Piecing together how Russian trolls have been using social media to disrupt authentic electoral democracies is a slow and disjointed business. But, little by little, help is available to try to understand what has been happening. Societies that value the many purposes of public debate cannot be passive about this potent threat to their information health.
Assisted by data analysts, I have been looking into the vast data lake that forms under the Guardian’s international journalism, and into which flow comments from audiences worldwide. It is a modest effort to chart how far, in 2016-17 in particular, Russian trolls infected the Guardian and its online debate forums. The results are heartening, in the sense that they indicate a relative lack of impact – which is in part a credit to the mostly unsung work of the moderators of comment threads.
Sifting 47m comments received since late 2015, the analysts looked into the impact of a set of 55 companies, individual names (real and fake), Facebook, Instagram, Twitter and email accounts, and Twitter hashtags, which was gleaned from an indictment made public by US special counsel Robert Mueller during his investigation into Russian interference in the 2016 presidential election.
Bear in mind that an identified account or hashtag may not be Russian-made; it may be genuine, but its sentiments chime sufficiently with trolls’ aims for them to boost it. Coupled with other data about confirmed Russian activity, the material provides more evidence of how the Russians went about supporting Bernie Sanders over Hillary Clinton as Democratic candidate, and Donald Trump over Clinton for president.
Among 111 appearances in Guardian comment threads by Mueller-identified entities, the Twitter hashtag HillaryForPrison cropped up most, 53 times, followed by the hashtag TrumpTrain, 25 times. The data has to be interpreted with care, but early impressions are that the trolls seemed to cluster most around comments alleging criminality or warmongering by Clinton.
One commenter, who appeared in Guardian threads between March and December 2016, seemed to promote a particular hashtag in comments posted in quick succession. One aim appears to be to leverage a large media organisation’s online forums to spread exposure to hashtags, and also to other social media presences where there is less moderation, more extreme material, or both.
Trump’s current travails with Mueller permit a thin smile at the irony of this comment that cites the HillaryForPrison hashtag: “Hopefully Hillary will be indicted by the FBI and DOJ.”
The New York Times has used the indictment disclosures to chart the Russians’ manipulation of the Trump campaign via Facebook; the Daily Beast has reported that trolls also used Reddit and Tumblr; and NBC News has done substantial work to analyse the effects of troll Twitter accounts, including the ways they were unknowingly amplified by celebrities.
At least 1,062 more impugned Twitter accounts are yet to be released by the intelligence committee of the US House of Representatives, which disclosed 2,752 last November.
One of the great advantages of democratic systems over authoritarian ones is the democracies’ ability – through techniques such as independent journalism, public inquiries and litigation in independent courts – to detect, expose, learn lessons and make necessary change for the better.The bigger the dataset examined, the greater the collaboration among the organisations’ technical experts, and the more useful these exercises could be to professional journalism and the public. The most telling patterns are more likely to emerge through large-scale retrospective analysis of what took place and when.
If journalism institutions do more of this kind of work and prudently share their findings, their audiences become more aware of tactics that can distort elections. The effort dovetails with a fundamental aim of journalism over the centuries – a better informed electorate.
• Paul Chadwick is the Guardian’s readers’ editor
