From fake news to hyper-partisan media
The term “fake news” began to be widely used in the press coverage of the 2016 US presidential elections. There is much debate in the social sciences and communications fields about the applicability of this concept. We believe that the term “hyper-partisan media”, which is used in part of the literature, provides a better understanding of the phenomenon. We also argue that it is important to understand that hyper-partisanship emerged as a result of the polarisation of the public sphere. In this article, we analyse the dissemination of hyper-partisan news one week before the judges’ vote in the trial of former president Lula at the TRF-4 (Tribunal Regional Federal da 4a região, or the Federal Regional Court of the 4th Region). We show how the headlines of the stories that were shared the most on social media during that period fit into one of the two polarised narratives currently dominating the Brazilian public sphere. We conclude the article with an assessment of possible approaches to regulation on this issue.
The concept of “fake news” is highly disputed and there is no broadly accepted definition for it in academic literature or media discourse. Even though we find earlier references to it, it was in the press coverage of the 2016 United States (US) presidential elections that the use of the term, with its current meaning, became widespread. It was adopted to refer to news sites that disseminated false information on Hillary Clinton on social media and whose impact is said to have contributed to Donald Trump’s victory.11. Hunt Allcott and Matthew Gentzkow, “Social Media and Fake News in the 2016 Election,” Journal of Economic Perspectives 31, no. 2 (2017): 211-236.
Today, the literature is divided up between those who defend the use of this concept, which was shaped by the political debate and news coverage, and those who believe that since it is so unprecise and leads to misunderstandings, it would be better to find a more appropriate term. We list below some of the most influential definitions to illustrate the nature of the debate.
An official Facebook report written by Weedon, Nuland and Stamos defines fake news as follows:
The term “fake news” has emerged as a catch-all phrase to refer to everything from news articles that are factually incorrect to opinion pieces, parodies and sarcasm, hoaxes, rumors, memes, online abuse, and factual misstatements by public figures that are reported in otherwise accurate news pieces. (…) We’ve adopted the following terminology to refer to these concepts: (…) News articles that purport to be factual, but which contain intentional misstatements of fact with the intention to arouse passions, attract viewership, or deceive.22. Jen Weedon, William Nuland and Alex Stamos, “Information Operations and Facebook.” Facebook Newsroom, April 27, 2017, accessed June 11, 2018, https://fbnewsroomus.files.wordpress.com/2017/04/facebook-and-information-operations-v1.pdf.
In the most commonly quoted study on this issue, Alcott and Gentzkow 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.33. “Social Media and Fake News in the 2016 Election,” 2017, 212.
Finally, in a recent article in the Science journal, Lazer et al. define “fake news” as:
…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).44. David M. Lazer et al., “The Science of Fake News,” Science 359, no. 6380 (2018): 1094-1096.
We can say, then, that in the debate on the term’s value for analytical purposes, there is considerable controversy over at least two points: i) whether the “fake news” concept should refer only to news content that has been proven false55. Soroush Vosoughi, Deb Roy and Sinan Aral, “The Spread of True and False News Online,” Science 359, no. 6380 (2018): 1146-1151. or whether it should also refer to other tactics to misinform and deceive others, such as exaggeration, omission, information taken out of context and speculation;66. Claire Wardle and Hossein Derakhshan, Information Disorder: An Interdisciplinary Framework for Research and Policy Making (Strasbourg: Council of Europe, 2017). and ii) whether the concept should only include false content that has been produced intentionally or whether it should also include any kind of verifiable factual error, even if unintentional, such as a simple error made when verifying the facts.
The choice of one definition or another should not be made arbitrarily. Instead, it should be guided and informed by the analysis of what has been commonly called “fake news sites”. And if we examine how these websites function, we will see that what best defines them is not the publication of fake news – which only happens occasionally – but rather the production of “combat information” in the form of news reports.
This combat information may be merely a clipping from the daily news selected to prove one’s point, a news report with a sensationalist headline, a fact mentioned out of context, an exaggeration of a story or some speculation on an event or an issue that is presented as factual information – and sometimes, it can even involve lies. Several of these tactics to distort facts are not limited to the so-called “fake news sites”; they have also been used by mainstream media. This is why it is impossible to trace a clear line to separate bad news outlets from trustworthy ones, the truth from lies.
