The following essay summarizes the state of research in spring 2018 and is only partly updated since them. A shortend version of the essay was published in the context of the Aspen Transatlantic Workshop “Present at the New Creation? Tech. Power. Democracy.” in Autumn 2018.
The question of whether and how the Internet is shaping democracy is one of the oldest debates surrounding the rise in importance of the Internet. Authors like Castells (2002) celebrated the egalitarian access to the Internet as a catalyst for democracy, predicting an uprising of the people against incumbent elites and institutions. The Internet was seen as the place where new political movements would coalesce and shake off the bonds of an entrenched and outdated system. However, recent discussions around new media like Facebook and Twitter are much more pessimistic. Characterized by major political events like the “Brexit” of the United Kingdom and the election of Donald Trump as president of the US, 2016 was called a turning point for western democracy (FAZ 2017). The Internet and its new platforms are at least partly blamed for the election outcomes in 2016. Media phenomena like echo chambers, political disinformation and psychological microtargeting are considered as core threats for democracy (Wired 2016).
This essay investigates the research question, “Have social media and information technologies had an impact on elections and referenda?” As the scope of this essay is limited, the question will be narrowed down in the following way. Three topics surrounding the impact of the Internet on democratic elections—namely, the Internet as a field for new political actors; Internet media and political disinformation; and finally Internet media and echo chambers—are analysed. All three examples received significant attention in public and academic circles in the last years and can therefore be considered highly relevant to the current state of democracy. The phenomena will be examined in terms of their (potential) impact on two elections in the recent years: the 2016 US presidential election and the 2017 federal election in Germany. These countries are chosen as they are both highly developed western democracies, but with significantly different political and media systems. The essay follows Chadwick’s approach (Chadwick 2006) to discuss the influence of the Internet and new media on the political processes, while acknowledging significant differences between political systems and avoiding technologically deterministic argumentation. The combination of three phenomena in two different countries allows us to examine the question from different standpoints and with different perspectives. At this point it is important to underline that, due the limited scope of the essay, it does not aim to examine and understand all three phenomena in their depth, but instead to use them as examples to answer the research question.
RQ: To what extent have social media and information technologies impacted recent elections and referenda in the US and Germany?
Setting the scene: media systems in the US and Germany
Scholars like Schroeder (2018) argue that the effects of new media on politics and society only can be assessed alongside an acknowledgement of national differences in media systems. This argument is based on the work of Hallin und Mancini (2004) who assert – despite globalizing patterns – significant differences in the media systems of western democracies. Following this argument, new media is unquestionable shaping the media systems worldwide, but its specific effects depend on the respective media systems (Schroeder 2018). Therefore, the media systems in Germany and the US will briefly be introduced and relevant differences regarding the research question will be outlined.
Hallin and Mancini (2004) categorize the US as liberal media system which is market-dominated and has several unique attributes. The media system is characterized by strong competition between various private news companies for audience shares and low state intervention in terms of public broadcasting or regulation. A current central challenge for the media companies is strong economic pressure resulting from the shortfall of advertising money and the decrease of print media. Commentators are afraid that this commercial pressure might influence the type and quality of news coverage. Another more recent development is a growing polarisation within the American system as a whole (Schweiger 2017), where the political neutrality of media agencies is increasingly questioned. This results in substantial distrust of media along party lines; while 53 percent of Americans trust the media they use themselves, only 38 percent trust the media in general (a remarkably low value compared to other countries; Reuters Institute 2017).
Contrarily, Germany is categorized as a democratic corporatist country characterized by strong state intervention and a tradition of public broadcasting (Hallin and Mancini 2004). German public broadcasting is as politically independent as possible and is obligated to serve public interest, support education and enhance diversity. In addition to the public services there are private newspapers and TV channels which play a central role within the media system (Schweiger 2017). As in the US, the decrease of advertising is considered as a central threat, but with the strong role of public service the system is considered more stable (Reuters Institute 2017). A major challenge of the last years was the rise of political right-wing movements like Pegida (Patriotic Europeans against the Islamification of the Occident) and AfD (Alternative for Germany) and their opposition to traditional media. Phrases like “Systempresse” (system press) or “Lügenpresse” (lying press) became popular and there was a general decrease in public trust of the press. Nevertheless, with around 50 percent of Germans claiming to trust the media most of the time, faith in media remains comparably high (Reuters Institute 2017).
