Is Public Ignorance a Problem? An Epistemic Defense of Really Existing Democracies

Abstract

Does good governance require citizens to be knowledgeable of basic facts and best policy ideas? Some scholars suggest that it does, and propose disenfranchising the most ‘ignorant’ voters. In contrast, we argue, political systems are complex systems inevitably exhibiting incomplete, imperfect and asymmetric information that is dynamically generated in society from actors with diverse life experiences, antagonistic interests and often profoundly dissonant views and values, generating radical uncertainty among political elites over the consequences of their decisions. Radical uncertainty, radical dissonance and power asymmetry are inescapable properties of politics. Good performance significantly depends on how political elites navigate through radical uncertainty to handle radical dissonance. Democracy, by offering citizens equal rights to participate in politics and talk freely, both enables and compels political actors to track social feedback regarding the effects of their decisions on a diverse public, and consider it in ways that mitigate these three problems.

Introduction

Most citizens are far from being well-informed about relevant political facts and best policy options. Many of them even lack basic knowledge of the political process. Is this type of ignorance a problem for the performance of democratic systems? Consolidated democracies tend to provide better outcomes in terms of both personal freedom and overall human flourishing compared to authoritarian regimes (De Mesquita et al. 2005; Estlund 1997; Feng 1997). This record is attributed to the fact that democracies, through elections, civil rights, public contestation, and checks and balances, prevent elites from fully capturing the political system (Bagg 2018; Cf.; Dahl 1971; Green 2016) and push them to be responsive to the views of the many (Arnold and Franklin 2012; Page 1994; Uslaner and Weber 1983). This line of argument points to democracy’s incentive structure regarding elite behaviour, the constraints and terms of competition set by democratic institutions and norms, but does not address questions over the epistemic qualities of democratic systems and the key objection, as old as Plato’s Republic, that a poor quality of citizens in terms of levels of knowledge would undermine the quality of political outputs and policies. This ‘input-output’ epistemic relationship has been a central concern in discussions regarding how to reform democracy and improve its performance and outcomes.

Assuming that widespread public ignorance undermines the quality of the judgements, choices and contributions of citizens poses a compelling challenge for instrumental arguments defending democracy (Estlund 2008; Knight and Johnson 2007; Landemore 2012, 2020). If we see the democratic process primarily as selection based on preferences – the election of political representatives and voting for political platforms and ideas – public ignorance must negatively affect the quality of the preferences that will be cast and counted as ballots, and represented in elected decision-making bodies. Alternatively, if we understand democracy as broader processes of participation, contestation and deliberation, or even ‘a mode of collective teaching and learning by which we discern and define problems of public interest and experiment with solutions to these problems’ (Anderson 2010, 96), public ignorance remains a problem that must be fixed through education and formative participatory processes in order to enhance the quality of these processes and their outcomes. Otherwise, if this avenue is too difficult or unrealistic, drastic reforms have been proposed about diminishing the input of ‘ignorant’ views into the decision-making process, such as ‘epistocracy’ (Brennan 2016), combinations of limited democracy, enlightened authoritarianism and preference for more technocratic governance (Bell 2006), or expert-led authoritarianism (Mittiga 2022).

Our account is fundamentally different in terms of how we conceptualize this question. We challenge the epistocratic critique of democracy on its own core premise: treating knowledge of facts and ‘good’ ideas as the relevant type of knowledge for citizens, and looking at policymaking and democratic decisions as primarily reflections and aggregation of citizens preferences, which should ideally be well-informed inputs. Our argument rejects this model of aggregation and recasts the problem of ignorance. We see political systems as complex systems consisting of myriads of interacting actors with distinct life experiences, different preferences, antagonistic interests and often irreconcilable values. Multitudes of actions and interactions both influence and limit others, trigger reactions and new interactions, and make the properties and developments at system-level ever-changing, multifaceted and, importantly, not adequately known and unpredictable. All actors, including political elites and experts, inevitably have limited information – incomplete, imperfect and asymmetric ¹ – about the dynamic evolution of such a complex system, which concerns the circumstances and potential effects of their decisions. Ignorance is thus an inherent property of complex systems, but it is a greater problem for political elites who take authoritative decisions affecting the whole of society and must consider multidimensional, dynamic and often unknown conditions with variable and unpredictable trajectories, particularly since their decisions interfere with complex interactions in society and could add uncertainty regarding ensuing developments.

In that regard, political elites face radical uncertainty (Cf. Kay and King 2020). ² They cannot predict the very nature, evolution and full scale of the consequences of their decisions. However, they are in positions of political power, and their decisions affect diverse and evolving life experiences of the citizens, and may clash with diverse political views, antagonistic preferences, irreconcilable values and competing ideas (Cf. Tebble 2016), what we call radical dissonance, often manifested as persistent disagreement (Knight and Johnson 2011). In political systems, radical dissonance accentuates radical uncertainty because political decisions and policies affecting large segments of society, if not the whole of it, can generate serious repercussions, public grievances, backlash and reactions, which could escalate to challenges for the legitimacy, stability and prosperity of the system itself.

