1 Introduction

Democracy is in deep crisis. Democracy’s durability and credibility are increasingly under strain, facing significant challenges in both structure and practice. The electoral success of leaders with authoritarian tendencies in democratic nations, such as the United States, underscores a troubling shift toward authoritarianism within democratic systems. Simultaneously, the hollowing-out of democracy has led to fragmented and disjointed governance, eroding public trust and the effectiveness of democratic institutions.

Scholars have extensively explored the theme of democracy's crisis, beginning with Crozier et al. (1975) Trilateral Commission report, followed by Habermas’s (1975) seminal work on the legitimacy crisis that same year. Subsequent contributions include Dahl’s (1989) Democracy and Its Critics in 1989, Huntington’s (2004) analysis of the political community’s crisis amidst multiculturalism in 2004, Tormey's (2015) examination of the end of representative politics, and more recent examinations by Adam Przeworski (2019).

To deal with the crisis of democracy, a lottocratic turn among theorists of democracy—a trend toward preferring democracy based on lotteries for the selection of decision-makers—has appeared in recent decades, with the intention of saving democracy from the flaws into which electoral systems have led (Brennan 2016; Achen and Bartell 2017). Lottocracy is a political system in which individuals are selected for political positions or decisions through a random lottery, placing ordinary citizens at the heart of governmental decision-making. It is believed by its proponents to decrease government corruption and increase citizen participation and trust in democracy (Waxman and McCulloch 2022).

The theory of sortition democracy raises the following critical questions. Is the theory of sortition democracy a correct remedy for the crisis of contemporary democracy, or is it a misguided prescription? Does the resurgence of sortition truly reflect an accurate understanding of Athenian direct democracy?

This paper aims to address the above questions through a theoretical and empirical enquiry. It critically evaluates the concept of lottery democracy as a purported remedy for the crisis in Western democracies. It also scrutinizes its political proposals and programs through examining the practical workings of the random selection process. My goal is to critically consider the assumptions and complex issues involved in random selection processes within their various proposals for sortition democracy, thus helping us to move away from radical alternatives and adopt a more moderate approach to sortition democracy.

To address the above questions, a key approach is to critically examine the conceptualizations of randomness and its underlying logic. The argument for sortition democracy can be found in the logic of randomness. Random selection is considered to provide everyone with an equal chance of being chosen. The inherent uncertainty and fluidity of randomness help mitigate the risks of corruption, domination, and manipulation. Both Oliver Dowlen (2008) and Peter Stone have emphasised the exclusion of all rational evaluation in every true lottery. In the word of Peter Stone (2009: 375): “A lottery contributes to decision making by preventing the decision from being made on the basis of reasons.” This logic of randomness has a sanitizing effect, contributing to the decision-making process. Celebrating this kind of randomness in the exclusion of reason might reflect a longstanding anti-intellectual tradition in democratic thought, envisioning a political society where there is no leader, or where every citizen is a leader. However, this conceptualization of the logic of randomness is problematic in real political life.

I propose an alternative conceptualization of the logic of randomness, which expands the understanding of randomness by examining all aspects of the random selection process. Rather than focusing solely on the purportedly sanitizing effects of random selection, I examine several significant issues that emerge throughout the process. The logic of randomness reveals the inherent problems associated with sortition democracy: undermining empowerment, weakening consent theory, compromising substantive equality, and bypassing the authorization question. An analysis of the logic of randomness in Sect. 5 provides a compelling explanation for why various proposals for sortition democracy, despite their originality and appeal, have ultimately failed to be implemented in practice.

My conceptualization of the logic of randomness draws on Aristotle's theory of the mixed regime and China's practice of mixed governance over the past two decades. By revisiting Aristotle, I demonstrate that sortition theorists misinterpret both his ideas and the historical practice of sortition. Sortition as a selection method cannot function in isolation; it must be integrated with other political institutions and mechanisms. Sortition, elections, and deliberation must be combined to create a stronger democratic system—much like how a good dish depends on the right balance of salt, sugar, vinegar, and other spices. The effectiveness lies in the complementary interplay of these elements. The theory of sortition democracy cannot become an ontologically independent theory. From this theoretical perspective, I challenge the views of radical proponents of sortition who assume the existence of a pure form of sortition democracy—a concept that did not exist in ancient Athens. If so, their proposal to replace electoral democracy with sortition democracy is flawed and lacks a solid foundation.

