The WhatsApp elections: Exploring WhatsApp’s position as
a political-communicative game-changer in Brazil
Master Thesis
By Jacob Samuel Lassar
UvA Student Number: 12436267
MA Television and Cross-Media Culture
University of Amsterdam
Page 1 of 94
Table of Contents
Dedication ... 2 Introduction ... 3 Theoretical Framework ... 8 Methodology... .28Brazil’s media environment ... 32
Analysis ... 37
Fake news ... 37
Media ecology ... 49
Spreadability………56
Credibility………64
Media political implications………68
Election outcome……….72
Conclusion ... 74
Page 2 of 94
Dedication
This thesis is dedicated to all the people who supported me throughout the journey.
To my parents for all their love, support and encouragement. Without you, I would have never gotten that far, thank you always and forever.
To my sisters Denise and Laura, for their support, advice and motivation. And for making me babysit the night before my submission. Danke für alles!
To my love Mayra, who always believed in me, never doubted my success for a second and was always there for me. And for letting me go to Amsterdam for a year to actually write this thesis. Love you!
Lastly, to my supervisor Markus Stauff, for all the helpful and essential feedback, and the academic support.
Page 3 of 94
INTRODUCTION
“There is an array of social media, which are double-edged: sometimes they’re used for constructive purposes, but they have also become major forces for undermining democracy. We see that everywhere. Right now, in Brazil, for example, a scandal’s erupting […] about the fact that millions of dollars were spent to produce defamatory information on the social media.” Noam Chomsky, 2018
The emergence of digital social media in the twenty-first century largely impacted all forms of communication. The new media create a digital space in which people are able to interact with others unrestricted by geographical boundaries. Anyone who has access to the internet can create a social media profile and share information within a networked space. Social media have also been instrumental in communicating activism and coordinating protests. The early 2010s Arab Spring protests and the 2014 Ukrainian Revolution for instance were both largely organized via social media platforms, most notably Facebook and Twitter (Metzger and Tucker, 2017; Hänska Ahy, 2016). Furthermore, the hierarchies associated with traditional news media greatly shifted. Social media platforms enable anyone to produce, narrate and share their own news within the online sphere, therefore shifting news production and dissemination away from a top-down towards a many-to-many approach (Castells, 2009). These new technological networks also transformed the way citizens access the news. A recent study by Pew Research Center shows that the majority of Europeans access the news via social media (Pew Research Center, 2018), thereby presenting a paradigm shift within news production and consumption. But since these platforms greatly impacted the news cycle, they have also become significantly relevant for political engagement with voters and political campaigning (Stieglitz and Dang-Xuan, 2013; Van Dijck, 2013; Stier et.al., 2018). The 2008 and 2012 U.S. presidential elections pioneered the use of social media for political campaign purposes through the strategic use of
Page 4 of 94
ethnic social media platforms and the personal engagement of Barack Obama on the then relatively new platforms Tumblr.com and Reddit (Pick, 2012).
Since then, election campaigning has strongly evolved within the digital sphere. One of the most striking developments occurred most notably during the 2016 U.S. presidential elections and the 2016 Brexit vote, both of which brought to surface a concern regarding the spread of so-called “fake news” into the social media communication streams of U.S. and UK voters (Carlson, 2018; Bastos and Mercea, 2019).
While these cases have put “fake news” into the broader media debate, the social medium WhatsApp, an instant smartphone messaging application owned by Facebook, has become an increasingly popular communication tool and finds itself in the midst of a “fake news” discussion. Many media reports express concern over the spread of fake news on WhatsApp in recent elections in India (Ponniah, 2019, Varma, 2019), Colombia (Uribe, 2018), and the 2019 EU elections (Moechel, 2019), where the chat app is perceived to have become fertile ground for disseminating false and misleading content for political campaign purposes (Resende et.al., 2019). The most prominent example in recent months were the 2018 Brazilian general elections, which propelled itself into the spotlight of media observers who perceived that fake news on WhatsApp might have strongly impacted the election outcome in favor of president-elect Jair Bolsonaro. With a population of over 200 million, Brazil has over 120 million daily WhatsApp users (Oliveira and Rossi, 2018), making the country one of the most active WhatsApp populations globally, and therefore an important object of study.
WhatsApp is a unique social media platform, because it differentiates itself through three main features. First, WhatsApp users need the contact number of others in order to send them direct messages. Facebook on the other hand only requires “adding” someone as a “friend”, and no personal number is needed. Second, all messages are end-to-end encrypted and as such not subject to any specific content rules and thus not visible to outsiders unless purposely publicized. Finally, WhatsApp users can easily create groups of contacts of up to 256
Page 5 of 94
members1 and only the group admins can choose who joins a group. On Facebook any posted content becomes public to the entire list of friend contacts, but only those who actively choose to access the posters’ profile or see the content in their news feed get to see the content. This is different on WhatsApp, where one receives a notification on the smartphone when someone, either directly or within a group, shares an article, video, voice message or other form of content2. Thus the receiver is lured into engaging with the content in some way or another.
Until the second run-off between the competing candidates in Brazil, most international media outlets did not recognize the key importance of WhatsApp on the political campaign trails of the Brazilian candidates. This is because most academic and journalistic research hitherto concerned itself more with fake news spread on Facebook, Twitter and other social media, and neglected the political and democratic potential of WhatsApp.
When on October 18th, 2018, just a few days before the second run-off, the biggest Brazilian daily newspaper Folha de São Paulo (Mello, 2018) revealed that a group of local
companies were paying for a campaign to bombard millions of Brazilians with fake news on WhatsApp, the focus of the media was finally turned towards WhatsApp.
While it has been found that the candidates on both political spectra have engaged in digital fake news campaigning, most media articles recognized the particular effort on the political right, associated with Jair Bolsonaro, in spreading and developing coordinated false news operations online (Nalon, 2018). This parallels the empirical findings of particular right-wing efforts to induce false information into social media channels during the 2016 U.S. elections (Guess et.al., 2019). While these previous findings were mainly associated with the general question of political mobilization via social media, the Brazilian case highlights the
2 It is possible to „mute“ notifications for a certain time period, but that requires WhatsApp users to actively
Page 6 of 94
specific debate of WhatsApp’s role within Brazilian society and the dynamics of fake news on the platform.
Considering the lack of reflective research into WhatsApp’s specific role within political communication and voter engagement, this thesis will look at the contribution of WhatsApp to the fake news debate, and the different aspects of WhatsApp’s role during the 2018 Brazilian general elections. More specifically, I will tackle the following research questions:
RQ1: How do media experts and average WhatsApp users make sense of fake news,
especially in regards to WhatsApp during the 2018 Brazilian elections?
RQ2: How did Brazil’s media ecology provide fertile ground for fake news circulation
via WhatsApp?
RQ3: What do media experts and WhatsApp users identify as specific features of the
chat app that enabled it to impact the election campaign?