We thus recast the problem of ignorance linked to democratic performance. Assessing the performance of a type of political system concerns how political elites handle radical ignorance and radical dissonance, and whether they do so in ways that foster, rather than disturb, peace, relative prosperity, human rights and human security. ³ This assessment concerns both why democracies exhibit generally superior performance compared to authoritarian regimes as well whether proposed alternatives such as epistocracy and technocracy will have any added value as reforms of democracy. Epistocracy as well as preference for substantially more technocratic modes of governance (technocratic preference) contain yet-to-be tested reform alternatives, either excluding some citizens as ignorant or reducing the weight of citizens’ inputs on the basis that they are laypeople with epistemically inferior contributions to a political process. However, we argue, these alternatives must rather account for how a reformed decision-making system following their directions can handle three key interrelated characteristics of political systems: radical uncertainty, radical dissonance and power asymmetry.

Our argument is summarized as follows: (a) political systems are complex systems inevitably characterized by incomplete, imperfect and asymmetric knowledge among all participants regarding important circumstances and consequences of their actions and decisions as they interact with a dynamic context of social interactions – it is thus incorrect to expect that a basic standard of knowledge for each citizen will help them overcome the problem of ignorance; (b) in complex systems, dispersed pieces of information are dynamically generated from ongoing expressions and interactions of actors and they simultaneously serve as ongoing feedback harnessed by those interacting with others; (c) the epistemic quality of democracy concerns how political elites as decision-makers track continuous feedback from society regarding how people experience situations and policies; (d) it thus matters if the political system facilitates and encourages citizens’ inclusion as credibly enfranchised citizens because their expressive participation continually generates widespread and updated feedback that political elites will pay attention to; the democratic process both legitimizes and enables free expression and participation by all citizens, allowing the generation of the widest possible range of uncensored social feedback to continually reach decision-makers; (e), at the same time, the democratic system provides incentives for decision-makers and their political contestants to consider feedback from citizens and respond to it. By contrast, restrictions of civil rights and the ensuing disenfranchisement and delegitimisation of genuine public expression, as proposed by the epistocratic proposal or the technocratic preference, will weaken rather than enhance the way decision-making in a democracy handles radical dissonance in conditions of radical uncertainty, and will undermine its prospects for exhibiting relatively good performance.

The Epistocratic Critique Under Scrutiny

Both critics and proponents of full citizens’ inclusion and democratic equality view collective decisions as processes involving citizens’ direct inputs and contributions – whether this is a process of preference aggregation through elections or a formative dialogue in deliberative processes – and reach different conclusions regarding how the problem of public ignorance should be addressed. Deliberative democrats place faith in public deliberation, hoping that the uptake of citizens’ views and reflection among participating citizens could elevate the quality of collective decision-making. Skeptics argue that citizens are ignorant about relevant facts and make bad political choices (Brennan 2016; Caplan 2008; DeCanio 2014; Somin 2016), mostly based on cultural identity and group membership, while falling prey to political opportunism (Achen and Bartels 2016), and that these choices negatively affect policymaking and policy outcomes.

This line of criticism has gone as far as to question the value of an equal and universal franchise. Brennan (2016) argues that, to perform well, citizens must behave like ‘Vulcans’: rational, reflective and informed individuals who consider the facts and the public good carefully when deciding how to vote and participate in politics. However, he claims, most people in politics behave more like ‘hobbits’ or ‘hooligans’ (cf. Vallier 2019). Brennan supports departing from formal democratic equality and establishing a system of ‘epistocracy’ that would privilege the ‘knowledgeable’ citizens and exclude the most ‘ignorant’ from the democratic process anticipating that suffrage restrictions would filter the very ignorant views at least, preventing them from influencing decision-making.

Similar arguments on public ignorance are invoked to privilege preference for technocratic management over public opinion and social protest, on the grounds that experts have superior knowledge compared to the many. This technocratic preference defends changes in the relative balance of decision-making, limiting the contribution of franchise relative to decisions by elites aided by experts (Jones 2020). Bell (2006) proposes a hybrid model, inspired by Confucian political philosophy with a greater and more autonomous role for those with scholarly merit (established through transparent competitive examinations) and less emphasis on active citizenship.

The exclusion of citizens on the basis of a standard of knowledge has been criticized on moral grounds for taking pre-existing inequalities and using them to justify differential treatment among citizens, especially at the expense of the most marginalized and vulnerable who lack the resources and the time required for gaining the sort of baseline ‘knowledge qualifications’ (E. F. Cohen 2015). People dubbed as ignorant will most likely be discredited as sources of information in the eyes of the public, experiencing testimonial injustice, and some of them may understand their own exclusion as justified, and abstain from publicly sharing their stories and views, thereby experiencing and reproducing hermeneutical injustice (Fricker 2007).