I employ several methods in this study. My theoretical analysis involves rediscovering the value of Aristotle's theory of the mixed regime (Sect. 3) and constructing an intellectual framework focused on the practical dynamics of the random selection process (Sect. 5). My primary method is participatory observation. I have been directly involved in organizing and facilitating several deliberative polling experiments in China. This hands-on experience offers valuable real-world insights into the functioning of sortition. My critique of radical sortition democracy theory is rooted in this practical, real-life local experience, leading me to believe that we need a moderate sortition democracy approach. This paper has also adopted a case study approach. Choosing China as a case study provides a significant contribution to the field of sortition democracy studies, which predominantly focuses on democratic innovations in Western societies. China serves as a unique social science laboratory for testing theoretical claims about the benefits of sortition democracy, offering opportunities to advance a more universal understanding of its principles and applications (He 2025). Through this case study, I develop an experiential, observation-based analysis of how sortition functions within the context of Chinese local deliberative democracy.

The paper is divided into six sections. The first section is Introduction. The second section offers a critical review of the literature on lottocracy, presenting both supporting and opposing arguments. The third section returns to Aristotle’s theory of the mixed regime and argues that a purely sortition-based democratic model lacks historical precedent, particularly when considering the mixed system of ancient Athens. In the four section, the case of China is explored in detail to illustrate how random selection has been employed in local deliberative democracy experiments (Fishkin et al. 2010). These theoretical and empirical discussions lay the groundwork for examining the logic of randomness, specifically analyzing the theory of sortition through the lens of sampling practices in the fifth section. The sixth section is Conclusion. 

2 Lottocracy and Its Discontents

The terms lottocracy, lottery democracy, and sortition democracy are often used interchangeably (as of course scholars might have different views of these terms) to describe a political system where individuals are selected for political positions or decision-making roles through a random lottery. Advocates of lottocracy believe that this system could reduce corruption while enhancing citizen participation and trust in the political process. In recent years, sortition has been experimented with by governments and institutions in the USA, Canada, Iceland, Ireland, the UK, Australia, France, and other countries using randomly selected citizens' assemblies to address diverse issues (See Table 1). Permanent Citizens’ Council chosen by lot have been established in Paris and the German-speaking community of Belgium. Hundreds of such experiments globally are challenging the notion that elections are the only or ideal method for achieving credible and effective governance.Footnote 1

Table 1 A selected list of sortition-based democratic experiments worldwide
Full size table

Among the various theoretical arguments and practical proposals for sortition democracy, two prominent variants stand out: moderate and radical. The moderate approach seeks to complement and enhance electoral democracy by incorporating sortition mechanisms. In contrast, the radical approach advocates for the establishment of an entirely new lottocratic system to replace the current electoral and representative democratic framework.

Moderate proponents of the democratic lottery focus on specific contexts where the lottery method might be applied, such as selecting a portion of a legislative body like the Senate or choosing members of a review panel by lot. Stone and Malkopoulou (2021) propose three variations of sortition democracy—weak, moderate, and strong—and tentatively endorse the moderate version. Central to this model is the principle of random selection, which employs lottery-based mechanisms to appoint citizens to decision-making roles, thereby reducing elitism and promoting inclusivity. Moderate proponents endorse civic lottery: a lottery-based approach to citizen selection, utilizing a modern form of sortition that incorporates stratification to ensure the selected participants are representative of the broader community. This method accounts for key demographic criteria, ensuring that the cohort reflects the diversity of the population in terms of age, income, geography, political views, and other relevant factors. A civic lottery is used to create a descriptive sample of the population, forming the basis for deliberative bodies such as citizens' assemblies.Footnote 2

Below is a selection of significant proposals that utilize various sortition-based mechanisms to enhance the existing democratic system (Table 2). These proposals are arranged chronologically, allowing readers to trace their development and applications across various domains and issues. With the notable exception of Dahl’s proposal, most have not been implemented. An analysis of the logic of randomness in Sect. 5 provides insight into why these ideas, despite their originality and appeal, have ultimately failed to materialize in practice.

Table 2 A variety of proposals
Full size table

In contrast, radical proponents (Dowlen 2008; Buchstein 2019; Stone and Malkopoulou 2021; Waxman and McCulloch 2022) argue that the lottery principle should be applied across the entire political sphere, including legislation, with some proposing the creation of a “House of Lottery,” a legislative chamber formed entirely by lottery selection. They propose replacing universal suffrage with suffrage by lottery, where a randomly selected subset of the population is granted the right to vote. Additionally, they advocate for representation by lottery, which involves abolishing both universal suffrage and traditional voting methods, instead relying on random selection to choose political representatives (Umbers 2021).

The radical proponents advocate for a government founded entirely on random selection and citizen deliberation, offering a novel model of governance where elected politicians are replaced by assemblies chosen by lot. Their case for this radical model draws inspiration from the successes of ancient Athens, the centuries-long tradition of sortition in Venice, the effectiveness of randomly selected juries, and the philosophical value of randomness. They critically reassess the prevailing narrative that lotteries in ancient democracies were purely democratic, providing a historical account of randomly selected juries and introducing new research on lesser-known examples from England, America, and revolutionary France (Dowlen 2008). Later, in Sect. 3, I will argue that their interpretation of Aristotle and ancient Athens is flawed, as it overlooks the role of sortition within a mixed system, and complex connections between sortition, elections and deliberation.

