Skip to main content Accessibility help


  • Access
  • Open access

  • Social Media and Democracy
  • The State of the Field, Prospects for Reform
  • Online publication date: August 2020
  • pp 139-162
  • Publisher: Cambridge University Press


      • Send chapter to Kindle

        To send this chapter to your Kindle, first ensure is added to your Approved Personal Document E-mail List under your Personal Document Settings on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part of your Kindle email address below. Find out more about sending to your Kindle.

        Note you can select to send to either the or variations. ‘’ emails are free but can only be sent to your device when it is connected to wi-fi. ‘’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.

        Find out more about the Kindle Personal Document Service.

        Available formats

        Send chapter to Dropbox

        To send content items to your account, please confirm that you agree to abide by our usage policies. If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your account. Find out more about sending content to Dropbox.

        Available formats

        Send chapter to Google Drive

        To send content items to your account, please confirm that you agree to abide by our usage policies. If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your account. Find out more about sending content to Google Drive.

        Available formats
Export citation


The move to a more digital, more mobile, and more platform-dominated media environment represents a change to the institutions and infrastructures of free expression and a form of “democratic creative destruction” that challenges incumbent institutions, creates new ones, and in many ways empowers individual citizens, even as this change also leaves both individuals and institutions increasingly dependent on a few large US-based technology companies and subjects many historically disadvantaged groups to more abuse and harassment online. This chapter aims to step away from assessing the democratic implications of the internet on the basis of individual cases, countries, or outcomes, but rather to focus on how structural changes in the media are intertwined with changes in democratic politics.

7 Democratic Creative Destruction? The Effect of a Changing Media Landscape on Democracy

Rasmus Kleis Nielsen and Richard Fletcher


The move to a more digital, more mobile, and more platform-dominated media environment represents a change to the institutions and infrastructures of free expression and a form of “democratic creative destruction” that challenges incumbent institutions, creates new ones, and in many ways empower individual citizens, even as this change also leaves both individuals and institutions increasingly dependent on a few large US-based technology companies and subjects many historically disadvantaged groups to more abuse and harassment online. That is the argument we advance in this chapter, where we will aim to step away from assessing the democratic implications of the Internet on the basis of individual cases, countries, or outcomes to focus on how structural changes in the media are intertwined with changes in democratic politics.

We will set aside considerations of (important) individual phenomena like the Arab Spring, the indignados movement, and #MeToo, or (important) individual outcomes like the 2014 Indian general elections, the UK (Brexit) referendum on EU membership, or the 2016 US presidential elections, and instead identify a few key changes at the institutional level and the individual level that are part and parcel of the rise of digital media and discuss how this rise is in turn changing the institutions and infrastructures that enable free expression. Inspired by James Reference WebsterWebster (2014) and his work on structuration, we examine structural change by considering the interplay between institutional change on the supply side and aggregate individual-level behavior on the demand side.

We will do so through the lens of news, first the news media as an institution and second news as part of how individual citizens engage with public life. We focus on news as one of several key aspects of democratic politics, key to how we imagine it in its ideal forms and key to how we realize it imperfectly in practice. The structural changes we analyze are not dictated by technology but influenced by it and will play out in different ways in different contexts as technological momentum is shaped by cultural, economic, political, and social factors. Despite the enormity of the question “what does digital media mean for democracy?” and the impossibility of giving a single, clear, definite, universal answer to it, we believe it is useful to try to identify key, common aspects of global structural changes underway in democracies worldwide by synthesizing recent empirical research. We summarize these changes under the rubric of “democratic creative destruction” (and because of our focus on democracy, we are primarily concerned with developments in countries that are considered more or less free – currently accounting for somewhere in between a third and just over half of the world’s population depending on one’s definition of democracy and the index on which one relies).

We take the notion of creative destruction from the Austrian political economist Joseph Schumpeter, who used the term to capture ongoing structural changes in capitalism, to help us think about how structural changes in the media impact democracy. In his classic 1942 book Capitalism, Socialism, and Democracy, he argued that, in the long run, the history of social change is a history of ongoing evolution, but that the history of economic change specifically is a history of revolutions, of an ongoing process of creative destruction that “revolutionizes the economic structure from within, incessantly destroying the old one, incessantly creating a new one” (Reference SchumpeterSchumpeter 1992, p. 83). Schumpeter suggested that, faced with this kind of dynamic change, any analysis that focuses on a given point in time (whether a high point or a low point) risks misunderstanding the structural changes underway, and that any analysis focused on any particular part (as important as it might be) risks missing the systemic implications. The moments and the parts are important, but we adopt Schumpeter’s systemic view here. Our democracies are changing in response to many other, sometimes more important, factors than this kind of creative destruction, including those that political scientists and sociologists traditionally focus on, from changes in party systems (Reference MairMair 1997) and forms of governance (Reference RhodesRhodes 1997) to changes in trust in institutions (Reference NorrisNorris 2011) and people’s value systems (Reference InglehartIngelhart 1997). Yet because our democracies are intertwined with the news, and because the news is in turn intertwined with the market forces and technology that are central to creating, maintaining, and changing the institutions and infrastructures of free expression, Schumpeter’s approach to thinking about structural transformations gives us a way of capturing a few key aspects of the big picture. We focus here on what it means for news, though by extension, as political scientists have long argued, information revolutions often involve fundamental change in how governments, political parties, interests groups, and social movements operate too, because most of the institutions that enable democracy are interdependent (see, e.g., Reference BimberBimber 2003).

News is important here because of the different roles it plays in democracies. We should not romanticize how well actually existing journalism in actually existing democracies has in fact lived up to the various ideals we might have for journalism and for democracy. Yet with its many imperfections, at least in North America and Western Europe, empirical research suggests that independent, professionally produced news has helped inform the public, helped people make sense of the world through analysis, interpretation, and the portrayal of contending forces, and helped members of the public connect with one another to see themselves as part of a community and act in concert to influence public affairs (Reference Van Zoonenvan Zoonen 1998; Reference Curran, Iyengar, Lund and Salovaara-MoringCurran et al. 2009; Reference Couldry, Livingstone and MarkhamCouldry et al. 2010). Beyond these, perhaps the most visible democratic role of news, investigative journalism specifically can also produce a range of “positive externalities” that benefit the whole public – even those who do not actually engage with a particular story – by ensuring more efficient local government, reducing corruption, and increasing how responsive elected officials are to their constituents (Reference Brunetti and WederBrunetti and Weder 2003; Reference Besley and PratBesley and Prat 2006; Reference Snyder and StrömbergSnyder and Strömberg 2008). Again, we do not want to romanticize journalism as we knew it. News is necessarily always imperfect, and for much of the twentieth century much of it was all too often by white men, about white men, for white men, uncomfortably close to political elites and oriented toward the affluent, often not trusted, and not always trustworthy. Yet while journalism arguably has been able to do less for democracy than many journalists and academics like to think (Reference GansGans 2003), and sometimes plays a more ambiguous role, news is integral to actually existing democracies, because it helps citizens navigate the (public) world beyond personal experience.