Paradoxically, defining who is knowledgeable is judged by the epistocratic critics themselves with reference to standards chosen by them, usually references to knowledge of basic procedural facts and, often, some mainstream tenets from research and theory in economics and other sciences (though some of them are disputed among the experts). The epistocrats pick these views and facts to determine which policy positions will be considered as ‘correct’ and ‘socially beneficial’ and then elevate them into the standard of selection of enfranchised citizens. In that regard, epistocrats exhibit a technocratic preference too but in a peculiar way: they select a technocratic view of ‘correct’ decisions and policies independently of the views of the citizens and impose it both as (1) the benchmark of what the voting process must ideally lead to – citizens should reach the correct policies as the epistocrats define it, and (2) the standard of selecting the enfranchised citizens, rather than giving power directly to the experts themselves. In short, epistocrats exclude the citizens whose policy preferences and qualities they disapprove with. They want to amend democracy by adjusting the body of citizens itself but, by imposing ‘their’ own chosen definition of correct policies, the epistocratic position rejects the essence of democracy: that the very definition of good policy and the public interest is, in the end, a judgement which most likely differs among those who experience reality and policies in diverse circumstances.

Moreover, the idea of epistocracy is inconsistently based on a model of preference aggregation which, in its most clear articulation, is derived from neoclassical economics. Neoclassical economics regards any deviations from conditions of rationality and perfect information in a choice framework as a liability for how the system will perform overall. In the neoclassical conception of democracy, there is a process of aggregating and measuring the preferences of individuals. However, ignorance of politics is rational. Downs’s (1957) ‘economic theory of democracy’ posits that, because a single voter cannot influence electoral outcomes and because gaining information is costly, few citizens will ever bother to gain a good level of knowledge required to develop plausibly reasoned preferences and voting intentions. In addition, even if we could model citizens as having perfect information and a consistent set of preferences – each of them becoming a rational and informed homo politicus – the electoral process will lead to cycles of voting coalitions producing collectively arbitrary outcomes (Arrow 1950; Riker 1982). In addition, as the next section explains in detail, the source of different views stems from conflicting ideas, interests and values. In short, epistocracy relies on a contradicting and elusive standard of rationality, the reductionist and inconclusive premise of preference aggregation and an arbitrary selection of factual claims which, if consistently followed, cast doubt on the value of any process of participation altogether, including the one supposedly amended by epistocracy.

Radical Dissonance

The idea that there is a ‘correct’ answer which can be approximately reached through voting or ‘jury-like’ processes is challenged by a proper recognition of the heterogeneity of life experiences among citizens and the ensuing dissonance over what is right, correct or just (Cf. Anderson 2008a, 130, 133; 2008b, 11; Estlund 2008, 99). In real contexts of human interaction such as politics, disagreements, tensions and conflicts are inevitable (Knight and Johnson 2007). This dissonance concerns the definition ‘of ‘socially beneficial’ outcomes. The real problem of a democratic system is not about aggregating and measuring preferences with a view to a correct outcome, but it is about how democracies handle this dissonance, which always leaves some, if not most, preferences, ideas, values and norms unfulfilled at any given time (Honig 1993, 13):

‘The biggest task of democratic governments is to construct policy options by investigating what problems merit public action and imagining a variety of possible policy responses to those problems. The hard work comes in crafting laws that tend to a wide variety of interests, values, and constraints’ (Anderson 2008a, 134)

Such dissonance intersects with the question of public ignorance and creates a dual challenge. Will misinformed, biased and opportunistic views and actions accentuate clashes and conflict (Cf. Klosko 2000, 2005)? Deliberative democrats sought to address the challenge of public dissonance and fallibility by placing greater emphasis on citizens engagement in dialogue, creating opportunities for engaged citizens to become aware of, consider and critically examine their own views and the views of others (Bhatia 2020; Bohman 2007; Chambers 2021; J. Cohen 1997; Dryzek et al. 2019; Estlund 2008; Habermas 1996; Landemore 2012), potentially leading to some convergence on policy priorities and solutions (Dryzek et al. 2019; Habermas 1996, 2003). They posit that the views of citizens are not fixed inputs to be counted and aggregated, but contributions that can be revised in the context of a dialogue once they are thoroughly reflected and debated (Bohman 2006, 175), potentially generating shared understandings (Dryzek and List 2003; Cf.; Knight and Johnson 1994, 2007). Building on an Aristotelian thesis, Landemore (2012) argues that greater inclusiveness tends to be correlated with greater cognitive diversity, which, in turn, is associated with better problem solving thanks to the formative qualities of dialogue (Cf. J. Cohen 1986, 36).

The value of deliberation should not be underestimated when it comes to how political systems handle radical dissonance. Mill warned against overconfidence in one's own knowledge and a refusal to consider the opinions of others, or an excessive deference to the opinions of the majority (Arellano 2021). Dewey (2017) suggested tackling this problem through the improvement of the methods and conditions of debate, discussion and persuasion.