Among radical proponents, David Van Reybrouck (2018)’s critique of modern representative democracy needs special attention. He argues strongly that the electoral system has become increasingly dysfunctional with systemic issues such as polarization, populism, and declining public trust in institutions as consequences of electoral politics. As an alternative, Van Reybrouck advocates for sortition—randomly selecting citizens for decision-making roles, inspired by practices in ancient Athenian democracy. He argues that sortition can revitalize democracy by fostering inclusivity, encouraging deliberation, and reducing corruption.

The concept of sortition is introduced into discussions of democracy in Asia. As part of a broader effort to democratize South Asia, Shah (2021) emphasizes the potential benefits of sortition in fostering a truly democratic society and advocates for developing both theoretical and practical frameworks for leaderless societies—where everyone could be leaders. Inspired by the concept of lottocratic democracy, Chao-Cheng Chen (2023) proposes concrete steps to integrate sortition democracy into Taiwan's Legislative Yuan. Wang (2018) has also introduced and advocated sortition democracy in China. The advocacy of sortition democracy in China appears peculiar in two ways. First, since China has not yet fully established electoral democracy, it is challenging to argue for a Chinese version of the replacement model—replacing electoral democracy with sortition democracy. Second, Chinese political management and governance in a meritocratic society have always emphasized that those in charge should possess virtues, talents, and competence. The idea of governance by randomly selected individuals seems alien to implement in China (though historically, some ethnic minorities did use random selection to determine their leaders).

The case for lottocratic democracy rests on its perceived advantages. It is believed that sortition has several democratic benefits including the belief in the people's ability to govern, ensuring equal opportunities, representativeness, difficulty in being manipulated by others, and fostering and developing a sense of equality in personality. It is also believed that sortition has other political benefits including reducing political division, curbing factional conflicts, promoting fairness, and enhancing the recognition of the political system (Wang 2018). Pope (2023) argues that reintroducing sortition could break through political gridlock, apathy, alienation, and polarization by restoring the voice of the citizenry. Sintomer (2023) points out that with the aid of sortition, it is unlikely for representatives to form a proper “elite” with significantly different interests to the citizens. Buchstein (2019) emphasises the advantages of lotteries for making efficient and productive decisions and for avoiding corruption. Akhil Reed Amar (1984), an early advocate of lottery voting and Sterling Professor of Law and Political Science at Yale University, identifies a significant flaw in electoral democracy: its inability to represent the entire community, often silencing minority groups within a majoritarian system. In this context, sortition is seen as a potential solution to this longstanding political issue. While sortition democracy can overcome two significant failings of modern-day systems of electoral representative government—the failure of responsiveness and the failure of good governance—Guerrero (2014) considers some possible problems associated with sortition democracy: competence, capture and participation.

Critics of sortition democracy frequently express skepticism regarding the claimed benefits of these proposals. Lachlan Montgomery Umbers argues that lottocratic systems necessarily involve the establishment of objectionable social and political inequalities in a way that democracies do not (Umbers 2021). Bagg (2024) argues that “replacing elected and appointed officials with randomly selected citizens would only exacerbate elite manipulation of political processes” All of this aligns with the classic critiques of sortition. Montesquieu (2022) believed that the value of sortition lies in providing all citizens with equal opportunities to serve the country. However, human nature tends towards laziness, making it difficult to manage public affairs. Rousseau (2018) thought that administrative positions are a heavy burden, and sortition allows this burden to be fairly distributed among everyone. Sortition can result in unqualified individuals becoming officials. Only those who obtain their positions through competition will be more careful in fulfilling their duties. Those who gain their positions by chance may not necessarily work with full dedication. While these critical perspectives are insightful, they have not been presented in a systematic way. My approach will be to offer a systematic analysis of the logic of randomness and the mixed system, demonstrating why the replacement argument is fundamentally flawed at the ontological, methodological, and empirical levels.

3 The Mixed System in the Ancient Athens

To support their theoretical claim, lottocratic theorists often cite Aristotle’s (1962: Chap. 9) famous remark that “the appointment of magistrates by lot is thought to be democratic, and the election of them oligarchical” (the quota in Guerrero 2014).However, this quotation overlooks two critical issues. First, in Aristotle’s time, property qualifications played a key role in determining whether a system was considered democratic or oligarchical; that is why Aristotle said that “democratic again when there is no property qualification, oligarchical when there is.” In the contemporary world, where there are no property requirements for citizenship and the right to vote, both election and sortition are regarded as democratic, and the election of magistrates is no longer considered oligarchical.