In the rest of the chapter, we identify key aspects of the move to digital, mobile, and platform-dominated media first for the institutions that underpin the professional production of news (Reference CookCook 1998), then at the individual level to see how it affects the “public connections” that news can enable (Reference Couldry, Livingstone and MarkhamCouldry et al. 2010), before finally turning to a discussion of the democratic implications. We argue that the result of the democratic creative destruction we describe is an explosion in the amount of public communication overall (from organized, self-interested actors as well as billions of individual end users) combined with a drastic reduction in the number of professional journalists and the emergence of a media environment where everyone can speak (resulting in a far more intense competition for attention and used for many and often ambiguous purposes) and where news media still serve as gatekeepers to the news agenda but are increasingly supplemented by platform companies serving as secondary gatekeepers in terms of reaching a wide audience. This is a media environment that challenges many established institutions, including news media, gives technology companies more institutional and infrastructural roles (and power), and in many ways empowers individual media users, making democracy relatively more demotic and popular, even as many people are also exposed to more abuse and harassment and many frequently use digital media in ambiguous ways, including ways that challenge established norms and values associated with liberal democracy.

Institutional Changes in the News Media

The ongoing move to a more digital, more mobile, and more platform-dominated media environment has made news more abundant and more accessible but existentially threatens the business models that funded professional news production in the twentieth century, resulting in a marked decline in news industry revenues and in newsroom employment, thus challenging incumbent institutions, even as the large technology companies that dominate this environment are growing rapidly and are gradually being institutionalized as they are forced to accept (and sometimes actively embrace) a wider range of formal and informal obligations than just that to their own bottom line and mission statement.

The destructive side of creative destruction is clear to see in the decline of legacy news media, the creative side too, with the rise of platform companies like Google, Facebook, and Twitter, the multitude of ways in which many different actors make use of digital technologies broadly and platform products and services specifically, and in how news media, both legacy and digital-born, are evolving and adapting to a new media environment.

Instead of tracing the fortunes of individual news media organizations, whether legacy ones like the New York Times or digital-born ones like BuzzFeed, we can consider the news media of a given country as an institution, a set of organizations generally seen within a society as presiding over a particular social sphere, in this case the production of news (Reference CookCook 1998, p. 70). Such institutions are defined by a set of cross-organizational formal and informal norms, routines, and procedures, in the case of news media broadly shared news values, workflows, and a common orientation toward the events, ideas, preoccupations, strategies, and politics of powerful officials, as well as a sense of professional ethics among journalists and an orientation toward public service. Not all news media are exactly alike, but they share many features, and not all (national) news institutions are identical (Reference Hallin and ManciniHallin and Mancini 2004). At least within high-income liberal democracies, however, they have much in common, including their integral role in politics – an institutional role, Timothy Cook (Reference Cook1998) has suggested, “without which … government could not act and could not work” (pp. 2–3) and without which individual citizens would struggle to stay informed about and engage with complex political processes far removed from personal experience.

News is thus intertwined with politics but also with the marketplace. In high-income democracies, news is a public good overwhelmingly produced by professional journalists working for private sector news media operating for-profit. Historically, news production has been funded in a variety of ways, including subsidies from political actors and media proprietors interested in power to pursue their own ends and through various forms of nonprofit models, backed by private philanthropists or politically mandated investment of public resources. Yet, in the second half of the twentieth century, news in North America and Western Europe was increasingly the province of journalists working for for-profit businesses based on selling content to audiences and selling audiences to advertisers (Reference HamiltonHamilton 2004). Even in a country like the United Kingdom, home to the license-fee funded public service BBC, an estimated 79 percent of investment in news production comes from for-profit private sector media companies (primarily newspapers), and, in the United States, where public media enjoy less political support, print publishers still account for the majority of reporters employed and broadcasters for another quarter, with online media and “information services” at just 10 percent (Reference NielsenNielsen 2019).

In the business of news, we see clearly the destructive side of creative destruction, especially among newspapers. Newspapers are central to the news institution and the business of news, but print has been in structural decline in many countries for decades as more and more different forms of media compete for attention and advertising. In a country like the United States, per capita print circulation has been declining for more than half a century, the number of daily newspapers dropping (especially dramatically with the disappearance of evening newspapers as television grew), and newspapers’ share of the overall advertising market shrinking in parallel (see Figure 7.1). Surviving newspapers remained profitable, especially because many were near-monopolies for some forms of local advertising in the growing number of one-newspaper towns. Yet newspapers were clearly becoming a less important part of the media overall.

Figure 7.1 US print newspaper circulation and advertising share

Data. 2012 US Census and Historical Statistics of the United States: From Colonial times to 1970. Douglas Galbi (on advertising). Note, last data point on advertising is 2007. Additional information provided by Reference NielsenNielsen (2019).

This long-term structural decline, accelerated in recent years by the rapid rise of digital media, has had dramatic implications for the business sustaining (and sometimes constraining) journalism. Newspapers provided the bulk of investment in news production in a media environment where individuals had low choice and news media high market power over advertisers (Reference HamiltonHamilton 2004). Yet we are increasingly moving to an environment where individuals have high choice and news media have low market power over advertisers (Reference NielsenNielsen 2019). In this environment, people are voting with their attention, and they are not voting for news. Precise figures are hard to come by, but one can roughly estimate that about 20 percent of the time people spend with print media is spent with news, 12 percent of the time spent with television is spent with news, and at most 7 percent of the time spent with digital media is spent with news.Footnote 1 (Consistent cross-platform measurements do not exist, and people often overreport their news consumption in surveys, so the share may well be even lower. For digital media specifically, one behavioral tracking source estimates news accounts for just 3 percent of time spent; see Reference HindmanHindman 2018.)

As attention is going elsewhere, advertising is following, especially as people increasingly embrace platform products and services like search, social media, messaging applications, and the like that offer cheap, targeted advertising at scale. This means large technology companies like Google and Facebook have quickly come to dominate digital advertising, capturing a majority globally, leaving literally tens of thousands of other media organizations, including the entire news industry, competing for the rest (see Figure 7.2 – note that some of Google’s and Facebook’s revenues are shared with third parties, including publishers).

Figure 7.2 Developments in digital advertising

Note. Both Google and Facebook share some of their advertising with partners through various revenue sharing arrangements.

Thus, advertising revenues that in the twentieth century helped fund content creation for a mass public increasingly help fund the provision of platform products and services to individual users and are tied in with pervasive data collection, especially by the dominant technology companies (Reference Turow and CouldryTurow and Couldry 2018).Footnote 2

From the point of view of billions of end users worldwide, as well as for millions of advertisers and other third parties like app developers, gaming companies, and the like, the rise of platforms is an example of the creative side of creative destruction – in high-income democracies, digital media now account for half or more of all media use and will come to account for a similar share of advertising – even as this rise is an existential challenge to the traditional business of news. In this changing environment, advertising alone will be a much smaller and much less lucrative part of the business of news than it has been in print and television. Figures from the United States in 2011 give a sense of how different the rates for digital advertising are from offline advertising, in terms of CPMs (“cost per mille,” the advertising cost per 1,000 views). That year, industry observers estimated a $60 average CPMs for print newspapers, $23 for primetime network television, between $3.50 and $2.50 for generic online advertising, $0.75 for mobile display advertising, and just $0.56 per 1,000 impressions on social networking sites (Reference NielsenNielsen 2019). The industry saying that analog dollars (100 cents) turn into digital dimes (10 cents) and mobile pennies (one cent) captures the stark differences in advertising rates between different media. As audiences and advertisers increasingly embrace digital media, the consequences are clear. In the United States alone, from 2008 to 2017, newspapers lost more than $20 billion in print advertising, close to half of total industry revenues, and cut newsroom employment by 45 percent. This decline of a net 32,000 professional journalists was far from offset by the growth in digital-born news media expanding their newsrooms from 7,000 to 13,000 in the same period (newsroom employment in radio and television was broadly stable) (Reference GriecoGrieco 2020). These business changes have little to do with politics directly but could have potentially profound political implications, as news organizations cut their newsrooms, in many cases specifically getting rid of specialist beat reporters and closing local and international bureaus in favor of focusing on general, national news.

