Scrolling Forward, Sliding Backward: How Social Media Threatens the Functionality of Democracy

Abstract

Political theorists have suggested that democracy is at odds with liberalism. Moreover, with fears about the recent rise in populism, there is growing skepticism about whether liberalism and democracy can continue to survive. In her recent work, Democracy Tamed: French Liberalism and the Politics of Suffrage, political theorist Gianna Englert argues that voters’ political capacity—rather than democratic political rights—kept nineteenth-century French liberalism open to democracy while fostering citizens’ capacity for democracy. The theorists she discusses anticipated the problems we face today, including citizens being manipulated by unscrupulous and unqualified influencers. Thus, the concern over an uninformed public in democracy is not new. In the meantime, students of comparative politics have found that people can rely on elite cues to make reasoned choices “as if” they had sufficient information, even when they are uninformed and inattentive. However, with social media overtaking traditional media as the primary source of information for many people, this democratic safeguard no longer functions as it should. In this article, to tackle the age-old challenge of ensuring that citizens in democracies are well informed enough to make reasoned choices, we first summarize the problems identified by the nineteenth-century French liberal theorists with the capacity of non-elites to make sound political judgments. We then explore how the comparative politics literature has responded to concerns about an uninformed public in democracy, suggesting that the same mechanism would not work if people get information from social media. We examine the impact of social media on the rise of anti-democratic leaders by manipulating public opinion, which has allowed illiberal, populist politicians to come to power.

1. Introduction

The widespread popularity of liberal democracies around the globe and the faith that their citizens and governments have in them would suggest that liberal democracy is an institutional framework worth betting on. The resurgence of democracy in the eighteenth century—being the era in which representative democracies were forged at the crux of emerging doctrines of liberalism and popular sovereignty and thus defining the framework of modern democratic institutions—paints a starkly different picture. Westerners turned toward liberal democracies as a counter-cultural response to monarchies that civilians viewed as abusive. The shift toward democracy, therefore, came from a sense of necessity and a search for an alternative to the existing system, not from a deep-seated conviction in the sustainability of a liberal democracy. In fact, political theorists have suggested that democracy is at odds with liberalism since the notion of representative democracy first emerged in popular discourse. Thus, many of the earlier thinkers discussed democracy within a framework of skepticism about the masses’ ability to govern themselves. Limiting the voting population to only those possessing political capacity was the theoretical response to the expected chaos of mass democracy and a justifiable means of hampering the incompetence of the masses.

Political theorist Englert (2024) elaborates on the historical foundations of the age-old concern with how to ensure that citizens in democracies are well-informed enough to make reasonable decisions when it comes to politics. She discusses the nineteenth-century French liberal theorists’ debates over the notion of political capacity (“capacité”) for non-elites to make sound political judgments. She highlights their variant interpretations of what political capacity entailed in order to differentiate between their visions of how to reconcile democracy with liberalism. Their arguments varied in how to define capacité and how to deploy a meritocratic system of enfranchisement. However, in almost every instance, they reasoned that establishing a democracy with a tiered structure of representation would offset the concern about the masses’ general lack of capacité by tending towards a system which would emphasize the participation of those possessing capacité. Thus, the people’s right to sovereignty would not be infringed, but governance would be placed in the hands of those capable of handling the complex issue of governance.

Nineteenth-century French theorists saw the flaws in democracy firsthand, as their home country stumbled through several flawed and short-lived republics attempting to implement democratic enfranchisement. They attributed these faults to the existence of a character trait that manifested how capable a person was of political participation, asserting that the masses had either very little or none of this trait. Modern discussions qualify this assertion by highlighting that human decision-making is flawed in general. The flaws that nineteenth-century theorists characterized as belonging to those lacking capacité have, in large part, been integrated beyond the realm of political discourse into the study of social psychology. Thus, the faulty reasoning tendencies that used to be considered a “lack of capacité” are due to the human faculty to reason and thus may be generalized to the entire population. The thought that everyone lacks capacité would suggest that establishing any effective system of democracy—or even of governance—would be difficult at best and impossible at worst. However, political scientists have found that the masses can rely on cues from more well-informed elites to reason through rational decisions with limited information “as if” they are sufficiently informed. While people are uninformed and inattentive to political issues, complete information is not needed to make reasoned choices. This finding has alleviated the fear of the nineteenth-century French liberal theorists. For example, Lupia (2016) suggests that we should make a distinction between capacity and knowledge and that voters can be sufficiently competent even if they are not necessarily knowledgeable.

However, in the digital age, the emergence of a new wave of communications systems is undermining the mechanisms of representative democracy (Singer and Brooking 2018). With social media overtaking traditional media as the primary source of information for many—if not most—people, the aforementioned democratic safeguard no longer functions as it should.1 The general lack of curiosity, diversity of thought, and open-mindedness that people tend to exhibit in their day-to-day lives, combined with the lack of knowledge that people possess regarding political issues, has enabled a cycle of “siloing” and polarization (French 2020; Guzman 2022; Li 2023; Shigoka 2023; Tett 2021). In addition, boosting the reach of all people through augmented communications systems has upended the hierarchies that institutions once propped up to separate liberal governance from mass democracy. The new form of political communication, combined with the methods of messaging used by populist leaders who benefit from the new system, has been disrupting the existing political infrastructure that allows representative democracy to function.

