Digital Threats to Democracy: Comparative Lessons and Possible Remedies

Caught Off Guard

This special issue of The International Journal of Press/Politics emerges out of a research development workshop convened by the Media & Democracy program at the Social Science Research Council in June 2019, preceded and followed by open calls for papers. The workshop, titled “Digital Threats to Democracy: Comparative Lessons and Possible Remedies,” was premised on the notion that many democracies around the world seemed unprepared to combat the threats related to digital media that have become all too common over the last decade. From surveillance to disinformation to coordinated hate and intimidation online, the once-heralded tools of digital liberation have been repurposed and redeployed in illiberal and antidemocratic ways that were, if not unimaginable, simply unimagined by many of the most prominent early theorists of digital politics.

For many observers, these threats seemed to culminate in the unanticipated electoral outcomes of 2016, which featured a U.S. presidential election and a U.K.-wide referendum on the country’s withdrawal from the European Union, colloquially known as Brexit. We now know that disinformation campaigns combined with the mass deployment of automated bots and trolls were coordinated with the purpose of interfering with democratic processes in each country (as discussed in this special issue by Golovchenko et al. (2020) with respect to Russian interference in U.S. elections and by Tenove 2020 in terms of the policy responses to these threats to national sovereignty).

Yet those events may have been canaries in the coal mine. Since 2016, it has become clear that disinformation is not simply a foreign threat; that successful campaigns are waged by domestic purveyors of disinformation and, indeed, depend heavily for their success on networks of communication that include traditional media organizations, such as tabloid newspapers (Chadwick et al. 2018), radio stations, and cable news (Benkler et al. 2018). Moreover, hate and toxicity are used by partisan activists and supporters to intimidate and silence political opponents in ways that disproportionately affect traditionally marginalized groups (Sobieraj 2019; Udupa et al. 2020; see also Whitten-Woodring et al. 2020 in this special issue). Mass electorates are increasingly concerned with the spread of misinformation and disinformation (as documented in this special issue by Koc-Michalska et al. 2020), to the point that many doubt whether news can be trusted, question whether truth can be discerned in public debate, or withdraw from the news altogether to stay out of trouble (Vaccari and Chadwick 2020). This may have detrimental effects on satisfaction for democracy and belief in the legitimacy of this complex project (as highlighted in this special issue by Stoycheff 2020). And today, as the world reels amid a global pandemic, the value of quality information has taken on new salience and the vulnerability of some sectors of our media ecosystems and of our societies have been exposed in unprecedented ways.

It has long been held that democracies cannot thrive in the absence of freely available and trustworthy information (Flynn et al. 2017). Yet when lives are at stake and when every decision—every minute—matters, democracies depend on quality information, not simply to function well, but to function at all. And while the research presented here could not have foreseen the emergence of the COVID-19 pandemic, it nevertheless offers support for an emerging consensus within political communication: that perhaps the most significant threat to democracy around the world is the fact that flows of information, and the democratic quality of such flows, vary significantly by political and social context. Indeed, given the ubiquitous use of algorithms to personalize and micro-target content—from advertisements to music recommendations to search results—flows of information vary widely even from individual to individual. Moreover, it is increasingly evident that a few, entrenched quasi-monopolist corporations exercise unprecedented amounts of power over online information flows (Hindman 2018; Wu 2017) and that such flows are channeled, amplified, and augmented by interactions with organized political groups and traditional media, in complex ways unaccounted for in utopian visions of a free and open Internet. In short, if perception is reality, democratic citizens may inhabit starkly different worlds, shaped in large part by the distinct information ecologies they engage with on a daily basis. And while contemporary news diets and experiences of politics may not necessarily be as fragmented and idiosyncratic as implied by the most pessimistic scenarios (as shown by Stier et al. 2020 in this special issue), some of the niche news ecosystems that have developed over the past few years raise concerns for democratic governance, which this Special Issue aims to address.