It would perhaps be better to look at the context in which “combat information” has been produced, by both alternative and mainstream news sources, and reflect on how the fabrication of facts and other distortive practices are the result of a broader process of lowering editorial standards to respond to a polarised political environment. We believe that this is why a portion of researchers and analysts have abandoned the concept of “fake news sites” and referred to the phenomenon as the emergence of a “hyper-partisan” media.77. Matt Karolian, “Hyper-Partisan Pages Still Outperforming Mainstream Media on Facebook.” Medium, February 1, 2018, accessed June 11, 2018, https://medium.com/@karolian/hyper-partisan-pages-still-outperforming-mainstream-media-on-facebook-4fcbd036ced3; Craig Silverman et al., “Hyperpartisan Facebook Pages Are Publishing False And Misleading Information At An Alarming Rate.” Buzzfeed News, October 20, 2016, accessed June 11, 2018, https://www.buzzfeed.com/craigsilverman/partisan-fb-pages-analysis; Craig Silverman et al., “Inside The Partisan Fight For Your Newsfeed.” Buzzfeed News, August 8, 2017, accessed June 11, 2018, https://www.buzzfeed.com/craigsilverman/inside-the-partisan-fight-for-your-news-feed.
Brazilian academic literature has addressed the issue of political polarisation more from the angle of voters’ party preferences than from the perspective of the alignment of political views.88. André Borges and Robert Vidigal, “Do Lulismo ao Antipetismo? Polarização, Partidarismo e Voto nas Eleições Brasileiras,” Opinião Pública 24, no. 1 (2018): 53-89; Fernando Limongi and Rafael Cortez, “As Eleições de 2010 e o Quadro Partidário,” Novos Estudos, no. 88 (2010); Jairo Nicolau, “Vermelhos e Azuis: Um Estudo sobre os Determinantes do Voto nas Eleições Presidenciais Brasileiras (2002-2010).” 9th Meeting of the Brazilian Political Science Association, Brasília, 2014, accessed June 11, 2018, https://cienciapolitica.org.br/system/files/documentos/eventos/2017/03/vermelhos-e-azuis-estudo-sobre-determinantes-voto-nas.pdf; F. W. Reis, “Eleição de 2014: ‘País dividido’ e Questão Social,” Periódico de Opinião Pública e Conjuntura Política, year 6, v. 6 (October 2014). In the US, however, a lengthy debate in the field of political science has shown that while there has been a polarisation of political opinions or positions in that country, this phenomenon is probably limited to people who are more involved in and who identify more strongly with political parties.99. Kyle L. Saunders and Alan I. Abramowitz, “Is Polarization a Myth?”, Journal of Politics 70, no. 2 (2008): 542-555; Paul DiMaggio, John Evans and Bethany Bryson, “Have American’s Social Attitudes Become More Polarized?” American Journal of Sociology 102, no. 3 (November 1996): 690-755; Morris P. Fiorina, Samuel A. Abrams e Jeremy C. Pope. “Political Polarization in the American Public,” Annual Review of Political Science 11 (2008): 563-588. It should be noted that even though individuals who are more engaged and more partisan are a minority in society, they are precisely the ones who interact the most in the political debate, which German sociology has called the political public sphere.1010. Jürgen Habermas, Mudança Estrutural da Esfera Pública (Rio de Janeiro: Tempo Brasileiro, 2003 {1961}); Jürgen Habermas, “The Public Sphere: An Encyclopedia Article (1964),” New German Critique, no. 3 (Autumn 1974). Therefore, when we look at society as a whole, polarisation is a limited phenomenon; but when we consider the public sphere, where private individuals discuss political issues outside the realm of the state, political polarisation is a dominant phenomenon.
Even though there are not many studies in Brazil on the polarisation of political views and positions, the analysis of social media gives a solid indication of the polarisation of public opinion in the country that is contaminating the entire public sphere. When we analyse the patterns of interaction of the 12 million Brazilians who like posts on the top 500 political pages, we note that since the first half of 2014, they have closed themselves off into two separate circuits. On one hand, we find the pages of left-wing parties and politicians amalgamated with the websites of feminist, black and LGBT movements and human rights NGOs; on the other hand, we find the pages of right-wing parties and politicians, together with those who defend economic liberalism and moral conservatism. When we examine the pattern of the users’ interaction, we find that they form two clusters (see figure 1 below) with very few connections between them. This clearly confirms that the users have become polarised.