Moreover, with the rise of anti-democratic leaders through the manipulation of public opinion, social media has played an important role for illiberal, populist politicians to come to power. Thus, we need to examine how social media has formed this emerging trend that threatens the functionality of democracy. Many countries have seen the rise in strongman leaders pushing political narratives against democratic institutions and norms as a response to the establishment’s ties to the “corrupt elite” (e.g., Frantz et al. 2024; Guriev and Treisman 2022; Levitsky and Ziblatt 2018). Despite the ideological parity that runs as a common thread through the majority of modern populist movements, it is important to keep in mind that populism is not a partisan issue. Rather, Müller (2016) notes that populism consists of political movements united by principles of social psychology. Populism employs the “us vs. them” mentality to promote anti-pluralism and justify the degradation of democratic institutions and norms. In these circumstances, social media has weaponized the populist mentality in the pursuit of economic benefit and threatens the functionality of democracy.

Combining the insights of cognitive science, economics, political science, and psychology, Lupia and McCubbins (1998, p. 5) once argued that the public “can use a wide range of simple cues as substitutes for complex information.” As a result, the public can make decisions “as if” they have sufficient information to make reasoned choices (Bartels 1996). Moreover, Lupia and McCubbins (1998, p. 7) write that while people can acquire knowledge by drawing on personal experience or by learning from others, personal experience is not sufficient but learning from others is necessary in many political settings. Thus, in this article, we argue that whether people can make reasoned choices depends on simple cues and social networks, both of which have been disturbed by people’s reliance on algorithmic social media platforms in informing and expressing themselves. We also suggest that the mechanism that enables uninformed voters to make reasoned choices would work if they shared the same set of information sources, but this assumption is no longer plausible in the age of social media. Even when people have access to the same set of information sources from traditional media, they do not get the same information because each person relies on their own cues. However, they at least believe that their information comes from the same set of information sources shared by everybody else. This belief has disappeared, as more people rely on social media, and the gatekeeping function of traditional media is substituted for a platform of complete freedom of publication. As a result, social media has made it more likely for people to be exposed to information that is deceptive, outright false, or published in bad faith, and less likely to receive credible information countering their beliefs. Therefore, to explore the influence of social media on current democratic backsliding, we recall the concerns that the nineteenth-century French liberal theorists had with the functionality of representative democracy.

As the old adage states, “those who cannot remember the past are condemned to repeat it.”2 Without naming any event in particular, it has become apparent that Americans are prone to repeating history, even if doing so presents a credible threat to the sanctity of American democracy and the very structure upon which the nation has founded. However, as our discussion in this article implies, as with other democracies, the functionality of American democracy has always been fragile, and we have been aware of this problem from its very conception. Most importantly, in order to address this recurring problem, it is helpful to integrate the debates of intellectual history into our discussion, so that we can highlight how democratic societies can ensure that citizens make reasoned political decisions, explore the deleterious impact of social media on public discourse and voters’ ability to make reasoned choices in democratic elections, and consider how to redress these concerns following the values that the founders of democratic republics championed.

2. Liberals’ Concerns with Democracy in the Nineteenth Century France

Englert (2024) tackles the complexities of theoretical discussions pertaining to the appropriate extent of democracy that was to be permitted within the blossoming liberalism that sprang up in post-revolutionary France. Early French political history exemplifies a snapshot of much of the initial discourse, which revolved around the concept of liberalism and how it was antagonistic toward democracy. The two concepts were paired and disputed because they were both the subject of a revolutionary anti-monarchical movement that came about in a reactionary way. French liberal theorists in the nineteenth century varied in their openness toward democracy and, for the most part, were appalled at the idea of mass democracy. Given France’s rocky history when mass democracy arose, they had an aversion to democratic systems. A common thread among the majority of liberal theorists at the time was their disbelief in the ability of the people to merit political participation. They described the competence needed to participate in governing as “capacité.” Limiting the voting population to only those possessing capacité was the theoretical response to the anticipated chaos of mass democracy and a justifiable means of hampering the incompetence of the masses. Englert (2024) discusses five principal figures integral in structuring the early bounds of French democracy—Benjamin Constant, Francois Guizot, Alexis de Tocqueville, Edouard Laboulaye, and Ernest Duvergier de Hauranne. Examining these theorists chronologically unveils the tension between the desire for democracy and the knowledge of human nature, which hence inspires the modern liberal institutions that protect democracies from the tendency toward factionalism and populism.

Constant’s main gripe with democracy was his perspective that it did not prevent strongmen; he argued that majority rule had been used by tyrants to justify their tyranny. He wished to decouple genuine democracy—where citizens’ political participation determined political outcomes—from the staged elections and guise of democracy that despots used to manipulate their subjects. However, that is not to say that he was opposed to democracy. In his view, political debate was the breeding ground for the development of intellectualism, and without variety, politics would grow stagnant. He argued that in lacking true representation and plurality, existing democratic systems had failed to encapsulate the “public spirit” (Englert 2024, p. 19). Thus, he advocated for a system that would preserve the benefits of democracy—the positive externality of raising political awareness through political participation—while simultaneously controlling for the vulnerability of the masses to be manipulated by charismatic leaders inclined toward despotism. He argued that people without capacité might not be able to identify qualified leaders but would at least be able to distinguish between those with and without sufficient rationality. Thus, a tiered system of government would allow for popular inclusion in government, though in a buffered form, giving preferences to those with capacité.

Guizot was most skeptical of democracy among the French theorists discussed by Englert (2024) and attempted to redefine democracy altogether to better explain his qualms. He thought that democracy was a state of being that could describe societies as well as governmental systems. He asserted that post-revolutionary France was democratic and that it should not be. In his view, democracy could never survive without descending into violent factionalism, because democracy let loose the reins of not just political authority but education and culture altogether, leaving room for morality to be corrupted and for intelligence to be lost. He advocated for prioritizing order over any inherent enfranchisement, stating that those lacking capacité threatened the integrity of society by virtue of their tendency to demonize one another and inability to cooperate. It was thus necessary to exclude them from enfranchisement altogether. In his view, a rational electorate, built upon hierarchy and inequality, was necessary for governance. For lack of a better measure, Guizot endorsed a socioeconomic standard for enfranchisement, which he justified on the grounds that education and property ownership had become accessible enough to filter out the rational and enlightened members of society via monetary metrics.