In what follows, we offer a necessarily cursory exegesis of the shift in narrative about digital media from tools of liberation to tools equally well suited to illiberal and antidemocratic ends. In short, it was once believed, not only that flows of information (or “speech”) on the Internet could not be controlled, but that it was precisely from this free flow of information that the liberating potential of digital media derived. Today, however, it is increasingly evident, not only that information is subject to control—by states and private actors alike—but that it is influenced by a variety of actors such that users may experience starkly different realities online depending on their identities, preferences, resources, and the social and political context surrounding them.

While our analysis will mostly focus on the technological and commercial development of the Internet over the last thirty years, these transformations have also accompanied, and been accompanied by, relevant changes in democratic polities. Political actors and states have learned that the flow of information online can be shaped, at least in some spaces and among some groups, and the commercial development of social media platforms has provided them with a vast array of tools to achieve that goal. As digital advertising has grown in reach and precision, the ability to effectively communicate online has become increasingly tied to the availability of financial resources, especially in high-profile national campaigns. And while the Internet has clearly helped challenger parties and candidates make a mark in electoral competition—from the Five Star Movement in Italy to Podemos in Spain, from Barack Obama in the United States to Emmanuel Macron in France—it has also empowered well-resourced outside actors, funded by nontransparent corporate or private interests, to acquire disproportionate visibility without corresponding accountability and responsibility. Moreover, evidence suggests that the financial crisis of 2007–2008 and the long recession that accompanied it, together with the policy responses by most national and international institutions (Hacker and Pierson 2010; Przeworski 2019), gave rise to a backlash against globalization and open societies that boosted the appeal of old and new authoritarian populists that have shown little respect, in words and deeds, for democratic norms and values such as separation of powers, the rule of law, and expertise (Norris and Inglehart 2019). While digital media have often enabled political entrepreneurs of nativist anti-elitism—from Donald Trump in the United States to Jair Bolsonaro in Brazil to Matteo Salvini in Italy—to bypass journalistic mediation to speak directly to their supporters, it is also the case that traditional media greatly contributed to their success (e.g., Boydstun and Lawrence 2020; Wells et al. 2016). In sum, while politics is constantly fraught with conflicts and challenges—arguably, they are politics’ very essence—the cleavages that have emerged in the present configuration of digital politics may be more complex, and more threatening to the democratic project, than in the recent past.

Information Revolution

Since its emergence in the 1960s and widespread adoption by the end of the twentieth century and through to 2016, the dominant narrative about politics on the Internet has revolved around a revolutionary imaginary. It was thought that the so-called “Information Revolution,” brought about by the development of Internet protocols and infrastructures, combined with advances in computer processing capacity, would transform passive consumers into the authors, curators, and publishers of mass-mediated content (Benkler 2006; Shirky 2009). This revolution promised new, “virtual” realities where people could recreate themselves outside the strictures of cultural norms, legal regulations, and political borders (Barlow 1996; Castells 2007; Negroponte 1995). This revolution heralded new ways of organizing social and political life that would shift power away from large, hierarchical organizations toward diffuse networks of autonomous individuals.

Despite such revolutionary rhetoric, the Internet originated within the halls of the Defense Advanced Research Projects Agency (DARPA) of the U.S. Department of Defense, and its purpose was to provide the U.S. military with a communications infrastructure capable of withstanding a coordinated nuclear attack. A feature of the Internet’s design known as the end-to-end principle dictates that functionality on the network reside in the computers that connect to the network, not in the network itself. In other words, the technical capacity to produce, store, and distribute data is held by computers, and the Internet itself is neutral, a latticework of “dumb tubes” that are exclusively concerned with transmitting information as efficiently as possible by following routing commands. Crucially, this distribution technology is unconcerned with the content of that information, or the physical location or identities of end users (either the client requesting information or the server hosting it).

Moreover, information is distributed on the Internet via a process called “packet switching,” in which data are broken up into smaller pieces and sent separately to their destinations, where they are reassembled as interpretable forms of information: images, audio files, or documents, for example. Packet-switching protocols determine that if any node in the network is destroyed or impaired, packets of data will still make their ways to their destinations: The network will simply reroute them along any number of alternative paths.