These institutional changes are not only about a structural shift of attention and advertising away from news media and to content and practices hosted, enabled, and moderated by large technology companies (with consequent cuts in newsroom employment). They also involve changes to the operations of news media themselves, as well as in other actors’ and organizations’ ability to communicate with a wide audience. The rise of platforms is thus the most obvious creative side of this process of creative destruction, but not the only one. New institutions are emerging, but old institutions – including the news media but also governments, political parties, interest groups, social movements, and many others – are adapting to change and renewing themselves (Reference ChadwickChadwick 2017).

In the news media, the rise of digital media means that newsrooms frequently have to produce more news across more channels, at a faster pace than before, as a shrinking number of journalists publish to a website, multiple social media channels, and often a legacy media platform too (whether print or broadcast) (Reference Bell, Taylor, Brown, Hauka and RashidianBell et al 2017). Increasingly, inherited news values are supplemented with new editorial considerations focused on audience analytics, social media optimization, and search engine optimization as news media seek to build digital reach and compete for attention online. Especially new digital-born news media like BuzzFeed, HuffPost, and Vice have often aggressively embraced platforms and “worked towards the algorithms” to quickly build wide audiences via search engines and social media (Reference CaplanCaplan and boyd 2018). News media are thus actively embracing many of the products and services offered by some of their main competitors for attention and advertising, increasingly empowered by but also dependent on platform companies like Google and Facebook (Reference Nielsen and GanterNielsen and Ganter 2017) that provide them with everything from audience research over analytics to ad tech solutions (Reference Libert, Graves and NielsenLibert et al. 2018) – even as they also challenge the business of news and the relation news media have historically had with their audience. The often sudden and sometimes dramatic impact of changes implemented unilaterally by platform companies – a tweak to an algorithm, the discontinuation of a product or service – has revealed how news, like other forms of cultural production, is increasingly contingent on the infrastructure offered by platforms (Reference Nieborg and PoellNieborg and Poell 2018) and involve them in, for example, data-collection practices that some critics see as fundamentally illiberal in their disregard for traditional conception of privacy (Reference Glasius and MichaelsenGlasius and Michaelsen 2018).

In parallel, other organizations too have embraced digital media to communicate with the public. The PR cliché that “every organization is a media organization” is increasingly becoming a reality (Reference DavisDavis 2013). Whereas the news institution in the twentieth century was the single most important gatekeeper due to mass media’s control over the channels of communication, the rise of digital media (enabling virtually everybody with internet access to publish) combined with the rise of new secondary gatekeepers in the form of search engines, social media, and other platforms that drive discovery and structure attention online represents a new situation (Reference Nielsen and PetersNielsen 2016a; Reference Kreiss and McGregorKreiss and McGregor 2019). Everyone from activists, advocacy organizations, and think tanks, over political candidates, campaigns, and parties, to a multitude of private companies aggressively try to engage people in public affairs online. Even as the number of professional journalists employed by news organizations is in clear structural decline, the number of PR professionals employed by government agencies, interest groups, and corporations is on the rise, and far outstrips reporters (Reference DavisDavis 2013). Even if many of these forms of strategic communication often struggle to capture attention and engage people in effective and meaningful ways, some individual campaigns – from Barack Obama to Donald Trump, from the Tea Party to #MeToo, and, more darkly, in the form of extremist groups both foreign and domestic, as well as disinformation and information operations by states trying to undermine the democratic process and trust in institutions in other countries – have clearly managed to circumvent traditional gatekeepers and work with platform products and services to reach a wide public and in turn influence mainstream political debate (Reference ChadwickChadwick 2017).

Finally, of course, billions of individual users are embracing digital media, not only to get news but also to express themselves, connect, and build communities. In the next section, we examine their aggregate individual-level choices, but it is important to recognize that these choices also have an informal, institutional dimension that helps structure our media environment and thus impact our politics. The rapid, global embrace of digital media was initially cast in very positive terms by some who saw it as a dawn of a more “participatory culture” (Reference JenkinsJenkins 2006) superseding older forms of more passive mass media consumption, with trolling and the like as unfortunate and ultimately marginal aberrations. More extensive empirical research has challenged this benign interpretation and argued that, in deeply diverse and disputatious, irreducible plural societies, open and permissive systems are open to abuse, and many of the ways in which we use them are deeply ambivalent. Much of what some users may think of as forms of harmless “cultural play” online often involves deploying highly corrosive forms of speech, such as racist memes, sexist tropes, and the like that antagonize, silence, and marginalize people, especially historically disadvantaged communities including women and ethnic, religious, and sexual minorities (Reference Phillips and MilnerPhillips and Milner 2017). If the move to digital media and the increasing centrality of a few large technology companies to much of what we do with media point toward what José van Dijck (Reference Van Dijck2013) has called “platformed sociality,” it is important to recognize that these formal structures come accompanied with informal structures we might think of as a “culture of connectivity” and that this culture is deeply ambivalent in enabling communication, engagement, and community-building for a wide variety of different ends and with a wide variety of different valences, refracting the values and activities of the societies they are used in.

Major platform companies like Facebook and Google thus clearly already operate a number of products and services with infrastructural properties (they are shared, widely accessible systems that are used by a wide range of actors for many different purposes) (Reference Plantin, Lagoze, Edwards and SandvigPlantin et al. 2016), so far in ways primarily governed by inherited legal and regulatory frameworks and individual companies’ own sense of mission and terms of service (Reference GillespieGillespie 2018). Yet we are beginning to see the institutionalization of platform organizations through an increasingly heated public debate over what formal rules and informal norms should govern their operations as well as the creation of joint trade groups and various multi-stakeholder forums (Reference DeNardisDeNardis 2014). In line with Schumpeter, the economic change has in many ways already been revolutionary, even as many of the social changes are more evolutionary.

Individual-Level Changes in How We Engage with News

At the individual level, the ongoing move to a more digital, more mobile, more platform-dominated media environment involves a rapid move from direct access to news sources to a heavy reliance on distributed discovery via platform products and services like search engines and social media that provide us with new, supplementary ways of accessing, consuming, and engaging with news that broadly point to more diverse and sometimes participatory news use while also bringing with it widespread online harassment, exposure to disinformation, and the potential for more political polarization and social inequality in news use.