Looking at the role of digital media within the two media systems, we see several comparable developments which nevertheless differ substantially in their degree. In general, the role of digital media as an information source is growing in both countries, particularly for younger, more educated and politically interested citizens. This is especially true for the youngest generation for whom digital media is the central information source in both countries. Nevertheless, there are substantial differences. Whereas in the US digital media is becoming a more and more prominent information source, in Germany the Internet still plays a subordinate role in news consumption and is placed fourth behind TV, radio and newspaper (Schweiger 2017). As the role of the Internet increases in the US, niche partisan outlets like Breitbart and Occupy Democrats can increase their audience significantly. On other hand, in Germany the public-service evening news programmes are considered especially important news sources (Reuters Institute 2017). To summarize, while the US is characterized by strong market competition and the central role of the Internet as a news medium, Germany retains a strong public broadcasting system and a lesser role of the Internet in news consumption.
3.1 Internet media as an action field for new political actors
There is long tradition of scholarship of the Internet and politics focusing on new political movements and actors and how they use the Internet to accomplish (or fail to accomplish) their goals. A common assumption within these field of study is that Internet media is used by these new political actors to bypass traditional media gatekeepers (Stier et al. 2017). While these new actors often perceive traditional media as biased against them, the Internet offers them a new media realm where they can promote their messages. When successful, the Internet can help them to become a “counterpublic” (Warner 2002) and finally also reshape the agenda of traditional media in their terms (Schroeder 2018).
Schroeder (2018) takes this position while analysing the success of Donald Trump in the US, especially in the primary elections. He argues that Trump started as an outsider without support of the party establishment but was able to push himself forward with clever usage of Twitter. While using this channel to promote controversial positions on a range of issues he received a disproportionable amount of attention within Twitter, but especially from the traditional media. The coverage was often negative, but free; Trump spent much less money than others on political advertising. This argument refers to the idea of a limited attention space within an environment of unlimited choice where a broad range of issues and actors compete for attention (McCombs 2014; Webster 2014). Traditional media has been seen as in a gatekeeper position for this attention space and Trump was successful in compelling traditional media to include him in the media agenda and in the limited attention space. This argument is supported by a correlated increase in media coverage and poll results. Following Schroeder (2018) further, Trump’s success in putting himself on the media agenda was reasoned by the characteristics of the US media systems as explained earlier. The US media market is driven by market competition for audience share and is focused on horse-race politics. Trump was excellent in playing to these characteristics as his controversial statements ensured audience attention and therefore served traditional media’s own needs. It is important to note at this point that this argument does not focus not on social media as a platform for direct communication with his supporters, but how it underlines the role of social media in its interplay with traditional media. To conclude, Trump used Internet media to push himself in the traditional media agenda and to succeed in the primaries, despite the fact that he was a political outsider and spent fewer resources than other candidates.
Also for the rise of the AfD within the German party system a similar argument can be made. The AfD is the most successful German party in terms of Facebook usage. They have significantly more follower than other much larger political parties (~320,000 vs ~185,000 for the second largest Die Linke), their posts achieve the highest interaction rate of all parties, and they spend the most money in their social media campaigning (Wired 2017). Within Facebook they are dominating discussion about mass migration, Islam and media bias. A turning point in the party history of the AfD was the so-called refugee crisis in 2015. While in the Summer 2015 major parts of the public discussion were characterized by the so called “Willkommenskultur”, there was a major mood swing in late 2015 with the rise of strong anti-immigrant sentiments, a backlash against the governing parties and strong accusations against the media about biased coverage. The AfD profited from this development, which is shown by a quote from the current party leader, Alexander Gauland, who acknowledged the refugee crisis as “present” for his party (Spiegel 2015). While the AfD had constant low poll results for most of 2015 after intense internal fights and was predicted to become irrelevant by many political commentators, there was a sudden improvement in poll results in Fall 2015. In terms of the role of Internet media, the crucial point is that improvement was not mirrored by traditional media for a substantial amount of time. The AfD was simply not represented in the coverage of the German traditional media at this point, although this changed dramatically in 2016 (Haller 2017). On the other hand, this development was reflected in social media. As Schelter et al. (2016) show, the rise of the AfD in polls strongly correlated with an explosion of user activity on the official AfD Facebook page. There are good reasons to conclude that the AfD used Facebook for building a counter-public during this major political event, which ultimately was the basis for their later successes in terms of election results and traditional media coverage. To sum this up, the AfD was able to use Facebook to create a counter-public during the refugee crisis 2015, which formed the basis for the later success of the party.