However, it is uncertain that these deliberative qualities will emerge in larger decision-making units (Cf. Gastil, Black, and Moscovitz 2008, 38; Landemore 2014). On a larger scale, it is unlikely that citizens who hold incompatible views will converge on a shared understanding of the ‘right’ policy path (Cf. Bagg 2018). They may even disagree over the basic norms and procedures that govern the settlement of their own disagreements (Trantidis 2017). Moreover, it appears unrealistic to expect that citizens tend to express themselves and cast their votes and make choices based on considerations of the public interest instead of considering and promoting their own personal interests (J. Cohen 1986, 30). At best, the value of deliberative democrats is found in continual rounds of processes of reflection and reform acknowledging, negotiating and reformulating radical dissonance (Knight and Johnson 2011, 261; Trantidis 2022).

In addition, even in deliberative processes, there will be ‘subsets of citizens that are wiser than the group as a whole’ in the typical sense of information over facts and ideas (Estlund 2008, 40). Skeptics of democratic deliberation may claim that uninformed, biased and aggressive stances will ‘contaminate’ a process of deliberation, sapping the energy of the others, demoralizing and alienating some, and possibly leading to a gridlock or new clashes. A proposed way to improve this situation is to make ordinary citizens – not just elites – cultivate issue expertise and specialization (Elliott 2019). However, many, if not most, citizens may remain silent observers and may even choose not to be active participants at all (Green 2010). Finally, power dynamics in agenda setting, and the risk of manipulation and domination remains a problem affecting deliberative processes too (Sanders 1997).

Additionally, assessing democratic deliberation on instrumental grounds requires considering real and imperfect conditions in democracies such as power asymmetries and persistent disagreement (Bagg 2018; Green 2016; Knight and Johnson 1997; Knight and Schwartzberg 2024; Sabl 2015). A more holistic approach to democratic performance is observed in the debate between Lippmann and Dewey. Lippman (1922) problematized the citizens’ limited access to, and interest in factual information together with the distortion of publicly available information due to cognitive biases, media manipulation, inadequate expertise, and cultural norms. Given the difficulty of fostering an informed public capable of addressing complicated policy issues, Lippmann suggested that a technocratic elite could be better positioned to serve the public interest. Dewet’s (1927) response to Lippmann acknowledged the problem of public ignorance but also highlighted the threat of oligarchy at the expense of the many. Dewey justified democracy as a way of resolving conflicts without violence and coercion, and emphasized the need for open communication, inclusive decision-making, and the cultivation of democratic values next to the importance of civic participation and active engagement.

However, insofar as the current defense of democracy relies on a view of political processes whereby citizens’ inputs influence outputs, the epistemic critique of democracy and the argument for technocratic preference return with a vengeance: does ‘garbage in’ mean garbage out’?

The first line of objection to these proposals is that there can be no agreement in the first place about which kind of policies would signify that democracy performs maximally well based on some expectations of good outcomes and aspirations on good performance (Cf. Sabl 2015, 355). This is partly because dissonance is inherent in political societies, an inevitable consequence of human diversity and interaction, and the assessment of democracy's epistemic value cannot be based on a standard of ‘correct’ decisions defined ‘objectively’ or reflecting the wisdom of the many (Cf. Schwartzberg 2015), or stemming from the good judgement of citizens (Cf. Schwartzberg 2016, 744). Democracy cannot be conceptualized as a process of citizens reaching a ‘collective verdict’ who must be ideally more ‘enlightened’ and ‘informed’ because, at least some of the citizens will be partial, disappointed or angry for a variety of reasons, and poised to challenge these decisions in the next available opportunity. ⁴ This conceptualization leaves key properties of human systems unaccounted for: radical dissonance and radical uncertainty.

Radical Uncertainty: Political Systems as Complex Systems

Radical dissonance extends beyond the challenge of how policymakers navigate disagreement and handle conflict; it is complicated by an epistemic problem: the fact that their decisions interfere with a complex social system. Social systems are complex systems of many interconnected parts whose interaction with each other in intricate and multiple ways continuously generate emergent behaviours and trajectories (Lewis 2010; Root 2013, 2020, 25–29; Wagner 2016, 73). The properties and behaviour of a complex social system continuously evolve from the interactions of its numerous actors who inhabit it, and its macro properties cannot be understood simply through aggregation and extrapolation from the properties of the individuals who constitute a society (Hall 2010; Root 2016). These interactions and the ensuing adaptations and co-evolution of the actors themselves who acquire new properties generate novel relations and developments continuously, often surprising and at times unimagined, which cannot be predicted by past observations (North 2005, 22; Cowen and Schliesser 2023).

This conceptualization is important for how we approach the problem of knowledge in politics. Politics resides in complex social systems characterized by three properties: (a) information is dynamically generated through ongoing interactions of the participants, ⁵ (b) information is dispersed, incomplete and imperfect among the system’s participants, regarding opportunities, consequences, risks and obstacles of their decisions (Simon 1957), (c) information is asymmetric; some are more informed than others in the circumstances of their interaction and can take advantage of it in their relations with them.