Second, this simplified quotation overlooks Aristotle’s emphasis on the importance of blending election and sortition. From Aristotle's theory of mixed government, both election and sortition are essential components of a political system and must complement each other. The advantage of sortition is that it can break the monopoly of aristocratic interests, while the advantage of elections based on property ownership is that it can prevent popular uprisings. The collapse of the Roman Republic stemmed from the breakdown of the dual power structure, leading to a loss of balance between the aristocrats and the common people (Wang 2018: 103–4). Similarly, the Republic of Florence experienced an imbalance between the forces of the aristocracy and the populace, which was closely tied to the disparity between sortition and election. A political system relying solely on sortition is impractical, just as an election-only system suffers from the issues identified by sortition theorists. Rather than focusing solely on the effectiveness of sortition, we should explore how a mixed model can more effectively prevent political cycles and enhance political stability, participation, and equality.

Aristotle (1962: 168) explored various methods to combine oligarchy and democracy. The first is to “take a characteristic piece of legislation from both [oligarchy and democracy]”. The method includes the poor being paid to serve on juries and the rich being fined if they do not, thus ensuring both classes participate in governance and thus share power. The second is to take a midway position between a high property-assessment in oligarchies and a small or a complete absence of property-qualification in democracies (Aristotle 1962: 168). It concerns the qualifications for citizenship, emphasizing that wealth should not be a barrier to participating in political life. It ensures that even those with limited financial means have the opportunity to engage in decision-making and share in governmental power. And the third method is to fill offices by lot (which is seen as a democratic process) and by selection (which is seen as an oligarchic process) (Aristotle 1962: 168).

Aristotle (1962: 253) also advocates the idea of the mixed regime which combines three political institutions—aristocratic law-observance office, oligarchy pre-council, and democratic council. It blends the rule of one, the rule of the few rich and the rule of the many poor. It is based on a mix of three social classes, the rich, the poor and the middle in order to manage and reduce the conflict between the rich and poor. The appearance of the mixed regime ought to “look like both [oligarchy and democracy] and like neither” (Aristotle 1962: 169); and ‘the aim of a mixture of oligarchy and democracy is merely to have regard to the interests of both rich and poor, both wealth and individual freedom’ (Aristotle 1962: 167). It prevents one group from dominating another, ensuring that both groups keep each other in check. Unchecked power is a root cause of decline and revolution. Aristotle certainly would reject Karl Marx’s revolutionary principle of the destruction of the capitalist class by the proletarian class, as well as today’s prevailing populist principle that justifies the people’s battle against so-called elites (Mansbridge and Macedo 2019).

Indeed, there was a mixed system in ancient Greek political life where sortition, together with election and deliberation, played an important role. Among the total 700–1200 official positions annually, around 500–700 officials were selected by lot, including the Boule (Council of 500), and about 74 positions were filled by popular vote, mostly for strategoi (generals), financial magistrates, and other roles requiring technical or military expertise. A small number (perhaps 3–10 positions) were appointed directly, often religious or ceremonial roles (Aristotle 1962). For the selection of generals, each tribe elected one. Under the ostracism law, 6000 people voted to decide on exile, with the exile period set at 10 years. During the period of tyranny, voting temporarily replaced sortition, but in 482 BC, sortition was reinstated. From this, it is evident that while voting was integrated into various political institutions and activities in Athens, it is difficult to characterize Athens as a system where officials were purely selected by sortition. On the contrary, the combination of voting and sortition was a hallmark of Athenian democracy. Athens did not merely implement sortition but also practiced rotation of public offices, short-term tenures, and a mix of sortition and voting depending on the context and the specific functions of the offices (Hansen 1991). Additionally, the ancient Greeks combined sortition with deliberative democracy to make public decisions. Through the Council of 500, the jury system, or by having randomly selected citizens engage in deliberation, citizens were granted significant public decision-making powers.

The Republic of Venice also developed a four-step method which combines election with sortition. First, the Great Council, composed of 480 members, would elect 11 people. Second, these 11 people would be reduced to 4 through sortition. Third, the 4 would then vote to elect 41 electors. Fourth, finally, the 41 electors would vote to choose the Doge (Wang 2018: 321–2). In 1268, this four-step method evolved into a more complex ten-step process. However, this ten-step method left room for high-level manipulation. Additionally, it violated the principle that political operations should be simple and feasible.

In the contemporary world, various proposals by the moderate school of sortition democracy (as outlined in Table 2) aim to establish a hybrid political system. Frank Hendriks (2010) argues that wherever democracy is successfully practiced, it is invariably as a form of hybrid democracy. This perspective challenges democratic reformers and theorists who pursue the ideal of democratic purity.