Figure 1: Pattern of Facebook users’ interaction with the top 500 political pages, March 2016
Source: Elaborated by the authors based on Facebook API
The dispute over the significance of the Workers’ Party’s (Partido dos Trabalhadores often referred to using the acronym PT) historical role and the impeachment of former president Dilma Rousseff are at the heart of the feud that has polarised the public sphere in Brazil. On one side, there is the anti-PT movement, which is made up of liberals, conservatives, people who glorify the military, political parties from the current government’s base of allies and a number of people and groups who see corruption in general, and the Workers’ Party in particular, as the country’s biggest problem. The narrative shaping this side of the dispute claims that the PT hijacked the Brazilian state to pillage public resources and replaced the party’s political project with a plan to remain in power. It alleges that the party sought to stay in office by resorting to a series of schemes, such as forming alliances with large corporations that took advantage of resources from the Brazilian Development Bank (BNDES), bloggers who received advertising funds from the federal government, artists who obtained public funds via the Rouanet Law1111. Federal Law on Incentives for Culture. Law nº 8.313, December 23, 1991. and social movements whose members received privileges disguised as social programmes. It also argues that this corruption has not only set off an unprecedented crisis of ethics, but it is also the main cause of the economic crisis and the fiscal deficit. To face this historical challenge, Operation Car Wash is said to have been launched as a virtuous process of persecuting criminals, which has had to face resistance from the National Congress, the media and the business sector.
In opposition to the anti-PT groups, another side has been formed, this time by social movements, NGOs, opposition parties and individuals and groups who identify with the left. The narrative that shapes this side goes like this: under the guise of an anti-corruption discourse, what truly moves the anti-PT side is their aversion to the poor who made major advances during the Lula and Dilma administrations. Discontent with the social and electoral success of the PT governments, these sectors of the business class, the judiciary, the conservative middle class and the media were the driving force behind a partial and abusive investigation that pretended to be fighting corruption in the Brazilian state, but actually only persecuted the main left-wing political party in Brazil. Then, as if that was not enough, they staged a parliamentary coup to illegitimately overthrow President Dilma and, shortly after that, they arrested and convicted former president Lula, who was leading the opinion polls for the 2018 presidential elections. Having ousted ex-president Dilma and dashed the PT’s hopes of winning the elections, the coup organisers implemented a regressive agenda, which includes measures such as the elimination of social rights and the privatisation of public assets.
These two structuring narratives serve as a reference for hyper-partisan news sites when they produce combat headlines to fuel the public debate. To illustrate this dynamic, we gathered all of the headlines produced by 96 news sites, plus the ones produced by or shared on nearly 500 Facebook pages selected as the most popular pages focused on national politics, between 21 and 27 January 2018. This was the week of the trial of former president Lula in the second instance court (the federal regional court, commonly referred to as TRF-4) in the case on the apartment in Guarujá.1212. Former president Luiz Inácio Lula da Silva was the target of an investigation by the Federal Prosecutor’s Office that affirmed that he, together with the other defendants, received an apartment in Guarujá (SP) as a bribe from the OAS construction firm. The prosecution accused Lula of facilitating the signing of irregular contracts between the construction company and Petrobras. The case was tried by first and second instance courts and ended in the ex-president’s conviction: Luísa Granato and Valéria Bretas, “A Linha do Tempo do Caso Triplex que Levou Lula à Prisão.” Exame, April 6, 2018, accessed June 11, 2018, https://exame.abril.com.br/brasil/a-linha-do-tempo-do-caso-triplex-que-levou-lula-a-prisao/. The sites studied produced close to 13,000 articles that together accounted for a little less than 12 million shares. We analysed the 150 headlines on national politics with the highest number of shares during the period. This sample, though small, was shared 3.5 million times, or 29 per cent of the total shares during the period.
The main topic of the week was the trial of former president Luiz Inácio Lula da Silva in the second instance court. Other issues that received broad media coverage that week can also be easily inserted into one of the two narratives in the dispute. For example, three stories alerted readers about politicians being investigated by Operation Car Wash, who if elected, would have privileged jurisdiction. On the other hand, the decision of the Supreme Court president, Carmen Lúcia, to suspend Cristiane Brasil’s as the Minister of Labour was celebrated by the left who saw in Cristiane Brasil an antagonist to labour rights.