The creative side of this development is clear to see, with digital media enabling new ways of accessing, consuming, and engaging with news, including opportunities for individuals to express themselves, connect with others, and remix cultural products for a potentially large audience, leaving individuals simultaneously empowered by and dependent on the platform companies that provide many of the products and services we all increasingly rely on (Reference Van Dijckvan Dijck 2013).

Even if many fears are expressed, the precise nature of the destructive side of creative destruction is perhaps less easy to systematically substantiate, at least at scale. The move to digital media is accompanied by the relative decline of the kind of mass media experience associated in particular with mid-twentieth-century broadcasting, a shared, simultaneous experience of millions watching the same program – a single imagined community around a single news agenda (you can still watch the TV news, but it will never again mean what it once did), just as the limited share of attention going to news online raises the prospect of a marginalization of public affairs in favor of private concerns and entertainment. Yet many of the most frequently articulated fears concerning the democratic implications of our changing media environment are not supported by empirical research. As we will show in this section, and as Pablo Barberá documents in his contribution to this volume (Chapter 3), fears over, for example, audience fragmentation and filter bubbles remain unsubstantiated and are frequently complicated or flat-out contradicted by empirical evidence. Instead, destructive impacts may come in the form of sustained online harassment especially of vulnerable communities (which in turn may silence them and reduce their opportunities to participate in public life) and exposure to various forms of online disinformation, as well as the potential for greater political polarization and social inequality in news use.

The shift toward online news consumption is clear and visible in every high-income country. As the number of people who use offline media for news falls, and online news consumption grows, the centrality of each medium also changes, albeit gradually. Printed news consumption has declined to such a point that in 2018, across thirty-seven markets globally we surveyed, only around 10 percent of internet users regarded it as their main source of news (Reference Newman, Fletcher, Kalogeropoulos, Levy and NielsenNewman et al. 2018). Television is still the most widely used news source in many countries, but its importance is waning. In the same survey, around 40 percent say television is their main source of news, and the audiences for most television news bulletins are both shrinking and aging. As especially print and increasingly television decline, online news use is growing. In 2018, around 45 percent say that it is their main source of news. In contrast to the television news audience, the online news audience skews younger (see Figure 7.3). Two-thirds (67 percent) of respondents between eighteen and twenty-four now say online is their main source of news, compared to just a quarter (24 percent) who name television. Social media alone – just one of the ways of getting news online – is more important for news for the youngest age group than television. With generational replacement, the shift toward the kind of digital-, mobile-, and platforms-first environment that most people below the age of thirty-five already live in (especially in high-income, highly connected societies) will only continue.

Figure 7.3 Main source of news, by age group

Note. Respondents across thirty-seven markets were asked: “You say you’ve used these sources of news in the last week, which would you say is your MAIN source of news?” Base: 72,192.

Of course, many of the most popular online news sources in most countries are legacy news brands that have their roots in broadcasting or print publishing – so the destruction only extends so far. Even as older channels of communication become less important, organizations built around them may still renew themselves and find a digital future (Reference ChadwickChadwick 2017). The creative side, for individual users, is apparent. The shift toward digital news use not only has led to a massive increase in the number of available news sources both old and new (and thus massive increase in different types of coverage, perspectives, and styles that people have access to) but, perhaps more importantly, has enabled a change from an almost complete reliance on direct access to an increased reliance on distributed discovery via search engines, social media, and news aggregators. In practice, people arrive at news in many different ways, but our survey research shows that just one-third of people said that their main way of arriving at news was by going directly to the websites and apps of news publishers like the BBC or the New York Times (Reference Newman, Fletcher, Kalogeropoulos, Levy and NielsenNewman et al. 2018). The remaining two-thirds say that their main way of arriving at news was via search (24 percent), social media (23 percent), or email, mobile alerts, or news aggregators (6 percent each) (see Figure 7.4).

Figure 7.4 Main way of accessing news online

Note. Respondents across thirty-seven markets were asked: “Which of these was the MAIN way in which you came across news in the last week?” Base: 69,246.

This is an epochal shift on how news is distributed and curated. Direct discovery has been a defining feature of the mass media environment in the twentieth century, but the twenty-first-century digital media environment is increasingly defined by algorithmically based forms of personalization, as people rely on products and services like search engines and social media that do not create content but help users discover content. Reliance on distributed discovery is even higher among the under thirty-fives, but there is considerable national variation. For example, in the Nordics and other countries characterized by democratic corporatist media systems (Reference Hallin and ManciniHallin and Mancini 2004), where loyalty to news brands is stronger and trust in the news media is higher, the importance of direct access has remained high. In much of Southern Europe and other polarized pluralist countries, however, where the opposite is generally true, distributed discovery is clearly more widely used.

The difference between direct discovery and distributed discovery is key to many of the fears concerning the democratic implications of digital media, especially the fear that algorithmic personalization through search engines, social media, and the like will lead to audience fragmentation and the creation of filter bubbles. It is worth considering briefly how direct versus distributed forms of discovery can result in patterns of news use that are potentially democratically problematic.

Take direct discovery first, where people go directly to a preferred source of news. This form of discovery gives the individual user full, active control. Because it is not possible to consume all of the news that is available online (even from just a small pool of the most popular sources), those who primarily go to news sources directly have to decide where to get their news online. If they want to directly consume news from a particular source, they have to know it exists and then make a conscious decision to either type the web address into their browser or download the app onto their smartphone. Such choices are often characterized by what communications researchers and psychologists call “selective exposure,” where people tend to favor sources that reflect their preexisting beliefs and interests while avoiding sources that might contradict them. (We use terms like “choice” and “selective” in a broad sense here; much media use is habitual and not necessarily based on deliberate, discrete decisions.) These dynamics will vary from context to context and country to country, depending on people’s dispositions. In terms of beliefs, for example, many US media users engage in partisan selective exposure, where people are more likely to select news sources that produce content aligned with their political views (Reference Iyengar and HahnIyengar and Hahn 2009; Reference StroudStroud 2011). In parallel, marked variation in how interested people are in news (relative to alternatives like entertainment) also shapes the media habits of the large number of Americans who are not very interested in politics. Looking specifically at direct discovery across offline media like print and broadcast and online media like going to websites and apps, this combination leaves the United States with a hyper-partisan and engaged subset of people who are avid news consumers (and some of whom live in self-contained “echo chambers” mostly consuming news from attitude-consistent sources) while a much larger group of Americans pay only sporadic attention to political news (Reference PriorPrior 2007). In contrast, empirical research from parts of Europe suggests that, in other contexts, levels of interests matter more than partisanship (Reference Skovsgaard, Shehata and StrömbäckSkovsgaard, Shehata, and Strömbäck 2016; Reference Castro-Herrero, Nir and SkovsgaardCastro-Herrero, Nir, and Skovsgaard 2018). The deeper issue here is that levels of interest in the news are likely to be reflected in other socioeconomic inequalities between different groups (Reference Shehata and AmnåShehata and Amnå 2017; Reference Kalogeropoulos and NielsenKalogeropoulos and Nielsen 2018), with important knock-on consequences for their news consumption.