In theory, the end-to-end principle in combination with packet switching protocols engender a decentralized or distributed network, in which it would be impossible to disrupt communication on the network simply by destroying one or several centralized nodes in the network (Baran 1962; Galloway 2004). And while it is easy to understand how such a distributed infrastructure might seem advantageous from a military-defense point of view, its broader social and political implications were less straightforward. Nevertheless, as this infrastructure was privatized and made available for commercial use, its revolutionary potential began to be theorized by the narrow groups of scholars, computer programmers, and hobbyists who had been influential in its development or in its early adoption for popular use.

Importantly, many early theories of digital politics assumed that the same technical features that would shield the Internet from attack would also limit—even obliterate—the capacity of governments to control flows of information through the network (Barlow 1996; Johnson and Post 1996; Negroponte 1995; Shapiro and Leone 1999). Insofar as the network was resistant to centralized attack or disruption, it would also be immune to centralized control or management. If the network remained neutral, data sharing on it would remain autonomous and it would be impossible for states or other centralized actors to interfere with free expression in this sphere of social life.

If the Internet was fundamentally resistant to control, collective decisions would have to be made by alternative means, thus ushering in an alternative model of governance. It was argued that liberal and anti-hierarchical values were embedded in the very protocols that defined how computers must operate on the network—even in the mechanisms by which protocols were designed and implemented. Problems of the Internet would not be resolved by ballot or by edict, but by mechanisms of consensus among interested parties—chief among them computer programmers. According to John Perry Barlow (2000), this was because solutions on the Internet must win “by virtue of adoption, and they don’t get adopted if they’re bad solutions.” Summarizing this value system, one of the Internet’s architects stated, “We reject: kings, presidents, and voting. We believe in: rough consensus and running code” (Clark 1992).

Such was the prevalent view of many early theorists of the political Internet: that the technical features of the communication infrastructure would necessarily militate against centralized control; that this function would inevitably create opportunities for free expression; and that these opportunities were necessarily conducive to political liberation. These were decidedly normative—indeed, utopian—theories, with strong influences from the American strand of liberalism known as libertarianism. It is notable, moreover, that despite the many contributions of nonwhite and non-male scientists, hobbyists, and early adopters of Internet-based technologies, the utopian-libertarian strands of early Internet political theory were disproportionately concocted by white, Anglophone men (McIlwain 2019).

By the turn of the new millennium, utopian theories were still dominant, and even refined. Fast expanding access to Internet connections and increasingly affordable personal computing capacity were deemed to further augment the liberating potential of free-flowing information. Such capacity would give voice to otherwise marginalized political voices, enabling them to circumvent the gatekeepers of traditional media (Drezner and Farrell 2004). Far from being drowned out in a cacophony of competing voices, individual end users would leverage the emerging constellation of online social networking sites to express themselves and connect with each other in new, potentially disruptive, but fertile and empowering ways. The “information exuberance” (Chadwick 2009) that would result from this free flow of interactions and exchanges was generally seen as a breath of fresh air from an elite-controlled political communication system increasingly absorbed in cycles of spin and counter-spin between journalists and campaign professionals, crowding out ordinary citizens’ agency and engagement (Entman 1990; Gurevitch and Blumler 1995; Patterson 1994). With Wikipedia as the archetype, digital utopians argued that we no longer had to choose between the elite and the masses: We could choose both. We could leverage the capacity of citizens to create open, accessible cultural and economic products of superior quality than their analogue counterparts (Benkler 2006; Tapscott and Williams 2006). We could harness the ability of activists to organize movements that would be nimble, inclusive, and could scale more quickly than traditional organizations (Bennett and Segerberg 2013; Rheingold 2002; Shirky 2009; Surowiecki 2004). We could build new online commons that would enable citizens to express themselves, engage with causes they cared about, and meaningfully contribute to democratic self-governance even if they could not devote vast amounts of time and resources to these endeavors (Blumler and Coleman 2009).

