The Disinformation Issue Isn't (Just) a Platform Issue

Last week, Institut Montaigne was invited to speak at a think tank summit on “Globalization and Democracy”, organized by the Swiss think tank Avenir Suisse. Whilst many fear the impact of new communication tools and social media platforms on social cohesion, the roundtable on “the impact of technology and digitalization on democracy” painted a different picture. Below is an outline of the discussion.

Are social media companies the bad guys?

The argument most often heard on this issue is that social media platforms favour the formation of social communities that tend to read and share articles that comfort their world views.This argument is put forward by Eli Pariser in his 2011 book The Filter Bubble: What The Internet Is Hiding From You. In this environment, emotions influenced by social relations take over rationality as a vector for debate: we tend to trust news that come from people we respect, and are thus less inclined to question their veracity (this is explained by researchers Claire Wardle and Hossein Derakhshan in their 2017 report for the Council of Europe, Information Disorder: Toward an interdisciplinary framework for research and policy making).

This makes social media a great arena for sharing highly partisan and opinionated news, that in some cases can be manipulated or simply false. Sometimes the number of interactions social media users have with such content (shares, likes, comments) even surpasses the number of their interactions with traditional media articles. This is what Craig Silverman and his analysts at Buzzfeed recorded in their paper: the “20 top-performing false election stories from hoax sites and hyperpartisan blogs generated 8,711,000 shares, reactions, and comments on Facebook. Within the same time period, the 20 best-performing election stories from 19 major news websites generated a total of 7,367,000 shares, reactions, and comments on Facebook”.

Other facts have also supported the growing narrative about the negative impact of social media on democracy. For example:

• Twitter and Facebook have officially recognised the presence of foreign bots on their platforms, which took part in the political debate during the 2016 election (Twitter said that Russian bots retweeted the current US president 470,000 times).
• These companies made money from advertisements sold to foreign users (Facebook delivered 3,000 Russian ads to the American Congress in October 2017). In consequence, the argument goes, social media companies are responsible for polarizing politics, helping populism ride, and thus have to be controlled… On this very last point, Twitter, Facebook and Google have been testifying in US Congress, stuck in a sensitive debate on content moderation.

Research suggests their impact has been exaggerated

There is no doubt that audiences are being fragmented on social media, and that some of the content circulating receives very high visibility because of this. However, some consider the pretended impact of such content to be overly exaggerated. Those people also tend to believe that blaming the polarization of politics and growing social tensions on social media encourage us to avoid addressing the relevant, more difficult questions.

As was stressed during the roundtable, there is scientific evidence that helps overcoming these fears and contextualizing them. For example, one panelist quoted researchers from Stanford University and the National Bureau of Economic Research, who demonstrated that the heightened polarization of politics in the United States is not linked to the use of social media, as polarization mainly increased amongst the groups that are less likely to use the Internet and social media. Moreover, when it comes more specifically to disinformation, researchers Hunt Allcott and Matthew Gentzkow demonstrated that the average US adult came across and remembered only 1.14 fake stories in the months preceding the election. Considering the total amount of information circulating during the elections, this number is rather unimpressive.

What role for traditional media?

An article from the Columbia Journalism Review sheds a different light on the US elections. According to its authors, the figures reported by Craig Silverman and the team at Buzzfeed (8,700,000 interactions with the top 20 fake-news stories, more than the 20 most shared “real news” articles) need to be relativized. If all monthly Facebook users performed one action a day, which is likely an underestimate, those interactions with fake news would amount to 0.006% of all user actions over the same timeframe.

They then analyse the coverage of the campaign by the New York Times during the last 69 days of the election cycle. And the authors found that, out of the 150 front page articles related to the campaign published over this period, only 16 mentioned policy issues. Similarly, the number of articles mentioning Hillary Clinton’s email scandal over a period of six days was equal to the total number of articles mentioning both candidates’ policy proposals during the last 69 days of the campaign... Clearly, they say, traditional media played a significant role in promoting sensationalism.

What do “fake news” tell us?

Thus, we might benefit from changing our mindsets when thinking about disinformation, and start considering it as a consequence, rather than than a cause, of growing social tensions. As Ethan Zuckerman, director of the Center for Civic Media at the MIT Media Lab, puts it, in his white paper for the Knight Foundation: the journalism crisis in the United States has to be looked at from the wider perspective of a forty-year-long decline in trust in institutions. 

Institut Montaigne wrote on the negative perception of democracy in an article commenting the publication of two reports, by the Pew Research Center and the Fondapol. Both think-tanks reached similar conclusions. While 70% of citizens in Northern European countries claim to be satisfied with the way democracy is working, this figure goes down to 28% for Southern nations. Additionally, 62% of citizens considering that their quality of life improved over the last years are satisfied by the democratic system (vs. 29% for those whose quality of life has worsened). These and other figures (concerning links between attitudes towards democracy and geography, urban size or education) suggest that the declining trust in institutions is correlated with the level of satisfaction people have towards these institutions - this comes as no surprise...

According to Ethan Zuckerman, the reason why the news environment is in turmoil is that institutions simply don’t work for everyone. And in today’s world, this has become more visible. Thus, the media made it difficult for large, renowned organizations, to hide any failure in their functioning. From hacking information systems to recording private conversations, it has become easier for citizens to shed the light over some of our institutions’ inefficiencies. In an era dominated by scandals, it seems as if writing about scandals has become fashionable.

In this context, there are no easy solutions for the media to address the challenges that they face. Cooperation between media organizations and technology companies, as is often suggested, is probably amongst the best solutions to prevent the circulation of disinformation. However, while dealing with those issues, it is important to keep in mind the necessity to restore a productive dialogue between governance and citizens.