While there seem to be compelling arguments for the influence of social media and information technologies for both examples in this section, it is important to see the limitations of these arguments. As Schroeder (2018) highlights, these developments cannot be separated from the political and structural conditions and developments in the respective countries, an argument which will be developed further in the discussion.
3.2 Internet media and political disinformation
Another major phenomenon which was frequently discussed in recent years is that of “fake news” or political disinformation. While the terminology “fake news” is avoided by most academics, Bennett und Livingston (2018) define disinformation as “intentional falsehoods spread as news stories or simulated documentary formats to advance political goals.” They argue that the spread of disinformation is grounded in declining confidence in political and media elites which results into an openness to alternative sources. Such sources are often associated with political actors who aim to undermine the institutional legitimacy of centre parties, governments and the democratic system itself through the spread of political disinformation. This is perceived as particularly harmful as a functioning democracy is based on rational public debate, grounded in the norms of reason and evidence. The channels for spreading political disinformation are most often found within digital media. All in all, political disinformation in Internet media is seen as a major threat to democracy.
Before coming to the actual discussion of the role of political disinformation in US and German elections, a short introduction into the current status of communication studies and media effects will be necessary. The rise of the Internet is influencing the area of communication studies significantly, Neuman (2016) speaks of a paradigm shift in the foundational concepts of this field. While the order of old concepts like mass communication vs. interpersonal communication is undermined by socio-technological developments, scholars like Bennett und Iyengar (2008), Chadwick (2013), Neuman (2016) and Schroeder (2018) argue for how communication studies should be theorized in the digital age and media effects should be assessed. As McCombs (2014) puts it, there is a lack of research and understanding especially in terms of media effects. Important steps towards more integrated research, such as those in Thorson und Wells (2016) demonstrate how challenging this kind of research is, as a broad variety of different kinds of data—often difficult to access—need to be collected. Returning to the actual question of this subchapter—whether political disinformation in new media might have had an impact on the elections in the US and Germany—these general challenges of communication studies set strong limitations on how this question can be answered as we will see below. To conclude, communication studies are undergoing a general transformation due to the rise of the Internet, where the assessment of media effects in particular is seen as challenging.
Looking at the US, the election was accompanied by a broad range of developments within the new media. Guess et al. (2018) summarize that one of four Americans visited a so called fake news website before the election. Silverman (2016) examines that the top 20 fake stories resulted in more engagement in terms reactions and comments than the top 20 real stories, an effect that increased towards the election. Looking at Twitter, Howard et al. (2018) conclude that Twitter users encountered more misinformation, polarizing and conspiratorial content than professionally produced news, and in some swing states even higher average levels of misinformation were measured. Analysing bot activities, the same research project estimated that between 20 and 30 percent of the general twitter activity was produced by automated accounts which also pushed political disinformation (Kollanyi et al. 2016). Faris et al. (2017) conclude, based on their analysis of various media types and platforms, that right-wing media network played a substantial role in setting the agenda for traditional media. Similarly political interference by Russia, like political advertising and negative campaigning on social media platforms, was measured (Mueller 2018). All in all, there was a massive occurrence of political disinformation during the US election.