Distributed discovery through search engines, social media, and other platform products and services that rely at least in part on various forms of algorithmically curated personalized recommendation could potentially lead to similar outcomes, increased partisan polarization, and growing inequality between news lovers and a less interested majority as automated systems feed us more of what we click on and less of everything else. In these environments, our individual choices are increasingly structured for us through various forms of algorithmic filtering and recommendation. This development has often been associated with the idea of “filter bubbles” (Reference PariserPariser 2011), whether in the form of audience fragmentation or narrowed news diets. So far, however, empirical research has not substantiated these concerns and has, in fact, often found that digital news and media use work very differently from how it is often assumed, or asserted, to work (again, see Barberá, Chapter 3, this volume).

First, despite the explosive growth in the number of sources available and the gradual erosion of many traditional mass audiences, empirical research suggests that the audiences of the most popular news outlets tend to overlap with one another (Reference Webster and KsiazekWebster and Ksiazek 2012; Reference Fletcher and NielsenFletcher and Nielsen 2017). Furthermore, when comparing offline environments (completely reliant on direct discovery and thus selective exposure) with online environments (where distributed discovery is far more important), online news audiences are in fact frequently less fragmented and more overlapping than their offline counterparts (Reference Fletcher and NielsenFletcher and Nielsen 2017), perhaps in part because search engines and social media often lead people to broadly used, widely shared major news brands.

Second, in contrast to the concern that platform products and services would be associated with a narrowing of news diets where people are only fed more of the same, in fact most empirical research finds use of social media and search engines associated with more diverse news use (Reference Flaxman, Goel and RaoFlaxman, Goel, and Rao 2016; Reference Dubois and BlankDubois and Blank 2018; Reference Fletcher and NielsenFletcher and Nielsen 2018a).

Our own research has suggested two central causal mechanisms that supplement the tendency for forms of selective exposure that dominate direct discovery (often in ways that may point to polarization and inequality, as well as fragmentation and echo chambers), namely incidental exposure and automated serendipity. Incidental exposure refers to situations where people use social media or video sharing services for reasons that have nothing to do with news (staying in touch with friends, sharing things, entertainment) and in the process are incidentally exposed to sources of news they would not have come across otherwise. We have documented how incidental exposure leads to significantly more diverse news diets across different platforms like Facebook, Twitter, and YouTube (Reference Fletcher and NielsenFletcher and Nielsen 2018a). Automated serendipity refers to how people who use search engines to access news in the process are led to more and different sources of news than those they access by going directly, and we have demonstrated empirically how people who use search engines for news both are likely to consume news from a left-leaning and right-leaning news source and have more politically balanced news repertoires than people who do not use search engines for news (Reference Fletcher and NielsenFletcher and Nielsen 2018b).

This does not mean that echo chambers and filter bubbles do not exist in some form, or that people who are heavily reliant on distributed discovery have perfectly diverse and healthy news diets.Footnote 3 Yet we need to keep in mind that direct access and distributed discovery both create personalization, either because people personalize their own news consumption or because an algorithm does it for them (Reference Zuiderveen Borgesius, Trilling, Möller, Bodó, de Vreese and HelbergerZuiderveen Borgesius et al. 2016). It appears that the self-selected personalization that results from selective exposure is more powerful than the preselected exposure that results from algorithmic selection, meaning that people who rely on distributed discovery are left with larger and more diverse news diets.

Whether people trust these news sources is a separate matter. In parallel to the growth of distributed discovery, we have also seen steady year-on-year declines in trust in the news media. Despite being stable in most countries from 1980 to 2010 according to the World Values Survey (Reference Hanitzsch, van Dalen and SteindlHanitzsch, van Dalen, and Steindl 2018), more recent data from 2015 onward suggest that trust in the news media is falling in both traditionally low-trust countries like France (down 14 percentage points since 2015) and traditionally high-trust countries like Germany (down 13 percentage points since 2015) (Reference Newman, Fletcher, Kalogeropoulos and NielsenNewman et al. 2019). In most countries, trust in news from search engines and social media is significantly lower than trust in the news media as a whole (Reference Newman, Fletcher, Kalogeropoulos and NielsenNewman et al. 2019), but at the moment it is far from clear that the move to distributed discovery is to blame for the overall declines. Indeed, when it comes to evaluating different types of news selection, people have what we call “generalized skepticism,” where those skeptical of primary gatekeepers (publishers) are often skeptical of secondary gatekeepers (platforms) too – and usually for similar reasons. People often do not understand how editorial processes, let alone algorithmic news selection, work, but that does not mean they uncritically accept what they come across (Reference NielsenFletcher and Nielsen 2019).

In parallel with the move from direct to distributed discovery, we have seen the move to a digital media environment that affords people with more opportunities for more participatory forms of news and media use, in the process also exposing many to widespread online harassment and potentially various forms of disinformation disseminated online and especially via platforms. Digital media offers everyone with internet access a range of both “web 1.0” and “web 2.0” ways of engaging in more participatory forms of news and media use, ranging from commenting on news sites and sharing via email to commenting and/or sharing via social media sites. While all internet users have access to this participatory potential, it is important to recognize that the majority does not in practice use it, though a large minority does, relying especially on social media (Reference Newman, Fletcher, Kalogeropoulos, Levy and NielsenNewman et al. 2018). Looking more closely at who comments and shares news online, across web 1.0 and social media, in most countries, the most active tend to be those who have most enthusiastically embraced a wide range of different social media platforms and often use them for many other purposes than news, specifically political partisans and people with high levels of interest in news (Reference Kalogeropoulos, Negredo, Picone and NielsenKalogeropoulos et al. 2017). Unlike other forms of political participation, which tend to skew toward older men (Reference Verba, Schlozman and BradyVerba et al. 1995), all these forms of participation are as widespread among the young and among women (Reference Valeriani and VaccariValeriani and Vaccari 2016; Reference Kalogeropoulos, Negredo, Picone and NielsenKalogeropoulos et al. 2017).

Not all forms of participation, of course, are equally benign or pro-social. The Internet, and the use we make of it, is profoundly ambiguous. Various forms of online harassment and trolling once thought to be relatively marginal and subcultural phenomena are now mainstream and widely experienced, enabled by digital technologies but fed by culturally sanctioned impulses and often amplified by political actors as well as the corporate interest of some publishers and platform companies (Reference PhillipsPhillips 2015). The scale and scope of these issues will differ from country to country and are experienced very differently by different groups (especially along lines of gender, race, and religion), but survey research from the United States can serve to illustrate how widespread they are (Reference DugganDuggan 2017). In 2017, 41 percent of Americans said they had personally been subjected to online harassment, and 66 percent has witnessed such behaviors directed at others. (These numbers are higher than the percentage of people who use digital media to actively comment on or share news online.) Almost one in five (18 percent) reported having been subjected to particularly severe forms of online harassment, for example, physical threats, harassment over a sustained period, sexual harassment, or stalking. Different groups are subject to different kinds of harassment – while only 3 percent of white Americans report having been subject to racial harassment online, one-in-four black Americans have. Similarly, while only 5 percent of men say they have been harassed online because of their gender, 11 percent of women have. Strikingly, the ambivalence about what exactly constitutes harassment is not exclusively on the side of the perpetrators but also clear among the victims. Twenty-eight percent of those whose most recent encounter with online harassment involved severe types of abusive behavior – such as stalking, sexual harassment, sustained harassment, or physical threats – answered in the 2017 survey that they do not think of their own experience as constituting “online harassment.” Yet, however defined, such widespread, sustained, and unequally distributed intimidation will lead some to take a less active part in online public life than they would otherwise want to, a dynamic only further compounded in political contexts where people may feel reluctant to discuss news openly or share their political views for fear of social or other repercussions (Reference Newman, Fletcher, Kalogeropoulos, Levy and NielsenNewman et al. 2018) or where some political leaders amplify and encourage various forms of harassment (see Siegel, Chapter 4, this volume.)