For democracies, complexity is translated into the following challenges: a) political decision-makers cannot gain adequate information in order to foresee developments with a good degree of accuracy or, importantly, what the impact of their interventions will be, including potential unknown and undesired consequences, new risks, and further side effects from significant policy interventions (Trantidis and Boettke 2022). Besides currently unknown factors and unexpected exogenous events in the future, they face radical uncertainty because of how the system would respond primarily endogenously to their actions: a major intervention affecting so many people will trigger not just diverse immediate reactions by these people at once, but it will unleash sequences of interdependent decisions by these actors and those they are interacting with, recurrently influencing one another, leading to reverberating behavioural adaptations to continually incoming feedback, which continuously change the behaviour of the system in unforeseen ways (Trantidis and Boettke 2022). Knowledge over how society evolves is inevitably limited and conjectural, and best involves tracking feedback as much as possible regarding these interactions.

Hence, the state of the epistemic debate on democracy has relied on an unrealistic model of knowledge in political systems erroneously assuming that good policymaking requires citizens with some elusive yet adequate knowledge of political facts and policy processes on the grounds that policymaking would reflect a distribution of public preferences. Contrary to this conceptualization, the way a democratic system works cannot be captured through the scheme of ‘knowledgeable’ citizens’ inputs – ‘socially beneficial’ outputs for two key reasons. First, there is radical dissonance regarding the definition of ‘beneficial’. Second, there is radical uncertainty concerning the actual outcomes and developments that stem from dynamic interactions.

In addition, the epistemic debate has aspired to a model of citizens that is also unrealistic and demanding. The problem is not just that citizens tend to choose policies and politicians following heuristics (Cf. Wittman 2004) and that their judgement can be contaminated by personal bias or manipulated (Kuklinski and Hurley 1994) but, more profoundly, that there is no standard of heuristics or basic knowledge (Cf. Lupia and McCubbins 1998) by which citizens can gain an accurate summary or good representation of a multifaceted and evolving reality. In a complex system, the problem is not that, when we vote, we lack some baseline knowledge or good cues to make an informed choice and envisage the outcomes of this choice – for instance, whether the political parties and politicians we voted for will behave as they expected – but importantly it is that we cannot even trust the available and reliable informational parameters of our own thought process as adequate in the very act of us entrusting others, because we simply cannot be certain whether the policies and politics we opted for, even if implemented, would actually lead to what we have expected to get. With limited and asymmetric information, we can make some guesses and then observe how developments will play out. Likewise, we cannot use the evaluation of a policy manifesto or the incumbent’s past performance as credible test or predictor for future performance (cf. J. Cohen 1986, 30 Woon 2012; Ashworth, Bueno de Mesquita, and Friedenberg 2018) because future decisions will cause novel and, probably, unexpected developments.

This type of radical uncertainty that we describe here, is a far more consequential problem for political elites because policies and some political decisions have a sizeable effect on society and interfere with diverse life experiences, different perspectives, conflicting values and antagonistic interests. This is where the significant problem of ignorance lies in politics: there is uncertainty over how a policy would affect all aspects of an evolving social environment. The reactions and ensuing interactions of interconnected agents will continuously generate unknown new situations across time and new puzzles for policymakers to figure out. This uncertainty extends to not even knowing the kinds of political reactions and developments that might happen. Political elites cannot predict the full range, scale and actual type of repercussions from decisions that interfere with a complex environment of interactions. Aided by experts, policymakers can make some guesses, but both policymakers and experts need to update these guesses and estimates with constant feedback over evolving effects and reactions from the social system they have interfered with.

Handling Radical Uncertainty and Radical Dissonance

Rather than reducing complex societal interaction to choices judged against an unspecified standard of individual knowledge – or adequate heuristics – and rather than asking how citizens can meet an elusive standard of rationality and optimality, the question of good performance is about how a decision-making system navigates through radical uncertainty to handle radical dissonance. First, this concerns how a political system comes to acknowledge what a matter of public concern is amidst the citizens (Anderson 2008b, 10). This refers to the extent to which policy decision-makers become aware of, reflect upon and respond to the diverse needs and preferences of various groups in society.

Second, we must expect that some policy consequences in society will be undesired and/or unintended. Feedback mechanisms are necessary for any decisionmaker to reconsider their course in light of new information about incoming and developing consequences. In view of dissonance, it matters how political elites can track and monitor the evolving impact on the diverse body of citizens, including how citizens feel and what they worry about these situations, just as it matters if political elites have an incentive to consider this feedback, negotiate it and mitigate any potentially escalating side effects. Because the effects of a policy cannot be adequately foreseen, political elites involved in policymaking as decision-makers or contestants must be receiving continuous and uncensored messages and signals about evolving reactions and interactions in society, including how different people are experiencing policies and politics, their voices, views and emotions.

Social feedback emerges and is communicated through expressive interactions and exchanges among those who participate in the political process. In democracies, some of these interactions occur in institutional structures such as the exercise of citizen rights, open deliberations, and issue advocacy while others take place through informal communication channels, ranging from the mass media, social media and protests to specific interpersonal relations (Cf. Elliott 2019; Mansbridge 2003; Pitkin 1967; Urbinati 2006).