Tocqueville agreed with his peers’ concerns over the tendency of people to favor strongmen and factionalism, but he thought that societies that elected incompetent politicians did it on account of preference rather than negligence. However, this did not mean that he endorsed direct elections. Rather, he argued that the best system to allow democracy within a society for the sake of its betterment—all the while mitigating its harmful tendencies—was a representative democracy. He encouraged public participation in elections and promoted politics as a means of facilitating the public’s intellectual growth in society—mirroring Constant’s idea that democracy fostered the “public spirit.” He promoted indirect elections and argued for their functionality even in a system with the lower classes included, believing that their inclusion would allow for needed diversity of thought within political discourse. In practice, he helped to draft the constitution of the Second Republic, which granted universal suffrage, and his political legacy was immortalized in his literature.

Laboulaye built upon Tocqueville’s ideas with an added emphasis on American political culture. His focus was on the protection of what he believed to be universal rights, and he ranked suffrage highest, giving rise to a faith in democracy. Acknowledging that human nature tended toward self-serving biases that bred patronage and anti-pluralism, he argued that liberalism was the natural means to resist the dangers of democracy. To guard against allowing those in power to have control over popular beliefs and values, he asserted that the government’s role should be restricted in public education and religious practices. His faith in social freedoms as a means of checking ambitions placed him in a similar vein to James Madison, who also agreed with Tocqueville’s pretense that enfranchisement enriches the soul and motivates political participation conducive to intellectualism, so he encouraged it as a means of improving the national culture.

In the meantime, Duvergier de Hauranne turned to the British school of thought—instead of American democracy—for inspiration. Drawing on John Stuart Mill, he was a consequentialist and a pragmatist, who hoped to move from the abstract discourse of theoretical liberalism toward a more practical liberalism rooted in policy. He did not have faith in the capacity or rationality of the whole society but argued for a representative democracy that mirrored society for the sake of maximizing the utility derived from political participation. He subscribed to democratic principles of universal suffrage, not because he thought they were sound, but because the alternative had proven ineffectual. He argued that electoral rights should belong in the hands of the informed, yet his argument was different from the capacité arguments of other French theorists. He preferred the term “enlightenment” to describe the informed people, and this was because he thought that rationality should come from education and that it was the duty of the enlightened classes to educate their fellow citizens. He was also drawn to the idea of parties, and although French society was averse to these types of organized political structures due to their contentious history with hierarchical systems, he nonetheless championed this institution. He vouched for the ability of parties to preserve order within democracy and organize the will of the people such that it may be heard without the need for civil disruption.

Throughout the post-revolutionary era in France, there was a great sense of concern regarding whether democracy could function. And yet, monarchical rule led to frustrations that necessitated a system that would allow people to feel heard. Thus, we can see the period of immense political turmoil in France going through several different ideations of a Republic, in light of the inherent distrust that educated people had toward democracy, although they supported a liberal government that secured power for the people. While the institutional changes that modernized and tempered American democracy were far less drastic than those in France, we may analyze the representative democracy constructed in America through the lens of the French liberal theorists, given the similarity between their concerns and those of the American Founders. The French theorists’ concerns were not particular to France, and we may consider them more universal approaches to examining the dangers that threaten democracies. Therefore, discussions of the nineteenth-century French theorists help to decontextualize current discussions surrounding existing democratic institutions by framing the goals that shaped them.

3. The Twentieth-Century Response to the Concerns

The concerns articulated by nineteenth-century French liberal theorists about the ill-informed public’s ability to make reasoned choices have shaped the debates regarding modern democracies. Institutionally, capacité has been considered in two primary senses: indirect democracy and limitations on suffrage. Thus, representative democracy based on universal suffrage—instead of the referendum—has become commonplace. The electoral college institution most directly draws from Duvergier de Hauranne. The system allows for a diverse representation of the political and demographic plurality of the populace that makes their voices heard, while also introducing a middleman that places a quality check into presidential elections—theoretically maximizing civilian utility. The appointment of executive and judicial authorities—rather than their election—has served to heighten the distance between the people and policy as Constant wanted. The representatives that citizens trust enough to vote into power are made responsible for allocating further power, again instituting a real, tangible baseline that citizens would have to meet. Certain groups, especially minors, that are perceived to be undeserving of even an indirect voice in politics have also been shut out from voting, as endorsed by Guizot. This is evident in the fact that suffrage regulation has been evolving as social prejudices are discovered to be unfounded, including the beliefs that women and people of African American descent were less deserving of a vote than white males.

Political scientists have argued that the greater issue of people lacking information—reducing their capacité—has been resolved through “elite cues,” which are simplified messages sent by the more well-informed messengers whom individuals trust. On the issue of how people learn from others, psychologists have found that who the sender of a message is affects, if not determines, how recipients judge the message (Petty et al. 1997). On the issue of how people could make reasoned choices with limited information, political scientists have found that complete information is not needed to make reasoned choices (Lupia and McCubbins 1998). On one hand, seven decades ago, Lippmann (1955) argued that public opinion tended to be inattentive, uninformed, unstructured, volatile, and moody. On the other hand, more than three decades after Lippmann’s assertion had been published, Page and Shapiro (1992) found that public opinion was neither moody nor unstructured, although they granted that the public was uninformed and inattentive.

