dangerous way (Cordier supports this conclusion with an analysis of the Icelandic experience with sortition). Tridimas (2019) approaches the problem from a different angle by considering the historic experience of the Athenian democracy, which notably had no organized political parties. Tridimas investigates the institutional conditions that prevented the emergence of parties – including sortition – and considers how political leadership functioned without them.

Sortition and Democracy

There is significantly less agreement regarding sortition’s relationship with democracy more generally. “Democracy,” as one paper writes, “means rule by the people, not by elites, and if this were fully realized, the people would themselves actually rule” (Gastil & Wright 2018, 304; cf. Malleson 2018, 406). Sortition theorists certainly agree that sortition by its very nature leads to decision-making bodies not dominated by elites. And for some sortition theorists, this bypassing of elites is one of the major accomplishments of the process. Sintomer (2018, 344-345), for example, extols the combination of mini-publics with direct democracy, apparently because of the extent to which this combination excludes elites in favor of ordinary citizens; “the coupling of a sortition chamber and direct democracy,” he writes, “seems promising” (ibid., 353). ⁶ But doesn’t popular rule require something beyond this? And if so, what? Sortition theorists disagree about this, and this disagreement informs the role they envision for sortition in a democracy (a point noted in Abbas & Sintomer 2021, 34-35).

As noted before, sortition can “select a microcosm of the citizenry, a group with the same diversity as the citizenry, but on a smaller scale” (Sintomer 2018, 342). This matters for purposes of cognitive diversity, but some theorists say it also matters for purposes of popular rule. Whenever a contemporary RSC generates a microcosm of the citizenry, it can create “a counterfactual opinion that is representative of what the larger public opinion could think” (emphasis in original; ibid., 343-344). This is a point Fishkin has stressed throughout his long history of work on the topic (e.g., Fishkin 1991); a properly-constructed RSC can demonstrate “what the people would think about the issue under good conditions for thinking about it” (emphasis in original; Fishkin 2018, 374).

To what extent, however, does this matter for popular rule? Malleson believes this sort of counterfactual opinion can generate a form of popular control, which he describes (following Pettit) as “indicative.” According to Malleson, “if the descriptive representation works well so that some of the members are ‘like me’ in their values, then there is indeed a kind of popular control here” (Malleson 2018, 406-407). But Malleson is alone in reading this as “control” in any meaningful sense. ⁷ Both Fishkin and Sintomer, for example, reject this association, and view this counterfactual opinion as insufficient for democracy. In Fishkin’s case, it is because of the tradeoff between different democratic values (more on this shortly). In Sintomer’s case, the appeal to this counterfactual opinion stands in contrast to Athens, which was characterized by “radical democracy and self-rule” (Sintomer 2018, 339; cf. Abbas & Sintomer 2021). Sintomer does not carefully define “radical democracy,” which he seems to regard as “true” democracy, but he does mention that “In Athens, sortition and the rapid rotation of offices enabled citizens to govern and be governed in turn” (ibid., 342). He thus appears to imagine radical democracy as a process that allows every citizen to take a turn as a “governor,” however defined.

Sintomer does not appear to endorse the elimination of elections; “democratic innovations are always hybridized,” he writes, and an RSC serving as a “third legislature chamber” is a more realistic possibility (Sintomer 2018, 352). But he seems to view sortition as providing the only truly democratic component of any such arrangement, especially if it makes radical democracy possible. Other sortition theorists, however, typically see a tradeoff between democratic values, which generates the need to combine election and sortition. Gastil and Wright (2018, 307), for example, motivate their proposal for a mixed bicameral system by appealing to the idea of democracy as “government of, by, and for the people.” Elections render possible government by the people, whereas sortition enables government of the people (cf. Sintomer 2018, 342). Direct democracy could in principle facilitate both, but it is infeasible beyond the local scale in any serious way. And “Which method best governs for the people – that is, in the public’s interest – is an empirical question ultimately settled only through experience” (emphasis in original; Gastil & Wright 2018, 307). ⁸