Some articles reported the refusal of the writer Marcia Tiburi to give an interview with MBL (Movimento Brasil Livre or Free Brazil Movement) activist Kim Kataguiri. For some, Tiburi’s decision was correct and exemplary, since “there is a limit to everything” and that limit would be the fascism of those who defended the coup. The other side criticised what it interpreted as a contradictory attitude by the author of the book “How to Talk to a Fascist?”
In another example, perhaps even more illustrative of the hyper-partisan dynamics of sharing in social networks, two headlines deal with the campaign promoted by the media network Globo: O Brasil que queremos (The Brazil that we want). Both claim how the population would respond to the campaign. In one, the defiance would be to show not what the country wants, but what is not wanted, namely a country ruled by corrupt politicians. The other article sees defiance in those who sent videos to the station declaring that they want a Brazil sem golpe (without the coup).
The headlines related to the trial of ex-president Lula in the second instance court that were shared the most can also be easily grouped into two narratives. On one side, the unusual speed with which an investigation against José Serra1313. Former PSDB governor of the state of São Paulo, José Serra, was mentioned in the confession of one of the directors of the J&F corporation that controls the JBS meat processing company. In his testimony, Joesley Batista confirmed having paid R$6.4 million to Serra, who was running for president at the time, through illegal transfers to corporations, which was in addition to the R$13 million declared officially. Renan Ramalho, “Dodge Pede ao Supremo Arquivamento de Investigação sobre José Serra.” O Globo, January 24, 2018, accessed June 11, 2018, https://g1.globo.com/politica/noticia/dodge-pede-ao-supremo-arquivamento-de-investigacao-sobre-jose-serra.ghtml. was conducted and filed appeared to corroborate the theory on the persecution of the PT. The other side argued that the TRF-4 gave a “display of composure, efficiency and justice”. One side praised the activists who went to Porto Alegre to defend the former president, while the other described the protest as a “waste of money”. One side urged the people to take to the streets and warned that the imprisonment of the PT leader “would set the country on fire”, and the other echoed the message of Attorney General Raquel Dodge, who stated that Brazil was experiencing its “greatest moment of institutional stability since 1889”. One side shared messages where actors, activists and soccer players showed their support for the ex-president, whereas the other shared messages from people commemorating the ruling against him.
Of the headlines analysed, 67 can be categorised in the “coup” narrative category and were shared 1.494 million times, and 45, as part of the anti-PT narrative, which account for 1.308 million shares. In other words, more than two thirds of the headlines analysed, which represent close to 80% of shares, are aligned with one of the two narratives that structure the national political debate (table 1).
| Issue | Number of articles | Number of shares |
|---|---|---|
| TRF4 (coup) | 49 | 1041 |
| TRF4 (anti-PT) | 38 | 1016 |
| Cristiane Brasil (coup) | 7 | 265 |
| Operation Car Wash (anti-PT) | 4 | 154 |
| Lula’s candidacy (coup) | 4 | 85 |
| Márcia Tiburi (coup) | 5 | 82 |
| Lula’s candidacy | 2 | 74 |
| Globo’s campaign (anti-PT) | 1 | 53 |
| Globo’s campaign (coup) | 1 | 21 |
| Márcia Tiburi (anti-PT) | 1 | 11 |
| Other issues | 38 | 724 |
It is in this polarised context marked by heated political debate that the techniques of selecting, distorting and fabricating facts emerge. As we can see below, in general, it is not a question of lies, strictly speaking, but rather of various levels of distortion. We analyse below four headlines that concretely illustrate the nature of the issue:
During the impeachment process against ex-president Dilma Rousseff, a group of 17 lawyers linked to the OAB (Ordem dos Advogados do Brasil, or the Brazilian Bar Association) in the state of Espírito Santo resigned collectively in response to what they interpreted as censorship of the directors of the association who had criticised Judge Sérgio Moro.1515. “OAB-ES Censura Comissões que se Posicionaram Contra Moro e Membros Pedem Renúncia Coletiva,” Justificando, March 23, 2016, accessed June 11, 2018, http://justificando.cartacapital.com.br/2016/03/23/oab-es-censura-comissoes-que-se-posicionaram-contra-moro-e-membros-pedem-renuncia-coletiva/. At that time, another version of the story attributed the lawyers’ quotes to PT deputies saying the congressmen would resign if Dilma were to be impeached.1616. Edgard Matsuki, “Deputados do PT Vão Renunciar se Dilma Sofrer Impeachment #Boato.” Boatos, April 16, 2016, http://www.boatos.org/politica/deputados-do-pt-farao-renuncia-coletiva.html. A third version of this article published on the The Folha newspaper website1717. The Folha, Homepage, 2018, accessed June 11, 2018, http://www.thefolha.com.br. began to circulate the week of the ruling on Lula’s case in the second instance tribunal. This time, it was said that the deputies were threatening to resign if the ex-president were convicted. The headline was reproduced by people on the anti-PT side to celebrate the fact that the imprisonment of Lula would get rid of other leaders who they said were also corrupt.