RQ4: What factors influenced WhatsApp users to trust and believe in misleading
content that circulated on the chat app during the election campaign?
RQ5: How does fake news on WhatsApp impact media professionals in Brazil?
To explore these issues I conducted a series of in-depth semi-structured interviews with five Brazilian media experts and one campaign manager. I also created an online questionnaire in Portuguese that I sent to my personal and professional contacts in Brazil. Since I expected greater reflection on fake news and WhatsApp in Brazil from the media experts, the findings from the interviews provided a framework of comparison and assessment for each of the research questions in relation to the questionnaire results.
The thesis is organized as follows. The first section provides a theoretical framework that reflects on the literature on social media’s impact on political campaigning, fake news and propaganda, and the social aspects of WhatsApp; The second section briefly sketches the media context of Brazil; Section three details the methodology used for the analysis; Section four
Page 7 of 94
outlines the results and analytical discussion; Finally, section five concludes the thesis. Considering the numerous media reports drawing attention to fake news on WhatsApp as a game-changer for the 2018 Brazilian election outcome (Belli, 2018; Magenta et.al., 2018; Nalon, 2018; Funke, 2018), it can be assumed that such political ripple effect will repeat themselves in other countries with high numbers of WhatsApp users. Therefore I believe it is important to analyze the factors the contributed to the debate in the Brazilian context.
Page 8 of 94
Theoretical framework
In the past, internet websites were characterized by their static, one-way communication model in which publishers and content owners communicated with audiences as if they were lecturing and simply passing over information (Thackeray et.al., 2008; Arora, 2015), a digital era more widely known as Web 1.0, or the “read-only” web (Ankerson, 2015). Textual and audiovisual content was put on the web for the end-user to “consume”, and thus there was hardly any interaction between the internet users and content providers. Through the emergence of the age of Web 2.0, users can share, link, collaborate and network interactively in a two-way communication process (Van Dijck, 2013, p. 5, Arora, 2015). Internet users can now produce and distribute content among their peers in real-time, and whatever is on the internet can be sent, adjusted and shared within collaborative groups. The participatory culture of Web 2.0 affordances has been found to have transformative effects on education (Asselin and Moayeri, 2011; DePietro, 2012), health-care (Chou et.al., 2012) and particularly on election campaigning (Lilleker and Jackson, 2010; Gibson and McAllister, 2011; Lilleker and Jackson, 2011). One specific affordance that emerged as a result of Web 2.0 technologies are social media, which are networked platforms that enable users across the globe to blog, microblog, connect and share content (Van Dijck, 2013). Van Dijck (2013, p. 11) suggests that “the very word ‘social’ associated with media implies that platforms are user centered and that they facilitate communal activities”. The current web environment in which social media have come to replace much of long-held communication tools, such as email and SMS, has come to be associated with increased citizen empowerment (Grauenkaer and Tufte, 2018; Gillespie, 2013; Shirky, 2011), transnational peer-to-peer communication unrestricted by geographical boundaries and more personalized messaging formats that gave way to an increasing number of people using social media applications, platforms and tools where users can produce and share videos (YouTube),
Page 9 of 94
network among themselves (Facebook, WhatsApp and Twitter) and share photos of their recent trips (Instagram) (Constantinides and Fountain, 2008; Thackeray et.al., 2008; Kaplan and Haenlein, 2010). The news media in particular have been found to be impacted by social media’s ascension as a primary communication tool (Gil de Zúñiga et.al., 2012; García Perdomo, 2017; Boczkowski et.al., 2017). A recent study by the Pew Research Center for instance found that two-thirds of north-American adults access news content from social media platforms (Gottfried and Shearer, 2016), and that the majorities of Western Europeans get their news from social media sites (Pew Research Center, 2018). Hence many argue that social media have an impactful transformative effect on the broader media landscape and journalism as a whole (Gil de Zúñiga et.al., 2012; García Perdomo, 2017; Boczkowski et.al., 2017), and as such they influence democratic processes and the public information exchange (Goode, 2009; Gil de Zúñiga et.al., 2012; Loader and Mercea, 2011). Tufekci for instance explains how Twitter, one of the most popular social media, empowered and fueled the 2013 Gezi Park protests and therefore helped shape how democracy is perceived and attained in Turkey (Tufekci, 2013; Tufekci, 2017), but she also finds that digital social platforms allow certain political actors to conduct “unaccountable campaigns of persuasion and social engineering in political, civic and commercial spheres” through the use of computational data practices (Tufekci, 2014). Due to social media’s omnipresence in both news discourses and within a broader interdisciplinary academic debate, it is widely argued that democracy as a whole is now defined and shaped by Facebook, Twitter and others. (Loader and Mercea, 2011; Pennington, 2014; Tucker et.al., 2017).
In order to look at democracy’s position within the sphere of social networked media, we must first explore the broader conception of democracy and the media’s role in that. Many argue that the democratic discourse and status of a nation and a society very much depends on the media in that environment. Media are considered to have key functions for democracy. According to Asp (2007) for instance, one of the functions of media in a democratic society is
Page 10 of 94
to inform citizenry. As such, media are responsible for contributing “to free and autonomous opinion formation by providing the citizens with a sufficient basis for the rational assessments that enable them to form opinions on issues of consequence to society and the actions of those in positions of power”. Asp’s notion finds its roots in the Habermasian idea of the public sphere (Habermas, 1989), a “realm of social life where the exchange of information and views on questions of common concern can take place so that the public opinion can be formed” (Dahlgren, 2002, p. 195). This concept of the public sphere contends that a strong democracy is built on the realm of social life in which citizens deliberate informally about all matters of society, regardless of their background or ideology (Habermas, 1989), and “where the exchange of information, positions, and opinions […] and questions of common concern take place, ultimately forming public opinion, which in turn guides the political system” (Graham, 2009). Habermas argues that the rational public discourse takes place both in public spaces, such as coffee shops, and through the mass media. Verstraeten (1996) is one of those who support Habermas’ conception, and maintains that “the public sphere requires a ‘forum’ that is accessible to as many people as possible and where a large variety of social experiences can be expressed and exchanged”. Considering the year in which his article was published, it is evident that he did not consider new media such as social media into consideration. Nonetheless, what Verstreaten notes is that the traditional media can be seen as such a ‘forum’ of public contestation. This idea resonates with Asp (2007), who suggests that the media form the bedrock of a democratic society, and that democracy works when a society’s citizens make free and largely independent choices between different political options. The ideals of the Habermasian public sphere have been challenged by many scholars, some claiming it portrays an idealized and unrealistic view of society, since it does not take social classes, inequality and political conditions into consideration (Lubenow, 2012, Dean, 2003). The emergence of new digital media in particular encouraged some scholars to question the initially conceived notion of the public sphere. Dahlgren (2018) emphasizes that the many-to-many communications
Page 11 of 94
paradigm inherent to digital media threatens and distorts the basis of a rationally deliberating public sphere. The ability for any citizen to produce his or her own media content eliminates traditional gatekeepers of mass media and removes any inclination of equal and balanced public informative media (Dahlgren, 2018; Gehl, 2011). I therefore want to call Rodny-Gumede (2017) to attention, who maintains that the links between [news] media and political communication and campaigning need to be re-modeled, particularly in light of what Dahlgren (2005) and WAN (2014) call a fragmentation of the public sphere in relation to an increased audience disengagement with traditional news media. According to Macnamara (2012), particularly young citizens who increasingly use and engage with social media are becoming increasingly disengaged from traditional media and therefore alter the dynamics of the public sphere. Based on this line of reasoning, Canel and Voltmer (2014) impart that “the unpredictable nature of the internet is fundamentally altering the way in which political matters are communicated in the public sphere” (p.1) and hence new media in the form of digital social media are shaping the impact of media on the democratic foundations.