In his recent work, Lupia (2016) suggests that given the sheer mass of information that exists about every aspect of what may be considered political, any person may only acquire a shallow understanding of a wide breadth of topics, a deep understanding of a very narrow subset of topics, or neither. In any case, all of these people are still uninformed about politics as a whole. Yet, Lupia argues that political educators who point out this lack of information have, for the most part, overemphasized the importance of knowledge over competence. Educators have underscored each person’s ability to use their limited knowledge in combination with their available cues to make competent decisions that they would have made as if they had been privy to a wider breadth of information regarding the subject at hand. Of these several factors, Lupia (2016, chap. 8) focuses on the importance of elite cues, which come from trusted sources, and argues that whether the message’s recipient is convinced or not depends not only on the content of the message but on the credibility of the messenger. In other words, people rely on elite cues to form their opinions because they cannot analyze policy or form their opinions independently of elite perspectives, and they choose elite cues among competing messages based on the credibility of the senders of the messages if multiple competing messages are available (Zaller 1992). Lupia (2016, pp. 48–52) reasons that cues allow people to approach political decision-making efficiently and engage through reasoned assumptions without having complete information. In other words, cues may satisfy the need for political competence, as he argues that even if knowledge is incomplete and insufficient, people can vote as if full information is given as long as they have access to cues. A key example of using cues is when a voter votes along party lines.

From a theoretical perspective, the concerns about democracy raised by the nineteenth-century French liberal theorists have been addressed in modern democratic institutions in the twentieth century. The hierarchies proposed by Tocqueville and Duvergier de Hauranne have been institutionalized to balance the representation of the people with limitations on their governing authority. Systems of checks and balances have been created in order to allow appointed members to oversee democratically elected members of the government, thus factoring in the people’s will as well as heightening the input from more politically attuned voices. The enfranchised population has been limited to exclude minors by virtue of their limited experience and education, amounting to a lack of sufficient capacity. Systems of interest groups and parties have established hubs (e.g., non-governmental organizations, nonprofit organizations, and political action committees) for the well-informed to signal their preferences to voters who lack the resources to inform themselves. This enables people to gather sufficient information to vote in line with their cues. In principle, these mechanisms would all boost the competence of the government to execute the wishes of the people in good faith, while also securing a distance between governments and the populace that guards against factionalism and populism.

4. The Twenty-First Century Challenge with Social Media

To secure the functionality of liberal democracy, the infrastructure of representative democracy has been planned to control for limitations in voters’ capacité. However, the institutions that should clamp down on the dangers of mass democracy have long outlasted the theoretical debate over representative democracy. Given that current society is dissimilar from the society from which its institutions originated, it is reasonable to doubt that said institutions have adapted to the vast changes to communication and social interaction that have occurred since they were formed. Revisiting the original intentions for which institutions were created provides an understanding of the overall framework that political theorists had in mind when they argued for what became modern democratic structures, including what faults they believed it had and what failures they might amount to. Furthermore, we may turn toward these faults and reflect upon whether or not they are present in modern times in order to gauge the security of democracy through the eyes of its creators. Looking back at the concerns raised by the nineteenth-century French liberal theorists, we can see that recent changes in the realm of communication have permeated the mechanisms that are supposed to make representative democracy work. In this section, we suggest that the spread of social media has made elite cues malfunction, led the personality bias of voters to political polarization, and undermined the institutional infrastructure that made liberal democracy work for more than a century.

As shown in the previous section, Lupia’s argument represents a modern take on the discussions of capacité, the lack thereof, and how people’s relationship with political information shapes the nature of democracy. By differentiating between knowledge and competence, Lupia (2016) would challenge the fundamental conception of capacité that the nineteenth-century French theorists used, wherein knowledge and competence were considered interchangeable. He suggests that people are not fully informed, but they are rational enough to make reasoned choices with the limited information given by elite cues. However, achieving this end through the use of cues requires that either all cues are objective, or people can see the validity of the cues that they are reading, which comes from the credibility of the person delivering them. While the former condition is never possible because all cues are by definition subjective, the latter can hold if people share the same set of information sources through traditional media.3 However, this condition does not hold anymore in the current political landscape, given the proliferation of social media. Lupia’s argument assumes that cues originate from the sources that contain inherent legitimacy embedded in some sort of gatekeeping function of information. Moreover, for cues to work, they should come from sources that not only you (the cues’ recipient) find to be credible but also those who disagree with your view find to be credible. For example, a person who is an advocate for the Second Amendment may turn to the National Rifle Association (NRA) to inform them about gun bills for good reason—the NRA has the time, personnel, and resources to look into the details of issues concerning guns and reach conclusions based on this information. The traditional archetype of political elites relies on the idea that these individuals or entities have greater access to political information due to these heightened resources, and thus, American voters turned toward them to inform their political engagement.

Thus, with the gatekeeping function that traditional media has, as Zaller (1992, chap. 6) argues, democratic leaders can provide guidance in the realm of public opinion—as opposed to the popular assertion that they are responsive to public opinion. If a leader wants to build public support for a certain policy, they do not have to persuade individual voters but rather have to build support among policy-making elites who may become messengers of elite cues. If elites are united in support of a policy, what Zaller calls the mainstream effect occurs: the public receives a unified message and evaluates news and political arguments based on the united elites’ framework, leading to a high level of public support for the policy. If a leader fails to build support among elites, they will be divided, and what Zaller calls the polarization effect will occur: the public receives contradictory messages and evaluates news and political arguments in the context of the division among elites, leading to a low level of public support for the policy—particularly among those most aware of the policy debate. In other words, as Figure 1 shows, traditional media with the gatekeeping function plays a role like a traffic signal, which sorts out the available information and makes its flow efficient, so that people can make reasoned choices with limited information as if they have processed all the available information.4 By contrast, as Figure 2 shows, social media does not serve this function but enables anybody to change the traffic signal—which then does not function as a traffic signal anymore—so that people cannot make reasoned choices because no individual person has the cognitive ability to view, let alone process, all of the available information.