4 The Case of China

Democracy remains a highly contested concept in the Chinese context. While some studies suggest that Chinese citizens tend to favor deliberative forms of democracy over electoral mechanisms, others underscore the enduring relevance of electoral democracy. For instance, employing a novel method of automated text analysis on open-ended survey responses, Yang et al. (2024) find that Chinese citizens predominantly associate democracy with values such as freedom of speech, public consultation, and minben (people-centered governance), rather than with competitive elections. In contrast, Chang and Su’s (2024) experimental study involving 4000 Chinese university students reveals a substantial segment of young citizens who endorse democratic competition among political parties for top leadership roles—posing a direct challenge to the officially sanctioned notion of a singular “Chinese-style democracy.” Similarly, Lo (2021) critiques the Chinese Communist Party’s (CCP) strategic appropriation of deliberative democracy, arguing that the concept has been redefined to serve party objectives and ideological conformity. Further complicating the picture, Hu and Wu (2024) examine the CCP's promotion of “whole-process people’s democracy” (WPD), finding that although it nominally enhances citizen participation, such involvement is largely state-orchestrated. Citizens are encouraged to express their opinions, exercise limited forms of power and supervise government actions—but only within parameters set by the state. Nevertheless, WPD has arguably contributed to more responsive local governance by encouraging officials to engage with public opinion during the policymaking process.

Setting aside scholarly debates on the role of deliberative or consultative democracy in China, many local officials have faced intense pressure from frequent protests and social unrest, prompting them to adopt various deliberative or consultative mechanisms—such as participatory platforms, public hearings, or local meetings—as pragmatic tools to mitigate and manage social conflicts (Zhang et al. 2021). In doing so, in recent decades, some local democratic practices in China have adopted the method of drawing lots. Four communities—in Wenling, Shanghai, Guangdong, and Yunnan—have implemented and adapted random selection and deliberative polling (DP) methods to enhance participatory budgeting (PB). From 2005 to 2016 (with the exception of 2007), annual deliberative polling experiments were conducted in Zeguo township, Wenling city (He 2019).

Taking the case of Zeguo as an example, several factors contributed to the adoption of random selection (sortition) in local deliberative practices. First, random sampling is widely regarded as a fair and egalitarian method, as it gives every individual an equal chance of being selected. This approach enhances the representativeness and legitimacy of deliberative polling. Second, random selection prevents government officials from handpicking participants who might simply echo official priorities, allowing the local government to project an image of procedural neutrality and genuine democratic participation. Among the participants who were randomly selected in the 2005 Deliberative Polling experiment, the author also employed an additional randomization method in 2006 by choosing these individual participants with a number of 9. These individuals—about a dozen in total—were subsequently interviewed about their attitudes toward random selection. They expressed excitement about being chosen and conveyed a strong sense of fairness associated with the process. Over time, the use of sortition-based experiments in Zeguo evolved and improved, creating a cumulative effect that contributed to more effective local governance and greater citizen empowerment.

However, not all local governments have embraced random selection. In many towns and townships, direct selection or appointment of participants remains the dominant practice in public meetings. For example, Xinhe Township—a neighboring jurisdiction to Zeguo—explicitly rejected the use of random selection. Furthermore, when deliberative polling based on sortition proved successful at the township level, discussions emerged about whether this method should be scaled up to the city level. Yet city-level officials resisted such adoption, citing concerns over representativeness and questioning the applicability of sortition in larger, more complex administrative settings (for a detailed discussion of random selection method in China, see He 2008: Chapters 5 and 9).

Numerous political and cultural barriers continue to hinder the widespread adoption of random selection in China. In everyday life, the practice of drawing lots is not typically associated with democratic ideals. For instance, it is commonly used in temples for fortune-telling or as a means of fair distribution when resources are scarce. As such, the cultural meaning of sortition does not automatically align with democratic expectations, posing additional challenges to its broader institutionalization.

Additionally, the sortition practice brought to light the challenge of illiteracy, with approximately 9% of the sample being illiterate. There were two differing approaches to this issue. The former party secretary, Jiang Zhaohua, saw this as a positive development, as in the past, illiterate individuals would not have been included. In deliberative polling, these participants were given an equal opportunity to express their views vividly. Jiang believed that an open democratic process could help cultivate them into modern citizens.

On the other hand, the new party secretary, Zhao Min, observed the entire deliberation process and noted that illiterate participants were often quiet, leading to lower-quality public debate. To address this, Zhao Min introduced a mixed method in 2008. Under this approach, half of the participants were still randomly selected from the general population, while the other half were randomly chosen from elected village representatives who had a legal mandate for political representation. When James Fishkin asked the author to convey his objection to the mixed method to local officials, Zhao Min defended the approach, arguing that it was superior to pure deliberative polling. He believed that it not only increased the level of political representation but also enhanced the quality of deliberation, as elected village representatives generally have stronger argumentative skills compared to illiterate participants. Here we see the political logic of returning to or strengthening the elected deputies.Footnote 3

The local officials refined the deliberative polling method through a process of experimentation. They raised a key question from a participation perspective: can randomly selected participants truly represent those who are not chosen? To address this concern, ensure broader participation, and enhance representation and legitimacy, they required all randomly selected participants to visit at least ten families and gather their opinions on the issues to be deliberated. They believed this approach effectively increased political participation and representation and ensured that a wider range of public opinions, concerns, and complaints were represented. Local officials dismissed advice from Professor Bob Luskin, who argued that this methodology violated statistical requirements, focusing instead on the measure's ability to enhance political participation and representation.