Other forms of problematic online participation may take the form of good-faith, bottom-up dissemination of false or potentially harmful information online by ordinary users or, more nefariously, foreign or domestic extremism and organized for-profit or political disinformation campaigns. The same digital media that have dramatically lowered the barriers to entry and allowed everyone to publish online, the same platform products and services that demonstrably lead people to more diverse news and enable more participatory forms of news and media use also serve to disseminate various forms of misinformation and disinformation (Reference Wardle and DerakhshanWardle and Derakhshan 2017). The risks are great, and we are only beginning to see significant amounts of empirical research in this area (see Guess and Lyons, Chapter 2, this volume). On the one hand, there are clearly real issues around bottom-up disinformation shared among users, for-profit disinformation from various unscrupulous low-quality publishers, and, perhaps most worryingly, various organized disinformation campaigns by both domestic and foreign political actors; it is also worth noting that the best available evidence from the United States currently suggests that overall exposure to much of this content is low, fleeting, and largely confined to those who already hold partisan views (Reference Guess, Nyhan and ReiflerGuess, Nyhan, and Reifler 2018; Reference Nelson and TanejaNelson and Taneja 2018). Digital media has undoubtedly provided a structure for large-scale misinformation and disinformation to exist, and very real and very worrying attempts to abuse these structures are clearly afoot, but this does not mean that there will inevitably be dramatic democratic consequences in terms of individuals’ attitudes and behaviors.

What Are the Main Implications of This Process of “Democratic Creative Destruction”?

It is clear that democracy has historically evolved in part through creative destruction in the media environment. Think, for example, of broadcasting – starting with radio, then terrestrial television, then post-broadcast television. Similarly, democracy has evolved through other dramatic societal changes, from institutional changes in party systems (Reference MairMair 1997) and forms of governance (Reference RhodesRhodes 1997) to changes in individual-level aggregate levels of trust in institutions (Reference NorrisNorris 2011) and in people’s value systems (Reference InglehartIngelhart 1997). So far, we have described more recent creative destruction in the media at both the institutional and the individual level, but what is the impact of this process of creative destruction on democracy?

This of course depends on what one means by democracy. Historically, communications researchers have offered a range of different theoretical conceptions based on normative political philosophy, including ideal models of procedural democracy, competitive democracy, participatory democracy, and deliberative democracy (see, e.g., Reference StrömbäckStrömbäck 2005). Increasingly, those interested in discussing the democratic implications of digital media are drawing on similar theoretical models, asking what we would want algorithmic curation to look like from different idealistic starting points, whether participatory or deliberative (Reference HelbergerHelberger 2019). These are important discussions but also often rather abstract and a bit removed from the hustle and bustle of actually existing and highly imperfect democratic politics, so we will instead follow Joseph Schumpeter a step further and focus on the implications in terms of a fairly minimal conception of democracy as liberal, representative, and based on relatively diverse and inclusive elite competition that offers a basic structure for making and implementing political decisions but no guarantee that these decisions will be “good” ones that ultimately ensure peace, prosperity, and personal fulfillment (Reference Nielsen, Peters and BroersmaNielsen 2016b). If we start with this more minimalist view of democracy, and consider the changes we have summarized, the systemic implications for democracy in our view are likely to (1) be often indirect and institutional through an (incessant) revolution of the institutions that enable popular government (news media but also, by extension, interdependent other institutions like political parties, etc.), (2) be limited in terms of short-term impact on people’s attitudes and behaviors, and (c) leave most individuals and many institutions simultaneously increasingly empowered by and increasingly dependent on a small number of large, for-profit, US-based platform companies like Facebook and Google.

At the institutional level, many of the institutions that animated twentieth-century democracy, including political parties, member-based interest groups, as well as the news media itself, are seriously challenged by the rise of digital media and platform companies. Some adapt and others implode, even as new actors that have for some time enjoyed power with limited responsibility are increasingly (and imperfectly) being institutionalized and bound by both informal norms and formal rules.Footnote 4 As individuals, we are empowered by digital media even as we are also becoming dependent on them, exposed to more diverse news, offered new potential for participation, but also exposed to online harassment and disinformation, as the same platforms that enable us to find or share baby photos, cat videos, and dad jokes are integral to everything from #MeToo to populist campaigns by domestic politicians and attempts by foreign states to interfere with our elections. Advertising spending that used to support the private (for-profit) provision of news as a public good now increasingly support the private (for-profit) provision of platform products and services we use to access information and connect with one another.

This diagnosis of democratic creative destruction moves us away from more utopian early visions about the impact of online digital media on democracy but also departs from the newer, more dystopian visions that see the impact of online digital technologies on democracy as profoundly negative. We might call this position “complex realism,” in recognition that the impact on democracy is neither fantastically benign nor completely disastrous – but profound, complicated, varied, and contextual. Much depends on the transition of old institutions and the institutionalization of new actors; and the political and social consequences are fundamentally political and social questions at least as much as they are about business and technology. Thus, for us as people, the move a more digital, more mobile, and more platform-dominated media environment may point to a more popular democracy, democracy with more focus on the demotic demos- part of demos-kratos, even as some of the institutions that are central to the -kratos are challenged and new institutions developing. For publishers, and many other twentieth-century legacy institutions integral to our democracies, there is still an important but more precarious role. For platforms, commercial success and rapid growth in use come with great complexity, enormous responsibility, and increasingly intense political and public scrutiny. For public authorities, there are questions about how they can ensure the protection of fundamental rights (free expression, privacy, etc.), the existence of an enabling environment for citizens and for news, while protecting the integrity of public debate and political processes in a changing media environment increasingly influenced by a few very large US-based technology companies.

Based on current trends, we can imagine three consequences specifically for the news media’s role in democracy that could result from the huge changes to how it is financed: (1) restoration, where the news media finds a way to fund itself that allows it to play the same democratic role, even as new challenges arise (such as the concern that the move to pay models may exacerbate information inequalities); (2) renewal, where the news media finds a way to fund itself to play a different democratic role, either because it now longer has the resources to play the previous role or because the wider context has changed to such a degree that a new role is required; or (3) retreat, where the news media is able to fund itself but cannot play any meaningful democratic role at the level of mass politics (e.g., beyond the role played by high culture). Some of these paths would clearly be better for democracy than others. Beyond this it is vital that we avoid a fourth scenario we might call (4) relegation, where the vestiges of media power are co-opted and relegated to serve various political or commercial ends that have little to do with news (Reference SchiffrinSchiffrin 2017) – a form of instrumentalization common in history and seen all over the world today too where media are weak. This will probably be bad for democracy and bad for almost everyone else. We as individuals would have plenty of choice, and many platform products and services, but what we would have to choose from would be captured media focused on manipulating us for ulterior ends, not independent news media serving the public, however imperfectly.