Unlike traditional media, social media uses a reflexive quality checking mechanism. Traditional media has to go through a lengthy process of editing and clearing, moving through several hands before the message is ever published for the public to engage with. While one cannot assert that bias, fake news, and hate speech never make their way through this process, there is at least some quality standard to which the media is held before the message is placed in the public’s hands. By contrast, social media content is not subject to such quality checking process before being published. There are safeguards built into algorithms to screen for fake news, hate speech, and other harmful rhetoric, but each of these checks are occurring at best in real time if not too late. Content is evaluated for its merit of publication at the same time as it is seen and shared by the general public. Thus, although there is some content quality standard on social media platforms, by the time content is recognized as harmful, it has already reached an audience, which may be quite a large one, depending on the popularity of the messenger. There is also an exponential nature to online popularity because the algorithm promotes popular accounts and content. Yet popularity is determined by user preference, which is influenced by the personability of the creator, the presentation of the ideology, and the randomness of the algorithm. This is consistent with Tocqueville’s argument that rationality is irrelevant in determining rulership because people support politicians based on their personalities. Therefore, the traditional progression of information flow is inverted. While elites form policy options and people base their preferences on the limited choice sets from their elite cues, social media enables people to create their own “elites” in their echo chamber by supporting content creators who align with their existing preferences or even publishing their own content and assuming the role of an information source.

The nineteenth-century French liberal theorists discussed these concerns at length, with Tocqueville being wary of personality biases and Guizot discrediting the idea that democracy would educate citizens, arguing that it would only encourage those lacking capacité to spread their uninformed ideas and unreasoned choices. We should also give them credit for their recognition that these problems with individual voters would be amplified in their institution through their constant discussion of factionalism and mob mentality. The vast majority of their work was dedicated to forming an approach to democracy that would inhibit the formation of a mob, with systems that included delegates who possessed capacité and would thus be more cooperative (Constant), as well as centralized education and public discussion that would force people to engage with differing opinions (Tocqueville and Laboulaye). In essence, these theorists recognized that certain cognitive biases were inherent in human behavior and amplified en masse. Thus, their solution was to create a system that would favor those least inhibited by their biases, and—where this inhibition was insufficient—defy people’s tendencies toward in-grouping and polarization by forced cooperation and exposure to a plurality of ideas.

What these theorists did not anticipate was that online digital communications through social media would unravel the infrastructure of representative democracy, which was put into place in order to correct for the human nature of giving a priority to processing the information the individual likes to hear, ignoring different opinions from their own, and forming their opinions regardless of what more well-informed people present. Social media algorithms were created to gather data regarding people’s preferences and reinforce them while limiting people’s engagement with different opinions from their own. Moreover, contrary to traditional media, people are able to be content creators regardless of how well-informed they are, and there are no limitations on who may become the source from which information originates, allowing anyone to be the “elite” from whom the masses gather information and opinions upon which they base their political assertions. There are very limited restrictions on who may use each platform, and they are not strictly enforced. The managers of social media have different incentives from the founders of democratic governing systems. They are motivated by the desire to maximize user engagement. Thus, they have no incentive to restrict who can use the platform, to show people the different views with which they would disagree, or to prevent people from sharing inflammatory, misleading, or untrue statements, because any of these actions would reduce user engagement and inhibit them from maximizing their economic benefit.

5. What Should Be Done?

So, what should be done to keep representative democracy functioning in the age of social media? In this section, we argue that individual people need to rebuild their social networks and regain their communication ability in order to adapt to the current circumstances, given that people use social media to receive political information. First, people’s capacité is enhanced when people can disperse and share their information and ideas through their social networks. Second, in the age of social media, where elite cues do not function, each individual’s communication ability determines their capacité.

5.1. Democracy and Social Networks with Social Media

Returning to Laboulaye’s concerns regarding centralized education, social media centralizes the process of informing the public by controlling information and idea flow within the closed social network of an echo chamber. Laboulaye feared that this type of centralization of knowledge would threaten plurality and tend toward a mob mentality. Pentland (2014, p. 4) discusses the power of social networks and provides a useful framework for exploring the mechanisms of digital communications, “social physics,” which he defines as “a quantitative social science that describes reliable, mathematical connections between information and idea flow on the one hand and people’s behavior on the other.” One critical aspect of social physics is the ability of the masses to disperse their information and ideas within networks. Pentland (2014, chap. 5) finds that the organization where individuals are less talented but more connected to each other tends to perform better than the one where individuals are more talented but less connected to each other. Echoing this finding, Fowler and Christakis (2009, p. 72) show that ties of the network that include “a specific set of connections between people in the group … are often more important than the individual people themselves … [as] they allow [the] group to do things that a disconnected collection of individuals cannot.”