Another strategy to address the exclusion inherent in random selection was the use of rotation. In ancient Greece, the rotation of citizens in courts and councils ensured equal opportunities for participation. Similarly, in Zeguo, the repetitive nature of deliberative polling (from 2005 to 2016) served a similar purpose: “if you are not selected this time, you'll have a chance next time.”

The Zeguo officials also addressed Parkinson’s (2003: 188) concern about "the legitimating bonds of authorization and accountability between participants and non-participants." To tackle this issue, they developed several strategies to involve the Zeguo Township People’s Congress in the deliberative process.

First, between 2008 and 2010, about 20–30 deputies were required to attend the panel sessions, allowing them to listen directly to the voices of the people. This provided the deputies with an opportunity to better understand the citizens they represent. In interviews conducted by the author, several People’s Congress deputies noted that this experience helped them gain a deeper insight into public opinion.

Second, ten randomly selected participants from the public deliberation were invited to observe the Township People’s Congress meetings where the 2008 and 2009 budgets were discussed and decided. This inclusion of citizens as observers helped to enhance government transparency.

Third, the outcomes of deliberative polling had to be discussed and approved by elected deputies in the local People’s Congress. This reflects a political logic of reinforcing the role of elected deputies, who are perceived as politically representative because they are chosen by the entire citizenry.

In summary, Wenling officials sought to combine random sampling with the existing representative system of the People’s Congress. They ultimately favored a hybrid approach that integrated elite-led deliberation and elected deputies with broad popular participation. This mixed method, as it has evolved in Chinese deliberative practice, addresses the challenges of legitimacy and representation in public forums.

5 The Logic of Randomness

The theoretical discussion above highlights that democratic politics often relies on a combination of sortition, election, and deliberation, providing a conceptual foundation for examining the logic of randomness. At the same time, the empirical case of China reveals practical challenges linked to sortition, prompting reflection on the real-world implications of applying the logic of randomness. Now I examine the theory of sortition democracy from the perspective of sampling practice through a detailed investigation of the logic of randomness. My analytical framework encompasses the entire process of random selection, starting with whether citizens themselves accept the principle of sortition and concluding with the performance of those randomly selected.

5.1 Do Citizens Support the Concept of a Sortition-Based Parliament?

Any proposal for a sortition parliament must first be approved by citizens, but it is likely that their opinions on this issue will vary. Support for sortition can be observed, for instance, in the Australian Citizen Party's advocacy for a sortition-based parliament. However, many other citizens may oppose the idea, perceiving it as a threat to their right to vote. Some will accept random selection of respondents for a survey, but see randomly selecting participants for political decision-making as profoundly troublesome, illegitimate and undemocratic.

In everyday life, those who buy lottery tickets are aware that their chances of winning are very small. They invest a little money with the hope of winning, bringing them enjoyment and excitement. However, when citizens give up their electoral voting rights, they lose their sovereignty; the outcome of a sortition-based parliament is uncertain.

Empirical evidence indicates that many citizens are hesitant to support a sortition-based parliament, primarily due to concerns about losing their right to vote. A study analyzed both citizens' and Members of Parliament's (MPs) attitudes toward random selection as a democratic reform proposal. The findings revealed that citizens generally exhibit low approval rates for replacing traditional electoral systems with sortition-based methods (Jacquet et al. 2022). Additionally, research found that supporters of citizens’ assemblies selected through sortition tend to be politically disengaged or dissatisfied. These individuals often have lower education levels and feel underrepresented by current political structures. This suggests that while some segments of the population are open to sortition, the broader public may be resistant due to concerns about losing direct electoral influence (Pilet et al. 2023). These findings collectively suggest that while sortition is proposed as a means to enhance democratic representation, significant public resistance exists, primarily due to concerns about diminished voting rights and skepticism toward non-electoral selection methods.

Lottocratic representation as a form of post-representation that challenges stable and fixed identities (Tormey 2006), does not resolve the issue of political representation. If a sortition parliament is to make decisions, why should a small group of randomly chosen citizens have that power? Why should I relinquish my rights and allow these randomly selected participants to decide on my behalf? Why are a large number of unselected citizens excluded from the process? There is also a question of accountability: are randomly selected participants truly responsible for the decisions they make? Sortition democracy undermines the very foundation of democratic consent (Parkinson 2003).