It is important to note that the utopian visions of much early writing on the Internet and politics did not go unchallenged in social science research. From the early days of digital politics research, supporters of what came to be known as the “normalization” hypothesis (Margolis and Resnick 2000) argued that the Internet was not going to fundamentally transform key aspects of politics. Even more nuanced empirical analyses of the limits of the transformative potential of the Internet emerged at the turn of the millennium, focusing on its tendency toward concentration of power (Hindman 2008), the limited effects of the democratization of information it enables (Bimber 2003); its ability to magnify rather than reduce political inequalities (Bimber and Davis 2003); its limited capacity to foster coherent, inclusive, and transformative public dialogue (Papacharissi 2002); the reluctance among political parties to embrace its participatory potential (Gibson et al. 2003); and the risk that politicians would use digital tools to collect unprecedented amounts of individual-level data that would enable them to segment, manipulate, and disempower citizens (Howard 2006). Nuanced reviews of the field of digital politics emphasized the need to strike the right balance between technological determinism—ascribing inherently transformative and societally beneficial powers to the Internet—and social constructivism—highlighting the role of long-term social structure and power imbalances in shaping any implications of technology (Chadwick 2006). It is fair to say that most scholars who set out to study the relationship between the Internet and politics from the 1990s onward did not uncritically share most of the utopian predictions of the theorists whose ideas we have briefly summarized here. Still, most empirical research—including our own (Vaccari 2013)—started from the argumentative premise that the Internet could indeed have a democratically enhancing transformative potential and that the possibility of such potential deserved empirical scrutiny, if only to circumscribe them, contextualize them, and identify the conditions under which they were more and less likely to materialize. In other words, the utopian techno-optimism of the first wave of digital politics theories did not shape the results of academic research, but it definitely informed the questions scholars set out to answer and the (often implicit) normative frameworks based on which they interpreted their findings.

Information Reaction

With a zeal that would seem out of place today, and an unwitting signal that digital utopianism had reached its peak, Clay Shirky titled his 2009 book Here Comes Everybody. Even more revealing is the book’s subtitle—How Change Happens When People Come Together. As the Internet developed, the amount of people who can come together and the reasons why they come together has continued to increase. In hindsight, we may have gotten more than we bargained for.

The so-called “Arab Spring” of 2011 seemed to confirm the digital utopian thesis that popular sovereignty would be enabled by digital tools of connectivity and access, streamlined by then-ascendant social media platforms. But even as revolutions in the Middle East and North Africa toppled dictatorships, and as scholars embarked on a contested debate about the extent to which digital media had really contributed to these revolutions (Howard 2010; Tufekci and Wilson 2012; Wolfsfeld et al. 2013), elsewhere in the world the story was quite different. Indeed, at the close of the first decade of the new millennium, the same technologies that briefly liberated citizens in Egypt and Tunisia were being deployed in novel ways to surveil and repress dissidents in authoritarian regimes such as China and Iran (Diamond and Plattner 2012).

The first cracks in digital utopianism started to appear. In the first case, access, so fundamental to the utopian thesis, was shown to be unevenly distributed. Not only was access not nearly as extensive outside of the wealthy capitalist democracies of the West, but even within these countries, it varied significantly by demographic characteristics such as age, race, gender, and education (Norris 2001). Even as the divide in access narrowed, scholars identified persistent inequalities in both technological skills and media literacy (DiMaggio et al. 2004; Hargittai and Shaw 2015; Van Dijk 2005). As an example, research has found that most disinformation in the 2016 U.S. Presidential election was spread by older conservative voters (Guess et al. 2019)—precisely the kinds of users who were less likely to use the Internet in the early stages of its mass diffusion (Hindman 2005). As research by Pennycook and Rand (2019) has shown, users who employ simple, “lazy” styles of thinking are more likely to believe online disinformation than those whose cognitive styles are more systematic and “analytical.” To the extent that the Internet has become increasingly dominated by platforms whose user-friendly interfaces are constantly refined to keep their users entertained, engaged, and glued to their screens, it seems to be more conducive to “lazy” thinking than the more cumbersome and self-directed “information superhighway” of yesteryear. And as smartphones with mobile broadband have further democratized access to the Internet, especially in poorer countries and among sectors of the population without personal computers, it has become apparent that smartphone screens are not ideally suited to consume complex content such as news (Dunaway and Soroka 2019). While these may be fair prices to be paid for a more socially inclusive digital ecosystem, these were arguably not prices that were fully anticipated, nor accounted for, in early discussions of the digital divide. Moreover, as critical race scholars have demonstrated, when different groups go online, they often utilize new media technologies in ways that early adopters had not—indeed, could not have—anticipated (Brock 2012; Steele 2016, 2018). Thus, as the cohort of early technological adopters was progressively outnumbered online by a larger and more diverse population of newer digital citizens, the online marketplace of ideas has arguably become more complex and even chaotic, complicating not only the story of the Internet as a space of universal access but also the simple dichotomy of Internet haves and have nots.