For Lippman (1922), media should thus provide all citizens with full and fair information about what they need to know and would not learn otherwise. The notion that media need to enable individuals to make sound political choices, or more precisely to retrospectively judge the performance of the representatives and to prospectively decide which candidates match their preferences best, resonates in many contributions about media and democracy (Christians et al. 2009: 30; Ferree et al. 2002; Strömbäck 2005). Schudson (2008) agrees with Lippman and argues that one of the core functions of the media is to provide ‘information’. Moreover, this idea of ‘information’ is closely related to a further concept often discussed with regards to media and democracy. Traditional media is often seen as an instrument for state control and a means to diffuse government propaganda (Jebril et.al., 2013), it’s traditional function of informing the public about political affairs can also bring impulses to provide transparency.
Page 12 of 94
The question remains as to how new social media in fact contribute or alter democratic principles. With far-reaching transformations in the global media ecology and
the Habermasian public sphere due to the global rise of social media use, much prior research
has focused on and highlighted the diverse effects of social media on democratic discourses and practices (Stieglitz and Dang-Xuan, 2013; Gayo-Avello, 2015; Van Dijck and Poell, 2015; Tucker et.al., 2017). Generally speaking, two schools of thought emerged from this research.
One emphasizes the positive democratic potential of social media. This new form of networked communication has been found to facilitate citizen engagement (Moura and Michelson, 2017; Santini et.al., 2017; Harris and Harrigan, 2015; Rosenthal, 2003), access to news and an enhanced ability to undertake collective action (Shirky, 2011; Santini et.al., 2017; Sandoval-Almazan and Ramon Gil-Garcia, 2014; Brennan, 2018). Brennan (2018) for instance argues that “social media sites help bring new voices into the political discussion and help people get involved with issues that matter to them”, and Harris and Harrigan (2015) found that these social digital media such as Facebook and Twitter enable wider citizen participation and involvement, and thereby serving to legitimize the democratic system. Researchers often take the early 2010s revolts of the Arab Spring as an example, where citizens connected with each other and collaboratively organized via social media to protest against oppression, and during which social media platforms helped disenfranchised citizens to have their voice (Markham, 2014; Mcgarty et.al., 2014). Increased political and activist participation enabled through social media has also been a hot topic in Zimbabwe (Mutsvairo and Sirks, 2015), Turkey (Tufekci, 2017; Mercea et.al., 2018) and embodied by the Occupy Wall Street movement (Suh et.al., 2017). Lee et.al. (2018) refer back to the Habermasian public sphere and hold that social media are examples of subaltern public spheres. According to the authors, subaltern public spheres have a democratic function within the broader public sphere of “providing alternative platforms for minorities and marginalized” societal groups that then “defy mainstream discourses in the public sphere”. In other words, social media can empower marginalized groups within society.
Page 13 of 94
Researchers pertaining to this school of thought envision social media as a positive force for democracy, in particular due to digital media’s ability to encourage greater political participation and greater networked activism (Lee, 2018; Ley and Brewer, 2018).
However, the other school of thought challenges the optimist vision of social media use. Loader and Mercea (2011), for instance, argue that the widespread use of social media does not bear the democratic potential it is attributed with, and that we must be wary overtly celebrating Web 2.0’s affordances. Conversi (2012) draws attention to the radical and xenophobic dialogue that has emerged in diasporic social media groups, an assertion supported by Alsaad et.al. (2018), who express concern over the increase of racist discourses diffused by social media. Tucker et.al. (2017) suggest that social media platforms can and have been used by authoritarian regimes to repress freedom of speech, freedom of the press and other democratic principles, and Couldry (2015) argues that social media platforms on their own do not necessarily provide sustainable political participation and activism. In light of this, Persily (2017) rhetorically and cynically posits “can democracy survive the internet?”, provoking the notion that the internet and its various applications present a threat to democratic participation. While Lee et.al. (2018) mainly highlight the positive aspects of social media, they also point out fundamental democratic issues related to social media use. The authors for instance highlight growing political polarization as a consequence of frequent social media use, further supported by a recent study suggesting that 94% of Facebook users and 88% of YouTubers were polarized and thus focused their online social behavior and reading on a specific narrative of a controversial issue (Uzzi, 2017, Lee et.al., 2018). Social media have thus been dubbed a “double-edged sword” (Stavridis, 2018), pertaining both positive and negative spill-over effects on the democratic status of nations.
One fundamental element of democracy is an election and a political campaign preceding an election. As aforementioned, one of media’s function within a democracy is to inform citizenry, and that includes political and election-related information (Asp, 2007). A
Page 14 of 94
substantive body of research has been conducted on the emergence of new social media platforms and their influences on political campaigning in particular (Harris and Harrigan, 2015; Shirky, 2011; Vesnic-Alujevic, 2012; Graham et.al., 2013; Carlisle and Patton, 2013; Ross and Bürger, 2014; Bossetta, 2018; Zeng et.al., 2010; Stieglitz and Dang-Xuan, 2013; Dimitrova and Matthes, 2018). Some have argued that the new digital network platforms brought about by the internet transformed the way political campaigns are run (Sunstein, 2017; Ross and Bürger, 2014; Dimitrova and Matthes, 2018). One argument put forth is that social media are ideal vehicles for promoting political policies and positions, as well as garnering community support for politicians (Zeng et.al., 2010; Stieglitz and Dang-Xuan, 2013). This new interplay between the internet and elections implies, according to Ituassu et.al. (2018) that we live in a world of “post-modern campaigning”, a notion suggesting that political campaigning in the current social media-defined era resembles a new and different reality to that of the twentieth century. Klinger and Russmann (2017)’s findings refute that claim. Studying the online and offline strategies political parties in Austria and Switzerland implement, the authors believe social media’s significance is exaggerated and digital strategies employed by political actors constitute merely a marginal part in campaigning. The shortcomings of their research of course is that they only focus on two specific countries, rather than exploring the impact of social media on a global scale. Ross and Bürger’s (2014) findings resonate with those of Klinger and Russmann. The authors assert that most politicians employ social media-based strategies mainly to distribute information in a one-way flow and to make themselves visible, rather than engaging with the voters. This study as well focuses on one specific country, but does put into question the extent of social media’s impact on election campaigning.