Social feedback is harnessed by political actors who act as entrepreneurs in the Schumpeterian sense (1943). In a democracy, politicians, parties and civil society organizations seek to strengthen their own competitive position by representing different segments of the population, including minority and disadvantaged groups. They may seek to find their own niche by representing a segment of the body of citizens, sometimes a neglected or frustrated one, as well as address open discontent by civil organizations. They may choose to represent a group perspective – following or helping to craft a descriptive account of group-related concerns and experiences and a normative expression of them (Cf. Lepoutre 2020, 50–52) – discern differences within these groups, and track the evolution of these perspectives across time. They may expose the policy failures, abuses and misinformation of their political opponents. While addressing a wider audience, they will most likely develop proposals that must invoke a notion of public interest that allow them to appeal to many more voters and form political alliances with others. These actions and strategies require exploring opportunities for cooperation and delineating a scope for negotiations, deals, concessions and compromise.

Political entrepreneurship is energized thanks to the competitive conditions present in a democratic system. But it can only be effectuated into concrete actions if political actors can track and assess social feedback to gain a better idea of opportunities as well as tensions and tradeoffs involved in policy decisions and political moves. The quality of this feedback arises not from citizens’s technical expertise and policy prescriptions and guidance, or ideas following a good degree of reflection upon, and knowledge of the system itself, but simply the expression of their own evolving and diverse needs, perspectives and emotions, including discontent and grievances.

We distil the first basis for our epistemic argument in favour of citizens inclusion and against epistocratic exclusions as follows. Political decision-making unfolds in conditions of radical uncertainty and dissonance. In politics, there are multiple and evolving life experiences, preferences, attitudes and views in society that governance must handle with inevitable tradeoffs, negotiations and compromises. Democracy is a system of feedback generation offering the best possible avenue for decision-makers to track and consider multifaceted and dynamic circumstances of a heterogeneous public in conditions of radical uncertainty. The discovery of social feedback is comprehensive and effective when people are recognized as citizens with the legitimacy and freedom to talk and reveal information about how they feel, what they experience, what they fear, what they want and how strongly they want it etc. Greater public inclusion and participation enhances the range and timing of social feedback.

It is important to highlight that democracies do not avoid mismatches between the expectations citizens have from their elected representatives and the actual performance of politicians once in office. Democracy does not exclude the possibility of failure to provide basic public goods, as in the clean water scandal at the city of Flint in Michigan. Still, this case of policy failure was not the result of civic ignorance but largely a result of asymmetric information, embedded political corruption and peculiar divisions in jurisdictional responsibility. In the end, it was democratic politics that led to the scandal’s exposure to public scrutiny, which helped initiate a response by the authorities (Pauli 2019). In the far-from-perfect democracy of the United States, the solution came thanks to the reactions of citizens themselves, first and foremost the residents of Flint once the situation was revealed to them, who gathered in town hall meetings and vocally attracted media attention nationwide. These reactions were consequential, not because the citizens of Flint had a good knowledge of politics or a technical grasp of the problem they faced, but because they were given a say and a vote, and the legitimate status to protest as citizens.

Handling Asymmetric Knowledge

Our reply to epistocracy and the technocratic preference continues as follows. Social feedback is helpful but how effective this feedback process is depends on the degree to which citizens and social groups freely and legitimately express their concerns, agonies, views, expectations and values. By granting citizens equal rights to express openly and legitimately their diverse and changing circumstances, feelings, ideas and views, all citizens are encouraged and welcomed to express their experiences, interests and values.

Bringing our epistemic argument even closer to the realities of contemporary democracies, we recognize that democracies do not abolish imbalances and pathologies stemming from the fact that they retain an elite structure in terms of the distribution of decision-making power (Sabl 2015). As Green stresses (2016) no ordinary citizen, either today or in a more enlightened future, can be expected to feel fully free and equal. When citizens express their views, emotions and concerns, they are addressing the politically powerful. These include leaders in positions of formal political power with personal aspirations as well as leaders of business groups and organizations closely networked with decision-makers. Some socioeconomic actors have more resources to trade for access to political power and gain relevant information. They can lobby decision-makers and bargain for their preferences behind closed doors. They can target politicians with decision-making power and use campaign contributions (such as PACs in the US context) to create quid pro quo transactions with them (Hall 2016). In that context, formal political equality is far from realizing substantive equality.

Power asymmetry can have deleterious effects on societies. Growing elite privilege leading to elite capture, if unchecked and unhinged, will lead to degraded system-level performance on basic benchmarks of good governance: growing dissatisfaction with the system itself, increasing instability inciting violence and repression. The question about plausible democratic reforms must engage with the challenge of how unequal power can be mitigated (Knight and Schwartzberg 2024) and, in that light, our epistemic argument will not be complete without bringing this concern with power asymmetry into the debate.