In August 2017, The New York Times published a profile of Judge Sérgio Moro.1919. Ernesto Londoño, “A Judge’s Bid to Clean Up Brazil From the Bench.” The New York Times, August 25, 2017, accessed June 11, 2018, https://www.nytimes.com/2017/08/25/world/americas/judge-sergio-moro-brazil-anti-corruption.html. Then, months later, during the week of the TRF-4 vote, the “Notícias Brasil Online” website evoked this article again with the headline “The New York Times Newspaper Praises Judge Sérgio Moro And Says That The Judge Has Become The Face Of Accountability in the Country”.2020. “Jornal New York Times Elogia Juiz Sérgio Moro e Diz Que Magistrado se Tornou o Rosto da Prestação de Contas no País,” Notícias Brasil Online, January 25, 2018, accessed June 11, 2018, https://www.noticiasbrasilonline.com.br/jornal-new-york-times-elogia-juiz-sergio-moro-e-diz-que-magistrado-se-tornou-o-rosto-da-prestacao-de-contas-no-pais/. The “Notícias Brasil Online” article was shared 88,000 times. In January 2018, an opinion piece by Mark Weisbrot, also published in The New York Times under the title “Brazil’s Democracy Pushed Into the Abyss”,2121. Mark Weisbrot, “Brazil’s Democracy Pushed Into the Abyss.” The New York Times, January 23, 2018, accessed June 11, 2018, https://www.nytimes.com/2018/01/23/opinion/brazil-lula-democracy-corruption.html. was the subject of a series of articles published by alternative left-wing media outlets (Brasil 247, Revista Fórum, Sputniknews, lula.com.br and Falando Verdades). Together, these articled totalled 123,000 shares. In both cases, the article in question was taken out of context and in the latter case, the headlines suggested that the criticisms of Moro reflected the newspaper’s editorial line when, in fact, it was a mere opinion piece.
The article affirms that a woman apparently named “Mrs. Angelina” complained on Facebook that she did not get paid for participating in a protest in defence of Lula, as promised. Shared 34,000 times, the article supports the anti-PT narrative, which claims that protestors who support Lula have been paid to do so.
The Brazilian army’s website contains a section with clippings from the daily news. On 17 January, one of the items posted in this section was an opinion piece by Bernardo Bittar published by the “Correio Braziliense” newspaper under the title, “Lula’s trip to Ethiopia cost public coffers over R$30,000”. In the column, the author criticises the fact that former presidents continue to have the support of advisors who are paid with public funds. In the case of Lula, the author estimated that R$3.1 million had already been spent. The article published by “Imprensa Viva” considered the decision to select this text and include it in the daily clipping as proof that the army was outraged with the cost of the trip. Furthermore, the site opted for including the total amount spent on advisors since the end of Lula’s mandate in the title.
Much of public debate today takes place on social media platforms such as Facebook and Twitter. The majority of the content that circulates on them is produced by news sites. Yet, an article’s reach depends primarily on users, as they are the ones who decide what to share with their “friends” and “followers”. The proliferation of so-called “fake news” is intimately linked to a dynamic in which the polarisation of the public sphere turns the public debate into a confrontation between two narratives. Regulating the production and sharing of fake news would be one way of mitigating this risk. Yet, what can be done and what are the risks of regulating the digital public sphere?
According to a typology established by the lawyer Lawrence Lessig, individuals are subject to the constraints imposed by society, the state, the market and technology. Thus, there are four possibilities for regulating behaviours: intervening in the perception of social norms; altering and enforcing laws; adjusting prices and fees; and modifying the architecture, which, in the digital environment, corresponds to making changes to the application’s code.