5.2 Does Random Selection Truly Ensure Equal Opportunities?

It is often assumed that, in sortition, all citizens have an equal chance of being selected. However, in practice, stratified sampling is deployed whereby the population is divided into strata (e.g. by gender or race), and then random selections are made from each subgroup. Additionally, cluster sampling will be used when it is not possible to sample from the entire population. The sample is divided into clusters that roughly reflect the overall population, and samples are then randomly selected from these clusters. In such a practice, equal opportunity may appear fairer and more democratic. But this results in equality in nominal or statistical terms, rather than in substantive terms. Stratified random selection assigns different percentages to different groups, justifying unequal chances. Various types of random selection can create disparities in the likelihood of being chosen, leading to unequal treatment where not everyone has the same chance of being selected. Of course, while random selection may not guarantee perfect equality, it remains a highly effective approximation of the ideal of political equality.

5.3 What is the Rate and Gap Between the Selected Participants and Unselected Citizens?

There is a significant gap between the participants selected by sortition and those not selected. At the town level, the randomly selected participants constituted only 0.002% of the town population. At the city level, it will be 0.0003%, meaning more than 99.99% percent of the citizens in a community are not selected. At the global level, the disparity increases. In 2021, the Global Citizens’ Assembly on Climate selected around 100 citizens worldwide through a computer program, with China having 18 representatives, while the Pacific Island countries in the world had no representative at all.

The percentage of selected participants is so low that it undermines the concept of democratic representation. This issue became particularly evident in Wenling city, where officials debated whether to implement deliberative polling at the city level following its successful application at the township level in 2005. Several officials questioned whether a few hundred people selected by sortition could truly represent a population of over one million.

Similarly, 150 participants were randomly selected in France to deliberate on global climate issues. This raised the question of whether these participants and their deliberative outcomes resonated with the broader public. Research found that the relationship between the citizens' assembly and the general public was marked by mutual skepticism. While the public expressed support for the assembly’s proposals, they remained doubtful about the government’s commitment to follow through (Giraudet et al. 2022). The extremely low participation rate raises critical concerns about the validity of lottocratic representation. Representation becomes problematic when decisions made by randomly selected citizens impact the lives of those who were not chosen. It calls into question the epistemic and moral privilege associated with making decisions on behalf of millions, as the inclusivity and representativeness of the process become compromised. In the end, in this case, the French President exercised final authority on two occasions: first, by invoking “veto cards,” and second, by having the government revise the citizens’ proposals before submitting them to Parliament (Giraudet et al. 2022).

5.4 What is the Implication of the Low Acceptance Rate?

There is also a noticeable gap between those who accept invitations and those who decline to participate among the randomly selected participants. The acceptance rate for larger citizen assemblies was below 4%. Philippe Van Parijs expressed surprise upon discovering that assembling a group of approximately 700 participants for a G1000 event in Brussels could require as many as 50,000 phone calls (Lever 2024). The Irish Citizens' Assembly, which addressed issues including abortion and climate change, had an acceptance rate of approximately 33% among those invited to participate (Paulis et al. 2024). During the first deliberative polling in China, the acceptance rate exceeded 90%, likely because many citizens perceived the random selection method as both fair and innovative. Additionally, the Zeguo government facilitated participation by providing several buses to transport all the selected participants.

The low acceptance rate carries significant implications for lottocratic democracy. Those who decline invitations may hold values or perspectives different from those who accept, and their absence means their voices and opinions are excluded from the deliberative forum. This could result in a “silenced selection bias,” where critical viewpoints are left unheard. In the case of China, the author observed that randomly selected businesspeople often attended only the initial open session of the citizens' assembly, then quickly departed to attend to the pressing demands of their business operations.

Moreover, the fact that many randomly selected respondents may choose to decline the invitation to join citizens’ assemblies introduces a self-selection bias into the randomization process. Random selection seeks to eliminate selection bias, yet ironically, the low acceptance rate introduces a different form of bias through the back door. Just as an elected chamber can be unrepresentative and fail to reflect society accurately, a randomly selected parliament also suffers from bias due to the self-selection of participants, undermining the confidence and assumptions of sortition theorists, particularly their claim that lotteries yield epistemically superior decisions compared to elections.

We can certainly explore various strategies to increase participation rates, such as conducting personal, face-to-face visits, including door-to-door conversations. Offering a stipend has also proven to significantly enhance participation. The high participation rates observed in China provide valuable lessons in this regard.

5.5 Is an Idealized Version of a Citizen Assembly Reliable?

Theorists of lottocratic democracy advocate for a radical democratization of decision-making by replacing traditional parliaments with assemblies composed of randomly selected ordinary citizens. They presuppose an idealized version of a citizen assembly where randomly selected citizens are free from the influence of party politics. The ideal model envisions lottocratic citizens as being capable of resisting corruption and engaging in genuine deliberation, akin to Jean-Jacques Rousseau’s notion of a citizen devoid of personal passions and interests. However, the notion of an ideal citizen model is flawed when applied to real-world scenarios.