Nonetheless, all authors agree that the dynamics with which campaigns are fought have changed due to social media. While in the past election campaigns were dependent on getting their messages out via traditional mass media channels, such as broadcasters, new social media require political campaigners to fragment their messages across a spectrum of digital media
Page 15 of 94
(Klinger and Russmann, 2017; Serazio, 2014). Moura and Michelson (2017) emphasize that in effect social media-based campaigning allows an increased engagement with potential voters, but that traditional gatekeepers in the form of broadcast journalists for instance lose societal authority and that this might damage the democratic discourse.
The first widespread political use of social media was seen during the U.S. presidential elections in 2008 and 2012, in which the Obama campaign has effectively exploited the various digital platforms, such as Twitter and Facebook, to mobilize, engage with and inform their voters (Cogburn and Espinoza-Vasquez, 2011; Harris and Harrigan, 2015, Bossetta, 2018; Pick, 2012). The Obama team was able to synergize SMS, social networking sites (Twitter, YouTube and Facebook) and blogs with a more traditional political marketing strategy and an organized grassroots movement (Pick, 2012; Moufahim and Lim, 2009; Bijmolt et.al., 2010; Harris and Harrigan, 2015). During the 2012 Republican National Convention for instance, Obama faced the online Reddit community, and was thus able to engage with a community of over 5 million people (Pick, 2012). Yussi Pick, a prominent campaign strategist based in Washington D.C. and in Vienna, emphasizes that while Obama focused a large part of his campaign on microtargeting and the mobilization of voters via digital networks, platforms and media, his opponent Romney’s campaign put their efforts to “influence public opinion with traditional television advertising” (Pick, 2012) in stark contrast to the digitally oriented Obama campaign. Pick’s reflection very much resonates with Stier et.al. (2018), who suggest that social media have become battlegrounds for political campaigning because they transformed both voter engagement and the broader media environment. The vast majority of research on social media’s impact on political campaigning is based on cases from the U.S., and thus Dimitrova and Matthes (2018) and van Aelst et.al. (2017) very rightly surmise that the effects of social media can only be studied and understood in the specific context of a country and its media environment, and therefore the results of studies from the U.S. cannot be generalized for other contexts. The authors for instance emphasize the differences in voter characteristics, the varied
Page 16 of 94
structural nature of social media systems and the overall media environments that require a country and culture-specific study of social media and politics. Nonetheless, it can be surmised that social media impact and influence the way political campaigns are fought.
But with more than 2.46 billion social media users worldwide (Clement, 2019), wide concern regarding social media manipulation was raised by both journalists and academics of detrimental effects on democracy and free elections, which will be discussed in the following section.
Propaganda and fake news
While the Obama campaign showed the productive and positive aspects of social media on a political election campaign, other electoral cases on the other hand have caused great concern over content manipulation of social media content. The debate of media manipulation is not new – in fact, older concerns relating to broadcasting in the past have brought the attention to the possible political influence on mass media broadcasting (Ožegović, 2018; van Dijck, 2011). This goes back to the relevance of media for democracy. In Chew’s (2018) opinion, for instance, social media act within an “information space”, and are thus subject to “information warfare” by governmental and citizen stakeholders. As the author explains, any warfare involves some kind of propaganda. Furthermore, social media, in a way, disable the gatekeeper function of traditional media, and are thus subject to manipulation (Westlund and Ekström, 2018). Consequently, exploring social media and its political (dis-)contents requires a more in-depth look into the developments of propaganda and, further within the debate, the recently emerged term ‘fake news’. As media researcher Christian Fuchs (2018) points out, today we have a wide array of social media platforms, such as Facebook, WhatsApp and Twitter that we access on
Page 17 of 94
mobile phones, laptops, computers and TVs, and those have become important means of information and communication. These new digital affordances and media platforms must therefore, according to Fuchs, be questioned in respect of propaganda theory, and how social media’s power structure defines the current status quo.
For Marlin, who studied propaganda theory more profoundly, “the organized attempt through communication to affect belief or action or inculcate attitudes in a large audience in ways that circumvent or suppress an individual’s adequately informed, rational, reflective judgement (2013, p. 12) constitutes propaganda. Erbschloe (2017) argues that old propagandist techniques of influencing large audiences reach voting citizens in their homes and pockets through smartphones, smart TVs and laptops, thus shedding light on the core technical characteristic of propaganda in today’s digital environment.
In 1988, Noam Chomsky and Edward Herman published the canonical work
Manufacturing Consent: The Political Economy of the Mass Media. In this study, they
presented the Propaganda Model, which dealt with the ways in which traditional American news media could be seen as propaganda machines. The underlying idea of this model is that the elite and powerful “manufacture” peoples’ consent and that there exist five filters that information has to pass through in order to be published and accepted: ownership, advertising, the media elite, flak and the common enemy. By doing so, mass media can steer and dominate public opinion, in other words disseminate propaganda. While this Propaganda Model was an important contribution to the study of the communication and power structure of news media in the late 20th century, some scholars began to challenge this model and suggest that it must be re-evaluated in the current digital era. Fuchs (2018) highlights for instance, how “sourcing” as the third of the original propaganda filters, is different in digital networks than in traditional mass media. Chomsky and Herman’s initially conceived filter “sourcing” conceptualizes mass media’s reliance on trusted and established sources of information, which the authors argue is all too often from government and corporate stakeholders who spend vast amounts on PR and
Page 18 of 94
message control and is thus a form of propaganda. Fuchs in turn suggests that the internet and its affordances act as a “universal machine, at once a technology or production, distribution and consumption”, and henceforth enable user-generated content and a more networked information sourcing. Manuel Castells (2009) argues along those lines when he speaks of self-communication and the simplicity of content production that is core to digital networks, allowing many producers of information to reach larger audiences online, as compared to mass media and broadcasting formats. What we see here is a common thread of arguments, wherein information is not produced and sourced in a downward structure, but rather in a multi-polar format. As Caroline Jack puts it, “state actors are involved in the current media moment, but so are a variety of other collectivities, with a variety of motives” (2017), and therefore Jack, Fuchs and Castells posit a criticism to Chomsky and Herman’s model based on both the technological makeup of social network platforms, but also reflect an understanding of the political economy of social media.