The analysis of the German election 2017 looks very different so far. Here experts similarly describe significant amounts of political disinformation especially from the far-right. The AfD itself, supporting Internet media networks like Epoch Times or right-wing blogs and groups like Infokrieg and Reconquista Germania, are accused of media manipulation in terms of spreading political disinformation and pushing right-wing content (Sängerlaub et al. 2018; Davey and Ebner 2017). Nevertheless, Neudert et al. (2017) as well Sängerlaub et al. (2018) conclude based on several data sources that professional news sources were much more widespread than falsehoods (4 to 1) and that in general political disinformation played a minor role in election, especially compared to the US. The comparably low extent of falsehoods is especially remarkable when acknowledging the systematic efforts of groups like Reconquista Germania to systematically push AfD and other right-wing content (while it is important to underline that this type of content cannot be equated with falsehoods, but, as elaborated earlier, falsehoods most often originate from this end of the spectrum). The differences between both cases are mostly reasoned by already elaborated substantial differences between both media systems. On the hand, the overall trust within traditional media is significantly higher in Germany than in the US. Thus traditional media content was much more widespread that falsehoods. On the other hand, the Internet and especially social media plays a much less important role in news consumption. To summarize, compared with the US political disinformation played a much lower role.
Based on this analysis, what is possible to conclude about the question whether political disinformation in new media might have had an impact on the election results? Following the previous examination of the challenges of assessing media effects, the analysis of the role of political disinformation above was not an analysis in terms of opinion formation or voter share, but of the widespread dissemination and engagement with it. So, what might be possible to conclude for the US so far is that looking at the omnipresence of political disinformation, it probably played a significant role in the election in one way or the other. By contrast, in Germany it seems like it only played a role within the right-wing spectrum and with minor influence in the overall debate. What is not possible to conclude in either case is a quantification of the actual impact. Did Trump win the crucial swing states because of political disinformation? How much more of the vote did the AfD get than if they hadn’t used disinformation? So far it is not possible to answer this question on a scientific basis. There are some attempts like Allcott und Gentzkow (2017) who suggest that Trump would have won anyway. But they state themselves that their methodology is based on many unproven assumptions and the results should be interpreted carefully. All in all, it is not possible to conclusively answer these questions at this moment.
3.3. Internet media and echo chambers
A third phenomenon which gained disreputable prominence in the last in both public and academic discourse is the emergence of so-called echo chambers (Sunstein 2001) an idea which relates to a multitude of different but closely connected concepts like “media enclaves” (Webster 2014) “filter bubles” (Pariser 2012) or “gated communities” (Sunstein 2001). While there is a lack of consistent definitions, most of these arguments refer back to the idea of a fragmentation of the media audience and a resulting increase of polarisation in society caused by the evolution of the Internet (Stark 2013). In terms of impact on the election, this potential fragmentation and polarisation might undermine public discourse, support the growth of parallel worlds for the political extreme and influence the opinion formation of substantial parts of the population. As stated in (Rau 2017), the central arguments for the fragmenting and polarising role of Internet media are the development of a high choice media environment and a potential problematic usage of this environment based on psychological conditions like the confirmation bias (the tendency to favour information in a way that confirms one’s pre-existing beliefs), the role of personalising algorithms which might only serve users information which confirms and supports their own beliefs and positions, and finally the fear of cyber-balkanisation, where users use the opportunities of social interaction through social network systems to connect only with likeminded users and reinforce each other in their beliefs. As with political disinformation, the assessment of the actual impact of echo chambers is very challenging as the research must deal with similar difficulties. As stated earlier, while there is a remarkable lack of consistently used definitions and a predominant research framework on this issue (Rau 2017), a major part of the research focuses on the question of selective exposure. Most of these studies only take one kind of media content or a single platform into account and cannot assess the most relevant audience perspective (Webster 2014). Some of them suffer from poor operationalisation (they look at interaction fragmentation instead of exposure fragmentation), there is a general lack of longitudinal studies, and the research field must deal with potentially significant differentiation from country to country. Most importantly, as with political disinformation, less is known about the actual effects. To sum this up, while the idea of echo chambers gained significant prominence in the last years, empirical research proves challenging.
