A network of Donald Trump supporters shares the widest range of 'junk news' on Twitter, and a network of extreme far-right conservatives on Facebook, according to analysis by Oxford University.
Researchers at the Oxford Internet Institute (OII) analysed three months of social media activity of US Twitter and Facebook users from November 2017 to January 2018 – the period leading up to President Trump's latest State of the Union Address.
They find that on Twitter, a network of Trump supporters shares the widest range of junk news and circulates more junk news than all other political audience groups combined.
On Facebook, extreme hard right pages – distinct from Republican pages – both share the widest range and circulate the largest volume of junk news compared with all the other audiences. Specifically, a group of 'hard conservatives' circulates the widest range of junk news and accounts for the majority of junk news traffic in the sample.
On average, the audiences for junk news on Twitter share a wider range of known junk news sources than audiences on Facebook's public pages.
Junk news sources are defined as deliberately publishing misleading, deceptive or incorrect information purporting to be real news about politics, economics or culture. This type of content can include various forms of extremist, sensationalist and conspiratorial material, as well as masked commentary and fake news.
Lisa-Maria Neudert, researcher at the Computational Propaganda Project in the OII, said: 'While the US election is well behind us, a year after Donald Trump has taken office social media users on Twitter and Facebook are still sharing a wide range of junk news.'
Working over 18 months, researchers working on the Computational Propaganda Project identified 91 sources of propaganda operating across the political spectrum. They used this information to compile a dataset of 13,477 Twitter users and 47,719 public Facebook pages consuming or sharing junk content in the three months leading up to the State of the Union Address. These audiences were then categorised by their political interests, using a supervised machine learning algorithm, to identify groups such as Trump supporters, conservative media fans, the progressive movement, libertarians, and members of the two main US political parties.
Analysis showed that the distribution of junk news content was unevenly spread across the ideological spectrum. On Twitter, a network of Trump supporters shared the widest range of junk news sources and accounted for the highest volume of junk news sharing in the sample, closely followed by the conservative media group. On Facebook, extreme hard right pages shared more junk news than all the other audience groups put together.
Lisa-Maria Neudert said: 'We find that the political landscape is strikingly divided across ideological lines when it comes to who is sharing junk news. We find that Trump supporters, hard conservatives and right-wing groups are circulating more junk news than other groups.'
On Twitter, the Trump support group shared content from 96% of the junk news sources on the researchers' watch list. Moreover, the Trump support group accounted for 55% of all the junk news traffic in the sample. On Facebook, the hard conservative group shared 91% of the junk news sites on the watch list, and accounted for 58% of junk news traffic in the sample.
Professor Phil Howard, director of the Computational Propaganda Project, said: 'There is an upside to all of this. It appears that only one part of the political spectrum – the far right – is really the target for extremist, sensational and conspiratorial content. Over social media, moderates and centrists tend not to be as susceptible.'