Random selection yields varied outcomes; while it can select citizens who enhance public spirit as in the case of Zeguo (Fishkin et al. 2010), it may also result in the inclusion of militaristic participants in the citizen assembly in Switzerland, or a higher proportion of illiterate citizens (9%) in China. One major concern is that individuals selected at random may lack the qualifications or expertise needed to govern effectively or make informed decisions on complex policy issues, potentially leading to ineffective or even detrimental policies (Landa and Pevnick 2020). To mitigate this, we might consider introducing minimal entry requirements, such as a high school education. However, implementing such entrance barriers is controversial and may undermine the broad representation of society. Alternatively, strategies to reduce the incompetence of ordinary citizens could include improving general political education, ensuring access to necessary information, and providing opportunities to consult experts.

While random selection can create a more diverse and representative sampling of the community, it has limitations in addressing the complexities of modern, divided, and pluralistic societies. If the selected citizens are divided among themselves, it is unlikely to reduce polarization. Even an assembly composed of the world's greatest philosophers cannot eliminate persistent divisions or guarantee the creation of optimal policies. The various proposals for a sortition-based parliament overlook two critical issues. First, randomly selected citizens cannot ensure a meritocratic government capable of addressing the complexities of modern governance. Second, executive power, along with its professional support teams, cannot be effectively chosen through random selection. This explains why Athens continued to use elections to select generals, financial magistrates, and officials with technical or military expertise—ensuring both democratic accountability and the necessary competence for specialized roles. In everyday life, it is unlikely that citizens would trust a randomly selected political body to tackle practical challenges such as controlling inflation, managing transportation systems, housing, and other pressing issues.

Table 3 summarizes the discussion on the logic of randomness. The left column presents various theoretical claims, while the right column outlines the empirical findings or analyses of how random selection operates in practice.

Table 3 Testing the lottocratic claims
Full size table

6 Conclusion

Lottocratic theorists perceive the logic of randomness as inherently positive and beneficial. They argue that lotteries have a "sanitizing" effect, preventing corruption, manipulation, and domination, and overcoming party-based divisions. However, this paper offers a critical re-evaluation of the logic of randomness. A lottery inherently deprives citizens of their electoral voting rights; and stratified cluster sampling, often used in sortition, assigns different percentages to various groups, creating disparities in selection probabilities. Additionally, the significant gap between selected participants and the broader citizenry raises concerns about representativeness. Only (and often under) 0.002% of selected participants are tasked with deciding on behalf of the remaining 99.9% of unselected citizens, a proportion that strains the concept of democratic legitimacy. Furthermore, when many randomly selected individuals decline invitations to join citizens' assemblies, self-selection bias is introduced, undermining the ideal of impartiality and fairness. Random selection also yields highly varied outcomes. While it can elevate participants who foster public spirit, it may also result in the inclusion of militaristic individuals or disproportionately higher representation of illiterate individuals or certain social groups like the poor in participatory budgeting. These complexities explain why, despite the proliferation of lottocratic proposals, very few have been implemented. When democratic theorists propose new models or initiatives, they should go beyond theoretical arguments and critically engage with the complex processes involved in random selection. Proposals for sortition democracy must address practical issues. Tackling these challenges is essential to ensuring the success and viability of sortition democracy.

This paper distinguishes between a sortition-based parliament as a replacement for an elected parliament and the use of sortition as part of a mixed system designed to enhance the existing democratic framework. Enhancement is both desirable and achievable, while complete replacement is neither desirable nor feasible. The notion of replacing voting with sortition is not consistent with historical facts, difficult to implement in reality; and normatively undesirable. Ontologically, lottocratic democracy assumes a new form of governing by randomness, where everything is uncertain. Pure sortition democracy without elections did not exist in the past, and today it cannot completely replace elections. Methodologically, sortition is not a panacea for the challenges faced by liberal democracy. Sortition is merely a method and approach. It must be integrated with both elections and deliberation to be effective. Elections, along with citizens' political rights and public deliberation, form the foundation for the development of sortition democracy.

One promising way forward is to combine sortition, voting, and deliberation to create a mixed or hybrid democratic system. For instance, when addressing contentious or constitutional issues, the process could begin with a randomly selected citizens' assembly that develops proposals or recommendations. These would then be submitted to the elected parliament, which has the mandate to debate and revise them. Finally, the revised outcomes could be put to a public referendum (He 2015: Chap. 11; Levy et al. 2021). Naturally, the design of such a hybrid system—how sortition, deliberation, and electoral mechanisms are combined—should be shaped by local conditions, contextual factors, and the specific nature of the task. These factors determine the mode, balance, and institutional form of the mixed system. (For a discussion of such mix in the Chinese context, see Han 2018; for a critique, see He 2019: 151). The right strategy is to learn how to balance them, in what proportions, and in what sequence—this is the democratic art that political science needs to carefully study.