Fuchs (2018) however also believes that some aspects of Herman and Chomsky’s model still hold true, most notably the “flak” filter. Herman and Chomsky originally refer to flak as negative responses to a media statement or program, and some scholars assert that this involves political and media lobbying (Fuchs, 2018; Zollmann, 2017). Fuchs sees automated lobbying, or rather automated bots that manipulate data and textual content used for political communication as a form of flak in the digital age. I find this argument convincing, especially when we look at how social media corporations such as Facebook and Twitter enable automated lobbying through echo-chambers and filter bubbles with algorithmic changes, particularly those of right-wing talk shows, blogs and social platforms, an argument that Chomsky and Herman also believe has become stronger (Mullen and Klaehn, 2010) and has been found salient in the U.S. media ecology (Guess et.al., 2019).
As outlined above, manipulation is discussed in relation to social media and propaganda. However, recent debate is focused mainly on the phenomenon of ‘fake news’.
Page 19 of 94
While manipulation has been shown to have existed before the emergence of digital social media, the concern over ‘fake news’ in fact is seen by many as a specific consequence of Web 2.0 affordances (Zannettou et.al., 2018; Goswami, 2018; Chen, 2017; Martens et.al., 2018). One argument is that the supposed participatory culture that was envisioned by the internet in fact meant that everyone can produce their own content, and thus the protections that traditional media provided were torn down (Chen, 2017, Marres, 2018).
Jack (2017) believes this debate is particularly contested in this decade due to the growing anxieties over the power of social media. Indeed, the relationships between actors and audiences altered due to the new digital social platforms (Lewis and Westlund, 2014). But as Tandoc Jr. et.al. (2018) suggest, the concepts of trust and credibility are key to the fake news debate. The authors propose that “trust in the media is […] based on the belief that news items are dependable, competent, and have integrity”, and that in the past media consumers did not need to assess much individual news items due to the inherent trust in media. Notions of a current “post-truth era”, as initially proposed by Keyes (2004) further supports the idea that there was a time in the past in which true and false information was easily distinguished. This line of argument however is dependent on Metzger et.al.’s (2003, p. 296) assumption that the ability of media audiences “to effectively and accurately discern credible from unreliable information” is crucial because attitudes and behaviors are shaped by information.
But this brings up the question of what “fake news” is, and whether the aforementioned argument holds valid in this context. Venturini (2019) points out that whenever the term “fake news” is mentioned, it is immediately connected to a certain imagery, involving Russian propaganda, sometimes having sexual implications and more often than not politically related. As such, the author believes it lost its original meaning. Ever since Trump brandished traditional television channels that criticized his policies as “fake news”, the term has attracted much public interest and some academic research. Two dominant definitions of fake news have
Page 20 of 94
recently been proposed. In one of the most widely cited paper on the topic, Alcott and Gentzkow (2017) define fake news as
articles that are intentionally and verifiably false, and could mislead readers. (…) Our definition includes intentionally fabricated news articles (…) It also includes many articles that originate on satirical websites but could be misunderstood as factual, especially when viewed in isolation on Twitter or Facebook feeds. (…) Our definition rules out several close cousins of fake news: 1) unintentional reporting mistakes (…); 2) rumors that do not originate from a particular news article; 3) conspiracy theories (these are, by definition, difficult to verify as true or false…); 4) satire that is unlikely to be misconstrued as factual; 5) false statements by politicians; and 6) reports that are slanted or misleading but not outright false.
Lazer et.al. (2018) on the other hand argue that the term refers to
…fabricated information that mimics news media content in form but not in organizational process or intent. Fake-news outlets, in turn, lack the news media’s editorial norms and processes for ensuring the accuracy and credibility of information. Fake news overlaps with other information disorders, such as misinformation (false or misleading information) and disinformation (false information that is purposely spread to deceive people).
It becomes clear that one point of controversy is whether fake news refers to news content (true or false) or whether it only needs to mimic it. In order not to arbitrarily decide on one definition and include any kind of content or information that just might not be fully truthful, I will follow Venturini (2018), who suggests to call “fake news” “viral news” instead, due to its addictive character and its spread that birthmarks this kind of content. Marres (2018) emphasizes that today’s changing architectures of the public sphere are suited to manipulate peoples’ opinions,
Page 21 of 94
and that while in the past news items had to go through a gatekeeping block and can be easily fact checked by experts, digital media circumvent or delay this process. On Facebook, for instance, anyone can post an article, blog posting or other content, avoiding any gate-keeper, and journalists therefore can only subsequently fact-check the information. This scenario is even intensified through messaging apps such as WhatsApp.
WhatsApp is characterized by its end-to-end encryption, which makes it impossible for outside actors to access the source, content or information produced, viewed and shared. Not even WhatsApp’s parent company, Facebook, can access users content. This entirely removes any attempt of journalists to fact-check the information, and if fact-checking is wanted, they are dependent on individual users publicly publishing content. Content that users post to their Facebook or Twitter profile on the other hand is not encrypted, and can only be shielded from outsiders with privacy settings, but system administrators can still access the content (Resende et.al., 2019).
One could argue that social media’s technical makeup, as a result, enables and encourages fake or viral news. This has encouraged past research into computational bots and the power of certain digital platforms (Tufekci, 2014; Venturini, 2018; Ribeiro and Ortellado, 2018; Machado and Konopacki, 2018). The scholarly research on this topic suggests that through big data and open data, political and commercial actors can inject a wide range of material into digital social networks for certain purposes, often using artificial bots to do the job. Venturini (2018) suggests that most often fake or viral news are spread through government agencies or related actors for political purposes to manipulate the public discourse. Venturini’s focuses on intentional political misinformation. Mihailidis and Viotty (2017) propose that this type of fake news should be categorized as “hoax-based stories that perpetuate hearsay, rumors, and misinformation”. Although the authors mainly explains this dynamic on the basis of the 2016 U.S. elections, Machado and Konopacki (2018), Eugenio Jr. (2018) and Bauerlein (2018) argue that automated technology, or bots, have been extensively used for political campaigning
Page 22 of 94
purposes in recent elections across the globe. Eugenio Jr. (2018) and Bauerlein (2018) base their research on the “Russian Firehose of Falsehood Model”, a digital propaganda tool originally coined by the Rand Corporation, and referring to digital tools employed by Russian hackers during the 2016 U.S. presidential elections. By disseminating a massive amount of fake news, doctored memes and manipulated voice-messages into a busy news environment, political actors ‘firehose’ a vast number of Facebook and WhatsApp groups and users with daily memes and false information designed to shape public opinion and bias. These messages then get shared by the social media users within social circles, thus acting as propaganda tools themselves. This multichannel messaging approach is, as Eugênio Jr (2018) asserts, used to distribute “falsehoods, lies and misrepresentations” and is seen as a “core campaign strategy” for political campaigning in a Web 2.0 media environment, again resembling the notion of intentional political content manipulation.
Fake news however are not solely understood as political interferences. Kiwi (2018) for instance argues that fake news can also be commercially motivated. The author cites joke sites such as The Onion as propagators of fake news, but clearly not politically related. This notion resonates with Burger et.al. (2018) who contend that “money-driven, highly shareable clickbait with low journalistic production standards” are particularly present on social media platforms such as Facebook. According to the authors, this commercial type of false and misleading social media content is more engaging with users than mainstream news content.