However, the more fundamental critique of the utopian consensus was that, contrary to popular belief, control of end user experience and behavior on the Internet was not only quite possible, but could be managed in ways that went beyond maintenance of the status quo, and actually enhanced state capacity to surveil and repress dissent (Goldsmith and Wu 2006; Morozov 2012). In many ways, utopian hubris about the liberating effects of digital media reflected a U.S.-centered modernist meta-narrative about technological progress. Perhaps it is not surprising then, that the most damning critiques of the utopian liberation technology thesis emerged from scholars who studied digital politics in settings where liberal democratic norms could not be taken for granted.

In many authoritarian regimes, for example, states were so central to the adoption and expansion of Internet infrastructures that they could ensure these technologies ultimately served state-defined priorities, even as some digital affordances enabled limited challenges to state authority (Kalathil and Boas [2003] 2010). In fact, information controls evolved a great deal over the decade that culminated in the Arab Spring. While “first and second generation controls” worked by blocking or filtering content through technical, legal, or regulatory mechanisms, “third-generation controls” focused less on denying access to Internet resources and more on waging propaganda and counter-information campaigns, as well as using surveillance as a means to create cultures of fear that would discredit, demoralize, and overwhelm opponents (Deibert et al. 2008, 2010, 2011). In an illuminating analysis of how the Chinese state domesticated the Internet and turned it into a tool of state control, Roberts (2018) identifies three types of censorship—“fear,” deterring people from speaking their minds by threatening sanctions; “friction,” making it hard for people’s voice to be heard by manipulating, or temporarily stopping, the flow of information; and “flooding,” or drowning out dissent with an avalanche of content focused on entertainment, lifestyle, and feel-good news. The “social credit system” that is being rolled out by the Chinese government is poised to dramatically extend state control over citizens’ lives by connecting digital trace data on their online activities—chief among them political self-expression—with a system of rewards and punishments that shape access to fundamental public services and opportunities (Liang et al. 2018). This would entail a full-blown reversal of the original utopia of the Internet as an inherently anarchic and libertarian force, putting it at the service of a pervasive project of authoritarian social control instead.

Yet despite an emerging consensus that governments could, in fact, deny access to—or otherwise disrupt flows of—information online, to many at the end of the first decade of the new millennium, the antidemocratic and illiberal threats of digital media still appeared as non-Western and nondemocratic phenomena (Deibert et al. 2011; Diamond and Plattner 2012). If there were threats to democracy borne of digital media, they were distinctly authoritarian threats—challenging democracy’s emergence, not its survival where it had long since taken root.

The first clue, or perhaps the first publicly visible sign, that digital threats to democracy might be homegrown in the West as well, was the 2013 Guardian scoop that exposed the U.S. National Security Agency (NSA) and the British Government Communications Headquarters (GCHQ) mass surveillance programs (Greenwald 2013). The scandal unveiled by NSA contractor Edward Snowden revealed that the U.S. and U.K. governments had not only the technical wherewithal but also the will to conduct dragnet surveillance of their own citizens (Geiger 2018). Moreover, the investigation closed the door rather tightly on optimistic takes on digital politics that assumed the individual liberty-enhancing qualities of digital media. Indeed, even within the field of surveillance studies, which was already disposed to more critical analyses of technology in society, the Snowden revelations represented a critical juncture, such that the field has been described in terms of “before” and “after Snowden” (Lyon 2015).