These findings articulate how social media may deprive an uninformed public of the capacity to make reasoned choices. As Constant, Tocqueville, and Laboulaye all warned, democracy would motivate people to voice and spread their ideas, a belief encapsulated in Constant’s discussion of the “public spirit.” This means that people would be eager to share political messages, utilize social networks to allow these messages to traverse the nation, and contribute to the formation of political “knowledge” originating from any online post. Thus, uninformed voters would not only form their opinions—upon which they vote with the confirmation biases and personality biases about which Guizot and Tocqueville were concerned—but also spread the opinions throughout their social networks. This fault is amplified by the astounding quantity of bot accounts, misinformation, and deceptive content that is circulated within political circles. Thus, social media, in spite of its onus to encourage communication and networking in good faith, rather stifles diversified communication and exposure to dissent that could improve individual rationality.

Far and away, the most pressing concern that the nineteenth-century French liberal theorists shared was a fear of factionalism. Particularly, they were concerned that the “winner take all” system in representative democracy would change what people think and how they process information, and it would generate division, polarization, and radicalism, mirroring the concern that social media does not just allow people to spread their beliefs but also reshapes those beliefs altogether. Both Lupia (2016) and Pentland (2014) discuss the echo chamber effect—where people are exposed to the same set of information and ideas that amplifies their preexisting beliefs—and the tendency of people to group themselves into like-minded circles, which in turn reinforces their biases and radicalizes their originally moderate stances. For example, Pentland (2014, chap. 1) shows that a group’s performance will improve if the flow of ideas increases, but such improvement will occur if and only if there is diversity in the flowing ideas. In other words, the group’s performance will deteriorate if the flowing ideas lack diversity and the flow of the same ideas increases (i.e., if the echo chamber effect occurs). As discussed in the previous section, people are better informed when the available information and ideas are sorted out by the gatekeeping function of traditional media. In the meantime, people are less informed when the flow of information and ideas is more limited and less diverse in the echo chamber of social media.

Singer and Brooking (2018) explore the negative impact of social media on the functionality of representative democracy, articulating several examples pointing to its genuine threat to people’s capacity to cooperate with each other and bridge partisan lines. They look deeper into how social media has become a key tool used to spread propaganda since it is so effective in promoting a single narrative and silencing opposition. Providing organizations and individuals with this capacity has revolutionized the way in which competing groups recruit new members and harm one another. For example, they detail how Hamas and the Israel Defense Forces have both found great utility in social media’s capacity to distort narratives and frame issues such that each respective group appears to be the victim of the other, regardless of the nature of the actual event that occurred (Singer and Brooking 2018, chap. 1). In effect, social media allows individual people and groups to create sufficient content that is then streamlined toward their target audience without interruption from competing accounts of the same event or issue, effectively instilling a synthetic censorship, clamping down on idea flow, and inhibiting the maximization of utility derived from the other people in social networks—as Pentland (2014) discusses. Now that social media allows anybody to create any content, people do not have to take accountability for the message they spread. Because anybody can be the “elite” in the process of elite cue formation, people do not have to listen to what more informed people discuss in traditional media. As a result, in the age of social media, people tend not to make their message convincing to those who do not share their views or biases, nor to ensure that their information sources are credible.

Additionally, these cues are spread to large audiences of people and receive consistent engagement. Lupia (2016) suggests that one of the factors that allows people to trust cues is their popularity. Because whether or not you are convinced by the message depends more on your trust in the messenger than on the content of the message, people tend to support opinions that are supported by more people. If people receive elite cues from traditional media, this tendency will also help what Zaller (1992) introduces as the mainstream effect work, where a leader needs to persuade policy elites—rather than the whole public—to build popular support for a certain policy. Consistent with Zaller’s (1992) argument, Lupia (2016, p. 51) finds that elite cues lead people to tend toward more moderate and agreeable stances, which appeal to the wider audience. If a leader is successful in uniting elites in support of a policy and the mainstream effect takes place, the agreed stance will be in a moderate position because it requires wider support for consensus building. If a leader fails to build unified support and the polarization effect takes place, each side’s position will still be moderate—even if they are divided—because they are required to compete with each other to gain wider support. In short, as long as people rely on traditional media, their political positions will be more moderate and agreeable.

By contrast, if people rely on social media to receive political information, each side of their views will be radicalized. Singer and Brooking (2018) demonstrate that social media breaks down the signaling effect of popularity through the algorithmic promotion of more radical content (as opposed to more agreeable content) because algorithmic promotion is based on the type of content that people are more likely to engage with, as opposed to agree with. Moreover, bot accounts themselves may synthetically create and support certain viewpoints. If a good number of accounts are run by machine—not by people—they will be programmed to use the same hashtag and/or spread the same message all at once, and “that action can fool even the most advanced social media algorithm, which will mark it as a trend” (Singer and Brooking 2018, p. 141). For example, in the 2016 U.S. presidential election, researchers found that as many as 400,000 accounts on Twitter (officially known as X since 2023) alone were identified as “bot accounts” with the sole purpose of creating and pushing content in favor of a then-presidential candidate Donald Trump (Singer and Brooking 2018, p. 143). This example suggests that it is difficult to discern the true popularity of electoral candidates through the legitimate news headlines, degrading the cues upon which voters base their voting decisions. Thus, even though voters intend to make sound decisions given the limited information they have, they may vote in ways that do not align with their true preferences because the cues they receive have been distorted and falsified. And this problem will be amplified, particularly among those most aware of the debate, if they take political information only from social media.