Other theorists analyze the democratic tradeoff between election and sortition in different terms. Abizadeh (2021, 792) takes “the democratic tradition to be defined by two constitutive commitments: to people’s political agency and political equality” (emphasis in original). In a representative (not a direct) democracy “these two constitutive commitments face an inherent trade-off; realizing the one to some degree unavoidably compromises the other.” Election can realize agency (as well as accountability) at the expense of equality (as well as impartiality); sortition promises the opposite tradeoff. A bicameral legislature with different selection methods, then, offers the best method for ensuring both constitutive commitments are honored. Fishkin (2018, 362-363) views electoral democracy as realizing the democratic values of political equality and non-tyranny, while deliberative democracy (via RSCs) enables political equality and deliberation. (Participatory democracy, through referenda and the like, can realize mass participation but is less reliable on the other three values.) Other authors are not as explicit, but echo the same general idea. Malleson (2018, 412) sees “the main strengths of the electoral mechanism as accountability and competency, whereas its main weaknesses are generating political inequality […] and a systemic propulsion towards partisanship that undermines deliberation and impartiality.” He concludes from this that “because neither election nor sortition by itself can satisfy the full range of democratic values, an optimal democratic system would need to combine both mechanisms,” through a bicameral system or the like (emphasis in original; ibid., 413). Verret-Hamelin and Vandamme (2021, 27) suggest that “the legitimacy generated by elections is imperfect and can be usefully complemented by a distinct form of representation, carrying new sources of legitimacy.” Importantly, while Bouricius (2018) – the sole proponent of replacing election with sortition completely – considers the presumed benefits of election, none of the benefits he considers reflect core democratic values. If elections are truly undemocratic to their core, it seems hard for any theorist to defend their continued use. ⁹

Different sortition theorists, then, offer different (but similar) lists of competing democratic values – some associated with election, others with sortition. It is one thing, however, to lay out the tradeoff; it is quite another to show how to make it. If an elected chamber accomplishes X and an RSC accomplishes Y, it does not follow that a mixed bicameral system would accomplish both X and Y. Abizadeh (2021, 792) acknowledges the need to explain why a mixed bicameral legislature would realize his preferred democratic values better than other hypothetical arrangements. Vandamme et al. (2018) pursue the same topic from a different angle. They present survey data from Belgium regarding the perceived democratic credentials of different combinations of election and sortition – two equal chambers, an elected chamber subordinated to an RSC, an RSC subordinated to an elected chamber, and a single mixed chamber (further analysis of this survey data is provided in Jacquet et al. (2022)). The data comes from both ordinary citizens and members of parliament (unsurprisingly, the former are more enthusiastic about sortition than the latter). They then use this data to explore the advantages and disadvantages of these combinations. They conclude (inconclusively) that “Elections and sortition have complementary virtues that offer a promising prospect for being combined in a bicameral legislature, but that “The reintroduction of sortition into modern representative systems will be difficult to achieve, and sortition will likely face continued resistance once installed” (Vandamme et al. 2018, 392).

Conclusion

There is now a broad consensus regarding the benefits that a body filled via sortition can provide if properly constructed: descriptive representation, especially of groups under- or unrepresented in elected legislatures; independence; and impartiality, to name just three. All of these can, in turn, contribute to improved decision-making. But proposals to embed such bodies directly into the political process – that is, proposals for RSCs enjoying real political power – raise new questions about the democratic legitimacy of sortition, especially as compared to election. As experimentation with sortition continues, so will the empirical data on the functioning of RSCs under various conditions develop. But no amount of data can establish the respective democratic credentials of election and sortition. To what problems with election-based systems can sortition provide the solution? Partisanship? Cognitive bias? Lack of deliberation? Partiality? Distrust of politicians? Different answers will point to different institutional schema involving RSCs. The inclusion of sortition governance in the political process can mean a radical overhaul of current decision-making systems in the name of popular sovereignty; or it can mean constructing small, consultative bodies lacking even agenda-setting powers. Further theoretical reflection on the place of these mechanisms in a democracy is therefore likely to prove very beneficial in guiding future experimentation with sortition, hopefully to the benefit of democratic practice. These reflections will both build on and improve our understanding of the nature of democracy.