Looking at the first part of the echo chamber argument—i.e., the fear of audience fragmentation—there is a growing body of literature that suggests that there is only limited evidence for echo chambers. In the US case, while for example Flaxman et al. (2016) examine a clear audience fragmentation along political affiliation, more comprehensive and recent studies like (Nelson und Webster 2017)and Fletcher und Nielsen (2017) cannot confirm this finding. Instead they recognize a strong audience overlap even for partisan news sources and a political diverse audience. Looking at the role of social networks like Twitter and Facebook, several studies like Conover et al. (2011), Mocanu et al. (2015) and (Aragón et al.) find clear patterns of fragmentation. These studies must be interpreted carefully; as already mentioned, a substantial number of studies in this area look not at the fragmentation of information flow (as this kind of data is not available), but on interaction fragmentation (e.g. retweets). The information value of these kind of studies in terms of audience fragmentation is limited. Moreover, these types of studies often look at single (often political) events. In contrast to the described results, a longitudinal study with a time period of two years and including millions of tweets and users can find such patterns of fragmentation only for specific political issues over short time periods but not in general (Barbera et al. 2015). The strongest argument against audience fragmentation within social networks are so-called network effects. This concept describes the idea that as larger networks gets, the more heterogeneous they become (Schweiger 2017). This is confirmed by empirical findings by Bakshy et al. (2015) as well Barberá (2015) who conclude that contact networks in social networks are politically heterogeneous. In combination with the findings of Messing und Westwood (2014) that social recommendations strongly influence whether people read news sources, these studies suggest that social networks increase the diversity of news exposure for the majority of the users. All in all, there is not much empirical evidence to support the idea of a fragmented audience in the US.
For Germany, we have a much less detailed picture about audience fragmentation as most of the communication research focuses on the US. Nevertheless, studies like (Fletcher und Nielsen 2017) indicate a strong overlap of media audiences while (Barberá 2015) finds heterogeneous contact networks for German social media as well. Moreover, it is fair to say that we can expect fewer issues with echo chambers than in the US. First, Germany in general is much less politically polarised than the US. (Trilling et al. (2016) supports this arguments in regard to the Netherlands); second, the already highlighted role of public broadcasting; and third, the comparably minor role of the Internet in information consumption as a potential driver of echo chambers. So, all in all, empirical evidence cannot confirm the idea of a fragmented audiences in Germany, either.
The second argument central to echo chambers is the assumption that the potential audience fragmentation caused by the Internet drives polarisation in society. Following this argumentation, echo chambers might be responsible for the dramatic increase in polarisation in the US. Boxell et al. (2017) disagree with this argument and show that polarisation has grown the most among the demographic groups which are least likely to use the Internet and social media. Only a small amount of the recent increase in polarisation could be explained by the Internet. Baberba (2015) finds in his year-long longitudinal study not polarising but de-polarising effects of social networks through network effects (which refers back to argument from second to last paragraph). An interesting exception and a recurrent finding in this context is the role of politically extreme users. Prior (2013) concludes that an increase in polarisation based on selective exposure is especially true for a highly politically involved segment of the population, while Bode (2016) concludes that such users delete other users with politically dissonant beliefs from their online contact networks (so they homogenise them in political terms). Referring back to the issue of political disinformation, Guess et al. (2018) summarize in their findings that there seems to be a case of selection exposure of fake news, where a minority of ten percent of the population is responsible for around 60 percent of visits of so called fake news websites. To summarize, there is not much evidence for the polarising impact of Internet media apart from a politically extreme minority.
As with political disinformation, it is not possible to assess the actual impact (for example in terms of opinion formation) of echo chambers on the elections in the US and Germany. Looking at the society as a whole, empirical evidence suggests that the importance of echo chambers is overrated. In general, it was not possible to find convincing evidence for large-scale audience fragmentation or polarisation caused primarily by the Internet in either the US or Germany. The exception seems to be for the political extremes where a narrow subset of the population shows indices of selective exposure and resulting polarisation, proving to be an especially fertile ground for harmful phenomena like political disinformation. This group is relevant as it can disintegrate the idea of a public debate, foster hostility toward political opponents and undermine trust in political and media institutions. Nevertheless, this group is a minority and not a sufficient basis to win an election. Their role must be assessed further before claims can be made about the actual impact on the election result. As for political disinformation, it is not possible to examine whether echo chambers had a decisive impact on the election results or not. Finally, and again similar to political disinformation, it is important to not overstate potential media effects compared to real world phenomena, as is argued in the discussion.