While many of Guizot’s arguments have not been reflected in the institutional formation of existing representative democracies, his exclusionary perspective on voting is reflected in the age-based restrictions on the rights of political participation. Guizot argued that the voices of those lacking capacité should be shut out of politics, which corresponded to the institution prohibiting minors from voting. The logic is that because they are not mature enough, their political engagement should be stifled. However, social media does not abide by the same restrictions. Any person with a social media account might promote political content and interact with anybody else. On most platforms, users might be as young as 13, with the implausible assumption that nobody should lie about their age when creating accounts. In fact, social media tends to be skewed in favor of this youth bracket. Where kids aged 13–18 (and oftentimes, even younger) do not have a voice at the ballot box, they could turn toward social media to publish their opinions, support policy, and even spread misinformation. Worse yet, there is no way of identifying or confirming the age of content creators—content published by a 12-year-old could not be differentiable from that published by a 40-year-old. Thus, if many people engage with political content through social media, minors might have a significant impact on electoral results and other political outcomes.

So, what should be done? While social media presents a threat to representative democracy, we could address these issues. The integration of the internet into individual life at almost every level makes it infeasible to argue that social media should be eliminated or separated from politics. Moreover, regulations that would most directly and adequately control for the harmful processes by which communication occurs through social media platforms inherently conflict with the freedoms guaranteed by the Constitution. There is also the concern of practicality. While it is easier to make a consensus on whether social media sites should flag fake news, it is more difficult to agree on how to determine what is true and what is not. However, this does not mean that the issue is not manageable. We might obtain desirable results through subtle changes in the relationship between the consumption and production of social media. These suggestions are by no means intended to be sufficient. We only intend them to represent the first of many necessary steps to change the way that people engage with social media in political contexts.

Beginning at the consumption side, increasing education regarding media literacy and the prominence of fake news would enhance the quality of cues derived from social media. If people were aware that the production of fake news is rampant because anybody can write whatever they want to write on social media, they would be more likely to take the content they see with an air of skepticism. Encouraging political discussions outside of social media would also help to combat the echo chamber effect by introducing people to the ideas that their algorithms have never shown them. Moreover, we might execute these measures in classroom settings, where people could learn how to get information, develop their ideas, and discuss them with other people who have different views, which would raise their level of capacité by enhancing their ability to build diversified social networks. In the meantime, from the production side, social media platforms could be held accountable to a greater extent for enforcing existing guidelines. Restricting violent, hateful, and obscene messages would help to limit the digital presence of extreme and radical voices, which would facilitate greater agreeability in online political discourse. The attempt to limit online hate speech has already started by companies like Meta, but the current situation shows that it is not sufficient. Limiting the number of bots would also help to align online political discourse with people’s true preferences. Meta platforms have attempted to discern bot accounts from real accounts, thus warning people that they are not interacting with a human. However, this solution is not sufficient, either. Now that the AI chatbots are introduced as “users,” it is harder to distinguish between real and false content. Furthermore, these bot accounts may still spread false information because they have retained the capability to “like,” comment, and share posts, which will perpetuate the distortionary effect by misrepresenting the popularity and validity of information.

In short, social networks have the potential to strengthen people’s capacité as long as the information and ideas flowing in the network contain diversity. However, social media creates the echo chamber effect, which radicalizes the information and ideas flowing in the network. Thus, social networks based on social media weaken people’s capacité because such a network lacks diversity of information and ideas. We argue that we (people) should build our own social networks that do not rely on social media. For many people, their “real” social networks may be segregated and lack diversity. However, their social media networks will be even more segregated and less diverse, because they can “purify” their members by blocking those who have different backgrounds and ideas from entering their social media networks. In the meantime, in the real social networks, they can still have more unexpected opportunities to encounter people who have different backgrounds and ideas. In other words, we should build and appreciate the real social networks where we can interact with people who have different backgrounds and ideas, while recognizing the deceptive nature of superficial network ties created by social media.

5.2. Democracy and Communication Ability with Social Media

The nineteenth-century French liberal theorists were concerned that the centralization of knowledge would bring uninformed voters to factionalism, which would make it difficult for democracy to function. In the age of social media, similar concerns emerge corresponding with the rise of populism. Müller (2016, p. 3) identifies the cornerstone of populism as—in addition to anti-elitism—anti-pluralism, which is characterized by the dismissal of people who oppose a given viewpoint, asserting that their own viewpoint is that of the “true people” of the nation, that anyone who counters it is “part of the immoral, corrupt elite,” and further that “whoever does not support populist parties might not be a proper part of the people.” This definition contains elements of factionalism, in-grouping, partisanship, and silencing opposing voices, which have all become standardized elements of social media algorithms.

Another identifiable populist trend is the presence of a popular leader who can reframe narratives to make their group into the heroic rivals of an elite group. Doing so requires a great communicator, like those whom Duhigg (2024) describes as “supercommunicators.” Populist leaders are charismatic and can pander to their audience to make those who listen to them feel heard. Typically, this type of communication skill is a valuable tool to foster collaboration and cooperation in business and life. However, if used with social media, it may make listeners feel that their connection to the speaker is vulnerable, imposing an undeserved sense of certainty and trust within a speaker’s narrative. Populist politicians who can use these rhetorical tactics are more marketable on social media, allowing their message to be more well-suited to the algorithmic path to virality. Thus, their popularity will be augmented by tapping into channels that spread their rhetoric and hide opposing voices since their supporters do not want to hear them. This makes anti-pluralism not just an ideology (a set of beliefs) but an institution (a set of rules) in the very medium through which communication occurs. Therefore, political scientists have identified populism facilitated by social media as the greatest threat to democracy because social media has undermined the institutional infrastructure built to make democracy work. Although it is difficult to trace direct causation between the spread of social media and the rise of populism, one should note that the unique characteristics of social media platforms have amplified populist leadership’s narrative-building.