The rules and workings of democracy partly constrain the power of the elites (Bagg 2018; Riker 1982). Equal civic rights in democracies create a more inclusive and responsive context of political participation and interaction by adding a field of political competition that extends outside the core exchange relations between the powerful few. This incentives structure matters but does it alone help mitigate the problems of radical uncertainty and radical dissonance?

Decision-makers must be both keen on, and capable of tracking public concerns. However, the fact that relevant information in politics is not only imperfect and incomplete but also concentrated in elite circles exacerbates the pattern of unequal access to decision-making processes that favours elite groups of insiders: elected leaders, appointed bureaucrats and experts and powerful private lobbies and wealthy donors. Elite capture can be effectuated through the manipulation of asymmetric information. Asymmetric information can lead to a situation similar to what Akerlof (1970) described as ‘a market for lemons’, when sellers try to sell used cars with a significant defect or malfunction (a.k.a. ‘lemons’) to technically ignorant buyers and, with lack of transparency regarding the quality of goods being traded, adverse selection can occur. Akerlof noticed a problematic aspect of asymmetric information: the manipulation of information by more knowledgeable actors to deceive less knowledgeable ones. For him, asymmetric information is mitigated by competition among multiple sellers as they interact with customers. Sellers can build credibility and reputation by offering warranties and third-party certifications while buyers’ reviews and word-of-mouth comments inform consumers that the products and services of some firms are reliable. Note that, in part, the solution entails both competition (incentives) and information generated by the interactions, participation and free expression of customers.

Likewise, in politics, citizens face the risk of falling prey to ‘political lemons’: they lack the means and the knowledge of facts to assess politicians' capabilities, performance, integrity and motives. One way they can address this asymmetry is by freely expressing their reactions, frustrations and disappointments. Collective action by civil society organizations, labour unions and social movements allow citizens to pool resources and be vocal enough to push politicians to wider social arenas of contention (Della Porta 2013, Medearis 2005). With public criticism and protest from multiple sources, no member of the political elite can fully control the agenda to manipulate citizens’ political views (cf. Achen and Bartels 2016) or the conditions of their own publicity (Green 2014, 21).

Our argument here refines the idea that society can hold governments accountable through what Dewey called ‘social inquiry’ (1927) and with what Rosanvallon (2008) identified as ‘social attentiveness’ ranging from civil society activism to judicial redress. We emphasize the contribution of social feedback from the public participation and citizens’ expression to the conditions of public debate and scrutiny. When citizens offer incoming feedback, decision-makers are more informed in processes where they try to negotiate conflicting values and antagonistic preferences and explore larger tradeoffs. Consequently, while asymmetric information is inherent in politics, the problem is mitigated thanks to the combination of political competition and the information generated by citizens’ participation, reactions and public expressions, including agonistic forms of democratic participation (Mouffe 1992, 14, 2000, 101) such as petitions, rallies, strikes, sit ins and assertively disruptive speech (Green 2016). Democracy’s open competitive conditions offer not just the strongest structure of incentives for political elites to consider policy adjustments, moderations, compensations, proportionate actions, concessions and compromises, which they must negotiate with others, but also the feedback that enable them to proceed with doing all that. Nevertheless, as the next section explains, this process is contingent on who has citizenship status.

Citizens’ Exclusion as an Instrumental Loss

How this imperfect and ambiguous balancing act between elite privilege and responsiveness intersects with the problem of knowledge fundamentally differs from the terms that seem to support the epistocratic argument. Democracies must be valued for the way democratic constraints on elites, competitive procedures and incentives work in synergy with a feedback process reducing radical uncertainty facing elites. Mitigating the problem of asymmetric information depends on the range of citizenship status. It is feedback from enfranchised citizens that enjoys political legitimacy and will be heard and considered by virtue of this status. Likewise, only collective voices and collective action by those who are legitimated with citizenship status can become consequential in processes of public scrutiny.

Epistocratic arguments ignore the history of voter suppression in which white political elites used epistocratic grounds and tools such as literacy and citizenship tests to silence and disenfranchise African Americans. Of course, there were no good results. On the contrary, when this practice eventually ceased, the American South became not just fairer but relatively better governed (Bernini et al., 2023). Likewise, today, the voices of non-citizens, such as immigrants or asylum seekers, more often than not do not count, even if the decisions will affect them more than anyone else, as in the Brexit referendum concerning the status and life plans of EU residents in the United Kingdom. Organizations advocating the concerns of non-citizen immigrants and asylum seekers often face suspicion or apathy by a public and politicians who see these groups as outsiders.

Consequently, the exclusion of citizens on epistemic grounds as ‘ignorant’ will discredit them and can even silence them as sources of social feedback. Such exclusion extends beyond their capacity to vote. With the label of ‘ignorance’, dominant group/majorities can discredit minority views as flawed, non-credible and untrustworthy, and dismiss these expressions as less relevant, even erroneous and deficient compared to more ‘enlightened’ interpretations and understandings by the enfranchised citizens and experts (Catala 2015, 427–28).