This essay aimed to investigate the question “To what extent have social media and information technologies impacted recent elections and referenda in the US and Germany?”. Overall, the analysis showed mixed results for the role of the Internet and social media on the elections. Looking at Internet media as an action field for new political actors, there were compelling arguments that Internet media was a precondition for the success of Donald Trump and the German AfD in entering the political landscape. Donald Trump used Internet media in the primaries to gain a disproportionate amount of attention and outcompete the other candidates, despite lacking the support of the party establishment and spending much fewer resources than the other candidates. Following Schroeder (2018), Trump used the dynamics of a market-orientated media system to circumvent the gatekeeping role of traditional media and push himself into the media agenda. The AfD used Internet media to create a counter-public during the so-called refugee crisis in 2015, while circumventing the gatekeeper position of traditional media. Social media remains the most important communication channel for the party.
However, assessing the impact of political disinformation on the election results turned out much more difficult. While in the US there was massive presence of political disinformation prior to the election, political disinformation in Germany only played a subordinate role. This was reasoned by the substantial differences in the political and the media systems between the US and Germany, where in Germany there is much less political polarisation and distrust in media and politics, as well the central role of public media versus the subordinate role of social media. Nevertheless, as communication studies is undergoing foundational transformations with difficulties assessing contemporary media effects, it was not possible to say whether political disinformation had decisive impact on the elections or not.
The analysis of the role of echo chambers showed similar results. In general, it seemed like the importance of echo chambers is overstated as empirical research does not support the idea of large-scale audience fragmentation or significant polarisation caused primarily by the Internet. There are signs that there might be an exception for the political extreme which in general might have harmful effects on the democratic culture, but this group is still a minority and their influence will remain unclear until further research is conducted. It was also not possible to say whether political disinformation had decisive impact on the elections or not, but there are good arguments to say that the overheated discussions were not in a reasonable relation to potential effects.
While assessing the potential impact of Internet media on election results, it became salient that these developments cannot be separated by the political and structural conditions and developments in the respective countries. It was clearly shown that the impact of the Internet was imbedded in the respective political and media systems and connected to major political events like the refugee crisis. This point is pivotal in understanding and classifying the impact of Internet media in general and therefore will be elaborated here in more detail. As shown in this essay, there is a tendency especially in the public discussion to credit the Internet great impact on major political events like election outcomes. As also was examined, the Internet without doubt plays a significant role in these events. Nevertheless, the influence of the Internet is limited compared to other major structural conditions or long-term developments and potential media effect should not be overrated in comparison to other factors (Schroeder, 2018). Trump could have gained as much attention as possible but he never would have won the election without the support of a substantial portion of the electorate. The issues which Trump pushed forward like distrust in the established state elites and the government, distrust in the media and anti-immigrant, -refugee and -Muslim prejudices might have been reinforced by him but did not originate in him. The same is true while looking at the impact of real-world phenomena and long-term developments like the aftermath of the financial crisis, fear of globalisation or security issues. Looking at the the AfD, they could not have not succeeded without crucial non-media factors. These include similar factors to Trump like globalisation fears and anti-immigrant sentiments but also specific German conditions like major changes in the political party landscape (left orientation of the Christdemocracts under Angela Merkel and declining party affiliation (Nollmann 2007)) or persisting disparities between former west and east Germany in terms of politics and economics (Arnold et al. 2015). All in all, while the Internet clearly is shaping society and politics, at the same tame it is important to not exaggerate its impact compared to other deeper and more influential factors.
To summarize, there are several points to make based on this analysis. First, it is important conclude that the Internet had had a significant impact on the outcome of elections and their results could not been understood without studying the Internet and its influences. This clearly shows that politics and society are shaped by the Internet and underlines the importance of this research field for disciplines like political science. Second, it became clear that the influence of the Internet cannot be assessed without taking the specifies of the respective political system and media system of the respective country into account. The context and the surroundings matter for the impact of this socio-technological development and an analysis of it must be embedded in this context. Finally, the analysis showed that there is a need for more research and understanding. To really assess the impact of the Internet on major political events like elections, a better theoretical framework must be developed to inform further empirical research. This includes media effects in general as well effects of political disinformation and potential echo chambers on opinion formation and on the democratic culture as a whole. Special attention should spend to the influence of the Internet on the political extremes and their impact on the political system.