How we use digital media is up to us individually and collectively (even if we do not use them under conditions of our own choosing). In deeply divided, unequal, and disputatious societies there are no guarantees we will use them in ways that various elites or parts of the establishment will consider democratic or that we will all find substantially benign or to our liking – but it will almost certainly be more demotic, popular, often diverse, and sometimes majoritarian (in addition to being incredibly profitable for the most successful platform companies). This interplay is a classic tension between, on the one hand, liberal values, norms, and established institutions that many see as key to democracy and, on the other hand, popular sentiments and forms of representation that sometimes challenge these values, norms, and institutions. The challenge to the norms, values, and institutions of the -kratos is real and serious, especially so in countries where many of these seem fragile, but the opportunities for many people, around the popular, demotic, demos- are real too, and unease with specific political outcomes, potentially disastrous as they may be in other ways, should not lead people to jump to conclusions about whether current changes in our media environment are fundamentally antidemocratic. We may be moving toward a version of democracy where traditional elites, the media, and others parts of the establishment are relatively disempowered and ordinary people are empowered (as are a few for-profit platforms that urgently need to be institutionalized). Those experiencing a sense of relative disempowerment may be surprised and frustrated when the democratic process does not produce the outcomes they want. Whatever each of us individually think about these outcomes substantially, even if they may be malign, that does not make them undemocratic. Liberal representative democracy, minimally conceived, makes no guarantees about outcomes. We may have to accept that politics and the media can be both more democratic (at least in the sense of being more demotic and more popular) and bad for many people, particularly parts of established elites as well as for various vulnerable groups exposed to newly empowered majority groups.

Schumpeter thought the process of (economic) creative destruction would lead to the ultimate collapse of capitalism. We are less pessimistic about democratic creative destruction and more hopeful we will be able to renew and develop institutions in the future that hold powerful platforms to account (as news media have been, imperfectly, in the past) and also enable an empowered public of individual citizens to take a meaningful part in self-government in the future. In democracies, elites and establishments come and go, and that is part of the point. The question is whether people play a substantive role in that process. We believe they have, can, and will.