In the case of WhatsApp, messages are not only encrypted, but content can be shared with just two clicks, and to more than ten contacts and groups at the same time. The use of “multichannel messaging approach” further articulates the two-way communication process central to the current Web 2.0 era, in which information is not only passed on in an instructive manner from ‘above’, but can be produced, shared and passed on to users internationally and instantaneously. The academic debate hitherto commenced with the political campaigns of Trump in the U.S. and the subsequent public discussion around fake news (Guess et.al., 2019).
Page 23 of 94
These developments are of course to a certain extent culture specific (Corner, 2017) and hence the way in which fake news plays out in a different national and cultural context must be studied separately. This begs the question of what makes social media content so engaging and possibly more trustworthy than other content, which will be discussed in the following section.
The socialness of social media
While the section before explored the character of fake news and its relevance for political campaigning and commercial objectives, in what follows I will discuss how social connections on social media platforms contribute to the quality and trustworthiness of news and the democratic impact of fake news. When looking at social media, Bouvier (2015) suggests that it is crucial to explore the ideas and values that are shared across the interconnected networks. This is where social theory comes in handy to discuss the social habits of people.
Santini et.al. (2017) suggest that an imitative contagion process is what spurs the shared habits, behaviors and beliefs within a group of people. The authors maintain that this theory underlines the notion that the role of traditional mass media in shaping the belief systems of collectives is limited, and that social circles and contacts are more influential in shaping common views, an argument further supported by Livingstone, 2006. Dale and Strauss (2009) further reveal that messages are noticeable particularly if sent by close family members and trusted friends, a notion they coined the Noticeable Reminder Theory. Underlining this theory is the notion that social connectedness is the key to increasing political participation (Malhotra et.al., 2011)
Although this research is mainly focused on voter mobilization, we can draw parallels to social connectivity across all platforms and for geopolitical campaigning and peer-to-peer communication at large. Moura and Michelson (2017) found similar results researching voter mobilization through Facebook messages which we can extend to the research of WhatsApp communication. Tandoc Jr et.al. (2018) remind us that the fake news debate particularly
Page 24 of 94
requires us to look back at the intrinsic characteristic of credibility as a key part of social media’s socialness, an environment in which users are confronted by information overload. In this context the authors suggest the heuristic-systematic model (HSM) of information as a starting point, which states that “a message will be perceived as credible when the source is perceived as credible” (Kang et.al., 2011; Tandoc Jr. et.al., 2018). This notion supports the idea of the Noticeable Reminder Theory. Moreover, WhatsApp in particular has been found to be exchanged more often, be of a more conversational nature and is used to communicate within closer social circles and is thus most often used for group-based communication (Church and de Oliveira, 2013). Aharony and Gazit (2018) also bring to attention the importance of WhatsApp groups among family and friends, and argue that particularly socially anxious people will give more importance to their social interactions via the messaging application. This is important research, as it supports the concept of the Noticeable Reminder Theory.
One fundamental principle of social media is that of information exchange. Underlying the information exchange is the mobility of information flows (Souza Reis et.al., 2017). According to Lemos and Josgrilberg (2009, p. 29), “communication is a way to ‘move’ information from one place to another, producing sense, subjectivity, spatialization”. Hence mobile phones, as interactive objects, provide the necessary mobility for a space of information exchange, regardless of user location (Souza Reis et.al., 2017). Shirky (2011) further points out that “opinions are first transmitted by the media, and then they get echoed by friends, family members, and colleagues.” Applying the social theory and Noticeable Reminder Theory to this idea demonstrates that, although mass media disseminates certain information, the repetition or specific highlighting of this same information by closer family members provides content with more credibility.
This begs the question of how reflective message recipients are of the social factors impacting their trust in messages. Previous social media research into credibility therefore has also highlighted the importance of confirmation bias (Alsaad et.al., 2018; Winter et.al., 2016;
Page 25 of 94
Shu et.al., 2017; Spohr, 2017). This concept rests on the notion that people tend to unconsciously “favor, recall, and interpret information in a way that confirms their previous beliefs and hypotheses, while giving disproportionately less attention to alternative possibilities” (Alsaad, 2017). While social media has become an increasingly important source of news for many users (Broersma and Graham, 2012; Westerman et.al., 2014), research suggests that based on the concept of confirmation bias, social media corporations use specific filter algorithms to provide users with information content that they want to see, and thus filtering out other sources of information, opinions and ideas, which in turn reinforces the belief system of individuals based on the notion of confirmation bias (Aslaad, 2017). The validity of such assumption becomes evident through the pervasive media reports focusing on so-called “filter bubbles” (Delaney, 2017; Pariser, 2011) which causes concern particularly in light of recent political elections where bots have been found to be used to alter the algorithmic makeup of Facebook and Twitter (Tufekci, 2014).
Another important concept discussed in the literature in regards to social media’s impact on democracy is that of so-called ‘echo-chambers’. The underlying idea is that of groups of people in which only certain information and beliefs are shared, and one particular narrative is being told (Jamieson and Cappella, 2008; Sunstein, 2002). The views of the members of said group are reinforced by the lack of exposure to opposing views. According to Dubois and Blank (2018), the consequences of echo-chambers became particularly salient in light of polarized social media groups during the 2016 U.S. elections, in which American citizens navigated primarily within Facebook pages and groups that coincided with either liberal or conservative political viewpoints. Dubois and Blank strongly emphasize that echo-chambers are detrimental for the public sphere, because the inherent consequences of echo-chambers are, according to the authors, a decline in the influence of mass media and as a result an erosion of public political discussion. This idea is further supported by the findings of Lee et.al. (2018) who find that social media-enhanced echo-chambers reinforce the political polarization of South Korean
Page 26 of 94
social media users. Grimes (2017) further argues that the consequences of echo-chambers are severe for democracy, which hinges on the public sphere and its inherent characteristic of public deliberation. Recent cases of election campaigns in India (Murgia et.al., 2019) and in Brazil (Boadle, 2018) particularly highlight the dangers of echo-chambers that emerge within WhatsApp groups. These have been found by journalists as specifically damaging to an open and free public debate.