If the first clue signaled a threat to the individual liberties of democratic citizens, the second signaled an intrusion on their popular sovereignty. In 2016, citizens in the West became aware, all at once it seemed, of coordinated disinformation campaigns seeking to disrupt the U.S. presidential election and the U.K. referendum on exiting the European Union. It is now known that agencies and operatives acting at the behest of the Russian government conducted large-scale propaganda campaigns, including the purchase of advertisements on major social media platforms and the mobilization of an army of bots and trolls, each seeking above all to sow mistrust and discord among citizens in two of the world’s oldest democracies (Golovchenko et al. 2020 in this special issue; DiResta et al. 2018; Howard et al. 2018; Jamieson [2018] 2019). While most scholars are skeptical that these efforts were enough, per se, to sway the outcome of these elections, the fact that these efforts were even possible on a large scale, often with the unwilling complicity of domestic news media (Lukito et al. 2020) and self-interested domestic politicians (Giglietto et al. 2019) contributed to dismantling the useful democratic myth of the “attentive public” (Karpf 2019). This is the idea that mass electorates, while incapable of and uninterested in constantly and closely scrutinizing politics, can be expected to see through blatant and pervasive acts of mass deception and defend themselves from them at the voting booth (Erikson et al. 2002; Lupia et al. 1998). Whether this state of affairs is empirically true or not (Achen and Bartels 2017), one of the bedrocks sustaining the legitimacy of democratic government is that the public is capable of ultimately discerning what is good for itself and that self-interested politicians have only limited ability to elude voters’ vigilance. As Abraham Lincoln—or possibly P.J. Barnum—once said, “You can fool some of the people all of the time, and all of the people some of the time, but you cannot fool all of the people all of the time” (Vaccari 2019). Another myth that 2016 contributed to shattering is what Hedrick and colleagues (2018) call the “earnest internet,” or the idea that most people can be trusted to mean well and behave fairly and honestly when they are online. This arguably naïve idea, which nonetheless informed decades of theories and prophecies around the wisdom of the crowds, the gift economy, and the self-correcting capabilities of distributed networks, just to name a few, has given way to concerns about a new “culture of indeterminacy” (Chadwick 2019) fueled by the widespread assumption that little of what is shared online can be trusted and, hence, that individuals have little agency, let alone responsibility, for improving the quality of public communication on the Internet (Vaccari and Chadwick 2020).

Taken together, these two signs—one, an intrusion on the privacy of democratic citizens; the other, an intrusion on democratic citizens’ ability to make free electoral decisions—put to rest the libertarian-utopian notion, once the consensus view among scholars and commentators in the United States and across most of the West, that digital media were necessarily tools of liberation. Moreover, they compelled Western scholars to take seriously the notion that inasmuch as digital media might be boons to the twin pillars of Western democracy—individual liberty and popular sovereignty—they might also threaten them, whether directly, by impeding the processes they require, or indirectly, by diminishing their legitimacy among the public and weakening the social norms that make democracy possible in the first place.

Today, scholars such as Bimber and Gil de Zúñiga (2020) and Bennett and Pfetsch (2018) refer to an “unedited” or “disrupted” public sphere, in which some forms of speech that enrich democracy have been strengthened, even as antiliberal forms of speech, such trolling, harassment, and expressions of hate make online discussion spaces inhospitable or unsafe (Boatright et al. 2019; Phillips 2015). Indeed, precisely because gatekeeping power has been redistributed (but not necessarily democratized), digital media users are freer than in the past to consume, produce, and distribute misleading, polarizing, or speech-suppressing content. What is more, organized political actors—both inside and outside the system of legitimate democratic representation—have become increasingly capable of harnessing and channeling bottom-up efforts by online activists, which can scale up to substantial coordinated campaigns amplifying elite messages (Giglietto et al. 2020; Zhang et al. 2018). Parties and candidates have also been enabled to reach susceptible groups by using sophisticated but nontransparent digital advertising tools that tech platforms developed to maximize their profits and with an eye to the consumer market, often without putting much thought into what would happen were the same tools to be deployed to influence democratic elections and decisions (Vaidhyanathan 2018; Wu 2017; Zuboff 2019).