Aragón, Pablo; Volkovich, Yana; Laniado, David; Kaltenbrunner, Andreas: When a Movement Becomes a Party. The 2015 Barcelona City Council Election. Online verfügbar unter http://arxiv.org/pdf/1507.08599v1.
Arnold, Felix; Freier, Ronny; Kroh, Martin (2015): Geteilte politische Kultur auch 25 Jahre nach der Wiedervereinigung? In: DIW-Wochenbericht 82 (37), S. 803–814. Online verfügbar unter http://hdl.handle.net/10419/118624.
Bakshy, Eytan; Messing, Solomon; Adamic, Lada A. (2015): Political science. Exposure to ideologically diverse news and opinion on Facebook. In: Science (New York, N.Y.) 348 (6239), S. 1130–1132. DOI: 10.1126/science.aaa1160.
Barbera, Pablo; Jost, John T.; Nagler, Jonathan; Tucker, Joshua A.; Bonneau, Richard (2015): Tweeting From Left to Right: Is Online Political Communication More Than an Echo Chamber? In: Psychological science 26 (10), S. 1531–1542. DOI: 10.1177/0956797615594620.
Bennett, W. Lance; Livingston, Steven (2018): The disinformation order. Disruptive communication and the decline of democratic institutions. In: European Journal of Communication 33 (2), S. 122–139. DOI: 10.1177/0267323118760317.
Boxell, Levi; Gentzkow, Matthew; Shapiro, Jesse M. (2017): Greater Internet use is not associated with faster growth in political polarization among US demographic groups. In: Proceedings of the National Academy of Sciences of the United States of America 114 (40), S. 10612–10617. DOI: 10.1073/pnas.1706588114.
Chadwick, Andrew (2006): Internet politics. States, citizens, and new communication technologies. New York, NY: Oxford Univ. Press. Online verfügbar unter http://www.loc.gov/catdir/enhancements/fy0635/2005050892-d.html.
Davey, Jacob; Ebner, Julia (2017): The Fringe Insurgency. Connectivity, Convergence and Mainstreaming of the Extreme Right. Hg. v. Institute for Strategic Dialogue. Online verfügbar unter https://www.isdglobal.org/wp-content/uploads/2017/10/The-Fringe-Insurgency-221017.pdf.
Faris, Robert; Roberts, Hal; Etling, Bruce; Bourassa, Nikki; Zuckerman, Ethan; Benkler, Yochai (2017): Partisanship, Propaganda, and Disinformation. Online Media and the 2016 US Presidential Election.
Fletcher, Richard; Nielsen, Rasmus Kleis (2017): Are News Audiences Increasingly Fragmented? A Cross-National Comparative Analysis of Cross-Platform News Audience Fragmentation and Duplication. In: J Commun 67 (4), S. 476–498. DOI: 10.1111/jcom.12315.
Frankfurter Allgemeine Zeitung (2017): Zeitenwende in Amerika. Accessed at 23.04.2018.
http://www.faz.net/aktuell/politik/trumps-praesidentschaft/donald-trump-als- uspraesident- bedeutet-zeitwende-in-usa-14668148.html
Guess, Andrew; Nyhan, Brendan; Reifler, Jason (2018): Selective Exposure to Misinformation: Evidence from the consumption of fake news during the 2016 U.S. presidential campaign. Online verfügbar unter https://www.dartmouth.edu/~nyhan/fake-news-2016.pdf.
Hallin, Daniel C.; Mancini, Paolo (2004): Comparing media systems. Three models of media and politics. Cambridge: Cambridge University Press (Communication, society and politics). Online verfügbar unter https://doi.org/10.1017/CBO9780511790867.
Howard, Philip N.; Kollanyi, Bence; Bradshaw, Samantha; Neudert, Lisa-Maria (2018): Social Media, News and Political Information during the US Election. Was Polarizing Content Concentrated in Swing States? In: arXiv preprint arXiv:1802.03573.