Thus far, only a handful of studies have been conducted on the impact of WhatsApp on society and politics. Sultan (2014) for instance, who investigated addiction to mobile text messaging, suggests that individuals who use WhatsApp turn to it for maintaining contact with family and friends, information, as well as for entertainment. In another study, O’Hara et al. (2014) showed that WhatsApp use is a major element in individuals’ ways of dwelling with others, associating it with forms of commitment, faithfulness, and knowledge that can be presented while using WhatsApp. Further, Church and de Oliveira (2013), who compared WhatsApp and SMS use, and assert that WhatsApp messages are exchanged more often, are more conversational, are used to communicate within closer social circles, and are used more often for group-based communication. Moura and Michelson (2017) explored the use of WhatsApp for mobilizing teen voters in Brazil, and much to the author’s surprise find that the traditional format of door-to-door canvassing remains on par with WhatsApp in young voter mobilization. Aharony and Gazit (2015) explored the importance of WhatsApp family groups and came to the conclusion that personality characteristics influence the importance of WhatsApp family groups. And lastly, Resende et.al. (2019) offer the most recent exploratory study on WhatsApp. The authors collected the content that was shared among WhatsApp groups during the Brazilian national truck drivers’ strike and the 2018 Brazilian general elections and used the data to analyze the network structures of WhatsApp groups. Thus far however there has not been conducted any
Page 27 of 94
comprehensive study or research on the dynamics of fake news on WhatsApp and the dynamics of WhatsApp that contributed to the Brazilian election outcome.
Page 28 of 94
Methodology
Since WhatsApp is a relatively new social media platform, its impact on election campaigns is uncharted territory in media studies. During the 2018 Brazilian general election, WhatsApp played a key role in both the campaign and influenced for whom the Brazilian voters cast their vote (Boadle, 2018; Belli, 2018; Santos; 2018). Fake news circulating on WhatsApp groups made headlines and caused great concern for political and media observers (ibid). Therefore, in seeking to answer my research questions, this thesis will explore the different aspects of WhatsApp’s role in the 2018 Brazilian general election, and the specific
contribution of WhatsApp to the fake news debate. This thesis relies on a series of qualitative in-depth, open-ended and semi-structured interviews with five leading Brazilian media professionals, and one former campaign manager, and an online questionnaire sent to a sample of Brazilian WhatsApp users. Much prior research into the convergence of social media and political communication relied on personal interviews. Enli and Skogerbø (2013), for instance, used the data collected from a series of post-election research interviews with Norwegian media experts for an analysis of social media platforms as political
communication tools, and found that politicians use social media sites and profiles as
successful methods of dialogue and voter interaction. Harris and Harrigan (2015) conducted a series of in-depth interviews with political candidates across the United Kingdom to explore how social media impact the public political debate in the 2010 UK elections, and similarly to the previous paper, surmised that social media were instrumentalized as efficient political relationship marketing tool. Klinger and Russmann (2017) performed a series of semi
structured qualitative interviews with Austrian and Swiss campaign strategists to find out how they used certain social media platforms for their political campaigns, and found political strategists to ascertain less importance to social media tools. The authors’ method allowed them to further understand the parties’ use of digital social communication tools for political
Page 29 of 94
campaign purposes. Qualitative research interviews thus allow me to find out more about how certain professionals in different fields perceive the use of social media within the political arena. The interviewees are: Eugênio Bucci (professor of journalism at the University of Sao Paulo and columnist at the daily O Estado de Sao Paulo), Caio Túlio Costa (founder of the largest news aggregate site in Brazil, UOL and journalist at several newspapers), Camila Rodrigues (journalist and editor at UOL), Letícia Cesarino (professor of anthropology at the Federal University of Santa Catarina), Joy Ernanny (journalist and TV producer at
GloboNews) and George Acohama (head of a creative agency and former campaign manager
of presidential candidate Henrique Meirelles). Due to a non-disclosure agreement, Acohama was not able comment much on his perceptions of fake news in Brazil.
In the process of interviewing, I tried to understand the different ways in which the respondents defined fake news, how WhatsApp impacted political campaigning during the Brazilian elections and how WhatsApp affects democracy, journalism and electoral behavior. Interviewing media experts and social media researchers allows me to disclose similar, and at times competing perspectives that these actors have on WhatsApp use in Brazil, and its political influence. While qualitative interviewing for research purposes in a small scale does not allow for a general and wide-scale application of the analysis, semi-structured in-depth interviews do encourage the respondents to share a rich description of phenomena while leaving the interpretation or analysis to the researcher (DiCicco-Bloom and Crabtree, 2006). Thus I expected the media experts to provide personal reflections on how WhatsApp affects their profession, and to what extent they perceive WhatsApp played a role in the election outcome.
The interviews lasted on average between thirty and sixty minutes, and were all conducted via Skype due to the geographical distance. All interviewees were briefed on the subject matter beforehand and agreed to participate in this research.
Page 30 of 94
Semi-structured interviews
In order to further explore the phenomenon of WhatsApp in Brazil, I have chosen to complement the semi-structured interviews with an online questionnaire, that was filled out by 108 Brazilian WhatsApp users from 13 states across the country. A series of questionnaires was used by Aharony and Gazit (2016) to examine the importance of WhatsApp groups among university students, and in effect confirmed the authors’ hypotheses that the personality characteristics of people impacts the individuals’ attainment of importance to WhatsApp groups. Some previous social media research also uses separate methodologies. DiStasio and Bortree (2012) for instance complemented their semi-structured interviews with an online questionnaire sent out to public relations employees at Fortune 200 companies to find out if social media improve corporate transparency. In their study, the researchers found that while the survey respondents believed their social media behavior greatly impacted transparency on a corporate level, the interviews and a content analysis revealed quite the opposite. A multi-method analysis, such as the one I will perform in the following thesis, therefore allows me to explore and possibly reveal differing opinions and views, on the one hand from Brazilian media experts affected by fake news on WhatsApp, and on the other hand from average Brazilian WhatsApp users who engaged with WhatsApp content non-professionally.
The choice of the sample of respondents of the questionnaire is justified by
circumstantial limitations and holds an acceptable degree of generalizability. To account for ethical considerations, all respondents agreed to participate in the study before starting the questionnaire and were briefed about the purpose and nature of it before filling it out. The answers were collected anonymously, no personal information such as name or address was asked, no IP address was collected, and the respondent's answers are kept as confidential
Page 31 of 94
information in line with ethical considerations. Several questions did not need to be filled out compulsory, which explains why some questions were not answered by all 108 respondents. Based on the notion that indirect questioning in surveys “enhance the chances that the respondent would be willing to participate in such a survey and provide honest answers” (Chaudhuri and Christofides, 2013), I applied qualitative indirect questions at times throughout the questionnaire in which I asked about the respondents’ reflections on fake news, without explicitly mentioning fake news. Thereby I hoped to get more honest and reflective answers from the respondents.
Page 32 of 94
BRAZIL’S MEDIA ENVIRONMENT
Much of recent research into social media in Brazil suggests that the transformation of the role of traditional media influenced the emergence of social media as both a source of information for the average Brazilian and as a prolific tool for political campaigning in Brazil (Aguiar, 2017). According to Machado and Konopacki (2018), the 2018 Brazilian elections were the first in which social media played a more important role for political campaigning and convincing the electorate than traditional news media. I therefore want to briefly sketch the Brazilian media landscape and present the context that allowed social media, and in particular WhatsApp, to become such a dominating force for political campaigning.