The result is that today we may inhabit wildly different information environments and that these communication microclimates are increasingly shaped by well-resourced organized political actors and their partisan supporters, as well as by digital platforms’ commercial imperatives, rather than by a genuinely free flow of ideas and information—or, for that matter, a flow organized according to widely agreed-upon norms of democratic citizenship. In short, and in contrast with the first generation of digital politics research, scholars are increasingly aware that some aspects and uses of the Internet may threaten key features of democracy—as a project, as a possibility, and as a system of government.

Clarifying Challenges, Assessing Solutions: This Special Issue

Against this backdrop, scholars of media, technology, and politics have a key role to play in asking and answering empirical questions about the extent and causes of digital threats to democracy, and in evaluating the potential costs and benefits of different remedies. The need to empirically address these challenges also reveals the limitations of some of the normative underpinnings on which research in political communication, and digital media and politics in particular, has rested so far (Bucy and D’Angelo 2004; Chadwick 2019; Strömbäck 2005). Faced with the concrete possibility that the current configuration of the affordances of digital media may exacerbate certain long-standing threats to democracy, and given the delicate balance between the need to protect societies from these threats and the risk of limiting fundamental rights such as freedom of expression, researchers have the right and the duty to contribute knowledge and ideas to a public debate that is often fraught with short attention spans by the public, indulgence to simplification by journalists, facile posturing by politicians, hasty solutionism by policymakers, and prevarication by tech companies (Nash 2019; Tromble and McGregor 2019; see Tenove 2020 in this Special Issue for a review of existing policy responses).

In this context, our call for papers asked international scholars from different disciplines to reflect and shed light on the challenges that digital media pose to democratic institutions and processes. What lessons might be learned from the different experiences of polities around the world, including nondemocratic ones, that have been facing digital threats to democracy for longer than their counterparts in Europe and the United States? How do such threats manifest in different political settings and are there generalizable insights about the downsides of digital media in democratic contexts? What do we know about the ways in which these threats can be tackled, and their most pernicious implications reduced?

The research collected in this issue suggests that one of—if not the—greatest digital threat to democracy derives from precisely the same mechanism that was previously expected to liberate the minds of anyone with access to an Internet connection and computing capacity. The free flow of information online has meant that anyone and everyone—be they independent learner, civic-minded activist, voter, agent provocateur, or bot—has some capacity to shape communication and, in turn, knowledge, public opinion, and social behavior. But just as does money in an economic market, unregulated flows of information get concentrated in ways that yield distinct advantages, even quasi-monopolies, to certain organizations, groups, and actors over others, and a digital divide—again, if not in access to resources, certainly in skills, literacy, and, ultimately, influence and power—looms large as a result. Traditional theories often assumed that an open Internet would necessarily result, not only in better informed users, but in societal consensus through the seamless real-time aggregation of individual contributions treated equally and evaluated solely based on their merits, as evidenced by support gathered among other users. Instead, in many instances digital media have facilitated the spread of misinformation and disinformation, or of opinionated content that, albeit not blatantly false, contribute to ongoing processes of social and political polarization and fragmentation, as well as distorting the aggregation of interests and preferences.

The eight contributions collected in this Special Issue advance our understanding of some potential sources of digital threats to democracy, of how citizens are affected by and may contribute to these threats, and of some possible solutions to these problems. These studies combine a variety of disciplinary perspectives—political science, media and communication, and sociology—and employ a variety of methods—computational analysis, in-depth interviews, focus groups, surveys, analysis of secondary data, and policy analysis—often combining multiple approaches and comparing different countries. The articles cover twenty-three countries in four different continents, including established Western democracies (Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, The Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the United States), a younger, and arguably backsliding, Eastern European democracy (Poland), an East Asian democracy (Taiwan), and two authoritarian regimes (Myanmar and Russia). Below, we provide short summaries of each contribution.