Bell, E. J., Taylor, O., Brown, P. D., Hauka, C., & Rashidian, N. (2017). The Platform Press: How Silicon Valley Reengineered Journalism. Tow Center for Digital Journalism, Columbia University report.
Besley, T., & Prat, A. (2006). Handcuffs for the grabbing hand? Media capture and government accountability. The American Economic Review, 96(3), 720736.
Bimber, B. A. (2003). Information and American Democracy: Technology in the Evolution of Political Power. Cambridge: Cambridge University Press.
Brunetti, A., & Weder, B. (2003). A free press is bad news for corruption. Journal of Public Economics, 87(7–8), 18011824.
Caplan, R., & boyd, d. (2018). Isomorphism through algorithms: Institutional dependencies in the case of Facebook. Big Data & Society, 5(1).
Castro-Herrero, L., Nir, L., & Skovsgaard, M. (2018). Bridging gaps in cross-cutting media exposure: The role of public service broadcasting. Political Communication, 35(4), 542565.
Chadwick, A. (2017). The Hybrid Media System: Politics and Power (2nd ed.). New York: Oxford University Press.
Cook, T. E. (1998). Governing with the News: The News Media As a Political Institution. Chicago: University of Chicago Press.
Couldry, N., Livingstone, S. M., & Markham, T. (2010). Media Consumption and Public Engagement: Beyond the Presumption of Attention (rev. ed.). Basingstoke: Palgrave Macmillan.
Curran, J., Iyengar, S., Lund, A. B., & Salovaara-Moring, I. (2009). Media system, public knowledge and democracy. European Journal of Communication, 24(1), 526.
Davis, A. (2013). Promotional Cultures: The Rise and Spread of Advertising, Public Relations, Marketing and Branding. Cambridge: Polity Press.
DeNardis, L. (2014). The Global War for Internet Governance. New Haven, CT: Yale University Press.
Dubois, E., & Blank, G. (2018). The echo chamber is overstated: The moderating effect of political interest and diverse media. Information, Communication & Society, 21(5), 729745.
Duggan, M. (2017). Online harassment 2017. Pew Research Center, July 10.
Flaxman, S., Goel, S., & Rao, J. M. (2016). Filter bubbles, echo chambers, and online news consumption. Public Opinion Quarterly, 80 (Special Issue), 298320.
Fletcher, R., & Nielsen, R. K. (2017). Are news audiences increasingly fragmented? A cross‐national comparative analysis of cross‐platform news audience fragmentation and duplication. Journal of Communication, 67(4), 476498.
Fletcher, R., & Nielsen, R. K. (2018a). Are people incidentally exposed to news on social media? A comparative analysis. New Media & Society, 20(7), 24502468.
Fletcher, R., & Nielsen, R. K. (2018b). Automated Serendipity: The Effect of Using Search Engines on the Diversity and Balance of News Repertoires. Digital Journalism, 8(6), 976989.
Fletcher, R., & Nielsen, R. K. (2019). Generalised scepticism: How people navigate news on social media. Information, Communication & Society, 22(12), 17511769.
Gans, H. J. (2003). Democracy and the News. Oxford: Oxford University Press.
Gillespie, T. (2018). Custodians of the Internet: Platforms, Content Moderation, and the Hidden Decisions That Shape Social Media. New Haven, CT: Yale University Press.
Glasius, M., & Michaelsen, M. (2018). Authoritarian practices in the digital age| Illiberal and authoritarian practices in the digital sphere. International Journal of Communication, 12, 37953813.
Grieco, E. (2020). U.S. newspapers have shed half of their newsroom employees since 2008. Pew Research Center, April 20.
Guess, A., Nyhan, B., & Reifler, J. (2018). Selective exposure to misinformation: Evidence from the consumption of fake news during the 2016 US presidential campaign.
Hallin, D. C., & Mancini, P. (2004). Comparing Media Systems: Three Models of Media and Politics. Cambridge: Cambridge University Press.
Hamilton, J. (2004). All the News That’s Fit to Sell: How the Market Transforms Information into News. Princeton: Princeton University Press.
Hanitzsch, T., van Dalen, A., & Steindl, N. (2018). Caught in the nexus: A comparative and longitudinal analysis of public trust in the press. The International Journal of Press/Politics, 23(1), 323.
Helberger, N. (2019). On the democratic role of news recommenders. Digital Journalism, 7(8), 9931012.
Hindman, M. (2018). The Internet Trap: How the Digital Economy Builds Monopolies and Undermines Democracy. Princeton: Princeton University Press.
Inglehart, R. (1997). Modernization and Postmodernization: Cultural, Economic, and Political Change in 43 Societies. Princeton: Princeton University Press.
Iyengar, S., & Hahn, K. S. (2009). Red media, blue media: Evidence of ideological selectivity in media use. Journal of Communication, 59(1), 1939.
Jenkins, H. (2006). Convergence Culture: Where Old and New Media Collide. New York: New York University Press.
Kalogeropoulos, A., Negredo, S., Picone, I., & Nielsen, R. K. (2017). Who shares and comments on news? A cross-national comparative analysis of online and social media participation. Social Media & Society, 3(4).
Kalogeropoulos, A., & Nielsen, R. K. (2018). Social Inequalities in News Consumption. Reuters Institute factsheet.
Kim, Y. M., Hsu, J., Neiman, D. et al. (2018). The stealth media? Groups and targets behind divisive issue campaigns on Facebook. Political Communication, 35(4), 515541.
Kreiss, D., & McGregor, S. C. (2019). The “arbiters of what our voters see”: Facebook and Google’s struggle with policy, process, and enforcement around political advertising. Political Communication, 36(4), 499522.
Libert, T., Graves, L., & Nielsen, R. K. (2018). Changes in Third-Party Content on European News Websites after GDPR. Reuters Institute factsheet.
Mair, P. (1997). Party System Change: Approaches and Interpretations. Oxford: Clarendon Press.
Nelson, J. L., & Taneja, H. (2018). The small, disloyal fake news audience: The role of audience availability in fake news consumption. New Media & Society, 20(10), 37203737.
Newman, N., Fletcher, R., Kalogeropoulos, A., Levy, D. A. L., & Nielsen, R. K. (2018). Reuters Institute Digital News Report 2018. Reuters Institute.
Newman, N., Fletcher, R., Kalogeropoulos, A., & Nielsen, R. K. (2019). Reuters Institute Digital News Report 2019. Reuters Institute.
Nieborg, D. B., & Poell, T. (2018). The platformization of cultural production: Theorizing the contingent cultural commodity. New Media & Society, 20(11), 42754292.
Nielsen, R. K. (2016a). Democracy. In B. Peters (Ed.), Digital Keywords: A Vocabulary of Information Society and Culture (pp. 8192) (Princeton Studies in Culture/Power/History). Princeton: Princeton University Press.
Nielsen, R. K. (2016b). News media, search engines and social networking sites as varieties of online gatekeepers. In C. Peters & M. Broersma (Eds.), Rethinking Journalism Again: Societal Role and Public Relevance in a Digital Age (pp. 8196). London: Routledge.
Nielsen, R. K. (2019). The changing economic contexts of journalism. In T. Hanitzsch & K. Wahl-Jorgensen (Eds.), Handbook of Journalism Studies.
Nielsen, R. K., & Ganter, S. A. (2017). Dealing with digital intermediaries: A case study of the relations between publishers and platforms. New Media & Society, 20(4), 16001617.
Noble, S. U. (2018). Algorithms of Oppression: How Search Engines Reinforce Racism. New York: University Press.
Norris, P. (2011). Democratic Deficit: Critical Citizens Revisited. Cambridge: Cambridge University Press.
Pariser, E. (2011). Filter Bubbles: What the Internet Is Hiding from You. London: Penguin.
Phillips, W. (2015). This Is Why We Can’t Have Nice Things: Mapping the Relationship between Online Trolling and Mainstream Culture (Information Society Series). Cambridge, MA: MIT Press.
Phillips, W., & Milner, R. M. (2017). The Ambivalent Internet: Mischief, Oddity, and Antagonism Online. Cambridge: Polity Press.
Plantin, J.-C., Lagoze, C., Edwards, P. N., & Sandvig, C. (2016). Infrastructure studies meet platform studies in the age of Google and Facebook. New Media & Society, 20(1), 293310.
Prior, M. (2007). Post-Broadcast Democracy: How Media Choice Increases Inequality in Political Involvement and Polarizes Elections. Cambridge: Cambridge University Press.
Rhodes, R. A. W. (1997). Understanding Governance: Policy Networks, Governance, Reflexivity and Accountability. Bristol: Open University Press.
Schiffrin, A. (Ed.) (2017). In the Service of Power: Media Capture and the Threat to Democracy. Washington, DC: Center for International Media Assistance.
Schumpeter, J. A. (1992). Capitalism, Socialism, and Democracy (5th ed.). London: Routledge.
Shehata, A., & Amnå, E. (2017). The development of political interest among adolescents: A communication mediation approach using five waves of panel data. Communication Research, 46(8), 10551077.
Skovsgaard, M., Shehata, A., & Strömbäck, J. (2016). Opportunity structures for selective exposure: Investigating selective exposure and learning in Swedish election campaigns using panel survey data. International Journal of Press/Politics, 21(4), 527546.
Snyder, J. M., & Strömberg, D. (2008). Press coverage and political accountability. NBER Working Paper No. 13878.
Strömbäck, J. (2005). In search of a standard: Four models of democracy and their normative implications for journalism. Journalism Studies, 6(3), 331345.
Stroud, N. J. (2011). Niche News: The Politics of News Choice. Oxford: Oxford University Press.
Turow, J., & Couldry, N. (2018). Media as data extraction: Towards a new map of a transformed communications field. Journal of Communication, 68(2), 415423.
Valeriani, A., & Vaccari, C. (2016). Accidental exposure to politics on social media as online participation equalizer in Germany, Italy, and the United Kingdom. New Media & Society, 18(9), 18571874.
Van Dijck, J. (2013). The Culture of Connectivity: A Critical History of Social Media. Oxford: Oxford University Press.
Van Zoonen, L. (1998). A day at the zoo: Political communication, pigs and popular culture. Media, Culture & Society, 20(2), 183200.
Verba, S., Schlozman, K. L., & Brady, H. E. (1995). Voice and Equality: Civic Voluntarism in American Politics. Cambridge, MA: Harvard University Press.
Wardle, C., & Derakhshan, H. (2017). Information Disorder: Toward an Interdisciplinary Framework for Research and Policy Making. Council of Europe Report No. DGI(2017)09).
Webster, J. G. (2014). The Marketplace of Attention: How Audiences Take Shape in a Digital Age. Cambridge, MA: MIT Press.
Webster, J. G., & Ksiazek, T. B. (2012). The dynamics of audience fragmentation: Public attention in an age of digital media. Journal of Communication, 62(1), 3956.
Zuiderveen Borgesius, F. J., Trilling, D., Möller, J., Bodó, B., de Vreese, C. H., & Helberger, N. (2016). Should we worry about filter bubbles? Internet Policy Review, 5(1), 116.

2 Our focus here is on news but it is worth mentioning that such cheap, targeted, and tailored digital advertising is of course also increasingly important for political communication, whether candidate campaigns, interest groups, political parties, and the like (Reference Kreiss and McGregorKreiss and McGregor 2019) or more problematic actors like astroturf groups, and fundamentally unidentifiable groups, including foreign entities (see, e.g., Reference Kim, Hsu and NeimanKim et al. 2018).

3 Beyond the specific issues of news diversity that we examine here, algorithmic and automated ranking systems can embed various forms of discrimination and reinforce oppressive social relations as, for example, Reference NobleNoble (2018) shows.

4 Importantly, the platforms that increasingly dominate media markets in most democracies are all US-based. They often seem and are seen as primarily responsive to American political and public pressures and disinclined to customize their content moderation practices, terms of service, policies, and so on to other countries.