So, what should be done? We argue that moderate politicians should become supercommunicators to counter radical rhetoric from populist leaders by adapting to the current political landscape using social media. To achieve this goal, what advice should advisors and media consultants of the moderate politicians give? Provocative content spreads well on social media platforms, and one can make a moderate message provocative. Controversy has been the primary means of increasing the provocative nature of messaging, but that does not mean that it is the only way to increase the popularity of content. For example, humor, playfulness, and wit might contribute to making a message popular. Thus, moderate politicians could amplify the marketability of their content by making it funny and interesting. In the age of social media, politicians have to communicate with voters directly. They cannot rely on the elite cues that traditional media have delivered. To be a supercommunicator, they are required to be a great listener. Then, when sending their message, they should tell the voters a narrative—not a fairy tale—reflecting their visions in a convincing way.

Duhigg (2024, p. 30) argues that anybody could be a supercommunicator by following a certain list of tips and improving their proficiency at delivering a message, suggesting that the conditions to be a supercommunicator are as follows: (1) pay attention to what kind of conversation is occurring; (2) share your goals and ask what others are seeking; (3) ask about others’ feelings and share your own; and (4) explore if identities are important to this discussion. The key to making your message influential is how to connect with its recipient. Lupia (2016) introduces an interesting example to which we may apply Duhigg’s conditions to examine how to communicate: that is, how to discuss climate change with conservative evangelical pastors who are skeptical about scientific findings. He suggests that instead of “attempting to persuade evangelicals to trust the science, … [we should develop] their criteria from a point of view about the importance of various issue attributes that many evangelicals could accept,” such as the standpoint that as Christians they should seek to learn how to respond to a dramatic effect on the earth’s climate (Lupia 2016, pp. 135–6). In other words, even if one is confident in the scientific consensus on the existence of human-made climate change, one should not start the discussion with climate science but with what to do as a Christian for God’s gift—which is the earth’s climate in this case. In sum, Duhigg (2024) and Lupia (2016) both argue that the delivery of a message is more important than the message itself. And this perspective suggests that for a moderate politician to be a supercommunicator to confront their radical counterparts in social media, they would not have to change or radicalize the content of their message, but instead they should focus on improving how to deliver their message, which they could do by examining the perspectives of social psychology regarding effective communication.

Political theorist Mouffe (2022, p. 23) argues that “only by acknowledging the contradictory drives set to work by social exchange is it possible to grasp the practices and institutions necessary to secure democratic order.” For her, as “the agonistic tension between the liberal and the democratic principles … is constitutive of democracy” (Mouffe 2018, p. 16), we should not suppress conflict by curtailing radical voices to foster agreeability. She sees agonism as necessary for passions for democratic mobilization. Thus, what should be done is not to delegitimize adversarial politics but to institutionalize adversarial contestation so that antagonism will be converted into agonism. We agree. However, we should note that with most people taking social media as the primary source of political information, it is much more difficult to set an effective institutional mechanism to grant what Mouffe appreciates as agonistic domestication. At the same time, building such a mechanism does not contradict elevating politicians’ communication capability.

6. Conclusions

As articulated in Englert’s (2024) discussion of the problems identified by the nineteenth-century French liberal theorists with the capacité of non-elites to make sound political judgments, the concern over an ill-informed public in democracy is not new. However, liberalism and democracy have been entwined to hold the reins of the populace and to avoid the inherent chaos brought by mass democracy. In the actual operation of democracy, concern about the competence of uninformed voters was mitigated via the institutional infrastructure of a representative democracy, wherein the voices of the people were still honored through electoral processes, but their raw force was tamed through intermediary influences. Thus, the institutional mechanism allowed for the creation of a tier of political elites in the form of representatives, parties, and government officials, all of whom were versed in politics beyond ordinary people and hence played the role as a source of information and influence guiding the hand of the voting tier.

However, the current exponential expansion of communication, both in scale and accessibility, has removed the barriers to entry for political action. Social media has created a system for political advocacy that lacks the mechanisms and regulations to order the flow of information and ideas. As a result, the ballot in mass democracy based on social media algorithms has replaced the institutional infrastructure that enabled uninformed voters to make reasoned choices. Not only is social media not reined in by democratic institutions, but its operation is also in direct contradiction with the functionality of representative democracy. Rather than attempting to mitigate factionalism, increase the diversity of ideas, promote collaboration and cooperation among the people with different preferences, and prevent less competent masses from being too influential, social media promotes extreme views, gives a platform for bot accounts and unreliable information sources, and inhibits the diversity of ideas by purging the people with different preferences.

Furthermore, the landscape of American politics is highly polarized along regional lines, which has made it more difficult for representatives to be disjoined from the majority preferences within their jurisdictions (e.g., Levitsky and Ziblatt 2023). With representatives’ voting patterns and beliefs increasingly aligning with those of their constituency, the distance between representative democracy and mass democracy has shrunk, which has ironically realized the concerns that the nineteenth-century French liberal theorists had regarding the lack of capacité of voters. Thus, we highlight the correlation between the shift in the landscape of political communication and the erosion of the institutional mechanism that has made liberal democracy work.

The proposals to counter the problems induced by the spread of social media discussed in this article would only represent a handful of measures that could better integrate social media into current political circumstances, generating a sense of hope for the future of political communication and democracy. It does not mean that we cannot mitigate the problems that social media pose for the functionality of representative democracy. Instead, we should take them as a reminder that a stagnant system could not function in a world that is anything but consistent. We should keep in mind that governmental systems ought to evolve alongside the societies. Current liberal democracies should not have to be victims of the spread of social media and the rise of populism. Rather, we should take this situation as an opportunity to mobilize people to reconsider how to work with the government and how to interact with other people, reflecting upon how contingent these interactions have become upon social media, and further, reconsidering our relationship with social media altogether.