Moving Forward

We hope this Special Issue will contribute to and inspire the work of researchers, practitioners, digital media companies, and policymakers around the world who care about democracy and are concerned about threats to its functioning—and even its existence—that have emerged out of largely unforeseen, but not unimaginable, developments in the functioning of the Internet and the ways in which different types of actors employ digital affordances to achieve political goals.

While no collection of articles can shed enough light on challenges so diverse, complex, evolving, and context-dependent, the studies included in this Special Issue highlight the benefits of research that is international, interdisciplinary, multimethod, and multiplatform. As we trust readers will agree after engaging with the contributions presented here, our knowledge is rapidly growing in breadth and depth. Nevertheless, many challenges lie ahead for those who aim to better understand and mitigate digital threats to democracy. We highlight three that we think are particularly relevant.

First, our understanding of the problem is still partial. For one, most ongoing research on digital threats to democracy has focused on disinformation, as this Special Issue itself testifies. While this is definitely an important challenge, it is not the only one worth considering. Hate speech and intolerance, mass surveillance, censorship, and perverse uses of microtargeting, just to name a few, are equally as troubling. We also do not know enough about how these phenomena unfold beyond high-profile moments such as elections, in private as well as public and semi-public channels of online communication, and outside of Western democracies. And while we have developed some useful knowledge of how politicians, journalists, and citizens are navigating, and sometimes contributing to or even taking advantage of, these challenges, we do not know nearly enough about how digital media companies make decisions—from grand strategies to temporary experimentations—that dramatically shape the context in which we communicate about public affairs.

Second, there is the danger that the pendulum has swung, or will swing, too far too fast toward digital pessimism. A renewed focus on the democratic challenges that the Internet has brought about should not distract us from the opportunities it continues to provide. There is no golden age of mass-mediated political communication that we should yearn for, as entire library stacks of scholarship critical of mass media’s contribution to democracy testify. Scholars will need to answer pressing normative and empirical questions on how contemporary media ecosystems contribute to democracy, without any bias in favor of optimistic or pessimistic narratives. Research on incivility in online political talk offers an illustrative example of a journey in the opposite direction to that currently taken by scholars of disinformation and other digital threats. While early studies considered most, if not all, forms of online incivility as inherently problematic for democracy (Mabry 1997; Ng and Detenber 2005), contemporary research has developed more nuanced conceptualizations of different types of antinormative online political talk, leading not only to more precise empirical measurements of a complex constellation of phenomena but also to more refined, and overall less pessimistic, normative assessments of their contributions to democracy (Masullo Chen et al. 2019; Papacharissi 2004; Rossini, forthcoming). It is too early to tell whether the currently burgeoning field of online disinformation will follow a similar path as scholars continue polishing their concepts, measurements, and underlying normative models, but the possibility should not be ruled out.

Third, the long-term and indirect collective implications of digital threats to democracy require as much, if not more, attention than their short-term and direct individual effects. For instance, while it is important to know how many people are directly exposed to and affected by disinformation or hate speech, it is just as important to understand whether the sustained public debates and news coverage around these problems may affect citizens’ attitudes and behaviors in the long run. Does discussion of digital threats to democracy risk exacerbating some of these problems by further eroding trust in other citizens, in the news, and in the viability of the democratic project of self-governance itself? How are the norms and expectations around what it means to be a producer, consumer, and distributor of publicly relevant information changing as a result of the negative climate of opinion around online political debate, and how are those developments affecting the ways in which individuals and groups behave politically? And how does the potential weakening of democratic communicative norms shape users’ behavior during exceptional moments, such as contested high-profile elections, constitutional crises, eruptions of social conflict, and public health emergencies such as the COVID-19 pandemic that looms large as we finalize this Special Issue?

We trust that these and other questions will keep our research communities occupied and engaged in lively, difficult, and important debates for the next few years. We hope this Special Issue can play a small part in this collective endeavor.