This brings us to our final point. Besides a moral objection to exclusion referring to testimonial and hermeneutical injustice (Fricker 2007), our analysis, puts forward an instrumental objection to epistocracy and a technocratic preference. The exclusion of a segment of the citizenry on any grounds will diminish the process of discovering of social feedback with two critical and interrelated consequences. It will actually increase decision-making uncertainty and undermine how the democratic system handles dissonance. Limits to political participation under any pretext will weaken the range of feedback for decision-making, thereby exacerbating conditions of radical uncertainty, and will simultaneously alter the terms of public scrutiny and accountability itself, which depends on participation and feedback.

We stress that the political exclusion of (some) citizens is an informational loss for the system as a whole, both limiting and distorting the process of discovery of social feedback as well as adding, rather than removing, uncertainty over the effect of decisions and actions with negative consequences for the system’s overall performance. For example, disenfranchising and discrediting the voices of what experts and epistocrats would classify as ‘ignorant’, when they express their experience, emotions and dissent, can overlook early signs of mounting discontent that may eventually escalate to conflict. Epistocracy will not only marginalize the most vulnerable groups who face chronic problems of social exclusion and stigmatization, but will weaken the synergy between feedback and accountability which is what makes the democratic system better positioned to overcome uncertainty and radical dissonance in ways that avoid conflict.

Only democracies, to the extent that they allow political contestation through the fullest and freest possible public inclusion, both enable and compel political elites to consider the concerns and needs of the many as they evolve, and expose political decisions to the widest possible scrutiny in light of this incoming feedback. This scrutiny does not require citizens to be fully informed to a given standard. Instead, public scrutiny and public contestation works by citizens asking questions, expressing their emotions, venting their frustration, anger and despair, and joining other in doing so as fully enfranchised citizens whose voice has value.

In our argument, the discovery of social feedback and subsequent responsiveness to social feedback due to public scrutiny are two necessary conditions for good governance. This argument does not suggest that democracies will not benefit from better informed and more engaged citizens. It is an argument against exclusion on epistemic grounds, as it explains good performance with current levels of engagement and knowledge in inescapable conditions of limited and asymmetric information, and stresses where the main problem lies – political elites facing radical uncertainty – and how it is mitigated through political inclusion.

Finally, our argument does not vindicate existing representative democracies for their flaws and imbalances. It rather associates the reality of power asymmetry with the problems of radical dissonance, radical uncertainty and asymmetric information as they could fuel growing discontent and potentially instability, conflict and democratic backsliding. Importantly, we emphasize the value of inclusion in imperfect conditions. Political elites and, unfortunately, other citizens tend to pay attention to those who have been recognized as citizens. If you are not included as a citizen, you usually do not count. We also cast new light on how we approach democratic pathologies. For example, gerrymandering or burdensome voter registration procedures weaken the epistemic value of electoral competition: legislators effectively choose whom they have to listen to so as to get re-elected. These restrictions do not just constitute a problem of fairness but also an information loss about developments and policy consequences in society. Our perspective endorses reforms in the opposite direction to exclusion and elitism: strengthening equal citizenship and free expression on instrumental grounds.

Conclusion

Summing up, our contribution to the epistemic debate on democracy starts with a criticism of the underlying premises of this debate and by offering a different perspective regarding the conception of knowledge itself. The debate has revolved around an unrealistic model of knowledge in political systems, falsely concluding that good policymaking closely depends on citizens with ‘adequate’ knowledge of political facts and ‘best’ policy ideas. In reality, we argue, political systems are complex systems whose members have limited, incomplete and imperfect information next to different views, antagonistic interests and discordant values. Political systems are a particular type of complex systems because of three properties: radical uncertainty, radical dissonance and power asymmetry. Policymakers face radical uncertainty over how their consequential decisions will reconfigure a complex system as they interfere with myriads of different actors interacting with one another. Policies can generate unintended and unpredicted consequences and can trigger types and range of backlash that remain unspecified and unknown until they emerge. We thus reverse the problem of knowledge: political elites are ignorant too and is their ignorance that matters for the performance of the system.

Consequently, the epistemic question is not so much about enhancing the wisdom of the many but about tackling the ignorance of the few: political elites not fully tracking how the system evolves and what developments their interference could trigger. Better performance in democracies cannot be solely attributed to the incentives structure alone that makes elites more responsive to the citizens – fair elections, checks and balances, constitutional protection of rights and rule-of law procedures – but also depends on how political elites navigate through radical uncertainty to handle radical dissonance as they respond to society.

Our epistemic argument is that democracy is an information-generating system that establishes the most effective mechanism for those who govern, or aspire to govern, to discover and track citizens’ evolving and diverse life experiences, values, interests, emotions and preferences. This feedback is necessary for the recognition of policy issues and priorities, for the early detection of discontent, for identifying new challenges, discovering policy implications and engaging in processes of negotiation and deliberation, which could foster periodic policy adjustments and timely readaptations through moderation, agreements and compromises. The more inclusive and continuous public expression is, the better the degree of awareness of social feedback becomes and the wider the degree of elite responsiveness towards society around considerations of general welfare standards and human security.