Among the traditional mass media, television is the most popular in Brazil (Globo, 2017). Traditional television networks dominated the public discourse and information exchange in Brazil since the 1960s and 70s (Santini et.al., 2017, van Dijck, 2011). Rede Globo in particular became the largest television broadcast network across the country by attracting top television programming talents and producing popular telenovela shows throughout the years (McGregor et.al., 2017). Over 91 million people reported to tune into the channel on a daily basis in 2014 (The Economist, 2014), highlighting the pervasiveness of Rede Globo consumption. In addition, Grupo Globo (Rede Globo’s parent company) acquires the rights to broadcast all major sports events, ranging from the World Cup to the quadrennial Copa América (MOM, 2017), thus further positioning itself as the main sports broadcaster, a strategy that allows Rede Globo to dominate all major football event broadcasts in one of the most valuable football markets. The recent decade however experienced a major transformation across the Brazilian media ecology (Passarelli and Angeluci, 2017) in which the smartphone became a central communication tool and information source in Brazil, alongside television. According to a study by Cetic.br (2016), the smartphone3 surpassed the computer as the most commonly
3 Smartphones are cell phones with an operating system that allow access to the internet, and the sharing of
Page 33 of 94
used internet access tool, and 35% of Brazilians exclusively use the cell phone to access the internet. Particularly lower income classes and those living in rural regions use the smartphone to access the internet. A working paper by Arnaudo (2017) found that there are roughly 26 million broadband subscriptions in the entire population of 200 million, juxtaposed to 250 million cellular subscriptions. The pervasiveness of mobile usage in Brazil is hereby highlighted by comparison with Pakistan, a country with a population of similarly 200 million people, but only 140.5 million mobile phone subscribers (Attique, 2017). Brazil was among the ten largest emerging markets for smartphones between 2014 and 2015 (Passarelli and Angeluci, 2017), and as a result, smartphone penetration in Brazil is among the world’s densest (Michelson, 2017). This is important, because among the affordances of Web 2.0, social media are the most popular in Brazil, and predominantly used over the smartphone (PagBrasil, 2019). This reflects the current state of internet access in Brazil, which, according to Belli (2018) is very expensive, where a “broadband connection can cost up to 15 percent of a household’s income”. Central to this context is that most Brazilian mobile phone contracts provide unlimited use of Facebook and WhatsApp, which takes away personal inhibitions of using these specific social media networks (see Figure 1 for examples of Brazilian network providers offering free use of WhatsApp), and encourages smartphone users to rely more on Facebook and WhatsApp than commonplace internet affordances. The result is that Brazil is now one of the only economies in which social media penetration can be seen as a synonym of internet penetration, illustrating the importance of online networks (Citounadin, 2018). Parallel to the ubiquity of smartphones among Brazilians is the increasing popularity of video-on-demand services such as Netflix and Amazon Prime, much to the chagrin of big broadcast networks, for example Rede
Globo and its main competitors Band and SBT, who suffered a decreasing viewership and as a
result have had to re-think their user engagement strategies (Época Negócios Online, 2018; Ariens, 2017). Many argue that Rede Globo is the most successful in adapting to the new television consumption trend, by having created its own streaming platform Globoplay (Ariens,
Page 34 of 94
2017). Although the major Brazilian television networks put in effort to appease the new consumer demands, the power traditional television networks yield as public opinion-drivers has fallen according to some observers (Época Negócios Online, 2018) This power shift is most predominantly highlighted by the impact of media on the Brazilian political campaign. According to a study conducted in early 2018 (Goldzweig, 2018), more than 56% of Brazilian voters claim that social media influenced their voting behavior, and for the first time in modern Brazilian elections since 1988, those political candidates with the largest allotments of party broadcasting times during the free electoral TV time slots have performed badly, losing to the one candidate who had in the first presidential run-off no free electoral TV time slots, Bolsonaro4 (Machado and Konopacki, 2018, Gomes, 2018).
Additionally, the country is mired in several corruption scandals and an economic downturn in recent years, leading to a general disillusionment with national and state institutions (Arnaudo, 2017, Spyer, 2017), as well as traditional TV networks, which are seen as supporting the Brazilian establishment (Aguiar, 2017). It is in this scenario, as a result of people’s dissatisfaction of day-to-day governance and economic issues that digital social networks emerge as an alternative communication form and news source. Young citizens have been found to be specifically attracted to social media networks (Moura and Michelson, 2017), but, as Spyer (2017) and Arnaudo (2017) both highlight, the prevalent use of mobile phones in Brazil results in a cross-generational use of social media, which have become major focal points for political campaigning and communication. News in today’s Brazil is not consumed entirely via traditional mass media anymore. Although Globo’s daily news program Jornal Nacional still enjoys wide popularity (van Dijck, 2011; Guazina et.al., 2018), social media have become one of the most important sources of news for average Brazilians (Reuters, 2017).
4 During the first electoral run-off in 2018, Geraldo Alckmin, of the Brazilian Social Democratic Party (PSDB), had
more than half of the total electoral broadcast time slots on television, and only received less than 5% of the total votes.
Page 35 of 94
The rising ownership of smartphones across Brazil coincided with the growth of WhatsApp as a popular messaging application. Since then, WhatsApp is used for business, political engagement and interpersonal communication. In 2014 for example, the shampoo brand Head & Shoulders shared short and entertaining WhatsApp videos of the Brazilian football coach Joel Santana, thereby attempting to create a marketing campaign (Teixeira 2014 (quoted by Moura Michelson), and in early 2018 the former Brazilian president Dilma Rousseff created her own WhatsApp channel titled Dilma Zap Zap, with the intent to engage with potential voters (Negrisoli, 2018). But while business and marketing practices changed due to the messaging application, the Brazilian economy itself was impacted by a movement driven by WhatsApp. In May 2018, Brazil experienced what the Washington Post called “the largest and most effective truckers strike in the nation’s history”, in which hundreds of thousands of truck drivers mobilized themselves through closed WhatsApp groups to bring large parts of the Brazilian economy to a standstill (Lopes, 2018). Gasoline, ethanol and diesel ran out within less than a week, and by leaking WhatsApp voice messages to the public from frustrated truck drivers voicing their demands about the government, more than 87 percent of Brazilians supported the truckers’ movement, according to one poll (Gielow, 2018; Fox, 2018; Lopes, 2018). It therefore came to no surprise that WhatsApp gained further prominence during the election campaign when then-candidate and president-elect Jair Bolsonaro showed his personal WhatsApp feed in a widely shared YouTube video (YouTube, 2017), thus positioning the messaging application at the heart of political campaigning.
Page 36 of 94 Figure 1: the three major Brazilian network providers (Vivo, TIM and Claro) offering free WhatsApp use as part of their mobile plans