McCombs, Maxwell (2014): Setting the Agenda. Mass Media and Public Opinion. 2. Aufl. s.l.: Polity. Online verfügbar unter http://gbv.eblib.com/patron/FullRecord.aspx?p=1631739.
Messing, Solomon; Westwood, Sean J. (2014): Selective Exposure in the Age of Social Media: Endorsements trump partisan source affiliation when selecting news online. In: Communication Research 41 (8), S. 1042–1063. DOI: 10.1177/0093650212466406.
Mocanu, Delia; Rossi, Luca; Zhang, Qian; Karsai, Marton; Quattrociocchi, Walter (2015): Collective attention in the age of (mis)information. In: Computers in Human Behavior 51, S. 1198–1204. DOI: 10.1016/j.chb.2015.01.024.
Mueller, Robert S. (2018): Case 1:18-cr-00032-DLF. US Department of Justice. Online verfügbar unter https://www.justice.gov/file/1035477/download.
Nelson, Jacob L.; Webster, James G. (2017): The Myth of Partisan Selective Exposure. A Portrait of the Online Political News Audience. In: Social Media + Society 3 (3), 205630511772931. DOI: 10.1177/2056305117729314.
Neudert, L. M.; Kollanyi, B.; Howard, P. N. (2017): Junk News and Bots during the German Federal Presidency Election. What Were German Voters Sharing Over Twitter?: Oxford (available at http://comprop. oii. ox. ac. uk/wp-content/uploads/sites/89/2017/03/What-Were-German-Voters-Sharing-Over-Twitter-v6-1. pdf).
Neuman, W. Russell (2016): The Digital Difference. Cambridge, Mass: Harvard University Press. Online verfügbar unter http://gbv.eblib.com/patron/FullRecord.aspx?p=4603293.
Pariser, Eli (2012): Filter Bubble. Wie wir im Internet entmündigt werden. 1. Aufl. München: Carl Hanser Verlag. Online verfügbar unter http://www.hanser-elibrary.com/action/showBook?doi=10.3139/9783446431164.
Sängerlaub, Alexander; Meier, Miriam; Rühl, Wolf-Dieter (2018): Fakten statt Fakes. Verursacher, Verbreitungswege und Wirkungen von Fake News im Bundestagswahlkampf 2017. Hg. v. Stiftung Neue Verantwortung. Online verfügbar unter https://www.stiftung-nv.de/sites/default/files/snv_fakten_statt_fakes.pdf, zuletzt geprüft am 23.04.2018.
Schelter, Sebastian; Biessmann, Felix; Zobel, Malisa; Teneva, Nedelina (2016): Structural Patterns in the Rise of Germany’s New Right on Facebook. In: 2016 IEEE 16th International Conference on Data Mining Workshops (ICDMW). 2016 IEEE 16th International Conference on Data Mining Workshops (ICDMW). Barcelona, Spain, 12.12.2016 – 15.12.2016: IEEE, S. 440–445.
Schroeder, Ralph (2018): Social Theory after the Internet. Media, Technology, and Globalization. London: UCL Press. Online verfügbar unter http://www.knowledgeunlatched.org/ku-select-2017-books/.
Spiegel (2015): AfD-Vize Gauland sieht Flüchtlingskrise als Geschenk. Accessed at 23.04.2018
Stier, Sebastian; Posch, Lisa; Bleier, Arnim; Strohmaier, Markus (2017): When populists become popular. Comparing Facebook use by the right-wing movement Pegida and German political parties. In: Information, Communication & Society 20 (9), S. 1365–1388. DOI: 10.1080/1369118X.2017.1328519.
Trilling, Damian; van Klingeren, Marijn; Tsfati, Yariv (2016): Selective Exposure, Political Polarization, and Possible Mediators. Evidence From the Netherlands. In: Int J Public Opin Res, edw003. DOI: 10.1093/ijpor/edw003.
Wired (2017): So verhalten sich die Parteien auf Facebook. Accessed at 23.04.2018
Wired (2016): Your Filter Bubble is destroying Democracy. Accessed at 23.04.2018