Lessig anticipated that for the internet, there is a risk that the fourth option, which is much less permeable, would be too tempting and overlap the others.2424. Lawrence Lessig, Code: And Other Laws of Cyberspace, Version 2.0 (New York: Basic Books, 2006). In the case of fake news, this option corresponds to the development of algorithms that help identify fake news. It may be possible to develop algorithms that use a base of examples and enable machines to do this task. This algorithm would replicate the biases of this base of examples at a larger scale, which would make constructing it a delicate task, especially in a context of intense polarisation. Moreover, as we are dealing with a relationship between language and the world, a software, which does not have access to the world, cannot assess the accuracy of a fact. This leaves us with indirect measures for evaluating content: format, occurrence of words, the dynamics of sharing. At best, an algorithm would only be capable of identifying something that has the format of a fake news report established according to a model produced by individuals with these biases.2525. Cathy O’Neil, “If Fake News Fools You, It Can Fool Robots, Too.” Bloomberg News, January 24, 2017, accessed on June 11, 2018, https://www.bloomberg.com/view/articles/2017-01-24/if-fake-news-fools-you-it-can-fool-robots-too.
The legislative alternative would be to somehow punish the producers and/or disseminators of fake news. In an article posted on 11 May, the Pública news agency found 19 draft bills going through the Chamber of Deputies and one in the Senate that aim to criminalise the production and sharing of fake news.2626. Pedro Grigori, “20 Projetos de Lei no Congresso Pretendem Criminalizar Fake News.” Publica, May 11, 2018, accessed on June 11, 2018, https://apublica.org/2018/05/20-projetos-de-lei-no-congresso-pretendem-criminalizar-fake-news/. The proposals apparently assume that the production or sharing of false news is an intentional act that seeks to contaminate the debate and bring harm to an individual or a group. The empirical analysis that we discussed in the sections above, however, suggests that the nature of the phenomenon is different. The public sphere is split into two poles that develop reductionist narratives to explain the delicate national political context. The dissemination of fake news is mainly done by users who are more involved in this dispute. Therefore, it is not a deliberate attempt to deceive someone, but rather to convince others that what appears obvious to the person disseminating the information is correct or true.
With regards to the economic dimension, we can think of information as a good that typically has a high fixed cost and a very low marginal cost. This difference became even more striking with the advent of digital technology. Moreover, once the depreciation costs necessary to create an information product for more than one company have been covered, competition between companies drives the price down to the marginal cost, which is zero in the digital world.2727. Carl Shapiro and Hal R. Varian, Information Rules (Massachusetts: Harvard Business School Press, 1999). In the case of producing news, the fixed cost varies according to the quality of the product. This environment has allowed media channels that produce news with little or no budget and very low quality standards to proliferate, generating a competition among media outlets whose impacts are detrimental to the political environment. What is more, the excess information that circulates on the internet competing for readers’ attention creates a situation where users have little capacity to discriminate what is worth reading and sharing from what is not.2828. Xiaoyan Qiu et al., “Limited Individual Attention and Online Virality of Low-quality Information,” Nature Human Behaviour 1, no. 0132 (2017).
As a result, on social media, media outlets with professional writing teams compete for readers’ attention with sites that have little or no concern with verifying the facts, for example. One way to regulate via the market would be to charge for access to news articles in order to artificially increase the marginal cost and pay news producers to cover the cost of remunerating their team of journalists.2929. In August 2017, Facebook announced that it would test a billing model that charges users for access to certain articles on its platform, which goes against this proposal of regulation through the market. “Testing Subscriptions Support in Instant Articles,” Facebook, October 19, 2017, accessed on June 11, 2018, https://media.fb.com/2017/10/19/testing-subscriptions-support-in-instant-articles/. The assumption behind this proposal is that the public would reward them for the quality of the journalism. What our analysis suggests, however, is that part of the interest in consuming and disseminating news in a polarised society is to support pre-established narratives, regardless of the quality of the work to investigate or verify facts that is required to produce the news.
Finally, the dissemination of false or unverified content may or may not be considered socially acceptable or unacceptable behaviour. If the problem of the spreading the so-called fake news is a social phenomenon, the solution must also be social. Perhaps the best regulatory approach would be to intervene directly in the public debate to increase social awareness on the damaging effects that the prevalence of combat information and the lowering of editorial standards of news sites have on the public sphere.
There is a need to shift the social responsibility for fake news, which is normally placed entirely on malicious actors who are out to “deceive the public”, to all of us who engage in political debates. The problem does not lie only with hyper-partisan media operators acting in bad faith, but rather with all of us who collaborate in the degradation of the public sphere by transforming the public debate into a thoughtless information war in which sharing low quality news materials is a socially acceptable act.
